Personality Disorders PDF

Although the long-term patterns of problem-behaviour that form
personality disorders are part of the major classifications of disease, they

have a generally unrecognized importance in clinical practice. These
disorders are the origins of many medical consultations and social ills,
perhaps associated with self-harm, suicide, neglect of children,
criminality, alcohol and drug abuse, HIV transmission and 'accidents'.
In a comprehensive summary and evaluation of the clinical literature,
this book seeks to dispel the myth that personality disorder is simply a
category for those problem individuals for whom nothing can be done.
With detailed coverage of recognition, assessment and management,
Drs Dowson and Grounds provide a co-ordinated empirically-based
account of those aspects of personality disorders that are of relevance to
psychiatrists and allied disciplines, including clinical psychologists, social
workers, criminologists, specialist nurses and primary care physicians. It
describes how many individuals with personality disorders can be helped
by health care services.
PERSONALITY DISORDERS:
RECOGNITION AND CLINICAL MANAGEMENT
PERSONALITY DISORDERS:
RECOGNITION AND
CLINICAL MANAGEMENT
J O N A T H A N H. D O W S O N , MA, MB, BChir, MD (Cantab); DPM, PhD

(Edin); FRCPsych
A D R I A N T . G R O U N D S , BMedSci, BM, BS, DM (Nottingham); MRCPsych

University of Cambridge, UK
CAMBRIDGE
UNIVERSITY PRESS
Published by the Press Syndicate of the University of Cambridge
The Pitt Building, Trumpington Street, Cambridge CB2 1RP
40 West 20th Street, New York, NY 10011-4211, USA
10 Stamford Road, Oakleigh, Melbourne 3166, Australia
Cambridge University Press 1995
First published 1995
A catalogue recordfor this book is available from the British Library
Library of Congress cataloguing in publication data

Dowson, Jonathan H.
Personality disorders : recognition and clinical management /
Jonathan H. Dowson, Adrian T. Grounds.
p. cm.
Includes bibliographical references and index.
ISBN 0 521 45049 7 (he)
1. Personality disorders. I. Grounds, Adrian. II. Title.
[DNLM: 1. Personality Disorders - diagnosis. 2. Personality
Disorders - therapy. WM 190 D752p 1995]
RC554.D69 1995
616.85'8-dc20
DNLM/DLC
for Library of Congress 94-40749 CIP
ISBN 0 521 45049 7 hardback
Transferred to digital printing 2004
SE
JHD: To John Heaton Dowson, 1902-1994. Businessman,
husband, father and gentle man
Margaret Blanche Hudson Dowson, 1907-1992. Teacher, wife,
mother and gardener
of Leeds and Brinton, Norfolk.
ATG: To Professor John Gunn CBE. Teacher and tireless

advocate of psychiatric care for the personality disordered.
\ .. if psychiatry gives up all its difficult patients, society will give
up psychiatry (JG, 1992)'
Contents
Part one: Recognition page 1

1 Personality disorders: basic concepts and clinical overview 3
Psychiatric disorders 3
Why are personality disorders important? 4
Personality 5
Personality disorders 9
Principles of classification 13
Methods of classification 15
Boundaries between personality disorders and other mental
disorders 22
Co-occurrence of personality disorders with other mental
disorders 26
Epidemiology overview 27
Causation 28
Treatability 40
2 Specified personality disorders: clinical features 43
Validity of personality disorder syndromes 43
Paranoid personality disorder 44
Schizoid (and schizotypal) personality disorder 47
Antisocial personality disorder 56
Borderline personality disorder 65
Histrionic (or hysterical) personality disorder 74
Narcissistic personality disorder 80
Avoidant personality disorder 86
Dependent personality disorder 91
Obsessive-compulsive personality disorder 96
Passive-aggressive personality disorder 101
Self-defeating personality disorder 104
Contents
Sadistic personality disorder 109
Depressive personality (or personality disorder) and allied
disorders 114
Organic personality disorder 126
Enduring personality change after psychiatric illness 127
Personality disorders: less specific clinical presentations and
epidemiology 128
Personality disorder not otherwise specified (DSM-III-R
and DSM IV) or mixed and other personality disorders
(ICD-10) 128
Depressive disorders and personality disorders 129
Suicide and personality disorders 130
Self-harm and personality disorders 132
Substance use disorders and personality disorders 133
Somatization disorder and personality disorders 137
Anxiety disorders and personality disorders 140
Schizophrenia and personality disorders 142
Obsessive-compulsive disorder and personality disorders 144
Eating disorders and personality disorders 145
Increased use of health care facilities and personality
disorders 148
Complaints of chronic pain and personality disorders 148
Bowel disorders and personality disorders 149
Neuropsychological impairment and personality disorders 149
Post-traumatic stress disorder (PTSD) and personality
disorders 149
Attention deficit hyperactivity disorder (ADHD) and
personality disorders 150
Antisocial behaviours and personality disorders 151
Dangerous, violent or seriously irresponsible behaviour and
Inadequate and abusive parenting and personality disorders 152
Sexual deviation disorders and personality disorders 153
Epidemiology 153
Longitudinal aspects of personality disorders 159
Introduction 159
Other personality disorders: longitudinal studies 178
Effects of interventions for personality disorders - general
trends shown by longitudinal studies 179
Contents xi
Studies related to gender and sexual orientation 180

Longitudinal studies related to childhood precursors of
Longitudinal studies of personality disorders related to
suicide and mortality 189
Studies of the effects of personality disorders on course and
outcome of other psychiatric disorders 189
5 Assessment of personality disorders 195
Nature of personality disorders 195
Classification of personality disorders 196
Information sources 199
Non-methodological variables that influence assessment 199
Assessment: an overview 200
Methodological variables 202
Structured interviews 204
Self-report questionnaires (SRQs) 213
Clinical ratings 220
Routine clinical assessment 222
Differential diagnosis and co-occurrence 228
Personality disorder assessment: future directions 228
Part two: Clinical management 231
6 Drugs and other physical treatments 233
Overview 233
Neuroleptics 234
Antidepressants 235
Lithium carbonate 238
Anticonvulsants 239
Anxiolytics 239
Psychostimulants 240
Other drugs 241
Drug treatment: borderline and related personality
disorders 242
Drug treatment: schizotypal and related personality
disorders 243
Drug treatment: other personality disorders 243
Drug treatment: target features 243
Drug combinations 244
Other physical treatments 244
Physical treatments: psychological aspects 246
Physical treatments: a summary 246
Future directions 247
xii Contents
1 Psychological management 249

Definitions and overview 249
Psychodynamic psychotherapy 254
Brief (psychodynamic) psychotherapy 260
Supportive and related psychotherapies 261
Psychotherapy of borderline personality disorder: methods
compared 265
Cognitive psychotherapy 267
Behaviour therapy 272
Good clinical care 274
Couple, family and group treatment 279
Treatment settings 280
Management of specific problem-behaviours related to
8 Group psychotherapies 294
Overview 294
Definition and therapeutic ingredients 296
Group-specific phenomena 298
Organization 299
Goals of group psychotherapy 302
The group conductor's role 303
Emotionality 308
Patterns of group interaction 310
The conductor's problems 312
Negative effects 313
Training and supervision for group conductors 315
Varieties of group psychotherapy 315
Group psychotherapy and borderline personality disorder 316
9 Management of offenders with personality disorders 321
Personality disorders in offender populations 321
Clinical assessment for treatment 323
Range of interventions 324
Community and outpatient settings 325
Cognitive psychotherapy and behaviour therapy 329
Statutory supervision in the community 330
Inpatient settings 333
Assessments and reports for legal proceedings 335
Social management 338
References 340
Index 386
Acknowledgement
We are most grateful to Mrs Anne Robson who prepared the original
manuscripts.
Part One
Recognition
Personality disorders: basic concepts and
clinical overview
J.H. DOWSON
Psychiatric disorders
'Personality disorders' consist of maladaptive patterns of motivated behav-
iour that are usually evident for at least several years. They form one of the
main categories of psychiatric disorders.
Psychiatric disorders are usually defined in terms of signs and symptoms.
'Signs' are phenomena related to biological functioning and, in this context,
consist of behaviour, while psychiatric 'symptoms' refer to the patient's
complaints of adverse experiences. These include 'mental' phenomena,
involving what has been called conscious awareness. The term 'patient'
usually refers to someone who is being seen by a medical practitioner (or
other health-care worker) in a professional context.
Behaviour can be defined as what an individual says or does, but only if it
is associated with conscious awareness, so that an involuntary movement
such as a tremor would not qualify. Also, behaviour includes what is not
said or done; for example, patients with phobias may avoid a range of
situations. The various mental phenomena which underlie motivated
behaviour can be described by such terms as thoughts, ideas, memories,
moods, attitudes and feelings, while the concept of conscious awareness can
be extended to include the 'unconscious' or 'subconscious'; this involves
higher mental activity that can influence conscious mental phenomena,
although the person is not aware (or not fully aware) of these processes. For
instance, a person may 'unconsciously' avoid keeping a series of appoint-
ments with his/her doctor because of a fear of having a serious illness
diagnosed, even though the individual does not seem aware of the reason.
'Problem-behaviour in clinical practice' is a simple operational definition
of psychiatric disorder, and can refer to problems from the patient's
perspective as well as to difficulties that the patient causes for others. But
this does not clarify which individuals with 'problem-behaviour' should
4 Basic concepts and clinical overview
receive psychiatric and other medical services. Unfortunately, there is no

satisfactory answer to this and there are two main attributes of a patient's
problem-behaviour that are generally considered to require medical inter-
vention. First, if there is an identifiable (or suspected) abnormal biological
factor affecting the patient's behaviour (e.g. alcohol intoxication or a brain
tumour) or, second, if medical care and treatment appears to be helpful,
even if the concept of illness, disease or disorder would not usually be
applied. The latter situation could occur if an individual has been involved
in a serious traffic accident involving the death of several family members,
and becomes suicidal. He/she may then be considered to need a period of
care, restraint and sedation in a psychiatric hospital, even though the signs
and symptoms appear as an understandable reaction to events.
Every branch of medicine involves the accumulation of knowledge, and
the classification of relevant disorders (or 'diagnoses') is an essential part of
this process. Classifications of psychiatric disorders can be based on groups
of signs and symptoms, on causal factors, or on a rather untidy combi-
nation of both these approaches, which is apparent when the four main
generally-accepted classes of psychiatric disorder are considered. The first,
'organic disorders', is based on the presence of causal factors, and consists
of disorders that are due (at least in part) to an identifiable biological
abnormality, such as the changes in brain structure in Alzheimer's disease.
However, the remaining classes are mainly based on signs and symptoms;
the second, the 'functional psychoses', includes schizophrenia and manic-
depressive disorder, while the third, the 'neuroses', contains an hetero-
geneous group of syndromes, such as anxiety disorders and some types of
depression. The fourth andfinalcategory is 'personality disorders', which,
in contrast to all the previous categories, consists of repeated maladaptive
patterns of behaviour which are usually present, in some degree, through-
out adult life.
Why are personality disorders important?

While the repeated patterns of problem-behaviour that form personality
disorders (PDs) are part of the major classifications of disease, they have a
generally unrecognized importance in medical practice, although they
provide the basis for numerous medical consultations. Also, PDs are often
the origins of various social phenomena with serious consequences to
society or to public health, such as homelessness (Scott, 1993), alcoholism,
drug abuse, neglect of children, criminality, HIV transmission, 'accidents'
and driving offences. Tiihonen & Eronen (1993), in a study of all subjects
Personality 5
who had committed homicide in Finland over a specified period, found that
males with antisocial PD were about 20 times more likely to commit
homicide than males in the general population, and that homicide in this
group was usually associated with alcohol dependence. Of 13 homicide
recidivists in Finnish prisons or high-security hospitals, 11 had PD com-
bined with severe alcoholism (Tiihonen & Hakola, 1994). In both primary
care and hospital services, PDs appear to be associated with a high rate of
consultations for physical symptoms without a clear diagnosis, and with
many instances of self-harm. There is evidence that PDs can be causal
factors (with varying degrees of specificity) for other psychiatric diagnoses
such as some syndromes involving anxiety, depression and substance
abuse, while PDs often complicate the diagnosis, management and out-
come of co-occurring medical and psychiatric conditions. For example, an
hypomanic illness may be mistakenly diagnosed as an antisocial PD, or
exacerbate relatively mild maladaptive personality features. Also, if
patients with an eating disorder have an associated PD, this often affects the
severity and response to treatment, and co-occurring PDs have been shown
to adversely affect the outcome of a range of non-PD psychiatric disorders
(Reich &Vasile, 1993).
Patients with PDs are often mishandled in medical practice, because of a
lack of knowledge of PD syndromes and a generally 'negative' attitude that
can be evoked in their doctors. This is understandable, as the patient's
behaviour in relation to a PD often involves difficult-to-manage behaviour
such as excessive dependence on others, repeated complaints, problems in
sustaining relationships and difficulties in many aspects of social function-
ing. Also, many health care workers believe that PDs are nothing to do with
biology and medicine, despite the increasing evidence from twin and family
studies that genetic as well as environmental factors are important for the
development of many aspects of personality and PDs. A further reason is
another erroneous belief that 'nothing can be done', but, in reality, many
patients with PDs can be helped considerably by health care services.
Personality
Response I habit I traitI temperament

The concept of personality involves behavioural consistencies (Zuckerman,
1991), but describing 'behaviour' provides many pitfalls, even for the wary.
An observed action (e.g. avoidance) may be given a label, such as 'defensive
behaviour', which may make an assumption of a motive without sufficient
evidence, while similar and related behaviours may be given several

different designations. Also, different behaviours may receive the same
label.
A specified 'response' in a specified situation can be considered as the
basic unit of behaviour, for instance, a laboratory animal pressing a lever to
obtain food in response to an auditory stimulus. But even this example is
not a simple model, as the delay in, or duration of, the response may vary,
and be accompanied by variable additional behaviour that is not measured.
A 'habit' involves some consistency of behaviour over a period of time,
and as the human environment is not analogous to the relatively controlled
conditions of a laboratory animal, and will always show some variation, the
concept of a habit implies that there is some consistency of response in
different situations. It must be remembered that although a habit consists of
behaviour, this can be considered as being derived from higher mental
activity.
A 'trait' has been considered as the basic unit of personality or PD, and
consists of a specified group of related habits; for example, going to parties
can be a habit contributing to the trait of sociability. Consistency of
behaviour in different situations is more likely to be found in relation to a
trait than to a habit, and the degree of a behaviour related to a trait is
usually distributed approximately normally in the population.
An even broader concept combines traits to form basic dimensions of
personality, or 'temperaments', which have been defined as 'inherited
personality traits present in early childhood' (Buss & Plomin, 1986), and as
'the relatively stable features of the organism, primarily biologically
determined ...' (Strelau, 1983). But, because of continuous interactions
between the individual and the environment, modification of broad pat-
terns of behaviour over time would be expected.
The hypothesis that there is a major genetic contribution to temperament
implies that there may be associated biological markers, and that behaviour
related to temperaments may have counterparts in non-human animals.
Behaviour in non-human animals can be categorized as 'emotionality',
'sociability' and 'antisocial tendencies' (for example, involving intraspecies
aggression) can be identified in non-human mammals, and Rutter (1987)
considered that these have human counterparts, which probably have
'fairly direct' neurobiological correlates. Patterns of emotionality, socia-
bility and activity (the latter can be associated with antisocial tendencies)
can appear early in childhood and show a degree of stability, although the
nature of the related behaviours may change during childhood and adult
life. Other temperaments that have been postulated include impulsivity,
Personality 7
tough-mindedness, constraint, the three factors of extraversion, neuroti-
cism and psychoticism (which resemble sociability, emotionality, and
antisocial tendencies, respectively; Eysenck, 1982), and a five factor model
with extraversion (versus introversion), neuroticism (versus stability),
agreeableness, conscientiousness and culture (i.e. openness to experience).
Cloninger (1986) has argued for three main temperaments, i.e. harm
avoidance, novelty seeking and reward dependence, each associated with
specific neurotransmitter activity, and this model has been extended to
include seven dimensions, by the addition of persistence, self-directedness,
cooperativeness and self-transcendence (Cloninger, Svrakic & Przybeck,
1993). The latter dimension relates to how the individual identifies him/
herself as autonomous, an integral part of humanity or an integral part of
the universe as a whole.
Definitions and concepts

The concept of personality has been defined as 'the sum total of a person's
actions and reactions' (Walton, 1973) and the 'enduring qualities of an
individual shown in his/her ways of behaving in a wide variety of circum-
stances' (Gelder, Gath & Mayou, 1983). Hall & Lindsey (1978) considered
that the organization and integration of behaviours was an important
aspect of the concept, and they defined personality as 'that which gives
order and congruence to all the different kinds of behaviour in which the
individual engages'. Thus, although most definitions and assessments of
personality focus on overt behaviour, the underlying higher mental
activity, involving motivational, attitudinal and emotional elements, is an
integral part of the personality concept. Rutter (1987) considered that it is
the inclusion of patterns of thought and feelings, together with the
meaningful patterns of behaviour that reflect motivation, which dis-
tinguishes the concept of personality from temperament.
Allport (1937) described the 'dynamic organisation' of personality,
which refers to mental processes that are in a constant state of interaction,
with the potential for change. These phenomena involve attitudes to self,
patterns of relating to others and planning for the future. Also, psychoana-
lytic theory has focussed on the 'dynamic' (i.e. involving movement) aspects
of higher mental activity and the meanings of experiences. These
approaches recognized that symptoms such as anxiety are often found in
the context of a personality abnormality that has been termed the 'neurotic
character' (Alexander, 1930). Dynamic aspects of mental functioning
include 'defence mechanisms', such as 'denial', which are the various
observable methods by which an individual may keep certain thoughts at

the 'back of the mind' and out of conscious awareness. However, the
suggestion that the use of different psychological defence mechanisms can
be used to classify personality has been considered to be unproven (Rutter,
1987).
Consistency
A 'trait' (i.e. a group of related habits) is considered to be the basic unit of
personality, and a 'summary of consistencies of behaviour or an average of
states over time, which will reflect the individual's characteristics, the
situation and the individual-situation interaction' (Zuckerman, 1991). (A
'state' denotes behaviour at a given moment.) Some consistency has been
demonstrated for sociability in children (Kagan & Moss, 1962) and for
various traits in adults, in particular extraversion-sociability.
In relation to the consistency of temperaments, Rutter (1987) has pointed
out that differences between children can be measured with reasonable
reliability in relation to activity level and sociability, and that such
assessments can have clinically significant implications; for example,
measures of temperaments in children have been correlated with subse-
quent emotional disturbance in different contexts. (Reliability is the extent
to which a test would give consistent results on being applied more than
once to the same people under standard conditions.)
Behaviour reflects interactions between traits, temperaments and the
environment, and data from several measuring instruments, as well as from
several situations, may be needed for useful personality assessment. Also,
the relationships between different behaviours may reflect a consistent
pattern. Consistency of behaviour can be influenced by the degree of
specificity of the environment and, in studies on monkeys, autonomic
reactivity was reasonably consistent only when evaluated in stressful
situations (Suomi, 1983). (This example has a direct parallel with clinical
practice, as some patients can show consistent but episodic maladaptive
behaviour, but only in the context of a life-crisis.) Another variable
affecting consistency of behaviour is the effect of 'developmental' changes
(i.e. changes related to increasing age), as a trait can give rise to different
expressions of related behaviour at different ages; for example, a low level
of activity may initially lead to the seeking of maternal closeness but later to
an avoidance of new situations (Kagan, Reznick & Snidman, 1986; Rutter,
1987). Also, passivity and lack of physical adventurousness in childhood
have been shown to predict adult non-competitiveness and certain charac-

teristics of sexual behaviour (Kagan & Moss, 1962; Ryder, 1967).
Approaches
Personality has been evaluated from three main perspectives. The first
focusses on the individual's social interactions, in particular early parent-
child, peer, and close confiding relationships (Wolkind & Rutter, 1985;
Hinde & Stevenson-Hinde, 1986); for example, Hodges & Tizard (1989)
reported that early institutional care was associated with certain patterns of
social interaction in later childhood. The second approach is cognitive,
involving examination of consistent patterns of thoughts, feelings and
attitudes about self and environment, for instance low self-esteem, which
has been claimed to be an important causal factor for some depressive
disorders (Brown & Harris, 1978; Peterson & Seligman, 1984). In the third
approach, personality traits (such as empathy, suspiciousness, and sensa-
tion-seeking), can be variously defined and classified.
Personality disorders
Historical developments
Early concepts of PD were developed in the 19th century (Berrios, 1993),
for example, Prichard (1835) defined 'moral insanity' as
a form of mental derangement in whom the moral and active principles of the mind
are strongly perverted or depraved, the power of self-government is lost or greatly
impaired, and the individual is found incapable, not of talking or reasoning on any
subject proposed to him, but of conducting himself with decency and propriety in
the business of life.
Subsequently, Maudsley (1868) described a similar but narrower cate-

gory for someone with 'no capacity for true moral feeling ...', and Koch
(1891) used the term 'psychopathic inferiority' to denote those with clearly
abnormal behaviour in the presence of a normal mental state. Kraepelin
(1905) described 'morbid mental states in which the peculiar disposition of
the personality must be considered the real foundation of the malady',
while Schneider (1923) termed all PDs as 'psychopathic', and described ten
different types. This broad category of'psychopathic personality' contrasts
with subsequent narrower definitions of 'psychopathy', with an emphasis
on antisocial behaviour, lack of remorse, unreliability and selfishness
(Cleckley, 1941), although Henderson (1939) also included those who were
predominantly passive or inadequate.
Individuals whose PD is associated with repeated antisocial behaviour
come into the province of both law and medicine, providing society with an
unresolved dilemma about the concept of personal responsibility for one's
actions. However, from a biological perspective, it appears that genetic
factors are one type of variable that can predispose individuals to commit
aggressive and other antisocial acts.
Definitions
Despite the complexities of the concepts of personality, PDs are established
diagnoses in medical practice and are found in the two main recent
classifications of mental disorders, the International Classification of Dis-
eases (ICD-9 and -10, World Health Organization, 1978, 1992) and the
Diagnostic and Statistical Manual of Mental Disorders (DSM-III, DSM-
III-R, DSM-IV, American Psychiatric Association, 1980,1987, 1994). But
as the characteristics of PD usually differ only in degree or frequency from
the features that are present in the majority of the population, the
boundaries between normal personality traits and those of a PD are
generally arbitrary.
Schneider (1923) stated that PDs are those 'abnormal personalities who
suffer through their abnormalities or through whose abnormalities society
suffers', while Rado's definition (1953) focussed on the organization of
higher mental activity, i.e. 'Disturbances of psychodynamic integration
that significantly affect the organism's adaptive life performance, its
attainment of utility and pleasure'. More recently, Rutter (1987) considered
that PDs are 'a persistent, pervasive abnormality in social relationships and
social functioning generally'.
The World Health Organization's definition of PD for ICD-9 (1978)
stressed that signs usually appear early in life, and PD was said to consist of
deeply ingrained maladaptive patterns of behaviour generally recognizable by the
time of adolescence or earlier and continuing throughout most of adult life,
although often becoming less obvious in middle or old age. The personality is
abnormal either in the balance of its components, their quality and expression, or in
its total aspect. Because of this deviation or psychopathy the patient suffers or
others have to suffer and there is an adverse effect upon the individual or on society.
The other main classifactory system, the DSM (i.e. DSM-III, DSM-III-
R and DSM-IV) defines personality traits as 'enduring patterns of perceiv-
ing, relating to, and thinking about the environment and oneself, and are
exhibited in a wide range of important social and personal contexts', and
continues: 'it is only when personality traits are inflexible and maladaptive
and cause either significant functional impairment or subjective distress
that they constitute personality disorders. The manifestations of personal-
ity disorders are often recognizable by adolescence or earlier and continue
throughout most of adult life, though they often become less obvious in
middle or old age' (American Psychiatric Association, 1987). In the DSM-
IV, the general diagnositic criteria for a PD consist of: 'an enduring pattern
of inner experience and behaviour that deviates from the expectations of
the individual's culture' (which is manifested in two or more of the
following: cognition, affectivity, interpersonal functioning and impulse
control); 'the enduring pattern is inflexible and pervasive across a broad
range of personal and social situations'; 'the enduring pattern leads to
clinically significant distress or impairment in social, occupational, or other
important areas of functioning'; 'the pattern is stable and of long duration
and its onset can be traced back at least to adolescence or early adulthood';
'the enduring pattern is not better accounted for as a manifestation or
consequence of another mental disorder'; and 'the enduring pattern is not
due to the direct physiological effects of a substance (e.g. a drug of abuse, a
medication) or a general medical condition (e.g. head trauma)'.
The 10th revision of the ICD (ICD-10, World Health Organization,
1992) defined a specific PD as
a severe disturbance in the characterological constitution and behavioural tenden-

cies of the individual, usually involving several areas of the personality, and nearly
always associated with considerable personal and social disruption. Personality
disorder tends to appear in late childhood or adolescence and continues to be
manifest into adulthood. These types of conditions comprise deeply ingrained and
enduring behaviour patterns, manifesting themselves as inflexible responses to a
broad range of personal and social situations. Personality disorders ... are
developmental conditions, which appear in childhood or adolescence and continue
into adulthood.
In addition, the ICD-10 includes the diagnosis of'enduring personality

changes' that are acquired, usually during adult life, 'following another
medical disorder or severely adverse environmental stress'.
The disease concept and PDs

The concept of disease implies maladaptive biological functioning and can
be defined in various ways. One of the most basic criteria is an association of
disease with excess mortality, and if this is applied to individuals with some
types of PDs, there is convincing evidence that the disease concept is
justified. For example, antisocial PD was associated with increased risk of
unnatural death in a 7 year follow-up of 500 psychiatric outpatients of a
psychiatric service; this was related to substance abuse, accidents and
suicide (Martin, 1986). But the traditional medical approach of applying no
more than a few categorical labels (i.e. 'diagnoses') to an individual patient
is not well suited to the identification of PDs, as many patients qualify for
several co-occurring PD diagnoses. Therefore, it may be appropriate to
modify the traditional format of medical diagnostic systems for the
consideration of PDs.
Scadding (1988, 1990) has pointed out that 'diagnosis is the process by
which a patient's symptoms and signs are assessed and investigated, with a
view to the categorisation of his/her case with others of a similar sort which
have been studied in the past, in order that the patient may benefit from the
application of established knowledge to the problem'. He has also noted
that the concept of'a disease' is 'logically heterogeneous', as it may refer to
a group of signs and symptoms (a syndrome), to the effects of a specific
disorder of bodily structure or function, or to the effects of a specified causal
agent. The aims of psychiatric research include the investigation of the
associations of behavioural syndromes with disorders of bodily structure or
function, causal factors, prognosis and efficacy of treatments. A diagnostic
label for a disease is always provisional (Mindham, Scadding & Cawley,
1992), as it forms an hypothesis that is 'a conjecture about the relationship
between two or more concepts' (Freeman & Tyrer, 1989). Research often
leads to revisions of disease concepts, and classification is an essential part
of this process.
There is no clear dividing line between normal and abnormal personali-
ties, but it is possible to make reliable arbitrary distinctions, perhaps based
on a cut-off score of dimensions that have been measured. However, in
practice, the clinician or health-care worker is not concerned with the
threshold for a PD diagnosis, but with the questions 'is this person's PD
significantly maladaptive?' (i.e. does the individual suffer or do other people
suffer?) and, if the answer is 'yes', 'what interventions, if any, are
appropriate?'
There have been many criticisms of the disease concept in relation to the
use of PD diagnoses; for instance that a PD diagnosis is unreliable and that
it is often used as a derogatory label on the basis of inadequate assessment,
i.e. 'little more than a moral judgement masquerading as a clinical
diagnosis' (Blackburn, 1988). Also, it has been claimed that the process of
assigning a PD diagnosis can be used as an excuse to deny patients
Principles of classification 13
appropriate help (Lewis & Appleby, 1988) as, in contrast to other mental
disorders, a PD diagnosis is not usually considered to take away significant
personal responsibility from the patient for his/her actions. As individuals
with PD are often disliked because of their behaviour, the use of the PD
diagnostic category may have an implicit meaning, namely that the person
is not deserving of care and treatment. But while such criticisms of the use of
PD categories may sometimes be justified, they should be directed to
clinicians and health care workers, rather than to the relevant concepts and
classificatory systems.
Assessment
In routine practice, the clinician assesses PD by an interview in which
various aspects of the patient's past life and conscious awareness are
explored. Various topics are routinely covered but questions are not usually
presented in a standard format. If possible, an informant is also seen. It is
particularly important to elicit information about the major problems
related to PD, which include antisocial behaviour, dependence on others,
problems in sustaining relationships and attitudes to other people.
Various methods can be used for a more detailed assessment, in
particular for research studies. These may involve self-report question-
naires (for patient and an informant), questionnaires for the clinician
(which may involve rating scales), and structured interview schedules.
Whatever the method of PD assessment, other co-occurring abnormal
aspects of the patient's mental state may produce distortions of the
findings, and these are known as 'trait-state artefacts'. For example, it has
been shown that patients who are depressed or anxious often do not
provide accurate reports of their previous functioning. However, some
assessment methods have been shown to be relatively reliable for some PD
features despite variations in anxiety and depression between separate
assessments (Loranger et a/., 1991; Brown et ai, 1992). A further problem is
a tendency for patients to acknowledge fewer maladaptive PD traits at a
subsequent assessment.
Principles of classification
Classification in medicine has been claimed to be a 'necessary preliminary
to almost any useful communication' (Kendell, 1983); however, any
classificatory system involves underlying assumptions, which, if incorrect,
may hold back the development of knowledge. The various systems have
been mainly concerned with disorders that people have but can also be
applied to the people (patients) themselves. Any classification involves the

definition of membership of a population (e.g. a population of mental
disorders) and then, usually, separating its members into two or more
groups, which can be called 'types', 'classes' or 'categories'. The main uses
of any classification of disorders are to record phenomena, to understand
causes and to predict outcome, and it is convenient if one patient is
associated with just one class of disorder, particularly in respect of the
predictive function. Although a classificatory process always consists of
categories, a categorical system (e.g. classifying individuals as below or
above a specified height) can be supplemented by complementary dimen-
sional information (e.g. the measured height). A dimension involves the
identification of one characteristic (e.g. height in centimetres) among a
group of related variables (e.g. a range of heights to the nearest centimetre).
PDs are associated with a wide range of behavioural patterns, and an
appropriate classification should aim to include all relevant behaviour, i.e.
to provide adequate coverage. Even if PDs consisted of just one class, this
would form part of a wider classification of mental disorders and there is
evidence that even such a broad diagnostic category of PD has 'satisfactory
reliability and reasonable predictive validity', while the 'reliability and
temporal stability of most trait-defined subtypes has been shown to be quite
poor', except for antisocial PD (Rutter, 1987). Also, for some categories of
PD reported in the literature, the internal consistency of the syndromes (i.e.
the average covariance among a set of scores) has been low and their
validity has been unproven.
The process of defining and classifying personality and PD has been
criticized for not paying sufficient attention to mental activity and interper-
sonal relationships, due to a focus on overt behaviour, and for ignoring the
unique aspects of an individual, which are so important in our concept of a
human person (Ferguson & Tyrer, 1988). However, a focus on the unique
aspects of an individual's PD would not produce knowledge that would be
useful to others.
A classification of PDs presents considerable difficulties because of the
range, complexity and intercorrelations of relevant behaviours, their
interaction with environmental variables and with other mental disorders,
and the lack of natural boundaries between the attributes of most people
and most of the PD criteria. In addition, boundaries are unclear between
various PD subtypes, and between PDs and some other mental disorders.
Also, there is some overlap between PDs, as some criteria occur in more
than one PD syndrome.
Criteria for PDs are not all constructed on the same basis; some are
examples of a trait, i.e. a group of related behaviours, while other criteria

consist of relatively specific examples of behaviour. As no list of behaviours
for a trait could be exhaustive, there are often difficulties in selecting specific
examples of behaviour as diagnostic criteria, as these may be subject to
relatively specific effects of variables such as sex or social class.
Methods of classification
Categories
In an ideal classificatory system, each category should contain members
that are identical (i.e. all criteria must be met by all members) and should be
mutually exclusive (i.e. a member must be in one class only). Also, all the
categories should be jointly exhaustive (i.e. there should be one category
that is suitable for every member of the population being studied). Disease
categories are best suited to populations of disorders in which there are
sharp boundaries between the disorders. When PD criteria relate to
behaviour that can be found in most individuals (e.g. aggression), an
arbitrary dividing line between normality and disorder can only be avoided
if there is a point of rarity in the severity or frequency of the behaviour.
Thus, if human beings were all either very docile (except in extreme
circumstances) or so aggressive that breaches of the law occurred regularly,
then a categorical approach could easily be applied because intermediates
would be rare. Unfortunately, points of rarity are not found for most PD
criteria or most PDs, except, perhaps, for a minority of paranoid, schizoty-
pal, borderline and antisocial PD criteria (Frances, 1982; Livesley &
Jackson, 1992).
A 'monothetic' category of disease involves all its members meeting all
the criteria, while just a minimum number of positive criteria are needed for
membership of a 'polythetic' category. Thus, for the latter, there is a
variable level of resemblance to the 'ideal' or 'prototype' member of the
class, which meets all the criteria. But a polythetic system is misleading, as a
single PD diagnosis is applied to each example of a group of disorders that
shows considerable heterogeneity; for example, there are 93 ways in which
the DSM-III-R's polythetic criteria for borderline PD can be met.
In a polythetic system it is difficult to determine an appropriate threshold
in relation to the number of criteria to be met for diagnosis, and the decision
is usually arbitrary. But a variant of a polythetic classificatory system can
involve abolition of the threshold concept for ^n all-or-none diagnosis, and
its replacement by a score for each category of disorder, based on the
number of criteria met. Also, certain criteria can be designated as essential

for class membership or given special weighting.
Dimensions
Data from a population of diseases can be recorded not only by a system of
categories but by dimensional scores or ratings, each of which may involve
a continuous distribution of the amount of a variable. However, it must be
noted that categories and dimensions can be complementary, and that
dimensional scores are easily converted into categories; for example, if
severity of aggression has been rated on a 10 point scale (10 = very severe),
an arbitrary threshold of a score of 6 or more can be used to define a
category of aggressive disorder.
PDs have been traditionally described in categorical terms in the
psychiatric literature, although the dimensional nature of most PD features
has led to an increasing awareness that a more dimensional approach to
measurement and classification will be required. But even within recent
categorical systems, such as the DSM-III-R, the total score of positive PD
criteria for each PD diagnosis can provide a dimensional assessment that
may correlate with the degree of social impairment for many patients,
although some may show relatively severe disorder restricted to a small
number of PD criteria. Threshold values for the number of positive criteria
required to produce a categorical diagnosis have been generally arbitrary,
but it may be possible to link a cut-off point with clinically-significant
variables such as a specified degree of social impairment.
Influential classifications of PDs

The third edition of the American Psychiatric Association's DSM classifi-
cation, DSM-III (American Psychiatric Association, 1980), was the first
internationally-used system that specified criteria for PD diagnoses,
although it was often unclear how they should be met. The DSM-III was
revised in 1987 (DSM-III-R) and 1994 (DSM-IV), and the first two versions
have been the anchor-points for much of the psychiatric literature con-
cerned with PD. The World Health Organization's 9th revision of their
International Classification of Diseases (World Health Organization, 1978)
provided only brief clinical descriptions for its diagnostic categories of PDs,
but the 10th revision, the ICD-10 (World Health Organization, 1992), is
similar to the DSM-III-R and DSM-IV in providing a set of criteria to be
evaluated for each specified PD. The ICD-10 uses a polythetic system in
which at least three of the criteria need to be met for each PD diagnosis.
Also, diagnostic criteria for research have been published, closely based on
the ICD-10 (World Health Organization, 1993).
A diagnostic criterion may require the clinician to make a judgement on a
wide range of behaviours (e.g. 'expects, without sufficient basis, to be
exploited or harmed by others') or involve relatively specific examples of
behaviour (e.g. 'questions, without justification,fidelityof spouse or sexual
partner') (American Psychiatric Association, 1987). Also, polythetic cri-
teria may vary in relation to the degree of association with the diagnosis of
the disorder, and to their role in distinguishing the syndrome from other
disorders.
There are advantages in minimizing the role of clinical judgement for the
application of a PD criterion, as this should increase reliability (i.e.
repeatability) of assessment, although this may lead to relevant criteria
being discarded if they rely heavily on clinical evaluation. This could
produce an undesirable reduction in the validity of the syndrome, if the full
range of clinically-important phenomena are not being assessed.
Despite the importance of categorical classifications of PDs, their
limitations have been recognized, and it is likely that such systems will be
increasingly supplemented by dimensional information, both in research
and clinical practice. Nevertheless, PD categories have stood the test of
time, as they provide simple, convenient, and useful summaries of predomi-
nant maladaptive behaviours. Although all-or-none diagnostic PD cate-
gories involve arbitrary thresholds and lose information, many decisions in
clinical practice are also binary, i.e. to treat/not to treat, or to admit/not to
admit, so that the use of all-or-none diagnoses can provide a functional
model for these management decisions. On the debit side, the use of various
diagnostic categories for PDs can lead to concepts of PDs that are restricted
to the various prototypes (i.e. 'ideal' examples of a category); this is
misleading as most patients with PD show a mixture of positive criteria
from more than one PD syndrome.
Categories ofPD in the main international classifications

The DSM-III and its revised version have been the main influences in PD
research since 1980 and, like every classification, has underlying assump-
tions. The DSM-III-R and DSM-IV, with their format of polythetic PD
categories, assume that some members of each class are better examples
than others, and that the categories reflect natural classes, i.e. 'real-world
correlational structure of frequently encountered problematic personality
Table 1.1 Categories of personality disorder common to the DSM-III-R,

DSM-IV and ICD-10 classifications
DSM-III-R and DSM-IV category Corresponding ICD-10 category

Paranoid Paranoid
Schizoid Schizoid
Antisocial Dissocial
Borderline Emotionally unstable
Impulsive type
Borderline type
Histrionic Histrionic
Obsessive-compulsive Anankastic
Avoidant Anxious (avoidant)
Dependent Dependent
Personality disorder not otherwise specified Mixed and other
features'. (Also, the PD categories are not assumed to be mutually

exclusive, as there is overlap between some of their characteristics.) Further
assumptions in DSM-III-R and DSM-IV are that each set of criteria for a
specified PD displays a degree of internal consistency, i.e. 'features of the
same syndrome should correlate more highly with each other than with
features representing a different syndrome' (Morey, 1988), and that the
various PD syndromes can be grouped together in three superordinate
classes. These are designated as: cluster A, odd or eccentric; cluster B,
dramatic, emotional or erratic; and cluster C, anxious or fearful, with the
further assumption that each PD syndrome shows greater internal consist-
ency relative to other members of its superordinate group, compared with
the other PDs.
DSM-III-R gives 11 criteria sets for PD diagnoses, together with a
residual category of 'Personality disorder not otherwise specified' (which
corresponds to the 'Mixed and other' category in the ICD-10 classification).
There are also two additional 'Proposed diagnostic categories needing
further study' - i.e. self-defeating and sadistic PDs. Table 1.1 shows eight of
the 11 specified DSM-III-R PD syndromes that closely correspond to
syndromes in the 10th revision of the ICD, i.e. paranoid, schizoid,
antisocial, borderline, histrionic, obsessive-compulsive, avoidant and
dependent PDs. The eight DSM-III-R PD categories shown in Table 1.1 are
also found in DSM-IV (American Psychiatric Association, 1994). Several
other PD syndromes have been described and are listed in Table 1.2.
Schizotypal, narcissistic, passive-aggressive, self-defeating and sadistic
Table 1.2 Various categories of personality disorder

(PD) additional to those shown in Table 1.1
Category of PD Source
Schizotypal *DSM-III-R
Narcissistic *DSM-III-R
Passive-aggressive DSM-III-R
Self-defeating DSM-III-R
Sadistic DSM-III-R
Organic ICD-10
Enduring personality changes ICD-10
Accentuation of personality traits ICD-10
Depressive personality Various
Note: * and DSM-IV.
PD syndromes are derived from DSM-III-R but do not have equivalent

categories in ICD-10. However, the ICD-10 recognizes three further
categories: 'organic personality disorder' due to brain disease, damage and
dysfunction; 'enduring personality changes', not attributable to brain
damage and disease (i.e. after catastrophic or excessive prolonged stress or
following a severe psychiatric illness); and 'accentuation of personality
traits'. Leonhard (1968) was the first to use the term 'accentuated' to denote
personality characteristics that are only maladaptive in an unusually
demanding or stressful environment, and this category can be conceptua-
lized as intermediate between normal personality and PD. It should be
noted that the ICD-10 categories of 'cyclothymia' and 'schizotypal dis-
order' are not classified as PDs, but appear elsewhere in sections related to
mood disorder and schizophrenia.
Opinions have varied about the validity of a 'depressive personality'
category. This was included in ICD-9 but not in the 10th revision, while it is
not a DSM-III-R category. However the DSM-III-R's category of'dysthy-
mia' (a type of depression) is described as chronic and sometimes beginning
in childhood, and early-onset dysthymia has been considered by some to be
synonymous with 'depressive personality' or 'charactological depression'
(Widiger & Shea, 1991). In DSM-IV, 'depressive personality disorder' is
included as a criteria set 'for futher study'.
The ICD-10 recommends that its PD categories should be used to reflect
the patient's most frequent or conspicuous behavioural manifestations of
PD.
Table 1.3 Possible modifications to a poly the tic classification of personality

disorders (PDs) as found in the DSM-III-R, DSM-IV and ICD-10
classifications
1. Changes in the use of existing criteria

e.g. Specify certain criteria as essential for diagnosis
Give extra weight to certain criteria
Change minimum number of positive criteria required for diagnosis
2. Changes in individual criteria or the categories of personality disorders
e.g. Change individual criteria
Change overlap between criteria in different PD syndromes
Change or add exclusion criteria (i.e. criteria which do not allow the
diagnosis of PD)
Change the number of PD categories
3. Additional strategies for the classification of personality disorders
e.g. Add criteria which differentiate PD from other disorders
Operate an hierarchy in recording co-occurring PDs
Use compound diagnostic categories (i.e. sub-groups defined by
additional criteria)
Add dimensional information
Add complementary information (e.g. global functioning)
Disadvantages ofpolythetic PD categories

As the DSM-III-R, DSM-IV and ICD-10 have provided a series of
polythetic categories for PD diagnoses, the consequences of such a system
must be considered.
A group of individuals who receive such a PD diagnosis are hetero-
geneous in respect of the relevant disorder, as there are many different
combinations of the minimum number of positive criteria required,
together with variations in the number of positive criteria above the
diagnostic threshold. Also, the minimum number of criteria for diagnosis is
arbitrary, and whatever diagnostic threshold is chosen, an all-or-none
diagnosis loses information. (However, if all criteria were required to be
positive before diagnosis, this would not reflect the heterogeneity that is
actually found.) In addition, information is lost by an all-or-none rating for
each individual criterion, while different criteria for each PD in DSM-III-R,
DSM-IV and ICD-10 are considered to be equally valid, despite apparent
variations in clinical importance. Also, the threshold for rating each
individual criterion as positive is usually arbitrary.
Many patients fulfil criteria for two or more PD diagnoses, while some
criteria contribute to more than one PD. Unfortunately, this is not suited to
routine clinical practice, as it is cumbersome to give several PD diagnoses to
one patient, and a list of several co-occurring PDs disguises relatively severe
maladaptive behaviour in just one category.
Potential modifications to PD categories

As any classification is an hypothesis that invites testing and revision, some
possible changes in the most influential polythetic classifications are
summarized in Table 1.3.
Firstly, there are several potential changes in the use of existing criteria in
DSM-IV and ICD-10. Some 'core' criteria could be specified as essential for
the diagnosis and/or given extra weight by increasing the score assigned to
their positive rating. Also, the minimum number of positive criteria
required for diagnosis could be changed. However, the identification of
core criteria is difficult, and core criteria may not exist for some syndromes.
Secondly, changes could be made to the individual criteria and the
number of PD categories. The total number of criteria for each PD,
together with the minimum number required for diagnosis, are major
variables affecting the degree of heterogeneity of a group of diagnosed PDs,
while a consideration of the nature of PD criteria must recognize that they
consist of a mixture of relatively broad traits and relatively specific
examples of behaviour. Some specific behaviours will be better examples of
a trait than others, and a choice of several behaviours to indicate a positive
criterion will usually give increased reliability compared with the use ofjust
one behavioural example. The assessment of a trait (in contrast to a
relatively specific behaviour) requires a greater degree ofjudgement, and, in
a set of criteria, the balance between those that rely heavily on clinical
judgement and those that are related to specific behaviour can be reviewed.
If the balance is changed in favour of specific behaviours, the reliability is
likely to be increased but perhaps at the expense of the validity, if aspects of
the disorder that rely heavily on clinical judgement are omitted. Overlap
may occur between the various sets of PD criteria; for example, in the
DSM-III-R, one criterion is identical in schizoid, schizotypal and avoidant
PDs. But, while it would be possible to reduce such overlap by omitting the
relevant criterion, this may not be appropriate, if the overlap reflects the
reality of these disorders. Thus, excessive pruning of criteria may artefac-
tually restrict the recognition of the syndromes. Exclusion criteria are
provided for some PDs in DSM-III-R, and the use of this strategy could be
extended. For example, antisocial PD could be excluded as a diagnosis in
the presence of manic-depressive disorder, as antisocial behaviour can
often be a feature of mania, and if mania is present there is a chance of an
incorrect diagnosis of antisocial PD being made. But, although this will

reduce co-occurrence of PD diagnosis with other mental disorders, the
increased homogeneity of those who receive the diagnosis may not reflect
the reality, as different mental disorders can co-occur. However, this
approach may be valid for research samples. Other changes in classification
might involve the number of PD categories, for instance, various syn-
dromes that have been classed as PDs, such as schizotypal PD and
cyclothymic personality, have been increasingly considered to be more
closely related to other mental disorders. The relationship between avoid-
ant PD and phobic disorders is another area of uncertainty. It may also be
appropriate to combine some of the existing PD categories.
Thirdly, additional strategies for the classification of PDs could be
introduced. Criteria could be added with the aim of differentiating a PD
syndrome from other mental disorders, and an hierarchy could be intro-
duced for the recording of co-occurring PDs; for example, only one PD
could be recorded, perhaps the one that is judged to have the greatest effect
on social functioning. The device of the compound diagnostic category
could be used to specify, within a group defined by each PD diagnosis, a
sub-category associated with another variable, for example, an history of
childhood physical or sexual abuse (Mindham et al.9 1992). Also, dimen-
sional information could be used to supplement the PD diagnoses, perhaps
based on the degree of specific behaviour, the range of related behaviours,
the frequency of specific behaviour, and the range of situations in which the
behaviour is displayed. For instance, the Personality Assessment Schedule
is a system of clinical assessment that is based on dimensional ratings of 24
traits judged to be relevant to PDs (Tyrer & Alexander, 1988), and its
ratings can also be converted to categories that correspond to those of the
ICD and DSM. This and other systems that have attempted to utilize
dimensional information may influence further revisions of the two major
international classificatory systems. However, the question of which
changes are 'correct' is misleading, as one classificactory system will suit
one purpose but not necessarily another. Finally, complementary infor-
mation such as the severity of the adverse effect of behaviours on social and
occupational functioning could be added.
Boundaries between personality disorders and other mental disorders

In addition to problems in defining the boundaries between different PDs,
there can be overlap between PDs and other mental disorders when the
syndromes share common features.
Boundaries with other mental disorders 23
The nature of the relationships between PDs and other mental disorders
is complex and variable; for example, certain PD characteristics, such as
antisocial behaviour, may be responsible for adverse life events that are
then causal factors for a depressed mood. Also, patterns of conscious
mental activity, such as persistent wariness, indecision, suspiciousness or
pessimism, may lead to mood disorder involving symptoms of anxiety and
depression. These are examples of PD apparently causing mood disorder
but the reverse can also occur, as specified in the ICD-10, which notes that
'enduring personality changes' can follow various adverse events, including
'serious psychiatric disorder' such as manic-depressive disorder. Also,
although PD and another mental disorder may co-occur independently, the
PD may influence the expression or the severity of the other disorder; for
example, a patient with an antisocial PD is likely to become increasingly
antisocial if he/she develops a schizophrenic illness. Another possible
association is when the PD and another co-occurring mental disorder may
reflect various features of the same disorder; for instance, it is likely that
similar genetic vulnerability can be related to both schizophrenia and
schizotypal PD.
Boundaries between PDs and other mental disorders may be unclear
because of overlapping diagnostic criteria for the two syndromes under
consideration, or because more than one diagnoses has been arbitrarily
imposed upon dimensions (e.g. frequency or severity) associated with the
relevant behaviours. These issues have been given considerable attention in
relation to the interface between PD and depressive or anxiety syndromes.
Digman (1990) has pointed out that some features of depression and
anxiety have also been generally accepted as features of PDs, for instance,
'neuroticism', involving 'trait depression and trait anxiety'. Some episodic
'states' with features of depression and/or anxiety have produced diagnos-
tic categories that provide meaningful distinctions between some examples
of depression or anxiety and PDs (for example, episodic depression in the
context of bipolar manic-depressive disorder), but distinct boundaries
between PDs and some other examples of depression and anxiety are
lacking (Chaplin, John & Goldberg, 1988). Depressive syndromes consist
of heterogeneous groups of signs and symptoms that reflect a range of
causal factors of varying specificity. For some individuals, genetically-
determined vulnerability to abnormal brain function appears to be import-
ant and is an example of an 'endogenous' or internal causal factor, although
such vulnerabilities may themselves be heterogeneous. There is evidence
that environmental life events may precipitate a depressive syndrome in
those who have an endogenous vulnerability, although in some patients,
endogenous mood changes can appear 'out of the blue'. Other hypothe-
sized types of causal factors for depressive syndromes include identifiable
biological factors (e.g. brain disease, drug administration, infections,
nutritional factors, and hormonal influences), psychologically meaningful
reactions to (and interactions with) life events, and personality or PD.
There is some evidence that certain features of PDs may make an individual
particularly prone to the development of some depressive symptoms. For
example, borderline PD is associated with emotional instability, a tendency
to develop transient depressed moods, irritability or anxiety, inability to
sustain relationships, and an unstable lifestyle. In addition to mood
disorder being an integral part of this PD, individuals with these character-
istics usually experience more stressful life events, which may have addi-
tional causal effects for depressed mood or anxiety (Seivewright, 1987).
Also, it has been proposed that certain personality characteristics should be
recognized as a category of'depressive personality disorder' (Phillips et aL,
1990), involving excessively negative and pessimistic beliefs about oneself
and other people, with relatively persistent depressed mood (Hirschfeld &
Shea, 1992).
The hypotheses that give a major causal role for PD in the development
of some depressive syndromes have led to several diagnostic labels for the
corresponding depressive sub-types, i.e. 'characteriological depression'
(Akiskal, Rosenthal & Haykal, 1980), 'affective spectrum disorder'
(Widiger & Shea, 1991; McElroy et aL, 1992) and the 'general neurotic
syndrome' (Tyrer et aL, 1992).
Before leaving the interface of PD and depressive disorders, mention
must be made of'cyclothymic' PD which, according to the ICD-9, involves
a pattern of changes of mood involving periods of persistent depression or
persistent elation. But in the subsequent 10th revision of the ICD, cyclothy-
mia was described as 'a persistent instability of mood involving numerous
periods of mild depression and mild elation'. It was considered to be a mood
disorder rather than a PD, and it is likely that this syndrome is genetically
related to the more typical forms of manic-depressive disorder.
While anxiety is often a part of a depressive syndrome, various anxiety
disorders can be diagnosed independently, and there can be difficulty in
defining a boundary between certain features of PDs and certain anxiety-
related syndromes such as social phobia. As described in DSM-III-R, the
latter involves 'a persistent fear of one or more situations in which the
person is exposed to possible scrutiny by others ...' while, in the same
classification, the features of avoidant PD include being unwilling to get
involved with people unless certain of being liked and being reticent in
Boundaries with other mental disorders 25
social situations because of a fear of saying something inappropriate or
foolish. It seems likely that these two syndromes are not clinically distinct
but reflect overlapping variants along a spectrum of related phenomena
(Brooks, Baltazar & Munjack, 1989).
Another group of mental disorders that can be difficult to distinguish
from PDs (in particular those PDs with antisocial behaviour) consists of
'impulse-control disorders' (McElroy et al., 1992). These include intermit-
tent explosive disorders (instability of mood with outbursts of anger or
violence), kleptomania (a persistent impulse to steal, often without econ-
omic motive), pathological gambling, and pyromania (an irresistible
impulse to start fires). Also, trichotillomania (an abnormal desire to pull
out one's hair) can appear to be similar to some features of obsessive-
compulsive PD.
Another boundary problem involves the frequent associations between
antisocial PD (variously defined) and substance abuse disorders (Docherty,
Fiester & Shea, 1986). The DSM-III-R criteria for antisocial PD include
failure to conform to social norms with respect to lawful behaviour and an
inability to sustain consistent work behaviour, but these can also occur in
syndromes involving substance abuse such as DSM-III-R's 'psychoactive
substance dependence'. It appears that either there is considerable overlap
between the features of these disorders or that one syndrome is being given
more than one diagnosis (Frances, Widiger & Fyer, 1990).
The next area of diagnostic uncertainty involves schizoid, schizotypal
and paranoid PDs, schizophrenia and Asperger's syndrome. Schizoid PD
involves a relative indifference to personal relationships with social isola-
tion, while schizotypal PD includes features that can appear to be mild
forms of some of the features of schizophrenia, for example, odd beliefs,
eccentric behaviour, unusual perceptual experiences and odd speech.
Evidence has been accumulating that schizotypal PD should be considered
as a variant of schizophrenia (Kendler, Eaves & Strauss, 1981) and is
classified as such in the 10th revision of the ICD. Asperger's syndrome is
characterized by abnormal social interactions and a restricted repertoire of
interests and activities in older children but without significant general
delay in language or cognitive development. Rutter (1987) noted a possible
association between schizoid PD and Asperger's syndrome (and childhood
autism), and there is some indication of a familial association between
paranoid PD and schizophrenia (Kendler & Gruenberg, 1984).
Finally, the distinction between obsessive-compulsive PD and obsessive-
compulsive neurosis can also be arbitrary, for instance, when the former
involves excessive attention to detail and organization that can differ only
in degree from a compulsive ritual involving clearly abnormal patterns of
behaviour.
Further revisions of influential classification of mental disorders may
improve the scientific utility of the various diagnoses. If possible, similar
syndromes with different names should be merged or the overlap between
criteria for two or more diagnoses should be reduced. But this may not
always be appropriate if two defined syndromes are both useful despite
considerable overlap.
Co-occurrence of personality disorders with other mental disorders

The process of PD diagnosis involves matching signs and symptoms to
definitions in a specified classification, and co-occurrence of a PD diagnosis
with other PDs, and with other mental disorders, is common. Co-occur-
rence does not necessarily imply any causal relationships, although there
can be complex interactions between PDs, other mental disorders and the
social environment.
Common examples of co-occurrence are responsible for some of the
problems of defining boundaries between PDs and other mental disorders
that have been described, in particular between PDs and some depressive
syndromes, anxiety states and substance abuse disorders. As PD can
complicate the management of individuals with any other mental disorder,
such co-occurrence is of major importance in clinical practice. One of the
most common psychiatric diagnoses is a depressive syndrome, and in
outcome studies of the treatment of depressive disorders, patients with PDs
responded less well to various forms of treatment (Shea et aL, 1990; Tyrer et
aL, 1990). However, as has been described, the problems of defining a
boundary between PD and some forms of depression or anxiety can be
difficult; Tyrer et aL (1992) have argued that, for some depressive syn-
dromes with PD, the PD characteristics of excessive timidity, poor self-
esteem, avoidance of anxiety-provoking situations and dependence on
others, are invariably associated with depression and anxiety, and that such
a patient should be considered to have just one syndrome.
PD has also been reported in association with many other mental
disorders, for example (using DSM-III-R terminology) eating disorders,
somatization disorders, schizophrenia, manic disorders, sexual disorders,
hypochondriasis, paranoid states, dissociative disorders and obsessive-
compulsive disorder.
Although there are a number of categories in the main classifications of
Epidemiology overview 27
PDs, Rutter (1987) has pointed out that the majority of patients with
clinically significant PD show a variable mixture of antisocial, borderline,
histrionic and narcissistic PDs. Patients with severe PD are often con-
sidered to be 'difficult patients' and frequently present with other co-
occurring mental disorders, in particular depressive syndromes, episodes of
self-harm, anxiety states, substance abuse and eating disorders (Higgitt &
Fonagy, 1992).
Epidemiology overview
The DSM-III-R definition of PD is somewhat vague, as it encourages the
clinician to decide when aspects of personality 'cause either significant
functional impairment or subjective distress'. Therefore the diagnostic
threshold is difficult to determine both in practice and in research studies.
However, it is generally agreed that PDs are of considerable clinical
importance: 'the personality disorders constitute one of the most important
sources of long-term impairment in both treated and untreated popula-
tions. Nearly one in every 10 adults in the general population and over one-
half of those in (psychiatrically) treated populations, may be expected to
suffer from one of the personality disorders' (Merikangas & Weissman,
1986).
Prevalence of PD in different settings, estimated by various methods, has
been reviewed by Casey (1988). Using a structured interview, PD was found
in 13% of an adult urban population (Casey & Tyrer, 1986), while
prevalence rates of up to 34% have been reported for populations of
patients in primary medical care settings. Bateman (1993) claimed a
prevalence rate of about 10% in the general population and 20-30% in
patients attending general practitioners. PDs have been commonly found
in association with anxiety states, depressive disorders, substance abuse,
self-harm episodes and criminal conduct. It was reported that, in England
and Wales, 7.6% of psychiatric admissions had a PD diagnosis (Depart-
ment of Health and Social Security, 1985), which is likely to be an
underestimate of the actual prevalence. Other studies have found high rates
(up to 50%) in psychiatric inpatient populations, as well as in outpatient
samples (20-40%).
Weissman (1993), in a recent review, claimed a lifetime rate for a DSM-
III-R PD of 10-13%, and of 2-3% for antisocial PD. For the latter, three
different community studies have produced similar rates, and this diagnosis
is much more common in males and in younger adults. A decreasing
prevalence rate with increasing age has also been found for some other PDs,
such as borderline PD, and there appears to be a relatively high prevalence
of PDs in urban samples.
Causation
The range of causal variables

Two main classes of causal factors in relation to PDs have been identified,
i.e. genetic and socio-cultural (Walton, 1973). But in addition to genetic
endowment, other biological influences may be important, such as diet,
perinatal trauma and cerebral viral infections. Evidence has been accumu-
lating to indicate the marked influence of genetic factors on both personal-
ity and PD, and the effects of such influences may involve the expression of
some genetic effects at different stages of life, and several kinds of gene-
environment interactions (Kendler & Eaves, 1986). The latter include the
effects of some consistent patterns of behaviour on the shaping of social
environments.
Various models have been proposed for the effects of environment on
personality and PD development, which have led to hypotheses that
adverse situations at certain stages in development may have specific effects
on the development of the adult personality; for example psychoanalytic
theories have postulated three main developmental phases that can be
adversely affected by events occurring at certain critical times (Freud,
1962). However, there is no clear evidence that links specific events and
situations in childhood to the development of specific adult PD traits.
Whitehorn (1952) has provided another model involving four developmen-
tal phases leading to maturity, i.e. infantile, when the person expects
limitless consideration; childish, when the person is capable of limited
commitment and the main need is for security; early adolescent, when the
main concern is personal significance; and late adolescent, characterized by
loyalty to a cause and to groups of people. Subsequently the mature person
is less stereotyped and more flexible in his/her social roles. This appears to
reflect the experience of many individuals and, despite the lack of evidence,
it seems likely that the timing of various environmental events is of
importance for the development of aspects of PD for many individuals.
While the traditional concept of PD is of a developmental disorder, (i.e.
with features showing some continuity with childhood behaviour), the
latest (10th) revision of the International Classification of Diseases, ICD-10
Causation 29
(World Health Organization, 1992), points out that abnormalities of
personality can be acquired later in life 'following severe or prolonged
stress, extreme environmental deprivation, serious psychiatric disorder, or
brain disease or injury'. But the occurrence of late onset or intermittent PD
has not been adequately addressed in the main classifications of mental
disorders, despite the fact that the current ideas regarding the interaction of
biological and environmental causal factors would predict that patterns of
maladaptive behaviour in some individuals would be restricted to periods
when the person is in an adverse environment. In clinical practice it is
common for individuals to show certain features of PD, such as changeable
mood and episodes of self-harm, on an intermittent basis associated with a
series of life-crises related to persistent problems with social relationships.
The concept of intermittent PD is similar to that of the vulnerable
personality (Gelder et al., 1983).
The genotype
The genetic characteristics of an individual reflect his/her genetic inheri-
tance, together with any mutations, and the genetic substrate that
influences the development of patterns of maladaptive behaviour is likely to
involve many genes whose effects can be additive. Therefore, a search of
markers of such genes would be likely to involve several biological and
neuropsychological variables.
Non-genetic physical factors

The constitution, which is the biological structure of the individual, may be
influenced by a variety of non-genetic physical variables such as brain
infections, traumatic cerebral damage, malnutrition, the effects of low birth
weight and the effects of alcohol on the foetus in utero (Beattie et al., 1983).
Socio-cultural environment
Environmental variables affecting PD, which are mediated by higher
mental activity, include those related to parents (or care givers), living
group, immediate social milieu and wider cultural influences. It has been
noted that, at the age of around 3 months, the human infant begins to
discriminate between care givers and to form attachments with a limited
number of individuals. This process seems to reflect behavioural patterns
seen in higher primates that are necessary for satisfactory social
development.
Constitution-environment interactions
Although the biological constitution of an individual may also reflect
variables other than genetic factors, particular attention has been paid to
genotype-environment interactions, for example, when individuals with
different genetic characteristics respond differently to specific environments
(Bergeman et al, 1988). It appears that there can be a marked effect of a
specified social environment on a minority of subjects who have certain
genetic characteristics.
Methods of investigation
Animal studies
Selective breeding for certain behavioural patterns in animals may be
relevant to the understanding of human behavioural genetics. For example,
strains of rats have been developed that differ in their 'emotional' response
to being placed in a brightly lit area, as measured by the frequency of
defaecation (Wimer & Wimer, 1985), and 'nervous' pointer dogs appear
frightened in the presence of humans, showing immobility (Reese, 1979).
Assessment of biological markers ofPD

Genetically or non-genetically determined constitutional factors may be
associated with biological 'markers' of personality or PD characteristics.
The identification of a biological marker usually involves direct measure-
ment of a biological variable, but can also be considered to include
psychological tests of cognitive functions and other higher mental activity.
The marker may be directly observable (such as abnormalities in smooth
pursuit eye movement in association with schizotypal PD), or be induced by
a drug (i.e. a psychopharmacological 'probe'). The latter can include a
response of a PD characteristic, such as impulsivity, to a therapeutic trial of
medication.
Siever & Davis (1991) have suggested that research into biological
markers of PD characteristics should be based on a model of the following
four dimensions of maladaptation: problems with cognitive/perceptual
organization (shown by some patients with schizoid, schizotypal and
paranoid PDs); impulsivity/aggression (shown by some patients with
borderline and antisocial PDs, and claimed to be related to an hypothesized
reduction in serotonergic neurotransmitter activity); affective instability
(shown by some patients with borderline PD and suggested to be related to
Causation 31
hypothesized abnormalities of cholinergic and catecholaminergic neuro-
transmission); and anxiety/inhibition. It was noted that functioning on
these dimensions can also be associated with some non-PD psychiatric
disorders (for example, abnormal smooth pursuit eye movements have
been associated with schizophrenia), and that such models must recognize
that interactions between dimensions are likely to be important. Particular
attention has been paid to the biological correlates of schizotypal and
borderline PDs.
The comparison of PD assessments in genetically related individuals

The relative importance of genetic and environmental causal variables has
been investigated by evaluating PD in identical and non-identical twins
reared apart or together, in adoptees, and in several members of the same
family. Although many aspects of PD have been considered, the emphasis
has been on subjects with prominent antisocial behaviour.
Twin studies provide the best methods for evaluating the effects of
genetic and environmental influences, in particular the method of compar-
ing monozygous (MZ) and dizygous (DZ) twins who have been separated
from each other and their parents at birth, and raised apart. However, there
can be several methodological difficulties in twin studies, including a bias in
sample selection and inadequate determination of zygosity. Also, separated
twins may have been subject to similar non-genetic perinatal physical
influences due, for instance, to poor living conditions, while twins reared
together may be exposed to relatively similar treatment compared with
twins reared apart, with greater expectations of behavioural similarity.
Results of twin studies on personality characteristics of 'normal' (i.e.
non-clinical) subjects have shown the potential of such methodology for the
study of clinical populations. Also, as it is generally accepted that most PD
characteristics are on a continuum of severity with 'normal' personality
traits, the main findings from non-clinical samples are likely to be of
relevance to the study of PDs.
Several large twin studies have tried to evaluate the proportions of the
variance of 'normal' personality assessments that are related to genetic
influences, to shared family environment and to non-familial environmen-
tal effects. Bouchard et al. (1990) studied 100 twins reared apart and
concluded that MZ twins raised apart were as similar as MZ twins raised
together. A similarfinding,namely, that there is a negligible contribution of
a common family environment to the similarity of most measured traits
between twin pairs, was reported by Tellegen et al. (1988), although there
was some effect of common family environment on 'positive emotionality'
and 'social closeness'. This study involved 217 MZ and 114 DZ reared-
together adult twin pairs, and 44 MZ and 27 DZ reared-apart adult twin
pairs. It was concluded that about 50% of measured differences were
caused by genetic influences, while the remaining 50% of environmentally-
determined variance included measurement error and the influence of
transient 'states', such as depression or anxiety. It was also claimed that this
study provided evidence of both additive and non-additive genetic effects.
In general, studies of normal personality in related individuals have
shown that genetic factors are responsible for between 35% and 60% of the
variance for several traits measured by questionnaire; for example, the
range of estimates for the genetic contribution to the variance of measure-
ments of extraversion is between 54% and 74% (Livesley et at., 1993).
A further twin study, in a volunteer non-clinical sample, examined
genetic and environmental influences on PD (Livesley et al., 1993). Subjects
were 90 MZ and 85 DZ twin pairs who completed a 290-item questionnaire.
The following dimensions of PD were assessed: affective lability, anxious-
ness, callousness, cognitive distortion, compulsivity, conduct problems,
identity problems, insecure attachment, intimacy problems, narcissism,
oppositionality, rejection, restricted expression, self-harm, social avoi-
dance, stimulus seeking, submissiveness, and suspiciousness. Estimates
were made for the proportion of the variance of each scale that was
accounted for by genetic factors, common (shared) environment and non-
shared environment. Additive and non-additive genetic factors were esti-
mated separately, based on comparisons of the correlations between the
scores of MZ and DZ twins. The estimates of genetic influence on the
variance for 12 of the 18 dimensions were 40% or above. The highest values
were 64% for narcissism and 59% for identity problems (i.e. ahedonia,
chronic feelings of emptiness, labile self-concept and pessimism), while the
lowest values for genetic heritability were for conduct problems (0%) and
submissiveness (25%). As in previous studies, the influences of common
environment were generally low or non-existent, with the exception of the
contributions to conduct problems (53%) and submissiveness (28%).
However, marked effects of non-shared environment were generally found.
It was concluded that most PD traits involve the extremes of normal
variation, and that most aspects of PD are causally related to substantial
genetic and non-shared environmental components. But it was pointed out
that the relative importance of various causal factors may be different in
samples from clinical settings with a diagnosis of PD. Nevertheless, the
demonstration that PD traits vary in respect of the relative proportions of
Causation 33
genetic and environmental effects has implications for treatment pro-
grammes; for instance, a strong genetic predisposition to a maladaptive
trait may limit the potential for psychological treatments.
Methods to evaluate interactions between causal variables

Evidence of genetic-environmental interactions for personality traits in a
non-patient sample has been provided by Bergeman et al (1988). In their
study of 99 pairs of MZ twins reared apart, perceptions of early rearing
environment appeared to have interacted with genetic factors to influence
personality as assessed during later adult life.
Identification of associations between possible causal variables and PD

features in groups of subjects
Many studies have investigated the correlations (or categorical associa-
tions) between hypothesized causal variables (such as abuse in childhood)
and measurement of PD characteristics. However, an association does not,
of course, necessarily reflect a causal relationship.
'Cluster A' PDs (paranoid, schizoid and schizotypal PDs) - evidence of

causation
Evidence from studies of genetically related individuals

A study of schizotypal PD in subjects with a MZ twin showed a concor-
dance rate of 33% (with an absence of borderline PD in co-twins),
compared with 4% in DZ pairs (Torgerson, 1984). Also, an increased
prevalence of schizotypal PD and other 'cluster A' PD psychopathology
has been found in first degree relatives of patients with schizophrenia
(Kendler et al., 1981a; Baron et al., 1985; Parnas et al., 1993). Only one
study has failed to demonstrate such a relationship between schizotypal PD
and schizophrenia (Coryell & Zimmerman, 1989). Familial aggregation for
schizotypal PD has also been demonstrated (Livesley et al., 1993). A recent
report of a joint US and Danish project involved 207 offspring of
schizophrenic mothers and 104 controls folio wed-up for between 15-42
years (Kendler et al., 1993a, b) and found that, as well as increased
prevalence in the at-risk offspring of schizophrenia and other psychoses,
there was an excess of schizotypal PD (18.8% versus 5%) and paranoid PD
(2.6% versus 0%).
Evidence from studies of biological markers

Various biological associations with schizotypal and other 'cluster A' PD
features have been described. In relation to cognition, attention and
information processing (which is believed to reflect dopaminergic neuro-
transmission), schizotypal PD features have been associated with impaired
performance of tests of eye movements, of'continuous performance' tasks,
and of 'backward masking', and with abnormalities involving 'event-
related EEG potentials' (Weston & Siever, 1993). An example of an eye
movement disorder is shown by the presence of small 'jumps' when a
smoothly-moving target is followed. A 'continuous performance' task may
involve monitoring a changing series of letters and making a response when
a particular sequence of two letters occurs. 'Backward masking' involves
the presentation of two visual stimuli in quick succession; the first, or
'target', stimulus can be more easily identified as the time between the
stimuli is increased, and the minimum time associated with recognition of
the target stimulus is increased in association with schizophrenia and
schizotypal PD. This reflects the speed of information processing. 'Event-
related EEG potentials', which are also related to information processing,
are EEG recordings of brain activity after, for example, an auditory
stimulus. Impaired eye-tracking accuracy has also been found to be
associated with schizophrenia (Siever et al., 1990a).
Other associations of schizotypal and related PD characteristics have
been with reduced platelet monoamine oxidase (MAO), raised homovanil-
lic acid in CSF and plasma, and a characteristic response to an ampheta-
mine 'challenge', in which subjects with co-occurring schizotypal and
borderline PDs became transiently worse with amphetamine administ-
ration (shown by abnormal cognitions and reduced feelings of well-being),
while patients with only a borderline PD diagnosis improved (Schulz et al.,
1988). Also, changes in ventricle-to-brain ratio on computerized tomogra-
phy (CT) scans have been related to schizotypal children of patients with
schizophrenia (Schulsinger et al., 1984), and treatment with dopamine
receptor antagonist drugs, which are effective in schizophrenia, appear to
reduce some of the abnormal cognitions in schizotypal PD.
Freund et al. (1992) have reported that, among women with normal IQ
who are carriers of the fragile X chromosome abnormality, the degree of
fragility (the percentage of cells exhibiting the fragile-X chromosome
abnormality) is associated with a range of social and psychological
disability, including schizotypal PD features.
Causation 35
Summary
Studies of biological correlates of schizotypal PD have indicated geneti-
cally-determined abnormalities in dopaminergic neurotransmission and
information processing. Schizotypal PD and schizophrenia appear to be
related disorders, and this is consistent with the finding that some patients
with schizotypal PD may be helped by low doses of 'neuroleptic'
medication.
'Cluster Bf PDs (antisocial, borderline, histrionic and narcissistic PDs) -

evidence of causation
Evidence from studies of genetically related individuals

Twin studies of antisocial behaviour have been reviewed by Blackburn
(1993) and McGuffin & Thapar (1992). Many reports have shown clear
differences between a greater concordance for criminal behaviour between
MZ twin pairs than between DZ twin pairs for adults, although this
difference has not been found in relation to juvenile delinquency. McGuffin
& Gottesman (1984) summarized data from seven studies of adult crimina-
lity and found mean MZ:DZ concordance of 5 1 % : 22%, although it was
noted that interpretation must take into account the different methodolo-
gies used.
Scandinavian studies have taken advantage of national registers to
obtain unselected twin samples; Christiansen (1974, 1977, and re-analysed
by Cloninger et al, 1978) studied 3586 twin pairs and reported MZ:DZ
concordance for criminality in males of 35%: 13% and in females of
2 1 % : 8%, while Dalgaard & Kringlen (1976) found a smaller (and non-
significant) difference between MZ:DZ concordance rates of 26%: 15%, in
31 MZ pairs and 54 DZ pairs.
Therefore, twin studies have shown a definite genetic contribution to
adult criminality, but this could not be demonstrated in relation to juvenile
crime. This suggests that criminality is, in general, the result of additive
combinations of genetic and (substantial) environmental factors. Also,
female criminals appear to have, on average, a higher genetic predisposition
than male criminals (Cloninger et ah, 1978).
Adoption studies in relation to criminality have had various designs,
such as identification of criminal parents and comparing their children with
those of non-criminal controls, or comparing criminal and non-criminal
adoptees in respect of criminality in biological and adoptive parents. The

former approach was used by Crowe (1972) in a study of the children of 52
imprisoned female offenders and 52 controls: eight out of the 'at risk'
children had arrest records compared with two of the controls. Also, a
history of traffic offences in the 'at risk' and control populations was 19:8,
and the presence of antisocial PD 6:0. Cadoret (1978) reported similar
findings.
Hutchings & Mednick (1975) compared criminal and non-criminal
(control) adoptees in their study of 143 subjects and 143 controls. Of the
males, 36.2% had become criminal when both biological and adoptive
fathers were criminal, 21.4% when the biological father was criminal,
11.5% when the adoptive father was criminal and 10.5% when neither
father was criminal. Blackburn (1993) has reviewed how the presence or
absence of alcoholism, and pre-adoption factors such as perinatal compli-
cations due to poor living conditions, may influence the interpretation of
such results.
A further study of 14427 adoptees included 981 males and 212 females
with a criminal record (Mednick, Gabrielli & Hutchings, 1984). When at
least one of each pair of biological parents was criminal, 20.0% of sons
became criminal compared with 14.7% when an adoptive parent was
criminal and 13.5% when no parent was criminal. Therefore, there was a
significant effect of a criminal biological parent but not of a criminal
adoptive parent, although this effect was related to property offences rather
than violent crime.
Cadoret et al. (1990), in their study of 286 adult male adoptees, have
claimed that their findings suggest that a genetic-environment interaction
can be important in the development of antisocial PD. Both alcohol
problems and criminality/delinquency in biological parents were associated
with increased incidence of antisocial PD in adoptees. Also, the 44 adoptees
with antisocial PD were more likely to have alcohol problems. Alcohol
problems in a biological parent predicted the same disorder in the adoptee,
as did criminality /delinquency, while adverse environmental factors in the
adoptive home, namely an alcohol problem, antisocial behaviour or low
socioeconomic status, were also shown to increase the risk of antisocial PD
in the adoptees, but only when the adoptee had a biological parent with
criminality/delinquency.
It appears that while genetic factors are relevant for both antisocial PD
and alcoholism, these are separate to some degree (Baron et ah, 1985; Dahl,
1993), and Berelowitz & Tarnopolsky (1993) concluded that family studies
confirm 'the separateness of borderline and schizotypal personalities, and
Causation 37
of borderline personality and major depressive illness'. But there is evidence

of a genetic relationship between antisocial PD in men and somatization
disorder in women (McGuffin & Thapar, 1992).

The biological (including neuropsychological) correlates of antisocial PD
and criminality have been reviewed by Kilzieh & Cloninger (1993) and
Blackburn (1993). There have been claims of frontal and left hemispherical
neuropsychological dysfunction in offender groups, including visual-
perceptual deficits and impairment of language-related skills (Miller, 1988;
Mungas, 1988), and differences from control groups have been reported in
relation to EEG findings (including evoked potentials and event-related
potentials, i.e. scalp EEG responses to stimuli or events); testosterone
levels; autonomic function (including skin conductance levels, skin conduc-
tance responses and heart rate) (Raine, Venables & Williams, 1990); blood
platelet MAO (monoamine oxidase) activity (Klinteberg et ah, 1987;
Lykouras, Markianos & Moussas, 1989; Aim et ah, 1994); CSF (cerebro-
spinal fluid) catecholamine metabolites (Virkkunen et aL, 1987; Gardner,
Lucas & Cowdry, 1990); pre-menstrual tension; minor physical abnormali-
ties (Kandel et aL, 1989) and abnormal glucose metabolism. A consistent
finding has been that offender populations show electrodermal hypores-
ponsiveness when anticipating or experiencing aversive stimulation (Black-
burn, 1993).
The reported associations between antisocial behaviour, structural brain
damage, epilepsy and a proposed 'dyscontrol syndrome' have been incon-
sistent. A dyscontrol syndrome has been considered to involve explosive
outbursts of emotion (usually anger, with minimal triggering events), but
with a history of satisfactory social behaviour between episodes. It has been
suggested that such a syndrome may sometimes be associated with focal
abnormalities of brain structure or function in the temporal lobes or other
parts of the limbic system. In relation to epilepsy, the question can
sometimes arise as to whether a crime can be committed without awareness,
and rare examples of such automatism have been reported. The evidence of
an association between temporal lobe epilepsy and non-seizure-related
violence is unclear (Blackburn, 1993).
Studies on chromosome abnormalities, such as Kleinefelter's syndrome,
suggest that such disorders can sometimes make a non-specific contribu-
tion to criminality (as can childhood hyperactivity), and may be mediated
by low intelligence, but the explanations for reports of various associations
of criminality with physical appearance and relatively muscular body build
are unclear (Kumar et al., 1989). In respect of facial appearance, the

reaction of others to unattractiveness may sometimes lead to a self-fulfilling
prophecy.
There have been many studies of the biological associations of borderline
PD (Steiner, Links & Korzekwa, 1988; Lahmeyer et al, 1989; Berelowitz &
Tarnopolsky, 1993), although it must be remembered that many subjects
with this diagnosis have features of other PDs, particularly the other
members of'cluster B\ Studies have involved the dexamethasone suppres-
sion test, the thyrotropic releasing hormone test, sleep studies, electro-
encephalography (EEG), platelet MAO (Schalling et al., 1987), platelet
adrenergic receptor binding (Southwick et al., 1990; Shekim, Bylund &
Frankel, 1990), catecholamine metabolite studies (Linnoila & Virkkunen,
1992), and 'challenge studies', such as the prolactin response to fenflura-
mine (Halperin et al., 1994), the mood response to methylphenidate (Lucas
et al., 1987) or to oral m-chlorophenylpiperazine (Hollander et al., 1994),
the growth hormone (GH) response to clonidine, and the effects of
arecholine on REM (rapid eye movement) sleep latency (Weston & Siever,
1993). Also, antipsychotic drugs, some antidepressant drugs and mood-
stabilizing agents such as lithium carbonate and carbamazepine have been
found to be of benefit to some patients with affective instability in the
setting of PD.
It has been claimed that impulsivity is the most consistent abnormality in
'cluster B' PDs, and may be associated with abnormalities of serotonergic
neurotransmitter function, shown by low serotonergic metabolites in the
CSF and reduced prolactin response to fenfluramine (Coccaro et al., 1989),
together with abnormalities of noradrenergic neurotransmission, shown by
an increased growth hormone response to clonidine. Affective instability
may also involve abnormalities in acetylcholine neurotransmission (Wes-
ton & Siever, 1993). Linnoila & Virkkunen (1992), in their study of an
offender group, have suggested that a 'low serotonin syndrome', associated
with low 5-hydroxyindoleacetic acid (5HIAA) in the CSF, is related to
impulsive, externally-directed aggressive behaviour in some violent
offenders, which often occurs under the influence of alcohol, while such
CSF changes have also been reported in relation to violent suicide attempts
in patients with PD and unipolar depression (Gardner et al., 1990). In the
former study, the offender group also showed associations with mild
hypoglycaemia, a history of early-onset alcohol and substance abuse, and
disturbance in diurnal activity rhythm.
Neuropsychological testing of patients with borderline PD has produced
claims of significant impairment, compared with controls, on memory tests
Causation 39
of uncued recall of complex recently-learned material (O'Leary et aL, 1991).
Sleep EEG studies have been claimed to distinguish patients with
borderline PD and a co-occurring depressive disorder from those with
borderline PD and other types of co-occurring psychiatric disorder (Lah-
meyeretaL, 1989).
Evidence from associations within groups of subjects with PD

A number of studies have examined associations and correlations between
possible causal variables and PD, in particular environmental variables in
relation to borderline and antisocial PDs. Risk of criminality has been
shown to be associated with a childhood history of prolonged institutional
care, placement out of the family, multiple placements and low socioecono-
mic status (McGuffin & Thapar, 1992), while both antisocial and border-
line PD features have been associated with a history of childhood sexual
abuse, physical and emotional abuse, multiple forms of severe and repeated
abuse, separation from or death of parents, neglect (i.e. lack of supervision,
lack of care and parental unavailability), harsh or rejecting parenting,
losses in childhood, maternal overinvolvement combined with inconsistent
responses, and witnessing domestic violence or other antisocial behaviour
(Bezirganian, Cohen & Brook, 1993; Crowell et aL, 1993). Also, a history of
sexual abuse has been linked with relatively lethal incidents of self-harm
(Van der Kolk, Perry & Herman, 1991; Wagner & Linehan, 1994).
Inconsistent parenting can involve a failure to notice and respond posi-
tively to desirable behaviour, together with an inability or unwillingness to
address and confront antisocial behaviour. In a study of 776 adolescents, it
was found that parental inconsistency was associated with subsequent
development of borderline PD but only if it was combined with maternal
overinvolvement (Bezirganian et aL, 1993).
Research has also indicated that consistent and appropriate discipline
can be a protective variable, as can a high intelligence (Kandel et aL, 1988).
Summary
Studies of related individuals have clearly indicated that genetic factors can
predispose an individual to criminal activity. Studies of biological variables
have reported that groups of subjects with antisocial PD tend to show
hyporesponsiveness when anticipating or experiencing aversive stimula-
tion, while 'challenge' studies, such as the prolactin response to fenflura-
mine, have shown abnormalities related to borderline PD. Both borderline
and antisocial PD have been associated with a history of various aspects of
adversity in childhood, in particular repeated abuse involving sexual,
physical and emotional aspects. Consistent and appropriate disciplinary

practices in the upbringing of children should be encouraged, and may
protect against the later development of some aspects of PD in vulnerable
individuals. Impulsivity may be helped by serotonergic enhancing drugs,
such as fluoxetine, while affective instability may benefit from drugs
affecting cholinergic and catecholaminergic neurotransmitter function,
such as neuroleptics, lithium and monoamine oxidase inhibitors.
'Cluster C PDs (avoidant, dependent, obsessive-compulsive PDs and, in

DSM-III'R but not DSM-IV, passive-aggressive PD) - evidence of
causation
Evidence from studies of related individuals

Evidence of causation is relatively sparse for 'cluster C PDs, but it has been
claimed that avoidant and dependent PDs cluster in families (McGuffin &
Thapar, 1992). Genetic evidence for obsessive-compulsive PD is conflict-
ing, but a genetically-determined vulnerability may be linked with Tour-
ette's syndrome (Dahl, 1993).

Data on biological variables related to the 'anxious' 'cluster C PDs is
limited. However, a few studies have investigated the effects of growth
hormone response to clonidine, of lactate infusion, and of inhalation of 7%
CO2 (Weston & Siever, 1993). Also, Joyce, Mulder & Cloninger (1994^, b)
found that temperament, especially dependence and extravagance, was
associated with hypercortisolaemia in a group of depressed patients.
Conclusion
A considerable volume of research has identified a range of genetic and
environmental causal factors, and their interactions, which must be con-
sidered in the development of all PDs. Further studies of biological markers
of PD characteristics may lead to improvements in the targeting of both
pharmacological and psychological interventions.
Treatability
Unfortunately, 'personality disorder' has often been used as a negative and
perjorative term to imply that the individual does not deserve help and that
the presenting problems are not the province of medical practice. Also, it
Treatability 41
has been commonly believed that there are no effective interventions; this
can provide an excuse to avoid those patients whose behaviour is unplea-
sant and difficult to manage. Some patients with PD become poorly
tolerated by health care staff as they frequently demand help but reject what
is offered. This may be accompanied by complaints, anger, threats and
violence.
However, patients with PD can benefit from a variety of interventions,
even though the long-term persistence of these disorders usually prevents
the clinician from receiving the gratifying feedback of a short-term
dramatic cure. However, as the presenting problems often result from an
interaction between the PD and the environment, it is often possible to
arrange or encourage environmental changes that can lead to a rapid
resolution of a crisis. In the longer-term, it is possible for ingrained aspects
of PD to become modified, although, in general, any significant changes are
gradual, while pronounced traits will never completely disappear or be
replaced. Nevertheless, even relatively modest changes (for example,
learning to avoid some situations that evoke an outburst of temper) can
have significantly beneficial effects.
For many individuals with severe PD, the most helpful strategy is to
provide a long-term contact with one or more professionals in the health
care and allied services, with back-up measures (e.g. intensive counselling, a
short admission to psychiatric hospital, social work intervention, or day
care attendance) in times of crisis. Although such a supportive strategy
appears deceptively simple, it is often very difficult for the health-care
worker to maintain the motivation to continue regular (even if infrequent)
contact with a person who is difficult to manage and may be unlikeable. The
nature of these contacts, when the conversation may often be very
superficial and the clinician may not be using any specific techniques of
intervention, may not appear to justify years of professional training. It is
tempting not to give another appointment and, of course, this may be quite
appropriate, as not everybody with PD needs to be seen over long periods
or indefinitely. But for those individuals who are constantly presenting to
the psychiatric and other medical services, longer-term support and crisis
intervention can be justified, not only because the individual may be helped
but also because this may minimize demands on medical services. The
specific tasks of the clinician are firstly, to decide on an appropriate format
and duration of contact with the psychiatric (or other) services; secondly, to
tolerate a lack of response or cooperation; and, thirdly, to override his/her
desire to discharge the individual if the contacts evoke anger and frust-
ration in the clinician, when this is not in the patient's interest.
In addition to this supportive approach, there is evidence that more
specific interventions can sometimes be of benefit. These include various

forms of short-term, intermittent and long-term psychotherapy, the provi-
sion of various forms of social and community support, cognitive therapy,
different types of group psychotherapy, behaviour therapy and drug
treatments, in particular low dose neuroleptics (usually combined with
psychological management) for borderline PD.
Specified personality disorders: clinical
features
J.H. DOWSON
Validity of personality disorder syndromes

For a syndrome to have validity it must be demonstrated that its features
are associated with other variables; descriptive validity involves an associa-
tion with definitions of PD, while concurrent and predictive validity relate
PD assessments to measurements of other variables and outcome,
respectively.
A PD diagnosis based on a categorical classification of PD is 'inherently
imperfect because we can only provide an approximate match to an
hypothetical construct' (Widiger & Frances, 1985a), and the main polythe-
tic diagnostic systems give rise to PD syndromes that have a variable degree
of resemblance to the 'ideal' or 'prototype' disorders. Also, there can be
overlap between diagnoses if similar or identical features occur in more
than one syndrome, and co-occurrence of several PDs may be present in the
same individual. Furthermore, when two or more PD syndromes co-occur,
it cannot be assumed that they are independent, as interactions will occur.
The DSM-III-R and DSM-IV claim that PD syndromes can be grouped
into 'clusters', which implies that if one PD member of the cluster is present,
features of the associated PDs are more likely to co-occur than features of
PDs in other clusters. The first cluster includes paranoid, schizoid and
schizotypal PDs, and people with these syndromes often appear odd or
eccentric. Antisocial, borderline, histrionic and narcissistic PDs form a
second cluster, leading to dramatic, emotional or erratic behaviour, while
the third cluster consists of avoidant, dependent, obsessive-compulsive and
passive-aggressive PDs (the latter is omitted in DSM-IV). Individuals with
the latter disorders often appear anxious or fearful.
This chapter will summarize the clinical features of categories of PD
based on various sources, but it must be remembered that all classifications
are provisional, and the major classifications of mental disorders are
subject to regular revisions. Although the many problems of classification
43
44 Specified personality disorders: clinical features
and assessment of PD might lead a pessimist to conclude that PD diagnoses

are unlikely to be useful, many have persisted in clinical practice.
A particular definition of a PD syndrome can be evaluated in several
ways, for example, by using the criteria to divide a group of patients into
cases (with the disorder) and non-cases; also, each individual diagnostic
criterion can be evaluated in relation to frequency of positive ratings in the
cases (sensitivity), rarity of positive ratings in non-cases (specificity) and
association of positive ratings with a diagnosis (positive predictive power).
Other ways in which the properties of a PD can be investigated include a
comparison of the defined syndrome with the descriptions in the literature
(i.e. descriptive validity), and examination of concurrent validity, reliability
(i.e. inter-rater agreement and temporal stability of diagnosis), the degree of
overlap with other PD syndromes and the frequency of co-occurrence with
other disorders. It is also important to consider if the application of the
diagnosis is subject to prejudice (such as racial or sexual) or if its use (such as
for sadistic or antisocial PD) could lead to unfortunate legal consequences.
Finally, the value of a PD diagnosis is partly reflected by its level of
acceptance among those who are professionally concerned with personality
disordered individuals.
Many of the syndromes to be described are useful to clinicians as a
convenient summary of the main PD features for individual patients.
Although most individuals with PD fulfil criteria for more than one PD,
many such individuals show predominant features (i.e. those judged to
cause the most problems) that relate to just one PD diagnosis. In psychiatric
practice, the features of borderline PD are important as a frequent origin of
many crises and of 'difficult to manage' patients, while in primary medical
care, other PDs, such as histrionic and obsessive-compulsive PDs, are
relatively more important as causes of common complaints such as anxiety
and some depressive symptoms. In forensic practice, antisocial PD is of
particular relevance.
Paranoid personality disorder
Main features
These include excessive sensitivity to rebuff, insecurity, suspiciousness and
mistrust of others.
Clinical origins
Kretchmer (1918) described a 'sensitive' personality, which involves excess-
ive suspiciousness and exaggerated reactions to setbacks, and these features
Paranoid personality disorder 45
have been incorporated into all the versions of the DSM classification. But
despite general agreement among clinicians that such a syndrome does
exist, it is relatively uncommon in clinical practice and has attracted little
research. However, there is evidence from family and adoption studies that,
together with schizoid and schizotypal PD, paranoid PD has a genetic link
with schizophrenia (Kendler et al, 1984). For example, increased risks of
schizophrenia, paranoid PD and schizotypal PD were found in relatives of
a sample selected on the basis of the presence of schizotypal PD (Siever et
al., 19906).
Main definitions
The DSM-III-R considers that the essential feature of paranoid PD is a
'pervasive and unwarranted tendency ... to interpret the actions of people
as deliberately demeaning or threatening' and four of the following are
required for the diagnosis:
(1) expects, without sufficient basis, to be exploited or harmed by others, (2)
questions, without justification, the loyalty or trustworthiness of friends or associ-
ates, (3) reads hidden demeaning or threatening meanings into benign remarks or
events, (4) bears grudges or is unforgiving of insults or slights, (5) is reluctant to
confide in others because of unwarranted fear that the information will be used
against him or her, (6) is easily slighted and quick to react with anger or to
counterattack, (7) questions, without justification, fidelity of spouse or sexual
partner.
The DSM-IV contains minor modifications; for example, the first
criterion becomes: 'suspects, without sufficient basis, that others are
exploiting, harming, or deceiving him or her'.
The ICD-10 criteria show considerable overlap with the above, but also
include the following: a combative and tenacious sense of personal rights
out of keeping with the actual situation; a tendency to experience excessive
self-importance, manifest in a persistent self-referential attitude; and
preoccupation with unsubstantiated 'conspiratorial' explanations of events
both immediate to the patient and in the world at large.
Clinical features
Paranoid PD is characterized by a marked distrust of others and sensitivity
to rebuff. These features appear to relate to an underlying insecurity and are
accompanied by a watchful alertness.
Such individuals are suspicious, tense, emotionally cold and rigid in their
thinking. There is often a fear of losing independence, a resentment of those
in authority and a fear of intimate relationships. They see themselves as

victims of mistreatment and abuse, although this is associated with a
disdain for weakness or disability in others. There is an excessive sense of
personal rights and self-importance. Jealousy may be marked and directed
towards those with power and influence as well as to colleagues and sexual
partners. It may be felt that others have been responsible for preventing the
achievement of the person's potential.
These attitudes lead to a variety of interactive behaviours. Such persons
seem touchy, argumentative and hostile with a tendency to counterattack at
minimal provocation; they are guarded and secretive, being reluctant to
disclose personal information and having an excessive concern about
confidentiality; they may appear devious, stubborn and unwilling to
compromise; they lack humour, capacity for enjoyment and tender feelings;
and they have difficulty in expressing doubts and insecurities. Anger may be
provoked by a positive regard as this is seen as only conditional, while a
grudge is accompanied by vindictiveness. Not surprisingly, these indivi-
duals tend to generate unease, fear and hostility in others. They are often
not fully aware of the effects of their behaviour. Faced with a new situation,
the person may search for confirmation of his/her expectations and become
convinced of hidden motives on the basis of inadequate evidence, without
being able to appreciate the total context of the situation. Criticism is very
difficult to accept and there is a low tolerance threshold with a tendency to
try and avoid any blame. Because of the combination of insecurity and self-
importance, groups of people are avoided, unless the person is in a
dominant position. Under times of stress they may experience transient
'psychosis-like' phenomena, such as ideas of reference, which may involve a
feeling that others are taking special notice of him/her.
Presentation to medical services

Individuals with paranoid PD seldom seek help for the features of their
syndrome; however this disorder may be the origin of secondary symptoms
such as anxiety, some depressive symptoms, or substance abuse. These may
be in the context of occupational problems such as conflicts with those in
authority, or relationship difficulties within the family. As these individuals
tend to be secretive and guarded, evidence of paranoid PD features may
only emerge gradually within a series of interviews or treatment sessions.
It appears that paranoid PD can, more specifically, lead to the develop-
ment of other psychiatric syndromes, although these are 'final-common-
pathways' that may also reflect other disorders such as schizophrenia.
Some individuals with paranoid PD, the so-called 'paranoid litigants',

become heavily involved with legal activities, tribunals or complaints
procedures. It is useful to be able to predict the likelihood of such
behaviour, so that particular care can be taken to provide detailed
explanations to the patient and to keep adequate written records. Others
with paranoid PD may develop a delusional preoccupation with a part of
the body, usually a facial feature, which is known as 'dysmorphophobia'.
For example, a chance remark about the size of a patient's nose at the age of
14 led to a discrete delusional preoccupation that his nose was severely
deformed and attracted ridicule. Such people may present to surgeons with
requests for cosmetic operations, but these are not indicated unless there is
a clear and widely accepted need for corrective surgery. Finally, the
syndrome of'morbid jealousy', which involves a delusional preoccupation
with a partner's supposed infidelity, may also arise in the context of
paranoid PD, although this can also occur in association with schizophre-
nia or alcohol abuse.
A ssociated features
Reference has been made to the evidence that suggests that individuals with
paranoid PD can share some genetic factors with those showing schizotypal
PD, schizoid PD or schizophrenia. The DSM-III-R has claimed that
paranoid PD is more frequently diagnosed in men.
Current status
Despite the relative lack of empirical research, the category of paranoid PD
has a long clinical tradition. The phenomena associated with this diagnosis
can undoubtedly be of clinical significance in patient groups. The genetic
relationships between paranoid PD, the other PDs in DSM-III-R's cluster
A (schizotypal and schizoid PDs) and various types of schizophrenia,
require further investigation.
Schizoid (and schizotypal) personality disorder
Main features
The main features of schizoid and schizotypal PDs will be considered
together; these include social withdrawal, lack of emotional rapport,
sensitivity, a preoccupation with an internal world of cognitive processes
involving fantasy, and odd thoughts, appearance and behaviour. There is

evidence that a group of individuals with such features is associated with a
genetic predisposition that is related to genetic causes of schizophrenia.
Clinical origins
Although Bleuler (1908) drew attention to that part of any personality that
involves the direction of the person's attention to the inner mental life and
away from the external world, the concept of schizoid PD originated from
observations that social withdrawal and eccentricity appeared to have an
increased prevalence among relatives of patients with schizophrenia and
among the patients' premorbid personalities (Widiger et ah, 1988; Vaglum
& Vaglum, 1989). The vivid internal mental world that can accompany
social withdrawal was recognized by Kretchmer (1925), who described
persons who loved books, were impractical and lazy, yet could act with
passionate energy in certain areas. Despite social withdrawal, deep friend-
ships could sometimes be developed with a select few.
The revisions of the third edition of the DSM classification in 1980 were
of critical importance to the more recent concepts of the schizophrenia-
related PDs, and, prior to 1980, the DSM subdivided a schizoid PD
category into schizoid, avoidant and schizotypal PDs; schizoid PD denoted
social withdrawal in the context of an apparent indifference to social
relationships, together with a reduced range of emotional experience and
expression, while, in contrast, individuals with avoidant PD were thought
to have the desire to relate socially with more capacity for social relation-
ships. In avoidant PD there was also a pervasive social discomfort due to
fear of humiliation and disapproval. The third of the pre-1980 subcategor-
ies, schizotypal PD, was reserved for those features that appeared, on the
evidence of the Danish adoption studies of schizophrenia, to be the most
specific indicators of genetic linkage with schizophrenia, namely odd or
eccentric thoughts and behaviour (Kendler et ah, 1984; Modestin, Foglia &
Toffler, 1989). However, detailed evaluation of the range of PD features in
relatives of patients with schizophrenia has also identified an increased
prevalance of emotional detachment, unsociability and suspiciousness
(Kendler, 1985), which indicates that features in the DSM-IIFs definition of
schizoid PD were also related to schizophrenia. Another controversial and
confusing aspect in the classifications of these PDs has been the claim that
'schizophrenia-like' features can also be found in borderline PD.
Other syndromes, which are described as 'developmental' as they become
apparent in childhood, appear to be linked to schizoid (and schizotypal) PD
in adults; these are schizoid PD of childhood (Wolff, 1991), autism, and
Asperger's syndrome (Tantam, 1988). (The latter incorporates schizoid PD
of childhood in the ICD-10.)
Schizoid PD of childhood occurs more in boys than girls in a ratio of 4:1.
While autism is usually evident before the age of 3, schizoid PD of
childhood occurs later and, in general, patients are less disabled than those
with either autism or Asperger's syndrome, although the boundaries
between these disorders are not clear. Schizoid PD of childhood involves
solitariness, impaired empathy, emotional detachment, rigidity of mental
set with single-minded pursuit of special interests, increased sensitivity, odd
styles of communication (including overcommunicativeness), odd gaze and
gestures, and unusual fantasies. This is usually in the setting of average or
superior intelligence but there may be specific developmental delays,
especially of language-related skills. Such children present problems to
teachers and parents; they do not conform socially and react to constraints
with outbursts of weeping, rage or aggression. Follow-up studies have
shown that features of this disorder persist to at least early adult life and it
appears that biological causal factors, such as genetic influences, are
implicated. The risk of the subsequent development of schizophrenia may
be increased but is still relatively low at an estimated 1 in 10, but in a follow-
up study (Wolff, 1991), in which the mean age of patients was 27, 75%
fulfilled DSM-III criteria for schizotypal PD and there were significantly
increased prevalences of both schizoid and schizotypal PDs compared with
a control group. Also, the patient group had more impairment related to
social integration, heterosexual relationships and sensitivity, together with
unusual modes of communication, involving odd thoughts, overtalkative-
ness, unguarded communications, abnormal eye contact, abnormal smil-
ing, single-minded pursuits of special interests and poor work adjustment.
There was a greater incidence of attendance at psychiatric services and two
patients had developed schizophrenia. Wolff (1991) has hypothesized that
schizoid PD of childhood may be related to autism, Asperger's syndrome
and schizophrenia, in an association that may involve the sharing of a
genetically based predisposition that requires additional genetic and other
factors for the development of the latter three disorders.
Autism is also more common in boys and is usually apparent before the
age of 3. Most affected individuals have significant learning difficulties and
there is poor communication with a pattern of restricted repetitive behav-
iour. It has been reported that parents of such children show increased
schizoid PD features (Wolff, 1991) and that follow-up indicates a probable

increased incidence of schizoid PD and schizophrenia (Szatmari, Barto-
lucci & Bremner, 1989).
Asperger's syndrome is also apparent in early childhood and more
common in boys but there is usually no widespread delay or retardation in
language or cognitive development; features include poor empathy,
poverty of non-verbal expression, pedantic repetitive and circumstantial
speech that may lack appropriate intonation, an inflexible routine some-
times involving intense interests, clumsy gait and posture, and difficult peer
relationships (Mawson, Grounds & Tantam, 1985). The ICD-10 notes a
tendency for these abnormalities to persist into adolescence and adult life.
Main definitions
DSM-III-R considers that the essential features of schizoid PD are a
pervasive pattern of indifference to social relationships and a restricted
range of emotional experience and expression. At least four of the following
criteria are required:
(1) neither desires nor enjoys close relationships, including being part of a family
(authors' note: the lack of desire for relationships has been questioned), (2) almost
always chooses solitary activities, (3) rarely, if ever, claims to experience strong
emotions, such as anger or joy, (4) indicates little or no desire to have sexual
experiences with another person (age being taken into account), (5) is indifferent to
the praise and criticism of others, (6) has no close friends or confidants (or only one)
other thanfirst-degreerelatives, (7) displays constricted affect, e.g. is aloof, cold,
rarely reciprocates gestures or facial expressions, such as smiles or nods.
The DSM-IV criteria combined criteria 3 and 7 to form the criterion:
'shows emotional coldness, detachment, or flattened affectivity', and has
given an additional criterion: 'takes pleasure in few, if any, activities'.
DSM-III-R classifies schizotypal PD as a separate disorder although
criterion 6 for schizoid PD is included in the schizotypal PD syndrome.
Also, criterion 8 for schizotypal PD: 'inappropriate or constricted affect,
e.g. silly, aloof, rarely reciprocates gestures or facial expressions, such as
smiles or nods', is very similar to criterion 7 for schizoid PD.
The essential features for schizotypal PD are considered to be peculiari-
ties of ideation, appearance and behaviour, as well as deficits in interperso-
nal relations. The DSM-III-R criteria are:
(1) ideas of reference (excluding delusions of reference), (2) excessive social anxiety,
e.g. extreme discomfort in social situations involving unfamiliar people, (3) odd
beliefs or magical thinking, influencing behaviour and inconsistent with subcultural
norms, e.g. superstitiousness, belief in clairvoyance, telepathy, or 'sixth sense',

'others can feel my feelings' (in children and adolescents, bizarre fantasies or
preoccupations), (4) unusual perceptual experiences, e.g. illusions, sensing the
presence of a force or person not actually present (e.g. 'I felt as if my dead mother
were in the room with me'), (5) odd or eccentric behaviour or appearance, e.g.
unkempt, unusual mannerisms, talks to self, (6) (same as criterion 6 for schizoid PD
- see above), (7) odd speech (without loosening of associations or incoherence) e.g.
speech that is impoverished, digressive, vague, or inappropriately abstract, (8)
inappropriate or constricted effect, e.g. silly, aloof, rarely reciprocates gestures or
facial expressions, such as smiles or nods (similar to schizoid PD criterion 7 - see
above), and (9) suspiciousness or paranoid ideation.
It should be noted that criteria 3 and 4 involve cognitive-perceptual
distortions which, although not included in the DSM-III-R criteria for
borderline PD, have been claimed to be found in this disorder. The DSM-
IV criteria are similar but more explanation is given for criterion 2, i.e.
'excessive social anxiety that does not diminish with familiarity and tends to
be associated with paranoid fears rather than negative judgements about
self.
The ICD-10 also separates schizoid PD and schizotypal disorder,
although the latter is classified with schizophrenia and not as a PD. The
criteria for ICD-10 schizoid PD show considerable overlap with those of
DSM-III-R, but also include 'excessive preoccupation with fantasy and
introspection' (an attribute that has often featured in clinical descriptions),
and 'marked insensitivity to prevailing social norms and conventions'.
The ICD-10 criteria for schizotypal disorder show some overlap with
DSM-III-R's schizotypal PD, but the former also specifies 'depersonaliza-
tion, derealization, occasional transient quasi-psychotic episodes with
intense illusions, auditory or other hallucinations and delusion-like ideas,
usually occurring without external provocation, and obsessive ruminations
without inner resistance, often with dysmorphophobic, sexual or aggressive
contents'.
Clinical features
The features of schizoid (and schizotypal) PD result in various degrees of
social withdrawal and eccentricity, but individuals in whom these charac-
teristics are predominant are relatively rare in psychiatric practice,
although Tantam (1988) has described 60 patients from the severe end of
the spectrum. In this group, men outnumbered women 6 to 1, 25% had
criminal convictions, while half had a history of antisocial behaviour.
Nearly half had another psychiatric diagnosis and only a few lived
independently. In contrast, individuals with less severe schizoid (and

schizotypal) PD features are usually capable of steady work and the traits
can sometimes be adaptive, for instance, in academic and other intellectual
pursuits that can be compatible with social detachment and a focus on
isolated mental activity. For the purpose of this chapter, schizoid and
schizotypal PDs have been considered together, despite their separation in
DSM-III-R, DSM-IV and ICD-10, as features from both syndromes
appear to be genetically related to schizophrenia; it has also been suggested
that if sufficient schizotypal features are present, schizoid features are
invariably present (Livesley & Schroeder, 1990).
Individuals can appear self-sufficient, detached, aloof, and self-
absorbed, seeming preoccupied with their thoughts. They are usually
socially withdrawn with little emotional responsiveness, tenderness,
warmth and humour. In extreme cases there can be a callous disregard for
the suffering of others. There is relatively poor empathy, with little apparent
concern for the opinions of others. Interests tend to be solitary, but at times
there can be an emotional investment in specific and narrow activities or
pursuits that may involve a passionate preoccupation, although this may
fluctuate or disappear.
It has long been recognized that a detached unemotional exterior can
hide a rich and complex pattern of mental activity and the following
features of schizoid (and schizotypal) PD have been described: extreme
sensitivity, suspicious vigilance, curiosity about others, rigidity of thinking,
omnipotent and vengeful fantasies, depersonalization and derealization
(i.e. as if self, body or world are unreal) and idiosyncratic moral or political
views with a tendency towards mystical and spiritual interests. There are
different views in the literature as to whether individuals with pronounced
schizoid (and schizotypal) features do not wish for close relationships or are
emotionally needy; but evaluation depends on the definitions and concepts
used, and it is likely that these attributes may vary. For example, an
individual with reduced capacity and desire for close relationships may still
feel the 'need' for positive feelings and concern for others in the context of a
rather immature dependency.
As previously described, schizoid (and schizotypal) PD originated with
an attempt to describe and identify features that seemed to be more
prevalent in the premorbid personalities of individuals who developed
schizophrenia and in their close biological relatives. Three main categories
of such features have been described: transient psychotic episodes, involv-
ing delusions and/or hallucinations; odd appearance and behaviour, which
can affect dress, hairstyle and mannerisms; and cognitive-perceptual distor-
tions. The latter include odd speech that is mildly disjointed, superstitious
ideas, an interest in clairvoyance and telepathy, odd or illogical reasoning,
ideas of reference, chronic suspiciousness and unusual religious beliefs.
Although these features are not specific to the schizoid (and schizotypal)
PD syndrome, as they can be found in the borderline PD syndrome, they
appear to be less prevalent and enduring in the latter (Sternbach et al.9
1992).
While social withdrawal is the main overt feature of schizoid (and
schizotypal) PD, a small number of enduring relationships may be formed
that may involve significant emotional reactivity; also, as has been des-
cribed, passionate interest in certain organizations or pursuits can occur.
Males tend not to be involved in enduring heterosexual relationships,
although females may passively accept courtship and the subsequent
relationship. An affected individual often appears quiet, suspicious, lethar-
gic, uninteresting and unassertive. There may be poor social skills, little eye
contact and a lack of self-disclosure. Speech may be vague, for instance,
involving a lack of clarity in the person's goals.
In severely affected individuals, especially if antisocial PD features are
also present, bizarre crimes may be committed. In men, the lack of sexual
relationships may be accompanied by deviant sexual fantasies, an interest
in voyerism and pornography, excessive masturbation and, at times,
associated criminal behaviour.

