Attachment Styles and Personality Disord PDF

Attachment Styles and Personality
Disorders: Their Connections to Each Other

and to Parental Divorce, Parental Death,
and Perceptions of Parental Caregiving
Kelly A. Brennan
State University of New York, Stony Brook
Phillip R. Shaver
University of California, Davis
ABSTRACT Attachment theory was explored as a means of understanding

the origins of personality disorders. We investigated whether adult attachment
styles and personality disorders share a common underlying structure, and how
both kinds of variables relate to family background factors, including parental
death, parental divorce, and current representations of childhood relationships
with parents. A nonclinical group of 1407 individuals, mostly adolescents and
young adults, were surveyed about their attachment styles, parental marital
status, parental mortality status, perceptions of treatment by parents in child-
hood, and 13 personality disorders. Results indicated substantial overlap between
attachment and personality-disorder measures. Two of the personality-disorder
Parts of this article were presented at the 101st annual convention of the American
Psychological Association, held August 22, 1993, in Toronto, Ontario, Canada. This
research was supported by grant #2806 from the Hogg Foundation for Mental Health
awarded to the first author. The authors would like to thank Drs. Kim Bartholomew, Keith
Davis, Deborah Jacobvitz, Daniel Klein, K. Daniel OLeary, and Janet Spence, and an
anonymous reviewer, for their invaluable comments on a previous draft. Address corre-
spondence to Kelly Brennan, Department of Psychology, State University of New York,
350 New Campus Drive, Brockport, NY 14420-2914; electronic mail: kbren-
nan@po.brockport.edu
Journal of Personality 66:5, October 1998.

Copyright 1998 by Blackwell Publishers, 350 Main Street, Malden, MA 02148, USA,
and 108 Cowley Road, Oxford, OX4 1JF, UK.
836 Brennan and Shaver
dimensions are related to the two dimensions of the attachment space; that is,
there is a two-dimensional space in which both the attachment patterns and most
of the personality disorders can be arrayed. The one personality-disorder factor
that is unrelated to attachment appears akin to psychopathy. Both personality
disorders and attachment styles were associated with family-of-origin variables.
Results are discussed in terms of encouraging further research to test the idea
that insecure attachment and most of the personality disorders share similar
developmental antecedents.
According to the most recent Diagnostic and Statistical Manual of the