Schizoid (and schizotypal) PDs alone, or as the predominant clinical
features, are rare in psychiatric practice, and when such patients do present,
they usually have other PDs and/or other psychiatric disorders such as
substance abuse, a depressive syndrome or anxiety. This means that the
schizoid (and schizotypal) features of those individuals who are identified in
family studies of schizophrenia may be different in many respects from
those who present to general medical or psychiatric services. Schizoid (and
schizotypal) features are often present in individuals who have committed
odd or bizarre crimes in the absence of a mental disorder other than PD.
Bornstein et al (1988) reported ten patients with DSM-III schizotypal PD
and found higher prevalences of drug abuse and major affective disorders
than in a comparison group with other disorders; the patients with
schizotypal PD also had a history of more self-harm and hospitalization.
A ssociatedfeatures
Associations between schizoid (and schizotypal) PD and schizophrenia
have already been noted. There appears to be an increased prevalence of
these PD features in the premorbid personalities of those who subsequently
develop schizophrenia, and despite the heterogeneity of the schizophrenia
category and the use of various definitions of schizophrenia and schizoid
(and schizotypal) PD, such an association has often been reported.
Cutting's review of retrospective studies (1985) found an overall prevalence
of premorbid schizoid (or schizotypal) PD in 26% of patients with
schizophrenia, while Foerster et a/. (1991) also found good evidence of this
association, especially in males. Another reported association is between
schizoid (and schizotypal) PD (and also paranoid PD) and schizophrenia in
genetically-related individuals. This has been found in studies involving
adopted and non-adopted subjects, which have indicated that genetic but
not familial-environmental variables are responsible for the relationship of
schizoid (and schizotypal) PD to schizophrenia (Kendler, 1988). In a large
study of relatives of patients with schizophrenia, and of controls, DSM-III-
R schizotypal PD had a significantly increased prevalence in relatives of
patients with schizophrenia, and there were more modest increased preva-
lences for paranoid, schizoid and avoidant PDs (Kendler et al., \993a,b).
The prevalence rate of schizotypal PD was greater in parents than in
siblings of schizophrenic patients. In another family study, certain features
of schizotypal PD, such as odd speech, inappropriate mood and odd
behaviour, together with excessive social anxiety, were more common in
relatives of patients with schizophrenia than in relatives of those with major
depression (Torgerson et al., 1993). Although Squires-Wheeler et al. (1989)
reported that the rates of schizotypal traits in children of parents with
affective disorders were as high as those found in the children of subjects
with schizophrenia, Pica et al. (1990) found a difference between premorbid
PDs in relation to bipolar affective disorders compared with schizophrenia.
But other studies have suggested that schizotypal PD is increased in
relatives of patients with psychoses other than schizophrenia (Silverman et
al., 1993; Thakere/a/., 1993).
The associations between schizoid (and schizotypal) PD and various
syndromes in childhood (i.e. schizoid PD of childhood, autism and
Asperger's syndrome), have also been described. Also, follow-up of
patients with these childhood disorders into adult life has led to the claim
that some individuals may develop increasingly narcissistic PD features,
which is of interest as some clinical descriptions of schizoid individuals have

also included grandiosity.
The 'schizophrenia-like' features of schizoid (and schizotypal) PD have,
as has been noted, caused some controversy in the major classifications of
psychiatric disorders, as it has been unclear whether some of these should
be considered as relatively specific as markers of a genetic vulnerability that
is shared with schizophrenia. Also, such features have caused problems in
defining borderline PD, as claims that brief 'psychotic' episodes are
important characteristics of borderline PD were not reflected in the criteria
in DSM-III and DSM-III-R (George & Soloff, 1986). But there is evidence
that so-called 'cognitive-perceptual distortions' as well as brief episodes
involving delusions and/or hallucinations, often occur in the setting of a
diagnosis of borderline PD, and that individuals can qualify for DSM-III-
R diagnoses of both schizotypal PD and borderline PD (McGlashan,
1987). However, cognitive-perceptual distortions appear to be more endur-
ing in schizoid (and schizotypal) PD than in borderline PD (Sternbach et
ai, 1992). In a follow-up study of individuals with a mixture of schizoid
(and schizotypal) and borderline PD features, it was found that odd or
illogical thinking, suspiciousness and social isolation (but not the other
features of borderline PD), predicted the development of schizophrenia
(Fenton & McGlashan, 1989).
Current status
Although ICD-10, DSM-III-R and DSM-IV separate schizoid PD from
schizotypal PD (or schizotypal disorder, in ICD-10), the original pre-1980
broad concept of schizoid PD incorporated both categories and there is
evidence that features from both syndromes are associated with schizo-
phrenia. It has been recently suggested that schizoid PD and schizotypal
PD 'may be slightly different variants . . . or different points along a
schizophrenia spectrum' (Widiger et al., 1988).
The controversy and confusion about the separation of certain 'schizo-
phrenia-like' features of schizoid and schizotypal PDs into categories such
as 'transient psychosis' (i.e. involving delusions and hallucinations) and
'cognitive-perceptual distortions' continues. Such features have also been
considered to be part of the syndrome of borderline PD, but may be less
enduring in this disorder. The status of those individuals with both schizoid
(or schizotypal) PD and borderline PD is unclear. It has been suggested
that, for schizoid (and schizotypal) PD, the criteria should be modified to
require that cognitive-perceptual distortions must not be limited to discrete
periods of affective symptomatology, such as depression, anxiety and
anger. Another suggested modification to the schizoid and schizotypal PD
DSM-III-R criterion 'has no close friends or confidants ...', is to add 'due
primarily to lack of desire, pervasive discomfort with others or eccentrici-
ties', which takes into account the claims that lack of desire for relation-
ships is generally found (Siever, Bernstein & Silverman, 1991).
The overlap between schizoid (and schizotypal) PD and the DSM-III-R
category of avoidant PD is another area of uncertainty. It has been stated
that there is no clear boundary between the syndromes as, contrary to some
descriptions, schizoid (and schizotypal) PD can also be associated with
anxiety and discomfort in social situations (Overholser, 1989).
Antisocial personality disorder
Main features
Enduring patterns of antisocial behaviour have been classified in a medical
context by various sets of criteria, identified by terms such as 'psychopathy'
(or psychopath or psychopathic PD), 'sociopathy' (or sociopath or socio-
pathic PD) and 'antisocial' PD. But there has been controversy about the
degree to which specified antisocial behaviours should be part of any set of
defining criteria, as it has been claimed that the most characteristic features
are traits such as lack of empathy and remorse, unreliability, failure to
make loving relationships, failure to learn from adverse experience and
impulsive actions (Cleckley, 1976). Craft (19656) considered that the two
main features are a lack of feeling for others and a liability to act on
impulse.
Despite the heterogeneous nature of the PDs of the individuals identified
by criteria sets based mainly on specified antisocial behaviours, follow-up
studies of children with 'conduct disorder' into adult life have established
the validity of the concept of enduring patterns of antisocial behaviour,
even though most children who are identified as having conduct disorders
do not go on to exhibit clinically or socially significant maladaptive adult
behaviour (Guze, 1976; Robins, 1978).
Clinical origins
Pinel (1809) has been credited with the earliest influentional suggestion that
repeated, aimless antisocial behaviour may be a mental disorder. This led to
the term 'manie sans delire' (mania without delusions) and was the
precursor of Prichard's 'congenital deficiency of the moral sense' or 'moral
insanity' (1835), which provided an hypothesis that a diseased 'moral
faculty' could account for antisocial behaviour. But Koch (1891) believed
that there could be biological causes, describing abnormal behaviour that
was not insanity as 'constitutional psychopathic inferiority'. Walker &
McCabe (1973) have noted that the term 'psychopathic' first appeared in
the mid-19th century German literature to mean 'psychologically
damaged', and encompassed all forms of psychopathology.
In this century, Cleckley has provided an influential clinical description
in his book The Mask of Sanity (5th edition, 1976) in which the following 16
features of psychopathy were identified:
superficial charm and good intelligence; absence of delusions and other signs of
irrational thinking; absence of nervousness or psychoneurotic manifestations;
unreliability; untruthfulness and insincerity; lack of remorse or shame; inadequa-
tely motivated antisocial behaviour; poor judgement and failure to learn by
experience; pathological eccentricity and incapacity for love; general poverty in
major affective relations; specific loss of insight; unresponsiveness in general
interpersonal relations; fantastic and uninviting behaviour with drink and some-
times without; threats of suicide, rarely carried out; sex life impersonal, trivial, and
poorly integrated; and failure to follow any life plan.
In spite of these features, even if pronounced, the person may present

with a convincing 'mask of sanity', i.e. with apparent normality. In
Cleckley's view, the psychopath (or sociopath) has an impaired learning
ability with relative unresponsiveness to social reinforcement, so that these
individuals do not react with appropriate affect (mood) to situations that
would normally produce an emotional response. Although there is aware-
ness of social values, these do not influence motivation, resulting in
irresponsibility, unreliability, poor judgement and a relative inability to
profit from experience.
Various subtypes of antisocial PD have been described; Karpman (1948)
separated 'primary' psychopaths, who were similar to those identified by
Cleckley's criteria, from 'secondary' psychopaths, who were characterized
by the presence of other disorders, particularly anxiety. It was claimed that
a generally low level of anxiety is a characteristic of the primary psychopath
and is associated with a relative lack of effect of social constraints and
punishments for selfish and impulsive acts, as these do not produce
significant anxiety or guilt. The category of secondary psychopathy has also
attracted other labels such as 'neurotic' or 'symptomatic' (Hare, 1970), and
it has been suggested that, for such individuals, the motivations, personal-
ity, response to treatment and outcome is often very different compared to
those with primary or 'true' psychopathy (Blackburn, 1988).
The classification of patterns of antisocial behaviour must also take into
account the prevalence of such behaviour in the population under study or
in certain subcultural groups. West (1983) has pointed out that in a non-
clinical sample of males from a low socioeconomic background in London,
over a third had criminal convictions by the age of 24. But the incidents of
conviction declined after the age of 18, and it appeared that delinquency in
young men may usually be self-limiting. Also, there are delinquent subcul-
tures in which antisocial behaviour (mainly defined in terms of criminal
behaviour) is condoned and rewarded. If antisocial behaviour is perceived
as 'normal' within the subculture, 'antisocial' individuals may still have
satisfactory relationships and show appropriate guilt and remorse in their
social context. Such persons have been called 'subcultural delinquents'
(Hare, 1970).
These ideas have led to definitions of two overlapping categories of
disorder, the first related to a lack of feelings for others with impulsivity,
and the second to specified antisocial behaviour. Sets of defining criteria for
PD syndromes associated with antisocial behaviour have incorporated
varying proportions of items from these two categories. Therefore, for any
criteria set for antisocial PD, in particular when mainly based on specified
antisocial behaviour, an heterogeneous group of individuals is identified in
respect of their personality. As there is a variety of PD psychopathology
that is associated with chronic antisocial behaviour, Blackburn (1988) has
concluded that a concept of antisocial or psychopathic PD that is even
partly based on socially deviant behaviour is unsatisfactory and 'little more
than a moral judgement masquerading as a clinical diagnosis'. However,
this appears to overstate the case, as there is evidence that impulsive,
aggressive and relatively unemotional personality characteristics charac-
terize a subgroup of individuals who present with repeated antisocial acts.
But such individuals will form part of a larger, heterogeneous group defined
by an history of antisocial behaviour, and this group will contain indivi-
duals with a variety of personality structures. There will also be variability
in relation to anxiety, the influence of a delinquent subculture, other co-
occurring mental disorders and other co-occurring PDs involving border-
line, histrionic and narcissistic features.
Despite the problem in the diagnosis of PD on the basis of antisocial
behaviour, in England and Wales the Mental Health Act 1959 defined
'psychopathic disorder' as 'a disorder or disability of mind... which results
in abnormally aggressive or seriously irresponsible conduct on the part of
the patient, and requires or is susceptible to medical treatment'. However,

this issue was addressed in the UK by the Butler Committee (Home Office,
Department of Health and Social Security, 1975), which stated: 'the class of
persons to whom the term "psychopathic disorder" relates is not a single
category identifiable by any medical, biological or psychological criteria'.
The first edition of the DSM in 1952 contained the term 'sociopathic
personality disturbance' but this was a broad category that also included
sexual deviations and addictions. 'Antisocial personality' as defined by
DSM-II in 1968 was, however, based on Cleckley's concept:
this term is reserved for individuals who are basically unsocialized and whose
behaviour pattern brings them repeatedly into conflict with society. They are
incapable of significant loyalty to individuals, groups or social values. They are
grossly selfish, callous, irresponsible, impulsive and unable to feel guilt or to learn
from experience and punishment. Frustration tolerance is low. They tend to blame
others or offer plausible rationalizations for their behaviour. A mere history of
repeated legal or social offences is not sufficient to justify this diagnosis.
The next revision of the DSM, DSM-III (1980), used a similar term, i.e.
'antisocial personality disorder', but this was derived from a broader
concept as the emphasis was on specified criminal behaviour which
captured a wide range of PD and other disorders. In the DSM-III-R (1987)
the balance between specified antisocial behaviour and PD traits was
partially redressed in favour of the latter, but did not include many of the
features derived from clinical descriptions, such as selfishness, egocentri-
city, callousness, manipulativeness and lack of empathy. As a result, DSM-
III-R antisocial PD appeared to be both too broad and too narrow; a wide
range of criminals and antisocial persons were included without the
characteristic PD traits, while some of those with these traits were excluded
if they had not shown examples of the specified antisocial behaviours (Hare,
Hart&Harpur, 1991).
An alternative to the DSM criteria has been 'The Psychopathy Checklist'
(PCL) (Hare, 1980) and the revised version, the PCL-R (Hare, 1991), which
were related to Cleckley's descriptions. This concept of psychopathy is
associated with early onset, long-term maladaptive functioning, and social
dysfunction or disability. Unstable personal relationships and poor occu-
pational record is accompanied by an increased risk of criminal activity.
Features are usually noticeable by middle to late childhood and generally
persist into adult life, although there may be some improvement in social
adaptation after the fourth decade (Hare, McPherson & Forth, 1988).
The PCL-R was developed for use in male forensic populations, and
factor analysis of its 20 items has identified two factors, each involving
associations between subgroups of items. The first, which corresponds to

Cleckley's concept, is associated with egocentricity, lack of remorse and
callousness, as well as with features of narcissistic and histrionic PDs, while
there is a negative association with empathy and anxiety. In contrast, the
second factor is positivity correlated with the DSM-III-R category of
antisocial PD, as both are associated with criminal behaviour. The second
factor is also positively correlated with socioeconomic problems and self-
report measures of antisocial behaviour in the context of an impulsive,
antisocial and unstable lifestyle (Hare et al., 1991). Therefore, the PCL-R
includes items related to both the DSM-III-R antisocial PD syndrome and
the PD traits such as those described by Cleckley. Most of the PCL-R items
require the clinician to make a judgement about the degree of a trait with
the help of a manual, which gives a description and some examples of
associated behaviour. Psychopathy as defined by PCL or PCL-R has been
associated with several aspects of criminality; such individuals are charged
with a greater number and variety of criminal offences compared with other
offender groups, and show a higher incidence of violent and aggressive
crimes, physical violence, verbal abuse, threats and intimidation (Hare &
McPherson, 1984). Murders and serious assaults committed by PCL-R
psychopaths were less likely to occur in the setting of a domestic dispute or
emotional arousal and their victims were more likely to be men and to be
relatively unknown to them. It has been estimated that PCL-R psycho-
pathy is a narrower concept than DSM-III-R's antisocial PD, and that a
majority of PCL-R items are present in about 25-30% of prison popula-
tions, compared with a prevalence for antisocial PD of about 60-70%
(Hare et aL, 1991).
Main definitions
The DSM-III-R criteria for antisocial PD require evidence of conduct
disorder with an onset before the age of 15, as indicated by a history of three
or more of the following: 'truancy, running away, fights, using weapons,
forcing sexual activities on others, physical cruelty to animals, physical
cruelty to people, destruction of others' property, fire-setting, lying,
stealing without confrontation of a victim, and stealing with confrontation
of a victim'. In DSM-IV there are three additional behaviours: 'often
bullies, threatens, or intimidates others; has broken into someone else's
house, building or car; and often stays out at night despite parental
prohibitions, beginning before age 13 years'. Also, a pattern of irrespon-
sible and antisocial behaviour since the age of 15 is needed, as indicated by

at least four of the following ten criteria:
unable to sustain consistent work behaviour; failure to conform to social norms
with respect to lawful behaviour; is irritable and aggressive; repeatedly fails to
honourfinancialobligations; fails to plan ahead, or is impulsive; has no regard for
the truth; is reckless regarding his or her own or others' personal safety; if a parent
or guardian, lacks ability to function as a responsible parent; has never sustained a
monogamous relationship for more than one year; and lacks remorse (feels justified
in having hurt, mistreated or stolen from another).
Various examples of some of these criteria are specified, the current age
must be at least 18, and the occurrence of antisocial behaviour must not be
exclusively during the course of schizophrenia or manic episodes. In the
DSM-IV, the number of adult criteria are reduced to seven; one of these,
'consistent irresponsibility, as indicated by repeated failure to sustain
consistent work behaviour or honour financial obligations', incorporates
DSM-III-R's criteria 1 and 4, while the DSM-III-R criteria which specify
behaviour as a parent or guardian and relate to a monogamous relationship
for more than one year, are omitted.
In the ICD-10 the equivalent category is 'Dissocial PD', which is much
closer to the PCL-R criteria and to Cleckley's concept than is the DSM-III-
R's antisocial PD. As with all the specified PDs in the ICD-10, a diagnosis
of dissocial PD requires:
a minimum age of 16; markedly disharmonious attitudes and behaviour usually
involving several areas of functioning; that the abnormal behaviour pattern is
pervasive and clearly maladaptive to a broad range of personal and social
situations; that the above manifestations always appear during childhood or
adolescence and continue into adulthood; that the disorder leads to considerable
personal distress but this may only become apparent late in its course; and that the
disorder is usually, but not invariably, associated with significant problems in
occupational and social performance.
At least three of the following specific criteria are required for a diagnosis
of dissocial PD:
callous unconcern for the feelings of others; gross and persistent attitude of
irresponsibility and disregard for social norms, rules and obligations; incapacity to
maintain enduring relationships, though having no difficulty in establishing them;
very low tolerance to frustration and a low threshold for discharge of aggression,
including violence; incapacity to experience guilt and to profit from experience,
particularly punishment; and a marked proneness to blame others, or to offer
plausible rationalizations, for the behaviour that has brought the patient into
conflict with society.
Persistent irritability is an associated feature, while conduct disorder

during childhood and adolescence may further support the diagnosis
although it is not invariably present.
The Revised Psychopathy Checklist (PCL-R) (Hare, 1991) consists of the
following 20 items: glibness/superficial charm; grandiose sense of self-
worth; need for stimulation/proneness to boredom; pathological lying;
cunning/manipulative; lack of remorse or guilt; shallow affect; callous/lack
of empathy (i.e. a lack of a normal range and depth of emotions); parasitic
lifestyle; poor behavioural controls (i.e. violent behaviour or threats and
verbal abuse in response to frustration, failure, discipline or criticism);
promiscuous sexual behaviour; early (i.e. childhood) behaviour problems;
lack of realistic, long-term goals; impulsivity (i.e. behaviour that is carried
out suddenly, without preparation or consideration); irresponsibility;
failure to accept responsibility for own actions; many short-term marital
relationships; juvenile delinquency; revocation of conditional release; and
criminal versatility (i.e. involving many different types of offence). Hare et
^r/. (1991) have developed a ten-item criteria set for psychopathic PD based
on the PCL-R, each being rated on a 3-point scale; the 10 items are: 'glib
and superficial; inflated and arrogant self-appraisal; lacks remorse; lacks
empathy; deceitful and manipulative; early behaviour problems; adult
antisocial behaviour; impulsive; poor behavioural controls; and
irresponsible'.
Clinical features
Hare et al. (1991) have distinguished between interpersonal, affective and
behavioural features of psychopathy. Individuals with antisocial PD often
seem grandiose, self-centred, manipulative, dominant and unfeeling, and
find long-lasting relationships difficult. They may show labile emotions,
which lack consistency or depth, and a lack of empathy, anxiety, guilt and
remorse. Their behaviour is characterized by impulsivity and sensation-
seeking which is often antisocial and may involve criminality, substance
abuse and social irresponsibility.
The common features of antisocial PD in adolescence have been des-
cribed by West (1983); aggressiveness may be shown by defiance of parents
and authorities, destructiveness, quarrels and fights with peers, cruelty and
bullying, disgruntled and resentful attitudes, fierce temper and intolerance
of frustration. Failure to acquire social skills is shown by carelessness and
slovenliness, poor educational attainment in comparison with measured
intelligence, clumsiness and uncooperativeness in team efforts, restlessness
and inattentiveness in school, and frequent absences from class. There is a

lack of patience and sustained mental effort so that they seem lazy and
poorly motivated. Social rules are broken, involving theft, vandalism,
driving away vehicles, reckless risk-taking, truancy, drug and alcohol use,
early and irresponsible sexual behaviour or misconduct, running away
from home, and staying out at night. There are difficulties in relationships
shown by chronic conflicts with parents, distrust of others, a restriction of
peer relationships to those with similar characteristics and an unwillingness
to take part in organized leisure activities. Discipline evokes hostility and
there is stubbornness, unresponsiveness to punishment, and lack of guilt or
remorse. The first signs of conduct disorder in females who develop
antisocial PD usually appear at puberty, while in males these are generally
obvious in earlier childhood (American Psychiatric Association, 1987).
In a small minority of individuals with features of antisocial PD or PCL-
R psychopathy, aspects of mental functioning and behaviour may be
severely dangerous or bizarre, for instance, involving serious and/or
unprovoked aggression, fire-setting or murderous fantasies. In those
individuals whose serious aggression or fire-setting is associated with PD,
men are much more common than women, and, within this group, men
show more aggression while women have a history of more arson and
property damage.
For individuals with PD involving antisocial behaviour in secure hospi-
tals, associations have been reported with a history of incest, being in
institutional care as a child, childhood conduct disorder, social deprivation,
criminality and other PDs. Borderline, paranoid and narcissistic PD traits
have been noted, while some severely dangerous or bizarre individuals with
PD do not have antisocial PD or PCL-R psychopathy but rather border-
line, schizoid and/or schizotypal PD features.
One of the characteristic features of antisocial PD is an inability to
sustain close relationships due to self-centredness, lack of empathy and
callousness. In severe cases, the latter may involve the regular infliction of
pain and degradation. The lack of commitment to sexual partners is often
associated with a pattern of violence and separation.

Although one of the characteristics of an individual with severe antisocial
PD can be lack of anxiety, many antisocial individuals present with
complaints of tension and depression. The motive for medical consultation
often comes from pressure from others such as the courts, an employer or
the person's sexual partner who may threaten to leave unless 'treatment' is
sought. Many presentations to medical services involve the complications
of substance abuse (including drunken violence), self-harm episodes and
demands for medication. When contact is made there may be threats,
verbal abuse or actual violence. It has been estimated that antisocial PD as
defined by DSM-III-R occurs in 3% of American men but less than 1 % of
women.
Differential diagnosis can sometimes be difficult; for example, mania can
be associated with antisocial behaviour and this form of mood disorder can
present in relatively mild and chronic forms. However, the absence of
conduct disorder in childhood and an episodic nature of the antisocial
behaviour can be distinguishing features. Also, when adult antisocial
behaviour is mainly restricted to events associated with substance use
disorder (such as theft to obtain drugs or assaults when intoxicated), a
diagnosis of antisocial PD is often not justified.
Males are at much greater risk than females, and other associations with the
diagnosis include young age (under 45), lower socioeconomic status, urban
residence and homelessness (North, Smith & Spitznagel, 1993). Race does
not appear to be related to risk (Jordan et al.9 1989).
Robins (1978) has shown that while antisocial PD is clearly associated
with a history of childhood conduct disorder, most children with conduct
disorder do not go on to develop antisocial PD. If a subject gave a history of
several types of conduct disorder, this was a predictor of antisocial PD,
while social class was a poor predictor of serious adult criminality. Loeber
(1990) found that the pattern of childhood conduct disorder can predict
later delinquency, and that a high risk was associated with relative
frequency and variety, occurrence in multiple settings and relatively early
onset. Other risk factors in children for later aggressive offences were
hyperactivity, impulsive behaviour, attention problems, poor social skills,
poor peer relationships, academic problems and male sex. Subsequent non-
aggressive crime was associated with a history of non-aggressive conduct
disorders and female sex. Also, the DSM-III-R has claimed that antisocial
PD, as defined in this classification, is associated with low socioeconomic
class, abuse as a child, removal from the home and growing up without
parental figures of both sexes.
Family studies of antisocial PD involving biological children and
adoptees have shown that 'both genetic and environmental factors con-
tribute to the risk of this group of disorders' (American Psychiatric

Association, 1987). It has been reported that antisocial PD as defined by
DSM-III-R is about 5 times more common among first-degree biological
relatives of males with the syndrome than among the general population,
while the risk to the first-degree biological relatives of females with the
disorder is nearly 10 times greater. Within a family that includes a member
with antisocial PD, there is an increase of antisocial PD, substance use
disorder and somatization disorder (involving multiple complaints of
physical symptoms without obvious cause) in both males and females,
compared with the general population (American Psychiatric Association,
1987). In addition, individuals with antisocial PD have a reduced life
expectancy due to increased rates of self-harm episodes, substance use
disorders, violence and risk-taking behaviour.
The relationship between substance use disorders and antisocial PD is
complex; Brooner et al. (1992), in a study of 237 intravenous drug abusers,
found that the 44% with DSM-III-R antisocial PD were generally more
deviant and impaired than the rest. Twenty-four per cent fulfilled only the
adult criteria for antisocial PD and it was suggested that this reflected a late-
onset and less severe form of antisocial PD. Seven per cent fulfilled criteria
only for the childhood criteria for antisocial PD, while the remaining 25%
did not fulfil sufficient criteria for the diagnosis of either childhood or adult
antisocial disorder.
Current status
Definitions of antisocial PD that are based mainly on specified antisocial
behaviour have been validated by long-term follow-up of children with
conduct disorder, although this concept, as exemplified by DSM-III-R's
category of antisocial PD, has been criticized for emphasizing identifiable
antisocial acts at the expense of personality traits (Hare et al., 1991). Also,
there is increasing support for the recognition of a subgroup of individuals
with repeated antisocial behaviour that is mainly restricted to the context of
drug and alcohol abuse (Ferguson & Tyrer, 1991).
Borderline personality disorder
Main features
Borderline PD involves widespread instability in behaviour, shown by
intense and stormy relationships, impulsivity with destructive or otherwise
antisocial actions, self-harm and suicidal behaviour, self-mutilation or

substance abuse, labile mood (in particular with anger and depression), and
an inconsistent self-image.
When such features are pronounced, a patient is often considered as
'difficult to manage' and generally displays various other psychiatric
syndromes, such as affective disorder, substance abuse and somatization
disorder.
Clinical origins
Stern (1938) used the term 'borderline' to refer to patients who were usually
receiving regular psychotherapy and appeared severely disturbed and
difficult to manage, although not persistently out of touch with reality due
to delusions and hallucinations. However, the severity of the disorder
'bordered on' the generally-recognized serious mental disorders, such as
schizophrenia and manic-depressive disorder, and some authors believed
that the behaviour of such patients was related to the 'psychoses', i.e. the
most severe forms of mental illness, in particular to schizophrenia (Knight,
1953). This hypothesis was supported by reports that many patients
developed transient 'psychotic-like' symptoms (i.e. involving a departure
from reality) when sufficiently stressed.
Kernberg's (1967) description of the 'borderline personality organiza-
tion' involved a different perspective; the focus was not on a syndrome of
behaviour but on characteristics of mental functioning, such as a poor
capacity to tolerate anxiety, to control impulses and to develop socially-
productive ways of behaving, together with a tendency to have irrational
thinking patterns and to use certain psychological defence mechanisms.
The latter refer to ways of thinking that have the effect of making a person
less aware of feelings and attitudes; for instance, other people may be
categorized into 'all-good' or 'all-bad', so that any relationship is either
idealized or devalued, perhaps associated with a denial of contradictory
ideas or memories. Such 'splitting' may be partly recognized by the
individual and this can contribute to a feeling of insecurity in which he or
she continually feels neglected, hating those on whom he or she is
dependent. It has been claimed that another common defence associated
with borderline PD is 'projection', in which one's own unacceptable feelings
and attitudes are perceived as belonging to someone else. The concept of
borderline personality organization also involved 'identity disturbance'
with a lack of consistency about interests, values and desired relationships,
and it was believed to frequently co-occur with various other psychiatric
disorders such as anxiety, obsessive-compulsive disorder, hypochondriasis,
sexual disorders, substance abuse and impulsive, antisocial or self-harm
behaviour.
The DSM-III (1980) incorporated two of the previous uses of the term
'borderline'; features that were considered to be associated with schizo-
phrenia contributed to the category of 'schizotypal PD', while 'borderline
PD' reflected Kernberg's linkage of 'borderline' with personality and was
also based on the work of Gunderson and colleagues (Gunderson & Singer,
1975). Gunderson's concept of borderline PD included low achievement,
impulsivity, manipulative suicide attempts, intense mood changes, tran-
sient loss of reality, high socialization, and disturbed close relationships,
which were evaluated by the Diagnostic Interview for Borderline Patients
(Gunderson, Kolb & Austin, 1981). This consisted of a semistructured
interview with 29 items, covering instability in social functioning, poor
impulse control, disturbances of depression and anger, interpersonal
difficulties, and transient 'psychotic-like' disturbances, such as derealiza-
tion, depersonalization, depressive delusions and paranoid delusions. The
core features of the DSM-III concept of 'borderline PD' were impulsive,
unstable and intense relationships with inappropriate and intense anger,
but the criteria did not include the 'psychotic-like' features that had been
considered as important in the earlier clinical descriptions and in the criteria
of Gunderson and colleagues. The revisions that led to the DSM-III-R
reduced the co-occurrence of the diagnosis of borderline and schizotypal
PDs, as, with the DSM-III-R, depersonalization and derealization, (which
can occur in borderline PD, as well as in other disorders and in non-clinical
populations), were no longer examples of unusual experiences for the
diagnosis of schizotypal PD, as had been the case in DSM-III. Also, 'odd
behavior' was added as a criterion for this latter diagnosis. But it is of note
that both DSM-III and DSM-III-R criteria for borderline PD did not
include 'psychotic-like' features.
Main definitions
DSM-III-R claimed that the main feature of borderline PD is a pattern of
instability involving relationships, mood and self-image. For the diagnosis
to be made, at least five of the following are required:
(1) A pattern of unstable and intense interpersonal relationships characterized by
alternating between extremes of overidealization and devaluation. (2) Impulsive-
ness in at least two areas that are potentially self-damaging, e.g. spending, sex,
substance abuse, shoplifting, reckless driving, binge eating. (3) Affective instability:
marked shifts from baseline mood to depression, irritability, or anxiety, usually
lasting a few hours and only rarely more than a few days. (4) Inappropriate, intense
anger or lack of control of anger, e.g. frequent displays of temper, constant anger,
recurrent physicalfights.(5) Recurrent suicidal threats, gestures, or behaviour, or
self-mutilating behaviour. (6) Marked and persistent identity disturbance mani-
fested by uncertainty about at least two of the following: self-image, sexual
orientation, long-term goals or career choice, type of friends desired, preferred
values. (7) Chronic feelings of emptiness or boredom. (8) Frantic efforts to avoid
real or imagined abandonment.
The DSM-IV contains an additional criterion: 'transient, stress-related
paranoid ideation or severe dissociative symptoms'.
ICD-10 has an equivalent category of'emotionally unstable personality
disorder' and describes two variants, impulsive and borderline types, which
both show impulsiveness and lack of self-control. The impulsive type is
characterized by outbursts involving anger or violence, while the borderline
type is a wider concept involving many of the features of the equivalent
DSM-III-R borderline PD. It should be noted that both these classifica-
tions did not specify the 'psychotic-like' features which were noted in early
clinical descriptions associated with the 'borderline' term, although these
are referred to in the DSM-IV criteria. However, the Diagnostic Interview
for Borderline Patients includes 'psychotic' phenomena such as derealiza-
tion, depersonalization, brief psychotic depressive episodes, and brief
paranoid experiences.
Clinical features
Patients whose predominant problems are related to the features of
borderline PD are common in psychiatric practice, and present consider-
able difficulties in management due to volatile moods, impulsivity, sub-
stance abuse and self-harm. It has been claimed that the associations
between these various features produce three clusters, i.e. self-damaging
acts with impulsivity; anger, labile mood and stormy relationships; and
identity disturbance, emptiness and intolerance of being alone (Hurt et al.,
1988). However, the last cluster is the least specific to the borderline PD
syndrome. Other psychiatric disorders often co-occur, such as affective
disorders, substance abuse, eating disorders and short-lived 'psychotic-
like' features including non-delusional, paranoid experiences (such as ideas
of reference and marked suspiciousness), dissociative experiences, supersti-
tiousness, magical thinking and a 'sixth sense' (Gunderson & Zanarini,
1987).
The instability of self-image (i.e. identity disturbance) consists of uncer-
tainty related to concepts of self in various circumstances, involving a lack

of clarity of goals, beliefs, interests and values. Also, uncertainty in sexual
orientation is claimed in DSM-III-R, although it is unclear if this has any
specific association with borderline PD. But if sexual difficulties co-occur
with any PD, they are likely to contribute to an exaggeration of problem
behaviour. Such features of 'identity disturbance' are often found in the
context of a fear of being alone, and the patient may also complain of often
feeling empty, bored, uninterested, lonely, hopeless, and despairing.
Personal relationships are characterized by feelings that alternate between
extremes of overidealization or devaluation. Devaluation involves a
tendency to discredit or undermine the 'important others'; there is often a
wish for an exclusive relationship, together with anger when this ideal is not
realized. Such anger may be triggered by limit setting, confrontations or
separations that are actual or anticipated. There may be a sustained fear of
being abandoned, which can contribute to episodes of extreme panic when
alone. As a consequence of a tendency to devalue relationships, unless in a
phase of idealization, it is difficult for the individual to build up a store of
positive memories or fantasies about important people in his/her life in the
past or present. Instead, feelings and concepts about others tend to be
negative and associated with anger. This can be contrasted with a type of
'aloneness' which may be experienced by a person without obvious PD,
when there is a longing to regain a lost relationship or experience,
accompanied by positive feelings in relation to the memories (Adler, 1981).
Mood is subject to rapid and frequent shifts, which may involve depression
with thoughts of self-harm and suicide, irritability, anxiety, and chronic,
intense anger. Chronic anger is accompanied by bitterness, sarcasm and
unreasonable demands (Gardner et al, 1991). The mood component of the
range of features of depressive disorders consists of a variety of experiences,
and it has been claimed that depression in borderline PD can have a
characteristic quality with thoughts of sustained emptiness and destructive-
ness, associated with a 'conflicted' dependency, involving mixed feelings of
hostility and dependency on an important person or situation. In contrast,
when an unconnected dependency (i.e. without the hostility) is affected by
loss (or anticipated loss), the resulting features of depression have a
different quality and a better outcome (Gunderson & Zanarini, 1987).
Various studies have investigated the prevalence and nature of 'psycho-
tic-like' phenomena in samples of individuals selected on the basis of
borderline PD or of various other psychiatric disorders, and in non-clinical
populations. But the term 'psychotic-like' is confusing; taken at face value it
indicates phenomena which resemble certain features, namely delusions
Table 2.1 Features of the mental state and/or history involving 'departures
from consensual reality' associated with borderline personality disorder
(PD)
Degrees of % of subjects
specificity of the with borderline
associations of PD who had a
mental state features history of the
with borderline PD phenomena shown
compared with (Zanarini et ai,
Feature other PDs 1990)
Cognitive-perceptual distortions: ++
(except in relation to
schizotypal PD)*
(i) Odd or illogical reasoning 68%
(e.g. marked suspiciousness, (Controls 24%)**
illogical thinking, belief in
telepathy and sixth sense)
(ii) Unusual perceptions (e.g. 62%
recurrent illusions, (Controls 20%)
depersonalization,
derealization, body-image
distortions)
(iii) Non-delusional paranoia 100%
(e.g. undue suspiciousness, (Controls 66%)
ideas of reference, other
paranoid states)
Transient psychotic experiences, ++++ 40%
i.e. delusions and/or (Controls 2%)
hallucinations for periods
usually less than 2 days (or if for
longer, may be atypical
compared with other 'psychotic'
disorders)
Prolonged delusions and + 14%
hallucinations (Note: usually (Controls 4%)
secondary to co-occurring
mental disorders other than
PDs)
Notes:
* Schizotypal PD has a higher prevalence of associated cognitive-perceptual
distortions compared with borderline PD (Sternbach et al., 1992).
** The control group consisted of patients with other PDs and contained 14.5%
with a paranoid, schizoid or schizotypal PD.
and hallucinations, of the so-called 'psychoses', such as schizophrenia,

manic-depressive disorder or dementia. However, a delusion, as usually
defined, can also occur in other disorders, for example, obsessive-compul-
sive disorder. It can be argued that a better term than 'psychotic-like'
phenomena is 'departures from consensual reality', when reality is defined
as a consensus opinion of a group of individuals (such as clinicians) who are
drawn from the non-patient population. Table 2.1 outlines the main
'departures from consensual reality' that have been considered to be part of
the syndrome of borderline PD. These are divided into three categories; the
first, 'cognitive-perceptual distortions', includes odd or illogical reasoning,
unusual perceptions and non-delusional paranoia such as undue suspi-
ciousness. These features have been described as occurring more frequently
with borderline and schizotypal PDs (especially the latter) compared with
other PDs. The second category consists of 'transient psychotic exper-
iences', and usually involves hallucinations and delusions for relatively
short periods of less than two days. If they last longer in borderline PD,
these phenomena would be likely to affect only a discrete part of the
patient's functioning or differ from the major psychotic disorders by having
a banal or fantastic content. Zanarini et al. (1990) found that, in their
sample with borderline PD, a history of transient hallucinations was more
common than of transient delusions, i.e. 26% and 20% respectively, and in
another sample with borderline PD, Sternbach et al. (1992) reported that
the ratio of transient auditory to visual to olfactory hallucinations was
90:30:9. Although histories of transient 'psychotic' experiences are relati-
vely specific to borderline PD (i.e. they are rarely found with other PDs or in
non-clinical populations), Zanarini et al. (1990) found that these pheno-
mena had been present in the previous two years in only 40% of the sample
with borderline PD. The third category of 'departures from consensual
reality' includes delusions and/or hallucinations which are prolonged.
Although these can occur in association with borderline PD, they are rare,
and are usually associated with a co-occurring mental disorder.
Because individuals with borderline PD find being alone difficult to
tolerate, they tend to seek out company and support. Gunderson &
Zanarini (1987) have described the developmental cycle that often affects
such relationships; a clinging dependent style develops in the context of
sustained dysphoria, and when the relationship is disrupted, perhaps by
limit-setting, confrontation or separation, this leads to threats of self-harm,
somatic complaints and misleading communications. The dependency,
which may be transiently idealized, is then replaced by an angry devalua-
tion. Perceived or actual loss of support then leads to panic, with attempts
to ward off feelings of aloneness by impulsive actions involving substance

abuse, promiscuity and recklessness.
Some common behaviours associated with borderline PD have already
been described. These include suicidal threats and impulsive self-damaging
behaviour such as drug overdoses, wrist cutting and episodic substance
abuse, together with promiscuity, sexual deviance, aggression and leaving
home. Also, self-mutilation may occur, commonly taking the form of
multiple small cuts on the forearms; individuals usually recognize that this
is not an attempt at suicide, but rather a response to feelings of extreme
tension and anger, which may be relieved by the episode. Often the threats
or episodes of self-harm appear to be designed to elicit a supportive and
saving response from 'significant others', but such episodes must not be
dismissed on this account. Whatever the underlying motivation, repeated
self-harm puts the patient at a significantly increased risk of completed
suicide.
Coid (19936) described 72 females detained in English maximum-
security hospitals under the Mental Health Act 1983 under the legal
category of'psychopathic disorder' who fulfilled criteria for borderline PD,
and suggested that such women had a distinct affective syndrome. The
women showed mood-related self-harm behaviour with a feeling of com-
pulsion, which seemed to relieve the original mood symptoms.

Patients with borderline PD often demand urgent help in a crisis and are
difficult to manage because of their instability of mood, threats of self-
harm, self-harm episodes including self-mutilation without obvious suici-
dal intent, and intense anger. Also, they often present with a variety of
somatic complaints for which no obvious cause or effective treatment can
be found, or with other psychiatric disorders including depression, anxiety
and substance abuse. Violence may be a feature of the presentation,
especially if associated with antisocial PD features and substance abuse.
Borderline PD has often been reported to co-occur with schizotypal,
histrionic, antisocial and narcissistic PD features, as well as with substance
abuse and affective disorders (Higgitt & Fonagy, 1992; Oldham et ah,
1992). Also, families of patients with borderline PD had greater prevalences
of depressive, substance abuse and antisocial disorders (Goldman et al.,

1993).
Co-occurrence with these other PDs is due, in part, to some overlap in the
criteria for definition, as antisocial PD involves unstable relationships,
histrionic PD is associated with rapidly shifting emotions, and schizotypal
PD includes cognitive-perceptual distortions. But, in addition, co-occur-
rence of the other features of these PDs is common in clinical populations
(Gunderson, Zanarini & Kisel, 19916).
In this chapter, schizoid and schizotypal PD have been considered
together, but even if they are separated, there is evidence from follow-up
studies that borderline PD is distinct from schizoid and/or schizotypal PD.
Borderline PD is not genetically related to schizophrenia (McGlashan,
1983), although borderline, schizoid and schizotypal PDs may show certain
features that can also occur in schizophrenia, such as suspiciousness, ideas
of reference, odd beliefs, illogical thinking, and unusual perceptual exper-
ience. Cognitive-perceptual distortions appear to be more common and
more enduring in schizoid (and schizotypal) PD, compared with borderline
PD (Sternbach et al.9 1992).
Co-occurrence of borderline PD with histrionic and antisocial PD
features has often been reported (Gunderson & Zanarini, 1987), and there
is evidence that this adversely affects outcome. Narcissistic PD features
have also been found to be common in individuals with borderline PD.
While there is an undoubted association between some features of
depressive disorders and borderline PD, the nature of such relationships
remains to be clarified (Tarnopolsky & Berelowitz, 1987).
Current status
Although the borderline PD syndrome was not introduced into one of the
main classifications of mental disorder until 1980, the subsequent decade
saw more publications related to this disorder than to any other PD. The
features of borderline PD are of considerable importance in clinical
populations, and appear to be the basis of some of the most severe and
enduring problems in routine clinical practice.
Follow-up studies suggest that the syndrome shows a degree of temporal
stability, although improvement generally occurs by the fourth decade.
Genetic studies suggest that borderline PD is not related to schizophrenia,
while the relationships with various types of affective disorder require
further study (Gunderson & Elliot, 1985). Future revision of the main
classifications will need to take into account the frequent and relatively
specific occurrence of transient, and possibly stress-related, delusions and
hallucinations, as well as the less specific cognitive-perceptual distortions,
in patients with borderline PD. Modifications to present criteria could
involve a recognition that mood changes usually appear reactive to
relationship and environmental changes (Gunderson et aL, 1991&).
Histrionic (or hysterical) personality disorder
Main features
Individuals with DSM-III-R's histrionic PD often show attention-seeking
behaviour, self-centredness, rapidly shifting and exaggerated emotional
reactions, sexual provocativeness and an excessively impressionistic style of
speech. But this diagnosis is a concept that has been criticized as possibly
representing a prejudiced male view of women - 'a caricature of femininity'
(Ferguson & Tyrer, 1988).
Clinical origins
Easser & Lesser (1965) have described how Sigmund Freud recognized a
relationship between what in the DSM-III-R would be termed 'somatiza-
tion disorder' or 'conversion disorder' (i.e. involving physical complaints
that apparently are not due to a recognizable disease) and the 'erotic
personality, whose major goal in life is the desire to love or above all to be
loved'. There is a resemblance between this description and DSM-III-R's
criteria for histrionic PD, which include: 'is inappropriately sexually
seductive in appearance and behaviour'. Also, the DSM-III-R claims that
'brief reactive psychosis' (which may involve hallucinations and/or
delusions in the context of a dissociative disorder), 'conversion disorder'
and 'somatization disorder' are possible complications. Thus, the idea of a
link between a specified personality disorder and a vulnerability to so-called
'hysterical disorders' (i.e. somatization disorder, conversion disorder and
dissociative disorders) has endured.
Histrionic PD represents a concept that has had an evolving clinical
tradition over many decades, based mainly on clinical descriptions,
although there are some supporting data (Pfohl, 1991). However, the terms
histrionic or hysterical PD have often been applied without precision;
Lazare (1971) noted that 'hysterical is commonly used in a perjorative sense
to describe a patient who is self-engrossed, incapable of loving deeply,
lacking depth, emotionally shallow, fraudulent in affect, immature,
emotionally inconsistent, and a great liar . . . The presence of just one of
these traits together with a tired resident, may result in the diagnosis of "just
hysterical'".
Kernberg (1967) outlined features that were considered to distinguish
histrionic or hysterical PD from other disorders; these were: emotional
lability; overinvolvement; dependent and exhibitionistic needs; pseudohy-
persexuality; sexual inhibition; competitiveness; and masochism. In a
review of pre-1966 literature, Lazare, Klerman & Armor (1966) identified
the suggested features of dependence, egocentricity, emotionality, exhibi-
tionism, fear of sexuality, sexual provocativeness and suggestibility, while a
factor analytic investigation of hysterical personality (Lazare et ah, 1966;
Lazare, Klerman & Armor, 1970), based on a self-report rating scale (the
Lazare-Klerman Trait Scale), obtained a factor that resembled clinical
descriptions of hysterical PD. This involved aggression, emotionality,
exhibitionism, egocentricity and sexual provocativeness. It should be noted
that such a factor is a construct that is not a single directly measurable
variable, but is derived from the association between the measurement of
other, directly observable variables.
The DSM-I of 1952 did not include an histrionic PD category, although
some of the above features contributed to the broader concept of
'emotionally unstable personality'. However, the DSM-II, in 1968,
included 'hysterical personality' (and the alternative term 'histrionic PD'),
which was claimed to be 'characterized by excitability, emotional instabi-
lity, overreactivity, and self-dramatization. This self-dramatization is
always attention-seeking and often seductive, whether or not the patient is
aware of its purpose. These personalities are also immature, self-centred,
often vain, and usually dependent on others. This disorder must be
differentiated from Hysterical neurosis'. But the DSM-III revision of the
histrionic PD category, in 1980, did not include seductiveness, and the
overlap with the criteria for borderline PD led to further changes in DSM-
III-R (1987) in which some of the overlapping criteria such as 'prone to
manipulative suicide attempts' were omitted. Also, two new criteria were
added i.e. 'is inappropriately sexually seductive in appearance or behav-
iour', and 'has a style of speech that is excessively impressionistic and
lacking in detail'. (The resulting DSM-III-R criteria are similar to the
syndrome obtained by Lazare and colleagues by factor analysis.)
The various definitions of hysterical or histrionic PD have been asso-
ciated with a greater prevalence of these diagnoses in women, and there has
been concern that this may sometimes reflect a prejudiced tendency for the
clinician to make the diagnosis more often in women, perhaps as a result of
not giving this diagnosis to similar features in men. Some studies have
indicated that the application of the histrionic PD diagnosis may indeed be
subject to such bias, although additional reasons may contribute to a
greater prevalence in women, and it has been suggested that there may be
sex-related differences in the features of histrionic PD (Pfohl, 1991). It is of
interest that a twin study indicated a genetic contribution for histrionic PD
in women but not in men (Torgerson & Psychol, 1980). But the possibility
that histrionic PD is a diagnostic category that is liable to be misused by
some clinicians as a perjorative label, reflecting negative attitudes and the
denial of appropriate treatment, needs to be kept under review.
Main definitions
For the DSM-III-R's histrionic PD, four or more of the following eight
criteria are required:
constantly seeks or demands reassurance, approval, or praise; is inappropriately
sexually seductive in appearance or behaviour; is overly concerned with physical
attractiveness; expresses emotion with inappropriate exaggeration, e.g. embraces
casual acquaintances with excessive ardour, uncontrollable sobbing on minor
sentimental occasions, has temper tantrums; is uncomfortable in situations in which
he or she is not the centre of attention; displays rapidly shifting and shallow
expressions of emotions; is self-centred, actions being directed toward obtaining
immediate satisfaction - has no tolerance of the frustrations of delayed gratifica-
tion; and has a style of speech that is excessively impressionistic and lacking in
detail, e.g. when asked to describe mother, can be no more specific than, 'She was a
beautiful person'.
The DSM-IV criteria omit DSM-III-R criteria 'constantly seeks or

demands reassurance, approval or praise' and 'is self-centred...', which are
replaced by two new criteria, i.e. 'is suggestible, i.e. easily influenced by
others or circumstances', and 'considers relationships to be more intimate
than they actually are'.
The ICD-10 also has a category of histrionic PD. This has six criteria, five
of which have similar counterparts in DSM-III-R, i.e. self-dramatization,
theatricality, exaggerated expression of emotions; shallow or labile affecti-
vity; continued seeking for excitement, and activities in which the patient is
the centre of attention; inappropriate seductiveness in appearance or
behaviour; and over-concern with physical attractiveness. The remaining
ICD-10 criterion, i.e. 'suggestibility, easily influenced by others or by
circumstances', is also included in the DSM-IV criteria. Associated features
are described as egocentricity, self-indulgence, continuous longing for
appreciation, feelings that are easily hurt, and persistent manipulative

behaviour to achieve own needs.
Clinical features
Many features have already been noted. The characteristics of attention-
seeking behaviour incorporate exhibitionism (i.e. drawing attention by
dress, speech or behaviour) and exaggerated expressions of emotion, at
times resembling a theatrical performance. Such individuals may be good
company, welcome guests, good at public speaking and accomplished in
amateur dramatics. Some cases are reminiscent of Jaspers' description
(1946) of a personality that seems to have lost its core and consists entirely
of a series of shifting masks.
Other features include a craving for novelty and excitement, short-lived
enthusiasms, a tendency to become bored, a lack of consideration for
others, excessive vanity and concerns with physical appearance, self-
centred behaviour involving excessive demands of others with angry scenes
or demonstrative self-injury, and a readiness to display exaggerated
emotion easily, perhaps involving angry tantrums or dramatic despair with
rapid recovery. The latter behaviour is often manipulative, in that it is
designed to influence the feelings or behaviour of others. This may present
with overdoses or other forms of self-harm, which seem mainly motivated
by anger and a wish to punish partners or family members. Such individuals
may be seen in a very angry and distressed state, perhaps leading to hospital
admission, but this rapidly improves. Sometimes it is difficult to reconcile
the calm, composed individual in the hospital with the clinical notes of the
emotionally-fraught admission the previous day.
Inappropriate seductive behaviour, involving flirting and sexual
advances, may be part of a pattern in which the person with histrionic PD
finds it difficult to relate to the preferred sex in a non-sexual way (Livesley &
Schroeder, 1991). Such individuals tend to try and control the other person
in a relationship or to enter a dependent relationship, so that interactions
tend to have a child-parent or parent-child quality. A mature, adult
relationship involves considerable flexibility of roles with give-and-take,
which is difficult for those with histrionic PD to achieve. Such individuals
may be relatively successful in looking after or relating to old people,
children or adults with handicaps as, in all these instances, the relationship
is relatively structured, with a parent-child element. There is a tendency to
indulge in romantic fantasy, and the quality of sexual relationships varies
from promiscuity through normality to unresponsiveness. Sometimes the
recipient of seductive behaviour is a doctor or another health care worker

who is in a professional relationship with the patient. If these advances are
reciprocated the patient is likely to tell her (or his) spouse or partner, which
can then lead to a formal complaint. It seems that the motivation behind
this behaviour is often the gratification from the attention that this
situation provides.
An excessively impressionistic style of speech is not invariable but, when
present, can be strikingly obvious during interview, when it is very difficult
to obtain details about any topic. A further description is provided by Pfohl
(1991), i.e. a 'cognitive style dominated by formation of poorly delineated
global impressions without consideration of details that support or contra-
dict their opinion, e.g. believes a particular professor is a fantastic teacher
but cannot provide details about how the professor's technique differs from
others and is unable to identify areas that could use improvement'. Such
people may have creative talents but are unlikely to be good at thinking that
involves logical analysis.
The first DSM-III-R criterion for histrionic PD, 'constantly seeks or
demands reassurance, approval or praise', is often found when this
diagnosis is made but is also common in patients with other PDs. Pfohl
(1991) has recommended that this should be omitted in future revisions of
the DSM-III-R criteria (as, indeed, has occurred in the DSM-IV criteria) or
that a more detailed evaluation of the relevant behaviour is undertaken to
distinguish the behaviour in histrionic PD from similar behaviour in other
PDs.
The clinician will often notice that these patients take extra time and energy, as they
require reassurance and approval on matters that may go well beyond the expertise
of the clinician. The patient shows minimal interest in being taught how to
independently evaluate his/her performance and expresses the need to have
someone stronger provide a stamp of approval for his/her decisions. This contrasts
with the related criterion for narcissistic PD where the need for attention and
admiration suggests that the individual views him/herself as equal or superior to the
person providing the attention.
Pfohl (1991) also recommends that a second change be made to the
DSM-III-R criteria for histrionic PD, in relation to the criterion: 'is self-
centred, actions being directed toward obtaining immediate satisfaction;
has no tolerance for the frustration of delayed gratification'. Pfohl noted
that four studies have found that a positive rating for this criterion was
common in relation to other PDs and that it really consists of two separate
items. A suggested alternative was: 'is intolerant of, or frustrated by,
situations involving delayed gratification', which would refer to just one of
the two traits. However, both parts of the criterion have been omitted from
the DSM-IV criteria. In addition, Pfohl (1991) also introduced a proposed
new criterion based on clinical experience, i.e. 'tends to view relationships
as possessing a greater level of intimacy than is actually the case, e.g. refers
to someone he/she recently met as a "dear, dear friend", refers to doctors
and other professionals by first name and talks about a "special" relation-
ship when referring to a doctor known on a casual professional level'. (As
has been noted, this suggestion was included in the DSM-IV criteria.) It has
been pointed out that while a person with narcissistic PD may also talk
about 'special' friendships, the important status of these friends is empha-
sized rather than the special nature of the intimacy.

There is evidence that histrionic PD often co-occurs with other PD
diagnoses, in particular borderline, antisocial and narcissistic PDs, and
with somatization disorder, which involves a history of many physical
complaints with no biological changes to account for the symptoms
(Pollack, 1981; Kaminsky & Slavney, 1983; Lilienfeld et al.9 1986).
Individuals with histrionic PD may also present with symptoms of
anxiety and depression, manipulative self-harm in the context of relation-
ship difficulties and, as with many individuals with PD, substance abuse.
Also, it is probable that there is an increased vulnerablity to conversion
disorder and dissociative disorders, which can include certain types of
delusions and hallucinatory experiences.
The co-occurrence of histrionic PD with borderline, antisocial and narcis-
sistic PDs, and with somatization disorder, conversion disorder and
dissociative states, has been noted. Histrionic PD appears to be diagnosed
more frequently in women and has been claimed to be more common
among first-degree relatives of people with the disorder than among the
general population (American Psychiatric Association, 1987).
Although some features of histrionic PD overlap with those of borderline
PD (i.e. both are associated with unstable mood and dependent or 'clinging'
behaviour), histrionic PD is generally associated with more stability in
relationships and roles, relative absence of cognitive-perceptual distortions
and transient psychotic experiences, better self-esteem linked to sexual
desirability, less hostility and a more vague way of thinking (Gunderson &
Zanarini, 1987).
Current status
Despite a long clinical tradition in the descriptive literature, the diagnosis of
'histrionic' PD requires further research to determine whether it merits a
separate category. However there is no doubt that the above features
appear to be relevant to the problems of many patients with PD in
psychiatric practice, although it may be that, in general, they co-occur with
prominent features of other PDs. Poor test-retest reliability has been
reported, in particular with an unstructured clinical interview (Pfohl, 1991),
and sex-related differences in the use of the diagnosis require further
investigation.
Narcissistic personality disorder
Main features
The main features have been described as 'a pervasive pattern of grandi-
osity (in fantasy or behaviour), lack of empathy, and hypersensitivity to the
evaluation of others' (American Psychiatric Association, 1987).
Clinical origins
Narcissism has been defined as 'an interest in (or focus on) the self
(Bursten, 1989), and while Sigmund Freud's paper 'On narcissism'
appeared in 1914(Freud, 1955), it was not until 1968 that Kohut introduced
the term 'narcissistic personality disorder' to describe disturbances in
several areas including grandiosity and pronounced angry reactions. The
inclusion of narcissistic PD in DSM-III, in 1980, was largely due to reports
in the psychoanalytic psychotherapy literature, derived from descriptions
of subjects for whom narcissistic behaviour was predominant and discrete,
that is, not accompanied by other significant psychopathology (Akhtar &
Thomson, 1982). Despite its appearance in DSM-III, there had been little
experimental evidence to justify its inclusion, while subsequent research
using DSM-III and DSM-III-R criteria has mostly involved patients who,
although eligible for the diagnosis of narcissistic PD, are dissimilar from the
patients (many in fee-paying psychotherapeutic relationships), whose
characteristics led to the original descriptions of the syndrome. The 1987
revisions, which led to DSM-III-R, substantially increased the number of
patients who fulfilled diagnostic criteria for narcissistic PD (Gunderson et
aL, 1991a).
The DSM-III-R criteria for narcissistic PD have been assessed by a semi-
structured interview, the Diagnostic Interview for Narcissism (DIN)
(Ronningstam & Gunderson, 1989, 1990), by structured interviews for the
assessment ofall DSM-III-R PDs (Stangle/a/., 1985; Loranger etaL, 1987;
Zanarini et aL, 1987), and by the revised version of the Personality
Diagnostic Questionnaire (PDQ-R), which is a self-report instrument
(Hyler et aL, 1990). Additional methods of identification relate to other
definitions of narcissism (Dowson, 1992<z).
The DSM-III-R states that the essential feature is 'a pervasive pattern of
grandiosity (in fantasy or behaviour), hypersensitivity to the evaluation of
others, and lack of empathy', while a previous comparison of three
diagnostic systems, including DSM-III, concluded that 'in differential
diagnosis, the most outstanding and specific narcissistic characteristics are
grandiosity and entitlement' (Ronningstam, 1988). Studies with the DIN
and its precursor compared a group of patients, characterized by discrete
clinically-identified narcissistic psychopathology, to groups of subjects
with other PDs (Ronningstam & Gunderson, 1988, 1990, 1991). These
studies indicated that four criteria, i.e. 'reacts to criticism with feelings of
rage, shame or humiliation...'; 'is interpersonally exploitative...'; 'lack of
empathy ...'; and 'is preoccupied with feelings of envy ...', did not
discriminate significantly between the groups, while aspects of grandiosity
were the most distinctive aspects of the narcissistic group.
A study of 291 patients with various PDs, based on data from a
questionnaire completed by clinicians, found that 22% received the DSM-
III-R diagnosis of narcissistic PD (Morey, 1988), while a study which used
the PDQ-R self-report questionnaire reported a diagnosis of narcissistic
PD in 34% of 87 patients applying for inpatient treatment of severe PD
(Hyler et aL, 1990). Two semistructured interviews were also applied in the
latter study and produced narcissistic PD diagnosis in 17% and 22%
respectively. Also, there are many claims of co-occurrence of narcissistic
PD, as defined by DSM-III or DSM-III-R, with other PDs (Dowson,
1992a).
As one of the aspects of narcissism involves difficulties in the subject
viewing himself/herself realistically (Gunderson, Ronningstam & Bodkin,
1990), an informant's view might be expected to be important in assess-
ment. However, there is little empirical evidence; a study using DSM-III
criteria and semistructured interviews of 66 depressed patients and their
informants reported low diagnostic agreement for all PDs (Zimmerman et
ai, 1988). Informants reported more pathology, but while narcissistic PD
was not diagnosed on the basis of any patient's data, this diagnosis was
made in only three subjects on the basis of informants' data. But, in a study
of DSM-III-R narcissistic PD evaluated by patients' and informants' self-
report questionnaire, one criterion, i.e. 'has a sense of entitlement ...',
showed 'fair to good' reliability when patients' and informants' ratings
were compared (K = 0.62) (Dowson, 1992a). It was concluded that the
identification of a sense of entitlement by the patient may be a relatively
reliable and valid indicator of narcissism, when data are obtained from a
patient's self-report questionnaire or from the clinical interview.
Main definitions
Narcissistic PD is not represented in the ICD-10 classification but the
DSM-III-R provides nine criteria, five or more of which are needed for
diagnosis, i.e.
reacts to criticism with feelings of rage, shame, or humiliation (even if not
expressed); is interpersonally exploitative: takes advantage of others to achieve his
or her own ends; has a grandiose sense of self-importance, e.g. exaggerates
achievements and talents, expects to be noticed as 'special' without appropriate
achievement; believes that his or her problems are unique and can be understood
only by other special people; is preoccupied with fantasies of unlimited success,
power, brilliance, beauty or ideal love; has a sense of entitlement; unreasonable
expectation of especially favourable treatment, e.g. assumes that he or she does not
have to wait in line when others must do so; requires constant attention and
admiration, e.g. keeps fishing for compliments; lack of empathy: inability to
recognize and experience how others feel, e.g. annoyance and surprise when a friend
who is seriously ill cancels a date; is preoccupied with feelings of envy.
The DSM-IV criteria omit the first of the above criteria (i.e. 'reacts to
criticism...') and include a new criterion, i.e. 'arrogant, haughty behaviors
or attitudes'. Also, the fourth DSM-III-R criterion is expanded to: 'believes
that he or she is "special" and unique and can only be understood by, or
should associate with, other special or high-status people (or institutions)'.
Clinical features
Ronningstam & Gunderson (1990) identified the following characteristics
as significantly more common in a sample based on features of narcissistic
PD compared with another comparison group: a sense of superiority with
arrogant haughty behaviour and a sense of uniqueness; exaggeration of

talents, boastful and pretentious behaviour and grandiose fantasies,
although often with high achievement; self-centred and self-referential
behaviour; and need for attention and admiration.
A sense of superiority, with arrogant and haughty behaviour, involves an
unrealistic view that he/she is better than other people. Bad aspects of self
are forgotten, and there is a tendency to forgive oneself too easily, with an
exaggerated view of a perceived insult when the subject's superiority is not
acknowledged. Other people may be viewed with disdain as inferiors, while
their successes are downgraded (Livesley & Schroeder, 1991). Feelings of
superiority may also involve a sense of entitlement, in which special
favours, rights or privileges are expected. In the medical context this may
involve an unrealistic expectation of 'total cure' in which all problems will
be solved. However, such superiority is usually accompanied by a fragile
self-esteem and hypersensitivity to criticism. It is as if a completely
unconditional positive regard is expected and anything short of this is
unacceptable, giving rise to anger. Horowitz (1989) has described various
forms of anger, the most severe of which is an explosive (i.e. sudden and
severe) self-righteous even 'towering' or 'blind' rage, perhaps triggered by
an insult. This may be characterized by verbal abuse or physical violence. In
extreme cases the person feels completely justified in having such feelings, at
least during the state of anger, as others are viewed as subhuman without a
right to exist. But as some guilt and shame may be present at the time or
subsequently, this can lead to a mixture of anger and shame at aspects of self
(such as a tendency to irrational anger), together with anxiety that is related
to fear of loss of emotional control. Also, a state of chronic embitterment
may occur, with ideas that the person has been treated unfairly; this may
present as a 'blustery-outgoing' or 'sullen-withdrawing' behaviour pattern.
A sense of uniqueness describes a belief that the self is special and that few
others have much in common. Such individuals often complain of being
misunderstood, as they consider that only someone very special would be
able to understand them.
Exaggeration of talents and achievements is often accompanied by
boastful and pretentious behaviour, which can be the most obvious sign of
narcissistic PD features in social interactions, including the clinical inter-
view. The activities of such individuals are constantly referred to and they
may draw attention to themselves in ways that assume their importance.
Grandiose fantasies may involve unlimited success, power, beauty, wealth
or ideal love, and sometimes there is a history of periods of success that may
be seen as a justification of their superiority. Ronningstam & Gunderson
(1990) pointed out that such individuals are more likely to be seen in private
practice rather than hospital-based services.
Self-centred and self-referential behaviour involves the person appearing
self-preoccupied with a tendency to assign personal meaning to events that
are unrelated to them. It has also been observed that although they lack
interest in the personal opinions and reactions of others, this may be
associated with the previously-mentioned hypersensitivity to criticism.
Although the usual pattern is one of self-absorption and preoccupation
with ideas of superiority, this equilibrium can be upset by events, such as
criticism, which challenge these assumptions. Relationships are character-
ized by selfishness and a lack of input in the form of empathy and positive
regard for others (Horowitz, 1989). Other people are used to further the
person's aims, perhaps involving 'social climbing' or occupational advan-
cement, while another person may be exploited when, if the relationship has
adverse effects on the other individual, this is accepted or ignored by the
narcissistic subject. People are often discarded as friends when the relation-
ships are no longer useful. If the person with PD does not have special
abilities, another individual with talent may be selected for a dependent
relationship that has been termed a 'mirror transference' (Horowitz, 1989);
loss of such a person may lead to a severe, hopeless grief. Finally, a need for
attention and admiration can be reflected by an excessive concern with
grooming and appearance, 'fishing for compliments' or exhibitionist
behaviour. Sometimes, physical disfigurement can lead to a severe and
prolonged depressive reaction.
Evaluations of the specificity of individual DSM-III-R criteria for
narcissistic PD, in relation to patients with narcissistic PD and comparison
groups, has indicated that those criteria related to grandiosity perform best,
while the following three perform poorly: 'reacts to criticism with feelings
of rage, shame or humiliation'; 'lack of empathy'; and 'is preoccupied with
feelings of envy' (Gunderson et ai, 1991a). Angry reactions to criticism are
also commonly found in relation to paranoid and borderline PDs, and it
was suggested that the related criterion for narcissistic PD might be
reworded to: 'reacts to criticism, defeat, or rejection with sustained feelings
of disdain, shame or humiliation (even if not expressed)', although this
ignores the fact that rage can occur in narcissistic PD. With regard to lack of
empathy, Ronningstam & Gunderson (1989) noted that this is difficult to
evaluate, particularly in a single interview, and is also commonly associated
with antisocial and passive-aggressive PDs. The third problem criterion,
which is related to envy, was found to be relatively uncommon and also
associated with histrionic and avoidant PDs.
Ronningstam & Gunderson (1990) claimed that narcissistic PD often

includes three features that are not fully covered by DSM-III-R criteria,
namely 'boastful and pretentious behaviour, arrogant and haughty behav-
iour, and self-centred and self-referential behaviour'. Other non-DSM
features were described in a study by Morey (cited in Gunderson et al.,
1991fl) and consisted of the following: egocentricity, dominance, interper-
sonal disdain, preoccupation with status, petulant anger and fragile self-
concept. Also, the following associated features have been noted: acts
arrogantly; self-assured and confident; has sense of high self-worth; viewed
as vain and self-indulgent; and views self as gregarious and charming. It has
been suggested that a new criterion for narcissistic PD might be 'arrogant,
haughty behaviour and/or attitudes', which can often be identified in single
interviews and is not usually found with histrionic, antisocial or borderline
PDs (Gunderson et al., 1991a). This has been included in the DSM-IV
criteria.

Narcissistic PD, as defined in DSM-III-R, does not often occur in patients
who do not fulfil criteria for other PDs, although in some individuals,
particularly in private psychotherapeutic practice, narcissistic features may
be predominant.
Individuals with a PD which includes significant narcissism may seek
help because of depressive symptoms and anxiety in a social crisis or in
relation to events that challenge feelings of superiority. The presentation
may involve shame, panic, helplessness, depressed mood, hypochondriasis,
self-destructiveness, envy, rage (which occasionally can lead to very serious
physical violence), excessive demands, and substance use disorder. He or
she may feel humiliated about being depressed, and discomfort is not well
tolerated; the exceptional quality of the despair may be stressed. The faults
of others may be cited together with the uniqueness of the patient's
problems.
Often there are unrealistic expectations of what help can be given; the
patient appears to believe that because a special or unique problem has been
presented, he or she is then entitled to its complete resolution. Anything less
than this goal or the unlimited efforts of others is unacceptable; for
example, the patient may feel he or she should be seen on a daily basis and
will not accept a delay for a further appointment.
Gunderson et al. (1991<z) have reviewed 11 studies based on DSM-III and
DSM-III-R definitions of narcissistic PD, which showed that it is rare for
patients who meet criteria for this disorder not to meet criteria for other
PDs, in particular histrionic, borderline and antisocial PDs, but also
passive-aggressive, paranoid and schizotypal PDs. In a study of 60 patients
evaluated for DSM-III-R PDs by self-report questionnaire, narcissistic PD
scores (i.e. the number of positive narcissistic PD criteria for each subject)
were significantly correlated with histrionic, borderline and passive-
aggressive PD scores (Dowson, 1992a).
Current status
While many of the features of narcissistic PD occur in patients with various
PDs, it is likely that most of the patients who are eligible for the DSM-III or
DSM-III-R diagnosis are dissimilar from those whose characteristics
originally prompted the development of this diagnostic category. It
remains to be determined if the syndromes identified by DSM definitions
will be shown to be related to causal factors or outcome. But grandiosity,
which has been considered to be the most reliable and discriminating aspect
of narcissism, and which can be identified by a sense of entitlement, is
commonly found in clinical practice, and narcissistic features often appear
to contribute to the development of a depressed mood.
Avoidant personality disorder
Main features
The main characteristic is an avoidance of social contact, despite a wish for
relationships, in the context of social discomfort, low self-esteem and a fear
of failure, rejection, criticism, ridicule or of the experience of strong
feelings.
Clinical origins
The term 'avoidant personality' was introduced by Millon (1969), and
'avoidant PD' appeared in the DSM-III in 1980. But there had been several
previous descriptions involving the avoidance of other people because of
various fears; Kretschmer (1925) described those whose 'life is composed of
A voidan t personality disorder 87
a chain of tragedies . . . (the person) behaves shyly, or timidly, or distrust-

fully . . . seeks as far as possible to avoid and deaden all stimulation from the
outside'. Horney (1945) described a 'detached type' in whom 'there is
intolerable strain in associating with people and solitude becomes primarily
a means of avoiding i t . . . the underlying principle here is never to become so
attached to anybody or anything so that he or it becomes indispensable . . .
better to have nothing matter much'. Subsequently, Horney (1950) gave a
further description:
on little or no provocation he (or she) feels that others look down on him, do not
take him seriously, do not care for his company, and, in fact, slight him. His self-
contempt ... makes him ... profoundly uncertain about the attitudes of others
toward him. Being unable to accept himself as he is, he cannot possibly believe that
others, knowing him with all his shortcomings, can accept him in a friendly or
appreciative spirit'.
Another theoretical perspective for such behaviour was provided by

Burnham, Gladstone & Gibson (1969) who described a 'need-fear
dilemma'; i.e.
he (or she) has an inordinate need for external structure and control... existence
depends upon his maintaining contact with objects (which include social relation-
ships). The very excessiveness of his need for objects also makes them inordinately
dangerous and fearsome since they can destroy him through abandonment. Hence,
he fears and distrusts them. One way to avert or alleviate the pain for his need-fear
dilemma is ... object avoidance
Although 'avoidant PD' was retained in DSM-III-R, and 'anxious
(avoidant) PD' is the equivalent in ICD-10, the clinical usefulness of a
separate category for these features is uncertain, as other features of PD are
often present and there is overlap between the features of avoidant PD,
schizoid PD and social phobia (Millon, 1991). However, it has been claimed
that avoidant PD can be distinguished from these other disorders by the
presence of several features including a wish for relationships, a wider range
of feared social and interpersonal situations and low self-esteem.
Main definitions
The DSM-III-R provides seven criteria for avoidant PD, four or more of
which are required for diagnosis, i.e.
is easily hurt by criticism or disapproval; has no close friends or confidants (or only
one) other thanfirst-degreerelatives; is unwilling to get involved with people unless
certain of being liked; avoids social or occupational activities that involve signifi-
cant interpersonal contact, e.g. refuses a promotion that will increase social
demands; is reticent in social situations because of a fear of saying something

inappropriate or foolish, or of being unable to answer a question; fears being
embarrassed by blushing, crying, or showing signs of anxiety in front of other
people; exaggerates the potential difficulties, physical dangers, or risks involved in
doing something ordinary but outside his or her usual routine, e.g. may cancel social
plans because she anticipates being exhausted by the effort of getting there.
However, the DSM-IV criteria show several changes: two of the criteria,
i.e. 'avoids occupational activities that involve significant interpersonal
contact, because of fears of criticism, disapproval, or rejection', and 'is
unwilling to get involved with people unless certain of being liked' resemble
their precursors, but the following new criteria are suggested:
shows restraint within intimate relationships because of the fear of being shamed or
ridiculed; is preoccupied with being criticized or rejected in social situations; is
inhibited in new interpersonal situations because of feelings of inadequacy; views
self as socially inept, personally unappealing, or inferior to others; is unusually
reluctant to take personal risks or to engage in any new activities because they may
prove embarrassing.
Four of the six ICD-10 criteria for anxious (avoidant) PD are similar to
those in DSM-III-R, i.e. 'excessive preoccupation with being criticized or
rejected in social situations; unwillingness to become involved with people
unless certain of being liked; restrictions in lifestyle because of need to have
physical security; and avoidance of social or occupational activities that
involve significant interpersonal contact because of fear of criticism,
disapproval, or rejection'. The two remaining ICD-10 criteria are: 'persist-
ent and pervasive feelings of tension and apprehension'; and 'belief that one
is socially inept, personally unappealing, or inferior to others'.
Clinical features
The timidity and widespread avoidance of an appropriate degree and
quality of social contact is associated with a chronic state of anxiety when
with others, ideas of poor self-worth, an expectation of humiliation or
rejection, and, sometimes, a belief that strong feelings are unacceptable
(Beck & Freeman, 1990). But at the same time there is a strong wish for
affection and acceptance in relationships.
Several studies of the prevalence of DSM-III and DSM-III-R avoidant
PD in various patient groups gave an average of around 10% (Millon,
1991), which confirms clinical impressions that these features are not
uncommon in clinical settings. Millon noted that while avoidant PD shares
Avoidant personality disorder 89
social withdrawal with schizoid (and schizotypal) PD, shares low self-
esteem with dependent PD, and shares social anxiety with the non-PD
diagnostic category of 'social phobia', the motivations and pattern of
symptoms may differ between the various disorders. For example, although
social anxiety is found with both avoidant PD and social phobia, the
anxiety with avoidant PD is usually related to a wider range of interper-
sonal situations, while social phobia may coexist with a number of
satisfying social relationships. But a person with avoidant PD has few or
any close relationships, despite the desire for them, because others are not
trusted without excessive reassurance. This has been summarized by Millon
(1991): 'avoidant PD is essentially a problem of relating to persons; social
phobia is largely a problem of performing in situations', and several studies
suggest that groups of patients can be identified in which the two disorders
can be distinguished, the person with avoidant PD having more severely
impaired social skills and being likely to have more severe anxiety and
depression (Turner et al.9 1986).
Low self-esteem in the context of dependent PD is usually due to a feeling
that he/she is incompetent rather than to a fear of social interaction. Indeed,
social contact makes the person with dependent PD feel more secure and
not more anxious as it does with avoidant PD. The dependent PD person is
anxious at the prospect of interpersonal loss, while someone with avoidant
PD can be afraid of becoming too closely involved (Trull, Widiger &
Frances, 1987).
The lack of social engagement in both avoidant PD and schizoid (and
schizotypal) PD appears to reflect different underlying thoughts and
feelings. It has been generally believed that avoidant PD is characterized by
interpersonal anxiety, low self-esteem and self-consciousness, but with a
desire for relationships, while schizoid (and schizotypal) PD is not related
to poor self-esteem and reflects relative indifference to achieving closeness
to others.
Various recommendations have been made to improve the DSM-III-R
criteria for avoidant PD (Millon, 1991), which have influenced the changes
in the DSM-IV criteria. The first DSM-III-R criterion ('is easily hurt by
criticism and disapproval') could be amended to 'frequently anticipates and
worries about being criticized or disapproved of in social situations'. Such a
change would make the revised criterion more specific for avoidant PD, as
the original version also reflects a feature of dependent PD. The second
criterion ('has no close friends or confidants ...') could be modified to: 'has
few friends despite the desire to relate to others'. Thefifthand sixth criteria
('is reticent in social situations ...' and 'fears being embarrassed') could be
condensed and reformulated as: 'development of intimate relationships is
inhibited (despite desire) owing to the fear of being foolish and ridiculed, or
being exposed and shamed'. It was also proposed that the seventh criterion
('exaggerates the potential difficulties ...') be deleted, as it was relatively
uncommon in relation to the other features of avoidant PD and was often
found with obsessive-compulsive disorder, dependent PD and schizotypal
PD. Finally, a new criterion was suggested: 'possesses low self-esteem
because he or she feels socially inept and/or unappealing'.
Beck & Freeman (1990) have described some hypothetical cognitions in
relation to avoidant PD that are based on characteristic statements in
clinical practice, i.e.
I am socially inept and undesirable; other people are superior to me and will reject or
think critically of me if they get to know me; I can't handle strong feelings; you'll
think I'm weak; if I give in to these feelings they will go on forever, if I ignore them, it
might get better some day; I'm unattractive, boring, stupid, a loser and pathetic; I
don't fit in'.

Depressive features, anxiety and frustration with oneself are common in the
context of restricted social relationships and impairment of occupational
functioning. Also, in common with most PDs, avoidant PD may be a risk
factor for substance abuse.
Social phobia and avoidant PD may co-occur, but the nature of any causal
relationships are uncertain. DSM-III-R claims that avoidant disorder of
childhood or adolescence, and disfiguring physical illness, predispose to the
development of avoidant PD.
Current status
Avoidant PD is a relatively new category dating from 1980, consisting of
phenomena which are relevant to some patients with PD in clinical
populations who can present with depressive features, anxiety or substance
abuse. Features of other PDs are often present and it can be difficult to
distinguish some features of avoidant PD from other disorders, particularly
schizoid PD, dependent PD and social phobia.
Dependen t personality disorder 91
Dependent personality disorder
Main features
The main characteristic of dependent PD is a pervasive pattern of depen-
dent and submissive behaviour in the context of an excessive need to be
taken care of, and fears of separation.
Clinical origins
A degree of dependency and attachment is part of the normal range of
adaptive human characteristics but excessive dependency can cause serious
problems. Some theories have linked adult patterns of dependency with
past events related to the infant's instinctive initial reliance on the mother,
while others have claimed that elements of dependency should be viewed as
a learned behaviour in a social context.
Early clinical descriptions were derived from psychoanalytic psychother-
apy literature; for instance Abraham (1924) wrote: 'some people are
dominated by the belief that there will always be some kind person - a
representative of the mother, of course - to care for them and to give them
everything they need. This optimistic belief condemns them to inactivity...
they make no kind of effort, and in some cases they even disdain to
undertake a bread-winning occupation'.
In the DSM-I, in 1952, the concepts of dependent and passive-aggressive
behaviours were linked, so that passive-dependent personality was defined
as a subtype of passive-aggressive personality, and related to helplessness
and indecisiveness, with a tendency to cling to others like a dependent child
to a supporting parent. No distinct category or description was provided by
DSM-II, but the DSM-III in 1980 described three characteristics of the
'passive-dependent' PD, i.e. such a person passively allows others to
assume responsibility for major areas of life because of inability to function
independently; subordinates his or her own needs to those persons upon
whom he or she depends in order to avoid any possibility of having to rely
on self; and lacks self-confidence. In 1987, the DSM-III-R modified the
term to 'dependent' PD, and expanded the description by providing nine
criteria, five or more of which were required for diagnosis. Most were
related to passivity and subordination, which had been features in DSM-
III, but criteria associated with anxious attachment and sensitivity to
criticism were added.
An equivalent category of 'asthenic' PD was provided by the ICD-9 in
1978. This denoted an individual who could be described as weak-willed,

widely compliant and falling in with the wishes of others, having little
capacity for enjoyment, and avoiding responsibility with a lack of self-
reliance. It was noted that such persons may persuade others to help them
but may drift down the social scale if unsupported.
Main definitions
The DSM-III-R requires at least five of the following criteria for a diagnosis
of dependent PD, in the context of a pervasive pattern of dependent and
submissive behaviour:
is unable to make everyday decisions without an excessive amount of advice or
reassurance from others; allows others to make most of his or her important
decisions, e.g. where to live, what job to take; agrees with people even when he or she
believes they are wrong, because of fear of being rejected; has difficulty initiating
projects or doing things on his or her own; volunteers to do things that are
unpleasant or demeaning in order to get other people to like him or her; feels
uncomfortable or helpless when alone or goes to great lengths to avoid being alone;
feels devastated or helpless when close relationships end; is frequently preoccupied
with fear of being abandoned; and is easily hurt by criticism or disapproval.
However, the last criterion is omitted in the DSM-IV criteria, and the
seventh criterion becomes: 'urgently seeks another relationship as a source
of care and support when a close relationship ends'.
The ICD-10 also includes a dependent PD category, with similar
features, characterized by three or more of the following:
encouraging or allowing others to make most of one's important life decisions;
subordination of one's own needs to those of others on whom one is dependent, and
undue compliance with their wishes; unwillingness to make even reasonable
demands on the people one depends on; feeling uncomfortable or helpless when
alone, because of exaggerated fears of inability to care for oneself; preoccupation
with fears of being abandoned by a person with whom one has a close relationship,
and of being left to care for oneself; and limited capacity to make everyday decisions
without an excessive amount of advice and reassurance from others.
Clinical features
The DSM-III-R's concept of dependent PD emphasizes dependent and
submissive behaviour. Such persons are passive, docile, self-effacing and
tend to put their faithful trust in those they rely on. Excessive advice is
sought together with reassurance that their actions are correct. Responsi-
bility is transferred to others, they find it difficult to initiate tasks, and there
is a constant fear of being rejected or abandoned, associated with clinging
behaviour.
Although the features of dependent PD show some overlap with other
syndromes, in particular with borderline, avoidant and histrionic PDs, even
overlapping aspects can show distinctive characteristics if carefully evalu-
ated (Hirschfeld, Shea & Weiss, 1991). For example, while fear of abandon-
ment can be found with both dependent and borderline PDs, in the latter it
produces anger and manipulative behaviour, in contrast to submissive and
clinging behaviour in dependent PD. With dependent PD, feeling uncom-
fortable or helpless when alone may be related to 'exaggerated fears of
being unable to care for himself or herself (according to the DSM-IV
criteria), while in borderline PD, this characteristic is not usually associated
with such thoughts. Also, compared with those with borderline PD, those
with dependent PD are more tolerant of being alone if they know they have
access to support and they do not give a history of such intense and unstable
relationships.
Dependent and avoidant PDs share feelings of inadequacy and sensi-
tivity to criticism, but the excessive need for attachment, fear of separation,
submission and clinging behaviour in dependent PD is in contrast to the
fear of humiliation and rejection, timidity and withdrawal in avoidant PD.
A need for reassurance and approval is a feature of both dependent and
histrionic PDs, but, in the former, reassurance is required because the
person doubts whether his or her actions are correct, while in histrionic PD
it is associated with the need for approval and praise as an end in itself.
Studies of the DSM-III-R criteria for dependent PD have shown that the
criteria: 'is easily hurt by criticism or disapproval', and 'feels devastated or
helpless when close relationships end', are non-specific, as they often occur
with other PDs and are not consistently associated with the other features
of dependent PD. (As noted, the former criterion has been omitted from the
DSM-IV). Also, endorsement rates for these two criteria were high in
patients with no diagnosis of PD. The criterion with the lowest endorse-
ment rate in those who qualified for the dependent PD diagnosis was:
'volunteers to do things that are unpleasant or demeaning in order to get
other people to like him or her' (Hirschfeld et ai, 1991).
Several investigations have separated the features of dependent PD into
two associated subgroups related to emotional reliance on significant
others (attachment), and a more general lack of social self-confidence
(general dependency). Livesley, Schroeder & Jackson (1990) described five
features of attachment, i.e. fear of loss of an attachment figure; need for
affection; need for proximity to the attachment figure; feelings of security
based on physical presence of the attachment figure; and strong protest at

separation from the attachment figure. These authors also provided five
features of general dependency, i.e. low self-esteem; need for advice and
reassurance; need for constant reassurance and approval; need for care and
support; and submissiveness.
Hirschfeld et al (1991) have reviewed the DSM-III-R criteria for
dependent PD; the first two criteria were supported, but it was suggested
that the third, i.e. 'agrees with people ...' should be rewritten as: 'has
difficulty expressing disagreement with others because of fear of their anger
or loss of support'. A modification was also suggested for the next criterion,
i.e. that 'has difficulty initiating projects ...' be developed by adding 'this is
due to a lack of self-confidence in judgement and not to a lack of motivation
or energy'. The final criterion, i.e. 'is easily hurt by criticism or disapproval',
was recommended for deletion, as it is commonly found in other PDs and
other psychiatric disorders, while the remaining criteria were reworded as
follows:
has difficulty independently initiating projects or doing things on his or her own.
This is due to a lack of self-confidence in judgement and not to a lack of motivation
or energy; goes to excessive lengths to obtain nurturance and support from others,
to the point of volunteering to do things that are unpleasant; feels uncomfortable or
helpless when alone, because of exaggerated fears of inability to care for him/
herself; when close relationships end, indiscriminatingly seeks another relationship
to provide nurturance and support; and is frequently preoccupied with fears of
being left to take care of himself/herself.
These recommendations have contributed to the changes in DSM-IV.

A depressive syndrome is commonly associated with dependent PD (Reich
& Noyes, 1987), and although definitions of dependent PD and depressive
disorders can overlap, with both involving lack of initiative, feeling helpless
and difficulty making decisions (Zuroff & Mongrain, 1987; Overholser,
Kabakoff & Norman, 1989), it appears that there may be a causal
relationship between dependent PD and some forms of depressive disorder.
Also, an individual with dependent PD may have an increased vulnerability
to the development of other disorders, such as panic attacks and phobias,
alcoholism and substance abuse. Individuals with dependent PD are
particularly vulnerable to worry and anxiety about being unsupported, so
that panic attacks may be precipitated by new responsibilities, while
phobias are reinforced by avoidance and any increased care from others.
Reich, Noyes & Troughton (19876) found that dependent PD was the most
frequent PD diagnosis in patients with panic disorder, particularly in the
subgroups with phobic avoidance, when the prevalence was about 40%.
Associations have also been claimed between dependent PD and somatic
complaints. Hill (1970), in an early study of 'passive-dependent' individ-
uals, found that all subjects had reported somatic complaints which had
often led to a great deal of medical attention. Often medical treatments were
considered as their main source of support, when their problems became
viewed in bodily rather than in psychological terms. For those with
dependent PD, seeking comfort from a variety of sources may also lead to a
higher incidence of obesity and tobacco dependence.
In a counselling or psychotherapeutic relationship, the engagement of
such individuals is readily achieved, but thereafter there is a tendency to
resist the clinician's efforts to encourage more autonomy and to discharge
them from treatment. This may lead to what appears to be a permanent
commitment, which may not be compatible with the demands of a clinical
service.
Associated features
DSM-III-R's dependent PD has been shown to commonly co-occur with
other PDs, in particular borderline PD, but also with avoidant, histrionic
and schizotypal PDs (Reich, 19906). The proportion of patients with PD
who had a diagnosis of dependent PD ranged from 7 to 47% in five studies
with a median of 20% (Hirschfeld et aL, 1991).
The possibility of prejudicial sex bias in the use of the dependent PD
diagnosis has aroused controversy, but this should not prevent the evalu-
ation of any genuine sex differences. While Kass, Spitzer & Williams (1983)
found that dependent PD (based on clinicians' DSM-III diagnoses with no
standardized assessment), was diagnosed over 2.5 times more often in
women, Reich (1987) did not confirm this finding using standardized
interview and self-report instruments. This limited evidence suggests that
clinicians' attitudes to the association between dependent PD and gender
requires further study. However, there may be genuine gender-related
differences in the prevalence of submissive behaviour resulting from both
genetic and environmental factors.
Finally, the DSM-III-R has claimed that chronic physical illness may
predispose to the development of dependent PD in children and
adolescents.
Current status
Dependent PD is a common disorder in psychiatric practice. It is associated
with depressive and anxiety disorders and usually co-occurs with other
PDs, in particular, borderline and avoidant PDs. Excessively dependent
behaviour in medical settings can impose inappropriate and unsustainable
demands on available resources.
Obsessive-compulsive personality disorder
Main features
The essential feature has been described as 'a preoccupation with perfec-
tionism, mental and interpersonal control, and orderliness at the expense of
flexibility, openness, and efficiency' (Pfohl & Blum, 1991).
Clinical origins
In the early part of this century, Sigmund Freud (1959) described certain
individuals with so-called anal character as 'orderly, parsimonious and
obstinate ... Orderly covers the notion of bodily cleanliness, as well as
conscientiousness in carrying out small duties and trustworthiness ...
Parsimony may (include) ... avarice; and obstinacy can go over into
defiance, to which rage and revengefulness are easily joined'.
Abraham (1921) provided further descriptions, which have been sum-
marized by Pfohl & Blum (1991), i.e. pleasure in indexing, compiling lists,
and arranging things symmetrically; superficial fastidiousness masking
disarray or lack of cleanliness underneath; pleasure in possessions with an
inability to throw away worn out or worthless objects; a tendency to
postpone every action and unproductive perseverance; preoccupation with
preserving correct social appearances; in close personal relationships
refuses to accommodate to others and expects compliance; produces
exaggerated criticism of others and insists on controlling interactions with
others; and a generally morose or surly attitude. Some of these features, i.e.
perfectionism, excessive scrupulousness, rigidity and hoarding, contri-
buted to Kahn's description (1928) of the 'anankastic' person, a term which
is retained in ICD-10.
In 1952, the DSM-I included a description of 'compulsive' personality
and a similar account was retained in DSM-II, although the category was
termed 'obsessive/compulsive personality', involving excessive concern
with conformity and adherence to standards of conscience. Such individ-
uals were rigid, overinhibited, overconscientious and unable to relax easily.
Subsequently, DSM-III provided a fuller description in the form of five
criteria, of which four were necessary for diagnosis, while the DSM-III-R
expanded the features to nine criteria, of which five were required. This
increase in criteria produced a greater variety in the syndrome, and it has
been estimated that DSM-III-R criteria would produce about twice as
many diagnoses of obsessive-compulsive PD compared with those of
DSM-III, in a given population (Pfohl & Blum, 1991).
Most of the features described by Freud and Abraham are reflected by
the DSM-III-R criteria, with the exception of a pattern of anger and
hostility. The clustering of the various features was investigated by Lazare
et al. (1966, 1970) using factor analysis of a self-report rating scale. One
factor involved orderliness, strong conscience, perseverance, obstinacy,
rigidity, rejection of others, parsimony and emotional constriction. This
and other studies using self-report inventories have not provided evidence
for marked anger as part of an obsessive-compulsive factor (Pfohl & Blum,
1991).
Main definitions
DSM-III-R requires at least five of the following nine criteria:
perfectionism that interferes with task completion; preoccupation with details,
rules, lists, order, organization, or schedules to the extent that the major point of the
activity is lost; unreasonable insistence that others submit to exactly his or her way
of doing things, or unreasonable reluctance to allow others to do things because of
the conviction that they will not do them correctly; excessive devotion to work and
productivity to the exclusion of leisure activities and friendships (not accounted for
by obvious economic necessity); indecisiveness - decision making is either avoided,
postponed, or protracted... (do not include if indecisiveness is due to excessive need
for advice or reassurance from others); overconscientiousness, scrupulousness and
inflexibility about matters of morality, ethics, or values (not accounted for by
cultural or religious identification); restricted expression of affection; lack of
generosity in giving time, money, or gifts when no personal gain is likely to result;
inability to discard worn-out or worthless objects even when they have no
sentimental value.
The DSM-IV criteria omit the features related to indecisiveness and
restricted expression of affection, and include a new criterion: 'rigidity and
stubbornness'.
The equivalent category in ICD-10 is 'anankastic' PD, which requires

three of eight criteria, in addition to the general criteria for PD. Five of the
criteria closely resemble DSM-III-R features, i.e.
preoccupation with details, rules, lists, order, organization or schedule; perfection-
ism that interferes with task completion; excessive conscientiousness, scrupulous-
ness, and undue preoccupation with productivity to the exclusion of pleasure and
interpersonal relationships; excessive pedantry and adherence to social conven-
tions; and unreasonable insistence by the patient that others submit to exactly his or
her way of doing things, or unreasonable reluctance to allow others to do things.
The three remaining are: 'feelings of excessive doubt and caution; rigidity
and stubbornness; and intrusion of insistent and unwelcome thoughts or
impulses'.
Clinical features
The central feature of perfectionism is often associated with the feeling 'I
should', together with deliberate, purposeful activity, reflecting a focussed,
stimulus-bound style of thinking (Beck & Freeman, 1990). Beck & Freeman
have itemized various hypothetical characteristic thoughts, e.g.
There are right and wrong behaviours, decisions and emotions; I must avoid
mistakes to be worthwhile; to make a mistake is to have failed; to make a mistake is
to be deserving of criticism; I must be perfectly in control of my environment as well
as of myself; loss of control is intolerable; if something is or may be dangerous, one
must be terribly upset by it; one is powerful enough to initiate or prevent the
occurrence of catastrophes by magical rituals or obsessional ruminations; if the
perfect course of action is not clear, it is better to do nothing; without my rules and
rituals, I'll collapse in an inert pile.
Such individuals may appear preoccupied with unimportant detail,

humourless, rigid, dogmatic, obstinate and indecisive. However, it must be
remembered that obsessional traits that are relatively moderate can be an
asset in many occupations, and that many 'normal' individuals with such
attributes would attract positive descriptions such as dependable, precise,
punctual, having high standards, persistent, stable and determined.
Persons with problematic obsessive-compulsive traits lack adaptability
to new situations and fail to take opportunities. Their perfectionism may
lead to occupational inefficiency because they get side-tracked by trivial
detail or postpone actions due to indecision. One patient would spend
several hours trying to compose a short letter, although he could manage to
complete other parts of his work that involved a structured routine. There
may be a lack of humour, of expressions of affection and of enjoyment.
Opinions tend to be based on moral principles and others are often judged
harshly. Meanness is associated with a lack of enjoyment in giving or
receiving. There is a tendency to worry and to be sensitive to criticism as
they expect to be judged as severely as they judge themselves. Although they
appear unemotional, this may hide resentment, anger and, perhaps,
aggressive fantasies. Indeed, such persons can experience a range of
emotions such as love and loyalty to a select group of friends or family,
concern, guilt, mourning and sadness (Akhtar, 1987).
The various DSM-III-R criteria for obsessive-compulsive PD have been
reviewed by Pfohl & Blum (1991). Two criteria, i.e. 'unreasonable insistence
that others submit to exactly his or her way of doing things, or unreason-
able reluctance to allow others to do things because of the conviction that
they will not do them correctly', and 'indecisiveness: decision making is
either avoided, postponed or protracted ...' have often been found in
persons without a diagnosis of obsessive-compulsive PD, which indicated
that the sensitivity of these criteria was low. Pfohl & Blum recommended
that the latter criterion should be excluded in DSM-IV and that the former
be modified to identify the specific feature of the trait, i.e. 'relucant to
delegate tasks or to work with others'. Also, possible new criteria were
suggested, namely, 'periods of rigid adherence to authority are punctuated
by great irritation or overt anger when authority figures are perceived as
bending the rules; argues stubbornly with others who have a different
political or philosophical viewpoint and considers opposing opinions
utterly without merit; critical and judgemental of others, e.g. prone to make
biting and sarcastic remarks'. As has been noted, the DSM-IV has been
influenced by these suggestions.

The features of obsessive-compulsive PD are common origins for presen-
tation to primary care physicians or to psychiatric services, as associated
social and occupational difficulties may cause anxiety, depressive disorders
or hypochondriasis.
Associated features
There have been many claims that obsessive-compulsive PD is related to
obsessive-compulsive disorder (OCD), which, despite the similarity of the
terms, can be distinguished by the presence of obsessions, with or without
compulsions. Obsessions and compulsions consist of distressing and time-
consuming thoughts and behaviour, which may significantly interfere with

the person's normal routine, occupational functioning, usual social activi-
ties or relationships with others. Obsessions are recurrent and persistent
ideas, thoughts, impulses, or images that are usually experienced, at least
initially, as intrusive and senseless; while compulsions are repetitive,
purposeful and intentional behaviours that are performed in response to an
obsession, or according to certain rules or in a stereotyped fashion
(American Psychiatric Association, 1987).
However, such an association is difficult to evaluate as the boundary
between the two disorders can be indistinct and arbitrary; not all patients
with OCD give a history of experiencing their phenomena as 'intrusive or
senseless', and the distinction between, for example, a marked preoccu-
pation with detailed organization and a compulsion involving a series of
actions performed in a particular order or in a stereotyped manner, may not
be possible. Pollack (1979,1987) has reviewed earlier (pre DSM-III) studies
and considered that 'clearly there is no necessary one-to-one relationship
between obsessional personality (i.e. obsessive-compulsive PD) and obses-
sional neurosis (OCD), despite the occasional findings that more obsessive-
compulsive neurotics than would be expected by chance show evidence of a
premorbid obsessional personality'. The studies in relation to DSM-III and
DSM-III-R have led to a similar conclusion and have been reviewed by
Pfohl & Blum (1991), who stated
the majority of patients with obsessive-compulsive disorder (OCD) do not meet
criteria for obsessive-compulsive PD. Among patients with OCD who do have a
personality disorder, several other personality disorders may be just as common or
more common than obsessive-compulsive PD. It is not possible to determine at this
time whether patients with OCD are more likely to meet criteria for obsessive-
compulsive PD than patients with other Axis I diagnoses.
The other PDs that were commonly found as co-occurring with OCD
were dependent, mixed, histrionic and avoidant PDs, and it has been noted
that obsessive-compulsive PD is common among patients with depressive
and anxiety disorders.
Pfohl & Blum (1991) have also reviewed the co-occurrence of other PDs
with obsessive-compulsive PD. Co-occurrence rates involving paranoid,
histrionic, borderline, narcissistic and avoidant PDs were 'unexpectedly
high'; for example, for three studies related to DSM-III-R criteria, the
average percentages for the co-occurrence of the above five PDs with
obsessive-compulsive PD were 55%, 24%, 29%, 40% and 57% respect-
ively. The rates for borderline PD are surprising, as aspects of the
personality structures of the two disorders, such as impulsiveness and
perfectionism, appear to be incompatible. Pfohl & Blum (1991) pointed out
that other PDs may be associated with some behaviours which are
characteristic of obsessive-compulsive PD, such as lack of generosity, but
that these may be related to different motivations. However, it is possible
that co-occurrence of obsessive-compulsive PD with some features of other
PDs will be shown to contradict current concepts of the relevant cognitive
processes in this disorder.
In a non-patient community study, DSM-IIFs 'compulsive PD' had a
prevalence of 1.7% in the sample of the population studied, and it was
claimed that this was associated with a vulnerability to the development of
anxiety disorders (Nestadt et ai, 1991).
In the DSM-III-R it is claimed that obsessive-compulsive PD is more
frequently diagnosed in males, and is apparently more common among
first-degree relatives of those with this disorder than among the general
population.
Current status
Obsessive-compulsive PD consists of features that have long been part of
clinical descriptions and are common origins of anxiety, some depressive
disorders and impaired occupational functioning in patient groups. The
relationship with obsessive-compulsive disorder is uncertain, despite the
similarity of the terms, and there appears to be a high degree of co-
occurrence with other PDs.
Passive-aggressive personality disorder
Main features
The most characteristic feature is a pattern of passive resistance to external
demands, with obstructive behaviour and poor performance both socially
and occupationally. Affected individuals avoid being assertive, although
they may be irritable, sulky or argumentative. They resent interference and
control, and express their resistance by such behaviour as delaying tactics,
providing a poor quality of work, forgetfulness, stubbornness, and sabo-
taging the efforts of others.
Clinical origins
The term 'passive-aggressive' was used in a 1949 US Joint Armed Services
Technical Bulletin, to describe soldiers who displayed features such as
passiveness, obstructionism or aggressive outbursts (Beck & Freeman,
1990) and the first edition of the DSM, in 1952, contained a 'passive-
aggressive' category divided into three subtypes: passive-aggressive, pas-

sive-dependent and aggressive. Using these DSM-I categories in a study of
400 outpatients, Whitman, Trosman & Koenig (1954) found 19% with the
passive-aggressive subtype. DSM-II, in 1968, gave this term the status of a
separate diagnosis and this remained the situation in DSM-III. However, it
has also been claimed that passive-aggressive behaviour should be viewed
as a transient reaction that can be shown by most people in situations when
they feel helpless. Although the early concepts of passive-aggressive PD
were narrowly based, i.e. on resistance to external demands, Millon (1981)
noted associated characteristics such as irritability, low frustration toler-
ance, pessimism, and sulking behaviour. Some of these are represented in
the augmented criteria found in the DSM-III-R, but the category is not
represented in the ICD-10.
Main definitions
In the DSM-III-R, five of the following nine criteria are required for
diagnosis of passive-aggressive PD:
procrastinates, i.e. puts off things that need to be done so that deadlines are not met;
becomes sulky, irritable, or argumentative when asked to do something he or she
does not want to do; seems to work deliberately slowly or to do a bad job on tasks
that he or she really does not want to do; protests, without justification, that others
make unreasonable demands on him or her; avoids obligations by claiming to have
'forgotten'; believes that he or she is doing a much better job than others think he or
she is doing; resents useful suggestions from others concerning how he or she could
be more productive; obstructs the efforts of others by failing to do his or her share of
the work; and unreasonably criticizes or scorns people in positions of authority.
The category is omitted from the DSM-IV except as a criteria set

'provided for further study', characterized by: 'passively resists fulfilling
routine social and occupational tasks; complains of being misunderstood
and unappreciated by others; is sullen and argumentative; unreasonably
criticizes and scorns authority; expresses envy and resentment toward those
apparently more fortunate; voices exaggerated and persistent complaints of
personal misfortune; and alternates between hostile defiance and
contrition'.
Clinical features
The starting point for such behaviour in passive-aggressive PD is resent-
ment towards the demands of others, in the context of a wish for autonomy,
negative attitudes based on the assumptions that others are trying to

control them and take advantage of them, and a belief that they are
misunderstood or unappreciated. This is accompanied by a failure to
engage in direct assertion, conflict or expressions of anger, often due to a
fear of disapproval or to concern that this will lead to more interference and
control by others. Such people are often sulky, irritable, or argumentative
when asked to do something, and protest to others how unreasonable the
demands are. They resent useful suggestions and unreasonably criticize and
blame those in authority as being arbitrary and unfair (American Psychi-
atric Association, 1987).
Such people find it difficult to accept that it might be better to state their
case and try and modify requests from others, and that their own behaviour
has contributed to social and occupational difficulties. They tend to be
dependent, lack self-confidence and are pessimistic about the future. There
is a tendency towards a negative view of most situations based on the idea
that others are trying, unreasonably, to control them. While a person with
paranoid PD may suspect hidden motives, with passive-aggressive PD
anything that disturbs his or her autonomy is considered to be unfair and
unreasonable. Beck & Freeman (1990) have described various characteris-
tic thoughts: 'How dare they tell me what to do; nobody gives me credit for
the work I do; people take advantage of me; rules are arbitrary and stifle me;
being direct with people could be dangerous; people take advantage of you
if you let them; and it doesn't matter what you do - nothing works out
anyway'.
If his or her behaviour is challenged, the individual may respond with
further obstructive acts, although there may be occasional overt anger. In
clinical settings there may be lateness, non-attendance, poor cooperation -
perhaps with inadequate answers to questions, and irritability.

Common complications include depressive disorders and alcohol abuse
(American Psychiatric Association, 1987), often in the setting of relation-
ship difficulties (Small et al, 1970). Patients may present because of
complaints and pressure from employers or spouses. It is likely that a
chronically pessimistic outlook makes such individuals liable to develop
secondary depressive symptoms. As autonomy is important, such patients
may resent having to answer questions and can appear irritated and
uncooperative. They may complain that nothing works out and that this is
the fault of others.
Associated features
There is some overlap with the features of avoidant, paranoid and
dependent PDs. The diagnosis of passive-aggressive PD usually co-occurs
with other PDs and may contribute to the causation of certain types of
depression and alcohol abuse.
Current status
Passive-aggressive PD is rarely found as a discrete diagnosis for patients in
psychiatric practice, but the features often appear to contribute to presen-
tations to medical services.
Self-defeating personality disorder
Main features
The DSM-III-R describes the essential feature as 'a pervasive pattern of
self-defeating behavior . . . The person may often avoid or undermine
pleasurable experiences, be drawn to situations or relationships in which he
or she will suffer, and prevent others from helping him or her'. But the
diagnosis is not made if self-defeating behaviours occur exclusively in
response to, or in anticipation of, being physically, sexually or psychologi-
cally abused, or only when the person is depressed. Also, this diagnosis has
been suggested 'for those who repeatedly place themselves in abusive,
detrimental and injurious situations despite the opportunity to avoid them'
(Widiger et al.91988). Therefore, this disorder involves active participation
on the patient's part in producing and maintaining adverse situations.
Clinical origins
In 1983, 'masochistic' PD was proposed as a category for future revisions of
the DSM in an attempt to incorporate behaviour that had been noted in the
literature of several disciplines, including psychoanalytic psychotherapy.
However, this attracted criticism and controversy, related to concern that
such a diagnosis could be misused, for example by blaming a woman for
remaining in an abusive relationship if no alternative can be readily
achieved (Kass et al., 1989). There has also been anxiety that this concept
could be used unfairly to discriminate against women when fitness to look
after children is at issue. Additional controversy about the relevance of
psychoanalytic concepts of female masochism, with the implication that the

person may derive unconscious pleasure from suffering, led to the name of
the proposed new category to be changed to 'self-defeating' PD in 1985
(Fiester, 1991).
In the DSM-III-R, 'self-defeating PD', was defined by eight criteria, and
was one of the 'proposed diagnostic categories needing further study' (it is
not included in DSM-IV). But only four of the original ten criteria for
masochistic PD had been retained in a similar form. The term 'masochism'
had been used by Krafft-Ebing (1901) to describe a psychosexual disorder
in which pain and suffering coexisted with sexual gratification, while Freud
(1961) extended the concept to include self-imposed suffering, i.e. persistent
self-reproach and a pattern of contributing to unnecessary failures and
accidents, perhaps in the context of being exploited. Other chronic features
that have been claimed to be related to these attributes have included
unfounded or exaggerated feelings of suffering or of exploitation; frequent
complaints and demands about being mistreated; passive self-abasement;
an attitude of poor self-worth; a tendency to provoke rejecting responses; to
be rejecting of help; feeling ethically or morally superior; responding to
positive events by a negative reaction; rejecting opportunities for pleasure;
pessimism; self-punishing, self-destructive behaviour, involving self-
inflicted pain or adversity; sabotaging goals; a lack of interest in those who
treat him/her well; and sacrificial or martyred behaviour (Kass et ah, 1989).
Two self-defeating PD criteria (i.e. 'rejects or renders ineffective the
attempts of others to help him or her', and 'following positive personal
events, e.g. new achievement, responds with depression, guilt, or a behav-
iour that produces pain, e.g. an accident') can be reflected by specific
doctor-patient problems, namely, the patient sabotaging the clinician's
efforts and a 'negative therapeutic reaction', in which any progress may be
cancelled out by its subsequent negative effect on the patient.
The self-defeating PD criteria provide a variety of behaviours to be
evaluated, and Spitzer et al. (1989) invited each of 2000 clinicians to rate one
of their patients with self-defeating PD behaviour and another with other
PD features. Fifty-one per cent of respondents thought there was a need for
the proposed self-defeating PD category, and 222 patients who were
selected as examples of those who show some self-defeating PD criteria
were rated and compared with 222 patients with other PDs. Although the
eight self-defeating PD criteria were often found in the control group, all
were considerably more common in the self-defeating PD, in particular the
first criterion, i.e. 'chooses people and situations that lead to disappoint-
ment, failure or mistreatment...'. In another study of over 1000 patients
with PD (each clinician was asked to rate the first patient who came to mind
with PD), self-defeating PD was found in 42% of females and 28% of males
(Spitzer e/tf/., 1989). But it was the sole diagnosis in just 4% and was usually
associated with borderline and dependent PDs. Spitzer et al. (1989) also
reported on 85 patients 'who had at least one DSM-III-R PD' assessed by a
structured interview; the diagnosis of self-defeating PD was made in 21%.
Two further studies report on patients evaluated by self-defeating PD
criteria: 18% of 76 cocaine abusers who were assessed by a structured
interview received a diagnosis of self-defeating PD (Kleinman et al., 1990),
while in another study of 110 outpatients rated by clinicians, self-defeating
PD was a diagnosis in 14%, and one-third of these patients had no other PD
(Nurnberg et al., 1991). Although several other studies have reported self-
defeating psychopathology, these were related to earlier criteria for maso-
chistic PD (Kass, MacKinnon & Spitzer, 1986; Reich, 1990a).
Main definitions
Self-defeating PD is one of the 'proposed diagnostic categories needing
further study' in DSM-III-R, but is not found in DSM-IV or in ICD-10.
There are two exclusion criteria, namely that the disorder does not occur
exclusively in response to, or in anticipation of, being physically, sexually or
psychologically abused, and that it does not occur only when the person is
depressed. Five of the following eight criteria are required:
chooses people and situations that lead to disappointment, failure, or mistreatment

even when better options are clearly available; rejects or renders ineffective the
attempts of others to help him or her; following positive personal events (e.g. new
achievement), responds with depression, guilt, or a behaviour that produces pain
(e.g. an accident); incites angry or rejecting responses from others and then feels
hurt, defeated, or humiliated (e.g. makes fun of spouse in public, provoking an
angry retort, then feels devastated); rejects opportunities for pleasure, or is reluctant
to acknowledge enjoying himself or herself (despite having adequate social skills
and the capacity for pleasure); fails to accomplish tasks crucial to his or her personal
objectives despite demonstrated ability to do so, e.g. helps fellow students write
papers, but is unable to write his or her own; is uninterested in or rejects people who
consistently treat him or her well, e.g. is unattracted to caring sexual partners; and
engages in excessive self-sacrifice that is unsolicited by the intended recipients of the
sacrifice.
Clinical features
DSM-III-R describes how a person with self-defeating PD can repeatedly
enter relationships with others (or place him or herself in situations), that
are self-defeating and have adverse consequences, even when better options
are clearly available. For example, a woman may appear to choose to
repeatedly enter relationshps with men who have alcohol dependence, or a
man with employment skills may persist with jobs where these are not used.
Reasonable offers of assistance from others, including medical services, are
rejected, for example by not following-up on an agreed treatment plan. The
person's reaction to positive events, including the relief of symptoms, or
encouragement, may be a worsening of a depressed mood, perhaps with
feelings of guilt. However, such persons usually do not appear to be
engaged in deliberate sabotage, as they do not seem to be fully aware of the
motivations associated with their actions. An adverse event such as an
'accident', or losing something, may seem to others to follow a clear
pattern, but the individual is not necessarily aware of deliberate intent.
Often, a person's behaviour, such as making unreasonable demands from
medical services, invites anger and rejection, and this produces a feeling of
hurt and humiliation. Opportunities for pleasure may be avoided, or
enjoyment is denied in situations where this would be expected. He or she
may seem to avoid taking available opportunities, so that any achievements
are not in keeping with his or her potential, while people who treat the
person well may be experienced as uninteresting, so that friends or partners
are chosen who are bound to disappoint. There can be a tendency to do
things for others that involve excessive self-sacrifice, but this does not make
the person feel better and often makes the recipients feel guilty and
rejecting.
It has been noted that several criteria for this diagnosis require a high
degree of judgement for their application, for example, 'engages in excess-
ive self-sacrifice...', 'rejects people who consistently treat him/her well...',
and '... when better options are clearly available' (Fiester, 1991). The
evaluation of the latter is particularly problematic in a setting of poverty,
substance abuse, crime and violence.
Self-defeating PD should be distinguished from self-defeating behaviour
that is found only in the context of physical, sexual or psychological abuse,
as such behaviour can sometimes be viewed as a coping strategy for a
woman to avoid the further violence that would result if she asserted herself
or finished the relationship. Also, the diagnosis of self-defeating PD should
not be made if the behaviour is only found in the presence of a depressive
disorder, although this distinction can be difficult to make.
It has been suggested that to qualify as a 'diagnosis', a syndrome must be
capable of defining a group of patients in whom the features are relatively
discrete (or predominant) and that this group should be distinguished from
other groups, taking course and outcome into account (Robins & Guze,
1970). Evidence for the justification of self-defeating PD as a diagnostic

category on this basis is lacking, as few patients with a discrete self-
defeating PD have been described, and positive self-defeating PD criteria
have been commonly found in association with other PDs (Spitzer et al.,
1989). However, the concept of self-defeating PD as a syndrome could still
have a reality, even in a population of patients with co-occurring PDs, if
there is a greater correlation (or association) between its component items,
compared with the degree of correlation (or association) of its component
items with items from other PD syndromes. However, a study using a self-
report questionnaire based on DSM-III-R PDs indicated that the criteria
for self-defeating PD had a greater correlation with other PD syndromes
than with the self-defeating PD syndrome itself (Dowson, 1994).