American Psychiatric Association (APA), personality disorders are en-
during pattern[s] of inner experience and behavior that deviate markedly
from the expectations of the individuals culture, [are] pervasive and
inflexible, [have] an onset in adolescence or early adulthood, [are] stable
over time, and [lead] to distress or impairment (APA, 1994, p. 629).
Some personality disorders are characterized by patterns of disordered
cognitions (as in the Schizotypal and Obsessive-Compulsive disorders).
Others are characterized by problems with emotions (as in the Histrionic
and Borderline personality disorders). Virtually all personality disorders,
however, are characterized by persistent difficulties in interpersonal
relations, which is often their central feature (Widiger & Frances, 1985).
For example, the Schizoid character appears defensively devoid of any
interest in human interaction, whereas the Dependent character appears
incapable of functioning without the aid of a close other (Bornstein,
1992; Livesley, Schroeder, & Jackson, 1990). Similarly, the Avoidant
personality disorder is typically characterized by a simultaneous desire
for, and fear of, close relationships (Sheldon & West, 1990).
Despite the increasing attention paid to personality disorders since
publication of the third Diagnostic and Statistical Manual of Mental
Disorders (DSM-III-R) (APA, 1987), little is actually known about the
etiology of these disorders. Several factors have been implicated, ranging
from the genetic to the psychosocial (Benjamin, 1993/1996; Crowell,
Waters, Kring, & Riso, 1993; Dahl, 1993; Dutton, 1994; Heard & Lake,
1986; Laport & Guttman, 1996; Livesley, Jang, Jackson, & Vernon, 1993;
Mattick & Newman, 1991; Norden, Klein, Donaldson, Pepper, & Klein,
1995; Riso et al., 1996; Segal, 1988). There has been some discussion of
possible connections between personality disorders and other well-known
approaches to personality development (Ben-Amos, 1992; Bornstein,
1992; Glickhauf-Hughes & Wells, 1991; Paris, 1993; Vaillant, 1987),
including attachment theory (Gacano, Meloy, & Berg, 1992; Heard &
Attachment Styles 837
Lake, 1986; Sack, Sperling, Fagen, & Foelsch, 1996; Shaver & Clark,
1994; Sheldon & West, 1990; Torgersen & Alnaes, 1992; Van der Kolk,
Perry, & Herman, 1991; West & Keller, 1994; West & Sheldon, 1988;
West & Sheldon-Keller, 1994; Whitely, 1994), but the few empirical
studies have relatively small sample sizes. The purpose of the present
study was to investigate possible connections between personality disor-
ders and adult attachment patterns in a relatively large, nonclinical
sample of young adults. Before describing the study, we will review
relevant literature on attachment theory and research.
Literature Review
Research on attachment theory. Adult romantic love has been concep-
tualized as an attachment process that is conceptually parallel to most
peoples earliest relationship: infant to mother (Hazan & Shaver, 1987;
Shaver & Hazan, 1993). Hazan and Shaver (1987) took advantage of
Bowlbys (1969/1982, 1973, 1980) seminal writings on attachment the-
ory and Ainsworths discovery of three primary attachment types in
infancy (Ainsworth, Blehar, Waters, & Wall, 1978) to propose that these
same three attachment typessecure, avoidant, and anxious-ambiva-
lentexist in adulthood and color the ways in which adults experience
romantic love and behave in romantic relationships. Attachment style
therefore refers to characteristic patterns of experiencing romantic rela-
tionships. Hazan and Shaver (1987) devised a simple self-report measure
of the three adult attachment styles, based on extrapolations from
Ainsworth et al.s (1978) descriptions of the three primary infant attach-
ment patterns. According to the measure, secures are characterized by
comfort with intimacy and an ability to depend on their partners;
avoidants are characterized by fear of intimacy coupled with excessive
self-reliance; anxious-ambivalents are characterized as emotionally
labile, jealous, preoccupied with attachment issues and with partners,
and as desiring more closeness than their partners seem willing to allow.
Using this measure, Hazan and Shaver (1987) found theoretically
expectable attachment-style differences in how adults experience their
most important love relationships. People with different attachment
styles also reported characteristically different beliefs about romantic
love, different levels of loneliness, and different patterns of childhood
relationships with parents in ways congruent with attachment research
on parents and young children (Hazan & Shaver, 1987). Hazan and
Shavers (1987) simple measure (and later variants) has been shown to
be fairly stable across time periods ranging from eight months to four
years (e.g., Kirkpatrick & Hazan, 1994) and to be related to a wide variety
of variables, ranging from personality traits, self-esteem, and coping
styles, to relationship functioning and adaptation, to observable social
behavior. (See Feeney & Noller, 1996, and Simpson & Rholes, in press,
for a summary of research on adult romantic attachment.)
Due to attachment theorys relevance for the study of personality
structure and organization (Bowlby, 1988; Sroufe & Waters, 1977), much
of adult romantic attachment research has focused on normal personality
functioning (Bartholomew & Horowitz, 1991; Carver, 1997; Duggan &
Brennan, 1994; Mikulincer et al., 1990; Shaver & Brennan, 1992; Shaver
et al., 1996). Bowlby (1973) conceived of individual differences in
attachment as rooted in relatively stable, internal representations of the
self, important others (i.e., attachment figures), and relationships. These
internal representations, or working models, constructed gradually over
the course of infancy and immaturity, are considered to be fairly accurate
reflections of actual relationship experiences. Working models of attach-
ment theoretically account for continuity in attachment styles over the
lifespan.
Following Bowlby (1973), Bartholomew (1990) proposed a fourfold
typology of adult attachment to replace Hazan and Shavers threefold
typology, and designed self-report and interview assessment techniques
to classify individuals into one of the four categories. Conceptually,
secures possess relatively positive models of self and others. Anxious-
ambivalents, whom Bartholomew renamed preoccupied, possess a
positive model of others, along with a negative model of themselves.
Notably, Bartholomew distinguishes two types of avoidant attachment.
Fearful avoidants, akin to Hazan and Shavers avoidant type, possess
negative models of both self and others; hence they both desire and fear
intimacy. Dismissing avoidants possess a positive model of the self but
a negative model of others. According to Bartholomew (1990), Hazan
and Shavers avoidant type seemed too vulnerable and low in self-esteem
compared to Mains (Main, Kaplan, & Cassidy, 1985) description of
dismissing adults, which emerged from coded transcripts of the Adult
Attachment Interview (AAI) (George, Kaplan, & Main, 1984/1985/
1996), a technique for classifying adult parents of infants. Bartholomew
and Horowitz (1991) demonstrated that the two kinds of avoidants
differed as expected. Although both kinds of avoidants identified themselves
as socially avoidant, fearful avoidants scored lower in self-esteem

compared to dismissing avoidants. In addition, based on self- and peer
reports, dismissing individuals appeared cold, and were described as
competitive, autocratic, and introverted. Fearful avoidants, in
contrast, appeared submissive, and were described by themselves and
peers as sub-assertive, introverted, and exploitable.
Brennan, Shaver, and Tobey (1991) empirically documented the ex-
pected association between the three- and four-category attachment
measures and found a common two-dimensional structure underlying the
two measures.1 Other researchers (see Griffin & Bartholomew, 1994;
also see Brennan, Clark, & Shaver, 1998), using other measures, have
documented the same two-dimensional structure. There are three impor-
tant implications of the two measures shared underlying structure. First,
the two dimensions can be used to classify individuals into a three- or
fourfold typology, so it does not matter much which quantitative measure
is used. Second, the two-dimensional structure is further evidence that a
simple good/bad dimension is not enough to differentiate attachment
styles. This latter point is important in light of personality theorists
claims that negative affectivity underlies most self-report measures of
distress (e.g., Watson & Clark, 1984). Third, a fourfold, two-dimensional
typology in adulthood nicely matches what is emerging in the infant
literature (Crittenden, 1988; Main & Solomon, 1990). Following this
logic, Brennan et al. (1991) sought to test whether the fearful avoidant
types, who conceptually possess the worst of both worlds in terms of
mental models of self and others, might be the adult versions of the
disoriented and disorganized infants identified by Crittenden (1988)
and Main and Solomon (1990). Based on her research, Crittenden (1988)
asserts that this fourth type arises when an infants attachment figure is
extremely preoccupied or otherwise unavailable, as in the case of mater-
nal depression. As a consequence, the attachment figure is unresponsive
to the infants needs. As expected, fearful avoidants were more likely than
1. As predicted, Bartholomews secure and preoccupied types were closely akin to

Hazan and Shavers secure and anxious-ambivalent types, respectively. Bartholomews
fearful avoidant type was highly associated with Hazan and Shavers avoidant type.
Bartholomews dismissing avoidants classified themselves as either secure or avoidant
on Hazan and Shavers measure, indicating some defensiveness on the part of the
dismissing avoidants. A factor-analysis of the seven scales revealed a two-dimensional,
nearly orthogonal structure underlying both measures.
the other attachment types to report having grown up with problem-

drinking parents (Brennan et al., 1991).
Several theorists have begun to integrate attachment theory with the
literature on abnormal personality functioning and psychopathology in
infancy and early childhood (Belsky & Nezworski, 1988; Crittenden,
Parridge, & Clauseen, 1991). Sroufe and others have shown how attach-
ment assessed in infancy can account for later childhood patterns of ego
resilience, affect regulation, and problem solving in the face of stress
(Elicker, Englund, & Sroufe, 1992). The same principles thought to
underlie links between attachment and psychopathology in infancy are
being applied to adults (Shaver & Clark, 1994; Shaver & Hazan, 1993).
In adulthood, insecure adult romantic attachment styles have been asso-
ciated with loneliness (Hazan & Shaver, 1987), anxiety, depression, and
physical symptoms (Carnelley et al., 1994), negative affectivity (Simpson,
1990), neuroticism (Shaver & Brennan, 1992), low self-esteem (Brennan
& Bosson, in press; Brennan & Morris, 1997; Collins & Read, 1990;
Feeney & Noller, 1990), and maladaptive strategies for coping with
negative affect (Brennan & Shaver, 1995; Fuendeling, in press; Mikulincer
et al., 1993). These associations with insecure attachment are important
in light of the fact that the adult romantic attachment measures do not
refer to or include any phenomena relevant to individual adaptation.
Research on attachment and personality disorders. In the clinical lit-