As with many PD diagnoses, self-defeating PD features are seldom discrete,
but they can contribute to the development of some depressive disorders
and other common psychiatric symptoms.
Associated features
The limited evidence suggests that the diagnosis of self-defeating PD is
made more frequently in women than men in clinical samples (Fiester,
1991), and in view of concerns about the possible potential for the misuse of
this category, the reasons for the apparent sex difference in the frequency of
its use are of particular importance. If such a difference exists, it may be due
to three possible causes or their combinations (Sprock, Blashfield & Smith,
1990). The first involves a 'gender bias' on the part of the clinician,
reflecting socially-determined ideas about which types of behaviour are
healthy or pathological for males versus females. The second involves
gender bias in the way clinicians apply the criteria, while the third reflects
real differences in the prevalences of the disorders in males and females, due
to different susceptibilities, based on biological, genetic, social or environ-
mental factors. Fiester's review of the limited data (1991) concluded that
there was no clear evidence of gender bias in the application of the criteria,
although further studies are required.
The nature of an association between self-defeating PD and depressive
disorders has also been controversial, and it has been suggested that self-
defeating PD should be considered to be related to the spectrum of
depressive disorders, rather than to the other PDs. A concept of a
'depressive-masochistic' personality has been a feature of the psychoanaly-

tic literature, and it can be difficult to distinguish self-defeating PD features
due to personality traits from those secondary to chronic depressive
disorder.
Additional associations of self-defeating PD have also been claimed,
namely with other co-occurring PDs (in particular with borderline, avoid-
ant and dependent PDs), increased prevalence in the relatives of subjects
with the disorder, and an increased frequency of an history of physical,
sexual or psychological abuse in childhood (Fiester, 1991).
Evidence for concurrent or predictive validity for self-defeating PD is
sparse, but McCann et al. (1991), in a study of bulimic patients that
included PD assessment by a structured interview, claimed that those who
showed self-induced vomiting scored higher on the narcissistic, self-
defeating and borderline scales.
Current status
The features of self-defeating PD have been found in clinical descriptions
for several decades, and are present in many patients with PD in psychiatric
practice, although they usually co-occur with other PDs. While they should
be evaluated in a PD assessment, the provision of a distinct diagnostic
category cannot be justified as there is insufficient evidence that it describes
features that are relatively discrete or that the criteria are associated with
each other more than with the features of other PDs. Despite this, clinicians
in psychiatric practice are aware that, for a few patients who are among the
most difficult to help, their self-defeating behaviour appears to be the most
disabling aspect of their PD psychopathology.
Sadistic personality disorder
Main features
The DSM-III-R describes the essential feature as 'a pervasive pattern of
cruel, demeaning, and aggressive behaviour directed toward other people,
beginning by early adulthood. The sadistic behaviour is often evident both
in social relationships (particularly with family members) and at work (with
subordinates), but seldom is displayed in contacts with people in positions
of authority or higher status'. Also, T h e diagnosis is not made if the sadistic
behaviour has been directed towards only one person (e.g. a spouse) or has
been only for the purpose of sexual arousal (as in sexual sadism).'
Clinical origins
Sadism was a term originally used by Krafft-Ebing (1901) to describe the
desire to inflict pain upon a person who was the sexual 'object'. But while
Freud (1957) also considered that aggressiveness or a desire to subjugate
was associated with the biological substrate of sexuality in males, he
believed that there were two aspects of sadism, a generalized aggressive and
violent attitude towards the sexual object, and a more specific sexual
sadism, involving sexual satisfaction that is dependent on the humiliation
and mistreatment of the sexual partner. Subsequently, the psychoanalytic
literature has tended to consider that sadism is related to masochism (which
is the desire to have pain inflicted by the sexual object), and that both are
related to biological sexual instincts (Fiester & Gay, 1991). These ideas have
been extended beyond sexual relationships; for example Horney (1950) has
described the 'vindictive sadist' who shows a repeated tendency to domi-
nate, deprecate, humiliate, blame and cruelly criticize others, but not in the
context of sexual gratification.
As a category in a major classification, 'sadistic PD' first appeared in the
DSM-III-R in 1987, but only as one of the 'proposed diagnostic categories
needing further study', together with self-defeating PD. Both these pro-
posed disorders were controversial; some considered that self-defeating PD
reflected culturally-determined ideas of how women should behave, and
that sadistic PD had been included to placate such criticisms by 'balancing
the ticket', as cruelty, anger and aggression were supposedly associated
with a culturally-sanctioned male role. The DSM-III-R concept of sadistic
PD specifies that the associated behaviour has not been aimed solely at the
purpose of sexual arousal, and Fiester & Gay (1991) have reviewed several
studies that suggest that sexual sadomasochists generally limit sadistic and
masochistic behaviour to their sexual encounters, and are rarely involved
with more generalized sadistic interactions.
One of the main reasons why sadistic PD was included in DSM-III-R was
that several forensic psychiatrists believed that there was a need for such a
category, particularly in forensic settings, in relation to people who showed
a pattern of cruel and aggressive behaviour towards others but whose
behaviour was not adequately encompassed by the antisocial PD diagnosis.
Such individuals showed pronounced aggressive and domineering behav-
iour, in the relative absence of a history of childhood conduct disorder and
adult criminality.
The addition of this new proposed diagnostic category to DSM-III-R
gave rise to concern that it might be used to mitigate or excuse violent crime,
in particular repeated violence and sexual abuse to a child, spouse or
partner. As Spitzer et al. (1991) noted: 'the medicalization of evil deeds

becomes an avenue of excuses'. But many other diagnoses have the
potential to be exploited to interfere with the aims of a criminal justice
system and such a potential should not be an argument for the suppression
of a valid diagnostic category. As Spitzer et al. (1991) have pointed out, it is
possible to have specific laws precluding the use of PD diagnoses as
mitigating factors in criminal proceedings, and such laws have been enacted
in several states in the USA in respect of the diagnosis of antisocial PD. But
although it is understandable, and perhaps inevitable, that a society has
very different attitudes to a murderer with a schizophrenic illness involving
florid delusions, compared to a murderer with antisocial or sadistic PD, a
clear distinction between illness and normality in relation to this example
cannot be justified from a biological perspective, as genetic and other
biological factors are major causal variables in respect of both schizophre-
nia and PDs. It can be argued that a logical approach to the detention and
supervision of dangerous individuals with PD would not be based mainly
on concepts of blame and responsibility, and that the risk of the continua-
tion of antisocial behaviour should be given more consideration despite the
difficulties in prediction.
Main definition
DSM-III-R includes sadistic PD as one of the 'proposed diagnostic
categories needing further study' but it is not included in DSM-IV. The
behaviour must have been directed towards more than one person and not
solely for the purpose of sexual arousal. At least four of the following are
required:
has used physical cruelty or violence for the purpose of establishing dominance in a
relationship ...; humiliates or demeans people in the presence of others; has treated
or disciplined someone under his or her control unusually harshly, e.g. a child,
student, prisoner, or patient; is amused by, or takes pleasure in, the psychological or
physical suffering of others (including animals); has lied for the purpose of harming
or inflicting pain on others ...; gets other people to do what he or she wants by
frightening them (through intimidation or even terror); restricts the autonomy of
people with whom he or she has a close relationship, e.g. will not let spouse leave the
house unaccompanied ...; and is fascinated by violence, weapons, martial arts,
injury, or torture.
Clinical features
There is generally a pattern of cruelty and/or violence which establishes
dominance in relationships, and escalating violence may result from the
victim trying to assert him or herself, but not all persons with this disorder
are physically violent. Various behaviours, including verbal intimidation,
show a lack of respect for others, although this pattern may only involve
those people who can be subjugated or dominated. Such victims are often
humiliated in the presence of others, or, in the case of children (or others
under his or her control) discipline may be excessive. For example, a teacher
may insist that a student spends many hours in detention at school for a
minor offence. The person often restricts the autonomy of those who are
dominated, for example, by not allowing family members out of the house
for social occasions, and gets others to do what he or she wants by
frightening them. In severe cases, the person may enjoy producing physical
or psychological pain or suffering in others or in animals.
Fiester & Gay (1991) have reported on 235 adults who had been accused
of child abuse and found that 12 (5%) met the criteria for sadistic PD. These
individuals were functioning well in some areas, as eight were in regular
work and many were capable of remorse and sadness. Eight were men and
only two were considered to have another DSM-III-R PD (narcissistic and
antisocial). Nine gave histories of childhood physical abuse, often with
sexual or emotional abuse and neglect. In eight there was a history of early
loss of a parent or sibling by death or separation. The behaviour of this
group included demeaning verbal abuse, threats to kill, beating, murder
and abuse of spouse. There was a reluctance to acknowledge problems and
to receive help or treatment unless it was imposed.
The other main source of data is a survey of forensic psychiatrists by
Spitzer et al. (1991). Almost all cases described were male; there was a
frequent history of childhood abuse (emotional, physical and sexual) and
parental loss in 52%. Also, there was evidence that the diagnostic criteria
for sadistic PD were relatively sensitive and specific. Sensitivity is the
frequency of individual criteria in cases of the disorder, and ranged from
65% to 94%, while specificity (the frequency of the absence of the criteria in
non-cases) varied between 93% and 99%. Forty-seven per cent had co-
occurring narcissistic PD and 67% had co-occurring antisocial PD, but
20% had neither. (Other PDs were not evaluated.) Twenty-seven per cent
had a depressive disorder and 6 1 % a history of substance abuse. It was
concluded that the disorder may not be rare in forensic settings and that the
prevalences of a history of emotional, physical and sexual abuse (90%, 76%
and 4 1 % respectively), were much higher than in general population
studies, indicating an intergenerational transmission of violence.
Several other studies, reviewed by Fiester & Gay (1991), gave varying
prevalence rates (2.5-33%) in various populations, as assessed by different
methods. The highest prevalence, 33%, was in a group of 21 sex offenders,
mostly in prison. However, there are particular difficulties in assessing this
disorder unless there is independent information, as the person must admit
to various behaviours that are especially socially unacceptable.
A few studies have compared the performance of the various DSM-III-R
criteria in relation to the sadistic PD diagnosis (Fiester & Gay, 1991); the
sixth criterion ('gets other people to do what he or she wants by frightening
them') had the highest correlation with the diagnosis, while the lowest
correlations were with criteria 3 and 7 ('has treated or disciplined someone
under his or her control unusually harshly ...' and 'restricts the autonomy
of people ...').

In general, individuals with sadistic PD do not recognize their behaviour as
abnormal or problematic, and do not present to medical services unless this
is instigated following a court appearance, perhaps after being charged with
child or spouse abuse. But such people may also come to medical attention
because of associated substance use disorders, self-injury or pressure from
family and employers.
Associated features
Several studies have noted that sadistic PD often co-occurs with other PDs,
in particular narcissistic and antisocial PDs, but also with other PDs
(Fiester & Gay, 1991). In a study of 12 individuals with sadistic PD, two had
alcohol dependence, four a history of mixed substance abuse, and one had
current substance abuse. Spitzer et ah (1991) found that 27% of 113
subjects had a co-occurring depressive syndrome and 61% had substance
abuse. The disorder is far more common in males than females.
Current status
It is uncertain whether it will be useful to separate sadistic PD from
antisocial PD. But although PDs with predominant and severe features of
sadistic PD are rare presentations to the general psychiatrist, these attri-
butes are more common in forensic practice. It is possible that this category
represents a PD with a characteristic childhood history of abuse and
parental loss, but which, in comparison with antisocial PD, has a history of
less childhood conduct disorder and adult criminality outside the context of
the sadistic behaviour. A better understanding of any causal relationships

between adverse childhood events and subsequent sadistic PD may have
implications for the development of preventative interventions when child
abuse is identified.
Depressive personality (or personality disorder) and allied disorders
Main features
The main features of'depressive personality' (which is not a category that is
found in the DSM-III-R or ICD-10) are 'excessive negative and pessimistic
beliefs about oneself and other people' that are apparent from early adult
life (Hirschfeld & Shea, 1992). These are associated with pervasive dissatis-
faction, low self-esteem, feelings of inadequacy and worthlessness, and, for
most of the time, a persistent unhappiness. Also, such individuals appear
critical and negativistic. Depressive PD is a criteria set 'provided for further
study' in DSM-IV.
Clinical origins
Kraepelin (1921) described a depressive temperament, which he considered
was mainly caused by genetic predisposition, characterized by 'a perma-
nent, gloomy, emotional stress in all the experiences of life', and Phillips et
al. (1990) have noted that this concept emphasized 'persistent gloominess,
joylessness, anxiety and a predominantly depressed and despairing mood'.
Such individuals were also 'serious, burdened, guilt-ridden, self-reproach-
ing, clinging, and lacking in self-confidence'. Schneider (1959) termed a
similar disorder as 'depressive psychopathy', which also involved a persist-
ent sense of gloom together with the following characteristics: gloomy,
pessimistic, serious and incapable of enjoyment or relaxation; quiet;
sceptical; worrying; duty-bound; and self-doubting.
Subsequently, Akiskal's work (Akiskal, 1983; Akiskal, Hirschfeld &
Yerevanian, 1983), on patients with both PD and chronic depressive
disorders, led him to subdivide such 'characterologic depressions' into two
sub-types: 'sub-affective dysthymia' and 'character-spectrum disorder',
and it is the former that appears to correspond closely to the earlier
descriptions.
The psychoanalytic literature has provided a different perspective, in
which aspects of personality are considered to be determined by environ-
mental interactions during development. These traits then predispose the
Depressive personality (or personality disorder) 115
individual to some types of depressive disorder. It has been suggested that

relevant experiences include generally adverse circumstances, poor parent-
ing, and childhood loss or bereavement, and that relevant traits include
dependency, obsessionality, low self-esteem with helplessness, guilt, and
difficulty expressing anger (Phillips et al., 1990). However, theories of
environmental and genetic influences on depressive personality features are
not incompatible, as PDs are multidetermined, involving biological, cul-
tural and social variables.
The concept of depressive personality was included in the heterogeneous
category of'Depressive neurosis' in DSM-II in 1968, and Spitzer, Endicott
& Woodruff (1977) defined the subtype of depressive neurosis that corres-
ponded to depressive personality, i.e. 'individuals who for at least the past 2
years have been bothered by depressed mood much of the time, with
periods of normal mood lasting from a few hours to a few days . . . there is
usually no clear onset ...'. It should be noted that antidepressants were
increasingly being shown to be useful in the treatment of many patients with
chronic, relatively mild depressions, so that interest was developing in the
subtypes of this heterogeneous group of disorders (Klerman, Endicott &
Spitzer, 1979).
The consideration of the relationship of personality, PDs and depressive
disorders is complicated by the heterogeneity of depressive disorders in
respect of both clinical features and causation. However, various classifica-
tions of depressive syndromes have attempted to obtain aetiologically
homogeneous patient groups.
When considering this confusing area it can be helpful to bear in mind
that there appear to be three main interacting categories of aetiological
factors. The first is 'endogenous' (or 'biological') aetiology related to
episodes of depression, which involves hypothetical abnormal brain func-
tioning, for example, related to a genetically-determined vulnerability. The
resulting depression may appear 'out of the blue', but even if the disorder is
precipitated by life-events, the nature or severity of the symptoms may be
clearly out of the context of the patient's personality and environment,
corresponding to the lay person's concept of disease. Of course, the
definition of 'abnormal brain function' is a complex process, akin to the
definition of hypertension. But, although an arbitrary line between nor-
mality and abnormality may have to be drawn at one point in the spectrum
of hypertension, some cases of hypertension at the most abnormal end of
the spectrum are clearly due to qualitatively distinct abnormal biological
function, and there is good evidence that the same can be said for some
cases of depression. The second category of aetiological factor for depres-
sive disorders is past and present adverse environmental situations, which

often interact with the third category, which is PDs.
All three categories may contribute to a depressive disorder in some
patients. For example, certain PDs, such as borderline PD, have been
shown to be associated with genetic factors (Kendler et al., 1981; Baron et
al, 1985; Cadoret et al, 1985; Gunderson & Elliott, 1985), and, in the case
of a patient with borderline PD, environmental problems that may be the
consequences of the maladaptive behaviour in this disorder can contribute
to a depressed mood. This appears, at least in part, understandable as a
response to environmental adversity.
Such a three-category model of the causation of depressive syndromes
must, of course, be oversimplistic, but provides a framework for clinical
practice. A particular syndrome of depression may be considered to be
derived from one or a mixture of two or three of these aetiological
categories, although one may predominate.
Most categorical systems of classification of depression are based on
symptoms and these can sometimes correlate with aetiology, as 'endoge-
nous' aetiology related to episodes of depression appears to be associated
with several specified clinical features. These comprise the syndrome of so-
called 'endogenous depression', although these phenomena are not invaria-
bly present as markers of endogenous aetiology. The identification of
biological markers of endogenous aetiology, such as time from sleep onset
to the first period of rapid eye movements, and the dexamethasone
suppression test, have been used to improve patient selection for research
studies, and further advances in this area are to be expected (Gunderson &
Phillips, 1991).
One of the most enduring classifications of the features of depression
involves just two categories, 'endogenous' (or psychotic) and 'neurotic' (or
reactive) (Young et ah, 1986), and, despite its limitations, the clinical
usefulness of this approach has stood the test of time. However, these terms
are unsatisfactory. This classification is based on clinical description, and
although the features of the endogenous syndrome can indicate that there is
endogenous aetiology related to an episode of depression, endogenous
aetiology does not invariably present with the specified symptoms. (This is
evident from the features of some patients with regular and frequent mood
changes in the context of bipolar manic-depressive disorder.) The term
'psychotic', if it is used as an alternative to endogenous, is also misleading
because it infers the presence of delusions or hallucinations; however these
are not always part of an endogenous syndrome, which rarely presents with
all or even most of the various specified 'endogenous' features. Finally, the
term 'reactive', as an alternative to 'neurotic', makes an aetiological claim

that is not always justified; while the nature and severity of 'neurotic
depression' can often seem understandable (or meaningful) as a reaction of
that individual to events, endogenous aetiology may also contribute to the
development of a syndrome of 'neurotic depression', which would be
reflected by, for example, a degree of severity or chronicity that seems
incompletely understandable on the basis of personality and environment.
Many studies have examined the clinical features associated with the
endogenous syndrome, which can be identified relatively reliably in patients
with bipolar manic-depressive disorder; these include marked agitation or
psychomotor retardation with slowed thoughts and actions; absence of
mood reactivity to pleasant circumstances with a loss of interest in
pleasurable activity; self-reproach; early morning wakening; marked loss of
appetite and weight; a distinct quality of mood that is unlike normal
unhappiness; mood-congruent hallucinations and delusions; and repeated
diurnal variation in mood with morning worsening (Nelson & Charney,
1981). (The last two features seem to be relatively specific markers of
endogenous aetiology related to episodes of depression, while features such
as early morning wakening are much less specific.) However, the number
and severity of these features vary considerably and Parker et al. (1994)
have provided evidence that signs of psychomotor disturbance are also
relatively specific in defining a depression with significant endogenous
aetiology. These signs usually involve facial and bodily immobility, delay in
responding verbally, and impaired spontaneity of talk. In patients with
features of the endogenous syndrome there is less likely to be an identifiable
precipitating stress compared with those with a 'neurotic' syndrome, but
there are exceptions, as it seems that life events can sometimes precipitate
the biological processes underlying the endogenous syndrome.
'Neurotic' depression encompasses an heterogeneous group of clinical
presentations that tend to be less severe than endogenous depressions, but
again there are exceptions. Such syndromes consist of a varying and
inadequately defined number of a long list of symptoms, including anxiety,
phobias and somatic concomitants of anxiety, in addition to depressed
mood. The presentations of most patients with a depressive disorder are
within the category of neurotic depression, but progress has been made in
identifying clinically valid subtypes that reflect differences in aetiology. For
example, Akiskal (1983) identified four chronic subtypes of neurotic
depression; the first, 'primary depression with residual chronicity', occurs
in an individual with previously normal mood and appears to be a
syndrome with endogenous aetiology that usually responds to antidepres-
Table 2.2 Some categories of depressive disorders
1. Based on presence or absence of a previous history of mania:

Unipolar depression*
Bipolar depression (i.e. with a history of mania)
2. DSM-III-R classification:
(DSM-IV categories in brackets)
Bipolar disorder (Bipolar I disorder) - mixed or depressed
Major depression (Major depressive disorder)
Dysthymia* (Dysthymic disorder)
Cyclothymia (Cyclothymic disorder)
3. ICD-10 classification:
Bipolar affective disorder
Depressive episode (or recurrent depressive disorder)
Mild
Moderate
Severe without psychotic symptoms
Severe with psychotic symptoms
Persistent mood disorders
Cyclothymia
Dysthymia*
Note: *Heterogeneous categories each of which can be considered to include the

concept of depressive personality.
sants. The second subtype, 'subaffective dysthymia', is believed to have

similar endogenous aetiology, but can appear to be a PD as it begins before
the age of 25 and the patient may have seemed chronically depressed
throughout adult life. Such people may present as quiet, passive, gloomy,
self-critical and conscientious. It is claimed that this disorder also responds
to antidepressant medication. The third sub-type is 'chronic secondary
dysphoria', and is a chronic depressive disorder that seems understandable
as a reaction to some other chronic psychiatric disorder or medical
condition such as severe deformity (Winokur, Black & Nasrallah, 1988).
Akiskal suggested that some of these patients respond to a monoamine
oxidase inhibitor. Finally, there are the 'character-spectrum disorders', in
which the patient is chronically depressed since childhood or adolescence in
the context of various types of severe PD, for example with dependent,
borderline or sociopathic traits, which are often associated with self-injury
and antisocial behaviour. Akiskal claims that, in contrast to those with
'subaffective dysthymia', such patients do not usually respond significantly
to antidepressant medication.
Other classifications of depression have been based on the presence or
absence of another non-affective psychiatric disorder, of delusions (Coryell
et al., 1994), or of a history of mania, and the latter provides perhaps the
simplest and most robust classification. This involves the division of
depressive disorders into the two subtypes of unipolar and bipolar
depression (Table 2.2). There is evidence that genetic predisposition is
required for the development of disorders in the bipolar group, while
unipolar depression 'is usually considered to encompass a variety of
illnesses with different aetiologies and treatments; it is not clear whether
hypotheses relating to personality and depression are relevant to all forms
of depression or only to specific subtypes. Therefore, different findings from
different studies of depressed patients may be hard to integrate because of
diagnostic heterogeneity' (Hirschfeld & Shea, 1992). Table 2.2 also shows
the main categories of depressive disorders in the major international
classifications; the concept of depressive personality is subsumed in the
broader categories of 'dysthymias' in both DSM-III-R and ICD-10. In
DSM-III-R, another important category is 'major depression', partly
defined on the basis of increased severity and episodic course, when
compared with dysthymia, but Kendler et al. (1993c), in a twin study, found
that, in women, there is an association between 'neuroticism' and liability
to major depression and it is clear that PDs can be of aetiological
importance in some cases of 'major depression'. DSM-III-R's dysthymia
usually begins in childhood, adolescence, or early adult life and precedes
any superimposed major depressive episodes, if any, by years (Akiskal,
1991). (The return to a mild or moderate chronic dysthymia following
recovery from a superimposed major depressive episode has been termed a
'double-depressive' pattern.) Dysthymia is described by the DSM-III-R as
'a chronic disturbance of mood involving depressed mood (or possibly an
irritable mood in children and adolescents), for most of the day, more days
than not, for at least 2 years (1 year for children and adolescents)'.
Similarly, the ICD-10 describes dysthymia as
a chronic depression of mood. ... The balance between individual phases of mild
depression and intervening periods of comparative normality is very variable.
Sufferers usually have periods of days or weeks when they describe themselves as
well, but most of the time (often for months at a time) they feel tired and depressed;
everything is an effort and nothing is enjoyed ... It usually begins early in adult life
and lasts for at least several years, sometimes indefinitely. ...
However, dysthymia is a wider concept than depressive personality; the
latter should have an early onset, which is not essential for dysthymia, and
be more stable than some disorders that attract the dysthymia diagnosis.
Cyclothymia is also a category in both these classifications and has been
reviewed by Howland & Thase (1993); in DSM-III-R this refers to 'a
chronic mood disturbance... involving numerous hypomanic episodes and
numerous periods of depressed mood or loss of interest or pleasure of
insufficient severity or duration to meet the criteria for a major depression
or a manic episode'. ICD-10 describes a similar disorder with 'a persistent
instability of mood, involving numerous periods of mild depression and
mild elation. This instability usually develops early in adult life and pursues
a chronic course, although at times the mood may be normal and stable for
months at a time'. This category has also been called 'affective personality
disorder', 'cycloid personality' and 'cyclothymic personality', but is now
considered to be a relatively mild variant of bipolar affective disorder
(manic-depressive disorder), and not a PD. There is an increased prevalence
of cyclothymia in relatives of patients with bipolar affective disorder, and
some cyclothymic individuals will eventually develop bipolar affective
disorder or recurrent depressive disorder (using ICD-10 terminology).
Both cyclothymia and borderline PD are associated with unstable mood
and impulsivity, but cyclothymic patients, who may be misdiagnosed as
having borderline PD, may show a greater response to treatment with
lithium. Although it has been suggested that cyclothymia may have a
specific association with borderline PD (Levitt et ai, 1990), this is not
established.
Main definitions
Hirschfeld & Shea (1992) have proposed the following criteria for depres-
sive PD: 'tendency to dysphoria, dejection, gloominess, cheerlessness,
joylessness; prominent self-concepts of inadequacy, worthlessness and low
self-esteem; critical, blaming, derogatory and punitive toward oneself, and
prone to guilt; brooding and given to worry; negativistic, critical and
judgemental towards others; and pessimistic'. Subsequently, the DSM-IV
Work Group on Personality Disorders proposed that the core feature is
excessive, negative, pessimistic beliefs about oneself and other people, as
indicated by at least five of the following: usual mood is dominated by
dejection, gloominess, unhappiness, cheerlessness, joylessness; prominent
self-concept centres around beliefs of inadequacy, worthlessness, and low
self-esteem; is critical, blaming, derogatory, and punitive toward oneself; is
brooding and given to worry; is negativistic, critical, and judgmental
towards others; is pessimistic; and is prone to feeling guilt, remorse. These
must not occur exclusively during major depressive episodes (Hirschfeld &
Holzer, 1994). These features form the research criteria for depressive PD in
DSM-IV as a criteria set 'provided for further study'.
Table 2.3 Categories of relationship between personality

disorder (PD) and depressive disorders
1. Personality traits that predispose to the development of depressive disorders

2. Personality traits that may be the consequence of depressive disorders
3. Personality traits that modify the symptom pattern, course and outcome of
depressive disorders
4. 'Depressive PD' refers to:
(a) some examples of DSM-III-R's 'dysthymia'
(b) some examples of ICD-10's 'dysthymia'
(c) 'subaffective dysthymia'
(d) some examples of the 'general neurotic syndrome'
Hypothesis: some examples of 'depressive PD' (e.g. patients with subaffective
dysthymia) are variants of bipolar affective disorder
5. 'Character-spectrum disorder,' similar to:
(a) 'depressive-spectrum disease'
(b) 'hysteroid dysphoria'
Hypothesis: this category of depression is mainly a psychological reaction to
complications of behaviour related to PDs, in particular borderline PD
Clinical features
Depressive PD reflects one type of association between PD and depressed
mood, and there arefivemain relationships that have been identified. These
are shown in Table 2.3 (Phillips et aL, 1990; Hirschfeld & Shea, 1992).
Firstly, personality traits may predispose to the development of a
depressive disorder (Klerman & Hirschfeld, 1988) and claims have been
made in this respect for excessive dependency; an excessive need for
reassurance, support and attention; and high reward dependence, harm
avoidance and novelty seeking (Hirschfeld & Shea, 1992; Svrakic, Przybeck
& Cloninger, 1992). In a prospective study of antenatal subjects, high
interpersonal sensitivity appeared to be a risk factor of subsequent
depression 6 months after childbirth (Boyce et al., 1991).
Secondly, maladaptive changes in personality may follow as a conse-
quence of a depressive disorder. For example, in a recent review, Akiskal
(1991) identified the syndrome of 'residual major depression', in which
patients with an unremarkable premorbid adjustment develop residual
chronicity after one or more depressive episodes, which do not remit fully.
He noted that the course of an unipolar depressive illness beginning after
the age of 50 is often protracted, and secondary personality changes can
involve pessimism, passivity, a sense of resignation, generalized fear of
inability to cope, adherence to rigid routines and inhibited communication.
Also, more serious maladaptive behaviour may follow, involving unstable
relationships, promiscuous behaviour, and substance abuse. This syn-

drome should be distinguished from 'anxious or atypical depression', which
is another chronic depressive disorder and involves mild to moderate
anxiety and depression with fatigue (Akiskal, 1991). These disorders are
conceptualized as secondary to anxiety disorders although other PD
features may co-occur by chance. Clinical features in patients (who are
typically young) include fatigue, low self-confidence, feeling worse in the
evening, insomnia, oversleeping in the morning and overeating.
Thirdly, personality may have a pathoplastic effect on the symptom
pattern, course and outcome of depressive disorders. Obsessive-compulsive
traits or PD may contribute to the development of pronounced morbid
worry and preoccupation with guilt, and obsessionality has been associated
with poor outcome (Duggan, Lee & Murray, 1990), while features of
histrionic PD may be associated with hostility, demanding or manipulative
behaviour and complaints. Also, in a study of the relationship between PDs
and treatment outcome in 239 outpatients, those with PD had a signifi-
cantly worse outcome in terms of social functioning and residual symptoms
of depression (Shea et al., 1990). But although a clinician's attention is
usually directed towards maladaptive aspects of a patient's personality, the
presence of adaptive features (i.e. a person's strengths of character) are also
likely to exert pathoplastic effects. Another long-term study has shown that
personality variables can influence outcome of depressive disorders
(Andrews et al., 1990).
The fourth category of relationship between PD and depressive disorders
involves the concept of 'depressive PD', which corresponds to various
descriptions, including 'dysthymia' and Akiskal's 'sub-affective dysthymia'
(Phillips et al., 1990). The latter has been associated with a favourable
response to antidepressant drugs and sometimes with the following:
development of hypomania; shortened REM latency which is a characteris-
tic of severe primary major depression; family history of unipolar or bipolar
affective disorder; unremarkable developmental history; relatively good
social outcome; equal sex distribution; and indeterminate onset (usually
before age 25). Rihmer (1990) reported that such patients had abnormal
results on the dexamethasone suppression test, which are also associated
with severe primary major depression. Associated personality traits were
similar to those described by Schneider (see above), i.e. gloomy, pessimistic,
serious, incapable of enjoyment or relaxation; quiet, passive, indecisive;
sceptical, hypercritical, complaining; worrying, brooding; duty-bound,
conscientious, self-disciplining; self-doubting, self-critical, self-reproach-
ing, self-derogatory; and preoccupied with inadequacy, failure, and nega-
tive events to the point of morbid enjoyment of one's failures. Although
such individuals were generally introverted, with minimal social life, they
could pursue a limited range of interests.
The relationship between depressive PD and the 'general neurotic
syndrome' (GNS) as described by Tyrer et al. (1992), is uncertain, but there
appears to be some overlap. The GNS is a combination of anxiety and
depression, with features of avoidant, obsessive-compulsive and dependent
PDs. It also involves excessive timidity, poor self-esteem, avoidance of
anxiety-provoking situations and excessive dependence on others.
The fifth and final category of relationship between PD and depressive
disorders involves 'character-spectrum disorder', which has been asso-
ciated with the following: irritability; relatively poor response to anti-
depressant medication; normal REM latency; normal dexamethasone
suppression; family history of alcoholism and sociopathy; childhood
parental loss, separation or divorce; poor social outcome; greater preva-
lence among females; onset in childhood or adolescence; and polysubstance
or alcohol abuse (Akiskal, Rosenthal & Haykal, 1980). The associated
personality traits are variable but a mixture of dependent, histrionic,
antisocial and borderline PD features is usually found, and there is more
social and mood instability than is found with subaffective dysthymia. It
has been suggested that character-spectrum disorder involves a mood
disorder that is secondary to the complications of an heterogeneous
mixture of PDs, in particular borderline PD, whereas depressed mood is an
integral part of the various subtypes of depressive PD.
Although borderline PD with depression can be considered as just one
type of character-spectrum disorder, many claims that patients with
borderline PD have a high prevalence of depression have led to particular
attention to this association. Such depression tends to show features
specific to co-occurrence with borderline PD, i.e. complaints of chronic
emptiness and feelings of loneliness (Grinker, Werble & Drye, 1968). (The
term 'hysteroid dysphoria', which refers to repeated episodes of depressed
mood in response to relationship difficulties, is often associated with
borderline PD.) However, Gunderson & Phillips (1991) concluded that the
relationship between borderline PD and depression is probably non-
specific, as some other PDs are as likely, or more likely, to be associated
with depressive symptoms, in particular obsessive-compulsive, avoidant
and dependent PDs. Although depression with borderline PD may respond
to some degree to tricyclic antidepressants in a minority of patients, both
hysteroid dysphoria and depression with borderline PD may respond
preferentially to MAOI (monoamine oxidase inhibitor) antidepressants, in
particular tranylcypromine (Dowson, 1987). However, other drugs, such as

phenothiazine, fluoxetine, lithium and carbamazepine, have also been
shown to have some effect on depressive symptoms with borderline PD, and
it has been suggested that the effects of these drugs may be mediated by a
reduction in behavioural dyscontrol (impulsivity) rather than by a change
in mood. Gunderson & Phillips (1991) concluded that the affective symp-
toms of patients with borderline PD may respond as well to antipsychotic
drugs as to antidepressants, and while this does not imply that antidepres-
sants do not have a role in the treatment of depression with borderline PD,
their effect 'is not likely to be as specific or profound as in uncomplicated
depressions'. However, some patients with borderline PD appear to have,
by chance, co-occurring bipolar disorder or major depression with episodic
endogenous aetiology.
'Character spectrum disorder' is very similar to Winokur's concept of
'depressive-spectrum disease', which was defined by the presence of uni-
polar depression with a history of alcoholism and/or antisocial personality
infirst-degreerelatives, while, in contrast, 'pure depressive disease' showed
no such family history. Compared with the latter, depressive-spectrum
disease was associated with earlier age of onset, female gender, personality
problems and difficulties (particularly more frequent divorce and sepa-
ration), lifelong irritability, and more variable symptoms. Other studies
have reported similarfindings(Akiskal, 1991), but unlike Winokur's claim
of a favourable outcome for character-spectrum disorder, recent studies
have found that depressed patients with PDs appear to have a worse
outcome (Hirschfeld & Shea, 1992).
Character-spectrum disorder also appears to incorporate 'hysteroid
dysphoria' (Liebowitz & Klein, 1979) in which women are usually affected
and there are repeated episodes of depressed mood as a consequence of
feeling rejected within close relationships. Such individuals are over-reliant
on approval, attention or praise, and show an extreme intolerance to
personal rejection. They have been called 'attention junkies'. Mood reac-
tivity is preserved, and fatigue is a common complaint together with
overeating and oversleeping.
Although, in general, character-spectrum disorder appears to be less
responsive than other main categories of depressive disorder to antidepres-
sants, lithium or both, certain subtypes of this heterogeneous category and
certain individual patients may benefit from medication. For example,
phenelzine has been claimed to be useful in 'hysteroid dysphoria' (Kayser et
al., 1985), and Akiskal (1991) suggested that many of the character-
spectrum disorders may eventually respond to future new antidepressant

medications.
Character-spectrum disorder seems to be associated mainly with DSM-
III-R's cluster B PDs, which include borderline PD, but other PDs are also
common in depressed patients (i.e. those with DSM-III-R's major
depression or dysthymia), for example obsessive-compulsive, passive-
aggressive, schizoid, schizotypal, paranoid and dependent PDs (Alnaes &
Torgersen, 1988a, b, 1991). In patients with major depression (=197),
with dysthymia (n = 62) and with both disorders (n = 32), at least one PD
was found in 50%, 52% and 69% of the sample respectively, and the most
common were avoidant and dependent PDs (Sanderson et al., 1992).

DSM-III-R dysthymia (which can be considered to include depressive PD)
is common in clinical practice and appears to be more prevalent in females,
while character-spectrum disorder involves depressive features in the
context of various PDs and may present with social and relationship
problems, substance abuse and self-harm episodes.
Associated features
Depressive PD can sometimes be associated with a good response to
antidepressant medication, while the heterogeneous group of character-
spectrum disorders are generally less responsive to drug treatments.
However, there are claims that some individuals with character-spectrum
disorder, for example with hysteroid dysphoria, respond to medication, in
particular MAOIs (Tyrer, Casey & Gall, 1983).
Current status
The concept of depressive PD remains controversial and this term is not
featured in the main international classifications, although in DSM-IV it is
a criteria set 'provided for further study'. There is evidence to support such
a category (Widiger, 1989), such as reports in relation to the similar concept
of sub-affective dysthymia, involving young adults with chronic depression
and maladaptive personality traits (other than antisocial and borderline
features), who appear to respond to antidepressant medication. Such
individuals may have a variant of bipolar manic-depressive disorder,
mediated by genetic predisposition, and many patients who would have

been previously diagnosed as having a PD can be reclassified as having a
form of chronic reversible depression. This has led to successful treatment
with antidepressant medication and the claim that 'physicians should
therefore attempt vigorous pharmacological interventions with most forms
of chronic depression . . . psychotherapy alone proves inadequate for many
of these patients' (Akiskal, 1991). In addition, Hirschfeld & Holzer (1994)
claimed that the results of the DSM-IV Mood Disorders Field Trial show
that the DSM-IV criteria for depressive PD (previously described) identi-
fies a group of patients many of whom do not receive other diagnoses
involving depressive symptoms.
Organic personality disorder
Main features and clinical origins

This category is found in the ICD-10 classification and refers to an
alteration of personality and behaviour that 'can be a residual or concomi-
tant disorder of brain disease, damage, or dysfunction' and involves 'a
significant alteration of the habitual patterns of premorbid behaviour',
particularly in relation to the expression of emotions, needs and impulses.
Main definitions and clinical features

In addition to a history or other evidence of brain disease, damage or
dysfunction, the ICD-10 requires two or more of the following criteria to be
met:
consistently reduced ability to persevere with goal-directed activities, especially
those involving longer periods of time and postponed gratification; altered
emotional behaviour, characterised by emotional lability, shallow and unwarranted
cheerfulness (euphoria, inappropriate jocularity), and easy change to irritability or
short-lived outbursts of anger and aggression - in some instances apathy may be a
more prominent feature; expressions of needs and impulses without consideration
of consequences or social convention (the patient may engage in dissocial acts, such
as stealing, inappropriate sexual advances, or voracious eating, or may exhibit
disregard for personal hygiene); cognitive disturbance, in the form of suspiciousness
or paranoid ideation, and/or excessive preoccupation with a single, usually
abstract, theme (e.g. religion, right and wrong); marked alteration of the rate and
flow of language production with features such as circumstantiality, over-inclusive-
ness ...; and altered sexual behaviour
Some of these features, for example emotional lability and lack of
Enduring personality change 127
consideration of consequences, have been part of the so-called frontal lobe
syndrome, but these and other features of organic PD can also occur with
lesions to other parts of the brain.
Welch & Bear (1990) have reviewed many of the wide range of personal-
ity changes that can occur as a result of organic (i.e. identifiable) brain
disorders. A variety of disorders need to be considered if there is an
unexplained appearance of 'personality disorder', and these include tem-
poral lobe epilepsy, tumours, Huntington's disease, Parkinson's disease,
multiple sclerosis, AIDs encephalopathy and the effects of neurotoxins. For
example, some reports have linked apparent obsessive-compulsive PD with
temporal lobe epilepsy, paranoid and aggressive behaviour with Hunt-
ington's or Parkinson's diseases, and social withdrawal and indecisiveness
with AIDS encephalopathy. A 'pseudo-psychopathic syndrome', with
disinhibited, impulsive and socially inappropriate behaviour,- is associated
with orbital-medial frontal lobe damage, while a 'pseudodepressive syn-
drome', with lack of spontaneity, apathy and indifference is related to
dorsolateral frontal lobe lesions.
Enduring personality change after psychiatric illness
Main features and clinical origins

The ICD-10 includes this category to describe apparent personality
changes due to the experience of a severe psychiatric illness. These involve
such features as excessive dependence, reduced interests and social
isolation.
Main definition and clinical features

There should be no evidence of a similar pre-existing PD, although previous
personality vulnerability may be a predisposing factor. ICD-10 specifies the
following range of clinical features:
excessive dependence on and a demanding attitude towards others; conviction of
being changed or stigmatised by preceding illness, leading to an inability to form
and maintain close and confiding personal relationships and to social isolation;
passivity, reduced interests, and diminished involvement in leisure activities;
persistent complaints of being ill, which may be associated with hypochondriacal
claims and illness behaviour; dysphoric or labile mood, not due to the presence of a
current mental disorder or antecedent mental disorder with residual affective
symptoms; and significant impairment in social and occupational functioning
compared with the premorbid situation.
Personality disorders: less specific clinical
presentations and epidemiology
J.H. DOWSON
Personality disorder not otherwise specified (DSM-III-R and DSMIV) or

mixed and other personality disorders (ICD-10)
Main features and definitions

The DSM-III-R and DSM-IV category of 'personality disorder - not
otherwise specified' refers to any disorder of personality that cannot be
classified as one or more specific PDs, for example, features of one or more
PDs that do not meet the full criteria for any one PD, yet cause significant
impairment in social or occupational functioning, or subjective distress.
The ICD-10 category 'mixed and other personality disorders' is similar
but also includes the disorders of patients who have several PDs but
without a predominant set of symptoms. The latter would be expected to be
the most prevalent PD diagnosis in psychiatric practice, as many studies
have shown frequent co-occurrence of two or more PDs in various patient
populations.
Clinical features
Most studies of PDs have found that at least 50% of patients with PD had
two or more PDs (Oldham et al., 1992); for instance Nurnberg et al. (1991)
found 82% of patients with borderline PD had at least one additional PD
diagnosis from another DSM-III-R PD cluster. Although the term 'co-
morbidity' has generally been used to denote co-existing disorders that
have distinct causes, the term 'co-occurrence' is more appropriate in
relation to PDs, as it does not have any aetiological implications.
In a study of 106 applicants for long-term treatment of severe PD
(Oldham et al., 1992), which used structured interviews for PD assessment,
statistically significant associations were found involving six pairs of PDs,
128
Depressive disorders and personality disorders 129
namely histrionic and borderline, histrionic and narcissistic, narcissistic
and antisocial, narcissistic and passive-aggressive, avoidant and schizoty-
pal, and avoidant and dependent. It was concluded that frequent and
consistent patterns of co-occurrence had been shown.
Many other studies have reported that co-occurrence is common, as
evaluated by various methods in many clinical and non-clinical popula-
tions (Dowson, 1992ft). Zimmerman & Coryell (1989) assessed DSM-III
PDs with an interview method in a non-patient sample of 797 relatives of
patients and controls; about 17% had at least one PD diagnosis and, of
these subjects, about 25% had more than one. The most prevalent PDs were
mixed, passive-aggressive, antisocial, histrionic and schizotypal PDs. With
the increasing attention given to borderline PD as defined by DSM-III and
DSM-III-R, several studies have examined its co-occurrence with other
PDs (Grueneich, 1992). It appears that the most frequently-reported
association has been with histrionic PD, but co-occurrence has also been
found with other PDs including schizotypal, antisocial, dependent and
passive-aggressive. Also, several studies have provided evidence in support
of relationships between co-occurring PDs which reflect the three PD
'clusters' defined by DSM-III-R (Dowson & Berrios, 1991).
A ssociatedfeatures
The co-occurrence of PDs and non-PD psychiatric disorders may involve
significant associations; for example, Nestadt et al. (1991) evaluated 810
randomly-selected non-patients in respect of DSM-III compulsive and
antisocial PDs, generalized anxiety disorder, alcohol-related disorders and
simple phobia, and found that generalized anxiety disorder and simple
phobia were associated with obsessive-compulsive PD features, while
alcohol-related disorders (and a reduced occurrence of generalized anxiety
disorder) were associated with antisocial PD features.
Such associations can be investigated from the perspective of the non-PD
disorder by using the latter as the starting point to define a patient
population whose PD status is then determined. In the rest of this chapter
this approach will be used in an examination of the PD characteristics of
patient groups selected on the basis of other disorders.
Depressive disorders and personality disorders

There have been several claims that borderline PD is common among
patients with DSM-III-R's 'major depression' (Alnaes & Torgersen, 1990),
130 Less specific presentations and epidemiology
together with reports of the co-occurrence of major depression and several

other PDs. Sato, Sakado & Sato (1993a, Z>), in a study of 4-month outcome
in 96 outpatients with major depression, noted that while any co-occurring
PD worsened outcome, this was particularly found for schizoid PD and
other 'cluster A' PD features. Also, Marin et al. (1993), in a study
comparing PD in patients with 'dysthymia' and episodic 'major
depression', found a PD diagnosis in 51 % and 42% respectively. The most
common PDs in the dysthymia group were avoidant and NOS (not
otherwise specified) PDs, while the most common PDs in the major
depression group were borderline and NOS PDs. Joyce, Mulder & Clon-
inger (1994&), reporting a group of 84 patients with major depression
treated with antidepressants, found that temperament (as measured by the
Tridimensional Personality Questionnaire) accounted for 35% of the
variance in treatment outcome, while clinical variables predicted only 5%.
In relation to DSM-III 'dysthymic disorder', there were more co-occur-
rences with avoidant, self-defeating, dependent and borderline PDs, com-
pared with a non-dysthymic patient group (Markowitz et al., 1992).
Some of the associations between PD and certain types of depressive
disorder have been described (see depressive personality, Chapter 2), and
one of the main conclusions of clinical importance was that some young
people with chronic depressive disorders with apparent PD, who have
tended to be dismissed as pharmacologically untreatable, can respond to
antidepressant medications. It is also important for the clinician to be
aware that a non-PD psychiatric disorder, such as a recurrent depressive or
manic episode, can lead to an exaggeration of PD psychopathology that
may disguise the co-occurring episodic disorder. For example, a patient
with some antisocial PD features and a mild manic illness may become
considerably more antisocial during the course of the manic episode. The
antisocial behaviour may be the predominant feature and, if the person has
shown some previous antisocial tendencies, it may not be recognized that
an additional disorder is present, namely a treatable manic illness. Another
example would be a person with some histrionic and dependent PD features
who develops a depressive episode, when his or her behaviour may become
childish, demanding and manipulative; if this behaviour predominates, the
additional diagnosis of an episodic depressive disorder may be missed.
Suicide and personality disorders

Suicide and other forms of violent death or self-injury have been investi-
gated in relation to their associations with PD, and it should be noted that
Suicide 131
one of the DSM-III-R criteria for borderline PD is 'recurrent suicidal
threats, gestures, or behaviour, or self-mutilating behaviour'.
Several large long-term studies have reported a link between persistent
antisocial behaviour and excess mortality. In a follow-up of 500 psychiatric
outpatients, unnatural death was associated with antisocial PD as well as
with affective disorders, substance abuse and homosexuality, and in a
retrospective study of 1056 antisocial Swedish adolescents who had been
admitted to 'probationary schools', it was found that 13% of the boys and
10% of the girls had died over a 17-year period. Eighty-eight per cent of the
dead boys and 77% of the dead girls had died sudden violent deaths by
accident, suicide, death from uncertain causes, murder/manslaughter, or
alcohol/drug abuse (Martin, 1986). Similar associations were reported in a
prospective follow-up study of 50465 Swedish conscripts. There were 247
completed suicides during 13 years' follow-up and strong predictors of
suicide were early indications of antisocial PD (poor emotional control,
deviant behaviour, substance abuse, contact with a child welfare authority
or the police, and lack of friends), together with adverse social background
(Allebeck, Allgulander & Fisher, 1988). This study also showed that, as well
as PD, diagnoses of neurotic disorder or drug dependence at conscription
were associated with a significantly increased risk for future suicide
(Allebeck & Allgulander, 1990(3, b), although a subsequent period of
psychiatric inpatient care was the strongest predictor of suicide.
An approach from a different perspective compared the criminal history
of 181 suicides and matched controls: although the difference in the
frequency of a history of a criminal record (16% of the suicides and 11 % of
the controls) was not significant, there was a significant difference in the
nature of the offences, as a greater number (more than half) of the offences
committed by criminal suicides involved violations of the road traffic laws
(Modestin & Emmenegger, 1986). In a study of adolescent suicides in
Finland (44 males and 9 females), a history of antisocial behaviour was
found in 43%. Separation from parents, and alcohol abuse and violence in
the home, had been common in the males, who had shown more alcohol
abuse and co-occurring mental disorders compared with those without
antisocial behaviour (Marttunen et al., 1994).
Retrospective studies have examined the association between borderline
PD and suicide. A study of 58 consecutive suicides by adolescents and
young adults (aged 15 to 19) in an urban community, reported borderline
PD in 33% (which was the principal diagnosis in 28%), with antisocial PD
in 14% (Runeson & Beskow, 1991). Also, a comparison of individuals with
borderline PD who committed suicide, with a control group with borderline
PD, found that the suicide group had more frequent childhood loss, lack of
treatment contact before hospitalization, longer hospitalization and more
frequent discharges for violating a treatment contract (Kjelsberg, Eikeseth
&Dahl, 1991).
Self-harm and personality disorders

Repetition of self-harm behaviour is common, with reports of 10-15% of
patients being involved in further episodes within a year, and this is
associated with a higher risk of completed suicide. But although several risk
factors for suicide are known, such as previous self-harm, previous
psychiatric treatment, personality disorder (in particular antisocial and
borderline PDs), alcohol and drug abuse, unemployment and lower social
class, accurate prediction is difficult for the individual patient.
'Self-harm' is an heterogeneous group of behaviours, and several sub-
types have been identified, i.e. self-poisoning, self-wounding (which is
usually of low lethality), self-mutilation (involving serious injury, usually in
the context of a non-PD psychiatric disorder), and the person who makes 'a
highly lethal suicide attempt', for example, 'one deep and dangerous cut'
(Tantam & Whittaker, 1992). Self-wounding is associated with borderline
PD and with a history of sexual or physical abuse; episodes are usually of
low lethality, are often chronic and repetitive, are not typically associated
with death-orientated thoughts and commonly involve multiple methods of
self-injury (Russ, 1992). The patient usually carries out self-wounding
behaviour when alone, often in response to a perceived rejection or loss and
in a context of emotional arousal, i.e. with anxiety, tension, depression,
anger or emptiness. Self-wounding behaviour is not aimed at mutilation or
death and the motivation usually appears to be an attempt to influence
others or relieve distress, in particular a feeling of tension. The most typical
form of self-wounding appears to be wrist-cutting involving superficial
wounds, but burning or hitting oneself is also common. However there is a
continuum of behaviours, which can progress to severe self-mutilation.
Favazza & Conterio (1989) reported 240 women who regularly wounded or
mutilated themselves (although more men than women do it), in whom
cutting was the commonest method, in 72%. Other behaviours included
skin burning in 35%, hitting or punching parts of the body in 30%,
interfering with wound healing in 22%, scratching in 22%, hair pulling in
10%, and breaking bones in 8%. The wrists were involved in 74%, the legs
in 44%, the abdomen in 25%, the head in 23%, the chest in 18% and the
genitalia in 8%. Implements included broken glass, needles, scissors, razor
Substance use disorders 133
blades, knives, hammers and cigarettes. A study which compared a group

of patients with PD and an history of self-wounding, with a control group
without self-wounding but matched for PD, reported that the degree of self-
wounding was significantly correlated with impulsivity, chronic anger and
somatic symptoms of anxiety, and that the self-wounding group showed
more severe PD, with greater aggression and antisocial behaviour. It has
been suggested that self-wounding is related to underlying serotonergic
dysfunction (Simeon et al., 1992). Another study of self-wounding com-
pared these behaviours with self-poisoning, and found that the former was
more associated with youth, male sex, single status, PD, previous psychi-
atric inpatient status, use of alcohol at the time of the self-harm episode,
alcohol problems, violence against others and violence received from
relatives in the preceding 5 years, criminal record and unemployment
(Robinson & Duffy, 1989).
Substance use disorders and personality disorders

Many studies have described high rates of PD in samples selected on the
basis of substance use disorders, in particular involving antisocial, border-
line and narcissistic PDs (Nace, 1989). In a study of alcoholic out-patients,
64% had PD, in particular paranoid PD (in 44%), antisocial PD (in 20%),
avoidant PD (in 20%), passive-aggressive PD (in 18%) and borderline PD
(in 16%) (Nurnberg, Rifkin & Doddi, 1993), and PD was associated with
poorer outcome. Also, the trait of 'alexithymia' has been associated with
substance abuse (and with post-traumatic stress disorder), involving an
inability of the subject to identify a depressed mood, together with a
concrete style of thinking (Wise, Mann & Shay, 1992). Also, in 501 patients
seeking treatment for alcohol and other drug problems, the most common
other 'lifetime' disorders in their past histories were antisocial PD, phobias,
psychosexual dysfunctions and depressive disorders (Ross, Glaser &
Germanson, 1992). Co-occurring PD can affect the symptom pattern of
substance use disorders as well as treatment response and outcome; of 160
patients admitted to a private inpatient substance abuse programme, the
57% who had PDs differed from the rest in being more impulsive, isolated
and depressed (Nace, Davis & Gaspari, 1991). Also, it has been noted that
features of PD may occur secondary to (i.e. after) substance abuse, such as
when antisocial behaviour results from efforts to obtain money for illicit
drugs or alcohol, or when pre-existing mild personality vulnerability
becomes more pronounced.
Cloninger and colleagues (1988) have made claims for two main subtypes
of alcoholism, which differ in respect of alcohol-related symptoms, person-
ality traits, ages of onset and patterns of inheritance. Type 1' alcoholism is
associated with later onset, anxiety and dependent PD features, rapid
development of tolerance, dependence on the anti-anxiety effects of alco-
hol, loss of control, bingeing on alcohol, difficulty stopping binges, guilt
feelings and liver complications, while 'type 2' alcoholism is related to
earlier onset and antisocial PD traits. In the type 2 pattern, alcohol is
persistently sought for its euphoriant effects, there tends to be an inability to
abstain (rather than bingeing), and various complications occur earlier;
these include aggression and criminal involvement when drinking. Also the
type 2 disorder appears to be associated with a higher incidence of familial
alcoholism.
Although this classification appears to have identified various associa-
tions within patient groups and their relatives, some investigators have
reported that these two symptom patterns may not be a satisfactory basis
on which to divide alcoholism; for example, in a study of 360 male alcoholic
patients in hospital, 9 1 % satisfied criteria for both symptom clusters, and
when the sample was divided on the basis of early or late onset of
alcoholism (which is a variable that can contribute to the 'types 1 and 2'
classification, i.e. <25 years associated with type 2, and >26 years
associated with type 1), 96% of the early onset and 83% of the late onset
subgroups still showed a history of both symptom clusters. It was con-
sidered that the 'types 1 and 2' classification mainly reflects age of onset of
problem drinking, at least in hospital settings (Penick et ah, 1988). A similar
conclusion resulted from a study which applied Cloninger's subtypes to
alcoholic women (Glenn & Nixon, 1991). These subjects were also divided
into early and late onset groups (<25 and >26 years), and group
membership did not correlate with the symptomatology of types 1 and 2 as
predicted. Most of the women were positive for both types on the basis of
symptoms, and this study also suggested that the criteria for grouping
patients with alcoholism should be based on age of symptom onset,
regardless of symptom type.
In a study which grouped patients on the basis of age of alcoholism onset,
those with onset before their 20th birthday had a significantly higher
incidence of paternal alcoholism (Buydens-Branchey, Branchey & Nou-
mair, 1989), and patterns of inheritance appear to make a valid contribu-
tion to the criteria for the classification of alcoholism. Bohman et al. (1987)
examined the inheritance of alcohol abuse and PD in 862 men and 913
women adopted by non-relatives, and found that both male and female
adoptees were at increased risk of developing alcohol abuse if their
Substance use disorders 135
biological, but not their adoptive, parents were alcoholic. Three disorders
appeared to be associated in biologically related individuals, namely
alcohol abuse, somatoform disorders in women and antisocial behaviour in
men, and evidence for both genetic and environmental risk factors for
alcoholism were found. One of the associations could be explained by a
common vulnerability that was expressed as antisocial, violent and criminal
behaviour with recurrent alcohol abuse in males, but as frequent somatic
complaints (somatoform disorders) in females.
Despite the difficulties in separating some patient populations with
alcoholism into subgroups based on types 1 and 2, various features of these
subtypes have been related to outcome variables. In a 2-year outpatient
treatment programme, all those with an early stable improvement were type
1 alcoholics, while most who improved during the later part of treatment
were type 2 (Ojehagen & Berglund, 1986<z). Also, in a 20-year follow-up of
84 male alcoholics, those of type 2 showed a relatively high frequency of
sporadic abuse and a lower frequency of abstinence and social drinking
(Nordstrom & Berglund, 1987).
Many studies have reported on the frequent associations between
alcoholism and antisocial PD. In a study of 260 men starting inpatient
treatment, those with antisocial PD were distinguished by an earlier onset
of alcoholism, involvement with illegal drugs and more problems with
control of drinking (Yates, Petty & Brown, 1988), while in another study of
241 alcoholic men, also in hospital (Penick et al, 1988), the most frequent
co-occurring disorders were depressive disorders, antisocial PD and drug
abuse. Also, a 1-year follow-up study of 266 patients with alcoholism found
that, for both men and women, having an additional diagnosis of antisocial
PD or drug abuse was associated with a poorer outcome (Rounsaville et aL,
1987),
The prevalence of antisocial PD in patients with alcoholism is less in
females than in males, but a survey of 50 women attending an alcohol-
treatment unit found a high frequency of other behaviours that are often
related to PD; about half had taken an overdose, or described a history of
impulsive physical violence, or had had a period of promiscuity, while
about 25% had an history of self-wounding, and at least 16% had had an
eating disorder (Evans & Lacey, 1992).
PDs other than antisocial PD have been claimed to be related to the
outcome of alcohol abuse. Griggs & Tyrer (1981) reported that outcome
was better for those with schizoid PD than for those with antisocial or no
PD, although another study came to the opposite conclusion (Zivich,
1981).
With regard to the abuse of other substances, a study of 76 outpatients

who abused cocaine and 'crack' found antisocial PD in 21%, but also
passive-aggressive PD in 21 %, borderline PD in 18% and self-defeating PD
in 18% (Kleinman et <?/., 1990). Subtypes of cocaine abusers have been
suggested by Weiss & Mirin (1986), and include patients with antisocial PD
who use cocaine as part of an overall pattern of antisocial behaviour, in
contrast to those with narcissistic and borderline PDs who use cocaine for
social prestige and to bolster self-esteem. In a comparison of cocaine users
and opioid addicts, cocaine users showed lower rates of borderline and
antisocial PD features (Malow et ai, 1989). Antisocial PD has also been
reported to commonly co-occur with chronic solvent abuse (Dinwiddie,
Zorumski & Rubin, 1987).
Intravenous drug abuse is of particular importance, as it is associated
with HIV transmission. In a study of 273 intravenous drug abusers, who
were examined by a structured interview (Brooner et aL9 1992), antisocial
PD that included a history of childhood conduct disorder was present in
44%, while another 24% met only the criteria for antisocial PD that related
to adult life. The former group had a more pervasive and serious pattern of
adult antisocial behaviour, and it was suggested that the antisocial behav-
iour of those who met only the adult criteria for antisocial PD have a late-
onset and less severe form of PD. Associations of the injection of illicit
drugs in relation to PD was investigated in a study of 92 intravenous drug
users who were compared with three other groups: a drug-free group, a
group that used cannabis, and a group that had used illicit drugs other than
cannabis but had never injected (Dinwiddie, Reich & Cloninger, 1992). It
was found that the main difference between the drug users who injected and
those that did not was the much higher prevalence (68%) of antisocial PD
among the intravenous drug users, and it appeared that this was not just a
secondary effect of drug-related antisocial behaviour. Another study of
intravenous users showed that those with antisocial PD reported signifi-
cantly higher rates of sharing of injection-equipment, and of the number of
people involved with sharing, than those without antisocial PD (Brooner et
ai, 1990). There is also evidence that relatively high scores on borderline
and antisocial PD features is associated with multiple substance depen-
dence. However, needle-sharing is also influenced by social relationships
and gender (Barnard, 1993), and women injectors may be at greater risk of
HIV transmission than males. Perkins et al (1993) found that PD was
common in the HIV-positive population, and that those with PD had a
reduced ability to cope with the threat of AIDS.
Somatization disorder 137
Somatization disorder and personality disorders

The features of DSM-III-R's 'somatization disorder' are 'recurrent and
multiple somatic complaints, of several years' duration, for which medical
attention has been sought, but that apparently are not due to any physical
disorder. The disorder begins before the age of 30 and has a chronic but
fluctuating course'. It is mainly diagnosed in women and estimates of
lifetime prevalence rates have varied between 0.2 and 2%. DSM-III-R also
notes that
the male relatives of females with this disorder show an increased risk of antisocial
personality disorder and psychoactive substance use disorders. Adoption studies
indicate that both genetic and environmental factors contribute to the risk of this
group of disorders, because both biologic and adoptive parents with any of the
disorders increase the risk of antisocial personality disorders, psychoactive sub-
stance use disorders and somatization disorder.
Therefore, somatization disorder has been considered as a female

equivalent to antisocial PD in men, both sharing a similar genetic substrate.
In a study of 25 women with somatization disorder and matched patient
controls with depressive or anxiety disorders, the prevalence of PDs among
the former group was 72% compared with 36% among controls. Depen-
dent and histrionic PD features were among those that occurred signifi-
cantly more often (Stern, Murphy & Bass, 1993).
Somatization disorder (which is also known as Briquet's syndrome) has
also been considered to be a variation of DSM-III-R's 'conversion dis-
order' (or 'hysterical neurosis, conversion type'), which involves 'an
alteration or loss of physical functioning that suggests physical disorder,
but that instead is apparently an expression of a psychological conflict or
need. The symptoms of the disturbance are not intentionally produced and,
after appropriate investigation, cannot be explained by any physical
disorder or known pathophysiologic mechanism' (American Psychiatric
Association, 1987). Examples of conversion symptoms can include appar-
ent paralysis, blindness and co-ordination disturbances. PD, in particular
histrionic PD, has been associated with somatization disorder and conver-
sion disorder, and has been suggested as a predisposing factor, but
histrionic and other DSM-III-R 'cluster B' PDs are not invariably present.
Chandrasekaran, Goswami & Sivakumar (1994), in a follow-up study of
'hysterical neurosis' found that premorbid 'hysterical personality' pre-
dicted poor outcome.
The association of PD with the related condition of DSM-III-R 'disso-
ciative disorders' (or 'hysterical neuroses, dissociative type') is less certain.
In dissociative disorders there is a change in the way identity, memory, and
conscious awareness are experienced and integrated. If identity is mainly
affected, the person's usual identity can be temporarily 'forgotten' (i.e. is
not available to conscious awareness), and, in an extreme variant of this
disorder, a new identity (or more than one new identities) may be evident.
The latter variant has been called 'multiple personality disorder' but this is
not grouped with the other PDs as it more closely resembles other
dissociative disorders. 'Multiple personality disorder' 'involves two or
more distinct "personalities", sometimes with unique memories, behav-
iours and social relationships. Such "personalities" can be aware of the
others but this is not invariable as the 'personality' that presents for
treatment may have little or no awareness of the existence of the other
"personalities'" (American Psychiatric Association, 1987). But concepts
and diagnosis in relation to multiple personality disorder are confusing and
controversial (Piper, 1994). Usually most 'personalities' of multiple person-
ality disorder report lost periods of time or periodic confusion about past
experience of time. Sometimes one or more of the 'personalities' shows
features of another disorder, such as borderline PD, but 'it is unclear
whether these represent coexisting disorders or merely associated features
of multiple personality disorder' (American Psychiatric Association, 1987).
This syndrome appears to have been diagnosed from 3 to 9 times more
frequently in females, and complications include self-wounding, violence
and substance use disorder. In such cases a PD may co-occur by chance,
and the existence of specific predisposing PD features is uncertain. How-
ever, as histrionic PD appears to be more frequently associated with
somatization disorder compared with the other PDs, this PD, together with
related 'cluster B' PDs, may also be a predisposing factor for the related
dissociative disorders. Other, more common, dissociative disorders include
a 'psychogenic fugue', i.e. 'a sudden, unexpected travel away from home or
customary work locale with assumption of a new identity and an inability
to recall one's previous identity', and 'psychogenic amnesia', i.e. 'a sudden
inability to recall important personal information' (American Psychiatric
Association, 1987). Another related syndrome is 'factitious illness' or
'Munchausen's syndrome', in which physical symptoms and signs of illness
are fabricated; for example, a cut may not heal because the individual
repeatedly opens the wound. While deliberate malingering can occur,
dissociative states may reduce awareness of the individual's actions in
Somatization disorder 139
factitious illness, and co-occurring PD may act as a non-specific predispos-

ing factor. A variant of this disorder is factitious illness 'by proxy'
involving, for example, a parent fabricating the disorder in a child (Sims,
1992).
In addition to an apparent association between histrionic PD and
somatization disorder, there has been accumulating evidence for an
association between somatization disorder, conversion disorder and anti-
social PD (Lilienfeld et al, 1986). Guze, Woodruff & Clayton (1971a)
found an increased prevalence of conversion symptoms associated with
antisocial PD, while Guze, Woodruff & Clayton (19716) compared 30
women with a form of'hysterical' disorder (i.e. conversion or dissociative)
with 33 women with anxiety neurosis, and found a higher rate of childhood
delinquency in the former. Also, Cloninger & Guze's study of 66 female
felons (1970) reported that 40% of those with antisocial PD had received an
additional diagnosis of an hysterical disorder. Follow-up studies have also
provided evidence for such an association; Robins (1966) found that of 76
girls referred to a child guidance clinic for antisocial behaviour, 20
subsequently developed an adult hysterical disorder, compared with no
cases in a control group. Family studies have also shown that antisocial PD
and somatization disorder are related; for example, Woerner & Guze (1968)
found a high prevalence of antisocial PD among thefirst-degreerelatives of
subjects with somatization disorder, while Guze, Wolfgram & McKinney
(1967) reported a high frequency of somatization disorder among the
female relatives of male criminals. Also, somatization disorder and anti-
social PD are often seen among the parents of hyperactive children.
Lilienfeld et al. (1986) reported the association of antisocial PD, histrionic
PD and somatization disorder within individuals and within families in 250
patients; the strongest relationship was between antisocial and histrionic
PDs in different individuals. In this study, a high prevalence of antisocial
PD was found in the families of patients with somatization disorder but not
in the families of patients with histrionic PD, and it was suggested that
somatization disorder in women and antisocial PD in men may represent
gender-related variants based on similar causal factors.
If histrionic PD and somatization disorder co-occur, the demanding,
dramatic behaviour of histrionic PD will focus attention on any unex-
plained somatic symptoms. This would be expected to increase the likeli-
hood that somatization disorder will be diagnosed, as patients are likely to
return relatively often for medical consultations (Morrison, 1989).
Anxiety disorders and personality disorders

The main categories of anxiety disorders in DSM-III-R and DSM-IV are
panic disorder (with or without agoraphobia), agoraphobia, social phobia,
specific phobia and generalized anxiety disorder.
In a review of the role of personality in anxiety disorders, Papp et al.
(1990) concluded that PDs are more common in patients with anxiety
disorders than in the general untreated population, although the signifi-
cance and nature of the associations are uncertain; for instance, it is often
unclear whether the apparent PD features are premorbid, or the result of
the chronic experience of anxiety. Sanderson et al. (1994) studied a group of
347 patients with anxiety disorders and found that 35% had at least one
PD. PD was rare with simple phobia, and the most common PDs were those
of 'cluster C, in particular avoidant and obsessive-compulsive PDs.
It might be expected that there are relationships between anxiety
disorders and schizotypal, borderline and avoidant PDs, as all these PDs
have criteria related to anxiety. Schizotypal PD can involve excessive social
anxiety, borderline PD criteria include affective instability which may
involve anxiety, and avoidant PD is associated with fear of negative
evaluation. The ICD-10 category that corresponds to avoidant PD is
termed 'anxious PD', which involves 'persistent and pervasive feelings of
tension and apprehension ...', so that it has been argued that avoidant (or
anxious) PD cannot be usefully distinguished from social phobia, a view
that is supported by the finding that most patients with social phobia
qualify for the diagnosis of avoidant PD (Schneier et al., 1991). One model
for the relationship between avoidant PD and anxiety disorders is that
anxiety exaggerates the features of avoidant PD, so that the number and
degree of the features of avoidant PD are positively correlated with the
severity of anxiety, and this hypothesis was supported by the finding that
avoidant PD scores fell in association with successful treatment of anxiety
symptoms. But although certain avoidant PD features appeared to be
highly correlated with an improvement in anxiety symptoms (i.e. they were
'state-dependent'), others showed little change (Noyes et al., 1991).
Several studies have examined groups of patients with panic disorder
and/or agoraphobia, and reported rates of co-occurring PDs which ranged
from 27% to 58% (Pollack et al., 1992). Patients with frequent panic
attacks often have high anticipatory levels of anxiety, and seem dependent,
avoidant and unassertive. Also, Hoffart et al. (1994) found that avoidant
and dependent traits were related to symptom severity in patients with
agoraphobia and panic disorder. Although some PD features, such as those
Anxiety disorders 141
of dependent PD, may become less marked if panic symptoms are
successfully treated, others, such as those of histrionic PD, appear to be
more resistant to change (Mavissakalian & Hamann, 1986). This suggests
that some PD traits as measured by current methods are 'epiphenomena' of
panic disorder, while others may represent a premorbid predisposition for
panic disorder and agoraphobia. However, predisposing PD features may
not all have a high degree of specificity; for example, in a study of 187
patients with panic disorder and 51 with obsessive-compulsive disorder, the
most frequent PDs in the former group (avoidant, dependent, histrionic
and borderline PDs) were also more frequent in those with obsessive-
compulsive disorder (Mavissakalian, Hamann & Jones, 1990a, b). But it
has been suggested that panic disorder and features of certain PDs may
share causal factors (Mavissakalian, 1990), and it seems that behaviourally
inhibited and anxious children are at increased risk of developing adult
anxiety disorders. Complex interactions are possible, as childhood anxiety
may be a factor in the development of certain PD features.
Papp and colleagues' review (1990) concluded that, in eight studies,
about half the patients with panic and/or agoraphobia had one or more
DSM-III-R 'cluster C PDs, involving avoidant, dependent, obsessive-
compulsive and passive-aggressive PDs. But other associated PDs have
also been reported; for example, in a study of 55 outpatients with anxiety
disorders, 13% had borderline PD, 11% histrionic PD, 11% passive-
aggressive PD and 13% schizotypal PD (Alnaes & Torgersen, 1988&).
Although this study found a relatively low co-occurrence with borderline
PD, severe and chronic anxiety has been described as a characteristic
feature of this disorder, and affective instablity is featured in its DSM-III-R
criteria. Benjamin et al. (1989) compared groups of subjects on several
measures of anxiety; the groups included 11 patients with discrete border-
line PD, 31 normal subjects, and 16 with depressive disorders. Those with
borderline PD did not appear to be more anxious than the other patient
groups, but showed more hostility, interpersonal sensitivity and suspicious-
ness. It was suggested that marked mood changes, including anxiety, are
mainly related to interpersonal contexts in those with borderline PD, while
methods used to assess anxiety often relate to feelings that the patient has
when alone or out of the interpersonal context. Outside interpersonal
situations, those with borderline PD may experience mood changes such as
non-specific arousal or agitation, which are not described as anxiety, but
give rise to complaints such as being 'unreal', 'detached' or 'lost'. For those
with borderline PD, even in an interpersonal context, the usual terminology
of anxiety disorders may not be used; instead, a patient may say they are
angry, misunderstood or criticized. This calls into question the definition of
anxiety; if this is based on autonomic arousal, it may be that those with
borderline PD would often show high scores, but the usual assessment
methods may not identify the particular type of anxiety often experienced
by many patients with borderline PD.
Several studies have indicated that co-occurring PD can adversely affect
the response to treatment, course and outcome of anxiety disorders, as well
as of other syndromes such as major depression and obsessive-compulsive
disorder (Brooks et al., 1991; Reich & Green, 1991). In a study of the effect
of borderline PD on outcome in 110 outpatients with anxiety disorders or
(non-major) depressive disorders, borderline PD was present in 20%, and
those patients with co-occurring anxiety disorder and borderline PD had
the poorest outcome (Nurnberg, Raskin & Levine, 1989). It was noted that
PD must be recognized as a potential source of variance in treatment
studies of various disorders including anxiety disorders. A similar conclu-
sion was reported in a study of 52 patients with panic disorder, in which the
outcome of 8 weeks' treatment with benzodiazepines was negatively
associated with antisocial, borderline, histrionic and narcissistic PDs
(Reich, 1988). Also, the presence of a PD in a sample of 316 patients, many
with anxiety disorders, significantly impaired the outcome of treatment
with phenelzine (Tyrer et al., 1983). (Obsessive-compulsive PD was the
most prevalent PD diagnosis in this study.)
Schizophrenia and personality disorders

Eugen Bleuler (1911) introduced the term 'schizophrenia' in his book
Dementia Praecox or the Group of Schizophrenias, although he noted 'For
the sake of convenience, I use the singular, although it is apparent that the
group includes several diseases'. Dalen & Hays (1990) have pointed out that
subsequent authors have tended to 'vacilitate between assumptions of
homogeneity and heterogeneity', and that 'there is general agreement about
the importance of genetic factors in schizophrenia but no evidence shows
that genes are equally important in all cases'.
The division of schizophrenia into subtypes, together with their relation-
ships with unipolar depressions, bipolar affective disorders and schizo-
affective disorders, remains unclear. Crow (1986) has suggested that
schizophrenia and affective disorders are related to a genetic continuum,
while Cloninger et al. (1985) claimed 'a point of rarity' between symptoms
of schizophrenia and other psychiatric disorders. A review by Pope &
Lipinsky (1978) considered that 'schizophrenics of the good prognosis
category show typically two to three times as much familial affective illness
Schizophrenia 143
as schizophrenia; poor prognosis groups show a two to three fold difference
in the opposite direction'. Such claims support the view of Murray &
O'Callaghan (1991) that 'the appropriate genetic distinction is not between
schizophrenia and affective psychosis, but rather between acute schizo-
phrenia, schizoaffective disorder and affective psychoses on the one hand
and chronic schizophrenia and schizotypal personality on the other'.
The influential definitions of schizophrenia in the DSM-III-R and ICD-
10 each produces a relatively broad category which incorporates the more
chronic disorders, so that most research has been based on heterogeneous
patient groups. Broadly-defined schizophrenia has been shown to be
associated with a higher-than-expected history of obstetric complications
(which is itself associated with earlier onset of schizophrenia, male gender
and chronic course), of minor physical abnormalities (postulated to
indicate abnormal development during the second trimester of foetal life
and itself associated with earlier onset, male gender and cognitive impair-
ment), and of childhood behavioural abnormalities. The latter include poor
social adjustment and premorbid schizoid or schizotypal features. Patients
with schizophrenia and a history of such childhood abnormalities are
associated with an early onset of illness, male gender, a family history of
schizophrenia or a history of obstetric complications. (The findings in
relation to gender and schizophrenia have been reviewed by Goldstein &
Tsuang, 1990.)
Murray & O'Callaghan (1991) have reviewed the evidence for the
identification of a 'neurodevelopmental' or 'congenital' subtype of schi-
zophrenia resulting from a genetic defect (associated with smaller temporal
lobe structures and neuropathological changes in the hippocampus) or
from an early adverse effect on brain development in foetal or neonatal life,
perhaps due to obstetric complications or maternal influenza. It was
suggested that this neurodevelopmental subtype is particularly likely to
show social impairment in childhood, such as unusual patterns of thought
and communication as early as the age of 5, together with abnormal
premorbid personality in adolescence. It was also hypothesized that some
other subtypes of schizophrenia are genetically related to some affective
disorders.
Bleuler (1911) was the first to describe features reminiscent of schizo-
phrenia in some relatives of patients with schizophrenia, and the convincing
evidence for an association between schizoid (and schizotypal) PD features
and schizophrenia in genetically related families has been reviewed in
Chapter 2, together with the evidence for the association between premor-
bid schizoid (and schizotypal) PD and the subsequent development of
schizophrenia. Also, Dalkin et al (1994) found that premorbid explosive
and paranoid traits were commoner in patients with schizophrenia than in

patients with other non-organic psychoses, and these features were asso-
ciated with later onset of schizophrenia.
Obsessive-compulsive disorder and personality disorders

Obsessive-compulsive disorder (OCD) has a lifetime prevalence of about
2.5% (Karno, Golding & Sorenson, 1988). It is a disorder with maladaptive
recurrent or persistent thoughts (obsessions), with or without repetitive or
rigidly-patterned behaviours (compulsions) (Dowson, 1977). When severe,
this disabling disorder usually follows a chronic course with exacerbations,
and is associated with secondary anxiety and depression.
In the account of obsessive-compulsive PD in Chapter 2, it was noted
that there have been many claims that obsessive-compulsive PD is asso-
ciated with OCD. But despite the similarity of the terms, 'the majority of
patients with OCD do not meet criteria for obsessive-compulsive PD'
(Pfohl & Blum, 1991), although compared with normal controls, patients
with OCD are more likely to have a PD (Black et al., 1993). These are the
conclusions of most of the studies that have reported on retrospective
evaluation of PD in groups identified by the presence of OCD, although
Rasmussen & Tsuang (1986) found that 55% of 44 outpatients with OCD
met DSM-III criteria for compulsive PD. Joffe, Swinson & Regan (1988)
compared PD in 23 patients with OCD and in a matched group with major
depressive disorder, and found no significant difference between the two
patient groups in respect of the frequency or type of PD as assessed by the
self-report Millon Clinical Multiaxial Inventory. In the OCD group, a
mixed PD (i.e. with features of more than one PD) was the most common
finding; avoidant and passive-aggressive PDs were found in 44%, and
dependent PD in 35%. Only 4% met criteria for obsessive-compulsive PD,
while 17% had co-occurring schizotypal PD. An investigation with another
self-report instrument (the PDQ) also failed to provide evidence of a
specific association between obsessive-compulsive PD and OCD, and the
21 patients with OCD had more DSM-III-R 'cluster B' PDs than a 'normal'
control group (Black et al., 1989). Also, two groups of patients with OCD
and panic disorder respectively, had a similar PD profile, with avoidant,
dependent, histrionic and borderline PDs as the most frequent (Mavissaka-
lian et al., 1990a), and Sciuto et al. (1991), in another study of two
diagnostic groups based on OCD and panic disorder, also concluded that
no specific PD was associated with either group.
The various studies that have addressed this issue have used various
assessment methods, mainly involving self-report questionnaires, and the
Eating disorders 145
patient groups have been selected in different ways. Baer et al. (1990) have
provided one of the few reports based on a structured interview for DSM-
III PDs, which was used to assess 96 patients with OCD; 52% met criteria
for at least one PD, and mixed, dependent and histrionic PDs were the most
frequent. Obsessive-compulsive PD was diagnosed in only 6% of patients,
most of whom had an early childhood onset of OCD, which made the
distinction between PD and OCD difficult, while schizotypal PD was found
in 5%.
OCD appears to respond to drug treatments, such as some antidepres-
sants, and also to behavioural treatments, such as 'flooding' and 'response
prevention'. However, not all patients improve; in a 2-year follow-up study,
20% failed to show a treatment response, while many of the others
continued to show some symptoms and periodic exacerbations. Several
variables have been claimed to be related to outcome, such as depressive
symptomatology, lack of insight that some obsessional thoughts are not
realistic, and the presence of PD, in particular features of schizotypal PD
(Stanley, Turner & Borden, 1990).
Two retrospective studies have indicated that when schizotypal PD co-
occurs with OCD, this is a significant predictor of poor treatment response
to both drugs and behavioural treatment methods (Jenike et al., 1986;
Minichiello, Baer & Jenike, 1987). Jenike and colleagues reported schizoty-
pal PD (retrospectively assessed) in 33% of their sample with OCD, while
Stanley et al. (1990) studied 25 patients with OCD with the Structured
Clinical Interview for DSM-III-R PDs and found that 28% had schizotypal
features, although only 8% met criteria for schizotypal PD. Co-occurring
schizotypal PD features may identify a subgroup of OCD patients with
relatively severe social maladjustment, delusional experiences and a history
of abnormal visual experiences such as perceptual distortions. Such
patients may be relatively treatment-resistant, although some of the
apparent schizotypal features (such as delusional ideas) in OCD may be
part of the OCD itself and improve with treatment of the OCD.
Eating disorders and personality disorders

The main distinction between subtypes of psychogenic eating disorders has
been between anorexia nervosa (AN) in patients who have maintained a
low weight by a restrictive diet, perhaps with excessive exercise (who are
known as 'restricters' or 'dieters'), and AN in patients who have shown
various combinations and patterns of bulimia with binge-eating, self-
induced vomiting and laxative abuse (who have been classed as 'vomiters
and purgers', 'bulimics', 'bulimic anorectics', 'low-weight bulimics' or
'non-restricters'). It has also been suggested that a distinct subtype of

bulimia is bulimia nervosa (BN), which is associated with relatively small
deviations from normal weight (i.e. 'normal-weight bulimia'), and low-
weight bulimics may be more closely related to normal-weight bulimics
than to restricters. Most studies of AN subtypes have involved bulimia as
the distinguishing feature, although vomiting and purging have also been
used (Dowson, 1992c).
It has been suggested that the nature and degree of co-occurring
personality disorder (PD) can have major effects on the prognosis of eating
disorders and should be an important factor in deciding treatment strategy
(Rosenvinge & Mouland, 1990). Borderline PD has been reported to be
common in patients with eating disorders (Gartner et al., 1989) and several
studies have described a variety of PD psychopathology associated with
bulimia, i.e. histrionic, borderline, schizoid, schizotypal, antisocial, narcis-
sistic, obsessive-compulsive, avoidant, dependent and passive-aggressive
PD features (Dowson, 1992c), although Pope et al (1987) concluded that
their results argued against a specific relationship between bulimia and
borderline PD in patients with normal-weight bulimia. However, although
the latter study found no difference in borderline PD ratings between
patients with bulimia and another group with major depression, both these
groups had higher borderline ratings than the non-patient control group.
Also, a relatively high prevalence of obsessive-compulsive PD traits has
been described in 'restricters', and it was suggested that this can be of
aetiological significance (Beumont, George & Smart, 1976).
Several studies have reported personality differences between 'restricters'
and patients with AN together with a history of bulimia or vomiting and
purging. Beumont et al. (1976), using unstructured interviews, investigated
31 patients divided into 'dieters' and 'vomiters/purgers', although oc-
casional vomiting did not disqualify a patient from the former group.
Histrionic traits (considered as childish behaviour, affective instability,
evanescent enthusiasm, dramatic exaggerations and the impression of
emotional shallowness) were much more prevalent in the 'vomiters/
purgers', and obsessional traits (i.e. reliability, precision, scrupulousness,
conscientiousness and love of order and discipline) and social withdrawal
were more common in the 'dieters'. Casper et al. (1980) studied 105
inpatients with AN using the Minnesota Multiphasic Personality Inventory
(MMPI) and found that frequency of bulimia was associated with various
MMPI scale scores, including the psychopathic deviate scale. In contrast to
the fasting (i.e. 'restricter') AN patients, those with bulimia were more
extroverted and more likely to steal. Garfinkel, Moldofsky & Garner (1980)
Eating disorders 147
reported a series of 141 patients with AN who were divided into 'restricters'
and 'bulimics' on the basis of recent history at the initial consultation,
although some 'restricters' had a history of vomiting. Various intergroup
differences were found, including an association between bulimia, premor-
bid obesity and family history of obesity. Also, the bulimic group were
more outgoing, less isolated and more involved sexually. Their moods were
more labile and they showed more impulsive behaviour such as self-harm
(suicide attempts and self-mutilation), stealing and abuse of alcohol and
street drugs. Piran, Lerner & Garfinkel (1988) described similar findings
using the Diagnostic Interview for Borderline Patients as well as structured
interviews: 60% of the 'restricters' received a diagnosis of avoidant PD, and
the most common PD diagnosis in the 'bulimic' group (i.e. in those with AN
and bulimia) was borderline PD in 55.3% compared with only 6.6% in the
'restricters'. However, the consensus in the above studies was not sustained
by Gartner et al. (1989) in their report of DSM-III-R PD in 35 inpatients
using a structured interview, as they found no evidence of significant PD
differences between 'restricters' and those with 'low-weight bulimia',
although the number of 'restricters' was small (n = 6). Therefore, although
most reports that have compared subtypes of AN have found that 'low-
weight bulimia' is associated with PD characteristics related to impulsive
and harmful behaviour, this was not confirmed in the only study that used a
standard structured interview method for the assessment of all the DSM-
III-R PDs. But in a study of 55 patients with a history of AN, assessed by a
modified version of the revised self-report Personality Diagnostic Ques-
tionnaire (PDQ-R), based on DSM-III-R PDs, 'vomiters' showed signifi-
cantly higher scores than non-vomiters on self-report measures of border-
line and antisocial PD criteria (Dowson, 1992c).
PD has been assessed by a structured interview for DSM-III PDs in a
large sample of 210 women seeking treatment for AN (n = 31), BN (n = 91),
or a mixed disorder (n = 88) (Herzog et aL, 1992), in which the prevalences
of PDs were low compared with previously-reported similar samples.
Twenty-seven per cent had at least one PD, and the most frequent was
borderline PD in 9%. The highest prevalence of'at least one PD' was found
in the group with a mixed disorder at 39%, followed by 22% in the AN
group and 21 % in the BN group. There were higher rates of borderline PD
in the mixed and BN groups than in the AN group, and higher rates of the
'anxious' PD cluster (i.e. DSM-III-R's 'Cluster C ) in the AN and mixed
group. Those patients with co-occurring PD had significantly slower
recovery rates than those without PD, and the question of whether
assessment of PDs can have predictive value for the long-term course and
outcome of eating disorders is important, in view of the frequency of eating

disorders and the cost of providing treatment regimes for the most severely
affected. A high rate of eating disorders has also been found in a sample of
130 patients referred on the basis of PD to a therapeutic community (Dolan
etal., 1994).
Lacey (1984) has also found that, in patients with 'normal-weight BN', a
poorer prognosis was associated with co-occurring PD, particularly if the
PD was associated with alcohol abuse. This is in accord with thefindingsof
Yates et al. (1990), who reported that while BN and some examples of
alcohol abuse share intermittent behavioural loss of control of oral intake,
a comparison of 30 women with BN, and 30 age-matched women with
alcohol abuse, found that those with BN were less likely to have antisocial
or borderline PD features. Also, Fahy, Eisler & Russell (1993) found that
PD was related to poor treatment outcome with cognitive behavioural
therapy for BN.
Sohlberg et al. (1989) reported that impulsivity was the strongest
predictor of outcome in their study of patients with AN or BN, and that this
variable accounted for 25% of the variance of AN symptoms at 2 to 3 years'
follow-up, and for 14% at 4 to 6 years, as measured by the Eating Attitudes
Test. Also, in a study of 30 patients followed for 4-5 years, the results
suggested that borderline PD was predictive of poor eating disorder
outcome (Wonderlich et al., 1994). Sohlberg (1990) suggested that, for
those patients with eating disorders and co-occurring PD, non-specific
therapeutic efforts to help patients cope with stressful life events may result
in measurably reduced eating disorder morbidity.
Increased use of health care facilities and personality disorders

Many of the behaviours associated with PDs would be expected to be
reflected by an increased use of health care services and this was demon-
strated by Reich et al. (1989) in a non-patient survey of 401 randomly
selected subjects who showed a correlation between DSM-III 'cluster B' PD
features and visits to the family doctor with symptoms of other psychiatric
disorders. Also, Nestadt et al. (1990), in a study of DSM-III histrionic PD
in a community sample, found that individuals with this disorder tend to
use health care facilities more frequently than the rest of the population.
Complaints of chronic pain and personality disorders

The term 'idiopathic pain disorder' has been used when there is no clearly
identifiable somatic disorder (Williams & Spitzer, 1982), while Blumer &
Post-traumatic stress disorder 149
Heilbronn (1982) have suggested that this can be a feature of some
depressive disorders. Von Knorring, Almay & Johansson (1987) have
compared aspects of personality in patients with idiopathic pain disorder
and other groups, including healthy volunteers, and reported that certain
personality features were associated with idiopathic pain, including high
scores on an 'inhibition of aggression' factor and low scores on impulsivity
and suspicion.
It is likely that complaints of pain can be associated with various
psychiatric disorders in addition to depressive disorders, such as anxiety
states, conversion disorders and PDs. Jensen (1988) investigated chronic
pain in 73 non-psychotic psychiatric outpatients and found that in the
group with pain (n = 54), unskilled work status and a history of various
psychiatric diagnoses including PD were significantly overrepresented, but
it must not be assumed that idiopathic pain is always secondary to a
psychiatric disorder.
Bowel disorders and personality disorders

Several studies have claimed associations between personality traits, PDs,
and various bowel diseases but the interpretation of such findings is
unclear. Chronic inflammatory bowel disease has been associated with
introversion (Robertson et ai, 1989), while personality has been suggested
as a possible risk factor in large bowel cancer (Kune et ai, 1991). In the
latter study of 637 confirmed new cases of colorectal cancer and 714
controls, self-reported suppression of reactions that may offend others and
the avoidance of conflict showed a significant difference between cases and
controls.
Neuropsychological impairment and personality disorders

Malloy et al. (1990) have pointed out that those with antisocial PD may be
at risk for the development of neuropsychological impairment due to
alcohol or illicit drug abuse, and for a higher prevalence of head injury
associated with alcohol abuse, violence and risk-taking behaviour. Also,
Judd & Ruff (1993) have claimed that borderline PD is associated with
neuropsychological dysfunction, involving visuospatial tasks.
Post-traumatic stress disorder (PTSD) and personality disorders

This syndrome includes depression, anxiety, sleep disorder, distressing
memories and images, and irritability, and follows severe stressful life
events, such as combat, natural disaster, criminal assault or accident
(McCarroll et ah, 1993). Its medico-legal importance has been increasing,
and it can be presented as a disorder that justifies claims for damages or as a
mitigating factor for criminal behaviour.
Borderline PD, as well as a history of stressful life events and childhood
maltreatment, are often found to co-occur with PTSD, although reasons
for these associations are unclear. Famularo, Kinscherff & Fenton (1991)
have suggested that PTSD in children can present with a syndrome that is
similar to borderline PD in adults. Southwick, Yehuda & Giller (1993)
found PDs to be common in combat veterans with PTSD, in particular
borderline, obsessive-compulsive, avoidant and paranoid PDs.
Sierles et al. (1983) described 25 combat veterans in hospital for PTSD
and found that the majority had at least one additional psychiatric
disorder, in particular alcoholism, drug dependence, antisocial PD and
somatization disorder. A more recent study of 189 male Vietnam veterans
admitted to an inpatient treatment programme found that 72% had PTSD,
and that this group was significantly associated with passive-aggressive,
schizoid, avoidant and borderline PD features (Sherwood, Funari &
Piekarski, 1990). As in the previous study, drug and alcohol abuse
commonly co-occurred.
It is likely that co-occurrence of PDs, in particular borderline PD, may be
a predisposing factor for the development of PTSD and can affect the
course of the disorder (Gunderson & Sabo, 1993ft).
Attention deficit hyperactivity disorder (ADHD) and personality disorders

Hyperactivity in children is an heterogeneous syndrome which has been
termed 'hyperactive syndrome' or 'minimal brain dysfunction', as well as
ADHD. It has been suggested that ADHD is a predisposing factor for the
development of antisocial PD and criminality (Mannuzza et al., 1989), and
a review by Lie (1992) concluded that although 'hyperactives without
conduct problems do not have an increased frequency of delinquency,
ADHD in addition to another disorder indicates a more unfavourable
prognosis than for each of these disorders separately'.
In another literature review, Biederman, Newcorn & Sprich (1991) came
to the same conclusion, noting frequent co-occurrence of an history of
ADHD with conduct disorder, borderline PD and other psychiatric
disorders. In a prospective follow-up of 103 males with ADHD, Mannuzza
et al. (1989) considered that their findings also supported the view that
childhood ADHD is a risk factor for later criminality when co-occurring
Antisocial behaviours 151
with other disorders. ADHD in adults has been reviewed (Editorial, 1994)
in relation to a report that methylphenidate and d-amphetamine were
associated with significant improvements in behaviour in adults with the
diagnosis of ADHD (Matochik et al, 1994). However, this is a contro-
versial diagnosis in adults and is becoming increasingly frequent as a self-
diagnosis in the setting of chronic difficulties.
Antisocial behaviours and personality disorders

West (1983) has noted that some antisocial behaviours are common among
young males, and that unusually frequent or serious misbehaviour (as
admitted in response to self-report questionnaires), is associated with
having or subsequently obtaining a criminal conviction. Among males in
the UK, the incidence of convictions falls after the age of 18, although the
number of juvenile convictions, and a relatively young age at first convic-
tion, predict a greater risk of a persisting record of convictions. In a
community study of a sample of males from a region of London (West &
Farrington, 1977), most of those with a history of criminal convictions had
histories of at least four out of the following features: gambling, driving
after drinking, immoderate smoking, sexual promiscuity, unstable work
record, anti-establishment attitudes, involvement in acts of violence,
unconstructive use of leisure, association with antisocial groups, use of
prohibited drugs and having tattoos. But it has been argued that persistent
delinquents, whose antisocial attitudes and behaviour are pervasive and
frequent, are not just ordinary members of under-privileged groups in
society but have a disorder (usually a PD) that would bring them into
conflict with any society.
The relationship between PD and pathological gambling (which involves
chronic and increasing inability to resist impulses to gamble, with disrup-
tion or damage to family, personal or vocational pursuits), has also been
investigated. In a comparison between 19 male gamblers and controls, a
number of differences in measures of personality were found, such as a
higher total hostility score in gamblers (Roy et al., 1989#). Also, there have
been reports of biological correlates, involving abnormal functioning of the
central noradrenergic system (Roy, De Jong & Linnoila, 19896).
Other studies have identified groups of males with a history of involve-
ment in spouse (or partner) abuse, and examined possible correlates. A
literature review has suggested that there are associations with several PDs,
and subtypes have been defined (Hamberger & Hastings, 1988). The most
frequent co-occurring disorders in another study were sadistic, antisocial
and borderline PDs (Hart, Dutton & Newlove, 1993). Not surprisingly,
alcohol-abusing 'batterers' have been identified as a particularly difficult
group to manage. Some batterers may be so dangerous that the provision of
residential treatment programmes may be indicated.
Dangerous, violent or seriously irresponsible behaviour and personality

disorders
Many examples of dangerous, violent or seriously irresponsible behaviour
may involve both the police and health care services. Antisocial PD is partly
defined by such behaviour that may start in childhood and is more common
in men. Also, a study of 91 female offenders admitted for psychiatric
assessment found that early age of onset of criminal behaviour was
associated with adult PD (Brownstone & Swaminath, 1989).
Many examples of dangerous behaviour are unknown to police or
medical services or may be called 'accidents'. A study of male alcoholics
compared 57 who reported personal injury accidents with the 131 who did
not; the former group had more severe alcoholism, an earlier onset of heavy
drinking, and a higher prevalence of antisocial PD (Yates et al, 1987). Also,
in a study of several hundred male drivers referred by courts for a treatment
programme for alcoholism, the 'heavy' drinker subgroup had more anti-
social characteristics, a lower educational level and a greater tendency to
drop out of treatment programmes (McGuire, 1980). In 1990 there were
46 300 deaths from motor vehicle accidents in the USA, accounting for 50%
of all accidental deaths, and providing the fourth leading cause of death in
the USA, above infections. Many of these deaths are likely to result from
behaviour related to PDs.
Inadequate and abusive parenting and personality disorders

Evidence has been found for associations between antisocial PD in parents
and psychiatric disorders in their children, which is likely to be mediated, at
least in part, by bad parenting. Comparing child psychiatric clinic attenders
with other groups, an excess of a history of psychiatric hospital referral
among parents of behaviourally disturbed children has been noted,
together with a significant excess of parental PD, especially antisocial PD.
Also, it has been considered that parental behaviour, in particular marital
discord and violence, is often an important cause of conduct disorder,
particularly in boys (Wolff, 1983).
In a study of 54 mothers who had maltreated their children and 37
Epidemiology 153
controls, the former group showed a greater prevalence of PD as well as
mood disorder and alcohol abuse. Also, mothers who had maltreated their
children were more likely to have a history of post-traumatic stress disorder
(Famularo, Kinscherff & Fenton, 1992), while a highly abusive home
environment has been associated with one or more of the triad of enuresis,
firesetting and cruelty to animals (Prentky & Carter, 1984).
Sexual deviation disorders and personality disorders

Legal constraints on permissible sexual behaviour, and possible 'treatment'
for those individuals with sexual deviation disorders, have to be considered
when the well-being or rights of others are compromised (Berlin &
Meinecke, 1981). It does not appear that there are specific associations
between sexual deviation disorders and PD, but co-occurring PD, in
particular antisocial PD, is likely to be a risk factor in contributing to the
occurrence of criminal behaviour in the context of sexual deviation.
However, a study of transsexuals found a greater prevalence of PDs, mainly
within 'cluster B' compared with a control group (Bodlund et al., 1993).
Epidemiology
Methodological problems
The literature provides a range of prevalence rates and age or sex
distributions of the various PD syndromes, but this is to be expected in view
of methodological problems in their assessment and the use of different
selection criteria for groups of subjects. Morey & Ochoa (1989) compared
the DSM-III-R PD ratings of 291 clinicians for recent patients seen by each
(one patient per clinician), with their PD diagnoses based on global
impression. It was found that there was only a modest correlation between
the two procedures, producing a kappa of only 0.58 between the clinicians'
own diagnoses of borderline PD based on clinical impression, and their
diagnoses based on their own assessments of the individual DSM-III-R PD
criteria. This indicated that, in routine practice, there is often a failure to
systematically assess the individual DSM-III-R criteria, even though the
DSM diagnostic categories may be used. A clinician may often obtain an
initial impression of the most obvious feature of PD and then focus the rest
of the assessment on one or two probable PD diagnoses. However, even
though the DSM PD categories have often shown poor inter-rater re-
liability and stability, it is generally recognized that their features are
'highly prevalent in psychiatric settings and consume a significant propor-
tion of mental health resources' (Mulder, 1991). Also, epidemiological
studies have shown a surprising consistency for many aspects of PDs in
clinical practice.
Gender and PD
DSM-III and DSM-III-R made various claims in relation to gender and PD
syndromes, which were not all based on experimental evidence; males were
considered to be more prone to develop paranoid, antisocial and obsessive-
compulsive PDs, while increased prevalences of histrionic, borderline and
dependent PDs were associated with females. The clearest evidence for any
of these claims exists for antisocial PD in males, but an association between
male gender and paranoid or obsessive-compulsive PDs has also been
supported, as has an association between female gender and borderline,
histrionic and dependent PDs. But the evidence is conflicting: the latter
associations with females were not found in a series of 298 outpatients
(Alnaes & Torgerson, 1988a) in which passive-aggressive, schizotypal and
narcissistic PDs were more prevalent among men; these findings had not
been predicted by DSM-III/DSM-III-R.
Prevalence rates of individual PDs

Estimates of the lifetime prevalence of DSM-III-R antisocial PD have
ranged from 0.5 to 3.5% in non-patient community samples, and this
disorder has been associated with male gender, younger age (less than 45
years), lower socioeconomic status and urban residence (Jordan et al,
1989). For example, Swanson, Bland & Newman (1994) reported 104
subjects with antisocial PD out of 3258 randomly-selected adult household
Canadian residents; the overall lifetime prevalence rates for antisocial PD
were 6.5% for males and 0.8% for females. In those with antisocial PD, the
age at which conduct disorder appeared was usually under 10 and there was
an increased lifetime prevalence of co-occurrence with nearly every other
psychiatric disorder. Antisocial PD was most often found in the 25-34 year
age group, followed by the 18-24 age group.
For borderline PD, estimates of prevalences in non-patient community
samples and in the relatives of non-patients have been between 1 and 4%
(Loranger, Oldham & Tullis, 1982; Baron et al.9 1985; Maier et al., 1992),
but for inpatient groups higher rates of between 15 and 25% have been
Epidemiology 155
reported. In various studies it appears that two-thirds or more of subjects
with this disorder have been female (Gunderson & Zanarini, 1987). In a
study of 1583 inpatients there were no significant differences in the
prevalences of borderline PD among racial groups classed as whites, blacks
and hispanics (Castaneda & Franco, 1985).
Histrionic PD as defined by DSM-III criteria was found to have a
prevalence of 2.1% in a non-patient community sample (Nestadt et al.,
1990); males and females were equally affected, and those with this
syndrome tended to use health care facilities more frequently than others.
Data from the same study indicated a prevalence for DSM-III compulsive
PD (obsessive-compulsive PD) in 1.7%; male, white and employed subjects
received this diagnosis more often than the complementary groups, and it
was concluded that this PD is associated with a vulnerability for the
development of anxiety disorder (Nestadt et al., 1991).
Many studies have examined prevalence figures for at least one PD in
non-patient community samples. Srole et al. (1962) reported 1660 adults,
assessed by a self-report method, in whom 10% were considered to have
'probable PD', while more restrictive criteria have given rates below 4%
(Weissman, Myers & Harding, 1978; Myers et al., 1984). In the UK, 200
randomly chosen adults were assessed by a structured interview and PD
was found in 13% (Casey & Tyrer, 1986). In this study 'explosive' and
'anankastic' categories were most common (6% and 3% respectively),
which probably reflected borderline and obsessive-compulsive PDs. Sum-
marizing the literature, Casey (1988) noted that a lifetime prevalence for PD
of between 2 and 18% has been found depending on the method of
assessment and the populations studied, and that PD is associated with
young adult age and the male gender. Reich, Nduaguba & Yates' study
(1988) of age and sex distribution of DSM-III PDs suggested that these
generalizations conceal different relationships between age, sex and the
various PD syndromes. This study found that the prevalences of schizoid,
schizotypal and paranoid PDs did not show a change with age, while the
number of features of antisocial, narcissistic, borderline and histrionic PDs
(and to some extent the 'anxious' cluster C PDs) declined from younger to
older groups. The highest levels of PD features in men were found between
the ages of 18 and 30, while in women the highest levels were in the 31 to 40
age range. Although some studies found 'one or more' PDs to be more
prevalent in inner-city areas, Blazer et al. (1985) did not find any urban-
rural difference in the prevalence of antisocial PD. These authors studied
groups in five parts of the USA and also found a decrease in 'one or more'
PDs with increasing age and a higher prevalence in men. There was also an
association with poor educational attainment.
Maier et al. (1992) investigated DSM-III-R PDs by an interview assess-
ment in 452 non-patient 'community' subjects who they considered to be a
'good approximation' of a representative sample of the population. Of the
males, 9.6% (and of the females, 10.3%) had at least one PD; older subjects
were less likely to have PD than younger subjects; mixed PDs (i.e. more
than one PD) were found in 2.6% of females and 2.3% of males; obsessive-
compulsive PD was the most prevalent PD in males, while dependent and
passive-aggressive PD were the most frequent in females; only borderline
PD showed a clear tendency to co-occur with other PDs; and dependent,
passive-aggressive and histrionic PDs were more prevalent in females, while
borderline and avoidant PDs showed a similar but less marked association.
Obsessive-compulsive, schizotypal and antisocial PDs were found more
often in males. Also, anxiety disorders were associated with avoidant PD,
and affective disorders with borderline PD. It was noted that there was a
low prevalence of antisocial PD and schizophrenia in the sample, so that
other expected associations with these disorders could not be evaluated. It
was suggested that associations between PDs, major depressive disorders
and anxiety disorders may be less marked in non-patient samples, and that
the low prevalence of antisocial PD represented a selection bias in this
study.
These findings can be compared with those of Coryell & Zimmerman
(1989), who also examined a non-patient sample, although DSM-III
diagnoses were used. In this study, 9.2% of relatives of healthy controls had
at least one PD; but this is likely to be an underestimate of the prevalence of
PD in the general population as the sample was selected in relation to a
healthy control group. Nevertheless, the prevalence of PD in the general
population is similar in the two studies, despite the differences in method-
ology, and points to a frequency of about 10% for significantly maladaptive
PD in adult samples of the population.
Zimmerman & Coryell (1989) also reported a large non-patient sample
made up of 797 'normal' controls and offirst-degreerelatives of patients
with a variety of psychiatric disorders; of those with a PD, about 25% had
more than one, and the most prevalent diagnoses were mixed (3.6%),
passive-aggressive (3.3%), antisocial (3.3%), histrionic (3.0%) and schizo-
typal PD (2.9%). Half the subjects with avoidant PD also met criteria for
schizotypal PD. Paranoid, schizotypal, avoidant and borderline PDs most
frequently co-occurred with other PDs, and schizoid and dependent PDs
occurred relatively often in isolation.
Epidemiology 157
PD in primary medical care settings

When an individual consults a primary care physician, he or she can be
selected for assessment on this basis alone or because of the presence of
'conspicuous psychiatric morbidity'. Casey, Dillon & Tyrer (1984) used this
latter criterion in their study of a sample of patients in urban general
practice, in which the primary care physician rated PD as the primary
(main) diagnosis in 8.9%; the corresponding prevalence diagnosed by a
psychiatrist who made parallel ratings was 6.4%. But subsequent detailed
assessment by a structured interview suggested that no less than 33.9% had
at least one PD, in particular 'explosive' PD, which was diagnosed in 17%.
This suggests that a more detailed assessment of PD in primary care would
yield considerably more PDs than are apparent in routine assessments. It is
likely that this hidden PD psychopathology is the starting point for many
presenting symptoms in primary care settings (Casey & Tyrer, 1990).
PD in psychiatric inpatient and outpatient samples

In England and Wales in the mid 1980s, 7.6% of all psychiatric admissions
had a diagnosis of PD with the ICD classification (Department of Health &
Social Security, UK, 1985). This is likely to have been a considerable
underestimate as any non-PD psychiatric disorder would usually have been
recorded in preference to a co-occurring PD. The DSM-III and III-R have
encouraged the recording of co-occuring diagnoses, and co-occurring PD
has been identified in 40-50% of many inpatient psychiatric samples. PD is
the sole diagnosis in a minority of inpatients, and more often a PD co-
occurs with other psychiatric disorders; for example, Cutting and col-
leagues (1986) assessed a series of consecutive admissions with non-PD
disorders to an urban psychiatric unit and found co-occurring PD in 44%.
Mulder (1991) reported on admission trends for those with a principal
(main) diagnosis of PD admitted to New Zealand psychiatric hospitals over
a 7 year period. The most common diagnostic category was unspecified PD
(45%), followed by dependent PD (12%). Other rates included antisocial
PD (10%), histrionic PD (9%), explosive (i.e. similar to borderline) PD
(6%), schizoid PD (5%), paranoid PD (3%) and obsessive-compulsive PD
(2%). Antisocial PD was more common in men (x 3.2), as were paranoid,
schizoid and explosive PDs, while histrionic PD was more frequently
diagnosed in women. In general, the patients with PD were relatively
young; 73% were less than 35 years and more than 40% were under 25. Also
in relation to inpatients, Jackson et ah (1991) found that schizophrenia was
associated with antisocial and schizotypal PDs, manic disorder with

histrionic PD, and unipolar affective disorder with borderline, dependent
and avoidant PDs. The nature of these relationships is unclear but it is likely
that a subgroup of those with unipolar depressive disorder consists of
subjects in whom the PD is a major causal factor for the subsequent
development of mood disorder.
Groups of individuals from outpatient clinics in psychiatric services have
been associated with lower rates for the prevalence of PD, compared with
inpatients, but substantial rates of PD have been reported, often between
20-40% (Kass, Skodol & Charles, 1985; Morey, 1988). In an outpatient
sample selected on the basis of a diagnosis of'neurosis' (i.e. mainly anxiety
disorders and subtypes of depressive disorders), Tyrer and colleagues
(1983) found that 39% had PDs, in particular anankastic (obsessive-
compulsive) and passive-dependent (dependent).
PD in forensic samples
For antisocial PD, various prevalence rates up to 70% have been reported
in forensic samples (Gunn et al., 1978).
PD in self-harm populations
A wide range of prevalences of PDs have been found in self-harm
populations (Casey, 1988), in particular when the risk to life is judged to be
relatively low (Pallis & Birtchnell, 1977). Relatively serious risk has been
associated with obsessive-compulsive traits (Murthy, 1969) but this was not
confirmed in a subsequent study (Pierce, 1977).
4
Longitudinal aspects of personality disorders
J.H. DOWSON
Introduction
The longitudinal perspective

The course of a disorder is examined by data collection at a minimum of
two points in time, but an assessment on several occasions is required to
identify a fluctuating course; if only two sets of data are available, their
comparison is merely an outcome study. Many variables affect the course of
PDs, including various interventions, other co-occurring disorders and
secondary effects, such as a break-up of family relationships or the loss of
employment.
The evaluation of the course of PDs is adversely affected by the
complexity of PD assessment, so that even if a detailed evaluation has been
carried out at one time-point, it may be difficult to obtain the subject's
cooperation for further time-consuming interviews. Consequently, most
longitudinal studies of PD have relied on relatively crude measures such as
re-admission rates and suicide. Also, a high drop-out rate is a problem in
any longitudinal study and, in relation to PD, has been associated with
males and a diagnosis of antisocial PD.
Most PD traits appear to have a dimensional relationship with aspects of
'normal' personality in non-clinical populations, and it has been shown that
several such traits show stability for the decades over the age of 30
(Zuckerman, 1991). In a study of 769 male volunteers aged 17-97, the factor
structure of several aspects of personality was similar in different age
groups (McCrae et ah, 1980) and a 9-year longitudinal study of 4942
subjects aged 25-74 provided further support for the view that some
personality features are enduring; it appeared that, for most 'normal'
personality traits, there is stability from the age of 30, which follows a
159
160 Longitudinal aspects of personality disorders
period of significant change from adolescence for some attributes (Costa &
McCrae, 1992).
While the literature on clinical or forensic samples with PDs also
provides evidence for varying degrees of stability of some PD features,
others have been shown to be subject to change. Such instability is of two
main types, namely rapid short-term, and slow, linear long-term. Short-
term changes in PD features are often in relation to life-crises or other
co-occurring disorders, (in particular depression, anxiety disorders and
substance abuse, Bronisch & Klerman, 1991), while slower long-term
improvements in adaptation, which are characteristic of many individuals
with features of borderline and antisocial PDs, can provide realistic hope in
many young patients with seemingly intractable problems. Long-term
improvements in adaptation have also been found in a non-patient sample
in relation to antisocial and histrionic PD ratings, and it seems that those
features of normal personality and PD which involve lability (i.e. rapid
changeability) are particularly associated with age-related changes for the
better (Tyrer & Seivewright, 1988ft). Therefore, changes with age are most
associated with features of DSM-III-R's 'cluster B' PDs, (and perhaps with
passive-aggressive PD), in contrast with 'cluster A' PDs and obsessive-
compulsive PD. The outcome of avoidant PD is unclear.
Longitudinal studies of PD have mainly involved borderline, antisocial
and schizotypal PDs (Stone, 1993). These have generally reported on
inpatient or prison samples, so that the results relate to relatively severe
examples of these particular syndromes, and to individuals who usually
have several co-occurring disorders. There have been relatively few long-
term prospective studies, and while Robins (1978) found clear evidence of a
link between childhood conduct disorder and adult antisocial PD, most
long-term reports have been concerned with borderline PD (Perry, 1990).
Perry (1993) has reviewed 26 longitudinal studies, which were mostly
related to inpatient samples and borderline PD. There was a bimodal
distribution of follow-up periods, (between 2-4 years and 13-16 years), and
only two studies had more than two sets of assessment data. Of the nine
studies which looked at suicide in patients with borderline PD, there was a
mean suicide rate of 6.1 % after a mean follow-up of 7.2 years, and it was
found that the highest risk period was in the year after 'index' admission at
which the patient was identified for study. With regard to the stability of
diagnoses,fivestudies found that a mean of 57% of those with borderline
PD had retained the diagnosis after a mean follow-up of 8.7 years. Also,
most subjects had 'significant symptoms and impairment in social function-
ing on follow-up'.
Introduction 161
Associated variables
Longitudinal studies of PD need to evaluate a range of variables that affect
and reflect course and outcome, other than the features of PD. Therefore,
evaluation of the course of a PD should, if possible, consist of a variety of
measurements at multiple points in time. Measurements should include
global assessments, such as those involving occupational and social func-
tioning, while repeated data sets can evaluate the rate of change, which can
be an important characteristic of a PD that is not addressed in an outcome
study. Of the variables that may influence course and outcome of PDs,
some will co-occur by chance (e.g. a bereavement), some may share
aetiological factors with the PD being studied (e.g. social isolation due to
schizoid PD), while others may be secondary to aspects of PD (e.g. some
mood disorders and substance abuse). Variables affecting the results of
longitudinal studies include the patient's demographic characteristics,
adverse environmental experiences, co-occurring disorders, the individ-
ual's positive or adaptive potential, and treatments or interventions.
Relevant demographic variables include age, age when first in contact
with psychiatric services, sex, socioeconomic status of subject and parents,
marital/relationship status, cultural background, position in sibship, twin
status, being an adoptee and being born or raised without two parents in the
household. An adverse environment and co-occurring disorders may have
involved childhood abuse (physical, sexual or psychological), childhood
mistreatment or neglect, loss events, stress, illicit substance abuse, a history
of other psychiatric disorders such as eating disorders, childhood conduct
disorder, physical disease, low intelligence and sexual disorders or difficul-
ties. Positive or adaptive qualities are seldom considered in PD research,
but are important variables in relation to course and outcome; Kolb (1982)
has specified courage, flexibility, commitment, perseverance, responsibi-
lity, humour, empathy, trust, charm and likeability. Also, a very high IQ
may be of some advantage, while talent, female beauty, social position and
wealth appear to predict a relatively good outcome for borderline PD
(Stone, 1990).
The evaluation of these variables is important for the further understand-
ing of the range of causal factors that contribute to the development, course
and outcome of PDs. Variables that are potentially amenable to interven-
tion, such as the adverse care of children, are of particular interest.
Outcome measures
Longitudinal studies have mainly involved measures of aspects other than
those directly related to PD characteristics (Perry, 1993). These include
hospitalization, other interventions from medical services, global function-
ing, social and occupational functioning, general interests, marriage (or
long-term live-in relationships), child rearing, symptoms of other psychi-
atric disorders and mortality.
Epidemiological data
Longitudinal data can be deduced from epidemiological studies of preva-
lence and incidence of PD features and associated variables in specified
populations. For example, the prevalence of antisocial and borderline PDs
in different age groups can indicate an improvement in these disorders with
increasing age.
Methodological aspects
In addition to the formidable methodological problems in the 'cross-
sectional' evaluation of PDs, longitudinal studies involving two or more
assessments are subject to additional sources of error. For example, the
criteria for PDs have been subject to frequent revision, available data in
retrospective studies have often been inadequate, and social and cultural
changes may have affected the pattern of the disorders. For instance,
changing patterns of drug misuse have provided a variable background to
the expression of antisocial PD traits. Also, the course of co-occurring
psychiatric disorders has to be taken into account when evaluating the
course of PDs; for example, agoraphobia with avoidant PD, depressive
disorders with various PDs, and obsessive-compulsive disorder with
obsessive-compulsive PD (Perry, 1993). Improvement or worsening of a co-
occurring disorder may influence PD-related behaviour, while improve-
ment in PD (such as the gradual improvement in adaptation found in many
subjects with antisocial and borderline PD) can result in a corresponding
change in the co-occurring disorders.
Because of the lack of a 'gold standard' for PD assessment, the use of
more than one PD assessment method has been recommended to increase
reliability. Discrepantfindingsmay then be assessed further by additional
methods (Perry, 1990).
It must be remembered that nearly all longitudinal studies involve groups
who have presented to medical services, so that there is little knowledge

about those who do not seek help. Some individuals with severe PD
disorders who do not seek help may have a worse course and outcome
compared with those who receive some intervention, although for others,
not seeking help may reflect co-occurring positive attributes, with a better
course and outcome.
Borderline personality disorder
Longitudinal studies of borderline PD

Most longitudinal studies of borderline PD have used inpatient samples,
consisting of those who have usually been referred to psychiatric services in
adolescence or the early 20s. Improvement was generally identified by 8-10
years after initial diagnosis, but a mean of 6.1% of patients (range 3-9%)
had committed suicide by a mean of 7.2 years (range 1.75 to 16) (Stone,
1992; Perry, 1993). The risk was greatest in the year following index
admission. After 10-30 years, the outcome for inpatient samples varied
from 'very good' to 'suicide', but was generally favourable with about 60%
of subjects functioning at least reasonably well with global outcome
measures, despite many examples of residual mild psychiatric disorders.
About a third of the well-functioning group, particularly women, had a
very good outcome, with no symptoms and the eventual achievement of
satisfying relationships (Stone, 1993). (In Stone's opinion, the outcome for
outpatients with a borderline PD is at least as good, and perhaps better.)
Five studies reviewed by Perry (1993) found that, after a mean of 8.7 years,
57% of patients still received the borderline PD diagnosis. The negative
correlation between lengths of follow-up and percentage retaining the
diagnosis suggested a linear model for at least a degree of remission that
eventually disqualified the subject from the borderline PD diagnosis. Such a
diagnosis is rarely made for patients over the age of 40. However, some
patients may re-present to psychiatric services in their 40s, with depressive
disorders following the break-up of close relationships that have taken the
toll of borderline PD behaviour over many years.
Several studies have involved inpatients who received a diagnosis of
borderline PD on the basis of the DSM or Gunderson's Diagnostic
Interview for Borderlines (DIB) (Stone, 1992). For example, Stone and
colleagues (Stone et ai, 1987; Stone, 1990) have reported longitudinal data
on 96% of 206 patients with borderline PD (male: = 61, female: = 145),
followed up for 10-25 years. These individuals were mainly of middle to
upper-middle socioeconomic status in New York State. Ninety-one per

cent were single and the average age at index admission was 22 (range 13
39). Fifty-two per cent were Jewish. The average length of hospital
admission was 12.5 months. These patients were admitted to a unit of the
New York Psychiatric Institute, which specialized in intensive, psychoana-
lytically orientated psychotherapy of adolescents and young adults, but
individual and group sessions were also accompanied by drug treatments in
most cases. The 206 subjects were drawn from a wider sample of 550 (the PI-
500), 25.7% of whom were presenting with their first admissions. Of those
traced, 8.5% had committed suicide after a mean of 16.5 years, which was
similar to the rates for the PI-500 patients with schizophrenia and manic-
depressive disorder. Stone has pointed out that this is equivalent to a rate of
616 per 100000 of similar patients with borderline PD, which is 55 times
more than the rate for Caucasians in the US general population. Therefore,
borderline PD is likely to be an important causal factor in suicide for
individuals in their 20s, and it must be noted that, between 1950 and 1975,
there has been a considerable increase in the suicide rate for young adults in
the second half of their 20s, i.e. from 11.9 to 25 per 100000 for white US
males, and from 5.1 to 8.0 in females (Solomon & Murphy, 1984). Other
results from the New York State longitudinal study included the finding
that the marriage rate was half the US average for the equivalent cultural
group, and the rate for childbearing was about a quarter of average. The
majority eventually lost the borderline PD diagnosis, but most of these
subjects still qualified for other PD diagnoses, such as histrionic, avoidant
and obsessive-compulsive PDs. After the index admission, the subsequent
readmission rate was about one-third of the rate for those PI-500 subjects
with schizophrenia. Another contrast with the schizophrenia group was
that patients with borderline PD were more likely to have had a job and
have achieved more occupational responsibility.
McGlashan (1986, 1993) has reported the follow-up of 81 patients who
had received a diagnosis of borderline PD by DSM-III or Gunderson &
Kolb's criteria (1978) at index admission to a private residential unit,
Chestnut Lodge in Maryland, US. The average length of follow-up period
was 15 years, with a range of 2-32. Initial diagnoses were made retrospecti-
vely, and most follow-up contacts were by telephone. Patients with
borderline PD were compared with those with schizophrenia or affective
disorder. Most of those with borderline PD were single females whose
disorder had presented significant problems in late adolescence and wor-
sened during their 20s. The index admission to Chestnut Lodge was usually
Borderline personality disorder \ 65
after the age of 20 but there was generally a history of previous contacts
with psychiatric services. There was usually at least moderate impairment
in social, occupational and sexual adjustment.
Outcome was, in general, superior in relation to those groups with
schizophrenia and bipolar affective disorder, and comparable to subjects
with unipolar depressive disorder. At follow-up, most were living indepen-
dently from parents or institutions and had reasonable work records.
However, many had needed short 'crisis' hospital admissions, nearly half
had subsequent outpatient psychotherapeutic support, and other psychi-
atric disorders were common, in particular depressive disorders and
substance abuse. The subsequent pattern of relationships showed consider-
able variation; while one group avoided intimacy, others managed to
achieve stable relationships. While the latter had sometimes achieved
satisfactory sexual relationships, others had developed social relationships
only, perhaps supplemented by partial intimacy. The degree of improve-
ment was related to the duration of follow-up, when data for the first 9 years
was compared with those for the period 10-19 years.
The remaining inpatient studies are those of Plakun et al. (1985), Kroll et
al. (1985), Paris et al. (1987) and Links (1993). Paris and colleagues
followed up 100 subjects for a mean of 15 years and found a similar rate of
suicide (8.5%) to that reported by Stone and colleagues. Another finding
that was comparable with similar studies was that 75% no longer qualified
for the borderline PD diagnosis. Links reported a 7 year follow-up of 88
patients with borderline PD diagnosed by the DIB; of 57 who were DIB-
positive initially, and reinterviewed, 52.5% were no longer DIB-positive,
and males were more likely to refuse the follow-up interview.
Prognostic factors
Stone (1990) has identified several attributes which were associated with a
relatively good prognosis, including cooperation with the self-help group
Alcoholics Anonymous (when appropriate), the presence of obsessional
personality traits with a capacity for self-discipline and work-orientation,
unusual talent, very high intelligence and, in women, a high degree of
'attractiveness'. Also, anecdotal evidence suggested that both negative and
positive chance events can affect the course of borderline PD, as subjects are
very sensitive to environmental changes; for instance, a patient who
unexpectedly survives a determined suicide attempt may never again carry
out self-harm behaviour. Also, Plakun (1991), in a follow-up study of 33
inpatients for a mean of 14 years, found that a good outcome was
associated with self-destructive acts during the index admission.
Poor outcome for inpatients with borderline PD has been particularly
associated with a history of parental cruelty and also with childhood
neglect, childhood sexual abuse (especially father-daughter incest) (Paris,
Zweig-Frank & Guzder, 1993), and a history of fire-setting in childhood
(Stone, 1993). Relevant aspects of more recent history that predict poor
outcome include relatively long periods of inpatient management (around
10 months or more), rape, imprisonment and poor global functioning.
At index admission, poor prognostic features have been the presence of
positive ratings for all eight DSM borderline PD criteria, co-occurring
features of antisocial PD with marked irritability and anger, schizotypal
PD features, being a male without a major affective disorder, severe
features involving impulsivity or affective instability, low intelligence and
being a male who discharges himself against medical advice (Links et al,
1990; Links, 1993; McGlashan, 1993). Subsequent to index admission,
continuing substance abuse has indicated poor outcome (Stone, 1993). This
can be 'deceptively mild' in women (Stone, 1990) and requires careful
evaluation.
Prognostic factors in relation to suicide deserve particular attention as,
on the basis of one large series (Stone, 1990), a prediction can be made that
'4 out of 10 hospitalized "eight-item" borderline PD patients, in their early
20s or younger, will commit suicide within 5 years ...'. ('Eight-item'
patients are those who are rated positive on all eight DSM-III-R criteria.)
This study found a similar risk of suicide in the follow-up period (37.5%)
for patients with a combination of borderline PD, major affective disorder
and alcoholism. Also, the rate was 19% in a subgroup with continuing
alcohol abuse and was higher in those who refused to attend self-help
groups. Co-occurring antisocial PD increased the risk of subsequent suicide
by a factor of 3 compared with the risk for borderline PD alone. For the co-
occurrence of borderline PD and 'major depression', the suicide risk was
doubled for men only.
Other studies have reported similar findings; for instance Zilber et al.
(1989) found that in a group with various PDs together with drug or alcohol
abuse, the risk of suicide was 3 times (21 %) that of a group with PD alone.
Suicide in patients with borderline PD has also been associated with
previous imprisonment (in a white middle-class sample), more frequent
childhood loss, lack of treatment before hospitalization, longer hospitaliza-
tion, more frequent mandatory discharges by staff, previous self-harm and
higher education (Kullgren, 1988; Paris et al, 1989; Kjelsberg et al, 1991).
Co-occurring psychiatric disorders

The co-occurrence of borderline PD with combinations of substance abuse,
major depression and antisocial PD have been identified as indicators of
poor prognosis.
Affective disorders (mainly depression) are common with borderline PD,
and in five studies reviewed by Perry (1993) a mean prevalence rate for
affective disorders of 42% was found during the follow-up periods
reported. Such depressive disorders are aetiologically and clinically vari-
able, and in one study about 10% developed a manic-depressive disorder
(Stone, 1992). Also, borderline PD has been commonly reported in
association with agoraphobia (Stone, 1990), while Coid's study of a
forensic sample (1993ft) showed an association between borderline PD and
unspecified psychotic episodes as well as with mood disorders.
Other variables associated with course and outcome of borderline PD
Adverse childhood events

There is considerable evidence of a significant association between a variety
of adverse childhood events or circumstances and the later development of
borderline PD (Marziali, 1992). However, causal relationships are likely to
be complex, as, although the psychological effects of adverse events may be
of considerable importance, the behaviour of some children may contribute
to their adverse environment, and their genetic vulnerability in relation to
future PD may be shared by their living group if this includes their
biological relatives.
A wide range of events have been associated with borderline PD, such as
neglect (including emotional withdrawal by the caregivers), physical abuse,
parental loss (by death or separation), poor quality of parenting, and
various forms of sexual abuse, in particular father-daughter incest. It has
been claimed that an history of sexual abuse carried out by a single
individual is rare (Westen et al., 1990) and that it is common in groups with
borderline PD, i.e. 26-71% in various studies (Crowell et al., 1993).
However, it is also common in some other patient samples, and it has been
considered that an important causal factor for some examples of borderline
PD may be severe and repetitive abuse involving multiple traumas, rather
than a specific type of adversity.
In adult psychiatry, 'post traumatic stress disorder' (PTSD) has been
receiving increasing attention. This syndrome involves re-experiences of
traumatic events, avoidance of some situations, reduced emotional respon-
siveness and increased arousal, and it has been suggested that these features
may follow adverse events in childhood, in particular severe physical abuse
and incest, and contribute to the later development of borderline PD
(Stone, 1990).
Adverse events in adulthood

In addition to associations between borderline PD and childhood loss
events, there is evidence of a relationship between severity of borderline PD
features and rates for the recurrence of depressive disorders following
stressful life events in adulthood (Perry et al, 1992).
Duration of follow-up period

As has been previously noted, borderline PD tends to be associated with
gradual improvement, and this is reflected by findings of a correlation
between length of follow-up interval and scores on adaptive aspects of
global outcome (Perry, 1993).
Gender
Borderline PD has been more commonly diagnosed in females. However, in
males these features often co-occur with those of antisocial PD, and as this
PD is more common in males, it may often be given diagnostic precedence.
McGlashan (1993) in the Chestnut Lodge study, found that, at index
admission, females were more likely to be married with better heterosexual
adjustment, to be more depressed and be more prone to self-harm. In
contrast, males were more antisocial and uncooperative. Subsequently, of
those with relatively poor outcome, males were more likely to be involved
with alcohol abuse and females with self-harm episodes.
In Coid's study (1992) of 243 patients detained in forensic psychiatric
units, there was a higher prevalence of borderline PD in females than in
males (i.e. 90% to 50%), although several co-occurring PDs was the usual
finding.
Adverse childhood experiences have been suggested as causal factors for
some examples of borderline PD, and as girls are more likely than boys to be
sexually abused as children (Marziali, 1992), this may contribute to the
apparently greater prevalence of borderline PD in females.
Implications of longitudinal data for the management of borderline PD

Although about a third of hospitalized patients with borderline PD have a
poor long-term outcome, the outcome is very varied even for severe cases,
ranging from complete remission to suicide. This must reflect the hetero-
geneity of this disorder resulting from the variable interaction of a range of
causal factors. Therefore, it is not surprising that there are considerable
variations in what appears to be the appropriate management, while
improvement can occur without any interventions from medical services.
Nevertheless, certain principles of management can be recognized from
longitudinal studies (McGlashan, 1993). As the long-term outlook is
generally fairly good, even in severe cases, patience and optimism are
usually appropriate. Patients can be reassured that a significant (albeit
gradual) improvement in their adjustment is possible, and in the meantime,
the clinician can provide advice, support and protection in times of crisis.
This often involves short-term admissions to a psychiatric unit to protect
the patient from the harmful consequences of impulsive self-destructive
behaviour with self-injury, suicidal attempts or substance abuse. For a
small minority of patients, long-term asylum can seem appropriate,
involving continuous or intermittent institutional residence for up to
several years. It has been claimed that there is less danger of the negative
effects of an institution for those with borderline PD, compared with
patients with schizophrenia, and that the expected improvement in the 20s
and early 30s holds out a realistic hope of eventual rehabilitation to
independence from residential care.
For many patients, especially those who have pronounced problems with
intimate relationships, a focus on developing their capacity for work is
important. However, as individuals with borderline PD find it difficult to
live alone, there is a tendency to seek intimate relationships which may
regularly degenerate into an unstable hostile dependency. For those
patients who cannot improve their adaptation in intimate relationships,
some learn that they are able to cope better by avoidance. This may involve
a compromise, with strong attachments to groups of people associated with
work, religion, political parties or self-help organizations. Commitment to
a cause rather than one (or a few) individuals may be successfully combined
with relatively superficial relationships that can be sustained. Longitudinal
studies of inpatients with borderline PD have indicated that men are less
likely to get married than women, but, for both sexes, divorce or separation
is common. The principle that a patient should 'control the distance' of a
relationship to avoid a degree of intimacy that cannot be sustained, extends
to the relationships with medical services. Provision of treatment and
support often needs to be on aflexible,'as required' basis, which is under the
patient's control as far as is possible and reasonable. A further principle is
that a treatment 'package' should be designed in relation to the individual
patient.
A psychiatric service can plan its provision on the basis that the majority
of the most severely affected patients with borderline PD will require short
periods of hospitalization and flexible outpatient support for up to several
years, although, within this group, a few subjects appear to need long-term
residential care or intensive outpatient psychotherapy. The general pattern
of appropriate service delivery may change at different stages of the
disorder. In middle age, when many patients will no longer qualify for the
borderline PD diagnosis, a more sustained psychotherapeutic approach
may be indicated for the depressive symptoms that can follow the break-up
of a long-standing, stormy but supportive relationship.
Longitudinal studies of borderline PD can provide, from the public
health perspective, ideas for reducing the suicide rate in young adults, to
which this disorder contributes. A serious degree of childhood physical and
sexual abuse, which has been implicated as a probable causal factor for
some individuals with borderline PD, might be reduced by measures such as
improvements in social work services, in child-care provision and in
community recreational facilities. Also, as various studies have concluded
that self-help groups for those who abuse drugs and alcohol can have a
favourable impact on the course of these behaviours, the encouragement of
these organizations might also be expected to result in less maladaptive
behaviour, including suicide, shown by those with PD.
Antisocial personality disorder
Longitudinal studies of antisocial PD
Overview
DSM-III-R's category of 'antisocial PD' relies heavily on an history of
specified antisocial acts and identifies individuals who show considerable
heterogeneity, while Hare's criteria (1991) lay more emphasis on traits that
affect personal relationships. These include lack of remorse or guilt,
callousness, lack of empathy and failure to accept responsibility for one's
own actions. Such attributes can be considered as 'core' traits for antisocial
PD (psychopathy), although they are not prominent in many patients who
would qualify for the DSM-III-R diagnosis. Also, subjects can have
pronounced 'core' traits without an history of unlawful behaviour,
although there is a clear association between the 'core' traits and overtly
criminal or violent behaviour. For instance, Hare (1991) found that scores
on his rating instrument, the PCL-R, were associated with the degree of
violent and aggressive behaviour in prison settings.
The broad category of DSM-III-R's antisocial PD has to be evaluated in
the context of a high level of'sub-cultural delinquency' in some geographi-
cal areas and socio-cultural groups. West & Farrington's community
survey of males of low socioeconomic status in part of London (1977)
found that, at age 17, about 20% had a criminal record and that this had
risen to 33% by age 24. But although the incidence of convictions fell
considerably with increasing age over the age of 18, a minority showed a
pattern of repeated convictions, which persisted into adult life. This group
had a younger age at first conviction, a larger number of convictions up to
the age of 18 and showed antisocial behaviour in a wide range of contexts.
Such individuals can be considered to have a PD, and are often unpopular
with their peers, tending to stand out even within a group in which
lawbreaking is common. Despite the heterogeneity of the characteristics of
antisocial individuals, and the various criteria sets for diagnosis of 'anti-
social PD', recurrent or persistent antisocial behaviour has been validated
as a successful marker for PD by both genetic and longitudinal studies. The
demonstration of genetic factors associated with antisocial PD has been
considered to be 'convincing' (Dahl, 1993), as has the association between
childhood conduct disorder and adult antisocial PD.
Longitudinal studies that link childhood conduct disorder and adult
antisocial PD have also found that conduct disorder tends to be associated
with adverse and disordered family settings, involving violent and inconsis-
tent patterns of child-rearing and increased criminality (Coid, 1993a). This
is consistent with a model of causation that involves a variable interaction
of genetic factors, other biological variables and social environment. But
although most of those with the adult features of antisocial PD give a
history of significant childhood conduct disorder, (a minority do not show
problem-behaviour until mid or late adolescence), most children with
conduct disorder do not become adults with antisocial PD (Crowell et al.9
1993). However, for most seriously antisocial adults, antisocial behaviour
generally becomes apparent before the age of 10 (West, 1983).
As would be expected in relation to an heterogeneous disorder, the
course of adults with serious antisocial behaviour shows considerable
variation. One of the main defining characteristics of a 'disease' is increased
mortality, and this is found in relation to antisocial PD, mediated by
increased rates of suicide, accidental death, substance abuse and being a
victim of violence (Martin, 1986). Hare et al. (1988) have claimed that those
individuals with predominant 'core' features of psychopathy tend to show a
persistently high level of criminal activity in their early adult years but that
this decreases sharply after the age of about 40. However, this change is less
marked in relation to violent crimes, and, in some forensic studies, about
20% of those with antisocial PD in early adult life still meet criteria at the
age of 45 (Coid, 1993a). Nevertheless, it is encouraging that even in samples
selected on the basis of serious criminal activity, with a typical history of
childhood conduct disorder, most show considerable improvement during
middle age.
Studies related to epidemiology

Information about the course of a disorder can often be deduced from
epidemiological data, for instance when prevalence is related to age. In a
catchment area study in three US sites (Robins et al., 1984), the mean
lifetime prevalence rate for antisocial PD was 2.6% (range 2.1-3.4) with the
male: female ratio as 4.5:0.8. The diagnosis was relatively high in the 25-44
age group and was associated with early drop-out from education and
inner-city populations. A similar lifetime prevalence of 2.5% was reported
in another study of six US sites (Regier et al., 1988) and of 3.1% in New
Zealand (4.2% in males and 0.5% in females) (Wells et al., 1989).
Robins et al. (1991), in another epidemiological study, found that only
47% of subjects with a diagnosis of antisocial PD had arrest records, while
only 37% of a group who had been arrested met DSM-III criteria for
antisocial PD. Also, 51.6% of those who originally received the antisocial
PD diagnosis did not receive the diagnosis after follow-up at one year,
which may have been related to those subjects with substance abuse at the
initial contact whose behaviour had improved after a period of abstinence.
In Canada, Bland et al. (1988) found lifetime prevalence rates for antisocial
PD of 3.7% (6.5% in males, 0.8% in females), and this diagnosis was
associated with the 18-34 age group and lack of sustained intimate
relationships. However, lower rates have been reported in Taiwan (Hwe,
Yeh & Chang, 1989) and in two US cities (Zimmerman & Coryell, 1990). In
the latter study, this was probably due to an underestimate with a self-
report questionnaire assessment.
In prison populations, prevalence rates for antisocial PD have varied
from 39-76%, but of male prisoners whose behaviour caused serious
problems in the UK, 86% received this diagnosis (Coid, 1991, 1992). Also,
the pattern of violence has been found to be different between groups of
prisoners with and without antisocial PD. This diagnosis was associated
with bout drinking, violence directed more towards males, unknown
victims, feelings of revenge, callousness and displays of aggression. In
contrast, in a group without antisocial PD, violence tended to be associated
with emotional arousal, domestic argument and directed to female victims
who were known to their assailants.
Studies of childhood conduct disorder

The best-known study of the outcome of conduct disorder, especially theft
and aggression, is that of Robins (1966), who followed up children referred
to a child guidance clinic for up to 30 years. There was a significant increase
in the subsequent development of antisocial PD in comparison with a group
who had been referred for other problems.
Studies of non-clinical and non-forensic populations

Allebeck et al. (1988) and Allebeck & Allgulander (1990ft) reported a
prospective 13 year follow-up study of 50465 Swedish conscripts, which
examined personality and behavioural predictions of suicide in young men.
It appeared that several early indications of antisocial PD were strongly
predictive of the 247 suicides that occurred, i.e. poor emotional control,
substance abuse, contact with the police or child welfare authorities and
lack of friends. However, a subsequent psychiatric diagnosis in inpatient
care (in particular schizophrenia), was the strongest predictor of suicide.
Studies in psychiatric and offender populations

Various studies have reported different proportions of 'transient delin-
quents' and 'continuous antisocials', although many individuals do not
easily fit either of these broad categories.
Studies of hospitalized adolescents or young adults with PD (McGlas-
han, 1986; Stone, 1993) have shown that some of the males and most of the
adolescent females with less serious antisocial behaviour subsequently
made a good adjustment in their late 20s or 30s, despite often leaving the
unit against advice. The 'continuous antisocial' tended to show criminal
behaviour at an earlier age.
While the eventual outlook is 'fair to good' for many individuals with
antisocial PD who present as young adults to psychiatric services, there is
little evidence of benefit from short-term treatment programmes. Gabbard
& Coyne (1987) claimed that out of 33 patients with antisocial PD, 19 were
'completely unresponsive' to treatment.
The 'transient delinquents' show improvement relatively quickly, i.e.
within a decade, but improvement has also been described in more severe
and 'continuous' samples, often in the fifth decade. The sample described
by Tong & McKay (1959) consisted of 587 patients discharged from an
English maximum security hospital; it was found that 171 committed
further offences but that these generally involved a reduction in violence.
Therefore, even in such a severely affected group, a substantial number
appeared to improve (Whiteley, 1970). In a sample of male prisoners with
an antisocial PD diagnosis, 87% retained the diagnosis after 3 years and
72% at 9 years (Guze, 1976).
De Jong et al. (1992) followed up 348 male criminals who received a
forensic psychiatric assessment after committing manslaughter, attempted
manslaughter, or arson, at a mean age of 31 (SD 11.5); follow-up data were
obtained from the criminal registry of Finland for periods following release
ranging from 1 to 115 months with a mean of 41.4 months. While antisocial
PD was associated with both violent and non-violent recidivism, violent
recidivism in the groups that had killed or attempted to kill was associated
with the degree of impulsivity of the original crime. High impulsivity was
indicated by an unknown victim, no significant provocation, no premedi-
tation and lack of motivation for financial gain. In the arsonists' group, a
history of a suicide attempt was the strongest predictor of future violence.
It has also been found that prisoners with antisocial PD have a greater
incidence of parole violation and further offences after release.
As has been stated, criminal convictions mainly occur in males before the
age of 20. Most offenders do not continue their criminal behaviour beyond
their early 20s and, in various samples of offenders, around 5% are
responsible for a disproportionate amount of crime (Wolfgang et al., 1972;
Blumstein & Cohen, 1979). For the minority who engage in persistent
criminality for much of their adult life, various stages have been identified
although progression is not inevitable (Coid, 1993a). For this minority, a
common pattern between the ages of 10 and 18 involves theft (in particular
car theft) and burglary, often in the company of other adolescents, while
thrill-seeking and status within the peer group are frequent motivations.
The late 20s may coincide with a change for the better, but if the criminality
persists, the seriousness of the crimes may increase although they are
generally less numerous. Although improvement is still possible, usually in
the 40s, a few will continue with criminal activity throughout their active
adult life and it has been estimated that such individuals constitute a
lifetime prevalence of 0.5% of the US population.
Prognostic factors
In general, age and intelligence are inversely related to future violence for
those with antisocial PD (Klassen & O'Connor, 1988). It has been noted
that definitions of antisocial PD differ in their reliance on the nature of
interpersonal relationships involving the 'core' features of psychopathy,
and on specified antisocial acts, and it is clear that the best predictor of
future antisocial behaviour is not the degree of the 'core' features, but the
extent of previous antisocial behaviour (De Jong et al, 1992). For instance,
in prison samples, Hare (1991) has shown that scores for specified antisocial
behaviour using the PCL-R assessment schedule are related to revocation
of parole or re-offending to a greater extent than scores related to the 'core'
features.
Other 'poor prognosis' variables for re-offending have included the
impulsivity of the original crime (in relation to males convicted of man-
slaughter or attempted manslaughter), previous suicide attempts (in male
arsonists), the degree of severity of offences, persistent lying and conning
behaviour, and early onset of childhood conduct disorder (Stone, 1993).
When evaluating characteristics of childhood conduct disorder, the
number and range of antisocial behaviours has been found to be more
predictive of adult behaviour than environmental influences (Robins,
1978).
Other predictors of violence have included various cognitive deficits,
poor social and coping skills, and alcohol and drug abuse. Aspects of
various biological characteristics have also been claimed to be predictive of
violence, involving levels of a metabolite of serotonin in the cerebrospinal
fluid, the EEG, heart rate, skin conductance characteristics, and mild
hypoglycaemia (Virkkunen et ai, 1989), although the latter was not
confirmed (De Jong et al., 1992).
Co-occurring psychiatric disorders

Antisocial PD significantly co-occurs with other PDs and other psychiatric
disorders, which influence course and outcome in longitudinal studies.
In a sample of 243 subjects in special hospitals and prisons with the legal
category (in England and Wales) of 'psychopathic disorder', 86% of the
male prisoners had antisocial PD and 67% paranoid PD (the latter was
present in 45% of those with antisocial PD) (Coid, 1991, 1992). Only two
subjects with antisocial PD had this as the only PD diagnosis, and common
associations were with paranoid, narcissistic, borderline and passive-
aggressive PDs. The paranoid features were of particular note as they
seemed to reflect a subgroup who were extremely suspicious and intolerant.
Violence in such individuals could be easily triggered even by a hint of
criticism. In two studies of female prisoners and patients with 'psycho-
pathic disorder', who also had co-occurring borderline PD (Coid et ah,
1992; Coid, 1993a), the severity of mood disorder led to the suggestion that
this should be considered as an affective disorder in its own right. Anxiety,
irritability, depression and tension were associated with certain destructive
behaviours which provided symptom-relief. This pattern usually appeared
around puberty.
A history of other psychiatric disorders was also common in the sample
of 243 subjects in English maximum security hospitals and prisons (Coid,
1993a), i.e. major depression in 50%, alcoholism in 35% and drug abuse in
27%. Other co-occurring disorders included sexual deviation (in particular
sadism), sadistic masturbatory fantasies of rape and homicide, and com-
pulsive homicidal urges which, if carried out, produced a sense of power,
pleasure and non-sexual excitement. Pyromania and kleptomania were
more common in females.
Other studies have reported that antisocial PD is also associated with
somatization disorder, involving frequent bodily symptoms and com-
plaints (Tyrer & Seivewright, 1988ft).
Other variables associated with course and outcome of antisocial PD
Adverse childhood events

Many studies have reported associations between adverse childhood
experiences and antisocial PD; Robins (1978) has suggested that the
following factors may be of causal relevance in some individuals: the child
placed out of the family, severe poverty, an absent parent, and antisocial
behaviour by parents or caretakers. Zanarini et al. (1989) have noted an
association between antisocial PD and separation from parents for a month
or more, and claimed that harsh and rejecting parenting can also be causally
implicated. Child-rearing practices involving harsh or inconsistent disci-
pline and lack of rewards may be of particular importance as causal
variables for antisocial PD in vulnerable individuals. This may involve
undesirable behaviour not being consistently challenged, while good
behaviour remains unacknowledged and unrewarded.
Loss events
There is some evidence that features of antisocial PD can make an
individual more vulnerable to depressive symptoms after stressful life
events (Perry et al., 1992).
Duration of follow-up period

Although the heterogeneity of antisocial PD has been repeatedly noted, it is
clear that the incidence of offending decreases with age, although the rate
and timing varies between different offender subgroups. The eventual fate
of the 'core' attributes of antisocial PD involving interpersonal relation-
ships is unclear, and it may be that they are often not significantly modified
with time, even though the individual has developed a more socialized
behaviour pattern.
Gender
Cloninger & Gottesman (1987), reviewing genetic and family studies,
concluded that while genetic factors for the development of antisocial PD
are similar in men and women, the lower incidence in women is due to a
higher threshold for expression. Therefore, if a woman is to develop the
overt antisocial behaviour associated with antisocial PD, there may have to
be a greater degree of genetic vulnerability and/or a more adverse environ-
ment, and there is evidence that when men and women with antisocial PD
are compared, the women have the more disturbed childhood environ-
ments (Coid, \993a). Cloninger (1978) has suggested that mild antisocial
PD in women may sometimes present as somatization disorder, and there is
an increased prevalence of antisocial PD and somatization disorder in the
relatives of subjects with either syndrome.
Childhood conduct disorder in males usually has an earlier onset than in
females, when it becomes apparent around puberty (Robins, 1986). Also,
males tend to have more traffic offences, a criminal record and more
promiscuity, while women are more likely to be involved with domestic
violence. Compared with males, women with antisocial PD who had been
previously assessed at a child guidance clinic had lower intelligence, a
history of increased institutionalization, parents who were more often
chronically unemployed and fathers who were more often alcoholic or
antisocial.
Implications of longitudinal data for the management of antisocial PD

The evidence that has indicated that an adverse childhood can be an
important, although variable, causal factor for antisocial behaviour,
suggests that educational, social and economic measures aimed at improv-
ing child-rearing practices may have beneficial effects. In adult life, the
long-term nature of the antisocial behaviour pattern in the recidivistic
subgroup is not satisfactorily managed by a series of custodial sentences. A
logical approach to an individual who repeatedly offends would be to
provide longer periods of external controls but of a variable nature. While
some individuals may require long terms of imprisonment, others might be
provided with sufficient external constraints by careful monitoring or by a

flexible system involving day, night or weekend attendance at designated
centres, with the sanction of custodial provision for relatively short periods
if lapses occur.
Individuals with antisocial PD have been generally considered to
respond poorly to various forms of medium- or long-term psychotherapy,
unless there is a co-existing depressive disorder (Perry, 1990). However,
many of these patients can be assisted in a crisis by advice, counselling,
psychotropic medication and short psychiatric admissions in relation to
self-harm or substance abuse. It is often necessary to impose firm con-
ditions on such contacts with medical services. In a few cases it may be
appropriate to consider recommending a placement for up to a year in a
residential setting (therapeutic community) involving individual and group
psychotherapy.
Other personality disorders: longitudinal studies
'Cluster By PDs (antisocial, borderline, narcissistic and histrionic PDs)

Most longitudinal data for specified PDs relate to antisocial and borderline
PD, but narcissistic PD features are often associated with these (and other)
PD syndromes (McGlashan & Heinssen, 1989). Plakun (1989) found that a
group of inpatients with predominant narcissistic PD fared worse than a
comparison group with borderline PD in respect of general level of
functioning and readmission rates. It has been noted that a combination of
borderline, antisocial and narcissistic PD features have a particularly poor
prognosis (Stone, 1993).
'Cluster A' PDs (paranoid, schizoid, and schizotypal PDs)

Patients with paranoid and schizoid PDs tend not to seek help or to
cooperate with follow-up, so that outcome data are sparse. Although there
is some information for schizotypal PD, various definitions and assessment
methods make interpretation difficult. McGlashan (1986) found that an
inpatient group with schizotypal PD had a worse outcome compared with a
borderline PD group, while patients who had a mixture of the two
syndromes occupied an intermediate position. In the PI-500 series, 12
subjects had a combination of schizotypal and borderline PD, and of the
ten traced, only three were self-supporting, while in a study of 97 patients
Interventions: general trends 179
attending a day unit, followed up for 3 years, schizotypal PD had the worst
outcome compared with borderline PD and 'cluster C PD groups (Stone,
1993).
'Cluster C PDs (avoidant, dependent, obsessive-compulsive and passive-

aggressive PDs)
Patients with 'cluster C PDs tend to appear anxious or fearful, as do those
with self-defeating and depressive PDs.
Patients with these PD features often have other disorders such as
anxiety disorders, agoraphobia, depressive disorders and obsessive-com-
pulsive disorder. While, in general, many patients with these features can be
helped by psychotherapeutic measures, severe passive-aggressive behav-
iour has been considered to be very resistant to intervention. But in a study
of patients with PDs at a day unit followed up for 3 years, those with 'cluster
C PDs had the best global outcome, which included reduced features of co-
occurring disorders (Stone, 1993).
Effects of interventions for personality disorders - general trends shown by

longitudinal studies
It appears that various forms of individual psychotherapy for PDs produce
the best results for those with 'cluster C (anxious and fearful) PDs, while
the more focussed and specific cognitive-behavioural psychotherapeutic
strategies are best for 'cluster B' PDs (Stone, 1993).
In a study of 357 patients with psychiatric disorders in primary care, 301
were followed up after 3 years (Casey & Tyrer, 1990). There had been
significantly more PD psychopathology in those who could not be traced,
and a diagnosis of PD had been made in 24% of those who were followed
up. In this subgroup there had been a greater referral rate to psychiatric
services and, of those admitted to hospital for psychiatric disorders, the
subgroup with PD had more and longer admissions.
It is clear that PD is present in many patients who receive various forms
of psychotherapy. Smith et al. (1980), in a review of 475 studies, concluded
that the psychotherapies had been beneficial, but that efficacy had not been
clearly associated with specific characteristics of treatment. Key outcome
variables were the presence of positive or adaptive personality traits, such
as motivation in relation to appropriate goals, candour, introspection, and
a willingness to accept responsibility for one's own actions. Woollcott
(1985) has noted the prognostic importance of likeability; this is reduced if
the patient is hostile, irritable and angry, so that the degree of anger in an
individual with PD may be an important outcome variable.
Studies related to gender and sexual orientation
Gender and 'normal* personality

Gender differences in normal personality development have been des-
cribed, for instance that young girls appear to be less selfish and aggressive
than boys, but although these differences persist into adolescence, they have
been reported to disappear by early adult life (Cohn, 1991). This suggests
that adolescent girls achieve developmental stages, such as a capacity for
moral judgement and empathy, earlier than boys, but that differences then
decline with increasing age as males 'catch up'.
Gender and PD - the longitudinal perspective

Antisocial PD has a considerably higher incidence in men, while borderline
PD is diagnosed more frequently in women. There are also gender
differences in age of onset of PDs and related behaviour; for instance, males
generally present with childhood conduct disorder at an earlier age than
females, whose disordered behaviour is often apparent around puberty
(Coid, 1993a). Also, in a prison sample of 25% of all women serving a
prison sentence in England and Wales, women showed a higher rate of PD
than a comparison group of male sentenced prisoners (18% versus 10%)
(Maden, Swinton & Gunn, 1994).
Gender can be associated with various behaviours related to PDs. For
patients with borderline PD in the PI-500 series (Stone, 1990), male patients
married less often, found a better niche for themselves in relation to work or
hobbies, and coped better with being alone. In contrast, females were more
likely to become married, to relate better heterosexually, and to have more
depression and self-harm. Also, they were more likely to develop an eating
disorder, which occurred in about 17%.
Bardenstein & McGlashan (1988) also noted that females with border-
line PD were subsequently more likely than men to get married, and, at
follow-up, females reported more stressful events, and more feelings of
worthlessness and guilt. In mid-life, females more often showed a decline in
overall functioning and became chronically depressed in their 40s. In
Studies related to gender and sexual orientation 181
contrast, men were more likely to leave hospital against medical advice and
obtain a forensic history. Males tended to show more alcohol abuse, while
females engaged in more self-harm (McGlashan, 1993).
In relation to antisocial PD, it has already been noted that there is a
gender-related difference in the age of onset of childhood conduct disorder.
Also, women with antisocial PD generally report a more disturbed child-
hood environment than men (Coid, 1993a). Gender also affects the pattern
of antisocial behaviour, with female gender associated with less physical
violence and more arson.
PD and gender-related variables - the longitudinal perspective

An interesting feature of reports of the PI-500 series (Stone, 1990) was that
there had been a courageous attempt to rate the physical attractiveness of
its female patients with borderline PD! In general, extreme attractiveness
was a good prognostic indicator, together with unusual creative talent. For
those who had a poor outcome despite a high rating on attractiveness, there
seemed to be an association with previous sexual contact involving a first-
degree male relative.
Other features that have been associated with aspects of outcome of
borderline PD included having children born out of marriage and com-
plaints of a premenstrual syndrome, when the features of borderline PD
appeared to be intermittent and related to premenstrual mood changes.
Homosexuality and PD - the longitudinal perspective

Twelve per cent of 118 men with borderline PD in the PI-500 series had
features of homosexual orientation such as homosexual fantasies, homo-
sexual activity, and problems with gender identity (Stone, 1990). This
appears to be a greater prevalence than in the general population, for which
rates of up to 5% have been claimed for male homosexuality. The rate for
female homosexuality in the PI-500, which was 1 % in the borderline group,
approximately mirrored that of the general population.
Perhaps the difficulties that some men can experience in relation to
homosexual orientation can be a non-specific causal factor in the develop-
ment of their patterns of PD behaviour, on the basis that any parallel
problems can lower the threshold for the expression of behaviour related to
PDs.
Longitudinal studies related to childhood precursors of personality

disorders
Overview
The literature on childhood precursors of PD is mainly concerned with
conduct disorders, attention deficit hyperactivity disorder (ADHD) and
schizoid or schizotypal disorder. Also, there has been a renewed interest in
the importance of a relatively small number (3-5) of basic temperaments
(Cloninger, 1986), which appear to be identifiable in very early childhood,
and are largely due to genetic factors. But while the importance of genetic
inheritance for the development of personality and PD has become
established by many research findings, it is also clear that adverse environ-
mental events and interactions between specific or non-specific genetic or
biological vulnerability and the environment, together with protective
factors, are all important in the development of PDs. For instance,
relatively low intelligence was a poor prognostic factor for the later
development of antisocial PD in women, in a child guidance clinic sample
(Coid, 1993a).
The landmarks of the literature are the associations between childhood
conduct disorder (and/or ADHD) and the development of antisocial PD,
and between solitariness, rigidity and hypersensitivity in children and the
later diagnosis of schizoid PD (Wolff & Chick, 1980).
Both 'conduct disorder' and 'antisocial PD' are broad concepts, and the
former has been defined as behaviour which violates the rights of others or
social norms. It has been estimated that between 410% of children develop
some significant degree of conduct disorder, and of those who receive this
diagnosis in a medical setting, up to 40% have been reported to show
serious behavioural problems in adulthood (Robins, 1978; Rutter & Giller,
1983). Aggression is often the most troublesome feature of conduct
disorder and may present in different ways; younger children usually show
arguments or tantrums, and these may develop into defiance and oppo-
sition. Later still, examples of aggression can include arson, theft, truancy,
vandalism and substance abuse.
Early predictors ofPD

Renewed interest in a small number of broad 'temperaments' has been
stimulated by studies in infants and very young children, which have linked
early behavioural patterns with later personality and PD features.
Childhood precursors of personality disorders 18 3
A community sample of 976 infants and children were first assessed (at
T l ' ) between the ages of 1 and 10, and followed up on two occasions, i.e. at
'T2' (n = 778, age 9-19) and T 3 ' (n = 776, age 11-21). At Tl, the assessment
consisted of structured interviews with the mothers, which focussed on
aspects of conduct problems, depression, anxiety and immaturity. Later
assessments also involved interviews with the subjects. Conduct problems
at Tl were related to subsequent maladaptive patterns of behaviour in
adolescence, although for children under the age of 5, this was only found
for girls, and related to subsequent 'cluster B' and ' C disorders (Bernstein,
1993).
Another area of longitudinal research in relation to personality has been
shyness, which has been investigated by rating the reaction and degree of
behavioural restraint of very young children to unfamiliar stimuli (Kagan
et ai, 1988). Two groups were identified at the extremes of reaction (i.e.
showing 'restraint', with fear and avoidance, or showing unconcern and
approach), and by the age of 7 the 'restraint' group were generally much less
talkative and interactive, and more socially avoidant with unfamiliar
people. It was suggested that the relevant differences in the underlying
biological substrate may involve the threshold of arousal to novel stimuli.
Conduct disorder and PD

DSM-III-R notes that the essential feature of childhood conduct disorder is
'a persistent pattern of conduct in which the basic rights of others and major
age-appropriate societal norms or rules are violated', while the ICD-10
specifies 'a repetitive and persistent pattern of dissocial, aggressive or
defiant conduct'. The DSM-III-R criteria for conduct disorder in children
form part of the criteria for antisocial PD.
Various subtypes of conduct disorder may have a predictive value, such
as when behaviour is confined to the family context, the presence or absence
of good integration with the peer group, and when problem behaviour is
restricted to 'oppositional defiant disorder', for which DSM-III-R provides
nine criteria, including 'often loses temper', and 'often argues with adults'.
In a community sample of 19482 subjects interviewed in the NIMH
Epidemiologic Catchment Area Program (Robins & Price, 1991), a history
of childhood conduct disorder was associated with increased prevalence of
a range of psychiatric disorders, including antisocial PD but also mania and
schizophrenia.
Course and outcome of conduct disorder varies considerably, mainly in
relation to criteria used and the characteristic of the sample (Lofgren et aL,
1991). However, most childen who receive a medical diagnosis of conduct
disorder have a favourable outcome, even though Robins' classical study
(1966) of 406 children referred to a child guidance clinic for behaviour such
as theft and aggression found that this group went on to develop a higher
rate of antisocial PD than did a comparison group of 118 referred for other
reasons. After a follow-up of up to 30 years, 75% of men and 40% of
women had been arrested for non-traffic offences. Childhood factors which
predicted repeated adult antisocial behaviour were a variety of types of
conduct disorder, many episodes, and behaviour involving strangers rather
than school and home. Of the 76 girls in this study, about 20% went on to
develop a disorder similar to histrionic PD. In a more severely affected
sample of 55 adolescent girls in a locked unit for conduct disorder, follow-
up for between 2-4 years gave a generally poor outcome (Zoccolillo &
Rogers, 1991); 6% had died and about half had re-offended.
The Kauai longitudinal study of 698 children on this Hawaiian island has
provided further evidence of the interaction of biological and environmen-
tal causal factors for the development of adult antisocial behaviour
(Werner, 1985; Dolan & Coid, 1993).
ADHD (attention-deficit hyperactivity disorder) and PD

ADHD is denned by DSM-III-R by the presence of at least eight out of 14
behavioural characteristics, with an onset before the age of 7. The equiva-
lent syndrome in ICD-10 is 'hyperkinetic disorders', which is a term chosen
to avoid assumptions about aetiology; early onset is also claimed to be
found with the latter, usually within the first 5 years, and the main feature is
excessive motor activity. However, diagnosis is more readily made after the
age of 5 in the early school years. ADHD occurs in around 3% of children,
but is much more frequent in boys, i.e. x 6-9 in clinic samples, and x 3 in
community samples.
In both classifications, the core features of the hyperactivity syndrome
are overactivity and marked inattention, associated with impaired ability to
complete tasks. Overactivity may involve restlessness, difficulty in remain-
ing seated, excessive talking, frequent running and jumping, excessive
fidgeting, and difficulty in playing quietly, while inattention is reflected by
distractibility, rapid loss of interest, a failure to listen or to complete tasks,
frequent changing from one activity or task to another, and a tendency to
lose objects. However, as these features are most obvious when sustained
attention is required, behaviour may seem normal in a relatively structured
situation such as a clinical assessment. Also, impulsivity is a common
Childhood precursors ofpersonality disorders 185
feature, shown by frequent interruptions and intrusions to others, careless
and poor schoolwork, and not taking turns or not following the rules of
games. This often makes the subject unpopular with both peers and
teachers. Recklessness is another feature that is commonly described,
involving lack of foresight as to dangerous consequences of actions,
together with behaviour which is 'accident-prone'. Finally, cognitive
impairment and motor clumsiness have also been considered as part of this
variable syndrome.
Almost as important as the features of the syndrome itself, is the common
co-occurrence of several other apparently related disorders, particularly
conduct disorder (which may have a later onset in childhood), but also
'oppositional defiant disorder' (Rey, 1993), 'specific development disorders
(e.g. learning disability)', Tourette's syndrome', 'encopresis' and 'enure-
sis'. Secondary complications include low self-esteem, lability of mood with
temper, depression and anxiety, a low frustration tolerance and academic
underachievement.
While relatively mild overactivity and inattention are common in
conduct disorder, it appears that it is important, in relation to adult
outcome, to distinguish subjects with ADHD alone from those who also
qualify for the co-occurring diagnosis of conduct disorder. When both
disorders are present, the ICD-10 provides the diagnosis of 'hyperkinetic
conduct disorder'.
The literature on ADHD has to be interpreted in the light of the
characteristics of each sample reported, in particular the origin, and the
degree of co-occurrence of conduct disorder. In general, the DSM-III-R
claims that about one third of those from clinic samples will eventually
retain some aspects of the syndrome in adult life, and there is evidence to
support the validity of the diagnosis of ADHD in adults (Biedermen et ai,
1993). A general trend to improvement has been claimed by Klein &
Mannuzza (1991), who found that one third of their sample showed
remission of ADHD by mid-adolescence, and most showed reduced
impairment by late adolescence. Also, Weiss (1985) found that most
hyperactive children grow up without showing serious adult disorder,
although about half continued to show one or more features in adolescence
and early adulthood, and there was an increased risk of antisocial PD.
Various adult outcomes of ADHD are possible, including: firstly, nor-
mality; secondly, significant problems with concentration, social difficul-
ties, problems with emotional control and maladaptive impulsivity; and,
thirdly, serious problems involving antisocial PD and substance abuse
(Hechtman, 1991). Of the variables that influence outcome, probably the
most important is co-occurring conduct disorder, which is related to the

development of antisocial PD, but other factors related to poor outcome
include living in an environment with a high crime rate, low measured
intelligence, psychiatric disorder in parents (particularly PD), criminality in
parents, loss of parents (particularly with divorce), large family, periods 'in
care', violent abuse, inconsistent discipline, low socioeconomic status, and
a lack of structured environment. In contrast, there are protective factors
such as relatively high intelligence, adaptive personality features, a sup-
portive family, an extended family, and a law-abiding social environment.
Mannuzza et al. (1991, 1993) followed up 101 young adult males, who
had been previously diagnosed as hyperactive in childhood, for a mean of
16 years, and found higher-than-expected rates of ADHD features, anti-
social PD and substance abuse, while a number of studies have dis-
tinguished between the outcomes of subgroups of hyperactive children
based on the co-occurrence or absence of conduct disorder. For instance, a
4-year follow-up study compared 34 boys with 'pure' ADHD with 42
hyperactive boys who were also aggressive and more unsocialized (August
et al., 1983); it was found that the latter group, whose parents showed more
PD and alcoholism, remained relatively antisocial, while inattention and
impulsivity continued to be the main problems in the former group. In a
recent review, Lie (1992) concluded that hyperactive children without co-
occurring conduct problems do not have a significant risk for future
delinquency or antisocial PD, but that if ADHD is accompanied by
conduct disorder, there is a worse outcome compared with conduct
disorder alone. It seems that some remnants of ADHD can be found in
early adult life in about a third of children with ADHD who are followed
up, and that there is a significant association between ADHD and conduct
disorder, i.e. their co-occurrence is not just due to chance. This was
indicated in a study of 283 male adoptees (Cadoret & Stewart, 1991), when
it was found that a history of criminality in a biological parent was
associated with an increased incidence in ADHD in their offspring.
Several studies have provided evidence that medication can benefit
children and adults with ADHD.
Schizoidfeatures in childhood and PD

Schizoid PD in childhood, as described by Wolff (1991), was noted to be
more common in boys, and to consist of features such as unusually solitary
behaviour, poor social conformity and emotional outbursts. Relative social
isolation was found to persist into early adult life.
Ch ildhood precursors of personality disorders 187
Other childhood-related variables associated with PD

Many aspects of an adverse childhood environment have been shown to be
associated with PDs in adults, in particular with antisocial and borderline
PD. But such relationships are complex and it is important not to assume a
simple cause-and-effect model. Poor parenting may be related to parental
genetic predispositions that may also be present in the child, while complex
interactions between genetic, other biological, and environmental variables
are likely to be involved, including the subject's behaviour influencing his or
her environment.
Various aspects of experience can be subsumed under the heading of an
'adverse childhood environment'. Robins (1978), in a child guidance
sample, found that girls who later had a diagnosis of antisocial PD had
more institutional care, more fathers with alcoholism or antisocial behav-
iour, and more parents who were chronically unemployed. Other studies
have confirmed associations between childhood conduct disorder and
disordered family backgrounds including increased parental criminality,
violence, neglect, death or absence of a parent, and erratic practices of
child-rearing.
Physical abuse, including neglect, violence and sexual abuse, is of
particular concern, and has been estimated to be of significance in about
1% of children in the US (Raczek, 1992). There is a clear association
between physical abuse in childhood and the later development of aggress-
ive, antisocial and criminal behaviour, with some children becoming school
truants, teenage parents, delinquents, criminals and, in turn, parents who
give physical abuse to the next generation. However, some children who are
subject to physical abuse become shy, socially inhibited and suspicious, and
may present in adulthood with low self-esteem and a pattern of failure.
Also, it has been suggested that overprotection and humiliation may be
related to subsequent development of avoidant PD.
Associations between PD and childhood environment can be
approached from the adult perspective by examining the histories of
patients with PD. Herman et al. (1989) found that 81% of a sample of
patients with borderline PD had a history of severely adverse childhood
events, such as physical abuse and witnessing serious violence, while Nig et
al. (1989) claimed that a history of sexual abuse was associated with
relatively severe borderline PD features. In a study of 50 servicemen
referred for psychiatric evaluation (Raczek, 1992) there was, once again, an
association between PDs (mainly antisocial and borderline) and a history
of physical abuse, but it was noted that some individuals with avoidant or
paranoid PD features also gave such a history, while some seemed relatively
unscathed by such traumas.
The evidence does not suggest that certain childhood experiences are
necessary for the development of a particular PD, or that the effects of
sexual abuse are readily predictable in an individual. It has been pointed out
that some children can be remarkably resilient in the face of serious
childhood adversity. As causal factors for a particular PD syndrome are
multiple, interactive and variable, it is not surprising that research has
shown a variety of outcomes that are correlated with a range of causal
variables (Crowell et al., 1993).
Implications for management and prevention of PD from longitudinal

studies of childhood precursors ofPD
As antisocial and borderline PD are the PDs that produce the most
apparent problems to society and medical services, these syndromes have
attracted considerable attention. But it will be necessary to increase our
knowledge of adverse and optimal childhood environments in relation to
all PDs, as well as to well-adjusted personalities. This area of research is
likely to be of particular importance during periods of rapid social change,
for instance involving an increase in the number of one-parent families.
Even with present knowledge, it is possible to make recommendation for
social policy in relation to antisocial PD and related behaviours. Measures
to limit or prevent physical or psychological adversity in child-rearing are
clearly important, and might involve the provision of more day-care and
nursery education to pre-school children, particularly for families with
identified problems. It is also important to realize that causal factors for
serious conduct disorder in children are likely to be variable. In some cases,
when a child's upbringing and parental background is unremarkable, it is
likely that genetic or other biological factors are relatively important. For
such a child, the emphasis may have to be on the management of the
disorder by the provision of an appropriate limit-setting and supportive
environment, while for others, who have experienced a seriously adverse
environment, a more intensive psychotherapeutic approach may have more
potential.
The recognition of children who are 'at risk' for adult PD provides an
opportunity for educational, social and clinical services to target such
groups for special help, in particular, children with both conduct disorder
and ADHD. Ideally, a range of options for living and schooling should be
available, so that an individually orientated approach can be attempted.
What can be achieved in a special school unit with a small number of
Effects on other psychiatric disorders 189
children may be impossible in the normal school system if one or more

children with serious conduct disorder are part of large classes. Children
and adolescents with behavioural problems may also benefit from help that
focusses on social skills training, learning how to control impulsiveness,
and the encouragement of self-esteem by trying to identify and reward
desirable behaviour.
Longitudinal studies of personality disorders related to suicide and

mortality
Some patient groups selected on the basis of PDs have shown a higher
mortality than groups with affective disorders, for both inpatient and
outpatient samples (Tyrer & Seivewright, 19886). This was mainly due to
suicide and accidents.
Patients with borderline PD have a particularly high risk of mortality
within the 5 years after contact with psychiatric services. As previously
noted, a review of nine studies of borderline PD found that a mean of 6.1 %
had committed suicide over a mean follow-up of 7.2 years, and the greatest
risk appeared to be during the first 2 years after an hospital admission
(Perry, 1993). This contrasts with a group with passive-aggressive PD of
whom 1.3% had committed suicide by 11 years. In relation to antisocial
PD, Martin et al (1982) found that 2% of female felons had died at a
follow-up after 6 years, and Guze (1976) reported that 3% of male felons
had died violently by 9 years. This indicates that suicide is a more common
outcome for severe borderline PD than for antisocial PD.
Another approach to the relationship between PD and mortality is to
identify the characteristics of a group identified by suicide or other causes of
death mediated by behaviour. For instance, when 62 suicides aged between
15 and 24, who had been previously admitted to hospital for self-harm,
were compared with a group previously admitted but still alive, the
diagnoses of PDs and substance abuse were significantly associated with
the former group.
Studies of the effects of personality disorders on course and outcome of

other psychiatric disorders
Overview
It has often been reported that the co-occurrence of PD is associated with a
poor outcome and response to treatment of various other psychiatric
disorders (Andreoli, Gressot & Aapro, 1989; Reich & Green, 1991).
Sometimes this may reflect the importance of the PD as a precursor and
causal factor for the other disorder in question, for instance in relation to
some cases of depressive disorders and substance abuse, but a chance co-
occurrence of a PD may affect recovery from an unrelated disorder, such as
brain damage or mania.
Substance abuse with PD (including alcohol-related syndromes)

Penick et al. (1988) have reported that the most frequent co-occurring
diagnoses shown by 241 inpatient men with alcoholism were depressive
disorders, antisocial PD and substance abuse, while other studies have
classified alcoholism on the basis of the presence or absence of co-occurring
disorders (Schuckit, 1985; Yates et al, 1988; Liskow et al, 1990; Powell et
al, 1992). Co-occurrence of antisocial PD and substance abuse had a worse
outcome compared with substance abuse alone (Schuckit, 1985), but not all
studies have found that such co-occurrence can predict outcome (Powell et
al, 1992).
Yates et al (1988), in a study of 260 inpatient men with alcoholism, found
that co-occurring antisocial PD was related to earlier onset of alcoholism
and additional abuse of other substances, while Inman et al (1985), in
another inpatient study, noted that borderline PD was associated with a
short duration of treatment. Another PD-related variable associated with
substance abuse was found by Malloy et al (1990), who reported that in 30
substance abusers with antisocial PD, there were more neuropsychological
deficits compared with a control group without this PD. The explanation
was assumed to be related to an earlier onset and increased severity of past
alcohol intake, together with an increased incidence of head injury. In a 5-
year follow-up study of subjects with repeated self-harm episodes, degree of
repetition was related to a combination of PD and substance abuse,
particularly alcoholism (Krarup et al, 1991).
Other outcome studies of alcoholism, or other forms of substance abuse,
have found that a relatively poor prognosis was associated with co-
occurrence of schizoid PD (Zivich, 1981), relatively heavy drinking in a
programme for those convicted of driving offences (McGuire, 1980), and
drop-out rate from a residential treatment programme (Pekarik et al,
1986). But in contrast to the above finding in relation to schizoid PD, there
have been reports which indicate that this disorder can be associated with a
better prognosis for substance abuse than when antisocial PD co-occurs
(Griggs & Tyrer, 1981).
Many studies have focussed on the abuse of drugs other than alcohol.
Twenty-seven per cent of a sample of opiate addicts had co-occurring

antisocial PD (Rounsaville et al., 1986), and about one third of 298 cocaine
abusers had a history of ADHD and conduct disorder (Carroll & Rounsa-
ville, 1993). Of this latter subgroup, 78% were male and 47% had antisocial
PD. When these subjects were compared with those without a history of
ADHD, those with ADHD and conduct disorder had presented for
treatment at a younger age, had an earlier onset and showed a more severe
cocaine abuse. It was suggested that this form of substance abuse could, in
some cases, be viewed as self-medication for residual ADHD features.
Studies of intravenous drug abusers are of increasing importance due to
the spread of HIV, and the co-occurrence of intravenous drug administ-
ration with antisocial PD has been shown to be associated with increased
risk of HIV transmission due to more needle sharing (Brooner et ai, 1993).
In addition, antisocial PD has been associated with more sexual contacts
(Gill et ai, 1992).
It is difficult to generalize from follow-up studies of substance abusers
because of selection bias for the samples and high drop-out rates. However,
of 60 subjects who presented to a London drug clinic in the 1970s, who were
followed up for 10 years, 11 had died (Gordon, 1983). Of the rest, 75% had
been abstinent for 5 years and 25% were still addicted. The latter group was
associated with early parental loss, poor academic achievement, criminal
conviction before the onset of substance abuse and longer terms of
imprisonment, suggesting a relationship with antisocial PD. This wide
disparity of outcome, with high mortality, has been shown in other studies
(Edwards & Goldie, 1987; Gill et aL, 1992), and appears to be strongly
related to the nature of co-occurring PDs.
As counselling and other forms of psychotherapy are in limited supply, it
is important to deploy such resources to their best advantage. Woody et al.
(1985), in a study involving 110 opiate addicts, concluded that co-
occurrence of antisocial PD alone (i.e. without a depressive syndrome)
predicted poor outcome with psychotherapy, but that if depression was
present, psychotherapy appeared to be beneficial despite a history of
criminal activity.
Depressive disorders with PD

The literature on the effects of PD on the course of depressive disorders is
particularly difficult to interpret because of the variable range of relevant
biological and environmental causal factors for both these categories of
psychiatric disorder. However, it is clear that PD is common in patients
with depressive syndromes, and can often influence response to treatment

and outcome. Tyrer et al. (1983) have reported that about 40% of 313
psychiatric outpatients (many with depressive syndromes) appeared to
have a PD diagnosis, and similar prevalences for co-occurring PDs have
been reported for anxiety states. Merikangas, Wicki & Angst (1994) have
stressed the importance of using the course of the disorder as a classification
criterion for depressive disorders.
Andrews et al. (1990), in a 15-year follow-up study of depressive
syndromes divided into 'endogenous' and 'neurotic', claimed that co-
occurring PDs appeared to account for 20% of the outcome variance in the
neurotic group but only 2% in the endogenous group. These results can be
explained if it is assumed that this simple classificatory system produces an
'endogenous' group in which genetically mediated biological vulnerability
to episodic depressive disorder is more marked, and a 'neurotic' group in
which PDs are more prevalent.
Shea et al (1990) described the outcome for 239 out-patients with a
diagnosis of DSM-III-R's 'major depression' in the NIMH 'treatment of
depression' programme. PDs were identified in 74% of the sample and
these patients had a worse outcome in terms of social functioning and the
degree of residual depressive symptoms. The results were similar in relation
to the three DSM-III-R PD clusters and frequent PD co-occurrence was
noted.
Several other studies have linked PD with poor outcome in depressed
patients (Tyrer & Seivewright, 19886; Duggan et al., 1990; Perry, 1990;
Reich, 1990c) and with poor response to antidepressant medication (Sato et
al., 1993a, b; Tyrer et al, 1983). Also, Perry (1990) found that if depressive
disorders co-occurred with borderline or antisocial PD, this was associated
with relatively high recurrence rates compared with control groups with
depressive disorders without PD, and that co-occurring borderline PD was
related to a longer duration of depressive episodes compared with anti-
social PD. Although Zimmerman, Coryell & Pfohl (1986) found that PD
did not impair response to ECT in some severely depressed patients, PD
was related to more frequent relapse.
When two groups of inpatients with 'major depression', with and
without PD, were compared after a week in hospital without drug
treatment, the presence of PD was associated with an improvement in
symptoms (Mazure et al, 1990). This suggests that the non-PD group in
this study consisted of subjects whose depressive disorder was less respon-
sive to a supportive environment, perhaps because of a higher prevalence of
genetically-mediated biological causal factors. But the results of a particu-
lar study may not be relevant to groups of patients selected on a different
basis (Tyrer et ai, 1993).
Schizophrenia with PD
It is difficult to differentiate an insidious onset of schizophrenia from a
premorbid schizoid (or schizotypal) PD, but it has been claimed that such a
premorbid PD is an indicator of poor prognosis (Tyrer & Seivewright,
19886). However, any co-occurring PD is likely to have an adverse effect on
the prognosis for schizophrenia.
Anxiety disorders with PD

In general, co-occurring PD has been found to be related to poor prognosis
for anxiety disorders. In a study of 165 agoraphobic patients, 90% had at
least one PD, most commonly avoidant and dependent PDs, and those with
avoidant PD had the worst outcome (Chambless et aL, 1992). Hoffart &
Martinsen (1993) reported a similar finding for avoidant PD in relation to
panic disorder and agoraphobia, and to unipolar depression. Also, Reich
(1988) found that for 52 patients with panic disorder who were treated for 8
weeks with anxiolytics, a poor outcome was related to PDs, in particular,
antisocial, histrionic, narcissistic and avoidant PDs. However, Marchand
& Wapler (1993) did not find that PD predicted the responses to cognitive
behaviour therapy for agoraphobia and panic disorder.
Obsessive-compulsive disorder (OCD) with PD

PD was found to be related to outcome in 55 patients with OCD treated
with clomipramine (Baer et al, 1992). The co-occurrence of a 'cluster A' PD
(paranoid, schizoid or schizotypal PD) was associated with a more severe
syndrome, while a worse outcome was related to schizotypal, borderline
and avoidant PDs, to the total number of PD diagnoses and to one or more
'cluster A' PDs.
Eating disorders with PD

Co-occurring PD has been shown to predict outcome for eating disorders,
in particular for bulimia nervosa. For example, Sansone & Fine (1992), in a
3-year follow-up, found that the degree of borderline PD features was
related to greater global distress, less life satisfaction and less perceived
194
improvement in eating disorder symptoms. Also, Johnson et al. (1990), in a
1-year follow-up of borderline and non-borderline PD groups of patients
with bulimia nervosa, found that the non-borderline group had a better
symptomatic outcome. Alcohol abuse is another indicator of poor progno-
sis (Lacey, 1984; Rossiter et al., 1993). However, these data relate to
relatively short-term outcomes that do not take into account the gradual
improvement in borderline PD over longer periods.
Stone et al. (1987), in a 10-year follow-up study, found that eating
disorders with co-occurring borderline PD did not have a worse outcome
compared with borderline PD alone, measured by a Global Assessment
Scale score, and that severe obsessive-compulsive PD traits with anorexia
nervosa may be associated with the worst outcomes.
Implications for the management ofpsychiatric disorders with co-occurring

PD
It is clear that co-occurrence of PDs with other psychiatric disorders is
common, and has implications for management and outcome, in particular
for the common syndromes of substance abuse and depressive disorders. It
is, therefore, most important to assess PD status in patients with all
psychiatric disorders.
5
Assessment of personality disorders
J.H. DOWSON
Nature of personality disorders

The ingredients of PDs consist of higher mental activity such as attitudes,
beliefs, motivations and feelings, and associated overt behaviour. How-
ever, only the overt behaviour, which includes speech, is available for
investigation. Behaviour related to PDs has three characteristics: a varying
degree of repetition, which means that future behaviour can be predicted;
some degree of generalizability in which specified behaviour is shown in a
range of situations; and some impairment or distress. But even relatively
uncomplicated behaviours, such as driving recklessly, are subject to
variability when repeated, and precise specification of behaviour is difficult.
A trait, such as sociability, is the most commonly used basic unit to
describe and define PDs, and consists of a group of related behaviours. But
each trait is also a theoretical entity which involves relationships between
aspects of higher mental activity. In the main current definitions of PDs,
traits vary in the number of related behaviours, in the degree of reliance on
specified behaviours and on the nature of associated aspects of higher
mental activity, such as motivation. The trait approach to PD also involves
higher-order theoretical entities, in which several traits are combined to
form a smaller number of broader dimensions, while lower-order or 'fine-
grain' examples of specified behaviour have been included in definitions of
some features of PDs. A trait has been considered as a summary of
consistencies of behaviour, reflecting the individual's characteristics, the
situation and the individual-situation interaction, so that a PD will always,
to some degree, reflect the individual's environment.
An approach to PD assessment also depends on concepts of the
relationships between PDs, normal personality and other psychiatric
disorders. Paranoid and schizotypal PDs may be relatively discrete and
related to schizophrenia (Grove & Tellegen, 1991; Livesley, 1991), while
195
196 Assessment of personality disorders
most PD traits appear to be on a severity-related continuum with higher
mental activity and behaviour shown by the normal population (Cloninger,
1987a; Schroeder&Livesley, 1991). For such traits a dimensional approach
to assessment would seem to be indicated.
Classification of personality disorders

PD assessment has to be evaluated in the context of the main classificatory
systems, which are based on polythetic sets of defining criteria, each
requiring a minimum number of positive ratings. But such a system is not
necessarily ideally suited for all the purposes of classification, which include
description of behavioural phenomena, distinguishing between PDs, or
differentiating PDs from other mental disorders. Different criteria might be
employed for different purposes; for instance the presence or absence of a
clear desire for affection and acceptance would be expected to discriminate
between avoidant and schizoid PDs.
Despite their shortcomings, categorical 'all-or-none' PD diagnoses,
made on the basis of a minimum number of positive criteria, have been in
widespread use for many years, suggesting that such an approach has some
value. But although a categorical PD diagnosis can convey a considerable
amount of information, the disadvantages of the main classification
systems have been frequently described. For instance, for each criterion,
and for each PD, the threshold between disorder and normality is often
arbitrary and difficult to assess. Also, an 'all-or-none' rating generally loses
information, although it has been argued that determining a clinically
relevant threshold is more useful than measuring the degree of a character-
istic (Widiger, 1992). Other problems with the DSM-III-R system for PDs
have included overlap between features in the various PD syndromes, the
time required to assess over 100 criteria, and the heterogeneity of each of the
individual PDs due to the polythetic system. Also, this type of classification
makes no allowance for the occurrence of very severely maladaptive
examples of a limited number of criteria, which are insufficient to produce a
PD diagnosis. Possible modifications to the DSM-III-R system include the
provision of essential core criteria, weighted criteria and a greater use of
exclusion criteria, perhaps involving the presence of normal personality
features.
'Core' criteria can be considered as those which are either necessary or
sufficient for a categorical diagnosis (Reich, 1992). In the case of borderline
PD, Hurt et al. (1990) have claimed, on the basis of statistical analysis of
previous studies, that there are three core dimensions, i.e. disorders of
Classification of personality disorders 197
identity, mood and impulsivity. However, although relatively few criteria

would be needed for diagnosis on this basis, a smaller number of criteria
may not do justice to the descriptive function of a classification. A variant
of the concept of core criteria consists of the identification of combinations
of two or more criteria, which have particular diagnostic significance when
both are rated positive. For example, Reich (1992) found that a combi-
nation of unstable and intense relationships, intense uncontrolled anger
and physically self-damaging acts, had high values for both positive and
negative predictive power for diagnoses which were based on all the
borderline PD criteria.
If some of a group of individuals have received the diagnosis of
borderline PD on the basis of the DSM-III-R criteria, the positive
predictive power of each criterion is the number of positive ratings in the
subgroup with borderline PD, divided by the number of positive ratings in
the entire group. Similarly, the negative predictive power of each criterion is
the number of negative ratings in those individuals who did not have a
borderline PD diagnosis divided by the number of negative ratings in the
entire group (Skodol et al., 1988). These ratios are one means by which each
individual criterion, or a combination of criteria, can be evaluated, when
two groups of subjects are being compared.
Fundamental changes in the assessment of PD for clinical and research
purposes have been suggested, in particular the use of dimensional ratings
for many PD features and for the severity of impairment. The theoretical
advantages are clear: this is more consistent with research findings related
to the identification of most PD features in both normal and clinical
populations; more information is retained; reliability of assessment should
increase; and arbitrary boundaries between normality and disorder are
avoided. A dimensional approach to PD assessment could also include
other traits and higher-order traits derived from the study of normal
personality; for example, the following five dimensions: neuroticism (with
persistent or recurrent anxiety, anger and depression), extraversion, open-
ness to experience, agreeableness and conscientiousness (Costa & McCrae,
1992). (The main adjectives defining these dimensions were calm/worrying,
reserved/affectionate, down to earth/imaginative, ruthless/soft-hearted,
and negligent/conscientious.) Also, the following eight dimensions of
personality functioning have been described (Frances, 1982): managerial,
responsible, cooperative, docile, self-effacing, rebellious, aggressive and
competitive. Each can present to an intense and maladaptive degree and
this model has two characteristics that appear to be relevant in any system
of description and classification: some features are more likely to be
associated than others, while other features are negatively correlated, for
instance, responsibility and rebelliousness.
Livesley (1991) has attempted to identify a dimensional classificatory
system for PDs based on clinical observations and considered that 'the
personality characteristics of general-population subjects differ from those
of patients with personality disorder only in degree'. Factor analysis of self-
report ratings of a range of items gave 18 dimensions involving affective
instability, anxiety, compulsivity, conduct problems, diffidence, identity
disturbance, insecure attachment, interpersonal disesteem, intimacy
problems, narcissism, passive oppositionality, perceptual-cognitive distor-
tion, rejection, restricted expression, self-harm, social avoidance, stimulus
seeking and suspiciousness. These gave four higher-order dimensions
involving generalized behavioural instability ('psychopathic entitlement'
with poor impulse control, poor socialization and need for stimulation);
social withdrawal or avoidance; dependence; and compulsivity (with
orderliness, conscientiousness, preoccupation with detail and a need to be
active) (Schroeder & Livesley, 1991). In relation to PDs, there have been
other descriptions involving four main dimensions; for example, Walton &
Presly's (1973) similar descriptions of hysterical personality, social avoid-
ance, submissiveness, and obsessional/schizoid characteristics. Tyrer &
Alexander (1979) also reported four similar major PD dimensions, involv-
ing sociopathy, schizoid features, passive-dependence and anankastic
(obsessional) features.
Cloninger (1987a) has described a method for description and classifica-
tion of both normal personality and PD, based on a theory that there are
three dimensions of stimulus-response characteristics, namely 'novelty
seeking', 'harm avoidance' and 'reward dependence'. The interactions of
varying degrees of these dimensions are hypothesized to explain the various
clinical presentations of PD. 'Novelty seeking' involves frequent explora-
tory activity with active avoidance of monotony; 'harm avoidance' is a
tendency to show marked response to aversive stimuli, including the
avoidance of novel stimuli; while 'reward dependence' is a tendency to show
marked response to rewarding stimuli, including approval and support.
This model is compatible with some categories based on clinical data; for
instance, antisocial PD can be considered to be a consequence of high
novelty seeking, while excessive dependence could result from high degrees
of harm avoidance and reward dependence.
Another concept which has been used in the context of the classification
of PD is the 'ideal type'. This refers to a theoretical approach to the nature
of, and relationships between, various PD traits, and can be considered as a
Non-methodological variables 199
hypothesis that gives a meaningful account of suggested relationships

between various PD characteristics. Therefore it provides a framework to
understand an individual patient (Livesley, 1991). An 'ideal type' might be
provided by a selected case vignette that is considered to have general
significance, although the presumed relationships between the features are
not based on empirical data.
Information sources
Data for PD assessment can be obtained from three main sources: the
patient, informants and observation in clinical settings over varying
periods. There is evidence that some PDs are associated with a higher
prevalence of the same or related disorders in first degree relatives, and with
some biological variables, so that assessment may also include family
history and physiological data (Reich, 1992).
Non-methodological variables that influence assessment

The evaluation of repeated behaviour is influenced not only by variables
associated with the methods used, including sample selection, but also by
several non-methodological variables, such as environmental changes and
the presence of other mental disorders. The latter would include a premen-
strual syndrome that can be associated with the intermittent appearance or
exaggeration of relevant PD behaviour.
Features of PD can show apparent remission associated with improve-
ments in any co-occurring mental disorders such as anxiety and depression
(McMahon, Flynn & Davidson, 1985; Reich et aL, 1987a; Joffe & Regan,
1988; Boyce et aL, 1989; Endicott & Shea, 1989; Mavissakalian, Hamann &
Jones, 19906; Stuart et aL, 1992), although certain features of antisocial or
narcissistic PD may become more pronounced if a co-occurring depressive
syndrome improves (Stankovic et aL, 1992). Peselow et aL (1994) applied a
structured interview for PD during a period when the subjects' mood was
depressed and after recovery, and found that PD features from clusters A
and C tended to decrease after successful recovery from depression, in
contrast to features of cluster B. Also, any specified PD may interact with
other co-occurring PDs or other mental disorders, with subthreshold PD
features and with normal PD characteristics.
In addition, the manifestations of many PDs, such as antisocial and
borderline PDs, change gradually over time. Such changes may be related
to social class, which requires matching in comparative studies.
Assessment: an overview
PD assessment routinely involves a clinician's 'present or absent' rating for
the predominant PD features in the history and presentation, based on a
relatively unstructured interview and data from various informants. Such
'routine clinical assessment' will be considered in a subsequent section of
this chapter. The interview with the patient should cover the main areas of
maladaptive behaviour relevant to PDs: antisocial behaviour, dependence
on others, problems in sustaining relationships, and attitudes towards
other people (Tyrer, 1989). It must be noted that some patients with PD
refuse to cooperate in any assessment method that involves any structured
procedure; in such cases, methods involving the clinician's ratings are
required.
There are three main types of assessment methodology: structured
interviews, self-report questionnaires, and consensus ratings (or diagnoses)
from various sources. This last type includes the 'LEAD standard', i.e.
'longitudinal expert evaluation using all available data' (Spitzer, 1983),
which is based on information from patient assessments, past records,
informants and observations over a period of inpatient or outpatient
contact. The final ratings or diagnoses are based on a consensus of several
experienced clinicians. The reliability of assessment is likely to be increased
by the use of more than one assessment method. Informants are often
important for PD assessment, in particular where the patient has another
mental disorder such as a depressive syndrome or substance abuse.
Zimmerman (1994), in a review of research methods, noted that inter-
rater reliability coefficients for the same interview, using standardized
interviews, is usually good, in contrast to poor reliability in relation to
unstandardized clinical assessments. As might be expected, test-retest
reliability coefficients are generally lower than those for joint-ratings of the
same interview, and decrease with increasing time between interviews.
Some PDs, such as antisocial PD, are less affected than other PDs by the
test-retest interval, while co-occurring disorders such as depressed mood
tend to influence the results of PD assessment by both self-report question-
naires and structured interview. Assessment is further complicated by a
tendency for subjects to report less psychopathology at subsequent inter-
views. Also, reliability of PD assessment would be expected to vary
depending on the selection criteria of the sample, for example it is likely to
increase if a sample of subjects with PD are selected on the basis of severity.
Information from the patient, from both interviews or self-report
questionnaires, may be associated with inaccuracies and 'response-bias'.
Assessment: an overview 201
But, although the questions that form part of interviews and questionnaires
usually have 'face validity', and are mainly designed to elicit a true
response, a question may still be useful if it produces a consistent response,
whether or not it is true.
There have been several indirect approaches to PD assessment, but these
have made little or no contribution to influential psychiatric research.
Several such methods, such as the Thematic Apperception Test (TAT), the
Rorschach inkblot test, and sentence completion tasks, present the patient
with stimuli that have ambiguous meanings. The patient's subsequent
response would be expected to reflect aspects of his/her higher mental
functioning. The TAT was originally introduced as a measure of personal-
ity, and consisted of a series of cards, each with a picture of a situation. The
patient was then asked to comment on what the situation might be. The
rationale for this had been described thus:
the test is based upon the well recognized fact that when a person interprets an
ambiguous social situation he is apt to expose his own personality as much as the
phenomenon to which he is attending. Absorbed in his attempt to explain the
objective occurrence, he becomes naively unconscious of himself and of the scrutiny
of others and, therefore, defensively less vigilant.
But although this is likely to be true, the evaluation and rating of the
responses present considerable problems.
Another focus for the assessment of PD is the cognitive aspect of
personality. Although current DSM-IV and ICD-10 criteria include some
motivations and attitudes, such phenomena have been insufficiently
addressed in the clinical literature. For example, an evaluation of repeated
acts of violence might explore the personal meaning of the violent acts to
the offender (Blackburn, 1989).
The disadvantages of the DSM-III-R and DSM-IV PD classifications
will affect any assessment method based on this system, and a frequent
problem has been that an individual has often received several co-occurring
DSM-III-R PD diagnoses that are cumbersome in clinical practice.
Perry (1992) has reviewed nine studies which compared two or more
assessment methods for DSM-III-R PD diagnoses, administered to the
same group of subjects. Reliability was low, with a median kappa of only
0.25. Agreement was lower between self-report questionnaire and struc-
tured interview than between two interview methods, and comparing
dimensional scores for the different methods made little difference. The
depressing conclusion was that, despite the good inter-rater reliability of
the various assessment methods (although there was a lack of test-retest
reliability data for longer periods than a few months), PD diagnoses were
not significantly comparable across methods. However, agreement may
improve in more severely ill populations (Reich, 1992).
It has been argued that the LEAD procedure, in particular if this involves
a period of observation and multidisciplinary clinical conferences in an
inpatient unit, is the best validity standard available with which to compare
other methods of PD assessment. This procedure was compared with two
structured interviews, and confirmed that diagnoses such as antisocial and
schizoid PDs, which depend mainly on readily-identified behaviour, are
relatively reliable between assessment methods. In contrast, passive-
aggressive, self-defeating and narcissistic PDs were diagnosed much less
often with structured interviews (Skodal et aL, 1990).
Methodological variables
Many relevant methodological variables affecting PD assessment have
been studied, especially in relation to procedures involving structured
interviews or self-report questionnaires. Such variables include the
patient's perception of confidentiality, the relationship of the clinician or
researcher with the subject, the duration of the period to be assessed, the
wording of the questions and various types of response-bias (Goldberg,
1972). The 'true/false' (or 'yes/no') format has been commonly used,
despite problems with arbitrary thresholds and loss of information, but if a
more complex rating scale is used, other types of response-bias and
different scoring methods need to be considered. These include an estimate
on a scale with fixed points or on a scale with a continuous line, or the
selection of alternative adjectives to describe higher mental activity or
behaviour. It appears that the optimum number of positions on a scale
should not generally exceed six. Also, if there are three or more points on an
interval rating scale, unwarranted assumptions may be made about the
distance between pairs of points, as the distances between the points is
difficult to quantify in terms of each PD variable being measured, and may
vary in clinical significance between different PD criteria. If several criteria
are being rated, it is possible to weight the score for those considered to be
essential to the diagnosis or of relative clinical importance, or to transform
scores. When there are three or more points on a rating scale, various types
of response-bias need to be taken into account, such as 'end-bias', 'middle-
bias', or a tendency to mark the left or right side of the scale. Methods to
counteract such problems include changing the pathological response side
of the scale to the left or right in a random manner, and to score a 4-point
Methodological variables 203
scale as 0011 instead of 0123. A 4-point scale would be expected to avoid a

central-tendency bias (Streiner & Norman, 1989).
When considering the wording of interview questions or questionnaires,
commonly understood words need to be chosen and the item should avoid
ambiguity, value-laden words, implied criticism and negative wording. For
example, 'I sometimes steal' might become 'I can't help stealing sometimes'.
A self-report method with a degree of reliability and validity requires an
underlying PD classification that covers both the cognitive and overt
behavioural aspects of a given trait. Also, when the presence of relevant
behaviour is being assessed, it is best to enquire about a range of examples,
and the underlying motivation may need to be evaluated; for instance,
behaviour related to anger may be in response to very different PD traits
such as sensitivity to criticism or fear of abandonment, while having no
close relationships may reflect a lack of motivation or, in the case of
avoidant PD, anxiety together with an underlying desire for social contact
(Shea, 1992).
Response-bias may affect assessment procedures, for instance a 'yes' (or
agreement) bias, a 'no' bias, a tendency to answer in a way that reflects
social desirability, and the 'hello-goodbye' effect. The latter consists of a
tendency to over-report problems at the start of a clinical contact and the
opposite bias at its conclusion. Specific PD psychopathology may also
affect response-bias, for instance, exaggerated answers may be associated
with histrionic, narcissistic, self-defeating and dependent PD features. An
additional problem with any response-bias is that it may vary depending on
the content of the question. Social class is an additional variable which may
contribute towards a response-bias, as there is an association between low
socio-economic status and general dissatisfaction. A particular PD trait
may be associated with a response-bias, for instance, a person with
obsessive-compulsive PD traits may be very meticulous and tidy but does
not rate him/herself in this way because of his/her impossibly high
standards. Also, individual patients may have an idiosyncratic bias, and
individual response-bias may be partly determined by the person's use of
various psychological defence mechanisms such as denial. Assessment of
response-bias can be attempted with additional questions which are
directed to characteristics such as exaggeration, acquiescence, denial and
lying.
Although self-report questions form the basis of both structured inter-
views and self-report questionnaires (SRQs), structured interviews may
have advantages over SRQs, as an experienced interviewer can omit
inappropriate questions, make a judgement as to the subject's concent-
Table 5.1 Structured interviews for the assessment of personality disorders

(PDs)
Assessment target Interview method

DSM-III or DSM-III-R SCID-II (Structured Clinical Interview for DSM-
PDs III-R Personality Disorder)
PDE (Personality Disorder Examination)
SIDP-R (Structured Interview for DSM-III-R
Personality Disorders)
DIPD (Diagnostic Interview for Personality
Disorders)
PIQ-II (Personality Interview Questionnaire II)
24 personality dimensions PAS (Personality Assessment Schedule)
and ICD PD categories
ICD-10PD categories SAP (Standardized Assessment of
Personality) - for informants only
18 subscales based on KAPP (Karolinska Psychodynamic Profile)
psychoanalytic theory
ration, cooperation and understanding of the questions, and can often

encourage compliance. Also, some subjects refuse to cooperate in question-
naire assessments, and their use and understanding of the spoken word is
often better than the written word. But SRQs may have significant
advantages for cooperative subjects: interviewer variables are reduced and
the application of a SRQ is a simple and cheap procedure that can be
applied to large samples in research studies. At present, SRQs are not
recommended for use in making categorical PD diagnoses in clinical
settings, but they can be used to compare groups of patients in research
studies and as screening instruments to identify areas for further evaluation
at interview.
Structured interviews
'Structured interview' is a term that incorporates 'semistructured' pro-
cedures in which 'probes' (i.e. further questions) can be added based on the
interviewer's clinical judgement. But as an interview can never be comple-
tely structured in respect to interviewer behaviour, the term 'semistruc-
tured' is more accurate even for the set questions. Such interviews can be
divided into those based on a comprehensive range of PD characteristics
defined by various classificatory systems (Table 5.1), and those directed
towards a relatively narrow range of specific features of PDs (Table 5.2).
Most interviews involve the patient (or subject), although some have
Table 5.2 Structured interviews for the assessment of specific features of

personality disorders (PDs)
Assessment target Interview method

'Borderline' personality disorder based DIB (Diagnostic Interview for
on various concepts Borderlines)
Kernberg's Structured Interview
SBP (Schedule for Borderline
Personalities)
BDP scale (The Borderline Personality
Disorder scale)
DSM-III-R Schizotypal PD and SIS (The Structured Interview for
related features Schizotypy)
SSP (The Schedule for Schizotypal
Personalities)
Symptom Schedule for the Diagnosis
of Borderline Schizophrenia
Narcissistic PD based on various DIN (The Diagnostic Interview for
concepts Narcissism)
Self-harm and aggression SAS (The Suicide and Aggression
Survey)
Social dysfunction APFA (Adult Personality Functioning
Assessment)
complementary informants' versions. However, the Standardized Assess-

ment of Personality (SAP, Mann et ai, 1981) is used only with informants.
SCID-II
The Structured Clinical Interview for DSM-III-R personality disorder
(Spitzer et al., 1987) has 120 items and involves a 60-90 minute interview
between clinician and subject. It is based on the DSM-III-R classification
and is designed to be used in conjunction with the SCID-I instrument (i.e.
related to the DSM-III-R axis I), which evaluates any co-occurring mental
disorders.
The items cover each PD in turn and each DSM-III-R criterion is
evaluated by a specified question (or questions) and subsequent specified
probes. But the clinician can also probe further and clarify, using his/her
judgement before rating each item on a 4-point scale: 'inadequate infor-
mation', 'negative', 'subthreshold' and 'threshold'. One disadvantage of
presenting all criteria for each PD in sequence is a risk of a 'halo effect',
when subsequent ratings are influenced by prior positive ratings for the
same PD (Widiger & Frances, 1987). A unique feature of the SCID-II
(which is also available as a computerized version) is the option to precede

the interview by a self-report questionnaire covering the same items.
Although such questionnaires have been shown to produce false positives
in comparison with interview ratings, there are very few false negatives, so
that the prior use of the questionnaire can identify a subset of items for
further evaluation at interview, while the others are omitted. The conse-
quent saving of time should encourage its use in routine clinical practice as
well as in research (Reich, 1992).
PDE
The Personality Disorder Examination (PDE) (Loranger et al, 1987;
Reich, 1989) can provide diagnoses and dimensional scores for each of the
DSM-III-R PDs, although the DSM-III-R structure is reorganized so that
items are presented in a sequence under six sections: work, self, interper-
sonal relations, affects, reality testing and impulse control. Each section
begins with open-ended questions and additional items relate to behaviour
at interview. The time required is usually up to 2 hours and there is a parallel
version for informants. At the beginning, the subject is told 'the questions I
am going to ask concern what you are like most of the time. I'm interested in
what has been typical of you throughout your adult life, and not just
recently. But if you have changed and your answers might have been
different at some time in the past, be sure to let me know'. Each of the 128
items is scored on a 3-point scale: 0 (absent or not clinically significant); 1
(present but of uncertain clinical significance); and 2 (present and clinically
significant). Also there is an 'uncertain' category.
Inter-rater reliability has been shown to be good with kappas of 0.7 or
above for individual PD diagnoses, while test-retest reliability for border-
line PD (as shown by testing by another rater prior to discharge from
hospital) gave a kappa of 0.57 (Widiger & Frances, 1987). But this
instrument has been criticized for relying too much on patients' opinions
and self-evaluations, as patients are not encouraged to provide examples to
support their answers. However, it has led to the development of a modified
version, the International Personality Disorder Examination (IPDE),
which has been used in several World Health Organization studies (Lor-
anger et ai, 1991).
SIDP-R
The Structured Interview for DSM-III-R Personality Disorders is an
updated version of the SIDP for DSM-III (Pfohl, Stangl & Zimmerman,
1982; Reich, 1989). It provides 160 questions for the subject, which are
arranged in 16 sections based on aspects of functioning, such as self-esteem,
dependency and social interaction. Questions are open-ended and the
interviewer can clarify and probe. Each criterion is scored on a 3-point scale
(0,1,2). Although the time required for the interview has been stated to be
up to \\ hours, this has been considered to be an optimistic claim. The
interviewer should be an experienced clinician, and a prior assessment of
other mental disorders is recommended. If a change in personality has
followed the onset of another mental disorder, the predominant features in
the last 5 years are accepted (Standage, 1989).
Studies of inter-rater reliability have given kappas above 0.6 for schizoty-
pal, histrionic, dependent, borderline and passive-aggressive PDs (Widiger
& Frances, 1987), although in most cases the same interview was assessed
by two raters. However, test-retest results after 6 months, involving the
same interviewer, gave an acceptable kappa of 0.68 for the presence of'any
PD', and of 0.7 for borderline PD, although there was poor agreement for
some other PDs (Reich, 1989).
In a study of 66 patients with depressive disorders (Zimmerman et al,
1988), there was poor agreement between SIDP ratings made on the basis of
the interviews with the patient and those involving a close informant; for all
individual PD diagnoses, kappas were below 0.35. Informants generally
reported more pathology for each individual PD, except for antisocial and
schizoid PDs; for example, borderline PD was diagnosed 3 times more
often on the basis of the informants' interview.
D1PD
The Diagnostic Interview for Personality Disorders (DIPD) (Zanarini et
al., 1987) has 252 questions in 11 sections, based on the DSM-III-R PDs.
Each item has 2-3 probes and it has been claimed, again optimistically, that
the interview usually takes up to 1 \ hours. Additional enquiries based on
clinical judgement are encouraged, and prior assessment of other co-
occurring non-PD psychiatric disorders is recommended. Each PD criter-
ion is scored as follows: 2 = present and definitely clinically significant;
1 = present and probably clinically significant; 0 = absent and clinically
insignificant. Inter-rater reliabilities for a group of inpatients were high (the
kappas were above 0.88) for all PDs except paranoid PD, although schizoid
PD was never diagnosed. Test-retest kappas after one week were above 0.6
for seven PDs, including antisocial PD (0.84) and borderline PD (0.85).
PIQ-Il
The Personality Interview Questionnaire II (PIQ-II, Widiger, 1987) is an
interview that has 106 questions divided into eight sections: self-descrip-
tion; self-confidence; work; relationships; emotions; social responsibility;
interpersonal sensitivity and aberrant behaviour; and perceptions and
beliefs. Each item is scored on a 9-point scale. Although it was originally
used by lay interviewers, clinical experience is preferred.
The following structured interviews do not generate DSM-III-R PD
diagnoses:
PAS
The Personality Assessment Schedule (PAS) was developed by Tyrer &
Alexander (1988) and subsequently revised. It is not based on either the
DSM or ICD classifications, but on 24 traits derived from the clinical
literature: pessimism, worthlessness, optimism, lability, anxiousness, suspi-
ciousness, introspection, shyness, aloofness, sensitivity, vulnerability, irri-
tability, impulsiveness, aggressiveness, callousness, irresponsibility,
childishness, resourcelessness, dependence, submissiveness, conscientious-
ness, rigidity, eccentricity and hypochondriasis.
Each is assessed on a 9-point severity scale after several specified
questions have been presented, although further questions are encouraged.
The interview takes up to an hour and can be administered to the subject or
an informant. Analysis of the scores can give one of three global categories
based on severity: 'personality difficulty', 'personality disorder' and 'severe
personality disorder'. Scores can be converted to the ICD PD categories,
but while some of the PD categories are similar to those of the DSM-III-R,
they are not strictly equivalent.
Good inter-rater reliability and acceptable test-retest reliability have
been reported (Ferguson & Tyrer, 1989), but there has been poor correla-
tion between ratings from patients' and informants' interviews (Widiger &
Frances, 1987). This instrument has been shown to have predictive validity
for improvement of depressive symptoms with drug treatment and for the
effects of an alcohol treatment programme (Reich, 1992).
SAP
The Standard Assessment of Personality (SAP) was reported by Mann et al.
(1981) and is unusual in that it is a short interview (about 20 minutes) with
an informant only. A revised version, which was modified in relation to the
ICD-10, has been recently published (Pilgrim & Mann, 1990; Pilgrim et aL,
1993).
The informant should have known the patient for at least 5 years during
which other mental disorders were not apparent. After open-ended
questions with prompts, the following ten probes are given:
how does he/she get on with people; does he/she have many friends; does he/she
trust other people; what is his/her temper like; how does he/she cope with life; is he/
she a calm sort of person; how much does he/she depend on others; how does he/she
respond to criticism; does he/she have unusually high standards - at work or home;
and what sort of opinion does he/she have of him/herself?
If the informant uses any of a set of key words suggesting PD features, the
following eight personality categories are evaluated further: paranoid,
schizoid, dissocial, impulsive, histrionic, anankastic, anxious and depen-
dent, with a set of seven questions for each category. The degree of
handicap is then assessed by three additional questions, and the informant
is also asked about the priority of handicaps if there are more than one PD
categories.
Satisfactory inter-rater reliability has been reported, and although test-
retest reliability after one year was variable and often poor, the kappa for
anankastic (obsessive-compulsive) PD was 0.74 (Widiger & Frances, 1987).
KAPP
The Karolinska Psychodynamic Profile (KAPP, Weinryb et ai, 1992) is a
structured interview based on psychoanalytic theory in which 18 subscales
are evaluated, involving the patient's self-image and relationships with
others. But these include items which would appear to present particular
problems of reliability, such as 'regression in the service of the ego', while
others, such as 'sexual functioning' would be influenced by many variables
other than those of PD.
The following structured interviews are directed towards specific features
of PD psychopathology (Table 5.2):
DIB
The Diagnostic Interview for Borderlines (DIB, Gunderson et ai, 1981)
was based on several concepts and meanings of the term 'borderline' in
previous literature. Although it reflects a similar concept to the definitions
of borderline PD in DSM-III and III-R, substantial 'convergent validity'
(i.e. agreement between DSM-III-R and DIB diagnoses) has not been
shown (Widiger & Frances, 1987).
The interview takes up to 90 minutes and covers five areas of functioning
(sections): social adaptation (e.g. school/work achievement, social activi-
ties, appearance/manners); impulse/action patterns (e.g. self-mutilation,
manipulative suicide threat or action, drug abuse, antisocial behaviour);
affects (e.g. depression, anger); psychosis (e.g. derealization, depersonaliza-
tion, brief paranoid experiences); and interpersonal relations (e.g. avoids
being alone, instability, devaluation of others, and dependency on others).
The 165 items involve questions (some related to specified behaviour),
together with observations during the interview. Most items are rated as
'absent', 'probable' (some) or 'yes' (much), and scored 0, 1 and 2 respecti-
vely. But two items have a maximum score of 4, i.e. 'repeated manipulative
suicide attempts' and 'severe dissociative experiences'. Each of the five
areas of functioning is then scored separately on a 0, 1,2 scale, and a
threshold score for the sum of the five subscores of at least 7 classifies the
patient as 'borderline' (Lewis & Harder, 1991).
The sections that relate mostly to symptoms (affects and psychosis) are
generally based on the preceding 3-month period, while the sections
involving social adaptation and impulse/action patterns are based on the
previous 2 years, and the 'interpersonal relations' section on the preceding 3
years.
Acceptable inter-rater reliability has been reported (Ferguson & Tyrer,
1989), as have examples of convergent validity involving associations with
other assessment methods and clinical diagnosis. But although the DIB has
not always discriminated between those with borderline PD and those with
other PDs (Reich, 1992), this may have been due to high rates of co-
occurring PDs. Widiger & Frances (1987) have discussed the various
strategies for borderline PD assessment in different settings and popula-
tions, involving adjustment of diagnostic thresholds and weighting of the
scores of individual sections.
Kernberg's Structured Interview (Kernberg, 1977)

This is based on three areas of'personality organization': namely, serious
distortions in reality testing, a predominant use of splitting defence
mechanisms, and serious difficulties in the formation of a cohesive identity.
But although this method provides a framework for the evaluation of
individual patients, the scores cannot be satisfactorily combined to provide
a diagnosis. However, ratings from the three areas have been associated
with DSM-III-R borderline PD diagnoses (Lewis & Harder, 1991).
SBP
The Schedule for Borderline Personalities (SBP) is part of a 70-item
'Schedule for Interviewing Borderlines', (Baron, 1981), but has not been
widely used. Each item is rated on a 5-point scale.
BPD scale
The Borderline Personality Disorder scale (BPD) (Perry & Klerman, 1980)
has 36 items involving questions for nine categories of relevant behaviour.
It takes up to 90 minutes to administer and the results can be related to
DSM-III criteria.
There is evidence for its reliability and validity but this instrument has
not been widely used (Reich, 1992).
SIS
The Structured Interview for Schizotypy (SIS) was designed as a research
instrument for assessing schizotypal symptoms and signs (Kendler, Lieber-
man & Walsh, 1989) and has been applied to 29 pairs of twins (Kendler et
al., 1991).
SSP
The Schedule for Schizotypal Personalities (SSP) (Baron, 1981) is part of
the 'Schedule for Interviewing Borderlines', together with the SBP as noted
above. This assesses ten areas based partly on DSM-III criteria for
schizotypal PD: illusions; depersonalization/derealization; ideas of refer-
ence; suspiciousness; magical thinking; inadequate support; odd communi-
cation; feedback from others; social isolation; social anxiety; and transient
delusions/hallucinations. (Abnormal scores in four or more areas identifies
DSM-III schizotypal PD.) However, many items rely too heavily on self-
evaluation or on previous 'feedback' from others to the patient. But good
inter-rater reliability has been shown (Widiger & Frances, 1987).
Symptom Schedule for the Diagnosis of Borderline Schizophrenia

This structured interview was reported by Khouri et al. (1980) and covers
features such as altered perceptions of body-image/self or of the world
around, auditory hallucinations, abnormalities of thought pattern and
content, ideas of reference, ideas of persecution, preoccupation with
aspects of sex and violence, and episodes of self-harm.
DIN
The Diagnostic Interview for Narcissism (DIN) (Gunderson et al, 1990)
takes up to 50 minutes and evaluates 33 aspects of narcissism by 105
questions within the following five sections: grandiosity; interpersonal
relations; reactiveness; affects and moods; and social and moral adap-
tation. The sections cover three time frames: the previous year (affects and
mood states); the previous 3 years (grandiosity, interpersonal relations, and
reactiveness); and the previous 5 years (social and moral adaptation). The
'grandiosity' section explores whether the person has unrealistic views
about special abilities, invulnerability, self-sufficiency and superiority. The
section on 'interpersonal relations' focusses on issues such as a tendency to
idealize others, lack of empathy, devaluation or exploitation of others, and
feeling entitled. 'Reactiveness' is concerned with unusually intense reac-
tions to criticism, defeat or disappointment, within a background of
extreme sensitivity. Relevant 'affects and moods' include sustained periods
of emptiness, boredom and a feeling of meaninglessness. 'Social and moral
adaptation' for the narcissistic individual may be associated with high
achievement, but the subject has self-serving values and morals. 'Grandi-
osity' is assessed first, because it was felt that open responses were more
likely to occur early in the interview, as some subjects can become more
defensive as the interview progresses. All items are scored as 2 (much), 1
(some), or 0 (none). Good inter-rater reliability has been reported.
SAS
The Suicide and Aggression Survey (SAS) (Korn et ai, 1992) is a structured
interview that was developed to evaluate various aspects of these complex
behaviours and to help the prediction of suicide and violence. The interview
provides information about recent and past aggression, as shown by
suicidal and violent thoughts, threats and actions. Other aspects covered
include predisposing factors, precipitating events, underlying emotions,
nature of the action, effects of the action and functions of the act.
The interview consists offivesections: general background; screening for
history of self-harm and aggressive behaviours; ratings of such behaviours;
the assessment of contextual and cultural factors; and further detailed
information related to self-harm and aggressive acts. Aspects that are
covered include the use of drugs in relation to an incident, and stressful life
events that occurred in the preceding month.
Self-report questionnaires 213
APFA
The Adult Personality Functioning Assessment interview (APFA) (Hill et
al.9 1989) is based on the concept that most individuals with PD share a
'pervasive and persistent abnormality in social functioning'. This instru-
ment was an attempt to provide a standardized assessment of the subject's
functioning over a period, in six different social circumstances (sections):
work; love relationships; friendships; non-intimate social contacts; nego-
tiations; and everyday coping. In each section, the interviewer rates level of
functioning on a 6-point scale (0 = unusually effective, to 5 = pervasive
failure). As social circumstances vary during life, APFA usually rates the
period from age 22 to 30, although other age periods can be used, and
behaviour is only rated for time periods when opportunities for social
interaction have been present. The interviewer is concerned with reported
behaviour rather than the subjects' evaluations or attitudes, and high inter-
rater reliability has been reported.
Self-report questionnaires (SRQs)