erature, there is increasing support for conceptualizing personality dis-
orders as disorders of attachment (e.g., Heard & Lake, 1986; Shaver &
Clark, 1994; West & Sheldon, 1988; West & Sheldon-Keller, 1994).
There is growing empirical evidence connecting Borderline personality
disorder with patterns of insecure attachment reflected in representations
of childhood relationships with parents (Patrick, Hobson, Castle,
Howard, & Maughan, 1994; Sack et al., 1996; Stalker & Davies, 1995;
West et al., 1994). Patrick et al. (1994) assessed internal representations
of attachment (via the AAI; George et al., 1984/1985/1996) and found
that individuals classified as Borderline evinced the mental organization
characteristic of preoccupied attachment. Stalker and Davies (1995)
found the same pattern in a small, clinical sample of sexually abused
women. West et al. (1994) have discriminated Dependent and Schizoid
personality disorders on the basis of an enmeshed (i.e., preoccupied)
versus a detached (i.e., dismissing) interpersonal orientation. Similarly,
researchers have observed associations between recollections of insecure
attachments with one or both parents and the Self-Defeating personality

disorder (Williams & Schill, 1993). Among those diagnosed with
Avoidant personality disorder, Sheldon and West (1990) reported that
heightened desire for and fear of attachment relationships were more
diagnostic of Avoidant personality disorder than were poor social skills.
Sack et al. (1996) compared individuals with a diagnosis of Borderline
with an unselected group of college students and found evidence of a
number of indicators of attachment-related distress (e.g., fear of loss,
separation protest, compulsive care-seeking, angry withdrawal) as well
as a mixture of general ambivalent and avoidant tendencies in romantic/
sexual attachment relationships. Their pattern of findings indicate that
both preoccupation and fearful avoidance may be most closely associated
with the Borderline disorder, although they did not include Bartholomew
and Horowitzs (1991) measure in their study.
To date, however, no one has empirically examined links between adult
romantic attachment styles and all 11 DSM personality disorders in a
sample large enough to reveal structural connections between the two
domains of interpersonal functioning. It seems likely that personality
disorders and attachment styles overlap. Furthermore, both kinds of
variables may moderate the impact of environmental stresses, such as
major losses, separations, or maltreatment by significant attachment
figures. Like insecure attachment styles, personality disorders may foster
increasingly maladaptive or inflexible patterns of coping. For example,
those with Borderline personality disorder, like preoccupied individuals,
usually experience extremes of emotions, among other symptoms: After
experiencing relationship loss due to death or abandonment, both kinds
of individuals are likely to be distressed enough to contemplate suicide.
Similarly, those with Avoidant personality disorder, like fearful avoidant
individuals, appear to desire closeness with others but fear rejection
(Bartholomew, 1990; Duggan & Brennan, 1994; Mattick & Newman,
1991). As a consequence, these individuals are likely to avoid the very
social interaction that would mitigate their loneliness. Most troubling is
that both kinds of individuals may counter feelings of depression by
drinking alcohol (Brennan & Shaver, 1995), possibly compounding
interpersonal difficulties with other problems (e.g., drinking and driving,
illicit drug use, risky sexual behavior). Dismissing avoidants, on the other
hand, who describe themselves as unsociable (Duggan & Brennan,
1994) and appear to derogate the importance of close relationships, are
likely to be Schizoid. That is, both dismissing and Schizoid individuals are
likely to stress self-sufficiency and downplay attachment needs. As men-

tioned, West et al. (1994) reported an association between the Schizoid
diagnosis and the attachment-related dimension of compulsive self-
reliance. This association is what one would expect if dismissing (but not
fearful) avoidance is most closely associated with the schizoid pattern.
The Present Study