Influential SRQs include the PD scales of the Minnesota Multiphasic
Personality Inventory (MMPI), the Millon Clinical Multiaxial Inventory
(MCMI), the Tridimensional Personality Questionnaire (TPQ) and
versions of the Personality Diagnostic Questionnaire (PDQ). However, the
MMPI was not designed in relation to ICD or DSM PD categories, and the
TPQ is concerned with Cloninger's three hypothesized PD dimensions.
The following group of SRQs attempt to evaluate a range of PD features
(Table 5.3):
MMPI
The Minnesota Multiphasic Personality Inventory has over 500 questions
and was developed by Hathaway & McKinley in the 1930s. It also rated
other non-PD mental disorders (Widiger & Frances, 1987), but as it was not
based on the DSM, relationships between the MMPI scales and DSM-III-
R disorders are variable and often uncertain. Therefore, this instrument has
not been designed for PD diagnosis using current classifications, although
Morey, Waugh & Blashfield (1985) developed 11 scales (the MMPI
Personality Disorder Scales) related to the DSM-III PDs.
A currently-available version of the MMPI (from NFER-Nelson, Free-
post, Windsor, Berkshire SL4 1BU, UK) takes up to 95 minutes and has 14
Table 5.3 Self-report questionnaires (SRQs) for assessment of personality

disorders (PDs)
Assessment target SRQ

A range of PD features:
PDs (based on various concepts) and MMPI (Minnesota Multiphasic
other mental disorders Personality Inventory)
PD traits that can be related to DSM- MCMI-II (Millon Clinical Multiaxial
III-R diagnoses Inventory - revised)
DSM-III/III-R PD criteria PDQ and PDQ-R (Personality
Diagnostic Questionnaire and the
revised version). Also, a modified
version: the STCPD (Screening Test
for Co-morbid Personality
Disorders)
WISPI (Wisconsin Personality
Inventory)
13 proposed PD traits SNAP (Schedule for Normal and
Abnormal Personalities)
Three proposed dimensions of normal TPQ (Tridimensional Personality
and abnormal personality Questionnaire)
'Extraversion/introversion' EPI and EPQ (Eysenck Personality
'neuroticism/stability' and Inventory and Questionnaire)
'psychoticism' dimensions
Proposed personality dimensions KSP (Karolinska Scales of
Personality)
16 proposed personality dimensions Cattell's 16 Personality Factor Scales
Specific PD features:
Borderline PD BSI (Borderline Syndrome Index)
Schizotypal PD Claridge's schizotypal and borderline
scales and a combined Schizotypal
Traits Questionnaire
Narcissistic PD Narcissistic Personality Inventory; the
Narcissistic Trait Scale; O'Brien
Multiphasic Narcissism Inventory;
scales from the California
Psychological Inventory and the
MMPI
Antisocial PD Self-Report Psychopathy Scale; Buss
& Durkee's Hostility Inventory
scales: hypochondriasis; depression; hysteria; psychopathic deviate; mas-
culinity/femininity; paranoia; psychasthenia; schizophrenia; hypomania;
social introversion and four validity scales.
MCMI-II
The MCMI-II is the updated version of the Millon Clinical Multiaxial
Inventory (Millon, 1987) consisting of 175 questions, which are answered
as true or false. The original version evaluated eight personality patterns:
schizoid, avoidant, dependent, histrionic, narcissistic, antisocial, obsessive-
compulsive, and passive-aggressive, together with some features of other
mental disorders (Lewis & Harder, 1991). The revised version was designed
to bring the scoring more in line with the DSM-III-R PDs and some 8-week
test-retest reliability data were described as acceptable (Reich, 1989,1992).
PDQ
The Personality Diagnostic Questionnaire (PDQ) is a 163-question, true-
false questionnaire, which was closely based on the wording of the DSM-III
PD criteria. A lie scale and five questions for an 'impairment and distress'
scale were also included. The revised version, with 155 questions (PDQ-R)
(Hyler et ai, 1988) reflected the changes in DSM-III-R and takes about 30
minutes to complete.
It is uncertain whether some items involving socially undesirable behav-
iour should be evaluated by relatively subtle questions, but it has been
generally considered that most patients are compliant and that 'items will
perform optimally when their content clearly relates to what they intend to
measure' (Wrobel & Lachar, 1982). In the early versions, adequate one-
month test-retest reliabilities were found for most scales but were low for
narcissistic, histrionic, dependent and passive-aggressive PDs (Hyler et aL,
1990). But few data are currently available for the PDQ-R in relation to
reliability. The PDQ-R has all questions for a single PD on one page, and
around 20% of pathological answers are 'false' in an attempt to reduce a
'true' response-bias. There is a consensus that the PDQ format tends to
overdiagnose PDs compared with other methods, especially schizotypal
PD (Reich, 1989). But the PDQ-R has been used to identify those questions
to be evaluated further at interview, and an informant's version has been
developed (Dowson, 19926). It has been claimed that, for many patients,
the total score of positive PD criteria is an indication of overall severity, and
that a shorter modified version (the STCPD - Screening Test for Co-morbid
PDs) can be used as a screening test for routine clinical evaluation (Hyler et
al, 1990; Dowson, 19926). Yeung et al (1993) found that improved
concordance with an interview method could be obtained for the PDQ-R
by adjusting the threshold for 'case' (i.e. diagnosis) identification.
WISPI
The Wisconsin Personality Inventory (Klein, 1985) has 360 items related to
DSM-IIIPD criteria, interactional styles, social desirability, response bias
and global ratings of work and social adjustment. Each item is rated on a
10-point scale. Validation studies are awaited.
SNAP
The Schedule for Normal and Abnormal Personality (SNAP) (Clark,
1989), contains 375 self-report true-false items. It is not based on DSM-III,
and has 13 scales with three higher-order dimensions.
TPQ
The Tridimensional Personality Questionnaires (TPQ) (Cloninger, 19876)
is based on Cloninger's proposed three dimensions of normal and abnor-
mal personality, i.e. novelty seeking, harm avoidance and reward depen-
dence. It has 100 true-false questions and has been used in a study of
patients with alcoholism (Cloninger, 1987c). A relationship between the
TPQ and DSM-III-R PD features has been demonstrated (Goldman et al.,
1994).
EPIandEPQ
The Eysenck Personality Inventory (EPI, Eysenck & Eysenck, 1964)
(currently available from NFER-Nelson) takes about 15 minutes to
complete, and measures the dimensions of 'extraversion/introversion' and
'neuroticism/stability'.
The Eysenck Personality Questionnaire (EPQ, Eysenck & Eysenck,
1975) (also currently available from NFER-Nelson), in addition to the
above dimensions, provides a measure of'psychoticism'. The questionnaire
was standardized on a large normal group and in various groups with
psychiatric diagnoses.
KSP
The Karolinska Scales of Personality (KSP) is a questionnaire based on
theories of biologically-based personality dimensions (temperaments)
(Weinryb ef a/., 1992).
CattelVs 16 Personality Factor Scale

This system has been used for many years in psychological research but has
had little impact on the study of psychiatric populations. Sixteen factors
were described in dimensional form, for example, 'outgoing' versus
'reserved' (Frances, 1982).
Various other SRQs relate to specific PD features; for instance, the
Borderline Syndrome Index (BSI) (Conte, Plutchik & Karasu, 1980), which
has 52 items. This method was not able to distinguish between groups
selected on the basis of one of various principal diagnoses, including
borderline PD, and was considered to offer a measure of general maladap-
tion with chronic illness (Remington & Book, 1993). Schizotypal and
borderline scales were described by Claridge & Broks, (1984); and schizoty-
pal PD can also be assessed by a Combined Schizotypal Traits Question-
naire (CSTQ) (Bentall, Claridge & Slade, 1989). Raine & Allbutt (1989)
have described five further scales for related features.
Several SRQs for narcissism have been reported: for example the
Narcissistic Personality Inventory (Raskin & Hall, 1979); the Narcissistic
Trait Scale (Richman & Flaherty, 1987) with ten items each rated on a 4-
point scale; a 41-item O'Brien Multiphasic Narcissism Inventory which has
been fully reported (O'Brien, 1988); and narcissism scales from the Califor-
nia Psychological Inventory and the MMPI (Wink & Gough, 1990).
SRQs related to antisocial PD include the Self-Report Psychopathy
Scale (Hare, 1985). This has 29 items scored on a 1 to 5 point scale, and
varying convergent validity with other assessment methods have been
reported (Widiger & Frances, 1987). But there was more agreement
between interview-based ratings than between ratings which included a
SRQ. (It has been noted that SRQs for antisocial PD have the disadvantage
of relying on an antisocial subject's insight and honesty.) Buss & Durkee
(1957) fully reported their Hostility Inventory with the following sections:
assault; indirect hostility; irritability; negativism; resentment; suspicion;
and verbal hostility. Other instruments have been reviewed by Blackburn
(1989).
Miscellaneous SRQs include the Lazare-Klerman Trait Scale (Lazare et
aL, 1970) for histrionic PD; Sandier & Hazari's (1960) questionnaire for
obsessional personality; the California Personality Inventory (McCormick
et al., 1987); the 54-item socialization scale of Gough's California Psycho-
logical inventory, which measures the degree to which a person has
internalized social values and considers them binding (Gough, 1969);
Quay's behaviour dimensions for delinquents (Quay, 1977); a shortened
version of the Michigan Alcoholism Screening Test (Pokorny, Miller &
Kaplan, 1972); the South Oaks Gambling Screen (Lesieur & Blume, 1987)
which is a 20-item SRQ for the identification of pathological gamblers; and
the Interpersonal Dependency Inventory (IDI) fully reported by Hirschfeld
et al. (1977). The IDI consists of 48 items based on three main aspects:
emotional reliance on another person, lack of social self-confidence and
assertion of autonomy; and the measurement of dependence was reviewed
by Birtchnell (1991). Other SRQs include the Locus of Control of Behav-
iour (Craig, Franklin & Andrews, 1984); the Interpersonal Sensitivity
Measure (Joffe & Regan, 1988) with scales for interpersonal awareness,
need for approval, separation anxiety, timidity and fragile inner-self; the
Perceptual Aberration Scale, which is a true-false measure of body-image
and perceptual abnormalities (Chapman, Chapman & Raulin, 1978), and
other scales related to schizotypal PD features (Widiger & Frances, 1987).
Various aspects of validity for SRQs have been studied, in particular
convergent validity (i.e. agreement between different methods of assess-
ment), criterion validity (i.e. an association between the assessment and
other variables) and predictive validity (i.e. an association between assess-
ment and outcome).
Several studies have compared the convergent validity of the PDQ or
PDQ-R with structured interviews, and Hyler et al. (1989) compared PDQ
diagnoses with clinicians' diagnoses obtained by post. In this study, it was
noted that the PDQ diagnosed many more PDs than did the clinicians, for
instance, for paranoid PD, 44 versus 8. A similar finding has been noted in a
number of studies, and Hunt & Andrews (1992) reported one or more PDs
in 67% of their sample as diagnosed by the PDQ-R, while the PDE
interview yielded only 7%; also, Reich et al. (1987a) found a poor
convergence between the PDQ and the SIDP interview in 121 patients with
panic disorder and depression. In addition, Dubro, Wetzler & Kahn (1988)
reported a high rate of 'false positives' (in relation to other methods) with
the PDQ, and Hyler et al. (1990) described many 'false positives' with the
PDQ-R although there were very few 'false negatives'. An exception to this
trend was in Zimmerman & Coryell's findings (1990) in 697 non-patients,
17% of whom received at least one PD diagnosis with an interview but only
10% with the PDQ.
However, these reports do not mean there is little or no relationship
between some parts of SRQs and other PD assessment methods. Pfohl and
colleagues (1989) obtained a kappa of 0.72 for the presence of at least one
PD comparing the PDQ and SIDP interview in 45 non-psychiatric patients,
and, although concordance between PD diagnoses is not generally good
between a SRQ and an interview method, this varies between PD diagnoses.
Hyler and colleagues (1990) found kappas of around 0.6 between SRQ and
interviews in respect of avoidant and dependent PDs, while Hunt &
Andrews (1992) reported correlations of around 0.6 between the dimen-
sional scores from SRQ and interview for antisocial, avoidant and depen-
dent PDs, based on the number of positive criteria for a specific PD. Also,
when assessment of individual criteria are compared by SRQ and interview,
some provide a good concordance: Clark (1992) has compared a 'true/false'
SRQ with the SIDP-R interview, and good convergent validity was found
for some criteria for borderline and histrionic PD, involving impulsiveness,
recurrent suicidal behaviour, and expressing emotion with inappropriate
exaggeration. Conversely, items which showed poor convergent validity
included lack of empathy, and unstable and intense interpersonal relation-
ships. Factor analysis of positive scores for PD criteria obtained by both
SRQ and interview have shown that factors vary in the degree to which they
are influenced by the method of assessment.
Further evidence of examples of good convergent validity between SRQs
and other assessment methods has been obtained using a modified version
of the PDQ-R, which retained the basic format with wording closely related
to DSM-III-R criteria, and provided an informant's version (Dowson,
19926). In a study of 60 psychiatric patients with a provisional diagnosis of
PD, and their informants,fivePDs gave correlations (between subject and
informant) for the number of positive criteria in each PD, which were
significant at the 0.05 level, and when individual criteria were considered,
for example for narcissistic PD, three of the nine criteria for this PD yielded
significant kappas between patients' and informants' ratings, although only
one (a sense of entitlement) showed 'fair to good' convergence at a level of
0.62 (Dowson, 1992a).
Because of the lack of a clear standard with which to compare the various
PD assessment methods, it is important to obtain other forms of validity (or
usefulness), namely 'criterion validity' when PD assessment is associated
with other types of variable, including theoretical relationships and pre-
Table 5.4 Clinical ratings for the assessment of personality disorders

(PDs)
Assessment target Rating method

DSM-III-R PDs 'LEAD' standard (longitudinal expert
evaluation that uses all data)
PD related to antisocial behaviour PCL-R (Hare Psychopathy Checklist)
Criminal behaviour Criminal Profile Scale
Aggression MOAS (Modified Overt Aggression
Scale)
Risk of aggression ARP (Aggression Risk Profile)
dieted outcome. For example, the modified PDQ-R was completed by 74

psychiatric patients, and a factor analysis of the scores of positive criteria
for each of the PDs yielded three factors which defined three PD clusters
(Dowson & Berrios, 1991). These were similar to the three DSM-III-R PD
'clusters' for seven out of the 11 PD categories, as the first cluster included
paranoid, schizoid and schizotypal PDs, the second antisocial and border-
line PDs, and the third dependent and obsessive-compulsive PDs. A further
example of criterion validity of SRQs has been shown in relation to patients
with a history of anorexia nervosa who were divided into two groups
(Dowson, 1992c), firstly 'vomiters', defined by a history of self-induced
vomiting, and, secondly, 'non-vomiters' or 'restrictors' who had main-
tained low weight mainly or only by diet or exercise. Several previous
studies had reported personality differences between such groups and, in
this study, vomiters showed significantly higher scores on self-report
measures of borderline and antisocial PD criteria. Finally, Pfohl and
colleagues (1987) demonstrated predictive validity for the PDQ by relating
PD assessment scores with outcome measures, both at discharge and at 6
month follow-up, in a group of inpatients with 'major depression'.
Clinical ratings (Table 5.4)
The'LEAD9 standard
The process of improving the methodology for PD assessment has been
handicapped by the lack of an accepted standard with which to compare a
particular method. Spitzer (1983) attempted to solve this problem by
Clinical ratings 221
proposing that a PD assessment should be compared with a 'longitudinal
expert evaluation that uses all data' (LEAD). The three essential ingredi-
ents were the opinion of one or more expert clinicians, observations over a
period of time, and the use of all available information from past records,
informants and other members of a clinical team. Support for this approach
has been provided by Skodol and colleagues (1988), who reported evalu-
ations of 20 patients with severe PD, admitted for inpatient treatment. The
LEAD procedure involved multidisciplinary conferences at which behav-
iour in various unit settings was considered. PDs and traits were rated by
consensus, on a 4-point scale: 1 = no or very few traits, 2 = some traits,
3 = almost meets DSM-III-R criteria, and 4 = meets DSM-III-R criteria.
Although the reliability of this approach is likely to be suspect at times, it is
a welcome reminder that an experienced clinician's judgement is also an
important assessment method; for instance, when the LEAD procedure
was compared with the SCID interview, the interview produced fewer
diagnoses of narcissistic PD.
PCL-R
The Hare Psychopathy Checklist (Hare, 1991) is a rating scale for the
assessment of antisocial PD. Information is collected from records and an
informant (if available) as well as from a structured interview with the
subject. This covers school adjustment, work history, career goals,
finances, health, family life, sex/relationships, drug use, childhood/adoles-
cent antisocial behaviour, adult antisocial behaviour, and 'general
questions' related to aspects relevant to antisocial PD, such as previous
guilt. Twenty items are then rated on a 3-point ordinal scale (0,1,2), based
on the degree to which the information matches descriptions provided in a
scoring manual. The following items are rated: glibness/superficial charm;
grandiose sense of self-worth; need for stimulation/proneness to boredom;
pathological lying; conning/manipulative; lack of remorse or guilt; shallow
affect; callous/lack of empathy; parasitic lifestyle; poor behavioural
controls; promiscuous sexual behaviour; early behavioural problems; lack
of realistic, long-term goals; impulsivity; irresponsibility; failure to accept
responsibility for own actions; many short-term marital relationships;
juvenile delinquency; revocation of conditional release; and criminal
versatility.
The PCL-R is the result of many years of research on thousands of prison
inmates and forensic patients, and there is considerable evidence for its
reliability and validity in research settings. The manual recommends that

more than one rating for each patient should be made independently and
that the scores should be averaged.
Criminal Profile Scale

The Criminal Profile Scale (Gunn & Robertson, 1976) is a rapid and simple
5-point rating (i.e. 0 to 4 based on severity) for each of the following: theft,
fraud, sex, violence, motoring, drink, drugs, financial gain and relationship
to drinking.
MOAS
The Modified Overt Aggression Scale (Kay, Wolkenfeld & Murrill, 1988a)
is a revised version of the method of Yudofsky, Silver & Jackson (1986)
designed for psychiatric inpatients. The following four categories are rated
on a 5-point scale over a specified observation period: verbal aggression,
aggression against property, autoaggression, and physical aggression. The
rater is asked to rate the most severe point on the scale relating to the most
serious aggressive act, and the scores are weighted to reflect the increasing
severity of the above categories.
ARP
An Aggression Risk Profile (ARP) for psychiatric inpatients, including
those with PD, was reported by Kay, Wolkenfeld & Murrill (19886) and
included data on age, sex, previous hospitalization, psychiatric diagnosis,
history of aggression and a clinical profile based on information about the
previous week.
Routine clinical assessment
Overview
Routine clinical assessment does not generally include a questionnaire,
structured interview or formalized rating. However, the clinician needs a
procedure and a set of guiding principles.
It can be helpful to keep in mind Walton's (1973) concept of a 'mature' or
'normal' personality, which is an ideal view of good adaptation. A 'mature'
person has correct self-perception, can adjust to new situations, can operate
with appropriate autonomy, has good reality testing, and has an integrated
pattern of attitudes, feelings and behaviour. (A lack of integration would be
shown by a policeman who engages in petty theft.) Thus, 'normal'
individuals have been able to find ways of expressing their behavioural
patterns in adaptive ways. A detailed history of the patient's close and
significant relationships is of particular importance in evaluating PD.
Some structure can, however, be appropriate to clinical interviews, for
example, the use of screening questions such as those suggested by Leff &
Isaacs (1990), i.e. in relation to ICD-10 categories, 'do you find you can
trust people?' and 'are you mostly treated fairly?' (for paranoid PD); 'can
you mix easily?' and 'do you prefer to be alone or with company?' (for
schizoid PD); 'have you been in much trouble with the police?' and 'how do
you get on with people in authority?' (for dissocial PD); 'do you frequently
lose your temper?' and 'have you ever hurt anyone or caused any serious
damage?' (for emotionally unstable PD); 'are you sometimes over-
emotional?' and 'do you like to be the centre of attention?' (for histrionic
PD); 'do you always try to follow a set routine?' and 'do you prefer things to
be very neat and tidy?' (for obsessive-compulsive or anankastic PD); 'do
you tend to feel very tense and self-conscious?' and 'do you live cautiously
and avoid taking unnecessary risks?' (for anxious or avoidant PD); and 'do
you tend to rely on others excessively?' and 'do you prefer others to make
decisions for you?' (for dependent PD).
An interview with an informant can often be important, but a judgement
has to be made about the accuracy of this account, particularly if the
informant's relationship with the patient has significant maladaptive
aspects.
It must always be remembered that a 'cross-sectional' evaluation of a
patient may not provide a typical example of the patient's behaviour,
particularly in the presence of other mental disorders (such as depression or
anxiety) or soon after stressful life-events. Under stress, a normally
considerate and reasonable individual may become selfish, irritable and
demanding, and it is important for the clinician not to allow his/her
emotional reaction to the patient to influence the assessment. Also, it is
important to remember that there is evidence that clinicians can inappro-
priately assign PD diagnoses (such as histrionic and antisocial PDs) partly
on the basis of the patient's gender (Ford & Widiger, 1989; Adler, Drake &
Teague, 1990). As with any patient in psychiatric practice, it is important to
understand the relevance of the presenting complaint and why the patient
has been referred at that particular time.
The assessment, including family and personal history, should be
explored in a systematic way, for example under the following headings:
description of patient; the patient's complaints; history of present
problems/symptoms; previous contact with medical services; family
history; personal history (school record, work record, sexual history, police
record, present domestic situation, past and recent social network, recrea-
tional habits, and alcohol, tobacco and drug history); aspects of personality
(relationships, mood, dependency on others, impulsive aggression, solitari-
ness, inability to cope with routine demands, undue sensitivity to criticism,
undue mistrust of others, conscientiousness, a tendency to be excessively
methodical and thorough, self-confidence, timidity, worry, level of activity,
fantasy, and reaction pattern to stress); and mental state at interview.
Routine assessment of 'cluster A' PDs (paranoid, schizoid, and

schizotypal)
Patients with paranoid PD traits may have a variety of presenting com-
plaints, such as tiredness, insomnia, irritability, anxiety and difficulties in
coping with work. It may take some time (and several interviews) before the
underlying psychopathology is revealed, perhaps by a tendency to blame
others, to be easily offended and to feel that others are generally hostile.
Initially, such individuals may be more than usually concerned with issues
of confidentiality. Accounts of various relationships may indicate a pattern
of hostility and varying degrees of conflict. The patient may complain that
he/she has been overlooked for promotion or that he/shefindsit difficult to
assert him/herself.
Those with schizoid and schizotypal PD traits usually appear detached,
unemotional, cold, quiet and shy. Persons with schizoid PD traits do not
often request medical help and some other 'significant' person may ask for
advice. Medical involvement may be precipitated by some event that
interferes with the person's long-standing pattern of social isolation, such
as a physical illness that makes him/her less independent, or the loss of a job
due, perhaps, to the person's maladaptive functioning (Turkat, 1990). At
interview, those with schizoid traits often appear unemotional, aloof,
uncomplaining and relatively unresponsive to social cues. Schizotypal
features may include the patient appearing to be odd, anxious and vague
with a bizarre appearance, dress and mannerisms. Some individuals with
these features may apparently wish for social relationships but find these
difficult, and their talk may have persecutory or grandiose themes.
Routine assessment of'cluster B' PDs (antisocial, borderline, histrionic

and narcissistic)
When individuals with antisocial PD traits present to clinicians, their
motivation for seeking advice and help must be evaluated. Sometimes
antisocial individuals can have superficial charm, which must not be taken
at face value. Turkat (1990) has described three types of presentation:
'clear' sociopaths, when the history of antisocial features is obvious and the
person may be requesting help to deal with legal consequences; 'clever'
sociopaths, who deliberately present the clinician with a non-existent
psychological problem (such as hallucinations or depression) to try and
obtain the status of a patient rather than a law-breaker; and the 'hurting'
sociopath, who wants relief from genuine symptoms such as anxiety and
depression. But of course, these are not mutually exclusive. If the patient
appears motivated to change (for instance, in relation to a pattern of
domestic violence when his wife has threatened to leave), it is important to
explore the patient's expectations and whether these are realistic. Some of
the core psychological features of antisocial PD may be indicated by the
person continually stating that his/her actions are justified despite the
adverse effects on others, and showing a lack of concern for the conse-
quences. If future dangerousness is being assessed, it is important to have a
detailed record of previous dangerous behaviour, any precipitating and
relevant environmental factors, and the characteristics of any victim
(Pollock, McBain & Webster, 1989).
The inconsistency and variety of presenting complaints are often charac-
teristic of individuals with borderline PD, and the complaints may appear
unusual. Other pointers to a 'borderline PD' diagnosis include frequent
crises, extra demands and phone calls to medical services, apparent
misinterpretation of the clinician's remarks or advice, strong reactions to
changes in the arrangements (for instance, to a proposal to reduce the
frequency of appointments), a low tolerance for direct eye contact or
physical contact or close proximity, and ambivalence about a continuing
clinical contact (Beck & Freeman, 1990).
Histrionic PD traits may be suspected if the patient draws attention to
him/herself in relation to general behaviour, mannerisms, gestures, flam-
boyant appearance or dress (sometimes with inappropriate sexual seducti-
veness), and shows some variability of mood in the interview, perhaps
becoming suddenly tearful but recovering rapidly. Such individuals often
appear generally lively with frequent and exaggerated smiles and gestures.
They appear to invite attention, and while some are concerned with eliciting
reassurance, approval and praise, others seem to want to control people

they come into contact with, by their manipulative, dramatic behaviour
(Turkat, 1990). Such persons tend to be self-centred, unaware of the effects
of their behaviour on others, and may talk in a vague manner, which may
involve exaggerated all-inclusive adjectives but lacks detail. This can be
very striking to a clinician trying to take an initial history, whofindsthe task
almost impossible. Patients with histrionic PD often present to medical
services when they are unhappy and anxious, perhaps related to 'sepa-
ration-anxiety' or a crisis in their relationships. This discomfort provides
the motivation for seeking help but when these symptoms have improved,
the motivation is lost. This explains why many such patients fail to keep
subsequent appointments that are made in a time of crisis. An experienced
clinician can often predict a high risk of non-attendance at a first appoint-
ment by identifying examples of the pattern of behaviour for 'histrionic PD'
in a referral letter. Also, presentation of histrionic PD may be associated
with substance abuse, conversion disorder and somatization disorder.
Useful avenues of exploration are previous close relationships, any com-
plaints that the other individuals in these relationships made to the patient,
and activities that are generally enjoyed. Subjects with histrionic PD
generally like situations in which they are the centre of attention.
Those with pronounced narcissistic PD traits can appear to be trying
hard to impress the clinician by showing 'how good I am' (Turkat, 1990)
and can be difficult to interview because of a demanding, arrogant manner
with expectations of special consideration. They may question the clini-
cian's qualifications and try to negotiate special arrangements for their
appointments. The clinician is either idealized or devalued, and a change
can occur rapidly. Attempts to explain the effects of their behaviour on
others are considered as criticism and met with anger. They appear to be
most comfortable when talking about themselves, often mentioning their
abilities, achievements and possessions, while complaining about the
shortcomings of others. They tend to pay a lot of attention to their
appearance and dress; this concern can be noticeable at interview if the
subject frequently smooths his/her hair, adjusts clothes or over-reacts to
minor physical problems which affect appearance. They may resent any
kind of formal testing, such as completing questionnaires, as this may be
taken to imply that their problems can be related to similar problems of
others and are not unique. Motivation for clinical consultation can include
a depressed mood in the context of relationship difficulties or the patient
may wish to show his/her partner that he/she 'is trying' in response to an
ultimatum about selfish behaviour.
Routine assessment of'cluster C* PDs (avoidant, dependent, obsessive-

compulsive, and passive-aggressive)
The individual with avoidant PD may complain of anxiety, depressed
mood, and associated somatic symptoms, related to a widespread fear of
criticism and disapproval, and the clinician must remember that such a
patient is particularly sensitive to evaluation.
The person with dependent PD appears submissive and self-effacing, has
low self-confidence, is self-critical and, like those with avoidant PD, seeks
to please and to receive reassurance. It is particularly important to try and
explore the details of the individual's current supportive relationships, how
decisions are made and how he/she feels about being alone. At interview,
the clinician may observe that the patient avoids conflict where there is
disagreement. Presenting complaints often involve anxiety and depressed
mood, and the patient's passive helplessness may produce an emotional
reaction in the clinician, which can signal the nature of the problem. In this
situation some clinicians feel impatient and irritated, while others feel
tempted to make special arrangements for extra 'help', because it seems
difficult to set a limit to the patient's demands.
Patients with obsessive-compulsive (obsessive) PD show variable combi-
nations of indecision and rigid attention to order and detail. They may be
very neatly dressed and initially cooperative, polite and unemotional,
giving their history in a very detailed manner, although sometimes there
may be hesitancy due to anxiety. But this can cause some problems to the
clinician, particularly at the first interview, as it is difficult to interrupt and
clarify points of clinical importance. The patient may have a written list of
notes and is obstinately determined to tell you what he/she feels is relevant,
however long it takes. Such persons try to be unemotionally logical, and
find it difficult to describe their feelings and attitudes. Presenting problems
include depressed mood associated with failure to achieve unrealistically
high standards, or worry about illness. In a clinical situation, obsessive
patients are generally unemotional, and are most comfortable when the
interaction is organized and when details are considered. In general,
cognitive therapy may be better tolerated than unstructured counselling or
joint sessions with his/her partner.
Persons with passive-aggressive PD traits may have difficulty in formu-
lating a complaint other than vague dissatisfaction, or that someone else
has insisted that he/she takes advice. Because an essential feature of this
disorder is poor performance in unwanted tasks, he/she will be a 'poor
patient' if not motivated. Even if relief from a depressed mood or anxiety is
sought, he/she may blame others for the problems and tend not to
cooperate in giving information. Such individuals may give short, incom-
plete answers and appear irritable. They may feel they should not have to
answer some questions and tend to talk about perceived interference from
others. But they are less wary than those with marked paranoid PD features
(Beck & Freeman, 1990). If they cannot see that they have contributed to
the previous problems, a poor response to advice would be expected.
Differential diagnosis and co-occurrence

Reference has been made to the frequent co-occurrence of PDs with other
psychiatric disorders. Such associations may involve causal mechanisms or
occur by chance but, whatever their nature, co-occurrence and interactions
can affect the presentation and outcome of each disorder. Even when there
is a clear history of PD, the clinician must be alert to the possibility that
there may be additional disorders requiring diagnosis, especially if the
history is inconsistent with an apparently understandable interaction
between PD and life events. Sometimes PD can mask disorders such as
hypomania, an unrelated depressive disorder, or neurological disorder.
The frequent co-occurrence of more than one PD has also been des-
cribed, and the clinician must be aware that certain PD features are more
likely to co-occur than others. For example, histrionic with borderline,
histrionic with narcissistic, narcissistic with antisocial, narcissistic with
passive-aggressive, avoidant with schizotypal, and avoidant with depen-
dent (Oldham et al.9 1992).
Personality disorder assessment: future directions

Despite the aspects of validity that have been demonstrated for self-report
questionnaires, those currently available do not appear suitable for the task
of assigning specific categorical PD diagnoses in a clinical context, as their
diagnostic thresholds are generally lower than those related to interviews or
clinical judgement. But despite this and other limitations, a variant of the
PDQ self-report questionnaire is currently being used as a screening
procedure to identify items for subsequent assessment by the SCID-II
interview method. Also, a shorter version of the modified PDQ-R has been
used, which assesses those PDs that appear to be most prevalent in routine
psychiatric practice, namely borderline, histrionic, avoidant and dependent
PDs (Dowson, 1992ft). This can be used before a routine clinical interview
as well as for research purposes. Even this simple approach in routine
practice would encourage more clinicians to look for a range of features
from several individual PDs. This appears to reflect clinical reality, as
discrete PDs are rare.
It is, perhaps, surprising that aspects of validity for such a relatively crude
instrument as the PDQ have been demonstrated in several studies. But this
is encouraging, and the potential for the increased use of self-report
methodology appears to be considerable. Also, an intermediate format
between self-report and interview, involving actively supervised self-report
ratings, with some explanation and encouragement from the interviewer,
deserves further attention.
Self-report questionnaires, structured interviews and clinical ratings will
all have a place as part of a range of different methods of assessment, and it
has been claimed that 'multiple methods of measurement should be the
norm ...'. Also, all methods of PD assessment will need to consider
variables such as other co-occurring psychiatric disorders, the time-frame
being evaluated, whether questions in structured interviews or question-
naires are organized by PD diagnoses or by topics, other personality
attributes, the level of impairment and aspects of the environment.
It must be remembered that many patients with PD in psychiatric
practice refuse to cooperate with structured assessment methods or even
with a detailed unstructured interview; rating procedures by clinicians on
the basis of all available information are then required.
Part Two
Clinical management
6
Drugs and other physical treatments
J.H. DOWSON
Overview
For a patient with PD, a drug may act in several ways: by a non-specific
beneficial effect on a range of PD features, by a relatively specific effect, or
by targeting a co-occurring psychiatric disorder. Also, a placebo effect can
be marked.
Drugs are prescribed for many patients with PDs, although their effects
are often unpredictable. The most commonly prescribed classes are neuro-
leptics and antidepressants, while lithium carbonate, anticonvulsants,
anxiolytics and psychostimulants have also been used. But an individual
patient may benefit from each of several drugs, while an individual drug
may have effects on a range of target symptoms and behaviours. Also, there
can be a considerable variability of response to a drug regime within a
group of patients selected on the basis of PD.
Some of the many factors that account for such variability include an
inconsistent placebo response and the heterogeneity of a patient sample in
relation to PD features and to co-occurring psychiatric disorders, particu-
larly co-occurring depressive disorders.
Investigations of drug treatments for PD are complicated by method-
ological problems in the assessment of PDs and by the need to assess a range
of outcome measures. Compliance with treatment is often poor and the
effects of some drugs may take time to develop, for example several weeks in
relation to the effects of neuroleptics on some features of borderline PD.
Also, self-reports of change do not always correspond with clinicians'
ratings.
However, there is a general consensus that drug treatment can have an
important place as part of a range of measures to help some patients with
PD (Ellison & Adler, 1990; Stein, 1993).
233
234 Drugs and other physical treatments
Most of the empirical evidence in support of the use of drugs relates to
borderline PD, in particular to the effects of low doses of neuroleptics, but
even for these drugs the results of long-term treatment are unknown. It is
generally advised that drug treatment should not be the only intervention
for PDs, but be accompanied by psychological treatments, in particular
supportive psychotherapy. As some patients with PD show a repeated
pattern of self-harm, any drug treatment must take into account the risk of
overdose and of paradoxical reactions to drugs, for example, when
depressive features can worsen.
Other physical treatments can have a limited role in the management of
patients with PDs, and ECT and psychosurgery will be considered.
Neuroleptics
Neuroleptic (i.e. 'antipsychotic') medication has been reported to be
associated with a range of beneficial effects in patients with borderline PD,
in particular reduced impulsive anger and self-harm, and improved depres-
sive symptoms. Such effects are more evident when the features are severe,
and recommended doses are usually considerably less than those used for
the treatment of schizophrenia. Also, benefits have been claimed for
neuroleptics in relation to schizotypal PD, as well as schizoid and paranoid
PDs.
Montgomery & Montgomery (1982), in a non-controlled study, claimed
that flupenthixol injections reduced the rate of self-harm episodes in a
group with various PDs, and three subsequent placebo-controlled trials
found beneficial effects with different neuroleptics on various features of
patients selected on the basis of borderline PD. Goldberg et al. (1986)
compared thiothixene (at a mean daily dose of 8.67mg) with placebo in 50
outpatients with either borderline or schizotypal PD over 12 weeks.
Subjects were recruited by a newspaper advertisement. A beneficial effect,
which was positively related to severity of PD, was found for several
features including impulsive anger and cognitive abnormalities such as
ideas of reference. Soloff, George & Nathan (1986) compared the effects of
haloperidol or amitriptyline (at mean daily doses of 7.24 and 147mg
respectively) on 90 inpatients for 5 weeks. Patients were identified by a
diagnostic interview before being assigned into borderline, schizotypal and
mixed (borderline and schizotypal) PD groups on the basis of DSM-III
diagnoses. It was found that haloperidol was superior to amitriptyline or
placebo, and clear improvements were noted for several outcomes, includ-
Antidepressants 23 5
ing behavioural dyscontrol (i.e. involving impulsive anger/self-harm),
depressed mood, hostility, anxiety and features of schizotypal PD. In
another placebo-controlled study that used a longitudinal cross-over
design, Cowdry & Gardner (1988) investigated 16 outpatients with border-
line PD and prominent behavioural dyscontrol, who received the following
drugs in varying sequence: trifluoperazine, carbamazepine, tranylcypro-
mine, alprazojam (at mean daily doses of 7.8 mg, 820mg, 40mg and 4.7 mg
respectively), and placebo. The trial for each drug lasted 6 weeks. There
were positive results in relation to carbamazepine and tranylcypromine,
ajid the outcome for those who did not discontinue the trifluoperazine in
the first 3 weeks was 'fairly favourable' for anxiety and self-harm as rated by
the clinician, and for depression, anxiety and sensitivity to rejection as rated
by the patients. However, the most recent placebo-controlled trial of
haloperidol for borderline PD, (at a mean daily dose of 4mg for 5 weeks),
did not confirm the efficacy of this drug (Soloff et aL, 1993). This study,
which also included a group who received phenelzine at a mean daily dose
of 60 mg, involved 108 consecutively admitted inpatients, but, in contrast to
the report of Soloff et al. (1986, described above), patients were encouraged
to leave hospital after a minimum of 14 days. Also, all groups received
considerable non-specific support. Although phenelzine had a significant
beneficial effect on anger and hostility, these effects were not marked, and
the findings were largely negative. Soloff and colleagues considered that the
failure to confirm their previously reported beneficial effects of haloperidol
may have been due to such factors as a high dropout rate in the second study
and different sample characteristics, as the former study involved more
severe disorders. It was suggested that, in general, neuroleptics may be
clinically useful only when relatively severe disorders are treated.
Antidepressants
Most studies of the effects of antidepressants on patients with PD relate to
borderline PD, which commonly co-occurs with an additional diagnosis of
a depressive disorder. Also, some of the mood states that are an integral
part of the borderline PD syndrome consist of depressive symptoms, which
have been considered to differ in nature and origin from other types of
depressive disorders. Depression with borderline PD is, therefore, hetero-
geneous (SolofT, 1993), perhaps involving chronic unhappiness related to
experiences of boredom and emptiness, a labile depressed mood related to
life crises, or a co-occurring syndrome such as 'major depression'. The
latter is related to a variable range of causal factors including genetic
vulnerability to episodic changes of mood.
Monoamine oxidase inhibitors

Cowdry & Gardner's study (1988, described above), examined the effects of
four drugs, including tranylcypromine 40 mg daily, in a placebo-controlled
longitudinal crossover study of 16 outpatients with borderline PD and
prominent behavioural dyscontrol. Tranylcypromine was found to be
better than placebo in relation to a range of effects including decreased
depressive symptoms, reduced anxiety, increased capacity for pleasure, and
improved impulsivity and self-harm. But it was considered that behav-
ioural benefits were the result of an improved mood rather than reduced
impulsivity. Also, Liebowitz & Klein (1981), in a non-controlled trial of
phenelzine, found improved mood in patients with 'hysteroid dysphoria',
most of whom had DSM-III borderline PD.
Parsons et al. (1989) compared the effects of imipramine, phenelzine
(90mg/day) and placebo in over 300 patients with 'atypical depressive
disorder' and examined the associations of three tofivepositive criteria for
borderline PD. For those withfiveor more positive criteria, there appeared
to be an association with a good response of depressed mood to phenelzine,
compared with imipramine or placebo. But these findings related to
depressed mood were not confirmed by Soloff et al. (1993, described above),
who compared the effects of haloperidol, phenelzine (60mg/day) and
placebo on patients with borderline PD. Although phenelzine was found to
be significantly better than placebo on self-report measures of anger and
hostility, these effects were small. However, as previously noted, all groups
received considerable non-specific support.
Tricyclics
Amitriptyline (at a mean daily dose of 147mg) was compared with
haloperidol and placebo in 90 inpatients with borderline PD for 5 weeks by
Soloff et al. (1986, described above). As previously noted, haloperidol was
superior to amitriptyline on a number of outcome measures, including
depression, although a small improvement in depression was found for
amitriptyline compared with placebo. But a most important observation
was that some patients on amitriptyline became worse, although others
appeared to do well. The negative effects included increased paranoid ideas
Antidepressants 237
and behavioural dyscontrol, involving impulsive behaviour and threats of
self-harm.
Despite generally negative reports of tricyclics, various case reports
indicate that these drugs may have a role in the management of some
patients with PD. For example, Satel, Southwick & Denton (1988) des-
cribed a young man with borderline PD and co-occurring attention deficit
hyperactivity disorder (ADHD) in whom imipramine produced a marked
decrease in anxiety, explosiveness, restlessness and lability of mood.
Bellak (1985) also advocated the use of imipramine in relatively small
doses (e.g. 10-30mg daily) for 'attention deficit disorder psychosis', and
noted that improvements in agitation, overactivity, impulsivity and atten-
tion span may be found within a few hours of the first dose. Also,
Biederman (1988) reported that adolescents with ADHD may respond to
tricyclic antidepressants such as desipramine, as well as to various other
medications.
Selective serotonin reuptake inhibitors (SSRIs)

There are theoretical reasons why a drug with a relatively selective action on
serotonin metabolism may be expected to help to reduce impulsivity in
patients with borderline PD, as impulsivity appears to be associated with
reduced CNS serotonergic activity, shown by low levels of a serotonin
metabolite in the CSF and a reduced serotonin-mediated prolactin res-
ponse to fenfluramine (Soloff, 1993).
This prediction is supported by several non-controlled studies of fluoxe-
tine in patients with borderline PD. Norden (1989) prescribedfluoxetinefor
12 outpatients with borderline PD for up to 26 weeks in doses of up to *
40 mg/day. For most patients, improvements were reported for a number of
features such as depressed mood, lability of mood, rejection sensitivity,
anger and impulsivity, and such changes usually appeared in the first few
days. Markowitz et al. (1991) studied eight outpatients with borderline PD,
four with schizotypal PD and ten with both disorders, and gave fluoxetine
at a dose of 80 mg/day for 12 weeks; clear reductions in self-injury were
reported, regardless of diagnosis. Also, Cornelius et al. (1991) studied five
severely-affected inpatients with DSM-III borderline PD, who had not
responded to phenelzine or neuroleptics, who received 20 to 40 mg of
fluoxetine daily for 8 weeks. Their findings suggested that this drug
improved depressive and impulsive features. Finally, Hull, Clarkin &
Alexopoulos (1992) reported the case of a severely-affected young woman
with borderline PD who was in hospital for 58 weeks. Whenfluoxetinewas
prescribed at week 40 there was a rapid and widespread improvement
affecting her symptoms and behaviour.
These preliminary findings are encouraging, and further placebo-
controlled trials are needed. However, six reported cases of'intense suicidal
preoccupation' with depressed mood in patients receivingfluoxetinehave
been described by Teicher, Glod & Cole (1990). Two of these patients had a
diagnosis of borderline PD but, in addition, they had co-occurring
diagnoses of major depression or temporal lobe epilepsy with episodic
alcohol abuse. Also, one patient had an history of attempts at self-harm,
and the other had described previous suicidal ideas. In all the six patients
there was a tendency for intense, 'obsessive' suicidal ruminations to
continue for about 27 days after the drug was discontinued, and these cases
suggest the possibility of a rare 'paradoxical' response to fluoxetine.
Lithium carbonate
An effect of lithium on violent behaviour has been claimed in US prison
populations. Sheard (1971), in a non-controlled study, reported the effect of
the drug on 12 young male prisoners with a history of violence, and it
appeared that there was a consequent reduction of serious violent episodes,
although most of the improvement occurred in just three subjects. Tupin,
Smith & Clanon (1973) studied the effects of lithium on 27 male prisoners,
about half of whom had PDs associated with impulsive violence. Their
results were encouraging, as 56% showed a clear reduction in aggressive or
violent episodes, often associated with a 'reflective delay' when the subject
did not appear to respond so impulsively to adverse events. A further study
by Sheard et al. (1976) involved 66 subjects with PDs and a history of
violent crime who continued to show a pattern of violent behaviour in
prison. The placebo-controlled design involved two groups treated in
parallel for 3 months, and the lithium-treated group showed a significant
and clear reduction in assaults or serious threatening behaviour. Also, this
behaviour returned when the lithium was discontinued. It has also been
noted that lithium can be of benefit in some patients with mental retarda-
tion who show recurrent aggression and self-mutilation. However, there is a
possibility that a few patients may become more aggressive.
There is scant information about the value of lithium for subjects
identified on the basis of PD, but a placebo-controlled cross-over trial in
female adolescents with spontaneous mood swings and 'emotionally un-
stable character disorder', found a significant association between treat-
Anxiolytics 239
ment with lithium and decreased 'within-day mood fluctuations' (Rifkin,
Quitkin & Carrillo, 1972). Other studies have examined subjects selected on
the basis of alcoholism, and it was considered that some may benefit,
particularly when there is co-occurring episodic affective disorder. Also,
Stone (1990) found that some patients with borderline PD and co-occurring
episodes of affective disorder, in particular bipolar disorder, benefit from
lithium. Although it can be difficult to predict which patients with
borderline PD may respond to lithium unless there is a clear history of
discrete episodic mood disorder, a trial of this drug may be indicated if there
is a family history of episodic affective disorder or alcoholism.
Anticonvulsants
As previously described, Cowdrey & Gardner (1988) included carbamaze-
pine for 6 weeks (at a mean daily dose of 820 mg) as one of the three drug
regimes in their placebo-controlled longitudinal cross-over study involving
16 outpatients with borderline PD and prominent behavioural dyscontrol.
A significant reduction in the severity of behavioural dyscontrol was found
in the carbamazepine group, and this was suggested to involve a 'reflective
delay', as with lithium in the studies of prison populations. But one patient
became severely depressed and four others developed skin reactions.
It has been claimed that episodes of impulsive behavioural dyscontrol in
borderline PD, involving anger and self-harm, may sometimes follow 'an
abrupt onset' of intense dysphoria (Monroe, 1982), perhaps triggered by
alcohol, drugs or stress. The relationship of these phenomena to epileptic
disorders has been a matter of speculation, and some patients have reported
prodromal experiences such as anxiety and agitation. Also, after an episode
of self-harm or anger, some patients may feel that tension has been relieved.
But EEG changes related to temporal lobe epilepsy are not usually found in
borderline PD (Stein, 1993). However, in a case report of a patient with the
uncommon EEG patterns of rhythmic midtemporal discharges and 6/s
spike and wave complexes, carbamazepine produced clear improvement in
episodic violent behaviour that was accompanied by anxiety and intrusive
self-destructive thoughts (Stone et al.9 1986).
Anxiolytics
Although rapidly-acting anxiolytics might be expected to reduce a build-up
of tension that can precede an episode of behavioural dyscontrol in
borderline PD, anxiolytics are generally contra-indicated for this and other
PDs. This is because of a risk of addiction and of paradoxical reactions,
involving disinhibition with an increase in impulsive behaviour. However, a
small minority of patients with borderline PD may benefit.
The reported effects have been varied. Faltus (1984) described three male
patients with both borderline and schizotypal PDs and noted a favourable
response to alprazolam. However, two of these patients were rather
atypical as examples of PDs as they had hallucinations. Also, Reus &
Markrow (1984) studied the effect of alprazolam in 18 inpatients with
borderline PD in a cross-over design, and claimed a favourable response in
over half the subjects. In contrast, Cowdrey & Gardner (1988) used
alprazolam, at an average daily dose of 4.7 mg, in their placebo-controlled
cross-over study of 16 outpatients with borderline PD. Those receiving
alprazolam showed a clear increase in the severity of impulsive behaviour
involving self-harm and aggression, compared with placebo, although two
patients showed improvement. It was considered that this drug often has an
adverse disinhibiting effect.
However, there have been case reports in favour of clonazepam for
borderline PD, and preliminary results for buspirone in borderline PD
indicate that beneficial effects may occur (Soloff, 1993).
Psychostimulants
Various psychostimulants, including methylphenidate, pemoline, amphe-
tamine and levodopa have been reported to produce beneficial effects in
some patients with borderline PD, or with PD co-occurring with present or
past ADHD (attention deficit hyperactivity disorder).
For example, in a case report, methylphenidate benefitted a patient with
borderline PD and co-occurring ADHD, and this drug has also been
claimed to be of value for the impulsivity of adult minimal brain dysfunc-
tion (Wood et aL, 1976), for residual ADHD (Wender et al., 1985), and for
ADHD in children (Pelham et al., 1985). In addition, it has been noted that
a history of ADHD in borderline PD predicted a favourable response to
tranylcypromine, which has a stimulant action in addition to its effect on
monoamine oxidase (Cowdry & Gardner, 1988). Also, Stringer & Josef
(1983) described two inpatients with seriously disturbed behaviour in the
context of antisocial PD and repeated aggression, who improved consider-
ably while on methylphenidate, 20 mg twice daily.
Pemoline was reported to be better than placebo for improving concent-
ration and reducing impulsivity in adults with a history of severe ADHD
(Wender, Wood & Reimherr, 1984), and levodopa was claimed to benefit
Other drugs 241
patients with borderline PD and histories of drug abuse (Bonnet &

Redford, 1982). The use of dexamphetamine for borderline PD has also
been the subject of favourable case reports, although this drug is generally
contra-indicated because of its addictive and abuse potential. But a 30-
year-old man with repeated aggressive and violent behaviour apparently
responded dramatically to 20 mg dexamphetamine twice daily. It has been
claimed that childhood hyperactivity and an history of drug abuse.are
predictors of response to both psychostimulants and monoamine oxidase
inhibitors (Stein, 1993).
Other drugs
Propranolol has been reported to benefit brain-damaged patients who
show unprovoked rage attacks (Ratey, Morrill & Oxenkrug, 1983), as well
as some patients with behavioural dyscontrol (Mattes, 1988). Also, cloza-
pine may be useful for borderline PD associated with severe or prolonged
'psychotic' features (Frankenburg & Zanarini, 1993).
It has been suggested that abnormal functioning of the endogenous
opioid peptide system is associated with repeated self-harm and feelings of
boredom and emptiness in borderline PD, which may predispose some
individuals to self-medication by substance abuse (Soloff, 1993). Increased
levels of plasma jS-endorphin were found in five subjects with borderline PD
who were bored, unhappy and drug abusers (Bonnet & Redford, 1982), and
as dopamine inhibits the release of pituitary /?-endorphin, these authors
prescribed levodopa and carbidopa, with apparently beneficial effects.
Further evidence for the involvement of endogenous opioids in borderline
PD was provided by Coid, Allolio & Rees (1983), who found high levels of
plasma metenkephalin in ten borderline PD patients with a history of self-
mutilation. These observations and associated hypotheses have led to an
interest in the effects of opioid antagonists such as naltrexone, in particular
for self-mutilation and narcotic drug abuse. A case report with apparent
benefit of naltrexone was described by Soloff (1993).
Drugs that influence hormones related to sexual drive can be prescribed
for subjects who show sexual violence or distressing sexual preoccupations,
and PDs may often co-occur with these phonemena in groups who
repeatedly commit criminal offences with a sexual motive (Bowden, 1991).
Cyproterone acetate competes with testosterone at receptor sites, and also
inhibits hormonal release by the hypothalamus and pituitary, leading to a
reduced production of testosterone. However, side-effects are often proble-
matic and, while sexual drive is reduced, sexual aggression is not necessarily
affected. Other agents that produce reduced levels of plasma testosterone

include medroxyprogesterone acetate, which is given by regular intramus-
cular injections, and goserelin acetate, which is administered by subcuta-
neous injection. The practical and ethical problems of using these agents
have been discussed by Bowden (1991). Despite such difficulties, such
interventions do appear to enable some individuals to interrupt a pattern of
criminal activity, although these drugs should be used in the context of
counselling and practical support.
Drug treatment: borderline and related personality disorders

As has been described, borderline PD often co-occurs with other features of
PDs, and with other psychiatric disorders that may merit treatment in their
own right. Within the range of borderline PD features, targets for drug
treatment include affective instability, chronic dysphoria and behavioural
dyscontrol involving anger and self-harm. It is important to recognize that,
in general, the prescription of a drug should not be the only intervention
and that the patient's expectations should not be raised too high. While
there are anecdotal accounts of apparently dramatic responses to drugs, a
more modest effect is generally anticipated.
The best evidence in favour of drug treatment for borderline PD relates
to neuroleptics in low doses, monoamine oxidase inhibitors (MAOIs) and
carbamazepine. But MAOIs are dangerous in overdose and require dietary
compliance, while carbamazepine has been associated with a worsening of
depressed mood in some subjects. Selective serotonin-reuptake inhibitors
(SSRIs) such asfluoxetinerequire further studies of their use for PDs, but
early reports are promising, and these agents are relatively safe if taken in
overdose. The effects of tricyclic antidepressants and anxiolytics appear to
be much less predictable, as a worsening of symptoms and behaviour have
often been reported, although a minority of subjects may benefit. Other
drugs that have been associated with favourable case-reports include
lithium, psychostimulants and /?-blockers.
Some guidelines for prescribing for patients with borderline PD have
been suggested: if there is a possibility of a co-occurring episodic depressive,
manic or bipolar disorder, then lithium may be indicated, unless the risk of
overdose or dangerous noncompliance appears significant. If behavioural
dyscontrol is associated with EEG abnormalities, then carbamazepine
should be considered, while if chronic depressive features are prominent, an
MAOI may be prescribed if the patient is compliant and not at risk of self-
Target features 243
harm. When behavioural dyscontrol predominates, a low dose of neurolep-
tic (perhaps given by depot injection), or carbamazepine, may be of benefit.
It should be noted that the evidence in favour of these agents relates to
short-term use of up to several months, and the effects of longer periods of
drug treatments are unknown. A therapeutic trial of a sequence of several
drugs may be appropriate.
Drug treatment: schizotypal and related personality disorders

Low doses of neuroleptics have been reported to produce improvement in
features of schizotypal PD, such as ideas of reference, anxiety, derealization
and depersonalization (Goldberg et al, 1986; Ellison & Adler, 1990). Also,
patients with paranoid PD may respond to low doses of neuroleptics,
although compliance is usually poor and empirical evidence is lacking
(Tyrer & Seivewright, 1988a).
Drug treatment: other personality disorders

Several case reports suggest that some patients with avoidant PD may
respond to a MAOI orfluoxetine.Delito & Stam (1989) recommend a 2-3
month trial of these drugs in this context.
Drug treatment: target features

Ellison & Adler (1990) have considered drug treatment of PDs in relation to
various target symptoms, namely transient psychosis (related to borderline
PD, schizotypal PD, or co-occurring substance abuse), cognitive abnor-
malities (related to schizotypal, paranoid and borderline PDs), impulsivity
(related to borderline, narcissistic and antisocial PDs), mood instability
(related to borderline and histrionic PDs), dysphoria (with chronic low
energy and depressed mood related to borderline, antisocial and some other
PDs), and anxiety (related to many PDs). Such an approach can provide an
additional framework to guide drug treatment for patients with PD.
It appears that transient psychosis, and some cognitive abnormalities,
respond to relatively low doses of neuroleptics, which act more quickly than
when these drugs are used for the treatment of schizophrenia. Suggested
regimes have included trifluoperazine 2-4mg/day.
Impulsivity may respond to neuroleptics, M AOIs or carbamazepine, and
preliminary reports offluoxetine,which is relatively safe in overdose, are
encouraging in this respect. Also, trazodone, which has some serotonergic

reuptake inhibiting activity, has been reported to reduce aggression in
patients with structural brain disorders (Simpson & Foster, 1986). Lithium
may also improve impulsivity in some subjects.
Mood instability, with vulnerability to environmental stress including
relationship difficulties (i.e. 'rejection sensitivity'), may respond to a
MAOI, or to other agents, such as neuroleptics or fluoxetine. For those
patients with a co-occurring pattern of episodic mood disorder, lithium
may be indicated.
Chronic dysphoria may also respond to a MAOI, and a MAOI has been
reported to be beneficial for avoidant PD. In addition, Reich, Noyes &
Yates (1989) found that alprazolam helped features of avoidant PD. Ellison
& Adler (1990) suggested that if anxiety is present with PD, a beta-blocking
agent should be considered if somatic symptoms are pronounced, but while
alprazolam is a treatment option, it should be avoided if there is a perceived
risk of impulsivity or the development of addiction. The use of an MAOI
requires compliance and a low risk of overdose.
Drug combinations
As has been described, indications for the prescription of drugs are often
unclear in patients with PD, and it is not surprising that the role of drug
combinations is obscure. However, anecdotal case reports have involved
drug combinations, and it is likely that some patients could benefit. For
example, Ellison & Adler (1990) reported a patient with impulsive behav-
iour who appeared to respond tofluoxetineand lithium.
Other physical treatments
Electro-convulsive therapy (ECT)

ECT is mainly indicated for certain types of severe depression, treatment-
resistant mania, and certain features of schizophrenia, such as excitement,
stupor or drug-resistant distressing hallucinations. ECT may act in differ-
ent ways in relation to these syndromes; for instance, while it appears to
have a relatively specific effect on certain mood disorders, such as bipolar
manic-depressive disorder and its variants, it may also have a non-specific
and short-lived effect on other types of depression. In relation to this latter
effect, ECT can be considered to have a place in the management of any
type of severe depressive disorder, if the patient's life is at severe risk.
Other physical treatments 245
Zimmerman et al. (1986) reported 25 patients with 'major depression'
who received ECT; ten had various PDs, while the remaining 15 had
depression only. The short-term effect of ECT was considered to be good in
both groups but, at follow-up after 6 months,fiveof the patients with PD
had been readmitted in contrast to only one of the remaining 15. While
Black, Bell & Hulbert (1988) also found an apparently good response rate
to ECT in patients with major depression alone, and with depression and
PD, Pfohl, Stangl & Zimmerman (1984) reported that the absence of PD
was associated with a better response.
In conclusion, ECT does have a place in the management of depression in
patients with PD, in particular borderline PD. This treatment is particu-
larly indicated for those patients who have PD and a co-occurring affective
disorder of the type that shows a relatively specific response to ECT. In such
patients, a pattern of episodes of severe depression may be superimposed
upon a more chronic dysphoria. But for any patient with borderline PD
who develops a severe episode involving repeated impulsive self-harm, ECT
may produce a rapid response. Such an effect is usually short-term, but,
nevertheless, it may interrupt a vicious circle of escalating risk.
Psychosurgery
In a recent review, Dolan & Coid (1993) concluded that there is 'no clear
justification' for the use of psychosurgical procedure for PD. However,
there are many uncontrolled reports of favourable outcomes, for example,
in relation to impulsive aggression, following a variety of procedures.
Andy (1975) described six patients with antisocial behaviour, four of
whom showed marked improvements after thalamotomy, and Laitinen
(1988) concluded that posteromedial hypothalamotomy could be indicated
for 'restless, aggressive and destructive behaviour'. Dieckmann, Schneider-
Jonietz & Schneider (1988) reported a follow-up of 14 subjects with
'aggressive sexual delinquency' who received unilateral ventromedial
hypothalamotomy. In general, improvements in several aspects of func-
tioning were claimed, van Manen & van Veelen (1988) described 54 patients
who had received psychosurgery for various psychiatric disorders including
self-mutilation, aggressive behaviour with temporal lobe epilepsy, and
aggression in the setting of mental retardation. Operative procedures
involved lesions in the fronto-basal region, cingulum, para-cingular white
matter, anterior corpus callosum, amygdala and thalamus. In Japan, Sano
& Mayangi (1988) reported 'good results' from stereotactic posteromedial
hypothalamotomy for 'violent, aggressive behaviour'.
Physical treatments: psychological aspects

A decision to prescribe medication for a patient with PD may produce a
range of reactions apart from compliance with realistic expectations. Some
patients may perceive this as the clinician 'giving up', or, in the case of an
individual with narcissistic PD features, as a severe blow to self-esteem
(Ellison & Adler, 1990). Others perceive this as a welcome message that they
can be dependent patients and wait for their problems to be resolved.
Several studies have noted that there is often a powerful placebo effect of
drugs in patients with PDs, and the patient's psychological reactions can
depend to some extent on the characteristics of the underlying PD.
Borderline PD may be associated with the drug having a 'soothing' effect
due to its association with the clinician, patients with paranoid PD may
become defensive and suspicious as they feel some control has been lost,
and histrionic PD may be associated with exaggerated complaints of side-
effects. Patients who are prone to manipulative behaviour may use their
medication in such interactions, for example by threatening an overdose,
while others may develop marked psychological dependence, or physical
addiction due to non-compliance.
One model of service delivery separates the psychological and pharmaco-
logical management of patients (particularly those with borderline PD), so
that two clinicians are involved. But although such an arrangement may be
necessary if the psychological management is provided by a non-medical
clinician, the two roles can be successfully combined.
It is important to stress to the patient that a course of medication is being
advised on a trial basis. The patient should be told that individuals respond
differently to a particular drug and that it is uncertain whether benefits will
be obtained. It may also be appropriate at the outset to outline the
possibility of future trials with other drugs. Patients should usually be
encouraged to take an active part in the evaluation of the effects of a drug
regime by noting improvements and side-effects, and, if appropriate,
reducing the dose accordingly. It is most important to assess the risk of
overdose, in particular in the context of impulsive self-harm, and if a
significant risk is identified the prescription of drugs which are relatively
toxic must be avoided.
Physical treatments: a summary

Drug treatments have a role in the management of PDs, in particular
borderline and schizotypal PDs. Most of the literature has been concerned
with borderline PD, and it must be noted that patients with this diagnosis
have a high rate of co-occurrence of other psychiatric disorders. Also,
compliance is often poor and there is a danger of overdose. As favourable
results of drug treatment are generally modest, the patient's expectations
must not be raised too high, and drug treatment should take place in the
context of other interventions, in particular supportive psychological
management.
In relation to borderline PD, low doses of neuroleptics can be of benefit
for a range of features, but Soloff (1987) has recommended that treatment
duration should be generally brief, i.e. up to 12 weeks. Although clinical
experience suggests that longer periods of 'continuation' treatment may
help in some cases, a placebo-controlled trial of haloperidol or phenelzine
for 16 weeks in patients with borderline PD did not provide clear evidence
in favour of continuation treatment, even though modest improvements
were reported (Cornelius et al.91993). Some antidepressants and carbama-
zepine can also benefit some patients, as can benzodiazepines, lithium and
psychostimulants, but the effects of several of these drugs are unpredictable
and may worsen symptoms and behaviour. In particular, behavioural
dyscontrol may be worsened by anxiolytics or tricyclics, while carbamaza-
pine may be associated with an increasing depression.
It may be appropriate to arrange a therapeutic trial of a sequence of drug
treatments for an individual patient and, as anecdotal evidence suggests
that drug combinations may be of benefit, combinations can be evaluated
in certain patients with severe chronic disorders.
ECT can be indicated for management of some patients with severe life-
threatening depression in the context of PD, as outlined above.
Finally, it is important to recognize that some patients with severe PD
have co-occurring affective disorders of the type that responds well to
medication or ECT. But episodic mood changes due to such co-occurring
disorders are often obscured by a chronic pattern of maladaptive behav-
iour. If such co-occurence is suspected, the patient should receive a trial of
appropriate medication such as lithium or an antidepressant.
Future directions
Future research may enable a range of psychotropic drugs and drug
combinations to be targeted more specifically in patients with PDs. The
evidence of cerebral serotonin abnormalities associated with impulsivity is
partly based on abnormal neuroendocrine responses to drug 'challenges'
that affect serotonergic function, such as the administration of m-chloro-
phenylpiperazine and fenfluramine (Coccaro et al., 1989), and future

studies are needed to investigate the associations between such challenges
and subsequent treatment response. This strategy may also prove useful in
relation to short-term changes in mood as a result of an 'amphetamine
challenge', in which 30mg dexamphetamine is given daily, for example for
2-3 days, and ratings of mood are compared with those associated with
placebo (Fawcett & Siomopouloo, 1971). A significant improvement of
mood in relation to amphetamines has been associated with a history of
childhood attention deficit hyperactivity disorder and may predict a
favourable response to other stimulants or antidepressants, in particular
MAOIs (Stein, 1993).
7
Psychological management
J.H. DOWSON
Definitions and overview

'Psychological management' is a broad term encompassing a range of
theoretical frameworks and techniques, such as psychoanalysis, various
forms of psychotherapy, behaviour therapy, supportive psychotherapy and
counselling.
Some methods focus on an evolving interpersonal interaction between
therapist and patient (i.e. a 'psychodynamic' psychotherapy approach),
while others are mainly concerned with a patient's thoughts (i.e. a 'cognitive
therapy' approach) or behaviour (i.e. a 'behaviour therapy' approach).
Psychological management can be defined as the alleviation of problems
by behavioural interventions designed to influence the patient's higher
mental activity. Such behavioural interventions include what the therapist
(or other relevant individuals, e.g. in group psychotherapy) says and does
(or does not say or do), together with any modifications that the therapist
may encourage in the patient's social environment, such as admitting a
patient to hospital or a day centre. Higher mental activity can be considered
as those mental processes that are experienced in conscious awareness, or
have this potential, and includes thoughts, memories and emotions. Higher
mental activity also involves so-called 'unconscious' mental processes, an
example of which would be an instruction given in an 'hypnotic state' that is
subsequently carried out without the individual being consciously aware of
the reasons for his/her actions.
The term 'psychotherapy' has usually implied an emphasis on interper-
sonal interactions between therapist and patient, and the American Psychi-
atric Association's Committee on the Practice of Psychotherapy has
defined psychotherapy as
a method of treating mental disorders, the cornerstone of which is the development
of experiential understanding of the meaning of symptoms and/or ... those
249
250 Psychological management
characteristic patterns that cause the patient to suffer limitations and pain. In
contrast to somatic or pharmacological intervention, psychotherapy relies on
verbal means and techniques to modify characteristic ways of thinking, feeling, and
behaving that interfere with the patient's capacity for maximal functioning.
In similar vein, the British Association of Psychotherapists (1993) note

that the 'method employed is verbal communication' for
those who are experiencing difficulties within themselves or in their relationships
with others... Psychotherapy is a process of exploration undertaken by two people,
therapist and patient, in the context of a personal relationship between the two. In
this exploration, the patient takes the lead by relating thoughts, associations,
feelings, memories and dreams. The therapist follows, clarifying what is happening
by drawing the patient's attention to habitual modes of behaving and by making
links between conscious and unconscious, present and past experience. All this has
the aim of promoting the patient's self-understanding and capacity to view the
world objectively and less coloured by personal wishes, fear and prejudices.
Such descriptions refer to a process that is generally time-consuming (i.e.

1-3 hours per week for up to several years), that is not readily available to
patients in psychiatric services, and is relatively inaccessible to evaluation
because of the degree of variability of this type of intervention.
PDs are characterized by long-term disorders of social functioning
together with associated features, and it has been claimed that 'the
personality disorders constitute one of the most important sources of long-
term impairment in both treated and untreated populations' (Merikangas
& Weissman, 1986). However, the provision of psychological management
for those with PDs must recognize that apparent short-term improvement
in symptoms or behaviour can be misleading and may reflect an environ-
mental change, as, by definition, these disorders consist of deeply ingrained
patterns of behaviour. Nevertheless, psychological management can often
help to relieve a patient's symptoms (e.g. a depressive disorder) that are
secondary to problems engendered by PD; this may then lead to a
temporary respite in a pattern of maladaptive behaviour such as substance
abuse or self-harm episodes. But although overt patterns of PD-related
behaviour may show apparently rapid changes, the patterns of cognitive
processes associated with PD, if they do respond to psychological manage-
ment, do so at a relatively slow rate. If a clear change takes place this is
usually over a time-scale of many months or years, and realistic goals of
treatment of PDs usually involve only modest cognitive changes. However,
this need not discourage the therapist, as even a modest change (e.g.
teaching a patient to act in a different way when he loses his temper) can
have major beneficial effects on social functioning.
Shea (1993) has reviewed the claims for several approaches to psycho-
logical management of PDs, including various types of psychodynamic
psychotherapy, in particular for patients with borderline PD; one approach
involved a focus on the use of interpretations by the therapist, another
advocated more supportive interventions, while some treatment pro-
grammes have been time-limited, in contrast to most 'psychodynamic'
psychotherapy. Another approach for those with borderline PD has been
to focus on interpersonal interactions in a series of 30 sessions of group
psychotherapy, while 'cognitive' interventions have provided very different
strategies for the treatment of PDs, as they have been directed towards the
maladaptive patterns of thoughts and feelings (also known as 'schemas' or
'cognitive structures'), which are claimed to produce biased judgements
and interpretations in selected situations. Cognitive approaches usually
incorporate a behavioural dimension, as the process of developing new
schemas, and the modifications of old schemas, includes planning changes
in the patient's behaviour. Cognitive therapy has been applied to the
treatment of some depressive disorders, with attention to the thoughts and
behaviours related to a depressed mood, and for PDs a wide range of
schemas need to be addressed. Although 'behavioural' treatments theoreti-
cally involve a specific focus, most specified psychological interventions are
invariably supplemented by other potentially therapeutic aspects, such as
the nonspecific ingredients of a therapist-patient relationship, involving
empathy, advice, and personal warmth. The 'behavioural' approach of
Linehan et al. (1991) for patients with borderline PD is of particular
interest, as its efficacy has been supported by experimental data; this
treatment programme involved training the patient in interpersonal skills,
in emotional regulation and expression, and in distress tolerance. Other
behavioural approaches such as desensitization with gradual exposure to
feared situations have been used to treat patients with avoidant PD.
For patients with borderline PD, uncontrolled studies of various forms
of psychological management, in particular psychodynamic psychother-
apy, have indicated a tendency for a high drop-out rate in the first months
of outpatient treatment sessions. However, ten controlled studies of a
defined psychological management for patients with various PDs have
evaluated a range of behavioural treatments and other forms of psycho-
therapy (Shea, 1993). For avoidant PD, three studies reported a significant
beneficial effect of behavioural treatments, which included graduated
exposure, social skills training and systematic desensitization (Marzillier,
Lambert & Kellett, 1976; Cappe & Alden, 1986; Alden, 1989). Also for
avoidant PD, Argyle, Bryant & Trower (1974) found beneficial effects of
either social skills training or psychodynamic psychotherapy, while im-

provements following social skills training have been found in other studies
(Trower et al.9 1978; Stravynski, Marks & Yule, 1982; Cappe & Alden,
1986). But, while these reports indicated improvement on such measures as
frequency and range of social contacts, avoidance, and pleasure in social
activities, such benefits were often modest, and the effects of social skills
training were generally limited to superficial relationships. Control groups
were often waiting-list patients or patients undergoing a different type of
treatment, and none of these studies controlled for non-specific factors of a
treatment situation, such as the placebo effect and the therapist's interest
and concern.
Linehan et al. (1991) have provided impressive empirical evidence in
favour of a form of psychological management in relation to patients with
borderline PD who had a history of repeated self-harm episodes. A regime
of group and individual psychological management sessions for one year
(Dialectical Behaviour Therapy, DBT) was compared with the effects of
'treatment as usual'. The clear benefits of DBT (which will be described
subsequently) included a reduction in frequency and severity of self-harm
behaviour, so that these results have important implications for the
appropriate provision of psychiatric services for a group of patients who
present considerable management problems. Another study of borderline
PD investigated the effects of intensive, brief psychological management (4
hours/day for 8 days, using either behavioural methods or a psychodyna-
mic psychotherapeutic approach) and found that both treatments had
some effect, while the group treated with behaviour therapy had less
frequent suicidal thoughts on follow-up (Liberman & Eckman, 1981).
Winston and colleagues (1991) investigated the effects of two forms of
brief psychodynamic psychotherapy on patients with various PDs, but
excluded those with schizoid, schizotypal, narcissistic or borderline PDs.
Patients were involved with weekly sessions for 40 weeks, and patients on a
waiting-list formed the control group. Both forms of psychotherapy were
associated with significant improvements.
Investigating the effects of treatments for patients with PDs presents two
main methodological problems. Firstly, a patient sample based on the
diagnosis of a named PD will show considerable heterogeneity, as another
psychiatric disorder (such as a depressive disorder or anxiety state) may be
present, more than one PD may co-occur, PDs defined by polythetic criteria
are heterogeneous, and a specified PD exhibits a range of severity.
Secondly, a co-occurring psychiatric disorder can not only influence the
assessment of PD but can also affect the evaluation of change in response to
treatment. Other problems include a high drop-out rate from a treatment
programme in relation to some PDs, variability in the skills of therapists for
psychological management techniques, the effects of various aspects of the
patient's life situation (or other uncontrolled variables) on PD behaviour,
and the difficulty in obtaining suitable control or comparison groups. As it
is unethical to form a 'no-treatment' control group (such as a waiting-list
group) over a period in excess of that which would normally be the waiting
time, alternative research strategies are needed, such as comparing the
effects of a standard treatment on groups of patients with different PDs, or
comparing the results of a standard treatment 'package' with those of a
similar regime that omits certain specific interventions. However, the
findings using the latter approach can be influenced by interactions between
the various ingredients of a treatment programme.
The evaluation of treatments for PD may be assisted by the identification
of 'core' PD features that could provide the focus for outcome measures.
This would be a change from the approach derived from current polythetic
criteria sets in DSM-III-R and ICD-10, in which equal weight is given to all
the criteria listed. The selection of patients partly on the basis of core
features might also increase the homogeneity of the sample. Outcome of
treatment programmes for patients with PD can sometimes be expressed in
monetary terms in a cost-benefit analysis, perhaps by taking into account
the cost of medical resources and the costs to society of antisocial
behaviour, but, in the organization of clinical services, such analyses have
to involve additional clinical and social judgements. In clinical practice, the
evaluation of monetary benefits have to be considered together with effects
that are measured in different units, such as symptom scores and levels of
social functioning (Krupnick & Pincus, 1992).
There have been many reports that a patient's personality and PD can
influence the outcome of various forms of psychological management in
psychiatric practice (Conte et al., 1991). Not surprisingly, poor outcome
has been associated with mistrust, hostility and problems with intimacy,
while good outcome has been related to a good 'therapeutic alliance', non-
aggressive problems, assertiveness and sociability. An approach that has
identified the importance of PD for the outcome related to various
psychiatric treatments, including psychological management, has been to
study the results of treatment of non-PD psychiatric disorders in patients
with and without co-occurring PDs. Tyrer et al. (1990) reported 210
patients with anxiety or depressive disorders, of which 30% also had a PD.
The latter subgroup had more severe symptoms, and those with a depres-
sive disorder and PD responded less well to treatment than those without
PD.
An overview of psychological management for patients with PDs must
consider some implications of the use of the PD diagnostic categories, in
particular the attitudes of professionals in psychiatric services to patients
whose symptoms and behaviour have attracted these diagnostic labels.
Clinicians' attitudes often relate to their concepts of a patient's personal
reponsibility for his/her actions and whether he/she is 'in control'. In this
context, patients with PDs are often considered as distinct from those with
other forms of psychiatric disorder, in that they have 'control' and
'responsibility' for their actions. But such an analysis is simplistic, inaccur-
ate and inappropriate, as shown by the evidence that genetic and other
biological factors are associated with personality and PD. Although the
behaviour of many patients with PD can be aggravating and unpleasant,
clinicians must not allow themselves to reject such individuals
inappropriately.
Psychodynamic psychotherapy
Overview
Psychodynamic techniques focus on the interpersonal interactions between
the therapist and patient and, in relation to PDs, have been mainly
investigated for those patients with borderline PD.
Certain hypotheses related to the development of borderline PD have
been influential in providing a rationale of treatment, in particular that a
person has developed borderline PD as a result of bad parenting and other
traumatic experiences such as physical and sexual abuse (Masterson, 1981).
But although many patients with borderline PD (and other PDs) do give
such histories, there is no good evidence that such experiences are essential.
However, despite the uncertainties surrounding the hypothesis that border-
line (and other) PDs are mainly the result of faults in the previous behaviour
of 'significant' others, which would always give the patient victim status, a
major role of psychodynamic psychotherapy has been conceived of as
allowing the patient to re-experience past adverse relationships and to build
new ways of thinking, feeling and reacting, in the context of an extended
relationship with the therapist. Sessions may be regularly provided for up to
several years and the patient's feelings towards the therapist (the 'transfer-
ence') are usually a major focus of discussion. Although this format is
Psycho dynamic psychotherapy 255
subject to certain techniques and theories, it is relatively unstructured,
compared with cognitive and behavioural regimes.
As there are considerable methodological problems in evaluating the
processes of psychological management, and of changes in PD, a lack of
evidence in favour of a treatment regime for PD does not exclude the
possibility that there are some benefits. But reviews of published studies
have not concluded that psychodynamic psychotherapy is superior to
placebo. Andrews (1991) considered that 'dynamic psychotherapy ... has
not been demonstrated to be superior to placebo in the neuroses or
personality disorders. The case reports of improvement during long-term
dynamic therapy for personality disorders may be due to the combined
effects of normal maturation and the non-specific effects of continued
clinical care'. In a review of 19 studies, it appeared that short-term
psychodynamic psychotherapy was associated with beneficial effects at the
end of treatment but not at follow-up after 6 or 12 months (Svartberg &
Styles, 1991). But even these improvements were not as beneficial as those
associated with other forms of psychological management, mainly involv-
ing cognitive-behavioural regimes.
Such studies have generally evaluated relatively short periods of treat-
ment, and it can be argued that they are not relevant to the considerably
longer periods of psychodynamic psychotherapy which are often under-
taken. But it has been pointed out that at least 50% of patients with
borderline PD will no longer meet criteria for this diagnosis after 10 years,
and that specific effects of treatment must be distinguished from natural
remission and placebo responses. These latter variables could account for
the modest improvements that were found in patients with borderline PD
who received 100 hours of psychodynamic psychotherapy over 2 years
(Stevenson & Meares, 1992).
Andrews (1993) has considered that apart from lack of evidence for
efficacy, additional arguments against the provision of long-term psycho-
dynamic psychotherapy (other than in private practice) are the high cost (in
an Australian survey each patient received an average of 330 hours of
treatment) and possible side-effects such as inhibiting normal remission,
encouraging undue dependency, encouraging emotional arousal which
cannot be satisfactorily dealt with by the patient, and sexual abuse of the
patient by the therapist.
'Psychodynamic psychotherapy' includes a range of techniques, and one
important variable is duration of treatment, which can vary from up to
several sessions a week for several years to once a week for a few months.
Another variable is the degree of the therapist's 'neutrality', which is
defined as how much he/she shows his/her thoughts, feelings and opinions,
and how much active encouragement and support is given. There is also
variation in the focus of strategy; some therapists mainly try to uncover and
explain hidden aspects of mental functioning with an emphasis on past
experiences, while others concentrate more on the 'here-and-now' pattern
of feelings and interpersonal reactions.
Intensive psychoanalytic psychotherapy

This strategy of psychological management has been advocated by a
minority of psychiatrists for patients with features of borderline and/or
narcissistic PD. At the most intensive end of the spectrum of such
intervention is 'psychoanalysis', which involves sessions of up to 1 hour,
four or five times a week, with the patient lying on a couch (i.e. not in face-
to-face contact with the therapist) in a setting that encourages the patient to
experience strong emotions (Higgitt & Fonagy, 1992). However, this
format is considered by many psychiatrists to be generally contra-indicated
for individuals with borderline PD, as the emotional experience in relation
to the therapist can often become excessively hostile and destructive, with a
degree of anger that can lead to aggressive or self-destructive behaviour
that is the opposite of therapeutic (Gunderson & Sabo, 1993a).
At a modest distance from this end of the spectrum, 'intensive psycho-
analytic psychotherapy' involves two or three hourly sessions per week,
involving face-to-face contact between therapist and patient, and can
continue for between 2 to 7 years. Such a regime is believed by some
therapists (but probably not by most psychiatrists) to be the treatment of
choice for patients with severe borderline PD. Within this format, the focus
is on the interpretation of the transference, which can be defined as aspects
of the patient's emotions, attitudes, motivations and behaviour that are
directed to the therapist and are derived in part from past experience.
Transference phenomena may be present at the beginning of treatment as
well as developing subsequently. The process of interpretation involves
providing the patient with links or associations between aspects of the
transference and hidden mental processes that relate to past experiences.
The psychological means by which the patient keeps such mental processes
hidden are known as 'defence mechanisms'; examples are 'denial' and
'avoidance', which will be described subsequently.
Transference reactions in patients with borderline PD can involve
sudden intense emotion, which often leads to anger, and Kohut (1977) has
described three common patterns, mainly in relation to features of co-
occurring narcissistic PD. In a 'merger' transference, a grandiose patient
may see the therapist as part of his/her omnipotent influence, and expect the
therapist to be owned by the patient and exclusively committed to his/her
treatment. This is similar to a 'mirror' transference, in which the therapist is
expected to focus on and admire the patient's special qualities, while an
'idealized' transference involves the therapist being perceived as the pro-
vider of unlimited safety and support. Another pattern that is often seen
with borderline PD patients is a tendency to see the therapist as 'all good' or
'all bad', and to change from one perception to the other (perhaps
alternating) in the course of treatment.
For borderline PD patients, some therapists have believed that the main
ingredient of treatment is for the therapist to identify and point out
(interpret) negative feelings in the transference (i.e. anger and mistrust), as
well as the patient's hidden motivations of sadism, destructiveness, aggres-
sion and the wish to control others (Higgitt & Fonagy, 1992; Gunderson &
Sabo, 1993a). However, there have been disagreements whether all these
'hidden' aspects are invariably present, about the timing of such interpre-
tations, and about their relative importance.
In routine psychiatric practice, several strategic guidelines have emerged
for psychological management, some of which contrast with the approach
of intensive psychoanalytic psychotherapy. Many psychiatrists believe
that, compared with some other patients, those with borderline PD require
a relatively active therapist, as excessive silence is not well tolerated. Also, in
general, a limit-setting approach with few interpretations appears to be
needed, in which the therapist repeatedly draws the patient's attention to
the adverse consequences of his/her behaviour. The therapist has to be able
to tolerate considerable anger and child-like or regressive behaviour,
involving tantrums and demands for instant gratification. Common mis-
takes include not challenging the patient's view of the therapist as 'heroic'
or 'good' when the patient is in a phase of idealization, or being excessively
rejecting in response to provocation.
Kernberg's expressive psychotherapy

Kernberg's treatment model is a type of intensive psychoanalytic psycho-
therapy and was developed from ideas based on the Psychotherapy
Research Project of the Menninger Foundation (Kernberg et ai, 1972). It
was considered that this study indicated that patients with borderline PD
did best with intensive psychotherapy aimed at hypothesized core psycho-

pathology, rather than with a supportive approach aimed at improving
social adaptation.
Kernberg advocated the importance of transference interpretations, in
particular in relation to the patient's mistrust, anger and aggression, which
were considered to form the core of borderline PD psychopathology.
Indeed, he considered that a supportive approach, which avoids or damps
down anger by positive and caring responses, may cause an increase in
hidden anger as it cannot readily be expressed.
The format involves sessions at least twice weekly, in which the patient
faces the therapist and there is a fixed timetable of appointments. This
precludes responding to phone calls or approaches at other times. Limits
are set in relation to destructive behaviour such as missing appointments,
physical aggression, substance abuse and self-harm. A written contract at
the start of treatment can be signed, and the therapist's options are to stop
or postpone treatment, or to refer to another clinician if the contract is
broken. Sometimes any attempt at this form of treatment may produce an
increase in destructive behaviour and if a decision is made to proceed, this
may have to be within the setting of a long-term inpatient admission. At
times, regular contact between the therapist and significant persons in the
patient's life may be indicated.
The therapist should be neutral in that he/she should not be directive or
display much of his/her own feelings and attitudes. But this is not a 'blank
screen' approach, as a caring attitude, personal warmth, showing respect
and being active in treatment sessions is necessary. Interventions include
clarification, confrontation and interpretations, and are aimed mainly at
current situations and relationships. Clarification includes inquiring, in a
non-challenging way, about incomplete, vague or contradictory aspects of
the patient's statements, while a confronting therapist may tell the patient
how such statements appear to others. Interpretations involve suggestions
about possible underlying (hidden) motives and feelings associated with the
patient's behaviour; at first these can be limited to 'here and now'
situations, but later may be linked with past experiences, particularly when
interpretations about the transference are involved. For instance, a
comment may be made about the patient's negative reactions to the
therapist that appear disproportionate to the reality of the situation, or
about inappropriately positive feelings. But appropriate positive feelings
can be accepted without interpretation, and the patient must not be made to
feel that their every feeling and reaction is invalid and unreal. Often it may
be useful to point out a similarity between the transference relationship and
other current relationships with 'significant others'. If the patient distorts
the nature and meaning of the therapist's interventions, this may form the
basis for further interpretation, as can contradictory or maladaptive views
of self and others that can be expressed at different times. For example, a
patient may variously consider him/herself as a child, a victim, as all-
powerful, or as a person with a need to control others (Goldstein, 1990).
Patients often have difficulty in integrating different views of themselves
and others, and need to learn that an individual has different facets and
modes of behaviour, so that 'all-or-none' concepts are inaccurate.
While countertransference can be viewed as 'the therapist's emotional
reactions to the patient that are based on the therapist's own unresolved
unconscious conflicts' (Goldstein, 1990), Kernberg had a broader concept
that incorporated both conscious and unconscious reactions to the patient,
including those based on reality as a result of the patient's behaviour; the
latter component has been called the 'objective countertransference' (Win-
nicott, 1949). It is important that the therapist monitors his/her reactions
and uses this information objectively to try to understand the patient. If
objectivity is lost, countertransference can be antitherapeutic in many
ways; limits may fail to be set because of the therapist's inability to tolerate
anxiety, or the therapist's anger may lead to a harmful rejection or an
inappropriate encouragement for the patient to experience a destructive
anger. This could increase the risk of suicide.
Masterson9s confrontative psychotherapy

Masterson (1976) has described another variant of psychodynamic psy-
chotherapy for borderline PD patients (Goldstein, 1990). This involves the
therapist being ready to express a wider range of personal views and
attitudes than in Kernberg's approach, with a focus on the depressed mood
such patients have when they feel alone and abandoned. It was considered
particularly important to help the patient to cope with the ending of
treatment, so that any efforts to develop autonomy are encouraged and
supported. Various countertransference problems were identified which
apply to psychological management in a range of settings. These include
inappropriate anger if a patient shows destructive behaviour (this may
precipitate further self-harm), an inability to set limits that are necessary for
treatment to continue, an excessive need to have the patient's approval, and
an excessive need to be in control and directive, thus preventing auton-
omous behaviour.
Conclusion
The role of psychodynamic psychotherapy for PD, in particular for
borderline PD, is controversial. Gunderson & Sabo (1993<z) have noted that
there is increasing evidence (Wallerstein, 1986; Kolb & Gunderson, 1990)
that less intensive and more supportive strategies appear to be associated
with changes that have been believed by some to result only from more
intensive transference-based regimes; 'Such observations cast doubt on
accounts of psychoanalytic psychotherapy that propose that change relies
on insight without recognizing the corrective role of such unspecific
"supportive" interventions'. Although it is possible that, for certain
patients, their unique experiences of intensive psychodynamic psychother-
apy are of benefit, it seems that there is insufficient evidence to recommend
such expensive and time-consuming procedures.
However, many of the strategies derived from these approaches may be
of benefit in relatively brief, more supportive forms of psychological
management, which will now be described.
Brief (psychodynamic) psychotherapy

There have been several attempts to develop brief programmes of psycho-
logical management for use in routine psychiatric settings that incorporate
some theories and strategies derived from psychodynamic psychotherapy
(Weissman, Klerman & Prusoff, 1981; Luborsky, 1984; Horowitz, Marmar
& Weiss, 1986). Such interventions have been provided for patients with a
range of psychiatric disorders, including depressive disorders and PDs.
Brief (short-term or time-limited) psychotherapy involving 10-20 ses-
sions over 6-12 months was at one time considered inappropriate for those
with borderline PD (Mann, 1973), but some clinicians consider that such
regimes can be useful if sessions are focussed on specific situations or
interpersonal problems, and the patient is not encouraged to develop
strong transference feelings. This is achieved by avoiding explanations of
transference phenomena, and providing direction and structure to the
content of the sessions.
Gunderson & Sabo (1993a) have considered that this strategy may be
appropriate for those patients with borderline PD who would drop out of
more intensive treatments and who would find it very difficult to cope with
becoming dependent on a therapist. Also, brief psychotherapy can be given
intermittently on a long-term basis (Silver, 1985).
Winston and colleagues (1991) have examined the efficacy of two forms
Supportive and relatedpsychotherapies 261
of brief psychotherapy for patients with PDs, but excluded borderline,

narcissistic, paranoid, and schizoid/schizotypal psychopathology. Patients
were seen once a week for a maximum of 40 sessions. One treatment format
was called 'short-term dynamic psychotherapy' (STDP) and the other 'brief
adaptational psychotherapy' (BAP). Both therapies involved clarification,
confrontation and interpretation, and attempts were made to link present
patterns of thinking, feeling and behaving to past events and relationships.
In general, STDP was the more active method involving the confrontation
of psychological defences and the encouragement of emotional arousal,
while BAP was a more cognitive therapy which focussed on the origins and
determinants of maladaptive patterns of distorted thoughts and on the
patient-therapist relationship, with less pressure on the patient to remove
psychological defences. The findings suggested that STDP, with the
emphasis on confrontation and eliciting emotions, can be antitherapeutic if
inappropriately used, although if the patient has the resiliance to tolerate
the stress involved without dropping-out, or engaging in destructive
behaviour, higher levels of improvement may follow. Also, STDP seems
more effective than BAP in improving mood in some types of co-occurring
depressive disorder. This is consistent with the hypothesis that, for
depression which is causally associated with certain experiences, the
expression (release) of any anger can be therapeutic. But there is a great
danger of assuming that all depressed patients can be improved by the
therapist eliciting (or even provoking) anger; such a strategy can have
disastrous consequences.
The results of this study in relation to those PDs involved (which
excluded borderline PD), indicated that these treatment strategies were
associated with significant improvement in target complaints and social
adjustment, compared with a waiting-list control group. A subsequent
report from Winston et al. (1994), with a larger patient group, which also
excluded 'cluster A' and borderline PDs, concluded that BAP and STDP
were effective for patients with certain types of PD, but that the two
approaches did not differ in overall outcome.
Supportive and related psychotherapies
Overview
'Supportive psychotherapy' encompasses many theoretical perspectives
and techniques, including some derived from psychodynamic psychother-
apy. But in contrast to the latter, the main aim is to restore, maintain or
262 Psychological managemen t
improve adaptive functioning, at the expense of self-destructive behaviour.

The focus is on the current reality of the problems of daily life, and
psychological defences are generally strengthened unless they are seriously
maladaptive. Exploration of unconscious processes by interpretation is
generally avoided. The patient is not encouraged to develop the type of
relationship with the therapist that involves considerable emotional expres-
sion, excessive dependence or the experience of strong anger. But a patient
can be encouraged to realize that it is possible to deal with anger in a way
that is not always destructive.
The techniques used in supportive psychotherapy include allowing
patients to express their thoughts and emotions, appropriate reassurance,
rational discussion, education, advice, encouragement, promotion of
problem-solving behaviour, suggestion, and environmental manipulation.
However, some non-confrontational interpretations may also be useful.
There may beflexibilityabout timing of sessions or unscheduled contracts,
(for example, by phone), and a high degree of availability of the therapist
(or colleague) can be a successful strategy in responding to a crisis, without
allowing excessive dependency to develop.
For borderline PD patients, supportive psychotherapy may take place
once a week or less frequently, and continue for up to several years in
patients with severe disorders. For such patients a variety of schedules may
be appropriate, so that frequency of sessions may range from once a week to
once every 1-2 months, depending on the current level of social adaptation.
A 'holding environment' is often important for borderline PD patients,
which requires the therapist to offer a degree of personal warmth, consist-
ency, and flexible availability, within stated limits. 'What these patients
need is an experience, not an explanation' (Fromm-Reichmann, quoted in
Higgitt & Fonagy, 1992). Higgitt & Fonagy (1992) consider that, on
available evidence, techniques derived from both psychodynamic and
supportive psychotherapy may be useful for some patients with borderline
and narcissistic PD. Although Kernberg believed that supportive tech-
niques for some borderline PD patients were harmful, because dependence
on a therapist would be added to the problems without significant benefit,
he subsequently described a variant of supportive psychotherapy for
patients with borderline PD (Kernberg, 1984). In this approach, the aim is
to improve the patient's adaptation and to increase his/her control over
destructive behaviour. The therapist is less neutral than in psychodynamic
psychotherapy and the frequency of sessions can vary in response to a crisis.
The therapist's role is 'a teacher about the nature of reality and appropriate
behaviour' (Goldstein, 1990), which involves the identification of negative
Supportive and relatedpsychotherapies 263
feelings in the transference and maladaptive psychological defences, and
the process of relating these to the problems they have caused. But further
interpretations are avoided which would link them with past events and
relationships. For example, if the process of 'splitting' is identified (i.e. a
type of psychological defence, perhaps involving an attitude that a person is
either 'all good' or 'all bad'), such a distortion is challenged; i.e. 'is she really
as bad as you think she is?'. But even in this treatment format, Kernberg
believes that attention must also be given to the negative transference
involving feelings of mistrust and anger towards the therapist, and that
some insight into this tendency, with the encouragement of some adaptive
strategies for coping, is necessary if the patient is to be able to form and
maintain more positive attachments. Thus, the therapist tries to get the
patient to recognize such negative feelings, even if they may not be
apparent; for example, a patient may be reminded that although he/she
feels positive today, a few weeks ago he/she was full of complaints.
Kohut's psychotherapy
Kohut (1977) stressed the importance of focussing on the patient's inner
world in an attempt to understand and empathize with his/her thoughts and
feelings. However, it was suggested that this approach should be based on
theories that are controversial, namely that anger in individuals with
borderline PD is due to 'severe and repeated assaults they experienced in
early childhood', and that such a patient 'should have had a lot more
attention and encouragement ...' (Goldstein, 1990). It is important to
remember that an indiscriminate use of such unproven statements may not
only be inaccurate but could set up additional family conflicts. Some
interpretation of psychological defences, such as lateness, is advocated by
Kohut, although this should not usually be attempted in the early stages of
treatment, while empathy should not extend to approving dysfunctional
behaviour. Indeed patients should be confronted with the effects of their
behaviour.
Adler and Buie's model ofpsychotherapy (Adler & Buie, 1979; Adler,
1985)
This approach, for patients with borderline PD, is closer to psychodynamic
than supportive psychotherapy, but is moreflexiblethan the former, as it
draws on a wider variety of techniques. The main aim is to help patients
develop their impaired capacity to evoke and experience positive attitudes
to others in their absence or when frustrated by them. The therapist tries to

be empathic and supportive, in the hope that the patient will develop a
positive image and attitude to the therapist, which is then 'internalized'.
Such a strategy is based on an assumption that a patient with borderline PD
does not have sufficient positive memories and attitudes to previously
significant persons, and that an 'internalized' positive role model of the
therapist will allow more positive feelings to develop in future relationships.
This is in contrast to Kernberg's view that good and bad feelings are 'split',
in that they cannot be experienced together about another person; in Adler
and Buie's model, the capacity for 'good' feelings is missing.
This strategy, emphasizing the provision of safety, security and trust,
involves two tofivesessions per week but the arrangements can be flexible
with a defined provision for unscheduled phone calls and extra sessions.
Additional security can be offered by such techniques as offering to return
any phone call, the provision of the therapist's holiday address so that
the patient can write, and sending the patient postcards when on leave.
But although the therapist is relatively flexible, limits still have to be set
for the continuation of treatment, for example in relation to physical
violence. Other techniques used include clarification, confrontation and
interpretation.
Problems related to the therapist's feelings and attitudes (countertrans-
ference) are similar, whatever form of psychological management is used.
The therapist can feel helpless, hopeless about the outcome, and guilt that
more has not been achieved. Such feelings can lead to inappropriate
withdrawal, excessive dedication with the setting of unrealistic goals, or the
denial of the patient's hostility.
The Blancks' psychotherapy

Another approach for patients with borderline PD, advocated by Gertrude
and Rubin Blanck (Goldstein, 1990), aims to build up the positive and
adaptive aspects of the individual's functioning.
The patient's achievements and abilities are identified and commented
upon, which can include pointing out the positive aspects of otherwise
problematic behaviour. For example, the frequent expression of inappro-
priate anger may be maladaptive but the patient is told that it is better to
express anger directly than drop out of treatment. Adaptational strategies
can be encouraged or developed, and patients can be helped to identify
triggers for maladaptive anxiety or anger, together with alternative ways of
coping. Care is taken not to encourage patients to express anger to the
Psychotherapy of borderline personality disorder 265
therapist early in treatment if it is not balanced by more positive feelings,
which are identified and reinforced. Patients are encouraged to tell the
therapist about the details of their thoughts, although many patients find
this difficult and have an unrealistic wish to be understood without the
effort of verbal explanation. Attention is also given to helping the patient to
recognize and describe their various mood states and to understand how
these may be part of their patterns of maladaptive behaviour. Patients can
be regularly asked for their suggestions, rather than allowing an implicit
expectation that all the wisdom and effort resides in the therapist. Also,
some psychological defences can be adaptive and should be recognized as
such; for example, it may be appropriate for a patient to want a break from
treatment sessions to prevent the development of excessive dependency.
The development of the patient's autonomy is encouraged, perhaps involv-
ing less dependency within the therapeutic relationship and the develop-
ment of new social and recreational activities. The therapist needs to
recognize that some patients become more demanding during the time that
improvement in symptoms or functioning is taking place, because of their
concern that their treatment and support will be lost if it is no longer
needed.
This approach may involve one to three sessions per week or less often for
longer-term maintenance. The therapist takes an active interest in the
patient, is flexible within defined limits, and positive feelings for the
therapist are encouraged. In addition, some explanation and interpretation
aimed at distortions of the here-and-now relationship with the therapist can
be appropriate.
Psychotherapy of borderline personality disorder: methods compared

Much of the literature on psychotherapy for PD has been concerned with
borderline PD, and the clinician is faced with a confusing array of
theoretical frameworks, techniques and formats for the provision of
services.
While individuals with borderline PD form an heterogeneous group that
requires a range of psychotherapeutic approaches, there is uncertainty
about aspects of the theoretical frameworks on which some types of
psychological management are based. For instance, Kernberg has argued
that the patient's anger needs to be addressed before more positive,
adaptive, feelings can develop. But excessive early confrontation and
encouragement of the expression of anger can be harmful, and this
approach may allow a therapist to ignore that some of the patient's anger to
the therapist may be justified and is not necessarily a distorted reaction.

Also, the hypothesis that the encouragement of positive feelings should be
delayed can be used to justify the therapist's relative non-involvement with
the patient, although an active approach is generally considered to be
helpful for most patients. According to Kernberg, an early focus on anger is
the most effective means of significant change for the most destructive
patients and, while this is possible, it is unproven. Such a strategy must take
into account the risks of excessive anger leading to violence to others or to
self, and drop-out from treatment. Other authors have noted the import-
ance of several ingredients in the therapist-patient relationships for the
management of borderline PD; namely, initial acceptance of childlike,
dependent behaviour, rather than confrontation (Masterson); encourage-
ment of the development of positive, adaptive, examples of the patient's
behaviour (the Blancks); a human empathic approach (Kohut); and
availability andflexibility(Adler & Buie).
In routine psychiatric practice, it would seem appropriate to adopt a
flexible integrated approach, which draws on many of these styles and
formats, for the majority of patients with borderline PD (Goldstein, 1990).
A positive, empathic relationship with the patient is attempted and, in
general, the emphasis is not on confrontation or on interpretations that
attempt to link hidden feelings to past experiences. However, a limited
number of interpretations that help the patient to recognize and manage
hidden emotions, attitudes and fears can be useful, particularly if insight
can be gained into current behaviour; for example, the patient may
recognize that unexpressed anger, or fear of being alone and abandoned,
can be a problem in close relationships. Some limits need to be set for
therapist availability, while so-called 'transitional objects' or situations can
be provided; for example the patient phoning the therapist (or a colleague)
under certain circumstances and times, the provision of vacation addresses
so the patient can write when appointments are not available, and a diary in
which the patient is encouraged to note thoughts and feelings to be shown
to the therapist at the next meeting. The degree to which the clinician should
gratify excessive dependency needs of the patient is subject to theoretical
disagreements, and the appropriate decision is likely to vary. For some
patients, acceptance of requests for frequent appointments and toleration
of destructive behaviour without initial confrontation may be necessary for
engagement in treatment, but there are limits to what the therapist can
tolerate both emotionally and practically. There has to be a limit to the
availability of the therapist's time, while physical destruction or violence to
others must trigger containment measures or discharge from treatment,
above a certain threshold. It may be important to pay attention to the
patient's family and social environment, by involving significant others in
psychotherapeutic sessions.
Cognitive psychotherapy
Overview
Cognitive therapy (CT) is a form of psychological management in which
the therapist talks with the patient, focussing on how recurrent maladaptive
conscious thoughts are used to assign meaning to events, how these may
contribute to the patient's current behaviour and problems, and how the
patient can develop skills to change such patterns of thinking and behav-
iour. Such an approach requires a good therapist-patient relationship with
shared goals. Maladaptive patterns of thoughts can give rise to emotional
changes, and CT has been shown to be useful in the management of some
types of depression and anxiety, but has also been applied to PDs. It is a
relatively short, time-limited treatment, perhaps involving from between 10
and 30 sessions, and patients are told in advance that the aim is for them to
develop problem-solving skills that will be available to them after treatment
ends.
Beck & Freeman (1990) have termed stable, ingrained patterns of
conscious thought as 'schemas', which select and synthesize information
and assign meanings to events. Schemas can be considered to consist of a set
of related cognitions, or smaller units each involving a relatively narrow
belief, and can be evaluated in relation to their breadth, flexibility,
prominence and threshold for activation. Schemas can be considered as one
type of basic unit of personality and most PDs; for example, 'I am helpless'
would relate to dependent PD. Most PDs have characteristic schemas,
although Beck & Freeman (1990) have noted that borderline and schizoty-
pal PDs appear not to have specific beliefs. Schemas have been called 'rules
for living', and maladaptive examples associated with achievement, accep-
tance by others and personal control are often found in individuals with
PDs.
Maladaptive schemas, involving a patient's dysfunctional thoughts,
attitudes and beliefs, are explored in CT, and the associations between the
schemas and the patient's problems are identified. Three types of cognition
are relevant: core assumptions (e.g. 'I am a helpless person'); conditional
assumptions (e.g. 'If a person does not show me a lot of concern, he/she
does not care about me at all'); and goals that are the result of core
assumptions (e.g. 'I want people in my life who can give me the help I need')
(Beck & Freeman, 1990).
The process of modification of such schemas involves a variety of
procedures: the patient prepares a written record and evaluation of his/her
cognitions and behaviour; irrational beliefs are discussed, disputed and
tested; and new skills and strategies are devised and practised in 'homework
assignments'. Evaluation of schemas may include examining the antece-
dents and consequences of behaviour, as well as the mediating thoughts,
beliefs and feelings. The goals of CT include not only a change in overt
behaviour, but also a modification in irrational or unproductive thoughts,
beliefs and emotional responses to various stimuli.
Despite its structured theoretical framework, CT is not a fully automated
treatment, and the relationship between therapist and patient is of crucial
importance, requiring an atmosphere of collaboration and trust. A posi-
tive, non-judgemental approach by the therapist is needed and, at times, the
transference relationship should be discussed if the patient's distortions of
the reality are interfering with the progress. Also, the therapist may serve as
a model for new patterns of behaviour.
Another dimension to this form of treatment, which is relevant to all
forms of psychological management, is the degree of emotion (anxiety,
happiness, depression or anger) experienced and displayed. As with other
clinical settings, it is believed that the degree of change in conscious thought
processes often depends on the patient experiencing a certain level of the
appropriate emotional accompaniments.
The patient's written notes of problems and associated mental activity
are usually a central focus of CT treatment sessions. For each problem the
'ABC method' can be used for the patient's record: in column A the
activating event which triggers a problem is noted; in column B, the beliefs,
feelings and behaviour in relation to the situation are recorded; while C, the
third column, records the consequences. A final column D, for the later
stages of treatment, notes the patient's plans to 'dispute' and modify the
sequence (Beck & Freeman, 1990).
Practical strategies
The first stage of CT, the exploration of schemas of maladaptive thoughts
and feelings, occurs in regular sessions supplemented by the patient keeping
a written record of relevant beliefs, feelings and behaviour (together with

triggers and consequences).
Relevant incidents that have an association with the patient's PD are
identified, and the patient is invited to talk about his/her associated
cognitive processes, perhaps helped by trying to relive an incident in the
imagination. The patient has to try and recognize 'automatic thoughts', in
certain situations and any core assumptions, conditional assumptions and
associated goals. A series of questions (cognitive probes) may be needed to
help the patient to arrive at a 'core assumption' which underlies other types
of automatic thought. Beck & Freeman (1990) have given the following
example:
'Linda is ignoring me'

Q: 'What did that mean?'
'I can't get along with people'
Q: 'What does that mean?'
'I will never have any friends'
Q: 'What does it mean "not to have friends"?'
'I am all alone'
Q: 'What does it mean to be all alone?'
'That I will always be unhappy'.
It is necessary to identify which automatic thoughts are 'core assumptions',

as these need to be the target of methods designed to modify cognitive
processes. These can be explored by repeatedly asking 'What does that
mean for you?'.
The second stage involves attempts to restructure (i.e. replace or
modify) irrational thoughts, beliefs, emotional responses and behaviour,
although a more modest aim may be to help the patient alter behaviour and
its consequences, without a significant change at the cognitive level.
Identified schemas that are inaccurate, distorted or wrong, can be repeat-
edly challenged in an organized process. This can involve writing down and
discussing alternative rational approaches. The patient can be encouraged
to ask the following: 'What evidence is there?'; 'what alternative views are
there?'; 'what are the advantages and disadvantages of this way of
thinking?'; and 'if my interpretation is correct, are things as catastrophic as
they seem?' (Moorey, 1991). Often there is a tendency towards all-or-none
thinking, so patients can be asked to rate certain variables on a 5-point
scale; for instance, whether a person is 'worthless'. It can also be useful to
explore with the patient various aspects of maladaptive cognitions and
behaviour, such as certain advantages (e.g. avoiding the risk of a close
relationship), as well as disadvantages (e.g. loneliness). Other procedures

include planning for the patient to test whether some of his/her attitudes are
correct (e.g. 'I can't cope with certain situations and something will go
wrong'); encouraging the patient to imagine the worst possible outcome
(deliberate exaggeration) before identifying any discrepancy between this
and what is likely to happen; asking the patient to list other possible
outcomes of a situation if a negative outcome is always anticipated; and
discussing various explanations for other people's behaviour. A diary
related to ongoing maladaptive thoughts and feelings, with their associa-
tions, can be supplemented by 'predictive diaries' in which the patient
anticipates the outcome of certain situations on the basis of their maladap-
tive automatic thoughts. Subsequently, the patient notes what actually
happened. Another diary method involves the patient analysing events in
terms of alternative schemas.
Cognitive therapy is often called cognitive-behavioural therapy as it can
be used with other techniques that focus on behaviour. These include
planning behaviour to minimize anxiety (anxiety management), and assist-
ing decision-making by setting priorities amd listing pros and cons, perhaps
weighting individual items. Other relevant behavioural techniques are
assertiveness training, role playing (including roles as the patient in various
situations or as other people), relaxation training, distraction techniques,
'in vivo' exposure to feared situations, encouragement to be involved in
activities that give pleasure or a feeling of achievement, the realistic
anticipation of future problems, and time management planning. The use
of imagination has also been claimed to be useful for patients with PD, and
can involve reliving certain childhood experiences and traumatic events.
This can be associated with some 'restructuring' of the experience, when
other methods of coping are rehearsed, perhaps with associated role
playing. For example, unfair criticism which had been previously accepted
without comment can be recalled but this time the patient argues in his/her
defence.
As with all forms of psychological management, non-collaboration may
be a problem due to factors such as lack of patient's capacity to develop new
skills, lack of experience and skill in the therapist, advantages to the patient
in maintaining the PD features, lack of motivation to change, the setting of
unrealistic or vague goals by patient or therapist, and frustration at the slow
rate of progress.
Andrews (1993) has reviewed the evidence from controlled trials that
cognitive and cognitive-behavioural treatments can benefit some depres-
sive disorders, generalized anxiety disorders, obsessive-compulsive dis-
orders, bulimia nervosa and social phobias. In relation to depressive

disorders, Robinson, Berman & Neimeyer (1990) found that cognitive-
behavioural treatments were better than other verbal psychotherapies,
which did not differ from placebo, while in a multicentre trial of treatment
for depressive disorders, cognitive-behavioural techniques were associated
with modest benefit in excess of improvement on 'routine clinical manage-
ment', which was the placebo condition (Elkin, Shea & Watkins, 1989).
These results are likely to have some relevance to PDs, which may present
with symptoms of depression and anxiety, and may lower the threshold for
the emergence of other disorders (such as eating disorders) in vulnerable
individuals. There is less clear evidence for the efficacy of CT or cognitive-
behaviour therapy for features of PD, but claims have been made in relation
to avoidant PD and recurrent self-harm behaviour (Linehan, 1987; Alden,
1989). Also, in a study of opiate addicts in a methadone maintenance
programme, Woody et al. (1985) found that those with both antisocial PD
and a depressive disorder responded to CT, showing improvements in
psychiatric symptoms, drug use, employment status and illegal acts. But
those with antisocial PD and no depression showed little response. It is
likely that CT is less prone to harmful effects such as a transferance with
excessive violence or dependence, and, compared with psychodynamic
psychotherapy, it is less expensive. Also, staff training to an acceptable
standard is more readily achieved.
Despite these encouraging reports and the belief among many clinicians
that some patients with PD can benefit from cognitive and behavioural
techniques, these treatments require further controlled evaluation. It is
important to remember that depressive disorders are heterogeneous from
both the clinical and aetiological perspectives.
Cognitive analytic therapy (CA T)

Ryle (1989) has described an approach to psychotherapy that draws upon
concepts derived from psychoanalytic, cognitive and behavioural treat-
ments, with an emphasis upon the cognitive elements. Although various
techniques are used, the focus is on helping the patient to identify and
understand 'procedures', i.e. 'linked sequences of mental and behavioural
processes', which overlap with the concept of 'schemas'. An example of
such a sequence would be: dislike of self, avoidance of others, being seen as
hostile by others, being rejected or ignored by others, a feeling by the
patient that he/she has deserved this, and reinforcement of dislike of self.
The patient is encouraged to identify 'procedures' associated with
problems, and to work out ways of changing the usual sequence. The first
three or four sessions involve the identification of relevant problem-related
'procedures', and the patient is asked to keep a daily self-monitoring diary.
Once a list of relevant procedures has been completed, written copies are
provided for both therapist and patient. A contract is then made, usually
for 12 sessions, with the aim of their modification. Although subsequent
sessions are relatively unstructured, the 'procedures' form a framework for
other techniques such as discussing current relationships or the transfer-
ence. During treatment, the patient is asked to continue daily diary self-
monitoring, based on the identified procedures. Ending of treatment
should be anticipated and discussed during thefinalthree or four sessions,
when attention is paid to the transference in this respect. The patient should
be reminded of his/her new strategies and ways of thinking, and the aim of
increasing self-control. A follow-up appointment after 2 months may be
appropriate.
Behaviour therapy
In the consideration of cognitive therapy, it has been shown that techniques
of psychological management that focus on thoughts, feelings and beliefs,
are also usually concerned with some aspects of the patient's behaviour. But
although the separation between cognitive and behaviour therapy is
somewhat arbitrary, techniques that are aimed primarily at the patient's
behaviour can be termed 'behaviour therapy'.
A particular regime for patients with borderline PD who show repeated
self-harm episodes has been developed and evaluated by Linehan and
colleagues (1991) and is known as 'Dialectical Behaviour Therapy' (DBT).
This treatment, for a clinically important and problematic patient group,
was compared with 'treatment as usual' at completion of treatment and at
1-year follow-up, and it was found that those who received DBT for 1 year
had a significantly lower drop-out rate and less self-injury, despite both
groups showing only slight improvement in depression. The superiority of
DBT over 'treatment as usual' was still evident at a 1-year follow-up
(Linehan, Heard & Armstrong, 1993).
Details of this treatment have been described by Shearin & Linehan
(1993) and involve an individual session and a group meeting each week.
Individual sessions last at least 1 hour, while the group meeting, with two
co-therapists, lasts for at least 2 hours. The initial emphasis in the group is
on the teaching of behavioural coping skills, but subsequently it can also
function as a general support group. In addition, phone contacts in a crisis
Behaviour therapy 273
(mainly with the individual therapist), are encouraged on an as-required
basis.
In the individual sessions, patients are encouraged to acquire new
behavioural skills, and particular behaviours are targeted for discussion;
for instance, self-harm behaviour, behaviour that interferes with treatment
or with quality of life, behaviour related to past stress, and behaviour
related to the patient's level of self-respect. It is particularly important to
identify self-harm behaviour (and thoughts), and patients are asked to
complete a daily diary card noting any self-harm items. At all times self-
harm behaviour is given the highest priority, while behaviour that interferes
with treatment must be targeted before focussing on other topics, as the
process of treatment requires the patient's collaboration.
In relation to self-harm behaviour, the therapist needs to obtain infor-
mation about preceding and subsequent events, so that a more adaptive
strategy to similar situations in the future can be discussed. At all times the
therapist emphasizes that self-harm behaviour is taken seriously and that
such behaviour reduces the opportunity to discuss other issues. The aim is
for the patient to feel that self-harm behaviour, in particular suicidal
behaviour, will not be ignored and that it should be thought of as a signal to
try out active coping strategies, such as phoning the therapist. The next
priority is to address behaviours that interfere with treatment, such as not
completing diary cards or avoiding other 'homework' tasks, and it has been
noted that reinforcement (e.g. by praise) of partial task completion
(shaping) can be effective. It is also important for the therapist to be aware
of the counter-transference, and to resist inappropriate acceptance that the
patient should discontinue treatment. Priorities will need to be set when
targeting behaviours that interfere with the patient's quality of life, and
may involve focussing on substance abuse, maladaptive sexual behaviour,
criminal behaviour and the selection of abusive partners. Further goals of
individual or group sessions in DBT are to help the patient acquire new
behavioural skills such as carefully observing the environment; learning to
put up with distress using methods of distraction and comfort; regulating
the expression of emotion; setting realistic goals; avoiding or controlling
stimuli that trigger maladaptive behaviour; and increasing the patient's
interpersonal skills, such as assertion, communication, or assessing the
motives of others. The latter may involve pointing out that the patient must
not habitually judge people or situations as either 'all good' or 'all bad',
which is commonly found in relation to borderline PD. Certain behaviours
related to past stress may be modified by discussing their antecedants, for
example, childhood abuse, which may have contributed to a patient's self-
blame and low self-esteem. Again, the tendency to have a judgemental 'all-
or-none' thinking needs to be challenged. Finally, behaviours are encour-
aged that increase the patient's self-respect; for instance, maintaining
opinions even when opposed, attempting problem-solving behaviour, and
improving self-care.
While the focus of DBT is the patient's behaviour, aspects of cognitive
therapy can be incorporated; for example, dysfunctional beliefs can be
challenged. Also, the therapist-patient relationship is of major importance,
as it is with all forms of psychological management. A degree of self-
disclosure and responsiveness on the part of the therapist is considered
appropriate and can include telling the patient about the negative effects of
the patient's behaviour on the therapist's motivation to continue. Flexible
availability of the therapist (or colleague) by phone is an important part of
the regime, and the patient is expected to call if he/she feels suicidal. The
therapist requires a high tolerance of criticism and hostility, and an ability
to experience a patient's distress without feeling forced to 'do something' as
a response to his/her own reaction, rather than to the reality of the
situation. Appropriate goals need to be set, and care must be taken to help
the patients prepare to end their contact with the treatment programme. An
inpatient regime for patients with borderline PD, based on DBT, has been
described by Barley et al (1993).
Good clinical care
Overview
Andrews (1993), on the basis of Australian data, has estimated that, at any
time, 1600 persons from a catchment area of 100000 would be receiving
some kind of intervention from medical services for non-psychotic psychi-
atric disorders, including PDs. While some of these patients may respond
well to medication, such as those with some types of depression and anxiety
disorders, psychological management would probably be appropriate for
the majority. But the provision of psychodynamic psychotherapy or even
shorter courses of cognitive behaviour therapy would only be available for
a minority.
Andrews has pointed out that before the era of a range of effective
treatments in medicine, doctors had a major role in explaining, reassuring
and providing regular contact with people to help them cope with chronic
disorders. Such a process can be called 'good clinical care' and involves
establishing a good professional relationship with 'therapist genuineness,
empathy and non-possessive warmth'. This 'holding' process involves non-
specific support, as well as the recommendation of any appropriate specific
treatment, and there is considerable overlap with various forms of suppor-
tive psychotherapy. As gradual improvement can occur in many PDs, an
explanation of what kind of problem the patient has can help him/her to
accept what cannot be significantly altered in the short-term, and to make
the best of adaptive aspects of personality. This can be combined with
advice given within a relatively simple problem-solving approach, which
initially involves the identification of any problems that trouble the patient,
the clinician or significant others. Some of these may be amenable to
improvement by certain actions on the part of the patient, which can be
specified. The clinician needs to be able to tolerate 'not doing anything'
(despite pressure from patients or relatives), if there is nothing to be done
other than to maintain concern about the patient with a chronic disorder
and to be prepared to see the patient again, for instance, 'to see how you are
getting on'. Faced with this task, clinicians whofindit difficult not to be in
control of the patient's disorder or not to be offering specific treatment, may
inappropriately discharge the patient, providing themselves with the excuse
'there is no further treatment I can offer'.
It is probable that the psychological management of most patients with
PD who are seen within clinical services consists of 'good clinical care',
which, in broad terms, has been considered to be 'more effective, cheaper
and less harmful than dynamic psychotherapy, and less efficacious but
more broadly applicable and comparable in safety and cost to cognitive
behavior therapy' (Andrews, 1993). Despite the limitations of such a
generalization, these claims concentrate the minds of those who have the
task of planning services and allocating limited resources. But although
'good clinical care' can appear to be deceptively simple, it requires
considerable skill and experience and deserves more attention in training
and research.
Practical strategies
Within the general heading of 'good clinical care', psychological manage-
ment in a routine medical setting can draw upon psychoanalytic, cognitive
and behavioural techniques, with considerable flexibility. Sessions with
partners and family, together with the patient or separately, may also be
appropriate. Although this might seem to be a recipe for everlasting chaos,
certain principles can be identified, which are then applied to the manage-
ment of each individual patient.
The main format for intervention involves one-to-one sessions, each

lasting from a few minutes to an hour. The frequency also varies considera-
bly, from several times a week to once every 3-6 months. Even an
infrequent appointment may provide an important 'holding' function for a
patient with a chronic disorder. Some of the goals may be that symptoms do
not increase, self-harm does not occur, and a series of unproductive
referrals and investigations is avoided. Sessions are generally unstructured,
although cognitive and behavioural elements can be included. The thera-
pist should not demand early extensive self-disclosure.
For psychological management in the common setting of 'good clinical
care', the main framework is usually 'supportive' psychotherapy, which
implies that the clinician provides a consistently active input. The therapeu-
tic ingredients include advice, explanation, instruction, suggestion, a caring
attitude with active listening, reassurance, encouragement, persuasion,
permission for the display of emotion (in particular depression related to
loss), and environmental manipulation if this is possible and desirable. The
focus is usually on the present and future, rather than on the past. However,
more 'explorative' techniques can be incorporated, directed towards the
past and an awareness or understanding of hidden (unconscious) mental
activity; this may involve an examination of transference, countertransfer-
ence, psychological defence mechanisms, and resistance. The latter is
defined as obstacles to the aims and procedures of treatment that arise from
within the patient. Techniques of asking the patient to 'free-associate' (i.e.
to talk about what comes into his/her mind) or to discuss dreams, may
occasionally be appropriate in this setting but are generally reserved for
more time-intensive psychodynamic psychotherapy.
Several studies have claimed that certain characteristics of a therapist are
associated with beneficial effects, such as an ability to empathize with the
patient and to be personally concerned about his/her welfare, together with
emotional warmth and genuineness (Malan et ah, 1976). The 'treatment
alliance' (Sandier et ai, 1970^), refers to the common aims of therapist and
patient, and its development can also be beneficial, as can the provision of
new learning situations; for instance the patient may learn that it is possible
to be assertive without loss of temper.
Cognitive and emotional components of a learning process can usefully
be distinguished, although, in practice, these are never completely separ-
ated. Cognitive learning may involve conscious mental activity, which
includes the registration of new experiences, or, by the process of insight,
the development of an awareness of attitudes, feelings or conflicts that had
not been fully available to consciousness. Insight can be induced or assisted
by suggestions (interpretations) from the therapist, which are directed
towards the defence mechanisms responsible for the patient's transference
or resistance. However, even though a patient may develop some insight,
perhaps by becoming aware of the maladaptive nature of a behaviour
pattern or attitude, this may have little impact if the learning process did not
involve an emotional dimension, which was concerned with motivation and
mood. If emotional arousal occurs during a series of learning experiences,
the effects are more likely to include an emotional as well as a cognitive
change, so that emotional arousal has been termed the 'motive power of
psychotherapy'. Psychological events that are associated with the exper-
ience and expression of emotion are usually necessary for significant
personality adjustment in a psychotherapeutic context. Although insight
accompanied by little or no emotion can be of benefit, some patients'
behaviour does not change even though they appear to understand why
they are behaving maladaptively and the consequences of their actions.
Psychological defences include denial (a special type of denial is known
as 'isolation', when different components of a reaction are separated, for
example, when an emotional reaction to bereavement is not consciously
experienced); avoidance; projection (when unconscious mental events in
the patient are attributed to others, for example, 'my sister distrusts
doctors'); sublimation (when unconscious mental events produce con-
scious experiences that are similar but more acceptable, for example, anger
can be harnessed to produce success in aggressive contact sports); intellec-
tualization (when mental events are disguised by complex argument and
discussion); rationalization (when mental events are hidden by apparent
explanations that are wrong, for example, 'I became manic again just
because I was upset, not because I stopped taking my lithium carbonate');
reaction formation (when the opposite of unconscious mental events
appear in conscious awareness, for example, an angry patient may appear
excessively deferential); displacement (when mental events remain hidden
because the patient concentrates on other aspects of life); and regression
(when a person may return to a childlike pattern of thinking and behaving,
for example, becoming excessively dependent with frequent temper tan-
trums). But not all psychological defence mechanisms are maladaptive, and
psychological management is often directed towards a strengthening of
these unconscious manoeuvres to hide or distort reality, if they provide the
best method of coping with an insoluble situation or a severely maladaptive
personality structure. This might involve encouraging the patient to
continue to avoid certain situations, even though this avoidance may cause
some problems.
Patients can often be helped to recognize what they can achieve by their
own efforts, while identification with the therapist, involving the patient
modelling him/herself on the therapist's attitude and behaviour, may also
be of benefit. Another therapeutic ingredient is the instillation of hope, but
this must not involve making wild promises and predictions which, if
unrealized, will produce despair and loss of confidence. A patient's hope is
often a reflection of a positive approach by the therapist and, even when
severe and intractable problems exist, it may be possible to comment on the
patient's fortitude, transient symptomatic relief or the hope of a natural
remission. Also, the therapist's continued concern and interest encourages
a positive attitude in the patient. Another important therapeutic ingredient
can be catharsis, involving the patient's expression of anger, depression,
anxiety or, more rarely, positive feelings. Catharsis can be therapeutic in
itself, as was shown by a young man with a severe depressive disorder and
anxiety following, and causally related to, the death of his alcoholic father.
Although he clearly recognized his feelings of intense anger towards his
father for ruining the life of his family, these conflicted with positive
feelings, and it was not until he had repeatedly expressed his anger that his
symptoms were relieved. Self-disclosure, particularly if this involves con-
fessions, may also provide relief of some forms of depression and anxiety in
association with PD, even if this does not generate specific advice or other
interventions.
For patients with chronic disorders who receive 'holding' and 'support'
on an indefinite basis, the clinician needs to remember the aims of such a
strategy. These may include the prevention of deterioration in functioning,
even when there is little or no 'treatment alliance', so that the patient may
have a different view of what is happening; for example, he/she may be
hoping for and expecting a new treatment. Sometimes this can be acknowl-
edged, by telling the patient that while his/her disorder is being monitored
at the regular visits and that no change in treatment is likely to help at
present, the possibility of a change in management will be regularly
reviewed. In relation to PDs, the patient's presenting complaint may be of a
mood disorder, perhaps resulting from social difficulties. After listening to
and sharing the patient's account of the various problems and symptoms, it
may not be possible to give any specific advice or treatment and the clinician
needs to be able to 'do nothing', while still taking an interest in the patient
and arranging a further appointment. For example: 'I am very sorry that
you have had such a bad time in recent weeks and I know how difficult it has
been for you to keep at work. But I am glad you were able to share this with
me. I am afraid we have run out of time this session; let me give you another
Couple, family and group treatment 279
appointment'. This approach can be difficult with those patients who
engender feelings of guilt in their therapists and demand action or specific
advice. If challenged directly, the clinician can only state the reality, namely
that he/she is not able to make any other specific suggestions but that a
further appointment is recommended.
Couple, family and group treatment

'Significant others', mainly those living with the patient, may be incorpor-
ated in programmes of psychological management, in particular for those
with borderline PD. There is a range of formats for such involvement,
including sessions with the patient and partner/family (conjoint sessions) as
the only input, and sessions with significant others in parallel with, or in
sequence with, individual sessions with the patient. One therapist may be
involved throughout, but a second therapist may be a co-therapist or sole
therapist for partner/family or conjoint sessions.
For borderline PD, couple/family sessions may be useful when the
relationships seem to be aggravating or perpetuating problems, or where
the significant others are participants in dysfunctional beliefs, attitudes and
behaviour (Goldstein, 1990). Such sessions are often appropriately offered
to younger patients who are emotionally orfinanciallydependent on their
family, who may seem overinvolved. In such cases the patient may not have
been given an appropriate degree of independence.
The patient may display transference feelings to a partner/family
member, which can be the focus of discussion and attempts at modification.
The variety of techniques available for individual sessions can be con-
sidered for conjoint treatment but, in general, experienced clinicians have
recommended that the focus should be on the 'here-and-now', and that the
therapist should be active in setting a clear structure with limit setting, the
facilitation of communication and trust, and the containment of anger. The
participants in conjoint sessions can be encouraged to identify maladaptive
interactions and sequences of behaviour within their relationships, to be
less defensive and to be more aware of underlying feelings and motives. This
may involve some interpretation of psychological defences, such as ques-
tioning whether there is just one good or bad member of the partnership/
family. Participants can be encouraged to understand the feelings of others,
and positive behaviour can be reinforced with the aim of increasing self-
esteem and cohesion. Sometimes it can be useful for the participants to
understand how their present behaviour is linked to past experiences.
If the patient has been the victim of childhood sexual or violent abuse, or
neglect, family therapy presents particular difficulties and risks. Strategy

has to be determined by an evaluation of the individual family by an
experienced clinician, and care must be taken not to produce an emotional
experience of an intensity that cannot be satisfactorily handled and
resolved. But patients are often ambivalent about their feelings to their
abusers and may direct their anger and dislike onto themselves; in such a
situation, some exploration of the abuse may be appropriate, although
most of this work may need to be done in individual (one-to-one) sessions.
The partner/family may be angry with the patient for being 'in treatment'
if this requires their participation, and be worried that they will be blamed
for the patient's problems. It may be necessary to see parents (or relevant
others) alone at least once to give as much reassurance on this aspect as is
possible.
Whatever the therapist's views about the distribution of blame and the
causes and effects within a family, an acrimonious and angry content of
sessions is generally counter-productive. A positive, educational, suppor-
tive atmosphere should usually be fostered.
Various forms of group psychotherapy may also be part of a treatment
regime for patients with PD. For those with borderline PD, a group
focussing on behavioural skills was part of the regime of 'Dialectical
Behaviour Therapy', and it has been suggested that parallel group sessions
can be used with sessions of individual psychodynamic psychotherapy for
some patients with borderline and narcissistic PD (Higgitt & Fonagy,
1992). Most clinicians with special experience have believed that, for
borderline PD, group treatment should not completely replace individual
sessions, but can be a useful adjunct (Gunderson & Sabo, 1993a). Some-
times a patient's peers in a group session are better able than the therapist to
effectively confront maladaptive behaviour, as other group members are
not subject to such strong transferences as is a single therapist. Also,
various self-help groups, such as Alcoholics Anonymous, may be helpful to
some patients with PD.
Treatment settings
Any psychiatric service must provide a range of interventions with in-
patient, daypatient, outpatient and non-hospital-based facilities.
Borderline PD is the most common PD diagnosis in psychiatric inpatient
units, and such patients are usually admitted for relatively short periods
(i.e. from a few days to less than 4 weeks) in response to a crisis involving
self-destructive or antisocial behaviour. A low threshold has been advo-
Management of specific problem-behaviours 281
cated for the admission of apparently suicidal borderline patients, with the
aim of defusing the crisis and organizing outpatient or other alternative
input (Gunderson & Sabo, 1993a). The length of inpatient stay will be
related to the level of social support and to the degree of problems in the
social environment. Long-term hospital treatment for those with PD, in
particular with borderline and antisocial features, is rarely available but,
for a small minority, a long-term residential treatment programme may
seem appropriate. Alternatives to a residential hospital place include hostel
accommodation combined with day or evening treatment programmes,
and such arrangements may be better for those individuals who tend to get
excessively dependent on long hospital admissions. Such dependency may
cause 'regression' to a childlike overdependence, self-preoccupation and
intense anger at the inevitable limits and controls of a hospital regime, with
demands for instant gratification. For those borderline PDs in hospital, a
structured timetable is advisable with psychotherapeutic sessions and other
occupational, social or recreational activities. Group meetings can be a
useful adjunct with, for example, discussion of abuse experiences, family
problems and work plans.
Management of specific problem-behaviours related to personality

disorders
Self-harm and suicide

Self-harm involving 'recurrent suicidal threats, gestures, or behaviour, or
self-mutilating behaviour' is part of the syndrome of borderline PD
(American Psychiatric Association, 1987). Although there are associations
between self-harm with other PDs, such as antisocial PD, the focus in the
literature of PD in relation to self-harm and suicide has been on patients
with predominant features of borderline PD. Self-harm is a broader
category than suicidal behaviour, as it includes self-mutilation, such as
inflicting multiple superficial cuts on the wrists and arms, or more severe
examples such as hitting limbs with a hammer. But there may be various
motives, or combinations of motives, for self-harm behaviour, including
the expression of anger, the relief of tension or a wish to influence the
behaviour of others. Self-harm must invariably be taken seriously, as
several studies have shown that the history of frequency of self-harm
episodes is significantly related to completed suicide (Paris, 1993).
Several long-term studies of suicide in groups of borderline PD patients
have indicated a significant risk, in the order of 1 in 10, for those who have a
history of repeated self-harm and have become known to hospital services

(McGlashan, 1986; Paris et ai, 1987; Stone, 1990). Also, Stone (1990)
found the co-occurrence of substance abuse with borderline PD was
significantly associated with suicide, and Rich, Fowler & Fogarty (1988)
have reported that the large (x 3) increase in youth (mainly male) suicide in
North America from 1960 to 1980 was associated with substance abuse.
Other data (Runeson & Beskow, 1991; Paris, 1993), from retrospective
studies of suicides, suggested that about 30% of male youth suicides
merited a diagnosis of borderline PD. Another factor that may be asso-
ciated with suicide is relatively high educational achievement, and it has
been suggested that this reflects higher expectations that have been
thwarted (Paris, 1993).
The outcome of severe borderline PD gives some hope as well as
confirming the serious risk of suicide, as about 75% of those who survive
are considerably improved after 15 years (Paris et al., 1987) and suicide
after the age of 30 is relatively uncommon. There is, as yet, no evidence that
interventions or treatment make any difference to the incidence of suicide in
borderline PD or to the process of natural remission, but our present
understanding suggests that interventions can prevent completed suicides
and that if the patient can be kept alive, in particular by help in a crisis, the
outcome of patients with even very severely maladaptive behaviour can be
surprisingly good.
The management of self-harm, particularly in those with borderline PD,
presents common practical problems and different theoretical approaches.
Controversial aspects are the role of extended inpatient or other residential
treatment for those who are chronically suicidal, and the emphasis on
suicidal thoughts and actions during treatment sessions.
Although the approach in Dialectical Behaviour Therapy (Linehan et al.,
1991), which has been shown to produce short-term improvements in self-
harm behaviour, gives priority to the consideration of self-harm thoughts
and behaviour, this is within the context of a structured programme. But in
routine psychological management some limits have to be set with regard to
how the therapist responds to talk of suicidal intent.
For some patients, displaying the power to commit suicide may seem the
only way to obtain some control over themselves and their environment; in
this sense it is a coping strategy that needs to be replaced before it can be
discarded. This might be achieved if the patient can develop successful
relationships or a career but this takes time, and the practical problem
remains: how should the clinician respond to threats of suicide? There is a
general consensus that relatively brief hospital admissions can be appropri-
ate for exacerbations in the severity of suicidal or self-harm ideation and

intent, usually associated with a life crisis, but for those who are chronically
suicidal, 'extended inpatient treatment is unlikely to be of benefit and may
perpetuate the behaviour by leading to regression and excessive depen-
dence on hospitalization' (Ennis, Barnes & Spenser, 1985). Such generaliza-
tions are deceptively simple as, in practice, the boundary between a crisis
and fluctuating chronicity is often unclear. What is certain is that there is no
strategy that eliminates a risk of completed suicide, and that the clinician
has only a limited ability to prevent suicide. It is often appropriate to make
this clear to the patient and the patient's family, who may encourage such
patients to recognize their own responsibility and independence (Ennis et
al., 1985). When in doubt about the severity of a particular crisis, it is
probably best to give the patient the benefit of such doubt by arranging a
period of inpatient management, but making it clear that the aim is for a
short stay (a few days or up to 2-3 weeks) in the context of planning
subsequent outpatient or other support. Sometimes a clinician may decline
to admit a patient to hospital, or may discharge a patient, despite the
persistence of threats of self-harm or suicide, if it is considered that the
further inpatient care would increase the risk of such behaviour still further.
Aggression and violence

Aggression has been defined as coercive action, involving the use of threats
or punishments to influence social interaction (Blackburn, 1989). Violence
denotes the use of physical force and is an extreme form of aggression.
Anger is often associated with aggression, but not invariably so, and this
definition of aggression could include self-defence.
Although any example of aggression can be viewed as a specific
interaction between the persons involved, a tendency to repeat aggression
in different situations over time can be identified as a characteristic of some
individuals. But such a concept distinguishes between the personality trait
involving a capacity for aggression and the aggression itself, which is shown
only in certain situations. This means that the identification of the trait of
aggression predicts a series of aggressive acts rather than specific examples
of such behaviour. Various studies have found that a tendency towards
aggression can be a stable trait, and Olweus (1979) stated that 'there is
substantial stability in aggression which cannot be attributed to situational
constancy'. This has been shown in various samples including schoolchil-
dren (Feshbach & Price, 1984), a community sample (Huesmann et al.,
1984), and those with a history of criminal violence (Farrington, 1978;
Robins, 1978). For example, Robins (1978) found that fighting in child-
hood predicted violence in later life.
But even when an aggressive trait has been identified in an individual, the
contexts in which aggressive acts occur are important as additional
determinants of the person's behaviour. Analysis of the interaction of
various causal factors for aggression has been based on hypotheses that
implicate a range of variables, including genetic factors (Rushton et aL,
1986), witnessing and experiencing aggression in childhood (the more
someone has been exposed to aggression, the lower the threshold may be for
the perception of aggression in others), inadequate parenting with lack of
appropriate punishments and rewards, and any social learning that pro-
motes aggression. Even if genetic influences are of major importance in an
individual, these are mediated by cognitive processes that interact with the
environment. Therefore, examination of the cognitive processes preceding
aggression is necessary to understand the relative importance of the various
causal factors in any example of such behaviour.
Whether aggression and violence occur in a given situation may depend
on such factors as the behaviour and status of the other person or persons
involved, anticipation of rewards and punishments, the capacity of the
person to show alternative coping behaviour, the degree of control in
relation to the expression of anger, personality attributes that oppose
aggression (such as a tendency to social anxiety and withdrawal, Black-
burn, 1989), and general intellectual ability (IQ). Heilbrun (1982) found
that, in prisoners with antisocial PD, low IQ was associated with a history
of relatively violent crime, while a degree of social withdrawal has been
found to be negatively correlated with violence in a prison environment
(Heilbrun & Heilbrun, 1985).
The potential aggressor's cognitive appraisal of a situation is an import-
ant variable in determining the subsequent behaviour. If a provoking event
occurs, such as an insult, threat or violence, this is usually judged in relation
to the person's own standards of what 'ought' to happen, whether the other
person's behaviour is intentional, and the degree of maliciousness in any
such intent. (If someone is insulting only when under the influence of
alcohol, this may be considered as less provoking.) Certain appraisals may
be related to a low self-esteem, and involve an emphasis on the need for
'respect' and 'rights', or on the importance of what is considered to be a
masculine self-image. If repeatedly aggressive individuals have some co-
occurring obsessive-compulsive personality traits such as perfectionism,
they may be particularly liable to bear a grudge for a perceived unfairness;
for example, in relation to the way the rules have been applied in a prison
regime. Other examples of relevant aspects of such cognitive appraisals can

include frustration due to delayed gratification if the delay is judged to be
unnecessary, and whether certain incentives such as status or money will be
achieved by not showing aggression.
A tendency to repeated aggression has been a feature of various
definitions of antisocial PD or psychopathy, but while there is a definite
association between antisocial PD and violent crime, and between aggres-
sion, hostility and impulsivity, a group of individuals selected on the basis
of violent crime does not have an homogeneous personality structure. Of
course, a single act of violence may not reflect an aggressive tendency.
Violent offenders have been divided into two broad types: 'undercon-
trolled' and 'overcontrolled' (Megargee, 1966). The undercontrolled
offenders have weak inhibitions and engage in more regular violence, while
the stronger inhibitions of overcontrolled offenders lead to rarer but more
intense episodes of violence. Lane & Kling (1979) found evidence that
overcontrolled offenders were more rigid in attitudes, and more reluctant to
admit to psychiatric symptoms, and Blackburn (1968) reported that
offenders with relatively severe examples of violence were more controlled,
inhibited and defensive, and were less likely to have a criminal record or a
diagnosis of antisocial PD.
Two subgroups of each of these categories have been identified, so that
four main types of violent offender can be described (Blackburn, 1971,
1975, 1986; McGurk, 1978; Henderson, 1982). Undercontrolled offenders
can be divided into 'primary' or 'secondary' psychopaths, while the two
types of overcontrolled offenders are those who are 'inhibited' (with social
withdrawal and anxiety) or 'conforming' (with sociability and lack of
anxiety). This classification is supported by Blackburn's finding (1984) that
52% of primary psychopaths (as described below) had a history of repeated
violence, in contrast to only 8% of the inhibited overcontrolled offenders.
Primary psychopaths are impulsive, aggressive, hostile and extraverted,
and while secondary psychopaths share the first three of these characteris-
tics, social anxiety and withdrawal is found instead of extraversion. Both
groups have antisocial PD traits, and it has been claimed that primary
psychopaths have more narcissistic and histrionic PD features, while
secondary psychopaths are associated with more features of borderline and
paranoid PDs.
The inhibited group of overcontrolled offenders appear introverted,
unaggressive and withdrawn, and these features have been associated with
low self-esteem, social anxiety, offences involving sexual coercion and
features of avoidant, schizoid and passive-aggressive PDs. The conforming
group of overcontrolled offenders seem defensive, unaggressive and

sociable. Such individuals do not appear anxious, deny anger and are often
dependent on approval. Associated PD features are from the dependent
and obsessive-compulsive PD syndromes, and their offences may be
triggered by perceived loss of status.
Management of aggression and violence in psychiatric inpatient units
involves a range of perspectives including how to respond to a particular
incident, the design and layout of offices and institutions, and specific
programmes of psychological management. In relation to aggression in
society, wider issues are also relevant, such as social supports and institu-
tions, the availability of illegal drugs and alcohol, the socio-economic
structure of society, and social alienation of groups such as the young or
members of ethnic or cultural minorities.
The immediate response to a patient's aggression requires a judgement of
the characteristics of the particular situation, but certain principles can be
used as guidelines. The first priority is to try to calm the aggression, which
often involves anger and agitation, using both verbal and non-verbal
techniques (Davies, 1989). The clinician should try to empathize with what
the aggressor is trying to communicate, perhaps by summarizing the
content of what has been said. The initial emphasis is listening, asking
questions and identifying the reason for the anger without arguing,
although some clarification of fact may be appropriate. Sometimes it may
be helpful to 'depersonalize' an accusation, by pointing out that a member
of staff has to act according to hospital policy or the limited availability of
services. Alternatively, if the anger is directed to the institution or to
clinicians in general, the clinician may try the opposite tactic by pointing
out that he/she is not just doing a job, but is also a person in his/her own
right. This can be reinforced by sharing some personal details with the
aggressor. Another technique that has been advocated in certain situations
is to repeatedly ask for the required behaviour (Tlease stop shouting at me')
with increasing force and authority, but maintaining emotional control and
courtesy. Other strategies include disengagement, when the patient is told
that the clinician is leaving and the reasons; offering to get a senior
colleague; promising to try and help; and giving the patient a choice, for
example to wait or be seen again later.
Although it is generally considered sound advice to 'keep calm', Davies
(1989) has pointed out that if this is perceived as unconcern and inaction,
the aggressor's anger may be further fuelled. The motivation of some
aggressors is to continue until a required response, such as fear, is obtained;
therefore it may be appropriate to contain aggression by a statement such
as: 'What you are doing is really worrying me'. Also, it may help for the
degree of arousal in the clinician to approach that of the aggressor to
demonstrate empathy and concern. If a very agitated and angry aggressor is
met with an apparently unconcerned and 'laid-back' approach, this may
provoke further anger, and it may be appropriate for the clinician to show
an increased activity and arousal level by non-verbal gestures (nodding, eye
contact) and frequent verbal statements.
The nature of bodily confrontation also requires consideration. The
clinician should, if possible, not be closer to the aggressor than would be
normal in social interaction, and it is best not to face the aggressor directly.
But if violence is still a possibility, or if the aggressor moves to a direct face-
to-face position, it can be helpful for the clinician to move to a non-
threatening stance, for example, moving the body's weight to one leg and
placing one hand on the chin (Davies, 1989). A normal pattern of eye
contact should be attempted, involving a variable repetition of holding gaze
for a few seconds and then changing gaze, even if the aggressor is engaged in
afixed,hostile stare.
After the immediate threat of violence has subsided, the next step is to
explore the background to the aggression and to apply a problem-solving
approach. However, if violence does occur, an organization requires
procedures to manage the after-effects on those who have been assaulted.
Such individuals may need support from colleagues and management and,
in general, should be encouraged to return to work as soon as possible to
prevent the build-up of fear and loss of confidence.
Various steps can be taken to reduce the risk of violence in clinical
settings. With regard to office design, chairs should not directly face each
other, as a confrontational bodily stance may enhance aggression, and the
clinician's access to the door should not be obstructed by the patient. There
should be an absence of heavy ornaments or other potential weapons, and it
should not be possible to lock the door from the inside. Ideally, the door
should open outwards so that it cannot be barricaded, and contain a non-
breakable glass panel to enable colleagues to see what is happening if a
situation develops when a good deal of noise can be heard by others in the
building.
More general organizational policies may include keeping the addresses
of staff secret, wearing safe clothing and jewelry (i.e. not earrings in pierced
ears), and letting colleagues know your schedule if home visits are planned
to potentially dangerous individuals. It may be appropriate to arrange to
phone a colleague at a prearranged time after a visit to confirm that all went
well. Of course, when a serious risk is perceived, home visits should not be
made alone.
Within a psychiatric unit, aggression and violence is reduced by a well-
functioning unit, with good communications, good handover between
different shifts, sufficient experienced staff on each shift, and clear con-
tingency plans for violent incidents involving a consistent but flexible
approach. Staff must try to allocate sufficient time for individual sessions
with each patient and manage each session with clear explanation and
courtesy. It is often advisable to tell patients how long you have got to see
them, so that resentment may be limited if a longer session is expected.
While various studies have examined the effects of variables such as time
of day, architectural blind spots, overcrowding and staffing levels on
violence in psychiatric inpatient units, no clear patterns have been demon-
strated. However, such factors appear to be of major importance in some
incidents. Other important variables related to ward violence are the
personalities and behaviour of staff and the nature of the social environ-
ment. Several studies have found that, in many short-stay adult psychiatric
units, patients are unoccupied for long periods, in which there is minimal
interaction with staff or with fellow patients, and there is some evidence to
support the view that there can be a link between levels of violence and lack
of planned activities.
Staff should be trained in techniques for physical restraint, and, if a
person is subject to compulsory treatment, have policies for the use of
sedation and seclusion. The latter may involve putting the person alone in a
'seclusion room' for periods from a few minutes to several hours. However,
the degree of seclusion may vary; the patient may be alone with the door
closed (but under regular observation through a glass panel), the door may
be ajar with a nurse just outside, or a nurse may be with the patient
continually or intermittently. Such arrangements provide some practical
and psychological containment in an environment that has reduced stimuli,
and can assist in the resolution of an aggressive episode. When physical
restraint is required, it is important that the aggressor is faced with
considerably superior force, as a more equal balance of force will lead to a
higher risk of injury to both staff and patient.
Various treatment programmes have been tried in the management of
aggressive individuals in both clinical and non-clinical populations, such as
inpatient 'therapeutic community' regimes, individual psychological man-
agement sessions including behavioural approaches, and drug treatments
(Suedfeld & Landon, 1978). But methodological difficulties allow only
broad conclusions to emerge from published studies: psychological man-
agement needs to be in a framework of clear rules and support; medication
can be a useful adjunct; therapeutic communities may be helpful for some
individuals; and improvement tends to occur with the passage of time.
Several studies have reported the effects of inpatient management in
specialized units (with emphasis on control, confrontation and peer group
support), outpatient group psychotherapy and a therapeutic community
(Vaillant, 1975; Carney, 1977). Although McCord (1983) reported an
association between reduced re-offending in young individuals and expo-
sure to a therapeutic community, empirical evidence of benefit for a
therapeutic prison regime was not found in another study (Gunn et al,
1978). There is evidence that various forms of psychological management
for aggression are more likely to be associated with improvement if the
individual is also depressed or has reasonable social skills (Copas &
Whiteley, 1976; Woody et al., 1985). Behavioural approaches to aggression
have included 'token economy' regimes (i.e. inpatient or custodial settings
in which certain behaviours are rewarded), social skills training, cognitive
therapy, and self-control procedures (Cavior & Schmidt, 1978; Templeman
& Wollersheim, 1979; Frederiksen & Rainwater, 1981; Moyes, Tennent &
Bedford, 1985).
Several studies have reported the effects of social skills training on self-
referred individuals with problems due to aggression. This approach is an
example of 'indirect' control measures, which aim to teach non-aggressive
alternative behaviour, and contrast with 'direct' methods, which try to
change the pattern of reinforcement of aggression. The procedures of social
skills training include identifying target behaviour and associated circum-
stances, teaching alternative responses and arranging practice. Desired
behaviour can be modelled by the therapist, new behaviour can be
rehearsed using role play, video recording and corrective feedback, and
subsequent encouragement can provide reinforcement. Such an approach
has been applied to various non-clinical and clinical populations, and
Henderson (1989) has concluded that some positive effects and generaliza-
tion have been reported, but that such results have been limited to verbal
aggression. Henderson (1989) has also described a social skills training
programme for male prisoners in a UK prison that involved group
meetings. Twelve prisoners with convictions for violence completed the
programme, which was aimed at increasing skills in the following three
areas: assertion (the ability to express negative feelings without aggression);
self-control (the ability to identify and control anger, and to prevent
irritability or loss of temper); and reduction of social anxiety, which may
provoke aggression when the person feels threatened or anxious. Six to

eight trainees met with two co-trainers for sessions that involved instruc-
tion, modelling, videotaped behavioural rehearsal and group discussion.
The results were assessed by ratings and self-report questionnaires, and, at
follow-up, improvements were found in assertion, social anxiety, self-
confidence and global measures of social skills. Although there was little
change in aggression for the group as a whole, individual prisoners
appeared to respond and it was considered that this programme may be 'an
effective intervention method for violent offenders'. Blackburn (1989)
considered that there are 'still insufficient empirical data to justify any
single approach to the treatment of aggressive individuals who are deviant
in personality. Examples can be found of successful outcomes for most
forms of psychological treatment, but the evidence suggests that none is
consistently successful'.
Chronic somatization
'Somatization' is a process in which 'there is an inappropriate focus on
physical symptoms and psychosocial problems are denied' (Bass & Benja-
min, 1993). Often this is associated with 'primary gain', when the patient is
relieved of responsibilities as he/she is 'ill', or 'secondary gain', such as a
gratifying response of caring and concern from others.
PD is often implicated as a causal factor. For instance, obsessive-
compulsive PD is associated with a tendency to worry (involving a reduced
threshold for somatic preoccupation), while those with dependent or
histrionic PDs obtain secondary gain from reassurance. Also, those with
narcissistic PD receive reinforcing gratification from medical attention.
Bass & Benjamin (1993) have outlined general management strategies for
somatization including the avoidance of a 'dualistic approach', which sees
patients as having physical or mental disorders but never both. Creed &
Guthrie (1993) have stressed the need for the clinician to develop a good
rapport with 'somatizing' patients, involving questions, empathic response
and comments that link somatic symptoms and psychological experiences.
Often patients can be persuaded to collaborate with psychological manage-
ment that has the stated aim of helping the patient to cope with the
symptoms. This may prevent an endless pursuit of a somatic cause by
repeated investigations. Sometimes a cognitive or cognitive-behavioural
approach is needed to help patients address dysfunctional beliefs, such as
the need to avoid all exercise. It is important to acknowledge that the
patient's experience of symptoms is 'real', and to state that a physical illness

can never be completely ruled out even if tests are negative, and that this will
be kept under review. Patients can then be told that people with chronic
symptoms can often be helped to feel better and cope with their disability
more effectively.
Substance abuse
A variety of programmes for alcohol abuse and other substance abuse
disorders can involve various formats: inpatient units, therapeutic commu-
nities, outpatient clinics and self-help groups. Many such individuals have
PD, in particular with antisocial and borderline features, and this needs to
be considered when the aims and outcomes of a service are being evaluated.
In a study of 20 social and personality variables in the treatment of heroin
addiction, the counsellor-patient relationship was related to final outcome,
and those who had a single counsellor did significantly better than their
peers who were transferred from one counsellor to another (Cohen et al.,
1980). Individual counselling was ranked the most important of several
treatment components by alcoholic individuals who accepted a 2-year
outpatient programme (Ojehagen & Berglund, 1986ft).
Sexual offenders
Sexual violence involves a victim who is unwilling or unable to consent.
Various studies of treatment interventions have been mainly concerned
with rape, offences involving children, and genital exhibitionism to adult
females (Marshall & Barbaree, 1989). Rapists have been subdivided in
relation to different forms of sexual gratification, displaced aggression, the
degree of sadism, and impulsive or opportunistic acts. The latter character-
istic is often associated with a history of other offences, and features of
antisocial PD (Perkins, 1991).
The first stage of management is a detailed assessment including the
individual's sexual preferences. The penile plethysmograph (PPG) can
indicate male arousal to a variety of stimuli, perhaps presented as projec-
tion slides, and can indicate inappropriate denial of specific sexual interests
or fantasy. Various behavioural techniques have been used in attempts to
modify sexual motivation, aimed at reducing deviant sexual interest,
increasing social skills and addressing relevant cognitive processes.
Deviant sexual interests can be the target of an aversive stimulus such as
an electric shock to the leg or an unpleasant smell (e.g. strong-smelling
salts), which are paired with deviant fantasies or stimuli. Related tech-
niques include masturbatory reconditioning, where masturbation is
encouraged when paired with non-deviant stimuli or fantasies, and sensiti-
zation, when the patient is encouraged to fantasize a sequence of images of
offending behaviour with unpleasant consequences. These techniques can
sometimes be extended from the consulting room to real life; if a sequence
of triggering stimuli and offending behaviour is identified, the patient can
be encouraged to interrupt this, for example, by the use of strong-smelling
salts (Perkins, 1991).
Social skills training can be provided in a group format, as for individuals
with aggressive behaviour. A problem-solving approach is indicated, which
may examine triggering situations and alternative behaviours, and involve
modelling, role play, rehearsal and feedback.
Many patients require basic education about sexual behaviour and
further understanding of their motivations for offending (together with
behavioural antecedents and the consequences), if a logical strategy for the
patient to avoid further offences is to be devised. Many subjects deny
aspects of their offences and four main types of offender have been
identified: 'rationalizers', who admit offences but deny harm is caused to
their victims (in particular involving homosexual offences against boys);
'externalizers', who blame others including their victims, who are typically
young women; 'internalizers', who admit the offence and the effects but
claim a temporary aberration of behaviour; and 'deniers', who present with
absolute denial (Kennedy & Grubin, 1992). It is helpful to encourage
subjects to reduce the level of denial so that behavioural approaches can be
more readily applied. Also, personal skills can be encouraged, involving
obtaining a job, reducing alcohol intake and organizing leisure activities.
Various evaluations of behavioural treatment for sex offenders have
demonstrated some positive effects within institutions (Burchard & Wood,
1982; Rutter & Giller, 1983) but there is little evidence of specific effects on
the reduction in subsequent offending. However, Davidson (1984), in a 2-5
year follow-up of individuals who had experienced a prison-based treat-
ment programme of psychotherapy and behavioural techniques, found
evidence that the programme was of benefit for those whose offences
involved children, although the recidivism rates for men who had sexually
attacked adults were not different from an untreated group. But recidivism
rates are unreliable data due to a high level of under-reporting (Furby,
Weinrott & Blackshaw, 1989), and the long-term outcome of institutional
programmes requires further research. While some subjects have been
shown to benefit within institutions, such treatment programmes may be a

necessary but not sufficient intervention to reduce further offending. The
provision of further outpatient and non-hospital-based individual and
group support may be an important development for the long-term
management of this group of offenders, as continuing intervention may be
able to build upon the gains in a prison or hospital-based programme.
8
Group psychotherapies
J.H. DOWSON
Overview
Historical development
Since Moreno introduced the term 'group psychotherapy' in 1931, many
theoretical frameworks and techniques have emerged for use in a wide
variety of settings. These include the short-stay psychiatric unit, the
outpatient clinic, the medium- and long-stay units of psychiatric hospitals,
specialist psychiatric units for alcoholism or forensic problems, and self-
help organizations (Ryle, 1976). Regular group meetings may also involve
non-patients, such as trainees in those professions where increased self-
awareness and an ability to communciate are particularly important.
Research in this field has been limited and, in general, theoretical
considerations have determined clinical practice (Malan, 1973). Three
approaches have been derived from psychoanalytic theory; the first exam-
ines the relationship of each individual patient to the group or therapist, the
second is concerned with the relationship of the group as a whole to the
therapist, while the third approach emphasizes treatment by the group
itself, focussing on general aspects of the group such as themes, communi-
cation patterns, values and cohesiveness. In this last approach, the therapist
ignores, as far as possible, his/her own relationship with individual group
members or with the group as a whole. The first two strategies have been
termed psychoanalysis 'in' groups and 'of groups (Wolf & Schwartz, 1962),
while the latter, the so-called group-analytic method (Foulkes & Anthony,
1965), encourages the therapeutic ingredients to be provided 'by' the group
(de Mare & Kreeger, 1974).
Despite the differences between these various approaches derived from
psychoanalysis, they share the underlying assumption that therapy largely
consists of exploring unconscious mental processes by means of interpre-
294
Overview 295
tations directed towards resistance and transferences. However, although
these approaches may be useful with carefully selected groups of out-
patients, other strategies are more commonly found in most types of group
psychotherapy that are routinely practised within a psychiatric service.
These involve the provision of more focus and instruction than is found in
groups that mainly aim to explore mental processes.
A major difficulty that is encountered by investigators of group psy-
chotherapy is that, like individual psychotherapy, the procedure is very
variable even if the same approach is used. In addition, the treatment goals
and criteria for the definition of improvement are varied. Beneficial effects
may be judged on the basis of changes observed in the group, disappearance
of symptoms, improved subjective well-being, the views of the conductor,
the opinions of friends or relatives of the patients, increased self-knowl-
edge, increased competence at work, or changes in social behaviour (Yalom
et aL, 1977). However, there is often a low correlation between various
outcome measures. Apart from the methodological problems in measuring
outcome, the results are often affected by spontaneous recovery and many
non-treatment variables. Also, if a particular psychotherapeutic method is
being investigated, many non-technique variables are relevant, such as the
'helping relationship' with the therapist, although this may be of less
importance in groups compared with individual psychotherapy (Gurman
& Gustafson, 1976). As each group is unique, control data must be
interpreted with caution.
Group psychotherapy and PD

Various forms of group psychotherapy have been advocated for individuals
with borderline PD, antisocial PD and for certain offender groups, as well
as for some other individuals with a PD diagnosis, who present with
secondary symptoms such as anxiety.
The literature contains a general consensus among clinicians that
individuals with borderline PD can benefit from a treatment regime that
includes relatively structured group methods (Blum & Marziali, 1988),
involving a format in which tasks of the group are defined, rather than just
telling each patient that his/her task involves self-disclosure, listening and
interacting with others.
Group treatment for patients who have all received the diagnosis of
borderline PD is often combined with individual psychotherapeutic ses-
sions and, with such a regime, there has been debate as to whether the
therapist for individual sessions should also be involved in the group
296 Group psychotherapies
treatment. Not surprisingly, there is not a clear consensus on this issue, as

individuals with borderline PD (and therapists) are heterogeneous. Also,
management decisions have to take into account many other aspects apart
from patient variables, such as the availability of resources and the level of
experience of staff in group treatment methods.
With regard to other PDs, a group of patients with avoidant PD were
reported to be helped by just five sessions involving either social skills
training or group sessions (Stravynski et al., 1989), and an intensive 4-day
group treatment for avoidant PD was associated with clinically-significant
improvements that were evident at 1-year follow-up (Renneberg et al.,
1990). Also, various programmes have been devised for offender groups
that can be assumed to contain many individuals with antisocial PD. For
instance, a London prison provided a regime for rapists, which involved 50
group sessions with a structured educational content, focussing on such
issues as attitudes to victims, effects on victims, how to avoid situations in
which offences may occur, adaptive relationships with women, and control
of alcohol abuse. Also, an 18-month therapeutic community experience has
been attempted for rapists within a prison system, although such an
approach does not appear to be indicated for those who refuse to take any
responsibility for their actions. Dolan & Coid (1993) concluded that the
literature on therapeutic community approaches to antisocial behaviour
have shown 'the most promising results', although there is a lack of
controlled studies.
Finally, patients who completed their treatment in a day hospital
substance abuse programme and continued to attend self-help groups, had
relatively low rates of substance abuse during follow-up (McKay et ai,
1994).
Despite the variety of selection criteria, formats, goals and strategies,
certain principles for group psychotherapy can be identified in relation to
therapeutic ingredients, group-specific phenomena, the organization of
groups and the therapist's (i.e. conductor's) role. These will be considered
in relation to groups that are relatively unstructured, but are also relevant
to groups that have defined procedures and tasks.
Definition and therapeutic ingredients

Group psychotherapy encompasses a variety of forms and goals, which
include relief of psychological symptoms, modification of social or family
relationships, and a permanent change in attitudes or other aspects of the
Definition and therapeutic ingredients 297
relatively stable organization of an individual's mental activity and behav-
iour, which comprise personality.
A group of individuals, who are appropriately called patients within the
setting of a psychiatric service, meet regularly with one or more trained
professionals to interact by verbal interchange and other communicative
behaviour (Frank, 1979).
All the phenomena that are found in individual psychotherapy may also
be present in group psychotherapy although some, such as the importance
of the relationship between patient and therapist, are attenuated, while
others, such as the provision of new interpersonal situations from which the
patient may learn, receive an extra dimension in groups.
As with individual psychotherapy there are two main approaches, which
can coexist. The 'supportive' approach implies that the therapist accepts a
degree of dependency from the patients and provides an active input to their
relationships, while the approach derived from psychoanalytic theory is
directed towards an increasing awareness and understanding of uncon-
scious mental activity, with a focus on the past.
Group psychotherapy can provide new learning situations that are
derived from a much wider range of interpersonal interactions than is
possible in individual psychotherapy. Insight may be obtained by observing
'unconsciously' motivated behaviour in other group members, while
increased self-awareness is produced by the patient finding out, by 'feed-
back', how he/she is perceived by others. Each patient is usually encouraged
to display his/her own psychopathology by ensuring that he/she feels 'safe'
enough to do so, and he/she may have the opportunity to develop and test
new patterns of social interaction. Insight can be variable in relation to a
beneficial instance of new learning, and may be minimal or non-existent.
Thus, a patient with limited capacity for insight or abstract thought may
increase his/her self-confidence by being able to advise or comfort other
patients, without being aware that a poor opinion of his/her value to others
was one of the main problems.
Group psychotherapy can also provide therapeutic events that are
specific to this form of treatment. An atmosphere of trust, openness,
concern, tolerance, and support, with controlled negative feedback, has
been termed 'group cohesiveness', and may be therapeutic in itself. It is also
well known that any group of individuals can exert powerful pressure to
conform to the attitudes of the majority, and this force may be harnessed to
the correction of maladaptive behaviour. The experience of universality, in
which the patient realizes that other people have problems, sometimes
similar to his/her own, has often been cited by patients to be therapeutic,

and the process of sharing problems encourages group members to become
less self-absorbed. The opportunity to be altruistic may also reduce self-
absorption and increase self-esteem. In some groups, re-enactment of
family relationships may take place if other group members resemble
various members of the patient's family, and this process may assist the
development of insight. For some patients, such as those with social
phobias, groups provide the feared situation and a setting in which coping
strategies can be developed, while a collective approach to problem solving
can sometimes lead to the optimum solution for certain problems.
Group psychotherapy usually provides various additional indirect bene-
fits, the importance of which should not be underestimated. If the group
meetings take place in a psychiatric unit, the interest and morale of the staff
may be increased, and the assessment and monitoring of the patient is
improved. For inpatients, the relief of boredom associated with this or any
other activity may also be of considerable, if non-specific, value.
Group-specific phenomena
Psychotherapy in groups incorporates various communication patterns
involving both verbal and non-verbal behaviour.
The possible benefits of so-called group pressures have already been
considered; groups have the capacity to define, by consensus, acceptable
attitudes or behaviour and then to exert pressure to conform to these
standards. Behaviour that is generally acceptable may be rewarded by
positive regard, while deviance can be punished by criticism, hostility or by
being ignored. But at times, the overt consensus attitude may be more
extreme than the most widely-held viewpoint. Another property of groups
is their ability to define what appears to be reality to the majority; for
example, the staff may be considered to be uncaring, or it may be assumed
that all patients have similar problems. Such assumptions may be far from
accurate. Resistance may involve the collaboration of group members to
prevent the consideration of reality, perhaps by means of silence or
changing the subject by directing attention to another group member.
Although at such times the patients may not be fully aware of what is
happening, on one occasion every member of a group made a deliberate
decision before the meeting that no one would utter a word! A common
example of an unconscious collaborative defence is that of scapegoating,
when one individual receives hostile feelings that have been displaced by
Organization 299
several group members from their appropriate target; this often occurs
when a new member enters a group.
When emotion is expressed in a group there is a tendency for it to
precipitate the experience of a similar feeling in several other group
members. Thus, a sharing of expressed emotion often occurs; one angry
complaint about the staff may initiate a chorus, while a weeping patient
may elicit several complaints of depressed mood.
Another group process has been called the 'chain phenomenon' when
several members contribute to a group theme, while an examination of the
pattern of interaction in a series of meetings invariably demonstrates a
tendency for the form and content of the meetings to undergo constant
change. Irregular cycles are often seen in which, for example, a meeting (or
series of meetings) characterized by the patients being constructive, caring
and active will be followed by withdrawal, hostility and hopelessness,
before the original form eventually returns.
Finally, a consideration of group-specific phenomena would not be
complete without mention of interactions that take place at times other
than the group meetings, but which would not have occurred if the patients
had not been members of the same group.
Organization
Groups may be 'closed' (and time limited) when the same individuals meet
for a set number of occasions, or 'open', when individuals join and leave the
group at different times.
The goals of group psychotherapy vary considerably and, although
many patients in a psychiatric service may be included in a group at some
stage, some are not suitable for this form of treatment and certain principles
of selection should be observed. Severely psychotic patients and those with
marked paranoid or schizoid PD traits should generally be excluded; the
paranoid patient often evokes hostility, which reinforces his/her pathologi-
cal ideas, while the severely schizoid person will usually find a group too
threatening. Also, excessively narcissistic individuals with a minimal capa-
city to take note of others should not usually be selected. Sociopathic
individuals, addicts of both drugs and alcohol and those whose problems
centre on sexual deviation must be included only after careful consideration
in a group of individuals with a range of psychiatric problems, and, in
general, there should not be more than one patient with any of these
problems in each group unless a group is selected on the basis of a
diagnostic category such as substance abuse, borderline PD or antisocial

PD (Bloch, 1979).
Certain characteristics of a patient imply suitability for relatively
unstructured group psychotherapy irrespective of the presenting psychi-
atric syndrome, and such a patient should be judged to have sufficient
psychological reserves to tolerate the anxiety-provoking situations that
invariably arise in such groups. The patient who is likely to derive
significant benefit from an unstructured group will usually have had at least
one 'meaningful' relationship, have the capacity to participate in discussion
and to use abstract concepts, be committed to active cooperation, be able to
experience and express emotion, and have a clearly defined area of conflict
(Marmor, 1979); however, these qualities are not essential for a patient to
be included in some types of group therapy. The group conductor should
avoid selecting a patient who would be conspicuously different from the
other group members on such dimensions as age, social class, intelligence or
verbal ability, and the inclusion of too many patients who are likely to say
very little must be avoided. Patients whose major difficulty is that they
become emotionally dependent on others to an excessive extent often do
better in groups than in individual psychotherapy, as a pathological and
disabling dependency upon one individual is usually avoided in a group
setting. The group conductor should interview each prospective member
and consider the current state of the group, the nature of the patient's
problems and personality, and the conductor's attitudes towards the
patient; if the conductor feels considerable hostility towards an individual,
this may preclude selection. The patient can be asked to provide a written
account of his/her history and present problems, which can indicate
motivation as well as providing information (Bloch, 1979). If a patient is
selected, the conductor must then consider the best time for entry to an
open group, as joining when patients are particularly hostile or apathetic
may not be appropriate.
After a patient has been selected, he/she can be seen individually by the
conductor or other staff member for a preparatory session (Yalom et aL,
1967). Certain rules are usually made: patients should not meet socially
outside the hospital or clinic, any significant change in life situation must be
reported to the group, notice must be given if attendance is not possible,
and confidentiality must be maintained. If necessary it should be made clear
that repeated disruptive behaviour other than verbal interchange may lead
to exclusion. For unstructured groups, the patient is told that the meetings
have no set agenda and that the tasks will be to talk about his/her problems,
feelings and thoughts, and to interact with other group members. It may be
Organization 301
appropriate to ask that the patient makes a commitment to attend a
minimum number of meetings.
The optimum number of patients is generally agreed to be about eight,
with a range of three to 12 (de Mare & Kreeger, 1974). Each meeting usually
lasts for 1 hour and takes place one to three times weekly, although
Linehan's regime for patients with severe borderline PD includes a weekly
group of 2\ hours (Linehan, 1987). Sessions should be as evenly spaced as
possible. Group members sit in a circle so that non-verbal communication
is apparent.
The group conductor should have some experience and must be assisted
by at least one other staff member. During each meeting one individual can
be clearly designated as the conductor, but two, or even three co-
conductors may alternate this responsibility. Not more than three staff
members should accompany a group of about eight patients. One, prefer-
ably two, staff should be regular group members, while the third may be a
student or trainee who can attend for a limited number of meetings.
Although two or three staff members may be sufficiently experienced to
share the task of conducting the group, one therapist may wish to retain
continual responsibility as the conductor; in this situation other staff
members can act as 'participant observers' who are relatively inactive. The
staff can meet for about 15 minutes after each session when the conductor
can compare his/her impressions with those of colleagues. A written
summary or report may then be prepared, which may be given to the
patients at the next meeting (Yalom, Brown & Bloch, 1975). For many
group members, group psychotherapy is part of a varied treatment
programme and the conductor must ensure that good communications
exist between staff in the group and other staff members in the various
professional disciplines.
For an unstructured group, each patient is told that his/her task in the
group involves self-disclosure, listening and interacting with others. How-
ever, for the many groups that have more structure, group members are
provided with defined tasks. For unstructured groups, the emphasis is on
verbal communication by which each patient reveals various aspects of
psychological functioning, such as attitudes, patterns of social interaction
and psychological defences. Transferences to several group members may
develop, involving unrealistic feelings and attitudes, which indicate uncon-
scious mental activity.
Awareness of non-verbal behaviour is important. When a patient is
speaking, the other group members will be engaged in various forms of
silent participation: emotions may be shared, self-examination may be
taking place, and the silent patient may be learning from events that involve
others. Overt non-verbal activity such as facial expressions, fidgeting,
moving restlessly in the chair, tapping the foot, or playing with an ashtray
or cigarette, should be noted and may be informative. More rarely,
unacceptable and dramatic behaviour may occur: objects are thrown,
physical violence or attempts at self-harm take place, or a patient rushes out
of the meeting.
Goals of group psychotherapy

Although the patient and therapist may not be able to agree the aims of
treatment, the conductor should formulate realistic goals for each group
member during the selection process, bearing in mind the likely capacity for
change and the time that will be needed to achieve the desired effects. It is a
common mistake to underestimate the time required to produce clinically
important changes in a patient's personality structure. A minimum of 6-12
months is usually necessary to encourage even modest changes in the
personality traits that comprise the major syndromes of PD.
Sometimes the treatment goal does not involve change but an attempt to
support the patient who cannot manage his/her life without some help.
Support may be provided for a limited time during a resolving crisis, but
group attendance may also play its part in the management of more chronic
problems. In a crisis, attention must be paid to the precipitating events,
which may have involved loss, change, interpersonal difficulties or a
conflict between alternatives.
Group treatment may involve discussion of the cause or effects of
psychological disorder; thus, attention may be directed towards a lack of
trust in others that has led to anxiety and unhappiness, while a patient who
has suffered repeated episodes of psychosis may be helped to adjust to
residual deficits and vulnerability to stress.
Treatment goals can involve the relief of symptoms that include what
Frank (1974) has called 'demoralization'. This involves a feeling of
isolation, shame, and a loss of long-term goals. Group psychotherapy may
be used to enhance a patient's capacity to relate to others, so that he/she
may be more able to achieve close personal relationships, perhaps by being
less assertive or histrionic, less uncomfortable in company, less mistrustful
of others or more independent. Another goal may be to modify various
aspects of personality, such as low self-esteem, lack of purpose and self-
confidence, an inability to express emotion, or a tendency to behave
impulsively with poor control of emotional expression.
At times, psychological defence mechanisms should be strengthened,
perhaps by encouraging avoidance of certain problems but, if defences are
maladaptive, another goal may be to assist the patient to discard these or to
substitute less harmful manoeuvres. The dismantling of defences usually
involves the development of insight; for example a patient may be encour-
aged to accept that a particular problem is insoluble, or that a decision has
been avoided.
Another important aim of group psychotherapy can be the regular
monitoring and assessment of the psychological state and life situation of
each group member. Group psychotherapy provides an opportunity for
this to be achieved in a way that is particularly valuable if the availability of
staff for individual interviews is limited. Group psychotherapy can also
encourage staff interaction with patients, improve staff morale and foster
cooperation across professional boundaries.
The group conductor's role

The conductor bears ultimate responsibility for the group and undertakes
the various administrative tasks that have been described. The conductor
must also open and close each meeting, monitor the proceedings, and
intervene appropriately. The relationship of each individual patient to the
group must be borne in mind and group processes should be identified.
Both verbal and non-verbal behaviour is noted and patterns are sought.
The conductor should also be aware of his/her own emotional reactions.
To achieve the treatment goals for the individual members, and to foster
the therapeutic activity of the group, the conductor can use a wide range of
interventions. The first priority must always be to do no harm, and a
vulnerable patient who cannot cope with certain stressful situations has to
be actively protected and, if necessary, withdrawn from the group. The
conductor must set a standard of behaviour involving tolerance, concern,
perseverance and, above all, calm in difficult situations. He/she should
encourage an atmosphere of openness, trust and hope that is 'safe' for self-
disclosure and will, by a variety of methods, further the specific treatment
goals that have been formulated.
The conductor's contributions are both verbal and non-verbal. Various
comments and questions can facilitate a climate of communication,
concern, mutual support and the sharing of experience or emotion.
Interventions may be relatively non-directive, or be aimed specifically at
one group member or a particular topic. At times the conductor should
encourage the sharing and expression of emotions, perhaps by drawing
attention to depression or anger in the meeting, and by inviting others to

share their own feelings or experiences. However, it is not a good policy
constantly to encourage the expression of anger or depression. The
opinions of the conductor and other staff are often an important class of
intervention, as these may help to counteract pathological opinions and
behaviour patterns that are shared by most of the group members. Other
types of intervention include clarification, the promotion of problem-
solving behaviour, confrontation, directives and interpretations. Clarifica-
tion of a situation may be the first step in solving a problem; Goldfried &
Goldfried (1975) have described subsequent stages in this process: the
identification of methods for the solution with an examination of their
consequences, the choice of one of these possible methods, and the planning
of the various steps for its performance. Confrontation refers to the
technique of repeatedly bringing the group face to face with a situation,
while directives, although usually kept to a minimum, can be essential.
Interpretations are the final major category of verbal input and refer to
comments designed to produce insight.
Non-verbal activity includes silent participation involving attentive
listening. Gestures, facial expression and movements will convey varying
degrees of information about the conductor's underlying thoughts, while
another type of non-verbal intervention is that of physical activity.
Although the need for this should be rare, the conductor and other staff
may occasionally have to stop a fight, or prevent an attempt at self-injury.
A flexible approach to the conductor's role is essential in this type of
group as changes of strategy are often required. Although the conductor
may have no alternative but to encourage a leader-centred communication
pattern, in which he/she interacts with several group members in turn, there
should be periods when most interactions are between group members.
Group communication is encouraged by such techniques as asking a
question of the group rather than of an individual, inviting the expression
of views from previously silent members and focussing on themes that are
of general relevance. In a series of meetings, there should be a balance
regarding the content of topics discussed; although events in the remote
past and a quest for explanations may be important, the 'here and now'
should predominate in most routine practice settings, both inside and
outside the group. Also, the future should be considered.
At any moment, the conductor has a choice from a range of overall
strategies for the conduct of the group. At one extreme the conductor is
authoritative, active, and freely gives his/her opinions and reactions; his/
her main aim is to provide a setting in which patients feel minimal anxiety,
where conflicts are diffused or avoided, and where sharing of problems,
mutual support and group cohesion are encouraged. The focus is on the
problems and symptoms of the here and now. At the other end of the
spectrum, the conductor is minimally active and non-directive; no attempt
is made to avoid anxiety and hostility, even though this adversely affects
group cohesion. With this approach, emotional expression is encouraged,
and the focus is on relationships, the awareness of the effects of behaviour
on others, defences, and group or individual transferences.
Most unstructured group meetings in a psychiatric service involve an
amalgam of strategies, and during each meeting the conductor's tactics may
undergo several changes in response to events in the group. However, in
general, strategy is weighted towards the former end of the spectrum in
routine practice settings, so that group psychotherapy should provide a
situation in which patients can usually feel safe and do not often experience
a high level of anxiety. The members should generally feel able to join in,
and have trust in the staff. Unfortunately, some staff members are not
suited to be group conductors in this setting by virtue of their personality or
theoretical bias; Yalom (1975) has described the aggressive, confronting,
excessively self-revealing, controlling and charismatic group conductor,
and it is likely that this type of conducting can often be harmful unless the
group is carefully selected. In general, the conductor must strive to ensure
that group meetings are usually characterized by attentive concern, so that
the members feel relaxed and able to contribute. Of course, emotions are
expressed and tensions appear, but these can be tolerated if the background
atmosphere is one of trust. The conductor must act naturally and should
generally be prepared to voice his/her own opinions or to demonstrate
personal concern when this is appropriate. A conductor who displays
nothing of his/her views or reactions encourages the development of
marked hostility which may not be adequately resolved unless the group
members are highly selected. However, the conductor must limit the degree
to which his/her own personality becomes apparent and some formality is
required, in keeping with the professional nature of the conductor's role.
The conductor must constantly review the degree of his/her activity,
whatever the overall strategy. This must be considered every few minutes,
and the staff should not be so active throughout a meeting that the patient
members of the group are seldom given a chance to interact. Often it is
appropriate for the conductor to say very little during the first part of the
meeting, but if interactions do not develop, or break down, a more active
and directive approach is usually necessary; in these circumstances a
subsequent attempt should be made to provide a period when the staff is
minimally active.
Resistance has been denned as all the obstacles to the aims and
procedures of treatment that arise from within the patient, and may include
both conscious and unconscious mental activity (Sandier, Holder & Dare,
19706). The latter may involve the various defence mechanisms and a
'negative therapeutic reaction' in which improvement, or the expression by
the therapist of an opinion that the patient has improved, is followed by a
worsening of the patient's condition. Resistance can involve a varying
proportion of group members; examples include persistently coming late,
refusing to discuss a particular topic, ignoring staff comments, meeting
socially, the formation of subgroups whose interactions persistently
exclude other group members, scapegoating, showing hostility to new
members, and (in unstructured groups) taking turns to speak or preselect-
ing topics for discussion. Although psychological defence mechanisms may
fulfil a useful function, and should not always be considered as resistance,
they can lead to the avoidance or distortion of reality and impede the
treatment process. Another phenomenon that can result from the use of
defences is that of'splitting', when the ambivalence of feelings to a person
or situation is not recognized and the opposing emotions or attitudes are
separated; one or more parts may be masked or distorted by defence
mechanisms, for example, people are considered to be 'all-good' or 'all-
bad'. Transference and counter-transference also involve unconscious
mental activity which, if unresolved, can interfere with treatment; for
example, one or several group members may develop unrealistic expec-
tations of treatment, or excessive emotional dependence on the conductor
(Sandier, Dare & Holder, 1970).
Several techniques are available for overcoming resistance either in the
group as a whole or in individual members. The conductor may ignore it,
encourage a safe, supportive group atmosphere, mobilize the group
members to share their experiences in relation to the topic under discussion,
and wait. But if aspects of a particular situation are persistently avoided or
distorted, the resistance can be side-stepped by asking a direct question
about the topic that is being avoided, if this is known. If this procedure is
not effective, the conductor may draw the attention of the group to the
resistance itself; for example, if a long silence is considered to be a
resistance, the group may be encouraged to consider their feelings about the
silence and the possible reasons why this situation has occurred. A fourth
technique for overcoming resistance is to provide an interpretation, if
unconscious mechanisms are involved. For example, a suggestion was
made to a husband who had been deserted by his wife, and who was
depressed and excessively protective to his children, that he may have mixed
feelings towards his children, as many people in his situation resent such
responsibility. This helped the patient to recognize that he felt intensely
resentful towards his children as well as towards his wife. But sometimes, all
the conductor's efforts to overcome resistance will be in vain, and if the
resistance is in the form of a long silence, this will continue. At this stage the
conductor has no option but to accept that, for the moment, he/she has
done all that is possible. The conductor does not bear all the responsibility
for the efficacy of the group, as this is shared to some extent between all its
members. Thus, in the case of a prolonged silence, the ball can be repeatedly
put back into the group's court by the conductor pointing out that he/she
can only do so much. Sometimes the anxiety provoked by an extended
silence is a useful catalyst to increased effort and self-reliance by the group
members, but, in general, the conductor should try and help the group to
avoid this situation.
The conductor may open each group meeting formally, and introduce
any new member. He/she may then state the reasons for any absences and, if
these are unknown, should encourage the group to consider if any action is
appropriate; in a hospital setting a patient may volunteer to look for a
missing individual. At this point in an unstructured group, the conductor
can wait for a patient to start the proceedings, as part of the overall policy
for the staff to be minimally active for at least part of the meeting. This
allows the group to assume some responsibility for selecting the topics to be
considered. However, if an initial silence continues for a minute or more,
interactions may be encouraged by non-directive interventions which
prompt the group as a whole.
During the final 15 minutes, the conductor should be aware that the end
of the meeting is near. He/she may then decide to prompt previously silent
members by directly asking them for their comments and, in the final few
minutes, the patients can be invited in turn to give their reactions to the
meeting or to each other. The conductor should try and ensure that the
scheduled time for ending the meeting does not arrive during an unresolved
period of emotionality but, if this occurs, the meeting may have to be
extended. This should be avoided if possible, and an alternative technique
for dealing with a patient who is very distressed or angry at the end of the
meeting is to say that, although it is time to finish, a staff member will
remain with the patient. This should also be a rare event, and patients
should not be encouraged to expect individual attention immediately after
a meeting.
The inexperienced conductor oftenfindsit difficult to decide on conclud-
ing remarks. At times, it may be appropriate to summarize the events of the
meeting or end on a hopeful and encouraging note. But the conductor must
not feel that it is his/her duty to do either, and when the mood is one of
hopelessness or resentment, or when the meeting ends in the middle of the
consideration of a problem, a statement can be made which merely clarifies
the situation; for example 'one of the main problems seems to be that
everyone has given up hope of any improvement', or 'although we have to
finish now, we must come back to this problem next time'. On other
occasions the conductor need not say anything other than to formally close
the meeting by a comment such as: 'I am sorry to interrupt you but we have
come to the end of our time'. This may be quite difficult, as he/she may come
under pressure to say something 'clever' or reassuring, or to prolong the
meeting. If group members ignore a concluding remark, the conductor may
need to repeat the closing statement. If necessary he/she should interrupt
the group by standing up with other staff members.
Emotionality
Although mental events have both a cognitive and emotional component, it
can be useful to isolate the emotional content of each group meeting for
consideration. Anger, depression and anxiety are of particular relevance,
but disgust, positive attraction, happiness and elation should also be noted.
If personality change is a goal of treatment for some group members, a
degree of emotional arousal may be facilitative and, at times, essential,
while varying degrees of emotionality are an inevitable part of all types of
group psychotherapy.
Anger may occur in many disguises, such as a show of boredom, sarcasm,
or the excessive politeness resulting from the use of the defence mechanism
of 'reaction formation', in which the opposite of an underlying feeling
appears in consciousness. The conductor should not necessarily draw the
group's attention to signs of anger, which can often be ignored, but if these
are persistent, exploration may be indicated. Underlying attitudes can be
considered, and group members may be helped to handle the experience or
expression of anger more appropriately. When anger is overt, the conduc-
tor or other group members may face considerable hostility. If the
conductor is on the receiving end he/she must remain outwardly calm, and
Emotionality 309
wait for the feelings to have been expressed before making any comments.
Even after a delay he/she must not be in too much of a hurry to defend; the
views of other group members should be invited, and he/she may decide not
to give a specific reply. However, if the anger is directed to another group
member, the conductor may need to be actively protective. If the staff are
being accused of having various shortcomings, it is usually better not to
argue, and an angry response on the part of the conductor is invariably
inappropriate, especially if it involves a lack of the self-control that is an
essential tribute of the role. It is often helpful to try and identify a pattern of
inappropriate emotional arousal or expression, as the group experience
may be representative of behaviour that repeatedly occurs in other
situations.
A patient who is depressed and tearful is often ashamed at exhibiting
emotion and if there are clear signs that a group member is upset it is often
reassuring if the conductor indicates that this has been noticed. This implies
that it is 'all right' to be upset and the conductor may then try and
encourage sharing of similar feeling, either past or present, and mobilize the
group's support. However, the tearful patient may find it very uncomfort-
able to remain at the centre of the group's attention, and the conductor
should consider shifting the focus without necessarily changing the subject;
for example, a patient who had been upset the previous week could be asked
how he/she has been feeling. Talk of suicidal ideas may reduce the
likelihood of subsequently putting these thoughts into action, but it may be
necessary to arrange to see the patient individually after the meeting, if, at
the end, he/she appears to be actively suicidal. However, the motivation
behind some patients' talk of possible self-injury is the expression of anger
or an attempt to manipulate the attitudes and behaviour of the group. In
such cases it may be necessary to state clearly that the group or staff cannot
always accept responsibility for a person's actions. At the same time it is
important to maintain a sympathetic approach, which can be difficult, as
the behaviour of some patients in such situations often arouses hostility.
Emotions can be 'acted-out' by disruptive non-verbal behaviour which,
if it is not modified or contained by group pressures, may require a patient's
exclusion from the group by the conductor. Physical intervention by the
staff may very occasionally be necessary to stop aggression or self-injury,
and, if a patient leaves the meeting, it may be appropriate for a member of
the group to try and persuade him/her to return.
The conductor's own emotions must not be forgotten, and a capacity for
self-control with regard to their expression is an essential skill, which may
require experience for its development. If the conductor expresses strong
emotions, this should only occur after a decision that this feedback would
be helpful to the group. If, occasionally, this ideal is not realized, and the
conductor's involuntary expression of strong feeling is apparent, it is
probably best to acknowledge to the group that he/she is feeling angry,
depressed or anxious. This tactic may defuse tension or embarrassment.
Positive emotions should not be discouraged, but laughter or jokes are
often defensive and inappropriate. They should be accepted at their face
value only with caution.
Patterns of group interaction

Various developmental stages have been described in 'closed' unstructured
groups (which rarely admit new members), and these can also be seen, albeit
less predictably, in 'open' groups (Schutz, 1958). In a closed group, an
initial 'forming' stage has been identified in which members are often
anxious, defensive and ambivalent. The group is dependent on the therapist
for guidance and takes few risks. Members are on their best socially
acceptable behaviour as they attempt to find out what is permitted and the
nature of their task (Bloch, 1979). The middle stage has been termed the
'working' group and involves several patterns of group activity: 'storming',
in which conflicts are expressed (often accompanied by criticism of the
conductor for not being sufficiently directive), 'norming', in which cohesion
develops, and 'performing', in which self-disclosure occurs and solutions to
problems may emerge. In addition there may be periods when defence
mechanisms are predominant (Bion, 1961). The final stage of a closed
group is that of 'termination', which is analogous to the discharge of
individual members from an open group. Termination should occur if no
further significant benefits are possible, and leaving the group may be, in
itself, a therapeutic process, during which further steps are taken towards
an appropriate degree of emotional independence.
Although the above stages are not delineated so clearly in an open group,
phases of group interaction, e.g. involving dependency, cohesion or
emotionality, are often experienced. Thus, if a group is constructively
discussing problems in a relaxed and supportive atmosphere, this does not
usually last for more than, at most, a few meetings; conversely, the
conductor can predict that meetings characterized by despondency, hos-
tility and apathy are likely to become more positive.
Although patterns of communication that involve several group
members should be encouraged, lengthy dialogues may also be appropri-
Patterns of group interaction 311
ate, if these do not lead to undue exclusion of other group members. Overt
or covert structuring of communications, such as turn taking or assuming
that certain questions must never be asked, should be generally avoided in
an unstructured group but, as previously described, patients may be asked
in turn for their comments near the end of the meeting. The conductor
should remember that the communication pattern of a group has a life of its
own, which is not entirely dependent on the staff activity, and that members
interact to a considerable extent outside groups if they are inpatients.
The degree of staff activity is a common topic for discussion in staff
meetings. As part of theflexibleapproach that is required, the conductor
must be prepared to vary the pattern of staff activity; constant intervention
by the conductor stifles the group processes, but a directive and active
approach by the staff may be needed when the group members lack the
capacity to provide a positive input.
Certain themes are commonly encountered in unstructured groups in
psychiatric practice, such as previous treatment, symptoms, problems of
being a patient, and the relevance, function and difficulties of the group
itself. Dependency, trust, self-confidence, loss of hope, future plans, coping
with emotions, loneliness, difficulties with authority, childhood exper-
iences, recent life events and marital problems are also often discussed.
Responsibility for one's actions is another recurrent theme and group
members may share unrealistic expectations of some kind of magical
treatment that will involve little effort on their part. Although many
psychiatric syndromes respond well to physical treatments, the kind of
problems that are considered in groups invariably require some effort and
motivation from the patients.
If membership has been beneficial, leaving the group can be a significant
event, even in patients who have not participated actively. The setting of a
target discharge date is necessary for those patients who have developed a
significant degree of dependency on the group, and the discussion of
reactions to discharge may be an important part of treatment. The
conductor should mention when it is a patient's last meeting and, if
appropriate, emphasize that follow-up arrangements have been made.
Psychological defences are often intensified before discharge, so that
emotionality and exploration of unconscious processes should be mini-
mized during a patient'sfinalmeetings. Timing of discharge is vital, as if it is
too late a damaging over-dependence may have resulted. The optimum
time for each patient will be determined, to a large extent, on the goals that
have been formulated.
The conductor's problems

Personality, knowledge and experience determine the efficacy of a group
conductor, and it must be recognized that certain traits, such as excessive
emotional coldness or obsessionality, can disqualify an individual for this
task. Conducting a group can be a stressful experience, and the relatively
inexperienced conductor should always have the assistance of at least one
other staff member, together with regular supervision. Various maladap-
tive behaviour patterns have been described in trainee psychiatrists who,
together with other professionals, often find that working in groups is one
of the most demanding of all their duties. The conductor is also involved in
the group processes and may be caught up in contagious emotion. He/she
may be the focus of hostility and be subjected to unreasonable demands,
which require a firm stand against pressures to conform. Faced with
stressful situations, the conductor's defence mechanisms may distort
judgement and he/she may take refuge behind a professional mask, by
showing his/her 'wares' in the form of long summaries, involved interpre-
tations or the use of jargon (Yalom, 1966).
The main attributes of an effective group psychotherapist are a degree of
self-awareness and self-control, together with a confidence to doubt and
question his/her own actions. The potentially harmful effects of counter-
transference or the conductor's defences can be minimized by supervision,
feedback from other staff members in the group, and the conductor's
increasing recognition of certain characteristics of his/her own personality.
Self-control involves an ability to endure difficult interpersonal situations
by 'sitting it out' and keeping relatively calm. This does not imply that the
conductor is not experiencing strong emotions, but their expression is being
controlled. Reacting with anger and getting caught up with a prevailing
feeling of despair are two of the most common pitfalls.
Various situations are commonly found in relatively unstructured group
psychotherapy, for instance, group members may sit expectantly, awaiting
direction from the conductor in the hope that someone else can sort out
their problems without any effort on their part. This attitude may need to be
discussed and the conductor should repeatedly clarify what can and cannot
be done without the members' active participation. This may reduce the
hostility that usually accompanies unrealistic and unfulfilled demands.
There are many kinds of silence that reflect apathy, active consideration,
a brooding hopelessness, perplexity, anger, tension, stubbornness, or,
especially with new members, uncertainty as to what they should be
discussing in unstructured meetings. In general, long silences are not
Negative effects 313
beneficial, but their resolution will depend on the underlying cause. Of
course, it should be remembered that although the predominant mood or
attitude of the group in a silence (or in any other situation) can often be
identified, it may not be shared by all the members. Some of those who do
not share the prevailing attitude or emotion understandably find repeated
references to the opinions of'the group' rather irritating, and the conductor
should always bear in mind that a group is composed of individuals.
If the conductor does not know what to do in any situation, the answer is
often 'nothing', apart from putting up with the situation while continuing
to try and help the group to function. The burden of responsibility for the
proceedings of the group is shared by all its members and, apart from not
giving up, the conductor is not necessarily obliged to 'do something'.
At times the conductor is aware of a major event, such as a recent
attempted suicide of one of the members, which is being avoided at the
expense of'safer' topics. In this situation the issue should usually be raised
at some time during the meeting.
Another common event is the repeated questioning of the conductor by
one or more group members; this poses a problem, because although it is
often appropriate to give direct answers, an excess of this type of interaction
fosters dependency and impairs group communication. It may be necessary
for the conductor to say that a question will be answered later in another
setting, or that an excess of questions and answers is unhelpful in group
meetings. However, it is rarely appropriate to completely ignore a question.
Group members may try and strike up a special relationship with the
conductor, perhaps by imparting 'confidential' information between meet-
ings, but the conductor should usually make it clear that he/she is not able
to accept information on this basis and that if it is confided, the right to
bring it to the attention of the group will be reserved. There are many ways
in which members attempt to manipulate the conductor to carry out
inappropriate actions.
The overtalkative group member who monopolizes meetings is another
common problem. Often group pressures and firm interventions by the
conductor are effective, but in the case of hypomanic patients in inpatient
units, temporary exclusion may be necessary.
Negative effects
A negative effect has been defined as a lasting deterioration in a patient,
directly attributable to therapy; however transient effects may also be
important (Hadley & Strupp, 1976).
Existing problems may be magnified, for example, the inability of the

dependent patient to refuse unreasonable demands may be exploited by
group members, the paranoid patient may consistently evoke hostility that
strengthens his/her beliefs, and social phobias may be intensified in a group
that lacks cohesion, concern and support. Psychotic patients or those with
limited intelligence may not be able to join in, thus compounding a feeling
of inadequacy, while exposure to an emotionally charged situation may
exacerbate a psychosis in vulnerable individuals. Also, maladaptive defence
mechanisms and behaviour patterns may be reinforced; for example, the
withdrawn individual may become more socially isolated, or the group may
provide an opportunity to display undue aggression.
New problems may also arise. A system of defence mechanisms can
represent the best possible way of coping and, if these are challenged,
decompensation may result in further impairment of the patient's adjust-
ment, perhaps involving depression, anxiety or excessive dependence on the
group. In addition, new defences may develop that are even more maladap-
tive than the originals. Symptom substitution may occur; for example a
preoccupation with somatic complaints may be replaced by the overt
development of a previously covert paranoid delusion. The setting of
unrealistic goals will lead to frustration, disappointment and loss of hope
for both the patient and therapist, while for some patients the pursuit of
'therapy' can become an unproductive end in itself, which protects the
individual from the need to face up to immediate problems. Breaches of
confidentiality may occasionally have serious consequences, for example if
they involve extramarital relationships or criminal activity, while for some
patients a relapse in the severity of their disorder, perhaps involving suicidal
intent, may not be recognized in a group setting.
Emotional arousal may have several harmful effects; heightened emo-
tions can give rise to destructive or antisocial 'acting out' behaviour, such as
taking an overdose, and can occasionally precipitate a relapse of manic-
depressive or schizophrenic features in patients who are predisposed to
these disorders.
The conductor must monitor the group sessions for harmful effects and
take active steps to stop their development. If necessary a patient should be
excluded, either temporarily or permanently, for his/her own good or to
prevent harm to others. If meetings are consistently characterized by anger,
distrust and lack of concern, the disadvantages will probably outweigh any
advantages, and disbanding the group should be considered. However, this
should not be necessary with an experienced conductor and careful
selection of patients.
Finally, the development of a sexual relationship between two group
Varieties of group psychotherapy 315
members may occur, usually with adverse consequences, especially if the

male patient has pronounced antisocial personality traits. This potential
problem should be anticipated when selecting the group members.
Training and supervision for group conductors

A group conductor should have certain attributes of personality, a degree
of self-awareness, and some theoretical knowledge and experience.
Self-awareness and theoretical knowledge can be acquired in various
ways, which often include a meeting of trainee conductors with their
supervisor in a group setting. In this group, the same phenomena that are
found in psychotherapeutic groups may occur; for example, the trainees
may show resistance by avoiding discussion of certain aspects of their work,
or by an inability to monitor group processes. Trainees' counter-transfer-
ences to the group members may be apparent to other members of the
supervision group; for instance, a trainee conductor may inappropriately
feel that the problems of a group member have a bad prognosis. A
transference on the part of the trainee to the supervisor may show itself in
an excessive need for approval, an unresolved hostility to authority, or
intense rivalry with fellow trainees.
Varieties of group psychotherapy

Various types of group psychotherapy, with different degrees of structure,
can be organized within a psychiatric service. Some groups may cater for
outpatients alone, or the conductor may select a relatively homogeneous
group with regard to PD diagnosis or to treatment goals. Although some
group activities for outpatients may involve little more than social contact
and a cup of tea, others may have goals which include significant personal-
ity change. These are often closed but 'open' formats may also be used.
Each session may last for between 60 and 150 minutes and may occur
weekly, perhaps for about 12-18 months for closed groups. Sometimes
groups may be selected with regard to the homogeneity of presenting
problems such as social phobia, agoraphobia, alcoholism, drug addiction
or sociopathy. Groups for sociopathic individuals should be in the context
of a specialist institution or service. Other groups may target couples with
marital difficulties, the relatives of patients, or several members of one
family.
Additional varieties of group occur outside the setting of a psychiatric
service. 'Marathon' groups have been described, lasting for between 6 and
48 hours, and peer self-help groups such as Alcoholics Anonymous and
Depressives Anonymous may not include a professional conductor. The

use of so-called alternate sessions involve the patient/client members of an
outpatient group meeting on their own, alternately with conventional
sessions.
Group psychotherapy and borderline personality disorder
Overview
It has been widely accepted that group treatment, in particular relatively
structured groups, can be useful for the management of patients with
borderline PD, although concurrent individual sessions are usually recom-
mended. But it is important to note the high drop-out rate from treatment.
Clinical practice has mainly developed on the basis of the opinions of
experienced clinicians, although there is some evidence that supports the
provision of weekly, task-orientated groups for patients who all have a
borderline PD diagnosis, running in parallel with individual treatment
sessions from a different therapist (Horwitz, 1987).
Waldinger & Gunderson (1984), in a retrospective study of 790 patients
with borderline PD, noted that they generally ended treatment earlier than
recommended. In relation to group treatments, Kretsch, Goren & Wasser-
man (1987) looked at the effects of a combination of individual and group
therapy over several years and concluded that the group element had been
useful. But the best evidence for the use of groups for those with borderline
PD comes from Linehan and colleagues (1991), who compared a combi-
nation of individual therapy and weekly group therapy with 'treatment as
usual'. The weekly group lasted 2\ hours and had a structured, educational
format, focussing on the teaching of behavioural skills such as managing
interpersonal interactions, distress tolerance, accepting reality and regulat-
ing emotional expression.
If patients drop out of treatment early, or have to be discharged because
of lack of benefit, it can be useful to present this to the patient as a failure of
the treatment to be helpful, rather than a failure of the patient to cooperate.
This can be combined with an offer to consider a re-referral to rejoin a
treatment programme at a later date.
Positive ingredients
Gunderson (1984) considered that groups can exert a socializing influence
on those with borderline PD, while Macaskill (1982) found that patients
who had experienced group treatment reported that increased self-under-
standing and the opportunity to try and help others had been the most
valued aspects. Also, groups may be a unique way of helping those
individuals who cannot tolerate or agree to individual psychotherapy
(Grobman, 1980).
If a group is selected with a view to an unstructured explorative
approach, it may be appropriate to limit the number of members with a
borderline PD to two, because of the danger of the development of
uncontrolled aggression in an unstructured group in which all members
have a borderline PD (Pines, 1990). In this situation, it has been suggested
that the same therapist should also be involved in providing parallel
individual sessions (Wong, 1980).
It often seems that a patient's peers in a group are better able than
clinicians to confront maladaptive and impulsive behaviour, as their
comments are accepted more readily. Another advantage of a group
compared with individual sessions is that some patients are able to identify
dependent and manipulative behaviour in others, which can be afirststep in
recognizing these aspects in themselves. Also, a group can provide a forum
in which to practise communication of feelings and the development of
empathy for others, while confrontation by peer pressure can be balanced
by support.
But perhaps the main advantages that a group can offer compared with
individual psychotherapy alone, is to dilute, or avoid, the intense feelings
(transference) of either idealization or devaluation that usually occur when
the patient with features of borderline or dependent PD is being seen
regularly by an individual clinician only. A group reduces the danger of an
uncontrolled escalation of childlike feelings of anger and dependence
(regression), which can lead to severe or repeated episodes of self-harm.
This is a particular risk for some patients with borderline PD who receive
frequent (once or twice weekly) long-term individual psychotherapy of an
unstructured, explorative nature, if they are not able to tolerate this
demanding intervention. Such a risk can be reduced by limiting the
frequency and duration of individual sessions, and by giving priority to the
discussion of external, current problems, as well as by arranging for
concurrent group therapy.
In general, an unstructured format is considered to be inadvisable for
those groups whose members all have a diagnosis of borderline PD, as
regression and self-destructiveness may dominate the picture. For such
patients, relatively structured groups can be used to promote reality testing,
self-esteem and skills for interpersonal relationships. Specific tasks can be
given, such as the discussion of decision making, problem solving, collabor-

ation with others, discharge planning, development of daily living or
improvement of social skills (Goldstein, 1990). Also, current external
problems should be a frequent topic of discussion, for example, the
situations that led to recent crises.
Multimodal treatments for borderline PD (and some other PDs)

Group psychotherapy is often part of several concurrent or sequential
treatment interventions in psychiatric practice, and combining treatments
is important in relation to borderline PD. Group therapy has often been
used as part of a treatment 'package' which, over time, may involve hospital
admission, long-term residential treatment, day care, night care, and
outpatient treatment. Group methods may play a part in management in all
these settings.
The importance of the social environment (or milieu) in psychiatric
hospitals and other institutions has long been recognized, and Main (1946)
introduced the term 'therapeutic community', which has come to imply
frequent group meetings, an emphasis on problems involving social
relationships, good interpersonal communication, flattening of the pyra-
mid of institutional authority, free emotional expression and examination
of institutional roles (Clark & Myers, 1970). However, this type of
organization produces its own problems (Zeitlyn, 1967).
Considerable and relatively exclusive emphasis on these principles results
in the therapeutic community 'proper', and has been advocated for some
long-stay wards of psychiatric hospitals and for some specialist units
catering for those with antisocial or borderline PDs and for those who
abuse drugs and alcohol. But this approach is not suitable for a typical
short-stay psychiatric 'admission' unit, which must cater for a wide range of
psychiatric disorders. However, principles derived from the therapeutic
community 'proper' are of considerable value in psychiatric 'admission'
units. A variety of group activities should be available, the patients are
encouraged to assume as much responsibility for themselves and their
fellow patients as is possible, while staff of all disciplines should be willing to
consider constructive and informed criticism, taking decisions in consul-
tation with their colleagues. However, in the case of disagreement, it should
be recognized that it is the responsibility of the psychiatrist to make the final
decision regarding the medical aspects of treatment or management. Staff
and patients of an admission unit may meet regularly for an unstructured
'community meeting', which may involve up to 30 people for 45 minutes,
but the patients should know that they are free to choose not to attend.
Although daily meetings of this type have been advocated, these can be
limited to the beginning and end of the week, with various structured
communal activities scheduled for other days.
A 'large group' has been defined as a meeting that includes more than 30
people (Kreeger, 1975), and Schiff& Glassman (1969) have described the
consequences of increasing group size: a higher proportion of patients are
silent, subgrouping increases, the participation of the less active members is
reduced, and the contribution of the conductor and staff is increased.
There is some overlap between the activities of community meetings in an
admission unit and group psychotherapy, but the former should focus on
day-to-day problems in the ward and administrative matters. Community
meetings also enable the staff to identify the main management problems of
the ward and assess a large number of patients in a short time. All the staff
who attend should be urged to participate, although not necessarily in every
meeting, and one of their members is designated as the conductor.
Various forms of structured group activity should contribute to an
admission unit's programme; for example, a weekly review meeting chaired
by a member of the nursing staff can discuss non-confidential aspects of
treatment, considering each patient in turn, but if patients prefer not to
participate, this wish should be respected. Administrative arrangements,
including those for weekend leave, can also be clarified. Other forms of
structured group activity may involve art, music, psychodrama, and
various forms of recreation.
Staff members require their own group activities, although there is a
danger that an excess of time is spent in this way. Inexperienced staff who
are involved in group psychotherapy can be supervised in weekly meetings
and all staff in a group can spend about 15 minutes after each meeting
discussing the session. The overall staff effectiveness in an admission unit
depends to a significant extent on adequate communication, and all staff
should meet regularly for an administrative meeting. This can be struc-
tured, with a chairperson and an agenda.
If hospital care is required for borderline PD it is usually short-term (i.e.
less than 4 weeks) and precipitated by a crisis. The aim may be to reduce
self-destructive and antisocial behaviour or to reduce the intensity of the
transference relationship if the patient is receiving individual psychother-
apy. Also, the ground may be prepared for future outpatient treatment. But
for the most severely-affected individuals, for instance those who show
repeated episodes of self-harm, longer-term residential care may be indi-
cated, although there is a risk that an intense dependence will develop, with
severe regression. If possible, less intensive long-term treatment should be
considered, such as day (or night) care.
When a patient with borderline PD is in an hospital or residential setting
(including day care), it has been generally recommended that, apart from
formal treatment sessions, the 'milieu' should have as much structured
activity as possible. Also, a balance has to be struck, particularly in relation
to borderline PD, between setting limits and being supportive and tolerant.
Clinicians and other staff must be as clear and consistent as possible with
regard to expectations, limits and sanctions. But it must be expected that
patients will display their maladaptive behaviour in treatment settings.
Many patients with PD can benefit from self-help groups such as those
provided by Alcoholics Anonymous (AA), Narcotics Anonymous, and
organizations concerned with eating disorders. In some cases, a support
group is backed-up by a 24-hour/day phone support.
The design of a multimodal treatment programme for each individual
patient is usually a matter of judgement. For those with borderline PD, a
crisis admission to hospital may be followed by individual outpatient
sessions running concurrently with attendance at weekly group therapy.
Subsequently, one or other of these interventions may be discontinued and,
if individual sessions remain, their duration and frequency may be gradu-
ally reduced.
Karterud et al. (1992) described a day hospital therapeutic community
programme for patients with various PDs, including borderline PD. The
mean time in treatment was 6 months and the dropout rate for schizotypal
and borderline PD patients was 38%. 'Cluster C PD did best, and results
were rated as 'modest' for borderline PD and 'very modest' for schizotypal
PD. However, it appeared that the 'containing capacity' of this regime is
considerable, and may reduce the need for long-term inpatient care for the
minority who would require this if an alternative is not available.
Management of offenders with personality
disorders
A.T. GROUNDS
Personality disorders in offender populations

The frequency of PDs amongst people who commit criminal offences is
difficult to determine. A high proportion of criminal offences are unre-
ported, and in only a minority of reported cases are the suspects appre-
hended. Only some of these result in criminal proceedings, and a smaller
proportion again will be cautioned or convicted. In the UK it is estimated
that only 3% of crimes result in cautions or convictions. Thus, by the time a
potentional population of offenders becomes identifiable and accessible for
research purposes they are unlikely to be representative of offenders as a
whole. Populations of prisoners, on whom most epidemiological research
has been carried out, are an even more highly selected group.
Figures from the UK British Crime Survey and from UK annual criminal
statistics indicate that about 15 million criminal offences are committed
annually in England and Wales with a population of around 50 million, of
which half are reported to the police. In 1992, the number of people
cautioned or convicted of criminal offences was 1.8 million and the number
of people sentenced to imprisonment was 58000 in 1992 (Home Office,
1993a).
A further difficulty in assessing rates of PD amongst offenders lies in the
variety of definitions and research instruments for identifying PD. In
offender populations, in whom antisocial behaviour is of particular import-
ance, a variety of diagnostic approaches have been used. Dolan & Coid
(1993) summarized the main contemporary systems, including the ICD-10
diagnosis of dissocial PD (World Health Organization, 1992), the DSM-
III-R diagnosis of antisocial PD (American Psychiatric Association, 1987),
and the Hare Psychopathy Checklist (Hare, 1991).
The most recent major study of the prevalence of psychiatric disorder in
321
322 Management of offenders with personality disorders
the sentenced prison population of England and Wales was carried out by
Gunn, Maden & Swinton (1991), and reported that 7.3% of sentenced adult
men and 8.4% of sentenced adult women had primary diagnoses of PD. But
these were clinical diagnoses based on interviews conducted by research
psychiatrists who did not employ standardized diagnostic instruments, and
the researchers' limited access to comprehensive information probably led,
as they acknowledge, to an underestimate of true prevalence rates. Coid
(1993a) reviewed the epidemiological studies that have employed stan-
dardized diagnostic instruments for antisocial PD, and noted that studies
of North American prison samples have found prevalence rates varying
between 39% and 76% of subjects. Thesefigurescompare with a lifetime
prevalence rate for DSM-III-R antisocial PD of 2.6% (4.5% for men and
0.8% for women) in the US Epidemiological Catchment Area study
(Robins et al, 1984; Robins & Regier, 1991). It is important to note that
although there is a strong association between criminality and antisocial
PD, the two are distinct, as a substantial amount of criminality is
committed by people who would not qualify for the diagnosis of antisocial
PD (Robins, 1993).
In clinical practice a careful and rigorous approach must be taken to the
diagnosis of PD among offenders who are seen within psychiatric services,
and a history of criminal offending, however heinous, does not in itself
justify a diagnosis of PD. The broad diagnostic principles outlined in ICD-
10 or the DSM-IV should be followed, and it must be kept in mind that the
basic feature of a PD is longstanding abnormality of personality that
adversely affects several domains of a person's life, and that has a disabling
effect on social functioning.
Particular care must be taken to avoid making unwarranted diagnoses of
PD amongst people with histories of psychoses, such as schizophrenia, who
later engage in recurrent offending behaviour. There is now good evidence
from cohort studies that serious mental illness is associated with increased
risk of criminal offending, and, in relation to schizophrenia, criminal
careers of such patients are likely to commence at a later age than in most
offenders (Hodgins, 1992). Not uncommonly, people with a diagnosis of
psychosis in early adult life may present subsequently with problems that
are dominated by troublesome offending behaviour in the absence of
prominent psychotic features. A careful history will often establish that the
behaviour represents a deterioration from the person's pre-morbid level of
functioning, and if this is the case the proper diagnosis is likely to be of
chronic psychosis rather than of primary PD.
Clinical assessment for treatment 323
Clinical assessment for treatment

Offenders with PDs may often be referred for psychiatric assessment
because they or the referring agency hope that psychiatric treatment will be
effective in achieving psychological change, but referrals may also be made
in a wider range of situations.
Defendants charged with offences and awaiting trial may be referred for
assessments by courts, the probation service or lawyers. Also, when faced
with the stress and crisis of court proceedings, defendants may turn to
primary care physicians who may then request psychiatric assessment and
support. Later on in the criminal process, offenders facing sentencing may
be referred for assessment in order to provide courts with advice about
specific treatment options that may be available as an alternative to a
custodial sentence. In addition, sentenced prisoners may need evaluation of
their treatment needs in prison or their suitability for transfer to hospital for
treatment during sentence. On release they may be referred for psychiatric
follow-up, and, in a few cases, psychiatric supervision may be made a
condition of a parole licence.
While most offenders with PDs will not receive a medical 'disposal' from
the courts, in England and Wales the Mental Health Act 1983 contains
equivalent powers in relation to 'psychopathic disorder' and 'mental
illness'. But there is the proviso, in the case of patients with 'psychopathic
disorder', that treatment is likely to alleviate or prevent deterioration of the
person's condition. In practice, however, only a small minority of offenders
with PDs are recommended for 'hospital orders'. Such individuals tend to
have committed grave offences and require detention in a secure hospital
setting and restrictions on discharge. When conditionally discharged from
hospital they are likely to be subject to statutory psychiatric supervision.
It is important to have a clear understanding of the context and purpose
of the psychiatric assessment in the case of an offender with PD, and the
person being assessed should also appreciate the uses to which the
assessment may be put. A patient referred by a primary care physician may
be willing to disclose more than would be the case if a report has been
requested by a court. If there is a possibility that a court report will be
requested at a later stage, the patient should be made aware of this, and
separate consent to prepare such a report should be obtained.
Assessment of offenders requires the identification of the range of the
patient's psychiatric and social difficulties. Patients with PDs may have
multiple, severe and chronic disabilities, and also suffer from episodes of
other psychiatric disorders, particularly in situations of stress. Usually,
psychiatric management will require agreement and commitment by the
patient, and a realistic approach has to be taken in assessing the prospects
of success and in setting goals that are achievable. Motivation should be
carefully evaluated, but its absence at a particular time should not
necessarily result in no help being offered. Some patients will require life-
long support, which will be provided by intermittent contact with a
psychiatric service over many years, during which they seek help when they
wish.
When preparing reports for courts in cases where psychiatric support
and treatment may provide an alternative to a prison sentence, it is
particularly important to ensure that the court clearly understands the
nature and likely success of the treatment, and the extent to which
treatment may or may not influence re-offending risk. In cases where the
clinical opinion is that treatment should be provided, but it will not
necessarily prevent further offending, this needs to be carefully explained.
For most offenders with PDs, treatment is an enterprise that involves the
offender working in partnership with clinicians; such treatment cannot be
compelled and major short-term improvement is not to be expected. As
with any patient who has chronic disabilities, the aim of treatment is long-
term and rehabilitative in approach, with a focus on measures that promote
and maintain the person's best possible level of functioning within their
limits, together with extra support in times of crisis or deterioration.
Range of interventions
A wide range of treatment interventions may be considered in the manage-
ment of offenders with PDs. These include a variety of psychotherapies,
social support and, in a minority of cases, medication. Also, counselling for
problems of drug or alcohol abuse may be needed. The treatment setting
will usually be outpatient-based, with brief admission to hospital as a
possible option at times of crisis. In a minority of cases longer-term
specialized inpatient treatment may be indicated, but such facilities are
scarce. In the UK, the Henderson Hospital provides 'therapeutic commun-
ity' treatment in an open hospital setting, while offenders convicted of grave
offences who are facing sentencing may be considered for treatment in
'special hospitals' if they meet the criterion for admission to maximum
security, namely that they constitute a grave and immediate danger to the
public. However, there must be careful consideration about what treatment
is proposed, its availability and the prospects for success.
People convicted of sexual offences may or may not merit a diagnosis of
Community and outpatient settings 325
PD or other psychiatric disorders, but convicted sex offenders receiving
non-custodial sentences may be considered for community-based treat-
ment programmes, which, in the UK, are usually organized by the
probation service. Residential group-based treatment in the UK is
provided at the Gracewell Clinic, Birmingham. Programmes for sentenced
prisoners serving 4 years or more have been established in a number of UK
prisons, and the 'special hospitals' also provide treatment for some patients
with histories of sexual offending.
Community and outpatient settings