The present study had two objectives: First, we sought to explore con-
nections between attachment styles and personality disorders and to
determine whether the two kinds of variables share a common underlying
structure. We predicted that attachment styles would be associated with
all of the personality disorders, using correlational or analysis-of-
variance procedures, and share most of their underlying structure. Sec-
ond, we sought to explore the connections between attachment,
personality disorders, and three other family-background factors: paren-
tal death, parental divorce, and current representations of early childhood
relationships with parents. Previous experience of loss due to parental
divorce has already been shown to predict risky health-related behavior,
lower educational attainment, and earlier transition to parenthood (Aro
& Palosaari, 1992), and should similarly affect personality disorders in
adulthood. It is not clear whether parental death should be associated
with any of the personality disorders, because parental death is a funda-
mentally different kind of loss than parental divorce. In any case, few of
our young adult subjects were expected to have experienced the death of
one or both parents. As noted, internal representations of early childhood
relationships with parents have been shown to be related to Borderline
personality disorder (e.g., Laporte & Guttman, 1996; Patrick et al., 1994).
Thus, it seems reasonable to expect that reconstructions (i.e., memories)
of early childhood relationships with parents should relate to other
personality disorders as well.
To accomplish these objectives, we questioned a large sample of
college-age individuals about their attachment styles, parental marital
status, parental mortality status, perceptions of childhood treatment by
parents, and 13 personality disorders: Schizoid, Schizotypal, Paranoid,
Avoidant, Dependent, Obsessive-Compulsive, Passive-Aggressive, Self-
Defeating, Histrionic, Narcissistic, Borderline, Antisocial, and Sadistic.
METHOD
Sample
The sample comprised 1407 introductory psychology students (587 males and
820 females) at the University of Texas at Austin who participated in exchange
for course credit. Fully 967 (69.1%) classified themselves as Caucasian; 169
(12.1%) as Mexican-American or nonwhite Hispanic; 130 (9.3%) as Asian
American; 69 (4.9%) as African American; 21 (1.5%) as white Hispanics; 21
(1.5%) as Indian Asian; 5 (.4%) as American Indian; and 3 (.2%) as Arabic; 22
(1.5%) failed to classify themselves or identified themselves as other. Fully
918 respondents reported that their parents were still married; 420 reported that
their parents were divorced or separated; and 69 reported that one or both of
their parents were deceased. The intact group included four respondents whose
parents divorced and then remarried each other, and one respondent whose
parents never married but stayed together. Three individuals experienced paren-
tal divorce followed by the death of a parent, and therefore were omitted from
the divorce/death analyses. Finally, three individuals were adopted and had no
knowledge of their biological parents, and so were excluded from the death and
divorce analyses. Thus, 417 individuals made up the divorce group, and 66
individuals made up the death group. Respondent age ranged from 18 to 50
(m = 18). For the divorce group, median age when parents divorced was 7; nearly
80% lived with mother post-divorce, and 13% lived with father. Over 80%
reported that one or both parents remarried post-divorce. Median age at mothers
first remarriage was 9; median age at fathers remarriage was 10. Mothers and
fathers remarried a median number of one time.
Materials
Participants were given a packet of questionnaires that began with queries about
demographic variables including sex, age, ethnic group, parental marital status,
and parental mortality status.
Attachment style. Subjects were asked to classify themselves into adult roman-
tic attachment-style categories (using Bartholomew & Horowitzs [1991] meth-
odology) and to rate the secure, fearful avoidant, preoccupied, and dismissing
avoidant descriptions on a 7-point scale to indicate the degree to which each one
characterized them. Everyone, regardless of self-classification, rated each of the
attachment-style categories, allowing correlational tests of associations between
the attachment-style ratings and other variables. Previous studies indicate that
the three Hazan and Shaver (1987) attachment-style categories are approxi-
mately 70% stable over testing periods ranging from 2 weeks to 4 years; the
three ratings testretest reliabilities average .60 over the same time periods
(Brennan & Shaver, 1995; Davis et al., 1994; Kirkpatrick & Davis, 1994;
Kirkpatrick & Hazan, 1994). Given the association between the three- and
four-category measures (Brennan et al., 1991), it is likely that Bartholomews
measure will demonstrate a similar level of stability (cf. Scharfe & Bartholomew,
1994). The proportions of attachment-style categories were similar in our sample
to those reported in previous research (Bartholomew & Horowitz, 1991): 47.9%
secure, 21.0% fearful, 15.2% preoccupied, and 15.9% dismissing.
Perceptions of childhood relationships with parents. The Mother-Father-Peer
Scale, developed by Epstein (see Ricks, 1985), consists of dimensions
measuring participants perceived quality of treatment by mother, father, and
peers while growing up. The following parental dimensions were included:
(1) acceptance versus rejection (e.g., [mother/father] gave me the feeling
that she liked me as I was; she didnt feel she had to make me over into
someone else); (2) fostering independence versus overprotectiveness (e.g.,
[mother/father] encouraged me to do things for myself); and (3) defensive
idealization (e.g., [mother/father] had not a single fault that I can think of).
Participants were asked first about their mothers, then about their fathers,
for a total of six scales. In the current sample, the three mother scales had
coefficient alphas of .87 (acceptance/rejection), .82 (independence/overpro-
tectiveness), and .88 (idealization). The three father scales had coefficient
alphas of .89 (acceptance/rejection), .79 (independence/overprotectiveness),
and .92 (idealization).
Personality disorders. Students also completed a measure of 13 personality
disorders, the Personality Diagnostic Questionnaire (PDQ-R; Hyler &
Rieder, 1987). The PDQ-R is a revised version of the original Personality
Diagnostic Questionnaire (Hyler, Rieder, Williams, Spitzer, Hendler, &
Lyons, 1988), which was amended along the lines of the third Diagnostic
and Statistical Manual (DSM-III-R; APA, 1987). The PDQ-R was designed
to assess the following 11 DSM-III-R personality disorders: Schizoid, Schizo-
typal, Paranoid, Avoidant, Dependent, Obsessive-Compulsive, Passive-
Aggressive, Histrionic, Narcissistic, Borderline, Antisocial, plus two
othersfrom the 1987 DSM-III-R AppendixSelf-Defeating and Sadistic.
The PDQ-R consists of 140 true-false items and takes about 35 minutes to
complete. It can be scored by a nonprofessional to determine an overall index
of personality disturbance and to determine the existence of each of the
specific personality disorders listed above. Approximately 30% of the state-
ments included in the PDQ-R are reverse-scored.
The PDQ-R has high testretest reliability (Hyler, Skodol, Kellman, Oldham,
& Rosnick, 1990) and is valid for use with both clinical and nonclinical
populations (Hyler et al., 1990; Hyler, Skodol, Oldham, Kellman, & Doidge,
1992).2 Although the PDQ-R should not be considered a substitute for clinical
interviews, its results agreed with judgments of each of two psychiatrists to the
same extent as the two psychiatrists agreed with each other (Hyler et al., 1992).
The PDQ-R has been shown to have high negative predictive power, but lower
positive predictive power (Hyler et al., 1990; Hyler et al., 1992). That is, the
PDQ-R has a relatively low rate of false-negative diagnoses (results indicating
the absence of a personality disorder when one actually exists) but a relatively
high rate of false-positive diagnoses (results indicating the presence of a per-
sonality disorder when none exists). In the present study of students who were
presumed to be mostly in the normal range on clinical measures, this bias was
considered acceptable, because we were looking for structural similarities
between personality disorder dimensions and attachment-style dimensions. We
were not seeking to make categorical clinical diagnoses. Each personality-
disorder scale will be described briefly below.
The Schizoid scale measures a persons tendency to retreat from others,
thereby isolating oneself from relationships with family and friends (8 items).
One item on this scale is Others see me as cold and unemotional. The
Schizotypal scale describes individuals who have experienced a break from
reality (12 items). These individuals may believe that they are receiving sub-
liminal messages or that they can communicate through ESP. An example is I
have an unusual ability to know that some things will happen before they actually
do. The Paranoid scale describes a person who is suspicious of others and fears
persecution by others (7 items). He or she is afraid of being tricked or betrayed
by others. A sample item is I often wonder whether the people I know can really
be trusted. The Avoidant scale describes a person who tends to lack self-assurance
and to fear ridicule by or disapproval from others (7 items). Therefore the
Avoidant person prefers to minimize contact with others. An example is I make
friends with people only when I am sure they like me.
The Dependent personality disorder scale refers to people who have an
excessive need to rely on others (9 items). For instance, they may agree with all
authority figures and be overly reliant on relationship partners. One item is I
want people to like me so much that I volunteer to do things that Id rather not
do. The Obsessive-Compulsive scale refers to an excessive focus on rules (9
items). This disorder often manifests itself in excessive concern for following
norms in social situations and for following a proper regimen for personal
hygiene, work assignments, and so on. The Obsessive-Compulsive individual
sees only one proper way to do things and becomes easily upset when rules are
broken. A sample item is I waste time trying to make things too perfect. The
2. Internal consistency is not the best indicator of the reliability of these scales because
the scales are essentially heterogeneous lists of items. Testretest reliability is a better
indicator.
Passive-Aggressive scale describes a person who uses passive means to accom-

plish aggressive behaviors (9 items). For instance, rather than directly refusing
to perform a task, a person with Passive Aggressive disorder might intentionally
make errors (to communicate that he or she did not really want to perform the
task). A sample item is When I resent doing something I tend to work slowly
or screw things up.
The Self-Defeating scale assesses the tendency to put oneself down and to
fear success and happiness (11 items). For example, I find myself more
interested in people who treat me badly than in people who are nice to me. The
Histrionic scale assesses a tendency to present oneself as friendly, accommo-
dating, and unselfish, when in reality one is vain and self-serving (10 items).
These people like to be the center of attention and often exaggerate actions and
emotions to attain attention from others. An example item is I often act very
emotionally when little things go wrong. The Narcissistic scale assesses the
tendency toward being pompous, egocentric, and self-absorbed (11 items). This
sense of personal greatness may be a defensive reaction to internal feelings of
worthlessness. For example, I often find myself thinking about how great a
person I am, or will be.
The Borderline personality disorder scale assesses the extent to which people
feel out of control of their emotions and behaviors (12 items). They may
experience mood swings and behave on impulse (drinking, shoplifting, or other
acts that often get them into trouble). An example is My feelings towards
another person can often change drastically. The Antisocial personality disor-
der scale characterizes people who are very focused on personal gain without
regard for the effects of their behavior on others (12 items). For example, I
dont care if other people get hurt so long as I get what I want. The Sadistic
scale describes a person who is physically and/or emotionally harmful toward
other people or animals (8 items). Such a person derives a sense of self-worth
and power from causing others harm. For example, I have enjoyed humiliating
other people.
RESULTS
Preliminary Analyses
Prevalence rates. Personality disorders were scored categorically
(presence vs. absence of a disorder) as well as on a continuum (i.e., the
degree to which one possesses attributes of a particular disorder).3 The
3. For the most part, the distributions of responses on the personality-disorder scales
were not at all skewed. The skewness of the Antisocial and Sadistic scales (1.32 and 1.57,
respectively) were by far the highest (average skew without these two scales was .67),
prevalence of various personality-disorder diagnoses ranged from 5.2%