The clinical management of offenders with PDs will often involve working
jointly with other agencies, particularly the probation service in the UK,
and as Joseph (1992) stressed, psychiatrists also need to work in close
partnership with primary care physicians. In psychiatric outpatient set-
tings, the medium through which care and treatment are provided is the
relationship with the patient, and attention must be given to how the
relationship is established and maintained. From the outset efforts should
be made to ensure that the patient feels understood, and that there is shared
agreement about realistic treatment objectives and mutual expectations.
Offenders with diagnoses of antisocial or psychopathic personality dis-
orders are particularly likely to have early histories of deprivation and
impaired early relationships, sometimes associated with rejection or victi-
mization of the child (Luntz & Widom, 1994). In adult life the patient may
therefore be abnormally sensitive to perceived provocation or rejection,
and particular vigilance and care is required at the outset in handling the
first interview.
The adoption of a realistic attitude towards treatment goals is important
clinically in order to ensure that the clinician and the patient commence
with appropriate expectations of each other. It is also consistent with the
available research evidence on the efficacy of treatment for psychopathic
and antisocial personality disorders. As West (1983) emphasizes in his
excellent brief review, treatment and rehabilitation programmes for
offenders have generally not shown strongly beneficial effects in terms of
reducing recidivism. However, he cautions against the adoption of a
nihilistic attitude by clinicians towards antisocial offenders. As he notes,
antisocial behaviour and delinquency cannot be separated from their social
context, and it is social and educational intervention that is primarily
needed. Thus, the relatively poor results of treatment programmes do not
necessarily imply that offenders have not benefited individually, but they
illustrate the fact that crime is overwhelmingly socially determined. Psycho-
logical changes in the individual may therefore have a relatively small
overall influence on likelihood of reoffending. Nevertheless, continuing
psychiatric interest and involvement in trying to help antisocial offenders
should be maintained. As West (1983) suggests, a careful diagnostic
approach is needed to ensure that any organic disorders and learning
disability are identified. The secondary consequences of antisocial behav-
iour, including substance abuse, relationship and family disturbances, and
any co-morbid psychiatric disorders need attention, and clinicians should
work in close liaison with social and criminal justice agencies:
A developed anti-sociality syndrome is a serious maladjustment and an unhappy

state of affairs for society and for the family as well as for the individual concerned
(West, 1983).
The recent review of the effectiveness of treatment for psychopathic
and antisocial PD by Dolan & Coid (1993) reached similar conclusions.
They summarized the lack of good studies of the outcome of treatment in
this area, although they noted some positive effects from studies of
cognitive/behavioural psychotherapies, community supervision, and, in
particular, therapeutic community approaches. Overall they concluded
that: 'there is still no convincing evidence that psychopaths can or cannot be
successfully treated' (Dolan & Coid, 1993). They urged caution in readily
accepting the view that these patients are 'untreatable', noting that: 'the
"untreatability of psychopaths" may, in part, result from the professionals'
inadequate assessment in the first place, followed by an inability to develop,
describe, research and adequately demonstrate the efficacy of treatment
strategies' (Dolan & Coid, 1993).
Health care services have responsibilities to provide care and treatment
for patients with PDs, and, in the UK, there has been an increasing
emphasis during recent years on the adoption of the 'Care Programme
Approach' (CPA) to the care of patients referred to psychiatric services.
The key elements are: first, systematic assessment of health and social care
needs; second, an agreed care plan; third, the allocation of a key worker to
ensure the programme of care is delivered; and fourth, the regular review of
the patient's progress and needs (Department of Health, 1990). The CPA
has been recommended in the management of certain patients with PDs
(Department of Health, 1994a), and such patients are also specifically
included in the requirements on health authorities in the UK to establish,
from October 1994, 'supervision registers' for patients who are at signifi-
cant risk of suicide, severe self-neglect or of causing serious violence to
others. The purpose of supervision registers is to identify and prioritize
these patients, to provide care plans for them, and to ensure careful
monitoring and planning for their care. The circular introducing these
arrangements notes that: 'Mental illness in this context includes people with
a diagnosed personality disorder, including psychopathic disorders, who
are receiving treatment from specialist psychiatric services' (Department of
Health, 1994ft).
Providing a supportive relationship is of central importance in the
management of offenders with PD. This needs to be consistent, tolerant and
potentially long-term. Patients with disorganized and unstable lifestyles
may have difficulty in maintaining regular outpatient attendance, and such
patients may best be managed by an open-ended commitment to see them
again on request at times of crisis. Contact may therefore be intermittent
over a long period, and it will be necessary to make a fresh assessment on
each occasion, and to work on any new problems that have arisen.
Relatively dependent patients may need to be seen regularly over a very
long period, and the management of such a relationship can be difficult.
Limits may have to be set to ensure that the frequency and duration of
contact is kept within reasonable bounds, but without losing an attitude of
acceptance, 'professional friendship' and a willingness to increase support
at times of crisis. For some patients, the relationship with the clinician may
be the only long-term, confiding relationship that they have, and its
importance should not be underestimated. A long-term perspective on the
part of a clinician is required; over short- and medium-term time scales,
support in a series of crises may be needed in the context of little change in
the underlying PD, but over a few years the emotional and social stability of
some offender patients can show unexpected improvement.
Psychodynamic psychotherapy
There is a consensus that engaging in psychodynamic psychotherapy with
personality-disordered offenders on an outpatient basis should be
approached with great caution; first, because of limited evidence of the
efficacy of this approach, and second, because of an increased risk of
disturbed behaviour. Malan (1979) recommended careful scrutiny of the
individual's history when assessing the suitability for such psychotherapy
because of the possibility that the most disturbed behaviour the person has
exhibited in the past might be re-enacted during the therapeutic process.
Main (1957) and Shapiro (1978) have provided good descriptions of the
pathological transference and counter-transference relationships that
patients with features of borderline PD may engender, with oscillation
between extreme idealization and denigration of the therapist, and promo-
tion of splitting and division between clinical staff (Kernberg, 1984).
Whilst most offenders with PDs are not suitable for individual 'psycho-
dynamic psychotherapy' as described in Chapter 7, the clinician should
retain an awareness of psychodynamic issues for three reasons. First,
psychodynamic theories may be necessary in order to develop a proper
psychological understanding of the patient's behaviour and the organiza-
tion of his or her emotional life. Second, because the relationship with the
patient is the medium through which the clinician works, and the dynamics
of that relationship need to be recognised and understood. The emotions
and attributions exhibited by the patient towards the therapist, and vice
versa, need to be considered in terms of what they reveal about the patient's
history and emotional life. Not uncommonly, patients who in early life have
been subjected to cruelty or rejection develop similar pathological patterns
of relating in adult life and these may begin to feature in the therapeutic
relationship with the clinician. Care needs to be taken to recognize this and
to avoid re-enacting rejection or punitiveness in the clinical relationship
and attitudes to the patient. Third, offenders with personality disorders
may sometimes present with a genuine desire for insight and self-under-
standing. They may be confused and distressed by their destructive
behaviour, the chaos of their lives, and their inability to change. In taking
the initial psychiatric history the clinician is likely to see links and
continuities between patients' early life experiences and their problems as
adults. Spending time over a limited number of sessions, taking the patient
in detail through his or her life history, may help them see for the first time
parallels between past and present. In reconstructing their biographies,
patients can experience relief and illumination through recognizing the
experiences that shaped their adult expectations and views of their social
world. For example, suspiciousness and paranoid sensitivity in adult life,
with a tendency to perceive threat and provocation in others and to react
explosively, may have understandable links with childhood experiences of
living in an early environment of genuine fear and threat of unpredictable
violence. Patients can be helped to see the parallels between the way they
view other people in their adult lives, and the way they experienced
significant others as children, and thus how their distorted attributions and
inappropriate responses arise. Although from a psychoanalytic perspective
this work is superficial, it may, nonetheless, be safe and useful: it can be
carried out over a limited period of time, it need not involve the develop-
ment of difficult and dependent transference relationships, and its limited
nature means that it is not too psychologically demanding of the patient.
Cognitive psychotherapy and behaviour therapy 329
However, it can also enable careful testing of the patient's capacity to

accept interpretations and to see links between past and present, and in
some cases this may help identify those patients who might benefit from
long-term and more demanding psychodynamic psychotherapy. An
important UK national centre specializing in outpatient individual and
group psychoanalytic treatment of offenders is the Portman Clinic, Hamp-
stead, London.
Cognitive psychotherapy and behaviour therapy

Short-term cognitive and behavioural treatments, focussed on specific
symptoms or problem behaviours, may have a useful part to play in the
outpatient management of some offenders with PDs. But such interven-
tions should be seen as an adjunct rather than an alternative to long-term
support. Clinical judgement is needed when considering such treatments, as
more empirical evidence is still required for their efficacy in the manage-
ment of offenders with antisocial PDs (Dolan & Coid, 1993).
Common problem-areas to which cognitive and behavioural treatments
may have a contribution to make include social anxiety with avoidance,
repeated aggression and violence, sexual offending and paranoid sensi-
tivity. Specific problem-behaviours such as repeated shoplifting and com-
pulsive gambling may also be helped by these approaches (Gudjonson,
1980; Greenberg & Rankin, 1982). The application of these techniques to
problems of violence and sexual offending will be briefly outlined (also see
Chapter 7).
In a cognitive-behavioural approach to anger management, patients can
initially learn to identify the circumstances in which responses of rage occur
and to keep records of their experiences. They can then be taught to view the
onset of these situations as tasks to be worked through, rather than
intolerable provocations, and they learn a variety of coping skills, including
relaxation training. Their skills are then progressively tested.
Community-based treatment programmes for convicted sex offenders
have been established in recent years by the probation service in many areas
of the UK, while some sexual offenders (or non-offenders with deviant
sexual behaviour) may be referred to psychiatrists. Treatment programmes
(as described in more detail in Chapter 7) may focus on four problem areas:
firstly, changing deviant sexual arousal; secondly, changing cognitive
distortions about offending and its impact on victims; thirdly, improving
social skills deficits; and fourthly, developing strategies for relapse
prevention.
Sexual offenders may have a variety of distorted cognitions about sexual
behaviour, sexual offending and the role of victims. Offenders against
children, for example, may have fixed beliefs that their behaviour is an
expression of affection or a harmless educational experience for the child.
Kennedy & Grubin (1992) distinguished several different patterns of denial
amongst imprisoned men convicted of sexual offences (see Chapter 7). The
treatment of such cognitive distortions may involve individual sessions or
be carried out in a group setting, and challenges may be particularly
powerful in the group context. In view of this, the individual's capacity to
cope with such pressure should be considered before embarking on such
treatment.
Deficits in social skills may also be addressed in the context of group
therapy, for example, using role-play, and may involve teaching appropri-
ate assertiveness and social interaction skills.
Strategies for relapse prevention are also considered. Offenders are
taught to identify the circumstances and events that may precede re-
offending, and they may learn coping strategies for avoiding and coping
with situations of risk. Offenders learn what constitutes a lapse in their
individual case (e.g. use of alcohol or the onset of a deviant fantasy), and
their task is to prevent this becoming a relapse. This approach is likely to be
strengthened if it is carried out in the context of long-term supervision in
which there is an expectation of open reporting by the offender, and the
supervisor and offender work in close partnership in devising agreed
strategies for the avoidance of re-offending.
It should not be assumed that a generic treatment programme can be
devized that is suitable for all sexual offenders, as such individuals are
heterogeneous in their sexual arousal patterns, cognitive sets, social skills,
and motivation to change. They may also have a variety of other associated
psychiatric difficulties. Other treatment methods may also help in some
cases, for example the use of antiandrogen medication, particularly in older
men who are distressed by high sexual drive, and whose consent and
willingness to accept the medication is unequivocal. Careful assessment of
the offender's needs, deficits and skills should be carried out in each
individual case.
Statutory supervision in the community

In the UK, psychiatrists may be involved in the statutory outpatient
supervision of two groups of offenders: sentenced prisoners released on
parole, and conditionally discharged patients who had been detained in
Statutory supervision in the community 331
hospital under the provisions of mental health legislation (Restriction

orders made under section 41, Mental Health Act, 1983).
Prisoners released on parole are under the supervision of a probation
officer until the period of parole ends, and a condition of psychiatric
supervision may be attached to the parole licence. Normally, the probation
officer will have an informal discussion with the relevant psychiatrist as to
whether psychiatric supervision would be appropriate and available before
the decision to grant parole with a condition of psychiatric supervision is
made, and the psychiatrist will liaise closely with the probation officer
during the supervision period.
Life-sentenced prisoners released on licence in the UK may also have a
condition of psychiatric supervision, or may be referred by a probation
officer for voluntary psychiatric help in relation to a complex range of social
and psychological difficulties following release from a long period in
custody. Most life-sentenced prisoners in the UK are sentenced for
homicide offences, and their victims are commonly partners or other close
acquaintances. Family break-up as a consequence of their offences or the
long-term imprisonment is common. Sapsford (1978) found that wives and
girlfriends of male 'lifers' had nearly all lost contact by the end of the fifth
year of sentence. Over half had no contact at all after conviction except to
serve divorce papers. Life-sentenced prisoners are also a group with high
rates of mental disorders (Smith, 1979; Taylor, 1986), and they have to face
major problems of adjustment after prolonged detention. As is the case
with other groups who have experienced prolonged captivity (Herman,
1992), they may present with chronic features of dysphoria, sleep distur-
bance, difficulties in coping with intimate relationships and loss of a sense of
purpose for the future.
Ex-prisoners have a variety of social needs in areas of accommodation,
employment, finance, social isolation, family difficulties and alcohol and
drug use, and it is likely that the ex-prisoner's social situation has a
significant impact on the likelihood of re-offending. A stable situation with
family support and constructive activities, in other words a situation in
which the individual has positive links with the community, probably helps
reduce the risk of recidivism (Haines, 1990).
In England and Wales, mentally disordered offenders who are detained
in secure hospitals under 'restriction orders' are likely to leave hospital in
due course either on trial leave or by being conditionally discharged by a
Mental Health Review Tribunal (or the responsible government minister,
who is the Home Secretary). The total number of 'restricted' patients in
hospitals in England and Wales was 2333 at the end of 1992, of whom just
under a fifth (419) came within the Mental Health Act category of
'psychopathic disorder'. Two-thirds of these had committed offences of
personal violence (including sexual offences). Just over half of all 'res-
tricted' patients in England and Wales are detained in maximum-security
hospitals (Home Office, 1994). In recent years approximately 200 'res-
tricted' patients have been discharged to the community annually (Home
Office, 1994). Conditionally-discharged 'restricted' patients remain liable
to recall to hospital, and the conditions of discharge are likely to include
supervision by a psychiatrist and a probation officer (or social worker), and
there may also be conditions about accommodation and the continuation
of treatment. The careful planning of community supervision and after-
care for this group is particularly important. Close liaison between the
psychiatric and social supervisors has to be maintained, and regular reports
are submitted to the appropriate government department, known as the C3
Division of the Home Office. This is the administrative unit that specializes
in mental health matters, and much of its work is concerned with the close
monitoring of conditionally discharged 'restricted' patients. As a matter of
good practice, the supervisors of restricted patients should keep in close
touch with the Home Office officials in C3, who may often be able to give
helpful information and advice. The Home Office also issues a useful set of
guidance notes for supervising psychiatrists (Home Office/Department of
Health and Social Security, 1987).
In terms of reconvictions, conditionally discharged patients with
'psychopathic disorder' have higher reoffending rates than patients with
other forms of mental disorder (Home Office, 19936). Bailey & MacCulloch
(1992) followed up a series of 112 patients discharged from England's Park
Lane secure 'special hospital' between 1974 and 1989. The follow-up
periods varied between 9 months and 14 years, (mean 6 years). During the
follow-up period, 13 of the 50 patients with 'psychopathic disorder' (26%)
were reconvicted of a serious criminal offence, compared with six of the 61
patients with mental illness (10%). MacCulloch and his colleagues
reviewed in detail the cases of serious reoffending (MacCulloch et al.,
1993<2, 19936, 1994) and drew particular attention to the lack of individual
treatment programmes during the hospital admissions, aimed at deviant
sexual behaviour, in those patients who were reconvicted of serious sexual
offences.
Inpatient settings 333
Inpatient settings
Crisis admissions
While treatment and support for personality-disordered offenders is pri-
marily outpatient based, short-term admissions may have a useful role to
play in the management of crises. Short hospital admissions can be a useful
safety net when individuals are seriously at risk. These patients may need
admission when suffering from severe situationally determined depressive
reactions, for example.
Brief admissions for assessment may also be appropriate when personal-
ity-disordered offenders are seen urgently at the request of the police or
criminal courts. The offences may be the culmination of serious social and
psychological difficulties and the offender may present in severe distress,
sometimes associated with suicidal ideation and substance abuse. There
may be a complex array of problems and it may not be possible to make an
accurate assessment of the individual's needs, including whether there is a
significant psychiatric disorder, without a period of observation in hospital.
In England and Wales, a remand for assessment order (section 35 of the
Mental Health Act, 1983) may be used in such cases, but recommendations
for local hospital admission should take account of the nature of the alleged
offence, to ensure that the admission is acceptably safe and not associated
with a serious risk of absconding.
Inpatient therapeutic community treatment

Specialist inpatient treatment using a therapeutic community model is
provided in the UK at the Henderson Hospital, Sutton, Surrey. The unit
has 29 places for adults, and admission has to be on a voluntary basis
following a selection interview. The ability to verbalize feelings and
problems and to engage in a group setting are important for selection. The
majority of the patients admitted have histories of drug abuse, alcohol
abuse or criminal convictions, and the maximum length of stay is a year.
Residents are expected to participate in a highly structured programme of
therapeutic groups, community meetings, practical tasks and other activi-
ties, with the aim of achieving 'a therapeutic mixture of supportive
containment and "reality confrontation" ...' (Norton, 1992).
Secure units and UK 'special' hospitals

Personality-disordered offenders constitute a minority of admissions to
secure units and 'special hospitals' in the UK, and those who are admitted
are generally sent by the criminal courts. Such individuals may be
remanded by the courts for assessment (section 35 of the Mental Health
Act, 1983), or for treatment under an hospital order (section 37 of the
Mental Health Act, 1983). Those admitted to secure hospitals with the legal
category of 'psychopathic disorder' have generally committed serious
offences, and will have additional restriction orders (section 41, Mental
Health Act, 1983). Some patients with 'psychopathic disorder' who are
ready to be discharged from 'special hospitals' will be transferred to
'regional secure units' as a first step towards gradual rehabilitation in the
community.
The question of whether offenders with serious PDs should be recom-
mended for hospital and restriction orders under the provisions of mental
health legislation as an alternative to sentences of imprisonment is contro-
versial (Mawson, 1983; Grounds, 1987; Dell & Robertson, 1988; Chiswick,
1992). Although when grave offences are committed in the context of
certain serious mental disorders such as schizophrenia, detention in a
secure hospital is often generally accepted as an appropriate way of
providing treatment that will reduce the risk of re-offending, in the case of
PDs, the prospects of effective treatment are usually uncertain. This makes
the selection of personality-disordered offenders for admission to a secure
hospital very problematic. In practice, younger patients with significant
anxiety, depression and distress, and who appear motivated to change, tend
to be judged as most suitable, but there is little satisfactory empirical
evidence about the factors that predict good and bad responses to clinical
interventions. But the specific treatment programmes that have been
developed use psychological methods that require the patient's active
participation and commitment, so that motivation and capacity for psy-
chological work is therefore an important practical consideration.
Grounds et al (1987) described the work of one unit in Broadmoor
Hospital (a UK 'special hospital') specializing in the treatment of young
men with 'psychopathic disorder'. The unit provides individual and group
psychotherapy using psychodynamic and cognitive/behavioural principles,
and a central focus is on seeking psychological understanding of the
patient's offence.
The general criterion for admission to a secure 'special hospital' is that
the patient constitutes a 'grave and immediate' danger to the public, and
that the patient could not be safely managed in less than maximum security.
In the UK each of the 'special hospitals' (Broadmoor, Rampton and
Ashworth Hospitals) covers a particular geographical region, and appli-
cations for admission are considered by an admissions panel in each
hospital.
Prisons
In the UK, it has been recommended that mentally-disordered offenders
who need treatment should generally receive this in facilities that are
outside the criminal justice system (Home Office, 1990). Imprisonment does
not have treatment as its primary purpose, and psychiatrists and sentencers
cannot assume that specific treatment will be provided if an offender is
sentenced to imprisonment. The amount of psychiatric input to prison
establishments is highly variable. Some prison health service doctors have
psychiatric experience and some establishments have visiting psychiatrists,
but many do not.
The frequency of violent and aggressive behaviour amongst prisoners is
related, in part, to the character of the prison regime (Bottoms, 1992), and
special prison units with radically different regimes have been developed in
the UK with the aim of dealing in a more humane and effective way with
'disruptive' prisoners, for example the 'Barlinnie Unit' in Scotland. In
England, Grendon Underwood Prison provides a specialist regime using a
therapeutic community approach in which prisoners engage in demanding,
confrontative group therapy. Although evidence is lacking that Grendon
has a substantial impact on re-offending rates, it has been argued that for
some well-motivated men the experience of Grendon does provide a unique
opportunity for change (Robertson & Gunn, 1987).
Sexual offenders are a group particularly likely to be victimized in prison
and may be isolated from other prisoners for their own protection. In recent
years new efforts have been made in the UK to introduce treatment
programmes for sex offenders, but only for those serving sentences of 4
years or more, and these programmes are available only in a limited number
of establishments (Home Office, 1991). Prisoners serving less than 4 years
may not receive any specialist help.
Assessments and reports for legal proceedings

The principles to be adopted when making assessments and writing reports
for legal proceedings are essentially the same, irrespective of the particular
issue being considered. The patient should be advised at the outset about
the purpose of the assessment, the uses to which it may be put and the likely
circulation of the report. The person should understand that what is said in
the interview may not remain strictly confidential because the report may be
seen by a variety of parties in the legal proceedings. Participation in the
assessment has to be voluntary.
It is essential that any report is based on full information. If the patient is
facing criminal proceedings, the reporting psychiatrist should see all the
prosecution witness statements. The patient's background history should
also be corroborated, if possible, by an independent informant. Reports
should not be undertaken if all relevant information is not disclosed. In the
criminal courts, if a report is based solely on what the defendant says
without account being taken of other evidence, the report is unlikely to
carry weight or credibility.
Reports should be clear and well organized in their layout and content.
Customarily there is an introductory paragraph outlining the purpose and
information sources of the report, followed by a summary of the relevant
background history, thefindingson psychiatric examination, and, finally,
the concluding opinion on the questions at issue. Accuracy, clarity, and
avoidance of technical vocabulary or emotive statements are necessary.
The concluding opinions should be substantiated with a clear explanation
of the reasons for the conclusions.
Psychiatric reports on personality-disordered offenders may be needed
for a variety of purposes. Most commonly, reports will be requested for use
in criminal proceedings when questions are raised about the person's state
of mind at the time of the alleged offence, or when advice is sought in
relation to a 'treatment disposal'. In England and Wales, reports may have
to be prepared for Mental Health Review Tribunals, and in such cases the
key questions are likely to be whether the patient still suffers from any form
of mental disorder warranting continued detention in hospital for treat-
ment, and whether they would pose a risk to others if discharged. Risk is
also likely to be a key issue when reporting in other contexts on people with
histories of offending, for example, for the parole board or for social
services departments in child protection cases.
In serious criminal cases the judgement about whether to recommend a
hospital order for subjects with PDs is, as noted above, a difficult one. If the
offender is regarded as being within the legal category of 'psychopathic
disorder', the requirement that treatment is likely to alleviate or prevent a
deterioration in their condition has to be fulfilled if an hospital order is to be
made. While, from a clinical viewpoint, motivation for psychological
treatment is clearly important, it should be recognized that the courts may
interpret a 'treatability' criterion more broadly than many clinicians. In the
'Andrews' case in the UK (R. v. Cannons Park Mental Health Review
Tribunal. The Times Law Report 2 March 1994) it was suggested that the
legal criterion of treatability would be fulfilled provided that hospital
treatment would prevent deterioration. It would not have to alleviate the
condition. Also, the absence of a patient's willingness to accept a specific
form of treatment would not prevent the patient being regarded as

'treatable', if there was a reasonable prospect of benefit from treatment at
some point in the future.
In cases of doubt, one possibility is to recommend an interim hospital
order (under section 38 of the Mental Health Act, 1983), which would allow
a trial period in hospital for 3 or 6 months before the court has to make a
final decision about whether to impose a 'hospital order' under the
provision of section 37 of the Mental Health Act, 1983.
Psychiatrists should also be aware of the possible consequences of
reporting that a personality-disordered offender is not treatable and
therefore unsuitable for a Mental Health Act order. Psychiatric evidence of
untreatable PD may be used by sentencers as a reason for considering a
longer-than-normal custodial sentence under the provisions of the Crimi-
nal Justice Act 1991 for offenders convicted of violent or sexual offences. In
the most serious offences, such evidence may be used as an argument for a
life sentence. The Court of Appeal has confirmed that the life sentence
should be reserved for people who have committed offences of substantial
gravity and who appear to be suffering from some disorder of personality or
instability of character that makes them likely to commit grave offences in
the future if left at large or released from a fixed term of imprisonment. (See
R. v. Hodgson (1968) 52 Cr. App. R. 113, and R. v. Billam (1986) Cr. App.
R. (S.) 347.)
In reports considering possible discharge from secure hospitals of
offenders with 'psychopathic disorder', three aspects merit particular
attention. Firstly, any current evidence of disorder and how much the
patient has changed psychologically. Secondly, if the patient was detained
because of a serious offence, there should be consideration of how well the
offence is psychologically understood, and how much insight the patient
has into the important factors that contributed to the offence. Thirdly, how
'supervisable' will the patient be? Do they recognize their need for
supervision? Are they likely to disclose openly how they are managing and
any problems that may arise? Do they recognize the difficulties that may lie
ahead and will they use the supervisory relationship well?
Assessment of risk depends predominantly on history. There needs to be
a thorough and detailed understanding of the patient's past history of
offending, the context in which it arose, its precursors and the patient's
psychiatric condition at the time. In looking to the future, and to the task of
ongoing supervision, what should be looked for is the likelihood of any
repetition of the context in which past offending occurred. The precise
factors of importance will vary in individual cases. For example, in one case
338 Managemen t of offenders with personality disorders
a key factor may be heavy alcohol intake, while in another it may be the
establishment of a sexual relationship in which pathological jealousy begins
to develop.
A useful checklist of key questions for the clinician when assessing the
risk of future violent behaviour was given by Monahan (1981):
1. Is it a prediction of violent behaviour that is being requested?
2. Am I professionally competent to offer an estimate of the probability of
future violence?
3. Are any issues of personal or professional ethics involved in this case?
4. Given my answers to the above questions, is this case an appropriate
one in which to offer a prediction?
5. What events precipitated the question of the person's potential for
violence being raised, and in what context did these events take place?
6. What are the person's relevant demographic characteristics?
7. What is the person's history of violent behaviour?
8. What is the base rate of violent behaviour among individuals of this
person's background?
9. What are the sources of stress in the person's current environment?
10. What cognitive and affective factors indicate that the person may be
predisposed to cope with stress in a violent manner?
11. What cognitive and affective factors indicate that the person may be
predisposed to cope with stress in a non-violent manner?
12. How similar are the contexts in which the person has used violent
coping mechanisms in the past to the contexts in which the person will
be likely to function in the future?
13. In particular, who are the likely victims of the person's violent
behaviour, and how available are they?
14. What means does the person possess to commit violence?
Social management
One of the key features of PDs, in particular when criminal offences are
involved, is their damaging and disabling effects on social functioning. Both
in their past histories and current situations, personality-disordered
patients may have serious difficulties in their family and social relation-
ships, and the clinical management has to include consideration of the
social context. Interventions may be needed to try to facilitate the patient's
social stability and, if possible, to prevent and mitigate socially damaging
outcomes.
Social management 339
Relatives and partners may find the patient's behaviour intolerable or
inexplicable, and they may also have unrealistic hopes or expectations
about the person's ability to change, and the influence of treatment. They
often need to be provided with explanation and support to develop a
realistic appraisal of the nature of the patient's disability, together with
guidance about how to manage difficulties and to protect themselves.
Parents may have particular problems of unrealistic guilt and self-blame in
relation to their children and they need sensitive advice and reassurance.
In cases where families and partners have been victims of violence, it is
vital that they are given advice about possible future risks, and where this is
significant they may need information about obtaining legal assistance and
about places of refuge.
Patients with PDs may also pose problems for other agencies such as
employers and housing authorities. The loss of stable accommodation and
employment may put some patients at increased risk of depressive symp-
toms, of offending or of other disturbed behaviour, and it may be important
to invest time in liaising with, and providing support to, these agencies,
provided that the patient gives permission for disclosure of the relevant
personal information. Residential provision may be difficult to secure when
there are worries about the possibility of unstable and hazardous behav-
iour, but the knowledge that psychiatric backup with rapid intervention
(for example, by a short crisis admission) will be provided may assist in
securing suitable residential placements.
The most disorganized and chaotic individuals with PDs are likely to
have recurrent breakdowns in their relationships, work and accommo-
dation, and clinical staff need to maintain an attitude that combines realism
and continuing commitment. Failure should not necessarily be greeted with
surprise and refusal of further help. A long-term perspective should be
maintained, as it is unrealistic to expect major short-term psychological
change, but even when there is slow improvement over one or two decades,
and clinical interventions over that time have helped to avoid serious harm
to the patient or others, much will have been achieved.
References
Abraham, K. (1921). Contributions to the theory of the anal character. In: K.

Abraham, On Character and Libido Development. New York: Basic Books,
published in 1966.
Abraham, K. (1924). The influence of oral eroticism on character formation. In:
Selected Papers of Karl Abraham, trans. D. Bryan & A. Strachy. New York:
Basic Books, published in 1948.
Adler, D.A., Drake, R.E. & Teague, G.B. (1990). Clinicians' practices in
personality assessment. Comprehensive Psychiatry, 31, 125-33.
Adler, G. (1981). The borderline-narcissistic personality disorder continuum.
American Journal of Psychiatry, 138, 46-50.
Adler, G. (1985). Borderline Psychopathology and its Treatment. New York:
Jason Aronson.
Adler, G. & Buie, D.H. (1979). Aloneness and borderline psychopathology: the
possible relevance of child development issues. InternationalJournal of
Psychoanalysis, 60, 83-96.
Akhtar, S. (1987). Schizoid personality disorder: a synthesis of developmental,
dynamic, and descriptive features. American Journal of Psychotherapy, 41,
499-518.
Akhtar, S. & Thomson, J.A. (1982). Overview: narcissistic personality disorder.
Akiskal, H.S. (1983). Dysthymic disorder: psychopathology of proposed chronic
depressive subtypes. American Journal of Psychiatry, 140, 11-20.
Akiskal, H.S. (1991). Chronic depression. Bulletin of the Menninger Clinic, 55,
156-71.
Akiskal, H.S., Hirschfeld, R.M.A. & Yerevanian, B.I. (1983). The relationship of
personality to affective disorders: a critical review. Archives of General
Psychiatry, 40, 801-10.
Akiskal, H.S., Rosenthal, T.L. & Haykal, R.F. (1980). Characterological
depressions. Clinical and sleep EEG findings separating subaffective
dysthymias from character spectrum disorders. Archives of General
Alden, L. (1989). Short-term structured treatment for avoidant personality
disorder. Journal of Consulting and Clinical Psychology, 56, 75664.
Alexander, F. (1930). The neurotic character. International Journal of
340
References 341
Allebeck, P. & Allgulander, C. (1990a). Psychiatric diagnoses as predictors of
suicide. British Journal of Psychiatry, 157, 339^4.
Allebeck, P. & Allgulander, C. (19906). Suicide among young men: psychiatric
illness, deviant behaviour and substance abuse. Ada Psychiatrica
Scandinavica, 81, 565-70.
Allebeck, P., Allgulander, C. & Fisher, L.D. (1988). Predictors of completed
suicide in a cohort of 50465 young men: role of personality and deviant
behaviour. British Medical Journal, 297, 176-8.
Allport, G.W. (1937). Personality: A Psychological Interpretation. New York:
Holt.
Aim, P.O., Aim, M., Humble, K. et al (1994). Criminality and platelet
monoamine oxidase activity in former juvenile delinquents as adults. Ada
Psychiatrica Scandinavica, 89, 41-5.
Alnaes, R. & Torgersen, S. (1988a). DSM-III symptom disorders (Axis I) and
personality disorders (Axis II) in an outpatient population. Acta
Alnaes, R. & Torgersen, S. (19886). The relationship between DSM-III symptom
disorders (Axis I) and personality disorders (Axis II) in an outpatient
population. Acta Psychiatrica Scandinavica, 78, 485-92.
Alnaes, R. & Torgersen, S. (1990). DSM-III personality disorders among
patients with major depression, anxiety disorders and mixed conditions.
Journal of Nervous and Mental Disease, 178, 693-8.
Alnaes, R. & Torgersen, S. (1991). Personality and personality disorders among
patients with various affective disorders. Journal of Personality Disorders, 5,
107-21.
American Psychiatric Association (1952). Diagnostic and Statistical Manual of
Mental Disorders, 1st edn. Washington DC: American Psychiatric
Association.
Mental Disorders, 2nd edn. (DSM-II). Washington DC: American
Psychiatric Association.
Mental Disorders, 3rd edn. (DSM-III). Washington DC: American
Psychiatric Association.
Mental Disorders, 3rd edn, revised, (DSM-III-R). Washington DC:
American Psychiatric Association.
Mental Disorders, 4th edn. Washington DC: American Psychiatric
Association.
Andreoli, A., Gressot, G. & Aapro, N. (1989). Personality disorders as a
predictor of outcome. Journal of Personality Disorders, 3, 307-20.
Andrews, G. (1991). The evaluation of psychotherapy. Current Opinion in
Andrews, G. (1993). The essential psycho therapies. British Journal of Psychiatry,
162,447-51.
Andrews, G., Neilson, M., Hunt, C. et al. (1990). Diagnosis, personality and the
long-term outcome of depression. British Journal of Psychiatry, 157, 13-18.
Andy, O.J. (1975). Thalamotomy for psychopathic behaviour. Southern Medical
Journal, 68, 437^42.
Argyle, M., Bryant, B.M. & Trower, P. (1974). Social skills training and
342 References
psychotherapy: a comparative study. Psychological Medicine, 4, 435-43.

August, G.J., Steward, M.A. & Holmes, C.S. (1983). A four-year follow-up of
hyperactive boys with and without conduct disorder. British Journal of
Baer, L., Jenike, M.A., Black, D.W. et al. (1992). Effect of axis II diagnoses on
treatment outcome with clomipramine in 55 patients with obsessive-
compulsive disorder. Archives of General Psychiatry, 49, 862-6.
Baer, L., Jenike, M.A., Ricciardi, J.N. et al. (1990). Standardized assessment of
personality disorders in obsessive-compulsive disorder. Archives of General
Bailey, J. & MacCulloch, M. (1992). Characteristics of 112 cases discharged
directly to the community from a new special hospital and some
comparisons of performance. Journal of Forensic Psychiatry, 3, 91-112.
Bardenstein, K.K. & McGlashan, T.H. (1988). The natural history of a
residentially treated borderline sample: Gender differences. Journal of
Personality Disorders, 2, 69-83.
Barley, W.D., Buie, S.E., Peterson, E.W. et al (1993). Development of an
inpatient cognitive-behavioral treatment program for borderline personality
disorder. Journal of Personality Disorders, 7, 23240.
Barnard, M.A. (1993). Needle sharing in context: patterns of sharing among men
and women injectors and HIV risks. Addiction, 88, 805-12.
Baron, M. (1981). Schedule for Interviewing Borderlines. New York: New York
State Psychiatric Institute.
Baron, M., Gruen, R., Asnis, L. & Lord, S. (1985). Familial transmission of
schizotypal and borderline personality disorders. American Journal of
Bass, C. & Benjamin, S. (1993). The management of chronic somatisation. British
Journal of Psychiatry, 162, 472-80.
Bateman, A.W. (1993). Personality disorders. Current Opinion in Psychiatry, 6,
205-9.
Beattie, J.O., Day, R.E., Cockburn, F. & Gary, R.A. (1983). Alcohol and the
fetus in the west of Scotland. British Medical Journal, 287, 17-20.
Beck, A.T. & Freeman, A. (1990). Cognitive Therapy of Personality Disorders
(with associates). New York: The Guilford Press.
Bellak, L. (1985). ADD psychosis as a separate entity. Schizophrenia Bulletin, 11,
523-7.
Benjamin, J., Silk, K.R., Lohr, N.E. & Westen, D. (1989). The relationship
between borderline personality disorder and anxiety disorders. American
Journal of Ortho-psychiatry, 59, 461-7.
Bentall, R.P., Claridge, G.S. & Slade, P.D. (1989). The multidimensional nature
of schizotypal traits: a factor analytic study with normal subjects. British
Journal of Psychology, 28, 363-75.
Berelowitz, M. & Tarnopolsky, A. (1993). The validity of borderline personality
disorder: an updated review of recent research. In: Personality Disorder
Reviewed, ed. P. Tyrer & G. Stein, pp. 90-112. London: Gaskell.
Bergeman, C.S., Plomin, R., McClearn, G.E. et al (1988). Genotype-
environment interaction in personality development: identical twins reared
apart. Psychology and Aging, 3, 399-406.
Berlin, F.S. & Meinecke, C.F. (1981). Treatment of sex offenders with
antiandrogenic medication. American Journal of Psychiatry, 138, 601-7.
Bernstein, D. (1993). Childhood precursors of adolescent personality disorders.
References 343
Paper presented at the Third International Congress on the Disorders of
Personality. Cambridge, Mass.
Berrios, G.E. (1993). European views on personality disorders: a conceptual
history. Comprehensive Psychiatry, 34, 14-30.
Beumont, P.J.V., George, G.C.W. & Smart, D.F. (1976). 'Dieters' and 'vomiters
and purgers' in anorexia nervosa. Psychological Medicine, 6, 617-22.
Bezirganian, S., Cohen, P. & Brook, J.S. (1993). The impact of mother-child
interaction on the development of borderline personality disorder. American
Journal of Psychiatry, 150, 1836-^2.
Biederman, J. (1988). Pharmacologic treatment of adolescents with Affective
Disorders and Attention Deficit Disorder. Psychopharmacology Bulletin, 24,
81-7.
Biederman, J., Faraone, S.V., Spencer, T. et al. (1993). Patterns of psychiatric
comorbidity, cognition, and psychosocial functioning in adults with
attention deficit hyperactivity disorder. American Journal of Psychiatry, 150,
1792-8.
Biederman, J., Newcorn, J. & Sprich, S. (1991). Comorbidity of attention deficit
hyperactivity disorder with conduct, depressive, anxiety and other disorders.
Bion, W.R. (1961). Experiences in Groups and Other Papers. London: Tavistock
Publications.
Birtchnell, J. (1991). The measurement of dependence by questionnaire. Journal
of Personality Disorders, 5, 281-95.
Black, D.W., Bell, S. & Hulbert, J. (1988). The importance of Axis II in patients
with major depression. Journal of Affective Disorders, 14, 115-22.
Black, D.W., Noyes, R., Pfohl, B. et al. (1993). Personality disorder in obsessive-
compulsive volunteers, well comparison subjects, and their first-degree
relatives. American Journal of Psychiatry, 150, 1226-32.
Black, D.W., Yates, W.R., Noyes, R. et al. (1989). DSM-III personality disorder
in obsessive-compulsive study volunteers. A controlled study. Journal of
Blackburn, R. (1968). Personality in relation to extreme aggression in psychiatric
offenders. British Journal of Psychiatry, 114, 821-8.
Blackburn, R. (1971). Personality types among abnormal homicides. British
Journal of Criminology, 11,1431.
Blackburn, R. (1975). An empirical classification of psychopathic personality.
British Journal of Psychiatry, 111, 456-60.
Blackburn, R. (1984). The person and dangerousness. In: Psychology and Law,
ed. D J . Muller, D.E. Blackman & A.J. Chapman, pp. 102-11. Chichester:
Wiley.
Blackburn, R. (1986). Patterns of personality deviation amongst violent
offenders: replication and extension of an empirical taxonomy. British
Journal of Criminology, 26, 254-69.
Blackburn, R. (1988). On moral judgements and personality disorders. The myth
of psychopathic personality revisited. British Journal of Psychiatry, 153,
505-12.
Blackburn, R. (1989). Psychopathy and personality disorder in relation to
violence. In: Clinical Approaches to Violence, ed. K. Howells & C.R. Hollin,
pp. 61-88. Chichester: John Wiley & Sons.
Blackburn, R. (1992). Clinical programmes with psychopaths. In: Clinical
Approaches to the Mentally Disordered Offender, ed. K. Howells & C.
344 References
Hollin. Chichester: John Wiley.
Blackburn, R. (1993). Biological Correlates of Antisocial Behaviour. In: The
Psychology of Criminal Conduct, pp. 136-59. Chichester: John Wiley &
Sons.
Bland, R.C., Newman, S.C. & Orn, H. (1988). Lifetime prevalence of psychiatric
disorders in Edmonton. Ada Psychiatrica Scandinavica, 11 (suppl. 338),
24-32.
Blashfield, R., Sprock, J., Pinkston, K. & Hodgin, J. (1985). Exemplar
prototypes of personality disorder diagnoses. Comprehensive Psychiatry, 26,
11-21.
Blazer, D., George, L.K., Landerman, R. et al. (1985). Psychiatric disorders. A
rural/urban comparison. Archives of General Psychiatry, 42, 651-6.
Bleuler, E. (1908). Textbook of Psychiatry, trans. A.A. Brill. New York: The
Macmillan Company, published in 1924.
Bleuler, E. (1911). Dementia Praecox or the Group of Schizophrenias. New York:
IUP, published in 1950.
Bloch, S. (1979). Group psychotherapy. In: An Introduction to the
Psychotherapies, ed. S. Bloch. Oxford: Oxford University Press.
Blum, H.M. & Marziali, E. (1988). Time-limited, group psychotherapy for
borderline patients. Canadian Journal of Psychiatry, 33, 364-9.
Blumer, D. & Heilbronn, M. (1982). Chronic pain as a variant of depressive
disease. The pain-prone disorder. Journal of Nervous and Mental Disease,
170, 381-406.
Blumstein, A. & Cohen, J. (1979). Estimation of individual crime rates from
arrest records. Journal of Criminal Law and Criminology, 70, 561-85.
Bodlund, O., Kullgren, G., Sundbom, E. & Hojerback, T. (1993). Personality
traits and disorders among trans-sexuals. Ada Psychiatrica Scandinavica,
88, 322-7.
Bohman, M., Cloninger, R., Sigvardsson, S. & Von-Knorring, A.L. (1987). The
genetics of alcoholism and related disorders. Journal of Psychiatric Research,
21, 447-52.
Bonnet, K.A. & Redford, H.R. (1982). Levodopa in borderline disorders.
Archives of General Psychiatry, 39, 862.
Bornstein, R.F., Klein, D.N., Mallon, J.C. & Slater, J.F. (1988). Schizotypal
personality disorder in an outpatient population: incidence and clinical
characteristics. Journal of Clinical Psychology, 44, 322-5.
Bottoms, A.E. (1992). Violence and disorder in long-term prisons: the influence
of institutional environments. Criminal Behaviour and Mental Health, 2,
126-36.
Bouchard, T.J., Lykken, D.T., McGue, M. et al. (1990). Sources of human
psychological differences: the Minnesota study of twins reared apart.
Science, 250, 223-8.
Bowden, P. (1991). Treatment: use, abuse and consent. Criminal Behaviour and
Mental Health, 1, 130-41.
Boyce, P., Hadzi-Pavlovic, D., Parker, G. et al. (1989). Depressive type and state
effects on personality measures. Ada Psychiatrica Scandinavica, 81,
197-200.
Boyce, P., Parker, G., Barnett, B. et al. (1991). Personality as a vulnerability
factor to depression. British Journal of Psychiatry, 159, 106-14.
British Association of Psychotherapists (1993). Information leaflet. 37,
Mapesbury Road, London NW2 4HJ, UK.
References 345
Bronisch, T. & Klerman, G.L. (1991). Personality functioning: change and
stability in relationship to symptoms and psychopathology. Journal of
Brooks, R.B., Baltazar, P.L., McDowell, D.E. et al (1991). Personality disorders
co-occurring with panic disorder with agoraphobia. Journal of Personality
Disorders, 5, 328-36.
Brooks, R.B., Baltazar, P.L. & Munjack, D.J. (1989). Co-occurrence of
personality disorders with panic disorder, social phobia, and generalized
anxiety disorder: a review of the literature. Journal of Anxiety Disorders, 3,
259-85.
Brooner, R.K., Bigelow, G.E., Strain, E. & Schmidt, C.W. (1990). Intravenous
drug abusers with antisocial personality disorder: increased HIV risk
behavior. Drug and Alcohol Dependence, 26, 39-44.
Brooner, R.K., Greenfield, L., Schmidt, C.W. & Bigelow, G.E. (1993). Antisocial
personality disorder and HIV infection among intravenous drug abusers.
Brooner, R.K., Schmidt, C.W., Felch, L.J. & Bigelow, G.E. (1992). Antisocial
behaviour of intravenous drug abusers: implications for diagnosis of
antisocial personality disorder. American Journal of Psychiatry, 149, 482-7.
Brown, G.W. & Harris, T.O. (1978). Social Origins of Depression: A Study of
Psychiatric Disorders in Women. London: Tavistock Publications.
Brown, S.L., Svrakic, D.M., Przybeck, T.R. & Cloninger, C.R. (1992). The
relationship of personality to mood and anxiety states: a dimensional
approach. Journal of Psychiatric Research, 26, 197-211.
Brownstone, D.Y. & Swaminath, R.S. (1989). Violent behaviour and psychiatric
diagnosis in female offenders. Canadian Journal of Psychiatry, 34, 190-4.
Burchard, J.D. & Wood, T.W. (1982). Crime and delinquency. In: International
Handbook of Behaviour Modification and Therapy, ed. A.S. Bellack, M.
Herson & A.E. Kazdin. New York: Plenum.
Burnham, D.L., Gladstone, A.I. & Gibson, R.W. (1969). Schizophrenia and the
Need-Fear Dilemma. New York: International Universities Press.
Bursten, B. (1989). The relationship between narcissistic and antisocial
personalities. Psychiatric Clinics of North America, 12, 571-84.
Buss, A.H. & Durkee, A. (1957). An inventory for assessing different kinds of
hostility. Journal of Consulting Psychology, 21, 343-9.
Buss, A.H. & Plomin, R. (1986) The EAS approach to temperament. In: The
Study of Temperament: Changes, Continuities and Challenges, ed. R. Plomin
& J. Dunn. Hillside NJ: Lawrence Erlbaum.
Buydens-Branchey, L., Branchey, M.H. & Noumair, D. (1989). Age of
alcoholism onset. Relationship to psychopathology. Archives of General
Cadoret, R.J. (1978). Psychopathology in adopted-away offspring of biologic
parents with antisocial behaviour. Archives of General Psychiatry, 35,
176-84.
Cadoret, R.J., O'Gorman, T.W., Troughton, E. & Heywood, E. (1985).
Alcoholism and antisocial personality. Interrelationships, genetic and
environmental factors. Archives of General Psychiatry, 42, 161-7.
Cadoret, R.J. & Stewart, M.A. (1991). An adoption study of attention deficit/
hyperactivity/aggression and their relationship to adult antisocial
personality. Comprehensive Psychiatry, 32, 73-82.
Cadoret, R.J., Troughton, E., Bagford, J. & Woodworth, G. (1990). Genetic and
346 References
environmental factors in adoptee antisocial personality. European Archives

of Psychiatry and Neurological Science, 239, 23140.
Cappe, R.F. & Alden, L.E. (1986). A comparison of treatment strategies for
clients functionally impaired by extreme shyness and social avoidance.
Journal of Consulting and Clinical Psychology, 54, 796-801.
Carney, F.L. (1977). Outpatient treatment of the aggressive offender. American
Journal of Psychotherapy, 31, 265-74.
Carroll, K.M. & Rounsaville, B.J. (1993). History and significance of childhood
attention deficit disorder in treatment-seeking cocaine abusers.
Comprehensive Psychiatry, 34, 75-82.
Casey, P. (1988). The epidemiology of personality disorder. In: Personality
Disorders, ed. P. Tyrer, pp. 7481. London: Wright.
Casey, P.R., Dillon, S. & Tyrer, P.J. (1984). The diagnostic status of patients
with conspicuous psychiatric morbidity in primary care. Psychological
Medicine, 14, 673-81.
Casey, P.R. & Tyrer, P.J. (1986). Personality, functioning and symptomatology.
Journal of Psychiatric Research, 20, 363-74.
Casey, P.R. & Tyrer, P. (1990). Personality disorder and psychiatric illness in
general practice. British Journal of Psychiatry, 156, 261-5.
Casper, R.C., Eckert, F.D., Halmi, K.A. et al (1980). Bulimia. Archives of
General Psychiatry, 37, 1030-5.
Castaneda, R. & Franco, H. (1985). Sex and ethnic distribution of borderline
personality disorder in an inpatient sample. American Journal of Psychiatry,
142, 1202-3.
Cavior, H.E. & Schmidt, A.A. (1978). Test of the effectiveness of a differential
treatment strategy at the Robert F. Kennedy Centre. Criminal Justice and
Behaviour, 5, 131-9.
Chambless, D.L., Renneberg, B., Goldstein, A. & Gracely, E.J. (1992). MCMI-
diagnosed personality disorders among agoraphobic outpatients. Journal of
Anxiety Disorders, 6, 193-211.
Chandrasekaran, R., Goswami, U. & Sivakumar, C. (1994). Hysterical neurosis -
a follow-up study. Acta Psychiatrica Scandinavica, 89, 78-80.
Chaplin, W.F., John, O.P. & Goldberg, L.R. (1988). Conceptions of states and
traits: Dimensional attributes with ideals as prototypes. Journal of
Personality and Social Psychology, 54, 541-57.
Chapman, L.J., Chapman, J.P. & Raulin, M.L. (1978). Body-image aberration in
schizophrenia. Journal of Abnormal Psychology, 87, 399-407.
Chiswick, D. (1992). Compulsory treatment of patients with psychopathic
disorder: an abnormally aggressive or seriously irresponsible exercise?
Criminal Behaviour and Mental Health, 2, 106-13.
Christiansen, K.O. (1974). The genesis of aggressive criminality. Implications of
a study of crime in a Danish twin study. In: Determinants and Origins of
Aggressive Behaviour, ed. J. de Wit & W.W. Hartup. The Hague: Mouton.
Christiansen, K.O. (1977). A preliminary study of criminality among twins. In
Biosocial Bases of Criminal Behavior, ed. S.A. Mednick & K.O. Christiensen.
New York: Gardiner Press.
Claridge, G.S. & Broks, P. (1984). Schizotypy and hemisphere function. I.
Theoretical considerations and the measurement of schizotypy. Personality
and Individual Differences, 5, 633-48.
Clark, D.H. & Myers, K. (1970). Themes in a therapeutic community. British
References 347
Clark, L.A. (1989). Preliminary Manual for the Schedule for Normal and
Abnormal Personality (SNAP). Dallas TX: Department of Psychology,
Southern Methodist University.
Clark, L.A. (1992). Resolving taxonomic issues in personality disorders: the
value of large-scale analyses of symptom data. Journal of Personality
Cleckley, H. (1941). The Mask of Sanity. London: Henry Kimpton.
Cleckley, H. (1976). The Mask of Sanity, 5th edn, St. Louis, MO: Mosby.
Cloninger, C.R. (1978). The link between hysteria and sociopathy. In: Psychiatric
Diagnosis: Exploration of Biological Predictors, ed. H.S. Akiskal & W.L.
Webb, pp. 189-218. New York: Spectrum Press.
Cloninger, C.R. (1986). A unified biosocial theory of personality and its role in
the development of anxiety states. Psychiatric Developments, 4, 167-226.
Cloninger, C.R. (1987c). A systematic method for clinical description and
classification of personality variants. Archives of General Psychiatry, 44,
573-88.
Cloninger, C.R. (19876). Tridimensional Personality Questionnaire (TPQ).
Department of Psychiatry and Genetics, Washington University School of
Medicine, The Jewish Hospital of St. Louis: 216 S. Kingshighway Blvd. PO
Box 14109, St. Louis, MO 63178.
Cloninger, C.R. (1987c). Neurogenetic adaptive mechanisms in alcoholism.
Science, 236, 410-6.
Cloninger, C.R., Christiansen, K.O., Reich, T. & Gottesman, I.I. (1978).
Implications of sex differences in the prevalence of antisocial personality,
alcoholism, and criminality for familial transmission. Archives of General
Cloninger, C.R. & Gottesman, I.I. (1987). Genetic and environmental factors in
antisocial behaviour disorders. In: The Causes of Crime: New Biological
Approaches, ed. S.A.Mednick, T.F. Moffitt & S.A. Stack, pp. 92-109.
Cambridge: Cambridge University Press.
Cloninger, C.R. & Guze, S.B. (1970). Psychiatric illness and female criminality:
the role of sociopathy and hysteria in the antisocial woman. American
Cloninger, C.R., Martin, R.L., Guze, S.B. & Clayton, P. (1985). Diagnosis and
prognosis in schizophrenia. Archives of General Psychiatry, 42, 15-25.
Cloninger, C.R., Sigvardsson, S., Gilligan, S.B. et al. (1988). Genetic
heterogeneity and the classification of alcoholism. Advances in Alcohol and
Substance Abuse, 7, 3-16.
Cloninger, C.R., Svrakic, D.M. & Przybeck, T.R. (1993). A psychobiological
model of temperament and character. Archives of General Psychiatry, 50,
975-90.
Coccaro, E.F., Siever, L.J., Klar, H. et al. (1989). Serotonergic abnormalities in
patients with affective and personality disorders correlate with suicidal and
impulsive aggressive behavior. Archives of General Psychiatry, 46, 587-99.
Cohen, G.H., Garey, R.E., Evans, A. & Wilchinsky, M. (1980). Treatment of
heroin addicts: is the client-therapist relationship important? International
Journal of Addiction, 15, 207-14.
Cohn, L.D. (1991). Sex differences in the course of personality development: a
meta-analysis. Psychological Bulletin, 109, 252-66.
Coid, J.W. (1991). Psychiatric profiles of difficult/disruptive prisoners. In: Special
Units for Difficult Prisoners, ed. K. Bottomley & W. Hay. Hull: Centre for
348 References
Criminology and Criminal Justice, University of Hull.
Coid, J.W. (1992). DSM-III diagnosis in criminal psychopaths: a way forward.
Criminal Behaviour and Mental Health, 2, 78-94.
Coid, J.W. (1993a). Current concepts and classifications of psychopathic
disorder. In: Personality Disorder Reviewed, ed. P. Tyrer & G. Stein, pp.
113-164. London: Gaskell.
Coid, J.W. (19936). An affective syndrome in psychopaths with borderline
personality disorder? British Journal of Psychiatry, 162, 641-50.
Coid, J., Allolio, B. & Rees, L.H. (1983). Raised plasma metenkephalin in
patients who habitually mutilate themselves. Lancet, 2, 545-6.
Coid, J.W., Wilkins, J. & Coid, B. (1992). Self-mutilation in female remanded
prisoners II: a cluster analytic approach towards identification of a
behavioural syndrome. Criminal Behaviour and Mental Health, 2, 1-14.
Conte, H., Plutchik, R. & Karasu, T. (1980). A self-report borderline scale:
discriminative validity and preliminary norms. Journal of Nervous and
Mental Disease, 168, 428-35.
Conte, H.R., Plutchik, R., Picard, S. & Karasu, T.B. (1991). Can personality
traits predict psychotherapy outcome? Comprehensive Psychiatry, 32, 66-72.
Cornelius, J.R., Soloff, P.H., Perel, J.M. & Ulrich, R.F. (1991). A preliminary
trial of fluoxetine in refractory borderline patients. Journal of Clinical
Psychopharmacology, 11, 116-20.
Cornelius, J.R., Soloff, P.H., Perel, J.M. & Ulrich, R.F. (1993). Continuation
pharmacotherapy of borderline personality disorder with haloperidol and
phenelzine. American Journal of Psychiatry, 150, 1843-8.
Copas, J.B. & Whiteley, J.S. (1976). Predicting success in the treatment of
psychopaths. British Journal of Psychiatry, 129, 388-92.
Coryell, W., Winokur, G., Shea, T. et al. (1994). The long-term stability of
depressive subtypes. American Journal of Psychiatry, 151, 199-204.
Coryell, W.H. & Zimmerman, M. (1989). Personality disorder in the families of
depressed, schizophrenic and never-ill probands. American Journal of
Costa, P. (1993). Some implications of longitudinal personality research for the
study of personality disorders. Paper presented at the Third International
Congress on the Disorders of Personality. Cambridge, Mass.
Costa, P.T. & McCrae, R.R. (1992). The five-factor model of personality and its
relevance to personality disorders. Journal of Personality Disorders, 6,
343-59.
Cowdrey, R.W. & Gardner, D.L. (1988). Pharmacotherapy of borderline
personality disorder: alprazolam, carbamazapine, trifluoperazine and
tranylcypromine. Archives of General Psychiatry, 45, 111-9.
Craft, M.J. (1965a). Ten Studies into Psychopathic Personality. Bristol: John
Wright.
Craft, M. (19656). The natural history of psychopathic disorder. British Journal
of Psychiatry, 115, 39^4.
Craig, A.R., Franklin, J.A. & Andrews, G. (1984). A rating scale to measure
locus of control of behaviour. British Journal of Medical Psychology, 57,
173-80.
Creed, F. & Guthrie, E. (1993). Techniques for interviewing the somatising
patient. British Journal of Psychiatry, 162, 467-71.
Crow, J.J. (1986). The continuum of psychosis and its implications for the
structure of the gene. British Journal of Psychiatry, 149, 419-29.
References 349
Crowe, R.R. (1972). The adopted offspring of women criminal offenders - a
study of their arrest records. Archives of General Psychiatry, 27, 600-3.
Crowell, J.A., Waters, E., Kring, A. & Riso, L.P. (1993). The psychosocial
etiologies of personality disorders: what is the answer like? Journal of
Personality Disorders, Spring suppl., 118-28.
Cutting, J. (1985). The Psychology of Schizophrenia. London: Churchill
Livingstone.
Cutting, J., Cowen, P.J., Mann, A.H. & Jenkins, R. (1986). Personality and
psychosis: use of the Standardized Assessment of Personality. Acta
Dahl, A.A. (1993). The personality disorders: a critical review of family, twin and
adoption studies. Journal of Personality Disorders, Spring suppl. 86-99.
Dalen, P. & Hays, P. (1990). Aetiological heterogeneity of schizophrenia: the
problem and the evidence. British Journal of Psychiatry, 157, 119-22.
Dalgaard, O.S. & Kringlen, E. (1976). A Norwegian study of criminality. British
Journal of Criminology, 16, 213-32.
Dalkin, T., Murphy, P., Glazebrook, C. et al. (1994). Premorbid personality in
first-onset psychosis. British Journal of Psychiatry, 164, 202-7.
Davidson, P.R. (1984). Behavioral treatment for incarcerated sex offenders: post-
release outcome. Paper presented at Conference on Sex Offender Assessment
and Treatment, Kingston, Ontario. March.
Davies, W. (1989). The prevention of assault on professional helpers. In: Clinical
Approaches to Violence, ed. K. Howells & C.R. Hollin, pp. 311-28. New
York: John Wiley & Sons Inc.
De Jong, J., Virkkunen, M. & Linnoila, M. (1992). Factors associated with
recidivism in a criminal population. Journal of Nervous and Mental Disease,
180, 543-50.
Delito, J.A. & Stam, M. (1989). Psychopharmacological treatment of avoidant
personality disorder. Comprehensive Psychiatry, 30, 498-504.
Dell, S. & Robertson, G. (1988). Sentenced to Hospital. Oxford: Oxford
University Press,
de Mare, P.B. & Kreeger, L.C. (1974). Introduction to Group Treatments in
Psychiatry. London: Butterworths.
Department of Health (1990). 'Caring for people': The care programme
approach for people with a mental illness referred to the specialist
psychiatric services. Health Circular (90) 23/LASSL (90) 11. London:
Department of Health.
Department of Health (1994a). Draft guidance on the discharge of mentally
disordered people from hospital and their continuing care in the community.
London: Department of Health.
Department of Health (19946). Introduction of supervision registers for mentally
ill people from 1st April 1994. Health Service guidelines (94) 5. London:
Department of Health.
Department of Health and Social Security (1985). Mental Illness Hospitals and
Units in England. Results from the Mental Health Enquiry. Statistical
Bulletin, Government Statistical Service. London: HMSO.
Dieckmann, G., Schneider-Jonietz, B. & Schneider, M. (1988). Psychiatric and
neuropsychological findings after stereotactic hypothalamotomy, in cases of
extreme sexual aggressivity. Acta Neurochirurgica, Suppl. 44, 163-6.
Digman, J.M. (1990). Personality structure: emergence of the five-factor model.
Annual Review of Psychology, 41, 41740.
350 References
Dinwiddie, S.H., Reich, T. & Cloninger, C.R. (1992). Psychiatric comorbidity
and suicidality among intravenous drug users. Journal of Clinical Psychiatry,
53, 364-9.
Dinwiddie, S.H., Zorumski, C.F. & Rubin, E.H. (1987). Psychiatric correlates of
chronic solvent abuse. Journal of Clinical Psychiatry, 48, 334-7.
Docherty, J.P., Fiester, S.J. & Shea, T. (1986). Syndrome diagnosis and
personality disorder. In: Psychiatry Update, vol. 5, ed. A.J. Frances &
R.E.Halls, pp. 315-55. Washington, DC: American Psychiatric Press.
Dolan, B. & Coid, J. (1993). Psychopathic and Antisocial Personality Disorders.
London: Gaskell.
Dolan, B., Evans, C. & Norton, K. (1994). Disordered eating behaviour and
attitudes in female and male patients with personality disorders. Journal of
Dowson, J.H. (1977). The phenomenology of severe obsessive-compulsive
neurosis. British Journal of Psychiatry, 131, 75-82.
Dowson, J.H. (1987). MAO inhibitors in mental disease: their current status.
Journal of Neural Transmission, suppl. 23, 121-38.
Dowson, J.H. (1992a). DSM-III-R narcissistic personality disorder evaluated by
patients' and informants' self-report questionnaires. Comprehensive
Dowson, J.H. (19926). Assessment of DSM-III-R personality disorders by self-
report questionnaires: the role of informants and a screening test for co-
morbid personality disorders (STCPD). British Journal of Psychiatry, 161,
344-52.
Dowson, J.H. (1992c). Associations between self-induced vomiting and
personality disorder in patients with a history of anorexia nervosa. Ada
Dowson, J.H. (1994). DSM-III-R self-defeating personality disorder criteria
evaluated by self-report questionnaire. Acta Psychiatrica Scandinavica, 90,
32-7.
Dowson, J.H. & Berrios, G.E. (1991). Factor structure of DSM-III-R personality
disorders shown by self-report questionnaire: implications for classifying
and assessing personality disorders. Acta Psychiatrica Scandinavica, 84,
555-60.
Dubro, A., Wetzler, S. & Kahn, M.W. (1988). A comparison of three self-report
questionnaires for the diagnosis of DSM-III personality disorders. Journal
of Personality Disorders, 2, 256-66.
Duggan, C.F., Lee, A.S. & Murray, R.M. (1990). Does personality predict long-
term outcome in depression? British Journal of Psychiatry, 157, 19-24.
Easser, B. & Lesser, S. (1965). Hysterical character and psychoanalysis.
Psychoanalytic Quarterly, 34, 390-405.
Editorial (1994). Attention Deficit Hyperactivity Disorder in adults. American
Edwards, J.G. & Goldie, A. (1987). A ten-year follow-up study of Southampton
opiate addicts. British Journal of Psychiatry, 151, 679-83.
Elkin, L., Shea, T. & Watkins, J.T. (1989). National Institute of Mental Health
treatment of depression collaborative research program. General
effectiveness of treatments. Archives of General Pychiatry, 46, 971-82.
Ellison, J.M. & Adler, D.A. (1990). A strategy for the pharmacotherapy of
personality disorders. In: Treating Personality Disorders, ed. D.A. Adler, pp.
43-64. San Francisco: Jossey-Bass Inc.
References 351
Endicott, J. & Shea, M.T. (1989). Measurement of change in personality
disorders. Psychopharmacology Bulletin, 25, 572-7.
Ennis, J., Barnes, R. & Spenser, H. (1985). Management of the repeatedly
suicidal patient. Canadian Journal of Psychiatry, 30, 535-8.
Evans, C. & Lacey, J.H. (1992). Multiple self-damaging behaviour among
alcoholic women. British Journal of Psychiatry, 161, 643-7.
Eysenck, H.J. (1982). Personality, Genetics and Behavior. New York: Prager.
Eysenck, H.J. & Eysenck, S.B.G. (1964). Manual of the Eysenck Personality
Inventory. London: University of London Press.
Eysenck, H.J. & Eysenck, S.B.G. (1975). The EPQ. London: University of
London Press.
Fahy, T.A., Eisler, I. & Russell, G.F.M. (1993). Personality disorder and
treatment response in bulimia nervosa. British Journal of Psychiatry, 162,
765-70.
Faltus, F.J. (1984). The positive effect of alprazolam in the treatment of three
patients with borderline personality disorder. American Journal of
Famularo, R., Kinscherff, R. & Fenton, T. (1991). Post-traumatic stress disorder
among children clinically diagnosed as borderline personality disorder.
Famularo, R., Kinscherff, R. & Fenton, T. (1992). Psychiatric diagnoses of
abusive mothers. Journal of Nervous and Mental Disease, 180, 658-61.
Farrington, D.P. (1978). The family backgrounds of aggressive youths. In:
Aggression and Antisocial Disorder in Children, ed. M. Berger & D. Schaffer,
pp. 98-122. London: Pergamon Press.
Favazza, A.R. & Conterio, K. (1989). Female habitual self-mutilators. Acta
Fawcett, J. & Siomopouloo, J. (1971). Dextroamphetamine response as a
possible predictor of improvement with tricyclic therapy in depression.
Archives of General Psychiatry, 25, 2447.
Fenton, W.C. & McGlashan, T.H. (1989). Risk of schizophrenia in character
disordered patients. American Journal of Psychiatry, 146, 1280-84.
Ferguson, B. & Tyrer, P. (1988). Classifying personality disorder. In: Personality
Disorders: Diagnosis, Management and Course, ed. P. Tyrer, pp. 12-32.
London: Wright.
Ferguson, B. & Tyrer, P. (1989). Personality disorders. In: The Instruments of
Psychiatric Research, ed. C. Thompson. Chichester: John Wiley & Sons.
Ferguson, B. & Tyrer, P. (1991). Personality disorder: the flamboyant group.
Current Opinion in Psychiatry, 4, 200-4.
Feshbach, S. & Price, J. (1984). The development of cognitive competencies and
the control of aggression. Aggressive Behaviour, 10, 185-200.
Fiester, S.J. (1991). Self-defeating personality disorder: a review of data and
recommendations for DSM-IV. Journal of Personality Disorders, 5, 194-209.
Fiester, S.J. & Gay, H. (1991). Sadistic personality disorder: a review of data and
Foerster, A., Lewis, S., Owen, M. & Murray, R. (1991). Premorbid adjustment
and personality in psychosis. Effects of sex and diagnosis. British Journal of
Ford, M.R. & Widiger, T.A. (1989). Sex bias in the diagnosis of histrionic and
antisocial personality disorders. Journal of Consulting and Clinical
Psychology, 57, 301-5.
352 References
Foulkes, S.H. & Anthony, E.J. (1965). Group Psychotherapy, 2nd edn.
Harmondsworth UK: Penguin Books.
Frances, A. (1982). Categorical and dimensional systems of personality
diagnosis: a comparison. Comprehensive Psychiatry, 23, 516-526.
Frances, A.J., Widiger, T.A. & Fyer, M.R. (1990). The influence of classification
methods on comorbidity. In: Comorbidity of Mood and Anxiety Disorders,
ed. J.D. Maser & C.R. Cloninger, pp. 41-59. Washington, DC: American
Psychiatric Press.
Frank, J.D. (1974). The restoration of morale. American Journal of Psychiatry,
131, 271-5.
Frank, J.D. (1979). What is psychotherapy? In: An Introduction to the
Psychotherapies, ed. S. Bloch. Oxford: Oxford University Press.
Frankenburg, F.R. & Zanarini, M.C. (1993). Clozapine treatment of borderline
patients: a preliminary study. Comprehensive Psychiatry, 34, 402-5.
Frederiksen, L.W. & Rainwater, N. (1981). Explosive behaviour: a skill
development approach to treatment. In: Violent Behaviour: Social Learning
Approaches to Prediction, Management and Treatment, ed. R.B. Stuart, pp.
211-34. New York: Brunner/Mazel.
Freeman, C. & Tyrer, P. (1989). Research Methods in Psychiatry, ed. C. Freeman
& P. Tyrer. London: Gaskell (Royal College of Psychiatrists).
Freud, S. (1955). On narcissism: an introduction. In: Complete Psychological
Works, standard edn. pp. 135-243. London, UK: Hogarth.
Freud, S. (1957). The Economic Problem of Masochism, standard edn. vol. 14.
London, UK: Hogarth Press.
Freud, S. (1959). Character and anal-eroticism. In: The Standard Edition of the
Complete Psychological Works of Sigmund Freud, vol. 9, trans. J. Strachey.
London: Hogarth Press (original work published in 1908).
Freud, S. (1961). Complete Psychological Works, standard edn. vol. 19. London,
UK: Hogarth Press (original work published in 1924).
Freud, S. (1962). Introductory Lectures on Psychoanalysis. Lecture 21. In: The
Complete Works of Sigmund Freud, vol. 16, pp. 320-338. London: Hogarth
Press.
Freund, L.S., Reiss, A.L., Hagerman, R. & Vinogradov, S. (1992). Chromosome
fragility and psychopathology in obligate female carriers of the fragile X
chromosome. Archives of General Psychiatry, 49, 54-60.
Furby, L., Weinrott, M.R. & Blackshaw, L. (1989). Sex offender recidivism: a
review. Psychological Bulletin, 155, 3-30.
Gabbard, G.O. & Coyne, L. (1987). Predictors of response of antisocial patients
to hospital treatment. Hospital and Community Psychiatry, 38, 1181-5.
Gardner, D.L., Leibenluft, E., O'Leary, K.M. & Cowdry, R.W. (1991). Self-
ratings of anger and hostility in borderline personality disorder. Journal of
Nervous and Mental Diseases, 179, 157-61.
Gardner, D.L., Lucas, P.B. & Cowdry, R.W. (1990). CSF metabolites in
borderline personality disorder compared with normal controls. Biological
Garfinkel, P.E., Moldofsky, H. & Garner, D.M. (1980). The heterogeneity of
anorexia nervosa. Archives of General Psychiatry, 37, 1036-40.
Gartner, A.F., Marcus, R.N., Halmi, K. & Loranger, A.W. (1989). DSM-III-R
personality disorder in patients with eating disorders. American Journal of
Gelder, M., Gath, D. & Mayou, R. (1983). Oxford Textbook of Psychiatry.
References 353
Oxford: Oxford University Press.
George, A. & Soloff, P.H. (1986). Schizotypal symptoms in patients with
borderline personality disorders. American Journal of Psychiatry, 143,
212-15.
Gill, K., Nolimal, D. & Crowley, T J . (1992). Antisocial personality disorder,
HIV risk behavior amd retention in methadone maintenance therapy. Drug
Alcohol Dependence, 30, 247-52.
Glenn, S.W. & Nixon, S.J. (1991). Applications of Cloninger's subtypes in a
female alcoholic sample. Alcohol Clinical Experimental Research, 15, 851-7.
Goldberg, D.P. (1972). The detection of psychiatric illness by questionnaire.
Institute of Psychiatry Maudsley Monographs, no. 21. London: Oxford
University Press.
Goldberg, S.C., Schulz, S.C., Schulz, P.M. et al. (1986). Borderline and
schizotypal personality disorders treated with low-dose thiothixene vs
placebo. Archives of General Psychiatry, 43, 680-6.
Goldfried, M.R. & Goldfried, A.P. (1975). Cognitive change methods. In:
Helping People Change, ed. F.H. Kaufer & A.P. Goldstein. New York:
Pergamon Press.
Goldman, R.G., Skodol, A.E., McGrath, P.J. & Oldham, J.M. (1994).
Relationship between the tridimensional personality questionnaire and
DSM-III-R personality traits. American Journal of Psychiatry, 151, 274-6.
Goldman, S.J., D'Angelo, E.J. & De Maso, D.R. (1993). Psychopathology in the
families of children and adolescents with borderline personality disorder.
Goldstein, E.G. (1990). Borderline Disorders: Clinical Models and Techniques.
New York: The Guilford Press.
Goldstein, J.M. & Tsuang, M.T. (1990). Gender and schizophrenia: an
introduction and synthesis of findings. Schizophrenia Bulletin, 16, 179-83.
Gordon, A.M. (1983). Drugs and delinquency. British Journal of Psychiatry, 142,
169-73.
Gough, H.G. (1969). Manual for the California Psychological Inventory. Palo
Alto, CA: Consulting Psychologists Press.
Greenberg, D. & Rankin, H. (1982). Compulsive gamblers in treatment. British
Griggs, S.M.L.B. & Tyrer, P.J. (1981). Personality disorder, social adjustment
and treatment outcome in alcoholics. Journal of Studies in Alcohol, 42,
802-5.
Grinker, R., Werble, B. & Drye, R.C. (1968). The Borderline Syndrome. New
York: Basic Books.
Grobman, J. (1980). The borderline patient in group psychotherapy: a case
report. International Journal of Group Psychotherapy, 30, 299-318.
Grounds, A.T. (1987). Detention of'psychopathic disorder' patients in special
hospitals: critical issues. British Journal of Psychiatry, 151, 474-8.
Grounds, A.T., Quayle, M.T., France, J. et al. (1987). A unit for psychopathic
disorder patients in Broadmoor Hospital. Medical Science and Law, 27,
21-31.
Grove, W.M. & Tellegen, A. (1991). Problems in the classification of personality
disorders. Journal of Personality Disorders, 5, 31-41.
Grueneich, R. (1992). The borderline personality disorder diagnosis. Journal of
Gudjonsson, G.H. (1980). Psychological and psychiatric aspects of shoplifting.
354 References
Medicine, Science and Law, 30, 45-51.

Gunderson, J.G. (1984). Borderline Personality Disorder. Washington DC:
American Psychiatric Press.
Gunderson, J.G. & Elliott, G.R. (1985). The interface between borderline
personality disorder and affective disorder. American Journal of Psychiatry,
142, 277-88.
Gunderson, J.G., Frank, A.F., Katz, H.M. et al. (1984). Effects of psychotherapy
in schizophrenia: II. Comparative outcome of two forms of treatment.
Schizophrenia Bulletin, 10, 564-98.
Gunderson, J.G. & Kolb, J.E. (1978). Discriminating features of borderline
patients. American Journal of Psychiatry, 135, 792-6.
Gunderson, J., Kolb, J. & Austin, V. (1981). The diagnostic interview for
borderline patients. American Journal of Psychiatry, 138, 896-903.
Gunderson, J.G. & Phillips, K.A. (1991). A current view of the interface between
borderline personality disorder and depression. American Journal of
Gunderson, J.G., Ronningstam, E. & Bodkin, A. (1990). The diagnostic
interview for narcissistic patients. Archives of General Psychiatry, 47,
676-80.
Gunderson, J.G., Ronningstam, E. & Smith, L.E. (1991a). Narcissistic
personality disorder: a review of data on DSM-III-R descriptions. Journal of
Gunderson, J. & Sabo, A.N. (1993a). Treatment of borderline personality
disorder: A critical review. In: Borderline Personality Disorder, ch.18, ed. J.
Paris, pp. 385-406. Washington DC: American Psychiatric Press Inc.
Gunderson, J.G. & Sabo, A.N. (19936). The phenomenological and conceptual
interface between borderline personality disorder and PTSD. American
Gunderson, J.G. & Singer, M.T. (1975). Defining borderline patients: an
overview. American Journal of Psychiatry, 132, 1-10.
Gunderson, J.G. & Zanarini, M.C. (1987). Current overview of the borderline
diagnosis. Journal of Clinical Psychiatry, 48, 5-11.
Gunderson, J.G., Zanarini, M.C. & Kisiel, C.L. (19916). Borderline personality
disorder: a review of data on DSM-III-R descriptions. Journal of Personality
Gunn, J., Maden, T. & Swinton, M. (1991). Mentally Disordered Prisoners.
London: Home Office.
Gunn, J. & Robertson, G. (1976). Drawing a criminal profile. British Journal of
Criminology, 16, 15660.
Gunn, J., Robertson, G., Dell, S. & Way, C. (1978). Psychiatric Aspects of
Imprisonment. London: Academic Press.
Gurman, A.S. & Gustafson, J.P. (1976). Patients' perceptions of the therapeutic
relationship and group therapy outcome. American Journal of Psychiatry,
133, 1290-5.
Guze, S.B. (1976). Criminality and Psychiatric Disorders. New York: Oxford
University Press.
Guze, S.B., Wolfgram, E.D. & McKinney, J.K. (1967). Psychiatric illness in the
families of convicted criminals: a study of 519 first-degree relatives. Diseases
of the Nervous System, 28, 651-9.
Guze, S.B., Woodruff, R.A. & Clayton, P.J. (1971^). A study of conversion
symptoms in psychiatric outpatients. American Journal of Psychiatry, 128,
References 355
643-6.
Guze, S.B., Woodruff, R.A. & Clayton, P.J. (19716). Hysteria and antisocial
behavior: further evidence of an association. American Journal of
Hadley, H. & Strupp, H.H. (1976). Contemporary views of negative effects in
psychotherapy. Archives of General Psychiatry, 33, 1291-5.
Haines, K. (1990). After-care Services for Released Prisoners: A Review of the
Literature. Cambridge: Institute of Criminology.
Hall, C.S. & Lindsey, G. (1978). Theories of Personality, 3rd edn. New York:
Wiley.
Halperin, J.M., Sharma, V., Siever, L.J. et al. (1994). Serotonergic function in
aggressive nonaggressive boys with attention deficit hyperactivity disorder.
Hamberger, L.K. & Hastings, J. (1988). Characteristics of male spouse abusers
consistent with personality disorders. Hospital and Community Psychiatry,
39, 763-70.
Hare, R.D. (1970). Psychopathy: Theory and Research. New York: Wiley.
Hare, R.D. (1980). A research scale for the assessment of psychopathy in
criminal populations. Personality and Individual Differences, 1, 111-17.
Hare, R. (1985). Comparison of procedures for the assessment of psychopathy.
Hare, R.D. (1991). The Hare Psychopathy Checklist - Revised. Toronto: Multi-
Health Systems.
Hare, R.D., Hart, S.D. & Harpur, T.J. (1991). Psychopathy and the DSM-IV
criteria for antisocial personality disorder. Journal of Abnormal Psychology,
100, 391-8.
Hare, R.D. & McPherson, L.M. (1984). Violent and aggressive behaviour by
criminal psychopaths. International Journal of Law and Psychiatry, 7, 35-50.
Hare, R.D., McPherson, L.M. & Forth, A.E. (1988). Male psychopaths and their
criminal careers. Journal of Consulting and Clinical Psychology, 56, 710-14.
Hart, S.D., Dutton, D.G. & Newlove, T. (1993). The prevalence of personality
disorder among wife assaulters. Journal of Personality Disorders, 7, 329-41.
Hechtman, L. (1991). Resilience and vulnerability in long-term outcome of
attention deficit hyperactive disorder. Canadian Journal of Psychiatry, 36,
415-21.
Heilbrun, A.B. (1982). Cognitive models of criminal violence based on
intelligence and psychopathy levels. Journal of Consulting and Clinical
Heilbrun, A.B. & Heilbrun, M.R. (1985). Psychopathy and dangerousness:
comparison, integration and extension of two psychopathic typologies.
British Journal of Clinical Psychology, 24, 181-95.
Henderson, D.K. (1939). Psychopathic States. New York: Norson.
Henderson, M. (1982). An empirical classification of convicted violent offenders.
British Journal of Criminology, 22, 1-20.
Henderson, M. (1989). Behavioural approaches to violent crime. In: Clinical
Approaches to Violence, ed. K. Howells & C.R. Hollin, pp. 25-38.
Chichester: John Wiley & Sons.
Herman, J.L. (1992). Trauma and Recovery. New York: Basic Books.
Herman, J.L., Perry, J.C. & Van der Kolk, B.A. (1989). Childhood trauma in
borderline personality disorder. American Journal of Psychiatry, 146,
490-95.
356 References
Herzog, D.B., Keller, M.B., Lavori, P.W. et al. (1992). The prevalence of
personality disorders in 210 women with eating disorder. Journal of Clinical
Higgitt, A. & Fonagy, P. (1992). Psychotherapy in borderline and narcissistic
personality disorder. British Journal of Psychiatry, 161, 23^43.
Hill, D.C. (1970). Outpatient management of passive-dependent women.
Hospital and Community Psychiatry, 21, 3841.
Hill, J., Harrington, R., Fudge, H. et al. (1989). Adult personality functioning
assessment (APFA). British Journal of Psychiatry, 155, 24-35.
Hinde, R.A. & Stevenson-Hinde, J. (1986). Relating childhood relationships to
individual characteristics. In: Relationships and Development, ed. W.W.
Hartup & Z. Rubin. Hillsdale, NJ: Lawrence Erlbaum.
Hirschfeld, R.M.A. & Holzer, C.E. (1994). Depressive Personality Disorder:
clinical implications. Journal of Clinical Psychiatry, 55, 10-17.
Hirschfeld, R.M.A., Klerman, G.L., Gough, H.G. et al. (1977). A measure of
interpersonal dependency. Journal of Personality Assessment, 41, 610-8.
Hirschfeld, R.M.A. & Shea, M.T. (1992). Personality. In: Handbook of Affective
Disorders, 2nd edn, ed. E.S. Paykel, pp. 185-94. Edinburgh: Churchill
Livingstone.
Hirschfeld, R.M.A., Shea, M.T. & Weiss, R. (1991). Dependent personality
disorder: perspectives for DSM-IV. Journal of Personality Disorders, 5,
135-49.
Hodges, J. & Tizard, B. (1989). IQ and behavioural adjustment of ex-
institutional adolescents. Journal of Child Psychology and Psychiatry and
Allied Disciplines, 30, 53-75.
Hodgins, S. (1992). Mental disorder, intellectual deficiency and crime: Evidence
from a birth cohort. Archives of General Psychiatry, 49, 476-83.
Hoffart, A. & Martinsen, E.W. (1993). The effect of personality disorders and
anxious-depressive comorbidity on outcome in patients with unipolar
depression and with panic disorder and agoraphobia. Journal of Personality
Disorders,!, 304-11.
Hoffart. A., Thornes, K., Hedley. L.M. & Strand, J. (1994). DSM-II-R Axis I
and II disorders in agoraphobic patients with and without panic disorder.
Act a Psychiatrica Scandinavica, 89, 186-91.
Hollander, E., Stein, D.J., De Caria, C M . et al. (1994). Serotonergic sensitivity
in borderline personality disorder: preliminary findings. American Journal of
Home Office (1990). Provision for Mentally Disordered Offenders (Circular 66/
90). London: Home Office.
Home Office (1991). Treatment Programmes for Sex Offenders in Custody: a
Strategy. London: Home Office.
Home Office (1993a). Criminal Statistics England and Wales 1992. London:
HMSO.
Home Office (1993&). Life Licensees and Restricted Offenders Reconvictions:
England and Wales 1990. Home Office Statistical Bulletin 03/93. London:
Home Office Research and Statistics Department.
Home Office (1994). Statistics of Mentally Disordered Offenders: England and
Wales 1992. Home Office Statistical Bulletin 04/94. London: Home Office
Research and Statistics Department.
Home Office/Department of Health and Social Security. (1975). Report of the
Committee on Abnormal Offenders. London: HMSO.
Home Office/Department of Health and Social Security. (1987). Mental Health
References 357
Act 1983. Supervision and After-Care of Conditionally Discharged
Restricted Patients: Notes for the Guidance of Supervising Psychiatrists.
London: Home Office.
Horney, K. (1945). Our Inner Conflicts. New York: Norton.
Horney, K. (1950). Neurosis and Human Growth. New York: Norton.
Horowitz, M.J. (1989). Clinical phenomenology of narcissistic pathology.
Psychiatric Clinics of North America, 12, 531-9.
Horowitz, M., Marmar, C. & Weiss, D. (1986). Comprehensive analysis of
change after brief dynamic psychotherapy. American Journal of Psychiatry,
143, 582-9.
Horwitz, L. (1987). Indications for group psychotherapy with borderline and
narcissistic patients. Bulletin of the Menninger Clinic, 51, 248-60.
Howland, R.H. & Thase, M.E. (1993). A comprehensive review of cyclothymic
disorder. Journal of Nervous and Mental Disease, 181, 485-93.
Huesmann, L.R., Eron, L.D., Lefkowitz, M.M. & Walder, L.O. (1984). Stability
of aggression over time and generations. Developmental Psychology, 20,
1120-34.
Hull, J.W., Clarkin, J.F. & Alexopoulos, J. (1992). Time series analysis of "
interventive effects, fluoxetine therapy as a case illustration. Journal of
Nervous and Mental Disease, 181, 48-53.
Hunt, C. & Andrews, G. (1992). Measuring personality disorder: the use of self-
report questionnaires. Journal of Personality Disorders, 6, 125-33.
Hurt, S., Clarkin, J., Widiger, T. et al. (1988). DSM-III and borderline
personality disorder: Decision rules and their implications. Paper presented
at the First International Congress on the Disorders of Personality,
Copenhagen.
Hurt, S.W., Clarkin, J.F., Widiger, T.A. et al. (1990). Evaluation of DSM-III
decision rules for case detection using joint conditional probability
structures. Journal of Personality Disorders, 4, 121-30.
Hutchings, B. & Mednick, S.A. (1975). Registered criminality in the adopted and
biological parents of registered male adoptees. In: Genetics, Environment and
Psychopathology, ed. S.A. Mednick, F. Schulzinger & J. Higgins.
Amsterdam: Elsevier.
Hwe, H.F., Yeh, E.K. & Chang, L.Y. (1989). Prevalence of psychiatric disorders
in Taiwan defined by the Chinese Diagnostic Interview Schedule. Acta
Psychiatrica Scandinavica, 79, 13647.
Hyler, S.E., Rieder, R.O., Williams, J.B.W. et al. (1988). The Personality
Diagnostic Questionnaire: Development and preliminary results. Journal of
Hyler, S.E., Reider, R.O., Williams, J.B.W. et al. (1989). A comparison of
clinical and self-report diagnoses of DSM-III personality disorders in 552
patients. Comprehensive Psychiatry, 30, 170-8.
Hyler, S.E., Skodol, A.E., Kellman, H.D. et al. (1990). Validity of the personality
diagnostic questionnaire-revised: comparison with two structured
interviews. American Journal of Psychiatry, 147, 1043-8.
Inman, D.J., Bascue, L.O. & Skoloda, T. (1985). Identification of borderline
personality disorders among substance abuse inpatients. Journal of
Substance Abuse and Treatment, 2, 229-32.
Jackson, H.J., Whiteside, H.L., Bates, G.W. et al. (1991). Diagnosing personality
disorders in psychiatric inpatients. Acta Psychiatrica Scandinavica, 83,
206-13.
Jaspers, K. (1946). Allgemeine Psychopathologie, 4th edn. Berlin: Springer.
358 References
Jenike, M.A., Baer, L., & Minichiello, W.E. (1986). Concomitant obsessive-
compulsive disorder and schizotypal personality disorder. American Journal
of Psychiatry, 143, 530-2.
Jensen, J. (1988). Pain in non-psychotic psychiatric patients: life events,
symptomatology and personality traits. Ada Psychiatrica Scandinavica, 78,
201-7.
Joffe, R.T. & Regan, J.J. (1988). Personality and depression. Journal of
Psychiatric Research, 22, 279-86.
Joffe, R.T., Swinson, R.P. & Regan, J.J.(1988). Personality features of obsessive-
compulsive disorder. American Journal of Psychiatry, 145, 1127-9.
Johnson, C , Tobin, D. & Dennis, A. (1990). Differences in treatment outcome
between borderline and nonborderline bulimics at one year follow-up.
International Journal of Eating Disorders, 9, 617-27.
Jordan, B.K., Swartz, M.S., George, L.K. et al. (1989). Antisocial and related
disorders in a southern community. Journal of Nervous and Mental Disease,
177, 529-32.
Joseph, P. (1992). Non custodial treatment: Can psychopaths be treated in the
community? Criminal Behaviour and Mental Health, 2, 192-200.
Joyce, P.R., Mulder, R.T. & Cloninger, C.R. (1994a). Temperament and
hypercortisolemia in depression. American Journal of Psychiatry, 151, 195-8.
Joyce, P.R., Mulder, R.T. & Cloninger, C.R. (19946). Temperament predicts
clomipramine and disipramine response in major depression. Journal of
Affective Disorders, 30, 35-46.
Judd, P.H. & Ruff, R.M. (1993). Neuropsychological dysfunction in borderline
personality disorder. Journal of Personality Disorders, 1, 275-84.
Kagan, J. & Moss, H.A. (1962). Birth to Maturity. New York: Wiley.
Kagan, J., Reznick, J.S. & Snidman, N. (1986). Temperamental inhibition in
early childhood. In: The Study of Temperament: Changes, Continuities and
Challenges, ed. R. Plumin & J. Dunn. Hillsdale, NJ: Lawrence Erlbaum.
Kagan, J., Reznick, J.S. & Snidman, N. (1988). Biological bases of childhood
shyness. Science, 239, 167-71.
Kahn, E. (1928). Psychopathic personalities. In: Bumke's Handbook of Mental
Diseases, vol. 5, p. 227. Berlin: Springer.
Kaminsky, M. & Slavney, P. (1983). Hysterical and obsessional features in
patients with Briquet's syndrome (somatization disorder). Psychological
Medicine, 13,111-20.
Kandel, E., Brennan, P.A., Mednick, S.A. & Michelson, N.M. (1989). Minor
physical anomalies and recidivistic adult violent criminal behavior. Ada
Kandel, E., Mednick, S.A., Kirkegaard-Sorensen, L. et al. (1988). IQ as a
protective factor for subjects at high risk for antisocial behavior. Journal of
Consulting and Clinical Psychology, 56, 224-6.
Karno, M., Golding, J.M. & Sorenson, S.B. (1988). The epidemiology of
obsessive-compulsive disorder in five US communities. Archives of General
Karpman, B. (1948). The myth of the psychopathic personality. American
Karterud, S., Vaglum, S., Friis, S. et al. (1992). Day hospital therapeutic
community treatment for patients with personality disorders. Journal of
References 359
Kass, F., MacKinnon, R.A. & Spitzer, R.L. (1986). Masochistic personality: An
empirical study. American Journal of Psychiatry, 143, 216-8.
Kass, F., Skodol, A. & Charles, E. (1985). Scaled ratings of DSM-III personality
disorders. American Journal of Psychiatry, 142, 627-30.
Kass, F., Spitzer, R.L. & Williams, J.B.W. (1983). An empirical study of the
issue of sex bias in the diagnostic criteria of DSM-III Axis II personality
disorders. American Psychologist, 38, 799-801.
Kass, F., Spitzer, R.L., Williams, J.B.W. & Widiger, T. (1989). Self-defeating
personality disorder and DSM-III-R: Development of the diagnostic
criteria. American Journal of Psychiatry, 146, 1022-6.
Kay, S.R., Wolkenfeld, F. & Murrill, L.M. (1988a). Profiles of aggression among
psychiatric patients. Journal of Nervous and Mental Disease, 176, 53946.
Kay, S.R., Wolkenfeld, F. & Murrill, L.M. (19886). Profiles of aggression among
psychiatric patients. II. Covariates and predictors. Journal of Nervous and
Kayser, A., Robinson, D.S., Nies, A. & Howard, D. (1985). Response to
phenelzine among depressed patients with features of hysteroid dysphoria.
Kendell, R.E. (1983). The principles of classification in relation to mental disease.
In: Handbook of Psychiatry, vol. 1, ed. M. Shepherd & O.L. Zangwill, pp.
191-8. Cambridge: Cambridge University Press.
Kendler, K.S. (1985). Diagnostic approaches to schizotypal personality disorder:
a historical perspective. Schizophrenia Bulletin, 11, 538-53.
Kendler, K.S. (1988). Familial aggregation of schizophrenia and schizophrenia
spectrum disorders: evaluation of conflicting results. Archives of General
Kendler, K.S. & Eaves, L.J. (1986). Models for the joint effect of genotype and
environment on liability to psychiatric illness. American Journal of
Kendler, K.S. & Gruenberg, A.M. (1984). An independent analysis of the Danish
adoption study of schizophrenia. VI. The relationship between psychiatric
disorders as defined by DSM-III in the relatives and adoptees. Archives of
Kendler, K.S., Gruenberg, A.M. & Strauss, J.S. (1981). An independent analysis
of the Copenhagen sample of the Danish adoption study of schizophrenia.
II. The relationship between schizotypal personality disorder and
schizophrenia. Archives of General Psychiatry, 38, 982-4.
Kendler, K.S., Lieberman, J.A. & Walsh, D. (1989). The Structured Interview for
Schizotypy (SIS): a preliminary report. Schizophrenia Bulletin, 15, 559-71.
Kendler, K.S., Masterson, C.C., Ungaro, R. & Davis, K.L. (1984). A family
history study of schizophrenia-related personality disorders. American
Journal of Psychiatry, 141, 4247.
Kendler, K.S., McGuire, M., Gruenberg, A.M. et al. (1993a). The Roscommon
Family Study. I. Methods, diagnosis of probands, and risk of schizophrenia
in relatives. Archives of General Psychiatry, 50, 527-40.
Kendler, K.S., McGuire, M., Gruenberg, A.M. et al. (19936). The Roscommon
Family Study. Schizophrenia-related personality disorders in relatives.
Archives of General Psychiatry, 50, 781-8.
Kendler, K.S., Neale, M.C., Kessler, R.C. et al. (1993c). A longitudinal twin
study of personality and major depression in women. Archives of General
360 References
Kendler, K.S., Ochs, A.L., Gorman, A.M. et al (1991). The structure of
schizotypy: a pilot multitrait twin study. Psychiatry Research, 36, 19-36.
Kennedy, H.G. & Grubin, D.H. (1992). Patterns of denial in sex offenders.
Psychological Medicine, 22, 191-6.
Kernberg, O. (1967). Borderline personality organisation. Journal of the
American Psychoanalytic Association, 15, 641-85.
Kernberg, O. (1977). The structural diagnosis of borderline personality
organization. In: Borderline Personality Disorder, ed. P. Hartocoullis. New
York: International University Press.
Kernberg, O.F. (1984). Severe Personality Disorders. New Haven: Yale
University Press.
Kernberg, O.F., Burstein, E., Coyne, L. et al (1972). Psychotherapy and
psychoanalysis: final report of the Menninger Foundation's psychotherapy
research project. Bulletin of the Menninger Clinic, 36, 1-275.
Khouri, P.J., Haier, R.J., Reider, R.O. & Rosenthal, D. (1980). A symptom
schedule for the diagnosis of borderline schizophrenia. A first report. British
Kilzieh, N. & Cloninger, C.R. (1993). Psychophysiological antecedents of
personality. Journal of Personality Disorders, Suppl. Spring, 100-17.
Kjelsberg, E., Eikeseth, P.H. & Dahl, A.A. (1991). Suicide in borderline patients
- predictive factors. Acta Psychiatrica Scandinavica, 84, 283-7.
Klassen, D. & O'Connor, W.H. (1988). A prospective study of predictors of
violence in adult male mental health admissions. Law and Human Behavior,
12, 143-58.
Klein, D.F. & Davis, J.M. (1969). The diagnosis of affective disorders. In:
Diagnosis and Treatment of Psychiatric Disorders, 1st edn, ed. D.F. Klein &
J.M. Davis. Baltimore: Williams & Wilkins.
Klein, M. (1985). Wisconsin Personality Inventory (WISPI), Department of
Psychiatry, 600 Highland Ave., Madison, WI 53792, USA.
Klein, R.G. & Mannuzza, S. (1991). Long-term outcome of hyperactive children:
a review. Journal of the American Academy of Child and Adolescent
Kleinman, P.H., Miller, A.B., Millman, R.B. et al (1990). Psychotherapy among
cocaine abusers entering treatment. Journal of Nervous and Mental Disease,
178, 442-7.
Klerman, G.L., Endicott, J. & Spitzer, R. (1979). Neurotic depression: a
systematic analysis of multiple criteria and meanings. American Journal of
Klerman, G.L. & Hirschfeld, R.M.A. (1988). Personality as a vulnerability
factor: with special attention to clinical depression. In: Handbook of Social
Psychiatry, ed. Henderson & Burrows, pp. 41-53. Amsterdam: Elsevier.
Klinteberg, B.af., Schalling, D., Edman, G. et al (1987). Personality correlates of
platelet monoamine oxidase (MAO) activity in female and male subjects.
Neuropsychobiology, 18, 89-96.
Knight, R.P. (1953). Borderline states. Bulletin of the Menninger Clinic, 17, 1-12.
Koch, J.L.A. (1891). Die Psychopathischen Minderwertigkeiten. Ravensburg:
Dorn.
Kohut, H. (1968). The psychoanalytic treatment of narcissistic personality
disorders. Psychoanalytic Study of the Child, 23, 86-113.
Kohut, H. (1977). The Restoration of Self New York: International Universities
References 361
Press.
Kolb, J.E. & Gunderson, J.G. (1990). Psychodynamic psychotherapy of
borderline patients. International Review of Psycho-Analysis, 17, 515-6.
Kolb, L. (1982). Assertive traits fostering social adaptation and creativity.
Psychiatric Journal of the University of Ottawa, 7, 219-25.
Korn, M.L., Botsis, A.J. Kotler, M. et al. (1992). The suicide and aggression
survey: a semistructured instrument for the measurement of suicidality and
aggression. Comprehensive Psychiatry, 33, 359-65.
Kraepelin, E. (1905). Lectures on Clinical Psychiatry, 2nd edn, trans. T.
Johnstone. London: Bailliere Tindall & Co.
Kraepelin, E. (1921). Manic-Depressive Insanity and Paranoia, trans. R.M.
Barclay, ed. G.M. Robertson. Edinburgh: E. & S. Livingstone.
Krafft-Ebing, R. (1901). Psychopathia Sexualis: A Medico-Forensic Study. New
York: Pioneer Publications (published in 1950).
Krarup, G., Nielsen, B., Rask, P. & Petersen, P. (1991). Childhood experiences
and repeated suicidal behavior. Acta Psychiatrica Scandinavica, 83, 16-19.
Kreeger, L. (1975). The Large Group, ed. L. Kreeger. London: Constable.
Kretchmer, E. (1918). Der Sensitive Beziehungswahn. Berlin: Springer.
Kretchmer, E. (1922). Korperban und Charakter. Berlin: Springer.
Kretchmer, E. (1925). Physique and Character. New York: Harcourt Brace;
London: Kegan-Paul, Trench & Trabner.
Kretsch, R., Goren, Y. & Wasserman, A. (1987). Change patterns of borderline
patients in individual and group therapy. InternationalJournal of Group
Psychotherapy, 37, 95-112.
Kroll, J.L., Carey, K.S. & Sines, L.K. (1985). Twenty-year follow-up of
borderline personality disorder: a pilot study. In: IV World Congress of
Biological Psychiatry, vol. 3, ed. C. Stragass. New York: Elsevier.
Krupnick, J.L. & Pincus, H.A. (1992). The cost-effectiveness of psychotherapy: a
plan for research. American Journal of Psychiatry, 149, 1295-1305.
Kullgren, G. (1988). Factors associated with completed suicide in borderline
personality disorder. Journal of Nervous and Mental Disease, 176, 40-4.
Kumar, H.V., McMahon, K.J., Allman, K.M. et al. (1989). Pericentric inversion
chromosome 9 and personality disorder. British Journal of Psychiatry, 155,
408-10.
Kune, G., Kune, S., Watson, L.F. & Bahnson, C.B. (1991). Personality as a risk
factor in large bowel cancer: data from the Melbourne Colorectal Cancer
Study. Psychological Medicine, 21, 29-41.
Lacey, J.H. (1984). Moderation of bulimia. Journal of Psychosomatic Research,
28, 397^02.
Lahmeyer, H.W., Reynolds, C.F., Kupfer, D.J. & King, R. (1989). Biologic
markers in borderline personality disorder: a review. Journal of Clinical
Laitinen, L.V. (1988). Psychosurgery today. Acta Neurochirurgica, Suppl. 44,
158-62.
Lane, P.J. & Kling, J. (1979). Construct validity of the overcontrolled hostility
scale of the MMPI. Journal of Consulting and Clinical Psychology, 41, 781-2.
Lazare, A. (1971). The hysterical character in psychoanalytic theory. Archives of
General Psychiatry, IS, 131-7.
Lazare, A., Klerman, G. & Armor, D.J. (1966). Oral, obsessive and hysterical
personality patterns. Archives of General Psychiatry, 14, 624-30.
Lazare, A., Klerman, G. & Armor, D.J. (1970). Oral, obsessive and hysterical
362 References
personality patterns: replications of factor analysis in an independent
sample. Journal of Psychiatric Research, 7, 275-9.
Leff, J.P. & Isaacs, A.D. (1990). History-taking: The personal background. In:
Psychiatric Examination in Clinical Practice, 3rd edn, pp. 13-26. Oxford:
Blackwell Scientific Publications.
Leonhard, K. (1968). Akzentuierte Persbnlichkeiten. Berlin: Verlag Volk und
Gesundheit.
Lesieur, H.R. & Blume, S.B. (1987). The South Oaks Gambling Screen (SOGS).
Levitt, A.J., Joffe, R.T., Ennis, J. et al. (1990). The prevalence of cyclothymia in
borderline personality disorder. Journal of Clinical Psychiatry, 51, 335-9.
Lewis, G. & Appleby, L. (1988). Personality disorder: the patients psychiatrists
dislike. British Journal of Psychiatry, 153, 44-49.
Lewis, S.J. & Harder, D.W. (1991). A comparison of four measures to diagnose
DSM-III-R borderline personality disorder in outpatients. Journal of
Liberman, R.P. & Eckman, T. (1981). Behaviour therapy vs insight-orientated
therapy for repeated suicide attempters. Archives of General Psychiatry, 38,
1126-30.
Lie, N. (1992). Follow-ups of children with attention deficit hyperactivity
disorder (ADHD). Acta Psychiatrica Scandinavica, suppl. 85, 1-40.
Liebowitz, M.R. & Klein, D.F. (1979). Hysteroid dysphoria. Psychiatric Clinics
of North America, 2, 555-75.
Liebowitz, M.R. & Klein, D.F. (1981). Interrelationship of hysteroid dysphoria
and borderline personality disorder. Psychiatric Clinics of North America, 4,
67-87.
Lilienfeld, S., Van Volkenburg, C , Larntz, K. & Akiskal, H. (1986). The
relationship of histrionic personality disorder to antisocial personality and
somatization disorder. American Journal of Psychiatry, 143, 718-22.
Linehan, M.M. (1987). Dialectical behavioral therapy: a cognitive behavioral
approach to parasuicide. Journal of Personality Disorders, 1, 328-33.
Linehan, M.M., Heard, H.L. & Armstrong, H.E. (1993). Naturalistic follow-up
of a behavioral treatment for chronically parasuicidal borderline patients.
Linehan, M.M., Hubert, A.E., Suarez, A. et al. (1991). Cognitive-behavioral
treatment of chronically parasuicidal borderline patients. Archives of
Links, P. (1993). Seven Year Follow-Up Study of Borderline Personality
Disorder: Factors Predicting Persistence. Paper presented at Third
International Congress on the Disorders of Personality. Cambridge, Mass.
Links, P.S., Mitton, J.E. & Steiner, M. (1990). Predicting outcome for borderline
personality disorder. Comprehensive Psychiatry, 31, 490-8.
Linnoila, V.M.I. & Virkkunen, M. (1992). Aggression, suicidality and serotonin.
Journal of Clinical Psychiatry, 53(10 Suppl.), 46-51.
Liskow, B., Powell, B.J., Nickel, E.J. & Penick, E. (1990). Diagnostic subgroups
of antisocial alcoholics: outcome at 1 year. Comprehensive Psychiatry, 31,
549-56.
Livesley, W.J. (1991). Classifying personality disorders: ideal types, prototypes,
or dimensions? Journal of Personality Disorders, 5, 52-9.
Livesley, W.J. & Jackson, D.N. (1992). Guidelines for developing, evaluating and
revising the classification of personality disorders. Journal of Nervous and
References 363

Livesley, W.J., Jong, K.L., Jackson, D.N. & Vernon, P.A. (1993). Genetic and
environmental contributions to dimensions of personality disorder.
Livesley, W.J. & Schroeder, M.L. (1990). Dimensions of personality disorder.
Livesley, W.J. & Schroeder, M.L. (1991). Dimensions of personality disorder.
The DSM-III-R cluster B diagnosis. Journal of Nervous and Mental Disease,
179, 320-8.
Livesley, W.J., Schroeder, M.L. & Jackson, D.N. (1990). Dependent personality
disorder and attachment problems. Journal of Personality Disorders, 4,
232-40.
Loeber, R. (1990). Development and risk factors of juvenile antisocial behaviour
and delinquency. Clinical Psychology Review, 10, 1-41.
Lofgren, D.P., Bemporad, J., King, J. et al. (1991). A prospective follow-up
study of so-called borderline children. American Journal of Psychiatry, 148,
1541-7.
Loranger, A.W., Hirschfeld, R.M.A., Sartorius, N. & Regier, D.A. (1991). The
WHO/ADAMHA international pilot study of personality disorders:
background and purpose. Journal of Personality Disorders, 5, 296-306.
Loranger, A.W., Lenzenweger, M.F., Gartner, A.F. et al. (1991). Trait-state
artifacts and the diagnosis of personality disorders. Archives of General
Loranger, A., Oldham, J. & Tullis, E. (1982). Familial transmission of DSM-III
borderline personality disorder. Archives of General Psychiatry, 39, 795-9.
Loranger, A., Susman V., Oldham, J. & Russaroff, L.M. (1987). The personality
disorder examination: a preliminary report. Journal of Personality Disorders,
1, 1-13.
Luborsky, L. (1984). Principles of Psychoanalytic Psychotherapy. A Manual for
Supportive-Expressive Treatment. New York: Basic.
Lucas, P.B., Gardner, D.L., Wolkowitz, O.M. & Cowdry, R.W. (1987).
Dysphoria associated with methylphenidate infusion in borderline
personality disorder. American Journal of Psychiatry, 144, 1577-9.
Luntz, B.K. & Widom, C.S. (1994). Antisocial personality disorder in abused
children grown up. American Journal of Psychiatry, 151, 6704.
Lykouras, E., Markianos, M. & Moussas, G. (1989). Platelet monoamine
oxidase, plasma dopamine /Miydroxylase activity, dementia and family
history of alcoholism in chronic alcoholics. Acta Psychiatrica Scandinavica,
80,487-91.
Macaskill, N.D. (1982). Therapeutic factors in group therapy with borderline
patients. International Journal of Group Psychotherapy, 32, 61-74.
MacCulloch, M., Bailey, J., Jones, C. & Hunter, C. (1993a). Nineteen serious
reoffenders who were discharged direct to the community from a special
hospital: 1. General characteristics. Journal of Forensic Psychiatry, 4,
237^8.
MacCulloch, M., Bailey, J., Jones, C. & Hunter, C. (19936). Nineteen serious
hospital: 2. Illustrated clinical issues. Journal of Forensic Psychiatry, 3,
451-70.
MacCulloch, M., Bailey, J., Jones, C. & Hunter, C. (1994). Nineteen serious
364 References
hospital: 3. Illustrated administrative issues. Journal of Forensic Psychiatry,
5, 63-82.
Maden, T., Swinton, M. & Gunn, J. (1994). Psychiatric disorder in women
serving a prison sentence. British Journal of Psychiatry, 164, 44-54.
Maier, W., Lichtermann, D., Klingler, T. et al. (1992). Prevalence of personality
disorders (DSM-III-R) in the community. Journal of Personality Disorders,
6, 187-96.
Main, T.F. (1946). The hospital as a therapeutic institution. Bulletin of the
Menninger Clinic, 10, 66-71.
Main, T.F. (1957). The ailment. British Journal of Medical Psychology, 30,
129-45.
Malan, D.H. (1973). The outcome problem in psychotherapy research. Archives
of General Psychiatry, 29, 719-23.
Malan, D.H. (1979). Individual Psychotherapy and the Science of
Psychodynamics. London: Butterworth.
Malan, D.H., Balfour, F.H.G., Hood, V.G. & Shooter, M.N. (1976). Group
psychotherapy. Archives of General Psychiatry, 33, 1303-9.
Malloy, R., Noel, N., Longabaugh, R. & Beattie, M. (1990). Determinants of
neuropsychological impairment in antisocial substance abusers. Addictive
Behavior, 15,431-8.
Malow, R.M., West, J.A., Williams, J.L. & Sutker, P.B. (1989). Personality
disorders classifications and symptoms in cocaine and opioid addicts.
Mann, A.H., Jenkins, R., Cutting, J.C. & Cowen, P.T. (1981). The development
and use of a standardized assessment of abnormal personality. Psychological
Medicine, 11, 839-47.
Mann, J. (1973). Time-Limited Psychotherapy. Cambridge, MA: Harvard
University Press.
Mannuzza, S., Klein, R.G., Bessler, A. et al. (1993). Adult outcome of
hyperactive boys: educational achievement, occupational rank, and
psychiatric status. Archives of General Psychiatry, 50, 565-76.
Mannuzza, S., Klein, R.G., Boragura, N. et al. (1991). Hyperactive boys almost
grown up. Replication of psychiatric status. Archives of General Psychiatry,
48, 77-83.
Mannuzza, S., Klein, R.G., Konig, P.H. & Giampino, T.L. (1989). Hyperactive
boys almost grown up. Archives of General Psychiatry, 46, 1073-9.
Marchand, A. & Wapler, M. (1993). The effect of personality disorders on the
efficacy of behavioral cognitive therapy of panic disorders with agoraphobia.
Canadian Journal of Psychiatry, 38, 163-6.
Marin, D.B., Kocsis, J.H., Frances, A.J. & Klerman, G.L. (1993). Personality
disorders in dysthymia. Journal of Personality Disorders, 7, 223-31.
Markowitz, J . C , Moran, M.E., Kocsis, J.H. & Frances, A.J. (1992). Prevalence
and comorbidity of dysthymic disorder among psychiatric outpatients.
Journal of Affective Disorders, 24, 63-71.
Markowitz, P.J., Calabrese, J.R., Schulz, S.C. & Meltzer, H.Y. (1991).
Fluoxetine in the treatment of borderline and schizotypal personality
Marmor, J. (1979). Short-term dynamic psychotherapy. American Journal of
Marshall, W.L. & Barbaree, H.E. (1989). Sexual violence. In: Clinical Approaches
to Violence, ed. K. Howells & C.R. Hollin, pp. 205^49. Chichester: John
References 365
Wiley & Sons.
Martin, H.P. (1980). The consequences of being abused and neglected: how the
child fares. In: The Battered Child, ed. L.H. Kempe & R.E. Heifer, pp. 347-
365. Chicago: University of Chicago Press.
Martin, R.L. (1986). Excess mortality among psychiatric patients. The
importance of unnatural death, substance abuse disorders, antisocial
personality and homosexuality. Acta Psychiatrica Belgica, 86, 553-4.
Martin, R.L., Cloninger, C.R. & Guze, S.B. (1982). The natural history of
somatization and substance abuse in women criminals: a six year follow-up.
Marttunen, M.J., Aro, H.M., Henriksson, M.M. & Lonnqvist, J.K. (1994).
Antisocial behaviour in adolescent suicide. Acta Psychiatrica Scandinavica,
89, 167-73.
Marziali, E. (1992). The etiology of borderline personality disorder:
developmental factors. In: Borderline Personality Disorder, ed. J.F. Clarkin,
E. Marziali & H. Munroe-Blum, pp. 2 7 ^ 4 . New York: Guilford Press.
Marzillier, J.S., Lambert, C. & Kellett, J. (1976). A controlled evaluation of
systematic desensitization and social skills training for socially inadequate
psychiatric patients. Behaviour Research & Therapy, 14, 225-38.
Masterson, J.F. (1976). Treatment of the Borderline Adult. New York:
Brunner-Mazel.
Masterson, J. (1981). The Narcissistic and Borderline Disorders. New York:
Brunner-Mazel.
Matochik, J.A., Liebenauer, L.L., King, A.C. et al. (1994). Cerebral glucose
metabolism in adults with attention deficit hyperactivity disorder after
chronic stimulant treatment. American Journal of Psychiatry, 151, 658-64.
Mattes, J.A. (1988). Carbamazepine v propranolol for rage outbursts.
Psychopharmacology Bulletin, 24, 179-82.
Maudsley, H. (1868). A Physiology and Pathology of Mind, 2nd edn. London:
Macmillan.
Mavissakalian, M. (1990). The relationship between panic disorder/agoraphobia
and personality disorders. Psychiatric Clinics of North America, 13, 661-84.
Mavissakalian, M. & Hamann, M. (1986). DSM-III personality disorder and
agoraphobia. Comprehensive Psychiatry, 27, 471-9.
Mavissakalian, M., Hamann, M.S. & Jones, B. (1990fl). A comparison of DSM-
III personality disorders in panic/agoraphobia and obsessive-compulsive
disorder. Comprehensive Psychiatry, 31, 238-44.
Mavissakalian, M., Hamann, M.S. & Jones, B. (19906). DSM-III personality
disorders in obsessive-compulsive disorder: changes with treatment.
Mawson, D. (1983). 'Psychopaths' in special hospitals. Bulletin of the Royal
College of Psychiatrists, 1, 178-81.
Mawson, D., Grounds, A. & Tantam, D. (1985). Violence and Asperger's
syndrome: A case study. British Journal of Psychiatry, 147, 566-9.
Mazure, C M . , Nelson, J.C. & Jatlow, P.I. (1990). Predictors of hospital outcome
without antidepressants in major depression. Psychiatry Research, 33, 51-8.
McCann, U.D., Rossiter, E.M., King, R.J. & Agras, W.S. (1991). Nonpurging
bulimia: a distinct subtype of bulimia nervosa. International Journal of
Eating Disorders, 10, 679-87.
McCarroll, J.E., Ursano, R.J. & Fullerton, C.S. (1993). Symptoms of
posttraumatic stress disorder following recovery of war dead. American
366 References

McCord, W.M. (1983). The Psychopath and Milieu Therapy. New York:
Academic Press.
McCormick, R.A., Taber, J., Kruedelbach, N. & Russo, A. (1987). Personality
profiles of hospitalized pathological gamblers: the California Personality
Inventory. Journal of Clinical Psychology, 43, 521-7.
McCrae, R.R., Costa, P.T. & Arenberg, D. (1980). Constancy of adult
personality structure in males: longitudinal, cross-sectional and times-of-
measurement analysis. Journal of Gerontology, 35, 877-83.
McElroy, S.L., Hudson, J.I., Pope, H.G. et al (1992). The DSM-III-R impulse
control disorders not elsewhere classified: Clinical characteristics and
relationships to other psychiatric disorders. American Journal of Psychiatry,
149,318-27.
McGlashan, T.H. (1983). The borderline syndrome: is it a variant of
schizophrenia or affective disorder? Archives of General Psychiatry, 40,
1319-23.
McGlashan, T.H. (1986). The Chestnut Lodge follow-up study: III. Long-term
outcome of borderline personality disorder. Archives of General Psychiatry,
43, 20-30.
McGlashan, T.H. (1987). Testing DSM-III-R symptom criteria for schizotypal
and borderline personality disorders. Archives of General Psychiatry, 44,
143-8.
McGlashan, T.H. (1993). Implications of outcome research for the treatment of
borderline personality disorder. In: Borderline Personality Disorder, ed. J.
Paris, pp. 235-60. Washington, DC: American Psychiatric Press Inc.
McGlashan, T.H. & Heinssen, R.K. (1989). Narcissistic, antisocial and non-
comorbid subgroups of borderline disorder: are they distinct entities by
long-term clinical profile? Psychiatric Clinics of North America, 12, 653-70.
McGuffin, P. & Gottesman, I.I. (1984). Genetic influences on normal and
abnormal development. In: Child Psychiatry: Modern Approaches, 2nd edn,
ed. M. Rutter & L. Hersov. London: Blackwell.
McGuffin, P. & Thapar, A. (1992). The genetics of personality disorder. British
McGuire, F.L. (1980). Heavy and light drinking - drivers as separate target
groups for treatment. American Journal of Drug and Alcohol Abuse, 7,
101-7.
McGurk, B. (1978). Personality types among normal homocides. British Journal
of Criminology, 18, 146-61.
McKay, J.R., Alterman, A.I., McLellan, A.T. & Snider, E.C. (1994). Treatment
goals, continuity of care, and outcome in a day hospital substance abuse
rehabilitation program. American Journal of Psychiatry, 151, 254-9.
McMahon, R.C., Flynn, P.M. & Davidson, R.S. (1985). The personality and
symptoms scales of the Millon Clinical Multiaxial Inventory: sensitivity to
post-treatment outcomes. Journal of Clinical Psychology, 41, 862-6.
Mednick, S.A., Gabrielli, W.F. & Hutchings, B. (1984). Genetic influences in
criminal convictions: evidence from an adoption cohort. Science, 224, 891-4.
Megargee, E.I. (1966). Undercontrolled and overcontrolled personality types in
extreme antisocial aggression. Psychological Monographs, 80, no. 611.
Mehlum, L., Friis, S., Irion, T. et al (1991). Personality disorders 2-5 years after
treatment: a prospective follow-up study. Acta Psychiatrica Scandinavica,
84, 72-7.
References 367
Merikangas, K.R. & Weissman, M.H. (1986). Epidemiology of DSM-III Axis II

personality disorders. In: APA Annual Review, ed. A.J. Frances & R.E.
Hales (vol. 5, Psychiatry Update, pp. 258-78). Washington DC: American
Psychiatric Press.
Merikangas, K.R., Wicki, W. & Angst, J. (1994). Heterogeneity of depression.
Miller, L. (1988). Neuropsychological perspectives on delinquency. Behavioral
Sciences and the Law, 6, 409-28.
Millon, T. (1969). Modern Psychopathology. Philadelphia: Saunders.
Millon, T. (1981). Disorders of Personality: DSM-IIL Axis II. New York: Wiley.
Millon, T. (1987). Manual for the Millon Clinical Multi-axial Inventory II
(MCM-II). Minnetonka: National Computer Systems.
Millon, T. (1991). Avoidant personality disorder: A brief review of issues and
data. Journal of Personality Disorders, 5, 353-62.
Mindham, R.H.S., Scadding, J.G. & Cawley, R.H. (1992). Diagnoses are not
diseases. British Journal of Psychiatry, 161, 686-91.
Minichiello, W.E., Baer, L. & Jenike, M.A. (1987). Schizotypal personality
disorder: a poor prognostic indicator for behavior therapy in the treatment
of obsessive-compulsive disorder. Journal of Anxiety Disorders, 1, 273-6.
Modestin, J. & Emmenegger, P.A. (1986). Completed suicide and criminality:
lack of a direct relationship. Psychological Medicine, 16, 661-9.
Modestin, J., Foglia, A. & Toffler, G. (1989). Comparative study of schizotypal
and schizophrenia patients. Psychopathology, 22, 1-13.
Monahan, J. (1981). The Clinical Prediction of Violent Behaviour. Maryland: US
Department of Health and Human Services.
Monroe, R.R. (1982). DSM-III style diagnoses for the episodic disorders. Journal
of Nervous and Mental Disease, 170, 664-9.
Montgomery, S.A. & Montgomery, D.D. (1982). Pharmacological prevention of
suicidal behaviour. Journal of Affective Disorders, 4, 291-8.
Moorey, S. (1991). Cognitive behaviour therapy. Hospital Update, September,
726-32.
Morey, L.C. (1988). The categorical representation of personality disorder: a
cluster analysis of DSM-III-R personality features. Journal of Abnormal
Morey, L. & Ochoa, E. (1989). An investigation of adherence to diagnostic
criteria: clinical diagnosis of the DSM-III personality disorders. Journal of
Morey, L., Waugh, M. & Blashfield, R. (1985). MMPI scales for DSM-III
personality disorders: their derivation and correlates. Journal of Personality
Assessment, 49, 245-51.
Morrison, J. (1989). Histrionic personality disorder in women with somatization
disorder. Psychosomatics, 30, 433-7.
Moyes, T., Tennent, T.G. & Bedford, A.P. (1985). Long-term follow-up study of
a ward-based behaviour modification programme for adolescents with
acting-out and conduct problems. British Journal of Psychiatry, 147, 300-5.
Mulder, R.T. (1991). Personality disorders in New Zealand hospitals. Ada
Mungas, D. (1988). Psychometric correlates of episodic violent behaviour. British
Murray, R.M. & O'Callaghan, E. (1991). Neurodevelopmental schizophrenia.
Schizophrenia Monitor, 1, 1-3.
368 References
Murthy, V.N. (1969). Personality and the nature of suicide attempts. British
Myers, J.K., Weissman, M.M., Tischler, G.L. et al. (1984). Six month prevalence
of psychiatric disorders in three communities, 1980-1982. Archives of
Nace, E.P. (1989). Substance use disorders and personality disorders:
comorbidity. Psychiatric Hospital, 20, 65-9.
Nace, E.P., Davis, C.W. & Gaspari, J.P. (1991). Axis II comorbidity in substance
abusers. American Journal of Psychiatry, 148, 118-20.
Nelson, J.C. & Charney, D.S. (1981). The symptoms of major depressive illness.
Nestadt, G., Romanoski, A.J., Brown, C.H. et al. (1991). DSM-III compulsive
personality disorder: an epidemiological survey. Psychological Medicine, 21,
461-71.
Nestadt, G., Romanoski, A.J., Chahal, R. et al. (1990). An epidemiological study
of histrionic personality disorder. Psychological Medicine, 20, 413-22.
Nig, J., Silk, H.R., Ogata, S. et al. (1989). Sexual abuse and borderline
personality disorder: a comparative study. Presented at the 142nd Annual
Meeting, American Psychiatric Association, San Francisco.
Norden, M.J. (1989). Fluoxetine in borderline personality disorder. Progress in
Neuropsychopharmacology and Biological Psychiatry, 13, 885-93.
Nordstrom, G. & Berglund, M. (1987). Type 1 and type 2 alcoholics have
different patterns of successful long-term adjustment. British Journal of
Addiction, 82, 761-9.
North, C.S., Smith, E.M. & Spitznagel, E.L. (1993). Is antisocial personality a
valid diagnosis among the homeless? American Journal of Psychiatry, 150,
578-83.
Norton, K. (1992). Personality disordered individuals: The Henderson Hospital
model of treatment. Criminal Behaviour and Mental Health, 2, 180-91.
Noyes, R., Reich, J.H., Suelzer, M. & Christiansen, J. (1991). Personality traits
associated with panic disorder: change associated with treatment.
Nurnberg, H.G., Raskin, M. & Levine, P.E. (1989). Borderline personality
disorder as a negative prognostic factor in anxiety disorders. Journal of
Nurnberg, H.G., Raskin, M., Levine, P.E. et al. (1991). The comorbidity of
borderline personality disorder and other DSM-III-R Axis II personality
Nurnberg, H.G., Rifkin, A. & Doddi, S. (1993). A systematic assessment of the
comorbidity of DSM-III-R personality disorders in alcoholic outpatients.
O'Brien, M.L. (1988). Further evidence of the validity of the O'Brien Multiphasic
Narcissism Inventory. Psychological Reports, 62, 879-82.
Ojehagen, A. & Berglund, M. (1986a). Early and late improvement in a two-year
out-patient alcoholic treatment programme. Acta Psychiatrica Scandinavica,
74, 129-36.
Ojehagen, A. & Berglund, M. (1986&). To keep the alcoholic in out-patient
treatment. A differentiated approach through treatment contracts. Acta
Oldham, J.M., Skodol, A.E., Kellman, H.D. et al. (1992). Diagnosis of DSM-III-
R personality disorders by two structured interviews: patterns of
References 369
comorbidity. American Journal of Psychiatry, 149, 213-20.
O'Leary, K.M., Brouwers, P., Gardner, D.L. & Cowdry, R.W. (1991).
Neuropsychological testing of patients with borderline personality disorder.
Olweus, D. (1979). Stability of aggressive reaction patterns in males: a review.
Psychological Bulletin, 86, 852-75.
Overholser, J.C. (1989). Differentiation between schizoid and avoidant
personalities: an empirical test. Canadian Journal of Psychiatry, 34, 785-90.
Overholser, J . C , Kabakoff, R. & Norman, W.H. (1989). Personality
characteristics in depressed and dependent psychiatric inpatients. Journal of
Personality Assessment, 53, 40-50.
Pallis, D.T. & Birtchnell, J. (1977). Seriousness of suicide attempts in relation to
personality. British Journal of Psychiatry, 130, 253-9.
Papp, L.A., Zitrin, C M . , Coplan, J. & Gorman, J.M. (1990). The role of
personality in anxiety disorders. Psychiatric Medicine, 8, 107-20.
Paris, J. (1993). Management of acute and chronic suicidality in patients with
borderline personality disorder. In: Borderline Personality Disorder, ed. J.
Paris, pp. 373-84. Washington DC: American Psychiatric Press Inc.
Paris, J., Brown R. & Nowlis, D. (1987). Long-term follow-up of borderline
patients in a general hospital. Comprehensive Psychiatry, 28, 530-5.
Paris, J., Nowlis, D. & Brown, R. (1989). Predictors of suicide in borderline
personality disorder. Canadian Journal of Psychiatry, 34, 8-9.
Paris, J., Zweig-Frank, H. & Guzder, H. (1993). The role of psychological risk
factors in recovery from borderline personality disorder. Comprehensive
Parker, G., Hadzi-Pavlovic, D., Wilhelm, K. et al. (1994). Defining melancholia:
properties of a refined sign-based measure. British Journal of Psychiatry,
164,316-26.
Parnas, J., Cannon, T.D., Jacobsen, B. et al (1993). Lifetime DSM-III-R
diagnostic outcomes in the offspring of schizophrenic mothers. Archives of
Parsons, B., Quitkin, F.M., McGrath, P.J. et al (1989). Phenelzine, imipramine
and placebo in borderline patients meeting criteria for atypical depression.
Psychopharmacology Bulletin, 25, 52434.
Pekarik, G., Jones, D.L. & Blodgett, C. (1986). Personality and demographic
characteristics of dropouts and completers in a nonhospital residential
alcohol treatment program. International Journal of the Addictions, 21,
131-7.
Pelham, W.F., Bender, M.E., Caddell, J. et al (1985). Methylphenidate and
children with attention deficit disorder. Archives of General Psychiatry, 42,
948-52.
Penick, E.C, Powell, B.J., Liskow, B.I. et al (1988). The stability of coexisting
psychiatric syndromes in alcoholic men after one year. Journal of Studies in
Alcohol, 49, 395-405.
Perkins, D. (1991). Clinical work with sex offenders in secure settings. In: Clinical
Approaches to Sex Offenders and their Victims, ed. C.R. Hollin & K.
Howells, pp. 151-78. Chichester: John Wiley & Sons.
Perkins, D.O., Davidson, E.J., Leserman, J. et al (1993). Personality disorder in
patients infected with HIV: a controlled study with implications for clinical
care. American Journal of Psychiatry, 150, 309-15.
Perry, J.C. (1990). Use of longitudinal data to validate personality disorders. In:
370 References
Personality Disorders: New Perspectives on Diagnostic Validity, ed. J.M.

Oldham, pp. 23^40. Washington DC: American Psychiatric Press Inc.
Perry, J.C. (1992). Problems and considerations in the valid assessment of
personality disorders. American Journal of Psychiatry, 149, 1645-53.
Perry, J.C. (1993). Longitudinal studies of personality disorders. Journal of
Personality Disorders, Spring suppl., 63-85.
Perry, J. & Klerman, G. (1980). Clinical features of borderline personality
disorder. American Journal of Psychiatry, 137, 165-73.
Perry, J . C , Lavori, P.W., Pagano, C.J. et al (1992). Life events and recurrent
depression in borderline and antisocial personality disorders. Journal of
Peselow, E.D., Sanfilipo, M.P., Fieve, R.R. & Gulbenkian, G. (1994). Personality
traits during depression and after clinical recovery. British Journal of
Peterson, C. & Seligman, M.E.P. (1984). Causal explanation as a risk factor for
depression: theory and evidence. Psychological Review, 91, 347-74.
Pfohl, B. (1991). Histrionic personality disorder: a review of available data and
Pfohl, B., Barash, J., True, B. & Alexander, B. (1989). Failure of two Axis I
measures to predict medication compliance among hypertensive patients.
Journal of Personality Disorders, 3, 45-52.
Pfohl, B. & Blum, N. (1991). Obsessive-compulsive personality disorder: a review
of available data and recommendations for DSM-IV. Journal of Personality
Pfohl, B., Coryell, W., Zimmerman, M. & Stangl, D. (1987). Prognostic validity
of self-report and interview measures of personality disorder in depressed
inpatients. Journal of Clinical Psychiatry, 48, 468-72.
Pfohl, B., Stangl, D. & Zimmerman, M. (1982). The Structured Interview for
DSM-III Personality Disorders (SIDP). Iowa City: University of Iowa
Hospitals and Clinics.
Pfohl, B., Stangl, D. & Zimmerman, M. (1984). The implications for DSM-III
personality disorders for patients with major depression. Journal of Affective
Disorders,!, 309-18.
Phillips, K.A., Gunderson, J.G., Hirschfeld, R.M. & Smith, L.E. (1990). The
depressive personality: a review. American Journal of Psychiatry, 147, 830-7.
Pica, S., Edwards, J., Jackson, H.J. et al (1990). Personality disorders in recent-
onset bipolar disorder. Comprehensive Psychiatry, 31, 499-510.
Pierce, D.W. (1977). Suicidal intent and self-injury. British Journal of Psychiatry,
130, 377-85.
Pilgrim, J. & Mann, A. (1990). Use of the ICD-10 version of the Standardized
Assessment of Personality to determine the prevalence of personality
disorder in psychiatric in-patients. Psychological Medicine, 20, 985-92.
Pilgrim, J.A., Mellers, J.D.. Boothby, H.A. & Mann, A.H. (1993). Inter-rater
and temporal reliability of the Standardized Assessment of Personality and
the influence of informant characteristics. Psychological Medicine, 23,
779-86.
Pinel, P.H. (1809). Traite Medico-Philosophique sur VAlienation Mentale, 2nd
edn. Paris, France: Brosson.
Pines, M. (1990). Group analytic psychotherapy and the borderline patient. In:
The Difficult Patient in Group Psychotherapy with Borderline and Narcissistic
Disorders, ed. B.E. Roth, W.N. Stone & H.D. Kibel, pp. 31-44. Madison,
References 371
Connecticut: International Universities Press Inc.

Piper, A. (1994). Multiple personality disorder. British Journal of Psychiatry, 164,
600-12.
Piran, N., Lerner, P. & Garfinkel, P.E. (1988). Personality disorders in anorexic
patients. International Journal of Eating Disorders, 7, 589-99.
Plakun, E.M. (1989). Narcissistic personality disorder. A validity study and
comparison to borderline personality disorder. Psychiatric Clinics of North
America, 12, 603-20.
Plakun, E.M. (1991). Prediction of outcome in borderline personality disorder.
Journal of Personality Disorders, 5, 93-101.
Plakun, E.M., Burkhardt, P.E. & Muller, J.P. (1985). 14-year follow-up of
borderline and schizotypal personality disroders. Comprehensive Psychiatry,
26, 448-55.
Pokorny, A.D., Miller, B.A. & Kaplan, H.B. (1972). The Brief MAST: a
shortened version of the Michigan Alcoholism Screening Test. American
Pollack, J.M. (1979). Obsessive-compulsive personality: a review. Psychological
Bulletin, 2, 225-41.
Pollack, J. (1981). Hysterical personality: an appraisal in light of empirical
research. Genetic Psychology Monographs, 104, 71-105.
Pollack, J. (1987). Relationship of obsessive-compulsive personality to obsessive-
compulsive disorder: a review of the literature. Journal of Psychology, 121,
137^8.
Pollack, M.H., Otto, M.W., Rosenbaum, J.F. & Sachs, G.S. (1992). Personality
disorders in patients with panic disorder. Comprehensive Psychiatry, 33,
78-83.
Pollock, N., McBain, I. & Webster, C D . (1989). Clinical decision making and
the assessment of dangerousness. In: Clinical Approaches to Violence, ed. K.
Howells & C.R. Hollin. Chichester: John Wiley & Sons.
Pope, H.G., Frankenburg, F.R., Hudson, J.I. et al. (1987). Is bulimia associated
with borderline personality disorder? A controlled study. Journal of Clinical
Pope, H.G. & Lipinsky, J.F. (1978). Diagnosis in schizophrenia and manic
depressive illness. Archives of General Psychiatry, 35, 811-28.
Powell, B.J., Penick, E.C., Nickel, E.J. et al. (1992). Outcomes of comorbid
alcoholic men: a 1-year follow-up. Alcohol Clinical and Experimental
Research, 16, 131-8.
Prentky, R.A. & Carter, D.L. (1984). The predictive value of the triad for sex
offenders. Behavioral Science and Law, 2, 341-54.
Prichard, J.C. (1835). A Treatise of Insanity. London, UK: Sherwood, Gilbert &
Piper.
Prichard, J.C. (1837). A treatise on insanity and other diseases affecting the mind.
Philadelphia: Harwell, Barrington & Harwell.
Quay, H.C. (1977). Measuring dimensions of deviant behaviour: The behavior
problem checklist. Journal of Abnormal Child Psychology, 5, 277-87.
Raczek, S.W. (1992). Childhood abuse and personality disorder. Journal of
Rado, S. (1953). Dynamics and classification of disordered behaviour. American
Raine, A. & Allbutt, J. (1989). Factors of schizoid personality. British Journal of
Clinical Psychology, 28,
372 References
Raine, A., Venables, P.H. & Williams, M. (1990). Relationships between central
and autonomic measures of arousal at age 15 years and criminality at age 24
years. Archives of General Psychiatry, 47, 1003-7.
Raskin, R. & Hall, C. (1979). A narcissistic personality inventory. Psychological
Reports, 45, 590-4.
Rasmussen, S.A. & Tsuang, M.T. (1986). Clinical characteristics and family
history in DSM-III obsessive-compulsive disorder. American Journal of
Ratey, J.J., Morrill, R. & Oxenkrug, G. (1983). Use of propranolol for provoked
and unprovoked episodes of rage. American Journal of Psychiatry, 140,
1356-7.
Reese, W.G. (1979). A dog model for human psychopathology. American Journal
Regier, D.A., Boyd, J.H. & Burke, J.D. (1988). One-month prevalence of mental
disorders in the United States. Archives of General Psychiatry, 45, 977-86.
Reich, J. (1987). Sex distribution of DSM-III personality disorders in psychiatric
outpatients. American Journal of Psychiatry, 144, 485-8.
Reich, J. (1988). DSM-III personality disorders and the outcome of treated panic
Reich, J.H. (1989). Update on instruments to measure DSM-III and DSM-III-R
personality disorders. Journal of Nervous and Mental Disease, 111, 366-70.
Reich, J. (1990a). Familiality of DSM-III-R self-defeating personality disorders.
Reich, J. (19906). Relationship between DSM-III avoidant and dependent
personality disorders. Psychiatry Research, 34, 281-920.
Reich, J.H. (1990c). Effect of DSM-III personality disorders on outcome of
tricyclic antidepressant-treated nonpsychotic outpatients with major or
minor depressive disorder. Psychiatry Research, 32, 175-81.
Reich, J. (1992). Measurement of DSM-III and DSM-III-R borderline
personality disorder. In: Borderline Personality Disorder: Clinical and
Empirical Perspectives, ed. J.F. Clarkin, E. Marziali & H. Munroe-Blum,
pp. 116-48. New York: Guilford Press.
Reich, J., Boerstler, H., Yates, W. & Nduaguba, M. (1989). Utilization of
medical resources in persons with DSM-III personality disorders in a
community sample. International Journal of Psychiatry in Medicine, 19, 1-9.
Reich, J.H. & Green, A.I. (1991). Effect of personality disorders on outcome of
treatment. Journal of Nervous and Mental Disease, 179, 7482.
Reich, J., Nduaguba, M. & Yates, W. (1988). Age and sex distribution of DSM-
III personality cluster traits in a community population. Comprehensive
Reich, J.H. & Noyes, R. (1987). A comparison of DSM-III personality disorders
in acutely ill panic and depressed patients. Journal of Anxiety Disorders, 1,
123-31.
Reich, J., Noyes, R., Hirschfeld, R. et al. (1987a). State and personality in
depressed and panic patients. American Journal of Psychiatry, 144, 181-7.
Reich, J.H., Noyes, R. & Troughton, E. (19876). Dependent personality disorder
associated with phobic avoidance in patients with panic disorder. American
Reich, J., Noyes, R. & Yates, W. (1989). Alprazolam treatment of avoidant
personality traits in social phobic patients. Journal of Clinical Psychiatry, 50,
91-5.
References 373
Reich, J.H. & Vasile, R.G. (1993). Effect of personality disorders on the
treatment outcome of Axis I conditions: an update. The Journal of Nervous
and Mental Disease, 181, 475-84.
Remington, G J . & Book, H. (1993). Discriminative validity of the borderline
syndrome index. Journal of Personality Disorders, 7, 312-9.
Renneberg, B., Goldstein, A.J., Phillips, D. & Chambless, D.L. (1990). Intensive
behavioral group treatment of avoidant personality disorder. Behavior
Therapy, 21, 363-77.
Reus, N.D. & Markrow, S. (1984). Alprazolam in the treatment of borderline
personality disorder. Paper presented at the 39th annual meeting of the
Society of Biological Psychiatry, Los Angeles, CA.
Rey, J.M. (1993). Oppositional defiant disorder. American Journal of Psychiatry,
150, 1769-78.
Rich, C.L., Fowler, R.C. & Fogarty, L.A. (1988). The San Diego Suicide Study
III: relationships between diagnoses and stressors. Archives of General
Richman, J.A. & Flaherty, J.A. (1987). Narcissistic Trait Scale (NTS),
Department of Psychiatry, University of Illinois, College of Medicine, 912 S.
Wood Street, Room 218, Chicago, IL. 60612, USA.
Rifkin, A., Quitkin, F. & Carrillo, C. (1972). Lithium carbonate in emotionally
unstable character disorders. Archives of General Psychiatry, 27, 519-23.
Rihmer, Z. (1990). Dysthymia - a clinician's perspective. In: Dysthymic Disorder,
ed. S.W. Barton & H.S. Akiskal. London: Gaskell (Royal College of
Psychiatrists).
Robertson, D.A.F., Ray, J., Diamond, I. & Edwards, J.G. (1989). Personality
profile and affective state of patients with inflammatory bowel disease. Gut,
30, 623-6.
Robertson, G. & Gunn, J. (1987). A ten-year follow-up of men discharged from
Grendon Prison. British Journal of Psychiatry, 151, 674-8.
Robins, E. & Guze, S.B. (1970). Establishment of diagnostic validity in
psychiatric illness. Its application to schizophrenia. American Journal of
Robins, L.N. (1966). Deviant Children Grown Up: A Sociological and Psychiatric
Study of Sociopathic Personality. Baltimore: Williams & Watkins.
Robins, L.N. (1978). Sturdy childhood predictors of adult antisocial behaviour:
replications from longitudinal studies. Psychological Medicine, 8, 611-22.
Robins, L.N. (1986). The consequences of conduct disorders in girls. In:
Development of Antisocial and Prosocial Behaviour. Research, Theories and
Issues, ed. D. Olwens, J. Block & M. Radke-Yarrow. Orlando: Academic
Press.
Robins, L.N. (1993). Childhood conduct problems, adult psychopathology, and
crime. In: Mental Disorder and Crime, ed. S. Hodgins. London: Sage.
Robins, L.N., Helzer J.E. & Weissman, M.M. (1984). Lifetime prevalence of
specific psychiatric disorders in three sites. Archives of General Psychiatry,
41, 949-58.
Robins, L.N. & Price, R.K. (1991). Adult disorders predicted by childhood
conduct problems: results from the NIMH epidemiologic catchment area
project. Psychiatry, 54, 116-32.
Robins, L.N. & Regier, D.A. (1991). Psychiatric Disorder in America. The ACA
Study. New York: Free Press.
Robins, L.N., Tipp, P. & Przybeck, T. (1991). Antisocial personality. In:
374 References
Psychiatric Disorder in America. The ECA Study, ed. L.N. Robins & D.
Regier, pp. 258-290. New York: Free Press.
Robinson, A.D.T. & Duffy, J.C. (1989). A comparison of self-injury and self-
poisoning from the Regional Poisoning Treatment Centre, Edinburgh. Acta
Robinson, L.A., Berman, J.S. & Neimeyer, R.A. (1990). Psychotherapy for the
treatment of depression: a comprehensive review of controlled outcome
research. Psychological Bulletin, 108, 77-84.
Ronningstam, E. (1988). Comparing three systems for diagnosing narcissistic
personality disorder. Psychiatry, 51, 300-11.
Ronningstam, E. & Gunderson, J. (1988). Narcissistic traits in psychiatric
patients. Comprehensive Psychiatry, 29, 545-9.
Ronningstam, E. & Gunderson, J. (1989). Descriptive studies on narcissistic
personality disorder. Psychiatric Clinics of North America, 12, 585-601.
Ronningstam, E. & Gunderson, J. (1990). Identifying criteria for narcissistic
Ronningstam, E. & Gunderson, J. (1991). Differentiating borderline personality
disorder from narcissistic personality disorder. Journal of Personality
Rosenvinge, J.H. & Mouland, S.O. (1990). Outcome and prognosis of anorexia
nervosa. British Journal of Psychiatry, 156, 92-7.
Ross, H.E., Glaser, F.B. & Germanson, T. (1988). The prevalence of psychiatric
disorders in patients with alcohol and other drug problems. Archives of
Rossiter, E.M., Agras, W.S., Telch, C.F. & Schneider, J.A. (1993). Cluster B
personality disorder characteristics predict outcome in the treatment of
bulimia nervosa. International Journal of Eating Disorders, 13, 349-57.
Rounsaville, B.J., Dolinsky, Z.S., Babor, T.F. & Meyer, R.E. (1987).
Psychopathology as a predictor of treatment outcome in alcoholics. Archives
Rounsaville, B.J., Kosten, T.R. & Kleber, H.D. (1986). Long-term changes in
current psychiatric diagnoses of treated opiate addicts. Comprehensive
Roy, A., Custer, R., Lorenz, V. & Linnoila, M. (1989a). Personality factors and
pathological gambling. Acta Psychiatrica Scandinavica, 80, 37-9.
Roy, A., De Jong, J. & Linnoila M. (19896). Extraversion in pathological
gamblers. Archives of General Psychiatry, 46, 679-81.
Runeson, B. & Beskow, J. (1991). Borderline personality disorder in young
Swedish suicides. Journal of Nervous and Mental Disease, 179, 153-6.
Rushton, J.P., Fulker, D.W., Neale, M.C. et al (1986). Altruism and aggression:
the heritability of individual differences. Journal of Personality and Social
Russ, M.J. (1992). Self-injurious behavior in patients with borderline personality
disorder: biological perspective. Journal of Personality Disorders, 6, 64-81.
Rutter, M. (1987). Temperament, personality and personality disorder. British
Rutter, M. & Giller, H. (1983). Juvenile Delinquency: Trends and Perspectives.
Harmondsworth: Penguin.
Ryder, R.G. (1967). Birth to maturity revised: a canonical reanalysis. Journal of
Personality and Social Psychology, 7, 168-72.
Ryle, A. (1976). Group psychotherapy. British Journal of Hospital Medicine,
References 375
March, 239-43.
Ryle, A. (1989). Cognitive-Analytic Therapy: Active Participation in Change.
London: John Wylie.
Sanderson, W.C., Wetzler, S., Beck, A.T. & Betz, F. (1992). Prevalence of
personality disorders in patients with major depression and dysthymia.
Psychiatry Research, 42, 93-99.
Sanderson, W.C., Wetzler, S., Beck, A.T. & Betz, F. (1994). Prevalence of
personality disorders among patients with anxiety disorders. Psychiatry
Research, 51, 167-74.
Sandier, J., Dare, C. & Holder, A. (1970). Basic psychoanalytic concepts:
transference. British Journal of Psychiatry, 116, 667-72.
Sandier, J. & Hazari, A. (1960). The 'obsessional'. On the psychological
classification of obsessional character traits and symptoms. British Journal
of Medical Psychology, 33, 113-22.
Sandier, J., Holder, A. & Dare, C. (1970a). Basic psychoanaltyic concepts: the
treatment alliance. British Journal of Psychiatry, 116, 555-61.
Sandier, J., Holder, A. & Dare, C. (19706). Basic psychoanalytic concepts:
resistance. British Journal of Psychiatry, 111, 215-9.
Sano, K. & Mayangi, Y. (1988). Posteromedial hypothalamotomy in the
treatment of violent, aggressive behaviour. Acta Neurochirurgica, Suppl. 44,
145-51.
Sansone, R.A. & Fine, M.A. (1992). Borderline personality as a predictor of
outcome in women with eating disorders. Journal of Personality Disorders, 6,
176-86.
Sapsford, R.J. (1978). Life sentenced prisoners: psychological changes during
sentence. British Journal of Criminology, 18, 128-45.
Satel, S., Southwick, S. & Denton, C. (1988). Use of imipramine for attention
deficit disorder in a borderline patient. Journal of Nervous and Mental
Disease, 176, 305-7.
Sato, T., Sakado, K. & Sato, S. (1993a). DSM-III-R personality disorders in
outpatients with non-bipolar depression. European Archives of Psychiatry
and Clinical Neuroscience, 242, 273-8.
Sato, T., Sakado, K. & Sato, S. (19936). Is there any specific personality disorder
or personality disorder cluster that worsens the short-term treatment
outcome of major depression? Acta Psychiatrica Scandinavica, 88, 342-9.
Scadding, J.G. (1988). Health and disease: what can medicine do for philosophy.
Journal of Medical Ethics, 14, 118-24.
Scadding, J.G. (1990). The semantic problems of psychiatry. Psychological
Medicine, 20, 243-8.
Schalling, D., Asberg, M., Edman, G. & Oreland, L. (1987). Markers for
vulnerability to psychopathology: temperament traits associated with
platelet MAO activity. Acta Psychiatrica Scandinavica, 76, 172-82.
Schiff, S.B. & Glassman, S.M. (1969). Large and small group therapy in a State
mental health centre. International Journal of Group Psychotherapy, 19,
2-10.
Schneider, K. (1923). Die Psychopathischen Personlichkeiten. Berlin: Springer.
Schneider, K. (1959). Clinical Psychopathology, trans. M.W. Hamilton. London:
Grune & Stratton.
Schneier, F.R., Spitzer, R.L., Gibbon, M. et al (1991). The relationship of social
phobias subtypes and avoidant personality disorder. Comprehensive
376 References
Schroeder, M.L. & Livesley, W J . (1991). An evaluation of DSM-III-R
personality disorders. Acta Psychiatrica Scandinavica, 84, 512-9.
Schuckit, M.A. (1985). The clinical implications of primary diagnostic groups
among alcoholics. Archives of General Psychiatry, 42(11), 1043-9.
Schulsinger, F., Parnas, J., Petersen, E.T. et al. (1984). Cerebral ventricular size
in the offspring of schizophrenia mothers. Archives of General Psychiatry,
41, 602-6.
Schulz, S.C., Cornelius, J., Schulz, P.M. & Soloff, P.H. (1988). The amphetamine
challenge test in patients with borderline disorder. American Journal of
Schutz, W.C. (1958). In: Firo: A Three-Dimensional Theory of Interpersonal
Behaviour. New York: Rinehart.
Sciuto, G., Diaferia, G., Battaglia, M. et al. (1991). DSM-III-R personality
disorders in panic and obsessive-compulsive disorder: a comparison study.
Scott, J. (1993). Homelessness and mental illness. British Journal of Psychiatry,
162, 314-24.
Seivewright, N. (1987). Relationship between life events and personality in
psychiatric disorder. Stress Medicine, 3, 163-8.
Shapiro, E.R. (1978). The psychodynamics and developmental psychology of the
borderline patient: a review of the literature. American Journal of Psychiatry,
135, 1305-15.
Shea, M.T. (1992). Some characteristics of the Axis II criteria sets and their
implications for assessment of personality disorders. Journal of Personality
Shea, M.T. (1993). Psychosocial treatment of personality disorders. Journal of
Personality Disorders, (Suppl.), 167-80.
Shea, M.T., Pilkonis, P.A., Beckham, E. et al. (1990). Personality disorders and
treatment outcome in the NIMH treatment of depression collaborative
research program. American Journal of Psychiatry, 147, 711-8.
Sheard, M.H. (1971). Effect of lithium on human aggression. Nature, 230, 113-4.
Sheard, M.H., Marini, J.L. & Bridges, C.I. (1976). The effect of lithium on
unipolar aggressive behaviour in man. American Journal of Psychiatry, 133,
1409-13.
Shearin, E.N. & Linehan, M.M. (1993). Dialectical behavior therapy for
borderline personality disorder: Treatment goals, strategies and empirical
support. In: Borderline Personality Disorder, ed. J. Paris, pp. 285-318.
Washington DC: American Psychiatric Press Inc.
Shekim, W.O., Bylund, D.B. & Frankel, F. (1990). Platelet alpha 2 adrenergic
receptor binding to 3H yohimbine and personality variations in normals.
Psychiatry Research, 32, 125-34.
Sherwood, R.J., Funari, D.J. & Piekarski, A.M. (1990). Adapted character styles
of Vietnam veterans with post-traumatic stress disorder. Psychological
Reports, 66, 623-31.
Sierles, F.S., Chen, J.J., McFarland, R.E. & Taylor, M.A. (1983). Posttraumatic
stress disorder and concurrent psychiatric illness: a preliminary report.
American Journal of Psychiatry, 140, 11779.
Siever, L.J., Bernstein, D.P. & Silverman, J.M. (1991). Schizotypal personality
disorder: a review of its current status. Journal of Personality Disorders, 5,
178-93.
Siever, L.J. & Davis, K.L. (1991). A psychobiological perspective on the
References ill
Siever, L.J., Keefe, R., Bernstein, D.P. et al. (1990#). Eye tracking impairment in
clinically identified patients with schizotypal personality disorder. American
Siever, L.J., Silverman, J.M., Horvath, T.B. et al (19906). Increased morbid risk
for schizophrenia-related disorders in relatives of schizotypal personality
disordered patients. Archives of General Psychiatry, 47, 63440.
Silver, D. (1985). Psychodynamics and psychotherapeutic management of the
self-destructive character-disordered patient. Psychiatric Clinics of North
America, 8, 357-75.
Silverman, J.M., Siever, L.J., Horvath, T.B. et al. (1993). Schizophrenia-related
and affective personality disorder traits in relatives of probands with
schizophrenia and personality disorders. American Journal of Psychiatry,
150, 435-42.
Simeon, D., Stanley, B., Frances, A. et al. (1992). Self-mutilation in personality
disorders: psychological and biological correlates. American Journal of
Simpson, D.M. & Foster, D. (1986). Improvement in organically disturbed
behaviour with trazodine treatment. Journal of Clinical Psychiatry, 47,
191-3.
Sims, A.C.P. (1992). Neuroses and personality disorders. Current Opinion in
Skodol, A.E., Rosnick, L., Kellman, D. et al. (1988). Validating structured
DSM-III-R personality disorder assessments with longitudinal data.
Skodol, A.E., Rosnick, L., Kellman, D. et al. (1990). Developments of a
procedure for validating structured assessments of Axis II. In: Personality
Disorders: New Perspectives in Diagnostic Validity, ed. J.M. Oldham, pp. 4 1 -
70. Washington DC: American Psychiatric Press Inc.
Small, I.F., Small, J.G., Alig, V.B & Moore, D.F. (1970). Passive-aggressive
personality disorder: a search for a syndrome. American Journal of
Smith, D. (ed.) (1979). Life Sentenced Prisoners. Home Office Research Study
No. 51. London: HMSO.
Smith, M.L., Glass, G.V. & Miller, T.I. (1980). The Benefits of Psychotherapy.
Baltimore: The John Hopkins University Press.
Sohlberg, S. (1990). Personality, life stress and the course of eating disorders.
Ada Psychiatrica Scandinavica, Suppl. 361, 82, 29-33.
Sohlberg, S., Norring, C , Holmgren, S. & Rosmark, B. (1989). Impulsivity and
long-term prognosis of psychiatric patients with anorexia nervosa/bulimia
nervosa. Journal of Nervous and Mental Disease, 111, 249-58.
Soloff, P.H. (1987). Neuroleptic treatment in the borderline patient: advantages
and techniques. Journal of Clinical Psychiatry, 48 (8 Suppl.), 26-30.
Soloff, P.H. (1993). Pharmacological therapies in borderline personality disorder.
In: Borderline Personality Disorder, ed. J. Paris, pp. 31948. Washington
DC: American Psychiatric Press Inc.
Soloff, P.H., Cornelius, J., George, A. et al. (1993). Efficacy of phenelzine and
haloperidol in borderline personality disorder. Archives of General
Soloff, P.H., George, A. & Nathan, R.S. (1986). Progress in pharmacotherapy of
borderline disorders: a double blind study of amitriptyline, haloperidol and
378 References
placebo. Archives of General Psychiatry, 43, 691-9.
Solomon, M.I. & Murphy, G.E. (1984). Cohort studies in suicide. In: Suicide in
the Young, ed. H.S. Sudak, A.B. Ford & N.B. Rushford. Boston: John
Wright/PSG.
Southwick, S.M., Yehuda, R. & Giller, E.L. (1993). Personality disorders in
treatment-seeking combat veterans with posttraumatic stress disorder.
Southwick, S.M., Yehuda, R., Giller, E.L. & Perry, B.D. (1990). Altered platelet
a2 adrenergic receptor binding sites in borderline personality disorder.
Spitzer, R.L. (1983). Psychiatric diagnosis: are clinicians still necessary?
Spitzer, R.L., Endicott, J. & Woodruff, R.A. (1977). Classification of mood
disorders. In: Depression: Clinical, Biological and Psychological Perspectives,
ed. G. Usdin. New York: Brunner/Mazel.
Spitzer, R.L., Feister, S., Gay, M. & Pfohl, B. (1991). Results of a survey of
forensic psychiatrists on the validity of the sadistic personality disorder
diagnosis. American Journal of Psychiatry, 148, 875-9.
Spitzer, R., Williams, J.B.W. & Gibbon, M. (1987). Structured Interview for
DSM-III-R Personality Disorders. New York: Biometrics Research
Department, New York State Psychiatric Institute.
Spitzer, R.L., Williams, J.B.W., Kass, F. & Davies, M. (1989). National field
trial of the DSM-III-R diagnostic criteria for self-defeating personality
Sprock, J., Blashfield, R.K. & Smith, B. (1990). Gender weighting of DSM-III-R
personality disorder criteria. American Journal of Psychiatry, 147, 586-90.
Squires-Wheeler, E., Skodol, A.E., Bassett, A. & Erlenmeyer-Kimling, L. (1989).
DSM-III-R schizotypal personality traits in offspring of schizophrenic
disorder, affective disorder, and normal control patients. Journal of
Psychiatric Research, 23, 229-39.
Srole, L., Langer, T., Michael, S. et al. (1962). Mental Health in the Metropolis.
New York: McGraw Hill.
Standage, K. (1989). Structured interviews and the diagnosis of personality
disorders. Canadian Journal of Psychiatry, 34, 906-12.
Stangl, D., Pfohl, B., Zimmerman, M. et al. (1985). A structured interview for
the DSM-III personality disorders. Archives of General Psychiatry, 42,
591-6.
Stankovic, S.R., Libb, J.W., Freeman, H.M. & Roseman, J.M. (1992). Post
treatment stability of the MCMI-II personality scales in depressed
outpatients. Journal of Personality Disorders, 6, 82-9.
Stanley, M.A., Turner, S.M. & Borden, J.W. (1990). Schizotypal features in
obsessive-compulsive disorder. Comprehensive Psychiatry, 31, 511-8.
Stein, G. (1993). Drug treatment of the personality disorders. In: Personality
Disorder Reviewed, ed. P. Pyrer & G. Stein, pp. 262-304. London: Gaskell.
Steiner, M., Links, P.S. & Korzekwa, M. (1988). Biological markers in borderline
personality disorders: an overview. Canadian Journal of Psychiatry, 33,
350-5.
Stern, A. (1938). Psychoanalytic investigation of and therapy in the borderline
group of neuroses. Psychoanalytic Quarterly, 1, 467-89.
Stern, J., Murphy, M. & Bass, L. (1993). Personality disorders in patients with
somatisation disorder. British Journal of Psychiatry, 163, 785-9.
References 379
Sternbach, S.E., Judd, P.H., Sabo, A.N. et al. (1992). Cognitive and perceptual
distortions in borderline personality disorder and schizotypal personality
disorder in a vignette sample. Comprehensive Psychiatry, 33, 186-9.
Stevenson, J. & Meares, R. (1992). An outcome study of psychotherapy in
borderline personality disorder. American Journal of Psychiatry, 149,
358-62.
Stone, J.L., McDaniel, K.D., Hughes, J.R. & Hermann, B.P. (1986). Episodic
dyscontrol disorder and paroxysmal EEG abnormalities: successful
treatment with carbamazepine. Biological Psychiatry, 21, 208-12.
Stone, M.H. (1990). The Fate of Borderline Patients. New York: Guilford Press.
Stone, M.H. (1992). Borderline personality disorder: course of illness. In:
Borderline Personality Disorder, ed. J.F. Clarkin, E. Marziali & H. Munroe-
Blum, pp. 67-85. New York: Guilford Press.
Stone, M.H. (1993). Long-term outcome in personality disorders. British Journal
Stone, M.H., Stone, D.K. & Hurt, S.W. (1987). Natural history of borderline
patients treated by intensive hospitalization. Psychiatric Clinics of North
America, 10, 185-206.
Stravynski, A., Lesage, A., Marcouiller, M. & Edie, R. (1989). A test of the
therapeutic mechanism in social skills training with avoidant personality
disorders. Journal of Nervous and Mental Disease, 111, 739-44.
Stravynski, A., Marks, I. & Yule, W. (1982). Social skills problems in neurotic
outpatients: social skills training with and without cognitive modification.
Streiner, D.L. & Norman, G.R. (1989). Selecting the items. In: Health
Measurement Scales, pp. 44-6. Oxford, UK: Oxford University Press.
Strelau, J. (1983). Temperament, Personality and Arousal. London: Academic
Press.
Stringer, A.Y. & Josef, N.C. (1983). Methylphenidate in the treatment of
aggression in two patients with antisocial personality disorder. American
Stuart, S., Simons, A.D., Thase, M.E. & Pilkonis, P. (1992). Are personality
assessments valid in acute major depression? Journal of Affective Disorders,
24,281-90.
Suedfeld, P. & Landon, P.B. (1978). Approaches to treatment. In: Psychopathic
Behaviour: Approaches to Research, ed. R.D. Hare & D. Schalling, pp. 347-
76. New York: Wiley.
Suomi, S.J. (1983). Social development in rhesus monkey: considerations of
individual differences. In: The Behavior of Human Infants, ed. A Oliverio &
M. Zappella. New York and London: Plenum Press.
Svartberg, M. & Styles, T.C. (1991). Comparative effects of short-term
psychodynamic psychotherapy: a meta-analysis. Journal of Consulting and
Clinical Psychology, 59, 704^14.
Svrakic, D.M., Przybeck, T.R. & Cloninger, C.R. (1992). Mood states and
personality traits. Journal of Affective Disorders, 24, 217-26.
Swanson, M.C.J., Bland, R.C. & Newman, S.C. (1994). Antisocial personality
disorders. Acta Psychiatrica Scandinavica, suppl. 376, 63-70.
Szatmari, P., Bartolucci, G.& Bremner,R. (1989). A follow-up study of high-
functioning autistic children. Journal of Autism and Developmental
Disorders, 19,213-25.
Tantam, D. (1988). Lifelong eccentricity and social isolation. British Journal of
380 References
Tantam, D. & Whittaker, J. (1992). Personality disorder and self-wounding.
Tarnopolsky, A. & Berelowitz, M. (1987). Borderline personality. A review of
recent research. British Journal of Psychiatry, 151, 724-34.
Taylor, P.J. (1986). Psychiatric disorder in London's life sentenced offenders.
British Journal of Criminology, 26, 63-78.
Teicher, M.H., Glod, C.A. & Cole, J.D. (1990). Emergence of intensive suicidal
preoccupation during fluoxetine treatment. American Journal of Psychiatry,
Ul, 207-10.
Tellegen, A., Lykken, D.T., Bouchard, T.J. et al. (1988). Personality similarity in
twins reared apart and together. Journal of Personality and Social
Templeman, T.L. & Wollersheim, J.P. (1979). A cognitive-behavioral approach
to the treatment of psychopathy. Psychotherapy: Theory, Research and
Practice, 16, 132-9.
Thaker, G., Adami, H., Moran, M. et al. (1993). Psychiatric illness in families of
subjects with schizophrenia-spectrum personality disorders. American
Tiihonen, J. & Eronen, M. (1993). Criminality associated with mental disorders
and intellectual deficiency. Acta Psychiatrica Scandinavica, 50, 917-8.
Tiihonen, J. & Hakola, P. (1994). Psychiatric disorders and homicide recidivism.
Tong, J.E. & McKay, G.W.(1959). A statistical follow-up of mental defectives of
dangerous and violent propensities. British Journal of Delinquency, 9,
276-84.
Torgerson, S. (1984). Genetic and nosological aspects of schizotypal and
borderline personality disorders. Archives of General Psychiatry, 41, 546-54.
Torgerson, S., Onstad, S., Skre, I., Edvardsen, J. & Kringlen, E. (1993). True'
schizotypal personality disorder: a study of co-twins and relatives of
schizophrenic probands. American Journal of Psychiatry, 150, 1661-7.
Torgerson, S. & Psychol, C. (1980). The oral, obsessive, and hysterical
personality syndromes: a study of hereditary and environmental factors by
means of the twin method. Archives of General Psychiatry, 144, 767-72.
Trower, P., Yardley, K., Bryant, B.M. & Shaw, P. (1978). The treatment of
social failure: a comparison of anxiety-reduction and skills-acquisition
procedures on two social problems. Behavior Modification, 2, 41-60.
Trull, T.J., Widiger, T.A. & Frances, A. (1987). Co-variation of criteria sets for
avoidant, schizoid and dependent personality disorder. American Journal of
Tupin, J.P., Smith, D.B. & Clanon, T.C. (1973). The long term use of lithium, in
aggressive prisoners. Comprehensive Psychiatry, 14, 311-7.
Turkat, I.D. (1990). The Personality Disorders. A Psychological Approach to
Clinical Management. New York: Pergamon Press.
Turner, S.H., Beidel, D.C., Dancee, C.V. & Keys, D.J. (1986). Psychopathology
of social phobia and comparison to avoidant personality disorder. Journal of
Abnormal Psychology, 95, 389-94.
Tyrer, P. (1989). Clinical importance of personality disorder. Current Opinion in
Tyrer, P. & Alexander, M.S. (1979). Classification of personality disorder. British
References 381
Tyrer, P. & Alexander, J. (1988). Personality assessment schedule. In: Personality
Disorders: Diagnosis, Management and Course, pp. 43-62. London: Wright.
Tyrer, P., Casey, P. & Gall, J. (1983). Relationship between neurosis and
personality disorder. British Journal of Psychiatry, 142, 404-8.
Tyrer, P. & Ferguson, B. (1988). Development of the concept of abnormal
personality. In: Personality Disorders: Diagnosis, Management and Course,
ed. P. Tyrer, pp. 1-11. London: Wright.
Tyrer, P. & Seivewright, N. (1988a). Pharmacological treatment of personality
disorders. Clinical Neuropharmacology, 11, 493-9.
Tyrer, P. & Seivewright, N. (1988&). Studies of outcome. In: Personality
Disorders; Diagnosis, Management and Course, ed. P. Tyrer, pp. 119-36.
London: Wright.
Tyrer, P., Seivewright, N., Ferguson, B. et al (1990). The Nottingham study of
neurotic disorder: relationship between personality status and symptoms.
Psychological Medicine, 20, 423-31.
Tyrer, P., Seivewright, N., Ferguson, B. et al. (1993). The Nottingham study of
neurotic disorder. British Journal of Psychiatry, 162, 219-26.
Tyrer, P., Seivewright, N., Ferguson, B. & Tyer, J. (1992). The general neurotic
syndrome: a coaxial diagnosis of anxiety, depression and personality
disorder. Acta Psychiatrica Scandinavica, 85, 201-6.
Vaglum, S. & Vaglum, P. (1989). Comorbidity for borderline schizotypal
Vaillant, G.E. (1975). Sociopathy as a human process: a viewpoint. Archives of
Van der Kolk, B.A., Perry, J.C. & Herman, J.L. (1991). Childhood origins of
self-destructive behavior. American Journal of Psychiatry, 148, 1665-71.
van Manen, J. & van Veelen, C.W.M. (1988). Experience in psycho-surgery in
the Netherlands. Acta Neurochirurgica, Suppl. 44, 167-9.
Virkkunen, M., De Jong, J., Bartko, J. et al. (1989). Relationship of
psychobiological variables to recidivism in violent offenders and impulsive
fire setters. Archives of General Psychiatry, 46, 600-3.
Virkunnen, M., Nuutila, A., Goodwin, F.K. & Linnoila, M. (1987).
Cerebrospinal fluid monoamine metabolite levels in male arsonists. Archives
Von Knorring, L., Almay, B.G.L. & Johansson, F. (1987). Personality traits in
patients with idiopathic pain disorder. Acta Psychiatrica Scandinavica, 76,
490-8.
Wagner, A.W. & Linehan, M.M. (1994). Relationship between childhood sexual
abuse and topography of parasuicide among women with borderline
personality disorder. Journal of Personality Disorders, 8, 1-9.
Waldinger, R. & Gunderson, J. (1984). Completed psychotherapies with
borderline patients. American Journal of Psychotherapy, 38, 190-202.
Walker, N. & McCabe, S. (1973). Crime and Insanity in England, vol. 2.
Edinburgh: Edinburgh University Press.
Wallerstein, R. (1986). Forty-two Lives in Treatment: A Study of Psychoanalysis
and Psychotherapy. New York: The Guilford Press.
Walton, H.J. (1973). Abnormal personality. In: Companion to Psychiatric
Studies, vol. II, pp. 15-37. Edinburgh: Churchill Livingstone.
Walton, H.J. & Presly, A.S. (1973). Use of a category system in the diagnosis of
abnormal personality. British Journal of Psychiatry, 122, 259-68.
Weinryb, R.M., Gustavsson, J.P., Asberg, M. & Rossel, R.J. (1992). Stability
382 References
over time of character assessment using a psychodynamic instrument and

personality inventories. Ada Psychiatrica Scandinavica, 86, 179-84.
Weiss, G. (1985). Hyperactivity. Overview and new directions. Psychiatric Clinics
of North America, 8, 737-53.
Weiss, G., Hechtman, L., Perlman, T. et al. (1979). Hyperactives as young adults.
Weiss, R.D. & Mirin, S.M. (1986). Subtypes of cocaine abusers. Psychiatric
Clinics of North America, 9, 491-501.
Weissman, M.M. (1993). The epidemiology of personality disorders: a 1990
update. Journal of Personality Disorders, Spring suppl., 44-62.
Weissman, M.M., Klerman, G.L. & Prusoff, B. (1981). Depressed outpatients
one year after treatment with drugs and/or psychotherapy. Archives of
Weissman, M.M., Myers, J.K. & Harding, P.S. (1978). Psychiatric disorder in a
US urban community. American Journal of Psychiatry, 135, 459-62.
Welch, L.W. & Bear, D. (1990). Organic disorders of personality. In: Treating
Personality Disorders, ed. D.H. Adler. San Francisco: Jossey-Bass Inc.,
Publishers.
Wells, J.E., Bushnell, J.A., Hornblow, A.R. et al. (1989). Christchurch
Psychiatric Epidemiology Study. Part I: Methodology and lifetime
prevalence for specific psychiatric disorders. Australian and New Zealand
Wender, P.H., Reimherr, R.W., Wood, D. & Ward, M. (1985). A controlled
study of methylphenidate in the treatment of attention deficit disorder,
residual type, in adults. American Journal of Psychiatry, 142, 547-52.
Wender, P.H., Wood, D. & Reimherr, F. (1984). Studies in attention deficit
disorder, residual type (minimal brain dysfunction in adults).
Psychopharmacology Bulletin, 20, 18-20.
Werner, E.E. (1985). Stress and protective factors in children's lives. In:
Longitudinal Studies in Child Psychology and Psychiatry, ed. A.R. Nicol.
Chichester: Wiley.
West, D.J. (1983). The childhood origins of delinquency and antisocial
behaviour. In: Handbook of Psychiatry, vol. 4, The Neuroses and Personality
Disorders, ed. G.F.M. Russell & L. Hersov. Cambridge, UK: Cambridge
University Press.
West, D.J. & Farrington, D.P. (1977). The Delinquent Way of Life. London:
Heinemann Educational Books.
Westen, D., Ludolph, P., Misle, B. et al. (1990). Physical and sexual abuse in
adolescent girls with borderline personality disorder. American Journal of
Orthopsychiatry, 60, 55-66.
Weston, S.C. & Siever, L.J. (1993). Biologic correlates of personality disorders.
Journal of Personality Disorders, Spring Suppl., 12948.
Whitehorn, J.C. (1952). Basic psychiatry in medical practice. Journal of the
American Medical Association, 148, 329-32.
Whiteley, J.S. (1970). The response of psychopaths to a therapeutic community.
Whitman, R.M., Trosman, H. & Koenig, R. (1954). Clinical assessment of
passive-aggressive personality. Archives of Neurology and Psychiatry, 72,
540-9.
Widiger, T.A. (1987). Personality Interview Questions II (PIQ-II). University of
Kentucky, Lexington, KY 40506, USA.
References 383
Widiger, T.A. (1989). The categorical distinction between personality and
affective disorders. Journal of Personality Disorders, 3, 77-91.
Widiger, T.A. (1992). Categorical versus dimensional classification: implications
from and for research. Journal of Personality Disorders, 6, 287-300.
Widiger, T.A. & Frances, A. (1985a). The DSM-III personality disorders.
Widiger, T.A. & Frances, A. (19856). Axis II personality disorders: diagnostic
and treatment issues. Hospital and Community Psychiatry, 36, 619-27.
Widiger, T.A. & Frances, A. (1987). Interviews and inventories for the
measurement of personality disorders. Clinical Psychology Review, 7, 49-75.
Widiger, T.A., Frances, A., Spitzer, R.L. & Williams, J.B.W. (1988). The DSM-
III-R personality disorders: an overview. American Journal of Psychiatry,
145, 786-95.
Widiger, T.A. & Shea, T. (1991). Differentiation of Axis I and Axis II disorders.
Journal of Abnormal Psychology, 100, 399^406.
Williams, J.B.W. & Spitzer, R.L. (1982). Idiopathic pain disorder: a critique of
pain-prone disorder and a proposal for a revision of the DSM-III category
psychogenic pain disorder. Journal of Nervous and Mental Disease, 8, 415-9.
Wimer, R.E. & Wimer, C.C. (1985). Animal behavior genetics: a search for the
biological foundations of behavior. Annual Review of Psychology, 36,
171-218.
Wink, P. & Gough, H.G. (1990). New narcissism scales for the California
Psychological Inventory and MMPI. Journal of Personality Assessment, 54,
446-62.
Winnicott, D.W. (1949). Hate in the countertransference. International Journal of
Winokur, G., Black, D.W. & Nasrallah, A. (1988). Depressions secondary to
other psychiatric disorders and medical illnesses. American Journal of
Winston, A., Laikin, M., Pollack, J. et al. (1994). Short-term psychotherapy of
Winston, A., Pollack, J., McCullough, L. et al. (1991). Brief psychotherapy of
Wise, T.N., Mann, L.S. & Shay, L. (1992). Alexithymia and the five-factor model
of personality. Comprehensive Psychiatry, 33, 147-51.
Woerner, P.I. & Guze, S.B. (1968). A family and marital study of hysteria.
Wolf, A. & Schwartz, E.K. (1962). Psychoanalysis in Groups. New York: Grune
& Stratton.
Wolff, S. (1983). The impact of sociopathic and inadequate parents on their
children (including child abuse). In: Handbook of Psychiatry, Vol. 4. The
Neuroses and Personality Disorders, ed. G.F.M. Russell & L. Hersov, pp.
410-4. Cambridge: Cambridge University Press.
Wolff, S. (1991). 'Schizoid' personality in childhood and adult life, I, II, III.
Wolff, S. & Chick, J. (1980). Schizoid personality in childhood: a controlled
follow-up study. Psychological Medicine, 10, 85-100.
Wolfgang, M., Figlio, R.F. & Sellin, T. (1972). Delinquency in a Birth Cohort.
Chicago: University of Chicago Press.
Wolkind, S.N. & Rutter, M. (1985). Separation, loss and family relationships. In:
Child and Adolescent Psychiatry: Modern Approaches, 2nd edn, ed. M.
384 References
Rutter & C. Hersov. Oxford: Blackwell Scientific.
Wonderlich, S.A., Fullerton, D., Swift, W J . & Klein, M.H. (1994). Five-year
outcome from eating disorders: relevance of personality disorders.
International Journal of Eating Disorders, 15, 233-43.
Wong, N. (1980). Combined group and individual treatment of borderline and
narcissistic patients: heterogeneous vs. homogeneous groups. International
Journal of Group Psychotherapy, 30, 389^04.
Wood, D., Reimherr, R.W., Wender, P. & Johnson, G.E. (1976). Diagnosis and
treatment of minimal brain dysfunction in adults. Archives of General
Woody, G.E., McLellan, A.T., Luborsky, L. & O'Brien, C.B. (1985). Sociopathy
and psychotherapy outcome. Archives of General Psychiatry, 42, 1081-6.
Woollcott, P. (1985). Prognostic indicators in the psychotherapy of borderline
patients. American Journal of Psychotherapy, 39, 17-29.
World Health Organization (1978). International Classification of Diseases, 9th
rev. (ICD-9). Geneva: World Health Organization.
World Health Organization (1992). The ICD-10 Classification of Mental and
Behavioural Disorders. Geneva: World Health Organization.
World Health Organization (1993). The ICD-10 Classification of Mental and
Behavioural Disorders. Diagnostic Criteria for Research. Geneva: World
Health Organization.
Wrobel, T. & Lachar, D. (1982). Validity of the Weiner subtle and obvious scales
for the MMPI: another example of the importance of inventory-item
content. Journal of Consulting and Clinical Psychology, 50, 469-70.
Yalom, I.D. (1966). Problems of neophyte group therapists. International Journal
of Social Psychiatry, 12, 52-7.
Yalom, I.D. (1975). The Theory and Practice of Group Psychotherapy. New York:
Basic Books.
Yalom, I.D., Bond, G., Block, S. et al. (1977). The impact of a weekend group
experience on individual therapy. Archives of General Psychiatry, 34,
399-403.
Yalom, I.D., Brown, S. & Bloch, S. (1975). The written summary as a group
psychotherapy technique. Archives of General Psychiatry, 32, 605-11.
Yalom, I.D., Houts, P.S., Newell, G. & Rand, K.H. (1967). Preparation of
patients for group therapy. A controlled study. Archives of General
Yates, W.R., Bowers, W.H., Carney, C.P. & Fulton, A.I. (1990). Is bulimia
nervosa related to alcohol abuse? A personality analysis. Annals of Clinical
Yates, W.R., Noyes, R., Petty, F. et al. (1987). Factors associated with motor
vehicle accidents among male alcoholics. Journal of Studies in Alcohol, 48,
586-90.
Yates, W.R., Petty, F. & Brown, K. (1988). Alcoholism in males with antisocial
personality disorder. International Journal of Addictions, 23, 999-1010.
Yeung, A.S., Lyons, M.J., Waternaux, C M . et al. (1993). Empirical
determination of thresholds for case identification: validation of the
Personality Diagnostic Questionnaire - Revised. Comprehensive Psychiatry,
34,384-91.
Young, M.A., Sheftner, W.A., Klerman, G.L. et al. (1986). The endogenous sub-
type of depression: a study of its internal construct validity. British Journal
References 385
Yudofsky, S.C., Silver, J.M. & Jackson, W. (1986). The Overt Aggression Scale
for the objective rating of verbal and physical aggression. American Journal
Zanarini, M., Frankenburg, F., Chauncey, D. & Gunderson, J.G. (1987). The
diagnostic interview for personality disorders: interrater and test-retest
reliability. Comprehensive Psychiatry, 28, 467-80.
Zanarini, M.C., Gunderson, J.G. & Frankenburg, F.R. (1990). Cognitive
features of borderline personality disorder. American Journal of Psychiatry,
141, 57-63.
Zanarini, M.C., Gunderson, J.G., Frankenburg, R.F. et al. (1991). The face
validity of the DSM-III and DSM-III-R criteria sets for borderline
Zanarini, M.C., Gunderson, J.G., Marino, M.F. et al. (1989). Childhood
experiences of borderlines. Comprehensive Psychiatry, 30, 18-25.
Zeitlyn, B.B. (1967). The therapeutic community - fact or fantasy. British Journal
Zilber, N., Schufman, N. & Lerner, Y. (1989). Mortality among psychiatric
patients - the groups at risk. Acta Psychiatrica Scandinavica, 79, 248-56.
Zimmerman, M. (1994). Diagnosing personality disorders. Archives of General
Psychiatry, 51, 2 2 5 ^ 5 .
Zimmerman, M. & Coryell, W. (1989). DSM-III personality disorder diagnoses
in a non patient sample. Archives of General Psychiatry, 46, 682-9.
Zimmerman, M. & Coryell, W. (1990). Diagnosing personality disorders in the
community. A comparison of self-report and interview measures. Archives of
Zimmerman, M., Coryell, W. & Pfohl, B. (1986). ECT response in depressed
patients with and without a DSM-III personality disorder. American Journal
Zimmerman, M., Pfohl, B., Coryell, W. et al. (1988). Diagnosing personality
disorder in depressed patients. Archives of General Psychiatry, 45, 733-7.
Zivich, J.M. (1981). Alcoholic subtypes and treatment effectiveness. Journal of
Consulting and Clinical Psychology, 49, 72-80.
Zoccolillo, M. & Rogers, K. (1991). Characteristics and outcome of hospitalized
adolescent girls with conduct disorders. Journal of American Academy of
Child and Adolescent Psychiatry, 30, 973-81.
Zuckerman, M. (1991) Basic dimensions of personality and consistency of
personality. In: Psychology of Personality, pp. 188. Cambridge, UK:
Cambridge University Press.
Zuroff, D. & Mongrain, M. (1987). Dependency and self-criticism: vulnerability
factors for depressive affective states. Journal of Abnormal Psychology, 96,
14^22.
Index
Page numbers for tables are in italics
ADHD see attention deficit hyperactivity and PDs 133-6

disorder and antisocial PD 135
Adler and Buie's model of psychotherapy onset, and genetic factors 134-5
263^ prognosis 194
adoption types of alcoholism 134
and antisocial PD 645 psychological management 291
comparison of PD assessments 32 alcohol, effects on foetus in utero, effects on
studies, criminality and PDs 35-6 PDs 29
Adult Personality Functioning Assessment alcohol problems and PDs 36-7
interview 213 alcohol-related syndromes and personality
affective instability and PDs, biological disorder, course and outcome 190-1
markers 301 alcoholism, Michigan Screening Test 218
affective spectrum disorder 24 alprazolam
aggression for borderline PD 235, 240
among prisoners 335 for schizotypal PD 240
causal factors 2845 amitriptyline
definition 283 for borderline PD 234, 236
modified overt scale 222 for schizotypal PD 234
physical restraint 288 amphetamine
psychological management 283-90 for ADHD 151, 240
calming the situation 286-7 effects on schizotypal and borderline PDs
general organizational policies 287-8 34
provoking further anger 287 anankastic personality disorder 98
reducing risk 287 anorexia nervosa
social skills training 289-90 and cluster C PDs 147-8
repeated, and antisocial PD 285 and PDs 145-8
treatment programmes 288-90 prognosis 194
see also offenders with personality anticonvulsant treatment for borderline and
disorders behavioural dyscontrol 239
Aggression Risk Profile 222 antidepressant treatment 235-8
aggressive behaviour, psychosurgery 245 antisocial behaviour
agoraphobia and personality disorders classification of patterns 58
140-1 co-occurrence with PDs 27
AIDS encephalopathy and organic PD 127 and excess mortality 131
AIDS and personality disorders 136 and PDs 151-2
alcohol abuse antisocial personality, definition 59
associated with spouse (partner) abuse antisocial personality disorder 56-65
152 and alcoholism 367, 135
and borderline PD, prognostic factors associated with criminality 37
165, 166 associated features 64-5
386
Index 387
biological markers, assessment 30-1 anxiolytics 242

childhood precursors 182 anxious personality disorder 87
and childhood sexual abuse 39 see also avoidant personality disorder
clinical features 62-3 Asperger's syndrome
clinical origins 56-60 relation to childhood PD 49, 50, 54
course and outcome, variables associated and schizoid PD 25
with 176-7 assessment of personality disorders
adverse childhood events 176 195-229
duration of follow-up period 176-7 classification 196-9
gender 177 clinical ratings 13, 220-2
loss events 176 'LEAD' standard 220-1
current status 65 consensus ratings (or diagnoses) 200
and drug abuse 136 core criteria 196-7
evidence of causation 35-40 future directions 228-9
gender prevalence 180, 181 information sources 199
lifetime prevalence 154, 155, 156 inter-rater reliability coefficients 200
lifetime rate 27-8 LEAD procedure 100, 202
longitudinal data, implications for methodological variables 202-4
management 177-8 response bias 202, 203
longitudinal studies 160, 170-8, 178 structured interviews or self-report
co-occurring psychiatric disorders questionnaires 202-3, 203-^, 204
175-6 non-methodological variables that
inpatient studies 173 influence 199
non-clinical and non-forensic overlap features 196
populations 173 overview 200-2, 222-4
overview 170-2 patient's information 200-1
prognostic factors 174-5 routine clinical assessment 222-8
psychiatric and offender populations of clusters A, B and C 224-8
173-4 self-report questionnaires 200, 217-20
related to epidemiology 172 severity of impairment 197
transient delinquents 173-4 structured interviews 200, 204, 204-13
main definitions 60-2 use of dimensional ratings 197-8
main features 56 assessments for legal proceedings 335-8
neglect, physical abuse, bad parenting 39 asthenic personality disorder 91-2
and neuropsychological dysfunction 149 see also dependent personality disorder
in parents, associated with psychiatric attention deficit hyperactivity disorder
disorders in childhood 1523 amphetamines 248
PCL-R 221-2 childhood, precursors of PDs 182
presentation to medical services 6 3 ^ and PDs 150-1, 184-6
routine assessment 225-6 psychostimulants 240
self-report questionnaires 217 attention seeking and hysteroid dysphoria
and somatization disorder 137, 139 124
and substance abuse disorders, autism, relation to childhood PD 49-50,
associations 25 54
subtypes 57-8 avoidant personality disorder 86-90
and suicide 131 and anxiety disorders 140
antisocial tendencies 6 associated features 90
anxiety disorders with personality disorders behavioural treatment 251-2
140-2 causation 40
boundaries between 24-5 clinical features 88-90
prognosis 193 clinical origins 86-7
anxiety management 270 current status 90
anxiety states, co-occurrence with PDs 26, group psychotherapy 296
27 longitudinal studies 179
anxiety/inhibition and PD, biological main definitions 878
markers 31 main features 86
anxiolytic treatment for borderline PD and presentation to medical services 90
behavioural dyscontrol 239-40 routine assessment 227-8
388 Index
basic concepts 3-42 longitudinal studies 160, 163-70, 178

behaviour adverse childhood events 16-8
consistency 8-9 adverse events in adulthood 168
effect of environment 8-9 co-occurring psychiatric disorders 167
influences 8 duration of follow-up period 168
behaviour therapy 272-4 gender 168
Dialectical Behaviour Therapy 272, 273, implications for management 168-70
274 inpatient studies 163-5, 168-9
behavioural treatment prognostic factors 165-6
overview 251, 252 main definitions 67-8
sex offenders 292 main features 65-6
beta-blockers 242 monoamine oxidase inhibitors 236
biological markers of PD mortality 189
assessment 30 multimodal treatments 318-20
evidence from studies 34-5, 37-9, 40 neglect, physical abuse, bad parenting 39
bipolar affective disorder, and borderline neuroleptic drugs 234-5
PD 165 and neuropsychological dysfunction 149
bipolar manic-depressive disorder 116, 117 and post-traumatic stress syndrome 150
ECT 244-5 presentation to medical services 72
Blancks' psychotherapy 264-5 prevalence 28, 154-5, 155, 156
borderline personalities 209-10, 211 psychodynamic psychotherapy 251, 252
and major depressive illness 37 psychotherapy, methods compared 265-7
borderline personality disorder 65-74 routine assessment 225-6
alprazolam 240 self-report questionnaires 217
antidepressant treatment 235-8 and suicide 131-2
and anxiety disorders 140 summary on treatments 246-7
associated with character-spectrum supportive psychotherapy 262-3
disorder 123-4 therapeutic community 318
associated features 72-3 transient psychotic experiences 70, 71-2
biological markers 38 treatment settings 280-1
assessment 30-1 Borderline Personality Disorder Scale 211
brief (psychodynamic) psychotherapy boundaries between personality disorders
260 and other mental disorders 22-6
buspirone 240 bowel disorders and personality disorders
carbamazepine 239 149
carbidopa 241 brain infections, effects on PDs 29
and childhood sexual abuse 39 brain tumours, and organic PD 127
clinical features 68-72 brief adaptational psychotherapy 261
clinical origins 66-7 brief (psychodynamic) psychotherapy
clonazepam 240 260-1
clozapine 241 Briquet's syndrome see somatization
co-occurrence with other PDs 27, 73 disorder
cognitive-perceptual distortions 70, 71 bulimia nervosa and personality disorders
couple/family sessions 279 145-8
current status 73-4 prognosis 193-4
departures from consensual reality 70,71 buspirone, for borderline PD 240
and depressive disorders 129-30
dexamphetamine 241 California Personality Inventory 218
Dialectical Behaviour Therapy 280 California Psychological Inventory
drug treatment 235, 239, 240, 241, 242-3 217
ECT 245 cannabis abuse and personality disorders
evidence of causation 35-40 136
gender prevalence 72 carbamazepine
group psychotherapy 316-20 for borderline PD 235, 239
group treatment 295-6 with behavioural dyscontrol 242
intensive psychoanalytic psychotherapy for dysphoria 239
256-7 for PDs 38
Ievodopa241 carbidopa in borderline PD 241
Index 389
catecholamine metabolite studies and physical abuse 187-8
borderline PD 38 schizoid features 186
Cattell's 16 Personality Factor Scale 217 sexual abuse 187-8
causation 28-9 violence 187-8
animal studies 30 chlorophenylpiperazine, future research 248
assessmnent of biological markers of PDs chronic pain, complaints, and personality
30-1 disorders 148-9
'cluster A' PDs 33-5 chronic secondary dysphoria 118
biological markers, evidence from categories 118, 119
studies 34-5 classification 196-9
evidence from genetically related influential 16-17
individuals 33 methods 15-22
'cluster B' PDs 35^0 categories 15-16
evidence from associations within dimensions 16
groups 39-40 principles 13-15
evidence from studies of biological Cleckley's concept of antisocial personality
markers 37-9 59-60
evidence from studies of genetically clinical care, good 274-9
related individuals 35-7 aims 278-9
'cluster C PDs 40 cognitive and emotional components
evidence from studies of biological 276-7
markers 40 instillation of hope 278
evidence from studies of related practical strategies 275-9
individuals 40 psychological defences 277
comparison of PD assessments in supportive psychotherapy 276
genetically related individuals 31-3 clonazepam for borderline PD 240
constitution-environment interactions 30 clozapine for borderline PD 241
genetic factors 28 'cluster A' personality disorders
genotype 29 causation 33-5
genotype-environment interactions 30 longitudinal studies 178-9
identification of associations between 'cluster B' personality disorders
possible causal variables and PD causation 35-40
features in groups of subjects 33 longitudinal studies 178
methods of investigation 30-3 'cluster C personality disorders
methods to evaluate interactions between causation 40
causal variables 33 longitudinal studies 179
non-genetic physical factors 29 clusters of symptoms in borderline
range of variables 28-9 PD68
socio-cultural factors 28 co-morbidity see co-occurrence
twin studies 31-3 co-occurrence of personality disorders
character-spectrum disorder 114, 118, and assessment 228
123-5 in borderline PD 73
associated with borderline PD 123-4 and eating disorders 146
characteriological depression 24 with histrionic PD 79
child abuse 152-3, 166, 167, 187-8 in obsessive-compulsive disorder 100
effect of family therapy 279-80 with other mental disorders 26-7
childhood precursors of personality other psychiatric disorders with
disorders antisocial PD 175-6
ADHD and PD 184-6 other psychiatric disorders with
adverse childhood environment 187-8 borderline PD 68
'at risk' for adult PD 188-9 sadistic PD 112
conduct disorder and PD 183-4 self-defeating PD 109
criminal behaviour 187-8 of two or more PDs 128
longitudinal studies 182-9 cocaine abuse and personality disorders
early predictors of PD 182-3 136
implications for management and course and outcome 191
prevention of PD 188-9 cognitive-behavioural therapy, and
overview 182 cognitive psychotherapy 270
390 Index
cognitive-perceptual distortions paranoid PD 45

borderline PD 70, 71 of personality 7-8
in schizoid/schizotypal PD 55 of personality disorders 10-11
cognitive psychotherapy 267-72 delinquency and antisocial behaviour 58
cognitive analytic therapy 271-2 departures from consensual reality,
and cognitive-behavioural therapy 270 borderline PD 70, 71
non-collaboration 270-1 dependent personality disorder 91-6
overview 267-8 associated features 95
practical strategies 268-71 attachment 9 3 ^
first stage 268-9 causation 40
second stage 269-70 clinical features 92-4
cognitive therapy, overview 251 clinical origins 91-2
cognitive/perceptual organization and PD, current status 96
biological markers 30 general dependency 9 3 ^
community-based treatment programmes, longitudinal studies 179
sexual offenders 292-3, 329 main definitions 92
see also sex offenders main features 91
community supervision of released overlap with other syndromes 93
prisoners with PDs 330-2 presentation to medical services 94-5
see also offenders routine assessment 227-8
concepts of personality 7-8 depression and depressive disorders
conduct disorder with borderline PD, treatment 235-8
and antisocial PD 64, 65, 171, 173 ECT 244^5
and PDs 182, 183-4 and PDs 129-30
congenital deficiency of the moral sense 57 boundaries between 23-4
consistency of behaviour 8-9 course and outcome 191-3
constitutional psychopathic inferiority 57 depressive neurosis 115
conversion disorder 74, 137 depressive personality (or personality
see also histrionic (or hysterical) disorder) 24, 114^26
personality disorder: somatization abnormal brain function 115
disorder anxious or atypical depression 122
core criteria 196-7 associated features 125
countertransference, effects 259 associated with other PDs 125
couple treatment 279-80 categories of relationship between 121,
'crack' abuse and personality disorders 136 121-5
Criminal Profile Scale 222 character spectrum disorders 114, 118,
criminality 123-5
and ADHD 150-1 chronic secondary dysphoria 118
and antisocial behaviour 58 classification 115-20
and antisocial PD clinical features 121-5
age factor 174 clinical origins 114-20
longitudinal studies 170, 171, 1 7 1 ^ current status 125-6
associated with PD 37 cyclothymia 119-20 dysthymias 114,
childhood precursors, and PD 187-8 118,119, 122
and PDs 151 classification 122-3
adoption studies 35-6 endogenous 115-17
twin studies 35 environmental situation 116
and sadistic PD 111 and general neurotic syndrome 123
see also offenders with personality history of mania 119
disorders main definitions 120-1
cyclothymia 119-20 main features 114
cyclothymic PD 24 neurotic depression 115, 116-17
obsessive compulsive traits 122
dangerous behaviour and personality personality disorder 116
disorders 152 presentation to medical services 125
defence mechanisms 7-8 reactive 117
definitions residual major depression 121
antisocial PD 60-2 subaffective dysthymia 114, 118, 119
Index 391
unipolar 119 classification of PDs 16-17
subtypes 119 polythetic categories 17-20
depressive syndrome dependent PD 91
associated with dependent PD 94 histrionic PD 75
co-occurrence with PDs 26, 27 narcissistic PD 80, 81-2
dexamethasone suppression test for obsessive-compulsive PD 97
depressive PD 122 passive-aggressive PD 102
dexamphetamine schizoid PD 48
for borderline PD 241 DSM-III-R 10-11
and mood changes 248 antisocial personality 59, 60, 60-1
diagnoses of personality disorders 10-11 avoidant PD 87-8, 88-9, 89-90
and disease concept 11-13 borderline PD 67-8
Diagnostic Interview for Borderline classification of PDs 16-17
Patients 67 disadvantages 201
Diagnostic Interview for Borderlines (DIB) polythetic categories 17-20
163, 209-10 disadvantages 20-1
Diagnostic Interview for Narcissism 212 potential modification 21-2
Diagnostic Interview for Personality conversion disorder 137
Disorders (DIPD) 207 dependent PD 91, 92, 93, 94
Dialectical Behaviour Therapy 252, 272, depressive disorders 118, 119
273,274, 280 depressive personality 114
differential diagnosis of personality histrionic PD 75, 76, 78-9
disorders 228 narcissistic PD 80-1, 86
disease concept and personality disorders obsessive-compulsive PD 97, 99
11-13 paranoid personality disorder 45
dissocial personality disorder, diagnosis 61 passive-aggressive PD 102
dissociative disorders 138 personality disorder not otherwise
co-occurrence with PDs 26 specified 128-9
see also somatization disorder sadistic PD 110-11
drop-out rate, high, for treatment 253 schizoid/schizotypal PD 50-1
drug abuse and personality disorders 133, schizophrenia 143
136 self-defeating PD 104, 105, 106
borderline PD, prognostic factors 166-7 DSM-IV 10-11
course and outcome 191 antisocial personality 60-1
and HIV 191 avoidant PD 88
intravenous, and HIV transmission 136 borderline PD 68
drug treatment 233-44 classification of PDs 16-17
combinations 244 disadvantages 201
overview 233-4 polythetic categories 17-20
specific drugs 234-42 disadvantages 20-1
see also names of drugs potential modification 21-2
specific PDs 242-3 dependent PD 92, 94
target features 243-4 histrionic PD 76-7, 79
DSM-I obsessive-compulsive PD 97
compulsive personality 96 passive-aggressive PD 102
dependent PD 91 schizoid/schizotypal PD 50
histrionic PD 75 dynamic organisation of personality 7-8
passive-aggressive personality disorder dyscontrol syndrome associated with PD
101-2 37
DSM-II dysmorphophobia 47
antisocial personality 59 dysphoria
histrionic PD 75 hysteroid, phenelzine 236
obsessive-compulsive disorder 96-7 MAOI 244
passive-aggressive PD 102 dysthymia(s)114, 118, 119
DSM-III 10-11 classification 119-20
antisocial personality 59 current status 125-6
avoidant PD 88-9 and personality disorders 130
borderline PD 67 presentation to medical services 125
392 Index
Eating Attitudes Test 148 and borderline PD 168
eating disorders, co-occurrence with PDs and 'normal' personality 180
26, 27, 145-8 and PDs 154
prognosis 193 longitudinal perspective 180-1
structured interview 147-8 gender-related variables and personality
subtypes of eating disorders 145-7 disorder, longitudinal perspective
EEG findings 37, 38, 39 181
electro-convulsive therapy 244-5 general neurotic syndrome 24
emotionality 6 relationship to depressive personality (or
during group psychotherapy 308-10 personality disorder) 123
enduring personality change after genetic factors and PDs 28-9
psychiatric illness 127 genetically related individuals
main definition and clinical features 127 causation of PDs 33, 35-7
main features and clinical origins 127 comparison of PD assessments 31-3
environment good clinical care 274-9
childhood, and PDs 187 aims 278-9
effect on PDs and personality 28-9, cognitive and emotional components
29-30 276-7
epidemiology of personality disorders instillation of hope 278
153-8 practical strategies 275-9
antisocial PD 172 psychological defences 277
data for longitudinal studies 162 supportive psychotherapy 276
gender and PD 154 Gough's California Psychological inventory
methodological problems 1 5 3 ^ 218
overview 27-8 group psychotherapy(ies) and personality
prevalence rates of individual PDs 154-6 disorders 294-320
epilepsy and criminality associated with PD basic rules 300-1
37 borderline PD 316-20
Eysenck Personality Inventory 216 multimodal treatments 318
Eysenck Personality Questionnaire 216 overview 316
positive ingredients 316-18
factitious illness 138 chain phenomenon 299
family doctor visits and personality 'closed' or 'open' 299
disorders 148 coping with anger 308-9
family studies of antisocial PD 64-5 coping with suicidal ideas 309
family treatment 279-80 coping with tearfulness and/or depression
fenfluramine 309
future research 248 definitions and therapeutic ingredients
response, PDs 38 296-8
fire-setting in childhood and borderline PD, emotionality 308-10
prognostic factors 166 expressed emotion 299
fluoxetine goals 299-300, 302-3
borderline PD 237 group cohesiveness 297-8
'cluster B' PDs 40 group conductor's problems 312-13
for depressive symptoms 237-8 group conductor's role 303-8
personality disorders 242 conductor's own emotion 309-10
flupenthixol treatment 234 constant review of own activity 305-6
forensic samples with personality disorders degree of staff activity 311
158, 160 emotionally disturbed members 307-8
fragile-X chromosome associated with the final 15 minutes 307-8
schizotypal PD 34 interaction between group members
304
gambling non-verbal activity 304
pathological and personality disorders resistance from the group 306
151 overcoming 306-7
South Oaks Screen 218 to do no harm 303
gender unstructured meetings 305
and antisocial PD 177 verbal contribution 303-4
Index 393
group conductors' training and gender prevalence 75-6
supervision 315 longitudinal studies 160, 178
group-specific phenomena 298-9 main definitions 76
hostility 298-9 main features 74
indirect benefits 298 overlap with borderline PD 79-80
leaving the group 311 presentation to medical services 79
meeting lengths and frequency 301 prevalence 155, 155, 156
negative effects 313-15 routine assessment 225-6
emotional arousal 314 HIV and personality disorders 136
existing problems magnified 314 and drug abuse
and selection of patients 314 course and outcome 191
symptom substitution 314 transmission 136
transient/lasting 313-14 Home Office officials (C3 Division)
new learning situations 297 concerned with restricted prisoners
number of patients 301 332
organization 299-302 homosexuality and personality disorder,
overview 294-6 longitudinal perspective 181
historical development 294-5 homovanillic acid, raised in CSF and
patterns of group interaction 310-11 plasma, in PDs 34
psychoanalytic theory 297 Hostility Inventory 217
repeated questions 313 Huntington's disease, and organic PD 127
reports 301 5-hydroxyindoleacetic acid and PDs 38
resistance 298-9 hyperactive syndrome and personality
staff members 301 disorders 150
staff members' own group activities hypochondriasis, co-occurrence with PDs
319-20 26
structured groups 301 hysterical disorders and somatization
suitable patients 299-300, 314-15 disorder 139
supportive approach 297 hysterical neurosis
treatment goals 295 conversion type 137
unstructured groups 300, 301 see also somatization disorder
unsuitable patients 299-300, 31415 dissociative type 138
varieties 315-16 hysteroid dysphoria 124-5
group treatment 279-80 phenelzine 236
Gunderson's Diagnostic Interview for
Borderlines (DIB) 67, 163, 209-10 ICD anankastic personality disorder 98
see also obsessive-compulsive personality
habit, definition 6 disorder
hallucinations, ECT 244-5 ICD-9 16-17
haloperidol dependent PD 91
for borderline PD 235, 236 diagnosis of personality disorder 10-11
for borderline or schizotypal PD 234 polythetic categories 1720
Hare Psychopathy Checklist 221-2 ICD-10 16-17
health care facilities, increased use, and antisocial personality 61
personality disorders 148 anxious (avoidant) PD 87
Henderson Hospital, Sutton, Surrey 333 avoidant PD 88
heroin addiction, treatment 291 borderline PD 68
historical developments 9-10 dependent PD 92
histrionic (or hysterical) personality depressive disorders 118, 119
disorder 74-80 depressive personality 114
associated features 79-80 enduring personality change after
associated with somatization disorder psychiatric illness 127
138-9 histrionic PD 76-7
clinical features 77-9 hyperkinetic disorders 184
clinical origins 74-6 see also attention deficit hyperactivity
co-occurrence with PDs 27 disorder
current status 80 mixed and other personality disorders
evidence of causation 35-40 128-9
394 Index
\CD-\ti(cont.) methodological aspects 162-3

obsessive-compulsive disorder 96 outcome measures 162
organic PD 126-7 perspective 159-60
paranoid personality disorder 45 longitudinal expert evaluation that uses all
polythetic categories 17-20 data (LEAD) standard 100, 202,
disadvantages 20-1 220-1
potential modifications 21-2 low birth weight, effects on PDs 29
schizoid/schizotypal PD 51
schizophrenia 143 ra-chlorophenylpiperazine, future research
idiopathic pain disorder and personality 248
disorders 148-9 malnutrition, effects on PDs 29
imipramine manie sans delire 57
for ADHD 237 manic disorders, co-occurrence with PDs 26
for borderline PD 236 Mask of Sanity 57
impulse-control disorders 25 masochistic personality disorder 104, 105
impulsivity see also self-defeating personality
/aggression, and PD, biological markers disorder
30 Masterson's confrontative psychotherapy
and borderline PD 65 259
and PDs 38 Menninger Foundation 257
information sources for personality mental disorders other than PDs
disorder assessment 199 co-occurrence with PDs 26-7
intensive psychoanalytic psychotherapy and PDs, boundaries between 22-6
256-7 Mental Health Act 1983 336
Interpersonal Dependency Inventory 218 Mental Health Act Review Tribunals 336
Interpersonal Sensitivity Measure 218 methylphenidate
interventions for personality disorders, in ADHD 151, 240
effects, general trends shown by and PDs 38
longitudinal studies 179-80 Michigan Alcoholism Screening Test 218
Millon Clinical Multiaxial Inventory 215
jealousy and paranoid PD 47 minimal brain dysfunction and personality
disorders 150
Karolinska Psychodynamic Profile 209 Minnesota Multiphasic Personality
Karolinska Scales of Personality 217 Inventory 146, 213-14,217
Kernberg's expressive psychotherapy 257-9 Modified Overt Aggression Scale 222
Kernberg's Structured Interview 210 monoamine oxidase
Kohut's psychotherapy 263 platelet, and borderline PD 38
reduced platelet in PDs 34, 37
Lazare-Klerman Trait Scale 217-18 monoamine oxidase inhibitors
LEAD (longitudinal expert evaluation that for borderline PD 242
uses all data) standard 100, 202, for borderline PD and depressive
220-1 symptoms 236
legal proceedings, assessments and reports for hysteroid dysphoria 125
335-8 mood disorders, ECT 244-5
levodopa moral insanity 57
for ADHD 240 mortality and personality disorders,
in borderline PD 241 longitudinal studies 189
life-sentenced prisoners, psychiatric support Multiphasic Narcissism Inventory 217
on release 330-1 multiple personality disorder 138
lithium carbonate 242 see also somatization disorder
for PDs 38 multiple sclerosis, and organic PD 127
for violent behaviour 238-9 Munchausen's syndrome 138
litigants, paranoid 47
Locus of Control of Behaviour 218
longitudinal aspects of personality diagnostic interview 212
disorders 159-94 self-report questionnaires 217
associated variables 161 narcissistic personality disorder 80-6
epidemiological data 162 associated features 86
Index 395
clinical features 82-5 antisocial PD, effectiveness of treatment
clinical origins 802 326
co-occurrence with PDs 27 assessments and reports for legal
current status 86 proceedings 335-8
evidence of causation 35-40 behaviour therapy 328-30
intensive psychoanalytic psychotherapy Care Programme Approach 326-7
256-7 checklist for assessment of future violent
longitudinal studies 178 behaviour 337-8
main definitions 82 clinical assessment for treatment 3 2 3 ^
main features 80 identification of psychiatric and social
presentation to medical services 85 difficulties 323-4
prevalence 155, 156 psychiatric follow-up on release 323
routine assessment 225-6 psychiatric report and re-offending
Narcissistic Personality Inventory 217 risks 324
Narcissistic Trait Scale 217 psychopathic disorder /mental illness
nature of personality disorders 195-6 323
neuroleptic drugs 234-5 recommendation for hospital orders
for paranoid PD 243 323
for schizotypal PD 243 uses of assessment 323
neuropsychological impairment and cognitive psychotherapy 328-30
personality disorders 149 community and outpatient settings
neuropsychological testing of borderline 325-7, 330-2
PD patients 38-9 crisis admissions 332-3
neurotic character 7 estimation of crimes resulting in caution
neurotoxins, effects, and organic PD 127 or conviction 321-2
not otherwise specified (NOS) personality frequency of PDs 321-2
disorders 128-9, 130 Home Office officials (C3 Division) 332
associated features 129, 130 identification of organic disorders 326
clinical features 128-9 inpatient settings 332-5
main features and definitions 128 inpatient therapeutic community
treatment 333
obsessive-compulsive disorder life-sentenced, psychiatric supervision on
association with obsessive-compulsive release 330-1
neurosis 25-6 long-term relationships with staff 327
causation 40 longitudinal studies 173-4
co-occurrence with PDs 26, 144-5 open-ended commitment by staff 327
prognosis 193 parole release, psychiatric supervision
relation to obsessive-compulsive PD 330
99-100 prevalence rates for antisocial PD 170,
obsessive-compulsive personality disorder 172, 174-5
96-101 prisons 334-5
associated features 99-101 providing supportive relationship 327
clinical features 989 psychodynamic psychotherapy 327-8
clinical origins 96-7 psychopathic disorder
co-occurrence with other PDs 100 assessment of risk 337
current status 101 reports for discharge 337
gender prevalence 101 psychopathic PD, effectiveness of
longitudinal studies 179 treatment 326
main definitions 97-8 range of interventions 324-5
main features 96 recidivism 292-3
presentation to medical services 99 relapse prevention 330
relation to obsessive-compulsive disorder released, community supervision 330-2
99-100 restricted 332
routine assessment 227-8 secure units and UK 'special' hospitals
obsessive-compulsive traits and depression 333^
122 serious re-offending 332
offenders with personality disorders, sex offenders see sex offenders
management 180, 321-39 social management 331-2, 338-9
396 Index
offenders with personality disorders (cont.) Perceptual Aberration Scale 218
statutory supervision in the community personality 5-9
330 adult, functioning assessment interview
treatability 336 213
variation of psychiatric treatment 334-5 approaches 9
violence among prisoners 335 consistency 8-9
violent definitions and concepts 7-8
checklist 337-8 response/habit/trait/temperament 5-7
four types 285 self-report questionnaires 213-20
management 283-90 standard assessment 208-9
see also criminality Personality Assessment Schedule 208
organic personality disorder 126-7 personality change, enduring, after
main definitions and clinical features psychiatric illness 127
126-7 main definition and clinical features 127
main features and clinical origins 126 main features and clinical origins 127
Personality Diagnostic Questionnaire
pain, chronic, complaints, and personality 215-16,218-19
disorders 148-9 Personality Disorder Examination 206
panic disorders and personality disorders personality disorder not otherwise specified
140-1 (DSM-III-R) 128-9
paranoid litigants 47 associated features 129
paranoid personality disorder 44-7 clinical features 128-9
associated features 47 main features and definitions 128
clinical features 45-6 or mixed and other personality disorders
clinical origins 44-5 (ICD-10) 128-9
current status 47 personality disorder syndromes, validity
diagnostic uncertainty 25 43-4
evidence of causation 33-5 personality disorders
longitudinal studies 178-9 basic concepts 3-42
main definitions 45 see also basic concepts: and headings
main features 44 throughout
neuroleptics 243 diagnostic interview 207
presentation to medical services 46-7 importance 4-5
prevalence 155 mishandled in medical practice 5
routine assessment 224 nature 195-6
underlying secondary symptoms 46-7 Personality Interview Questionnaire 208
paranoid states, co-occurrence with PDs 26 phenelzine
parental loss and borderline PD 167 for borderline PD 235, 236
parenting, inadequate and abusive, and for hysteroid dysphoria 236
personality disorders 152-3, 166, 167 physical abuse in childhood, and PDs 187
Parkinson's disease, and organic PD 127 physical factors and PDs 29
parole, prisoners released on, psychiatric physical treatments
supervision 330 ECT 244-5
passive-aggressive personality disorder psychosurgery 245
101-4 summary 246-7
associated features 104 post-traumatic stress disorder
causation 40 in borderline PD 167
clinical features 102-3 and PDs 149-50
clinical origins 101-2 posteromedial hypothalamotomy for
current status 104 restless, aggressive and destructive
longitudinal studies 179 behaviour 245
main definitions 102 predictors of personality disorders, early
main features 101 182-3
presentation to medical services 103 prevalence of personality disorders 27,
prevalence 156 154-6
routine assessment 227-8 primary medical care settings and
PDE 206 personality disorder 157
pemoline, for ADHD 240 prisoners see offenders
Index 397
prisons 334-5 psychopathic disorder
suitability for offenders with PDs 334 definition 58-9, 60
propranolol reports for discharge of offenders 337
for behavioural dyscontrol 241 psychopathology associated with PD 58
in brain-damaged patients 241 Psychopathy Checklist 59
pseudo-psychopathic syndrome 127 PCL-R 59-60, 61,62
psychiatric assessment and antisocial PD psychopathy, features 57
173^ psychostimulant treatment 240-1, 242
psychiatric disorders psychosurgery for personality disorders 245
basic concepts 3-4 Psychotherapy Research Project of the
see also basic concepts: and headings Menninger Foundation 257
throughout psychotherapy(ies)
co-occurrence with antisocial PD 175-6 Adler and Buie's model 263-4
with co-occurring PD, implications for Blancks' 264-5
management 194 borderline PD, methods compared 265-7
effects of personality disorders on course brief adaptational 261
and outcome 189-94 cognitive, see cognitive psychotherapy
overview 189-90 definition 249-53
enduring personality change after 127 group 294-320
main definition and clinical features see also group psychotherapies
127 Kohut's 263
main features and clinical origins 127 for PDs, general trends shown by
see also specific psychiatric disorders longitudinal studies 179-80
psychiatric inpatient and outpatient psychodynamic, see brief
samples and personality disorder (psychodynamic) psychotherapy,
157-8 psychodynamic psychotherapy
psychoanalytic psychotherapy, intensive short-term dynamic 261
256-7 supportive 261-5
psychodynamic profile 209 psychotic-like phenomena and borderline
psychodynamic psychotherapy PD 69-72
conclusion 260 psychotropic drugs, future research 247
definition 251-3
intensive psychoanalytic psychotherapy rapists, group psychotherapy 296
256-7 recidivism 292-3
Kernberg's expressive psychotherapy related individuals
257-9 and antisocial PD 64-5
Masterson's confrontative psychotherapy causation 40
259 comparison of PD assessments 31-3,
for offenders with PDs 327-8 35-7
overview 254-6 reports for legal proceedings 335-8
psychological management 249-93 response in specified situation, definition 6
behaviour therapy 272-4 Rorschach inkblot test 201
brief (psychodynamic) psychotherapy
260-1 sadistic personality disorder 109-14
couple treatment 279-80 associated features 113
definitions and overviews 249-54 clinical features 111-13
family treatment 279-80 clinical origins 110-11
good clinical care 274-9 co-occurrence with other PDs 112
overview 274-5 current status 113-14
practical strategies 275-9 distinction from antisocial PD 113-14
group treament 279-80 main definition 111
management of specific problem- main features 109
behaviours related to PDs 281-93 presentation to medical services 113
psychodynamic psychotherapy 254-60 prevalence 112-13
supportive and related psychotherapies Schedule for Borderline Personalities 211
261-5 Schedule for Normal and Abnormal
overview 261-3 Personality 216
treatment settings 280-1 Schedule for Schizotypal Personalities 211
398 Index
schizoid disorder, childhood, precursors of main definitions 50-1
PDs 182 main features 47-8
schizoid features in childhood and PD 186 neuroleptic drugs 234-5
schizoid personality disorder 47-56 and other PDs, boundaries between 23
associated features 54-5 presentation to medical services 53
in children 49 prevalence 49, 155
clinical features 51-3 relation to schizophrenia 25, 49
clinical origins 48-50 routine assessment 224
co-occurrence with other PDs 73 and schizophrenia 143
current status 556 self-report questionnaires 218
diagnostic uncertainty 25 summary of physical treatments 246-7
evidence of causation 33-5 Schizotypal Traits Questionnaire 217
longitudinal studies 178-9 schizotypy, structured interview for 211
main definitions 50-1 SCID-II 205-6
main features 47-8 secure units for offenders with PDs 333
neuroleptic drugs 234-5 selective serotonin reuptake inhibitors
presentation to medical services 53 for borderline PD 237-8
prevalence 49, 155 for PDs 242
relation to schizophrenia 49 self-defeating personality disorder 104-9
routine assessment 224 associated with depressive disorders
and schizophrenia 143 108-9
schizophrenia associated features 108-9
associated with paranoid PD 46 clinical features 106-8
borderline, symptom schedule for clinical origins 104-6
diagnosis 211 co-occurrence with other PDs 109
and borderline PD 165 concept, justification 107-8
co-occurrence with PDs 26, 1424 current status 109
boundaries between 23 distinction from self-defeating behaviour
diagnostic uncertainty 25 107
prognosis 193 patient studies 105-6
subtypes of schizophrenia 142-3 main definitions 106
neurodevelopmental (congenital) main features 104
143 presentation to medical services 108
ECT 244-5 self-harm populations, PD 158
relation to borderline PD 66 self-harm (self-mutilation)
relation to childhood PD 49 54, 55 behaviour therapy 273
in relatives 54 and borderline PD 66
schizotypal disorder, childhood, precursors co-occurrence with PDs 27, 132-3
of PDs 182 psychological management 281-3
schizotypal personalities, schedule for 211 Self-Report Psychopathy Scale 217
schizotypal personality disorder 47-56 self-report questionnaires 213-20, 214
alprazolam 240 evaluation and comparisons 217-20
and anxiety disorders 140 sentence completion tasks 201
associated features 54-5 seriously irresponsible behaviour and
biological markers 34 personality disorders 152
assessment 31 serotonergic metabolites and PDs 38
in children 49 sex offenders
clinical features 51-3 associated psychiatric difficulties 330
clinical origins 48-50 cognitive and behavioural treatment 329
co-occurrence with OCD 145 community-based treatment programmes
co-occurrence with other PDs 73 329
CT scans 34 drugs for 241-2
current status 55-6 posteromedial hypothalamotomy 245
diagnostic uncertainty 25 psychological management 291-3
drug treatment 243 basic education 292
evidence of causation 33-5 behavioural treatment 292
and fragile-X 34 four main types 292
longitudinal studies 160, 178-9 institutional programmes 292-3
Index 399
recidivism 292-3 supportive psychotherapy 261-5
social skills training 292 Symptom Schedule for the Diagnosis of
treatment intervention 325 Borderline Schizophrenia 211
victimization 335
sexual abuse temperament
and borderline PD, prognostic factors consistency 8
166, 167 definition 6-7
in childhood, and PDs 166, 167, 187, genetic contribution 7
187-8 temporal lobe epilepsy
sexual deviation disorders and personality anticonvulsant treatment 239
disorders 153 and organic PD 127
sexual disorders, co-occurrence with PDs 26 testosterone reduction for sexual offences
short-term dynamic psychotherapy 261 241-2
SIDP-R 206-7 Thematic Apperception Test 201
sociability 6 therapeutic community 318
socio-cultural environment and PDs 29 thiothixene for borderline or schizotypal
socio-cultural factors and PDs 28, 29 PD234
sociopathic personality disturbance 59 traits
somatization, chronic consistency 8
PD as causal factor 290 definition 6
psychological management 290-1 transference interpretations, importance
strategies 290-1 258
somatization disorder 74 transference reactions in patients with
co-occurrence with PDs 26 borderline PD 256-7
and PDs 137-9 transient delinquents and antisocial PD 173
hysterical neurosis transient psychotic experiences, borderline
conversion type 137 PD 70, 71-2
dissociative type 138 tranylcypromine
relation to conversion disorder 137 for ADHD in borderline PD 240
see also histrionic (or hysterical) for borderline PD 235, 236
personality disorder traumatic cerebral damage, effects on PDs
South Oaks Gambling Screen 218 29
special hospitals for offenders with PDs 333 trazodone 244
spouse (or partner) abuse, and personality treatability of PDs 40-2
disorders 151-2 tricyclic antidepressants 242
Standard Assessment of Personality 208-9 tricyclic treatment for borderline PD
Structured Clinical Interview for DSM-III- 236-7
R personality disorder 205-6 Tridimensional Personality Questionnaires
Structured Interview for DSM-III-R 216
Personality Disorders 206-7 trifluoperazine
Structured Interview for Schizotypy 211 for borderline PD 235
structured interviews 200, 204, 204-13 for schizophrenia 243
sub-affective dysthymia 114, 118 twin studies
substance abuse causation of antisocial, borderline,
group psychotherapy 296 histrionic and narcissistic PDs 35-7
with PD, course and outcome 190-1 causation of paranoid, schizoid and
psychological management 291 schizotypal PDs 33-5
substance use (abuse) disorders genetic and environmental influences on
and antisocial PD 65 PDs 31-3
co-occurrence with PDs 26, 27, 133-6
suicidal behaviour and borderline PD 66 UK Mental Health Act, 1983 323, 330, 331,
suicide 333
and PDs 130-2, 163, 164, 166-7 unipolar depressive disorder, and
longitudinal studies 189 borderline PD 165
preoccupation, use of fluoxetine 238
psychological management 281-3 validity of personality disorder syndromes
Suicide and Aggression Survey 212 43-4
supportive approach to PD patients 41 variables in longitudinal studies of PDs 161
400 Index
violence provoking further anger 287
among prisoners 335 psychological management 283-90
and antisocial PD, longitudinal studies calming the situation 286-7
170, 174 reducing risk 287
in childhood, and PDs 187 social skills training 289-90
definition 283 treatment programmes 288-90
effect of family therapy 279-80 see also offenders with personality
four types of offender 285 disorders
general organizational policies 287-8
lithium carbonate 238-9 Wisconsin Personality Inventory 216
of offender, report checklist 337-8 World Health Organization's definition of
see also offenders PD
and PDs 152 ICD-9 10
physical restraint 288 ICD-10 11

Personality Disorders PDF

Uploaded by

Copyright:

Available Formats

Personality Disorders PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Personality Disorders PDF

Uploaded by

Copyright:

Available Formats

Although the long-term patterns of problem-behaviour that form

personality disorders are part of the major classifications of disease, they

J O N A T H A N H. D O W S O N , MA, MB, BChir, MD (Cantab); DPM, PhD

A D R I A N T . G R O U N D S , BMedSci, BM, BS, DM (Nottingham); MRCPsych

Cambridge University Press 1995

First published 1995

A catalogue recordfor this book is available from the British Library

Library of Congress cataloguing in publication data

ISBN 0 521 45049 7 hardback

Transferred to digital printing 2004

ATG: To Professor John Gunn CBE. Teacher and tireless

Part one: Recognition page 1

Studies related to gender and sexual orientation 180

1 Psychological management 249

receive psychiatric and other medical services. Unfortunately, there is no

Why are personality disorders important?

Response I habit I traitI temperament

evidence, while similar and related behaviours may be given several

Definitions and concepts

observable methods by which an individual may keep certain thoughts at

have been shown to predict adult non-competitiveness and certain charac-

Subsequently, Maudsley (1868) described a similar but narrower cate-

a severe disturbance in the characterological constitution and behavioural tenden-

In addition, the ICD-10 includes the diagnosis of'enduring personality

The disease concept and PDs

applied to the people (patients) themselves. Any classification involves the

examples of a trait, i.e. a group of related behaviours, while other criteria

number of criteria met. Also, certain criteria can be designated as essential

Influential classifications of PDs

Categories ofPD in the main international classifications

Table 1.1 Categories of personality disorder common to the DSM-III-R,

DSM-III-R and DSM-IV category Corresponding ICD-10 category

features'. (Also, the PD categories are not assumed to be mutually

Table 1.2 Various categories of personality disorder

Note: * and DSM-IV.

PD syndromes are derived from DSM-III-R but do not have equivalent

Table 1.3 Possible modifications to a poly the tic classification of personality

1. Changes in the use of existing criteria

Disadvantages ofpolythetic PD categories

Potential modifications to PD categories

incorrect diagnosis of antisocial PD being made. But, although this will

Boundaries between personality disorders and other mental disorders

Co-occurrence of personality disorders with other mental disorders

The range of causal variables

Non-genetic physical factors

Assessment of biological markers ofPD

The comparison of PD assessments in genetically related individuals

Methods to evaluate interactions between causal variables

Identification of associations between possible causal variables and PD

'Cluster A' PDs (paranoid, schizoid and schizotypal PDs) - evidence of

Evidence from studies of genetically related individuals

Evidence from studies of biological markers

'Cluster Bf PDs (antisocial, borderline, histrionic and narcissistic PDs) -

Evidence from studies of genetically related individuals

adoptees in respect of criminality in biological and adoptive parents. The

of borderline personality and major depressive illness'. But there is evidence

Evidence from studies of biological markers

are unclear (Kumar et al., 1989). In respect of facial appearance, the

Evidence from associations within groups of subjects with PD