(Schizoid) to 42.5% (Paranoid). Most of the subjects had at least one
personality disorder (N = 1056 of 1407, or 75.0%), consistent with the
PDQ-Rs positive bias (cf. Moldin, Rice, Erlenmeyer-Kimling, &
Squires-Wheeler, 1994). The median number of disorders was 2, with an
average of 2.45. Frequencies and percentages of each disorder in the
current sample are displayed in the first column of Table 1.
Using the same measure and scoring procedures, two other sets of
researchers have reported similar prevalence rates in late-adolescent sam-
ples (Johnson & Bornstein, 1992; Johnson, Bornstein, & Sherman, 1996;
Maffei et al., 1995). In their random sample of first-year medical students
at a university in Milan, Maffei et al. reported an overall prevalence rate of
88.7% for all 13 disorders. Maffeis prevalence rates for individual disorders
were very similar to ours, except that they reported a higher rate for Paranoid
(66.0%) and Schizoid (10.2%), and a lower rate for Borderline (11.7%),
Schizotypal (12.1%), Passive-Aggressive (6.6%), and Sadistic (2.7%) (Maffei
et al., 1995, p. 335). In two samples of American undergraduates, Johnson
and his colleagues (Johnson & Bornstein, 1992, p. 453; Johnson et al., 1996,
p. 81) used the PDQ-Rs authors scoring criteria (Hyler et al., 1988, 1989)
and reported personality-disorder prevalence rates very similar to those
reported here. Frequencies and percentages of each disorder in the Johnson
et al. (1996) sample (which is larger than the Johson & Bornstein sample)
are shown in the second column of Table 1.
Such high prevalence rates prompted Johnson and his colleagues to
devise more stringent scoring algorithms for the PDQ-R, resulting in an
overall prevalence rate of 11.3% for 12 of the 13 disorders (Johnson et
al., 1996). Prevalence rates obtained using the more stringent scoring
criteria in both the current sample and the Johnson et al. (1996) sample
are listed in the third and fourth columns of Table 1. Note that in the
current sample the overall prevalence rate dropped from 75.0% to 14.3%
when Johnson et al.s more stringent scoring methods were used. None-
theless, we chose not to use Johnson et al.s (1996) scoring methods,
because these methods appear to rely too heavily on reports of distress.
but not high enough to violate the relatively robust assumptions of analysis-of-variance
procedures. Hence, although we analyzed the data primarily with the continuous ratings
of personality disorders (following the advice of Hyler, Lyons, Rieder, Young, Williams,
& Spitzer, 1990), we have provided information about personality-disorder categories
with respect to basic frequencies and associations with gender and attachment style.
Table 1
Frequency of Disorders (and % of Sample) Using Standard and
More Stringent Scoring Criteria in the Current Sample and
in Johnson et al.s (1996) Study
Frequencies (Percentages) of Each

Personality Disorder
Standard Scoring More Stringent Scoring
Current Johnson Current Johnson
Sample et al. Sample et al.
Schizoid 73 (5.2%) 17 (4%) 0 (0%) 0 (0%)
Schizotypal 258 (18.4%) 66 (15%) 52 (3.7%) 10 (2.3%)
Paranoid 596 (42.5%) 194 (44%) 38 (2.7%) 12 (2.7%)
Avoidant 245 (17.4%) 75 (17%) 83 (5.9%) 16 (3.6%)
Dependent 182 (12.9%) 53 (12%) 56 (4.0%) 4 (.9%)
Obsessive-Compulsive 384 (27.3%) 141 (32%) 52 (3.7%) 9 (2.0%)
Passive-Aggressive 254 (18.1%) 84 (19%) 38 (2.7%) 4 (.9%)
Self-Defeating 102 (7.3%) 35 (8%) 51 (3.6%) 14 (3.2%)
Histrionic 431 (30.7%) 150 (34%) 28 (2.0%) 9 (2.0%)
Narcissistic 236 (16.8%) 115 (26%) 13 (.9%) 9 (2.0%)
Borderline 426 (30.5%) 150 (34%) 142 (10.2%) 39 (8.8%)
Antisocial 103 (7.3%) 62 (14%) 0 (0%) 2 (.5%)
Sadistica 113 (8.1%)
At least one diagnosisb 1056 (75.0%) 201 (14.3%) 50 (11.3%)
a Johnson et al. (1996) did not provide any information with regard to the Sadistic scale,
so we could not provide alternative scoring for this scale. b Also, Johnson et al. did not
provide the overall prevalence rate obtained when using standard scoring methods.
According to Johnson et al., individuals cannot be classified as having a

disorder unless they report at least some distress, a criterion that may be
overly exclusionary. A person with a Schizoid disorder, for example, may
not acknowledge any distress. Note that, using the distress criterion, no
individuals were diagnosed as Schizoid in either sample. In any case, we
were interested in the associations between attachment styles and per-
sonality disorders, and these associations generally did not change when
the more stringent classification criteria were applied.4
4. The association of the Narcissistic personality disorder (present vs. absent) was not
associated with attachment style, perhaps because the percentage of individuals with this
Gender differences. The percentage of males with at least one person-

ality disorder was 76.8% and the percentage of females with at least one
disorder was 73.8%. Examination of the individual personality disorders
revealed six statistically reliable gender differences. Males were more
likely to qualify for a diagnosis of Schizotypal (2 [df = 1, N = 1402] =
4.83, p < .05), Passive-Aggressive (2 [df = 1, N = 1405] = 8.77, p < .01),
Narcissistic (2 [df = 1, N = 1402] = 17.26, p < .001), Antisocial (2 [df
= 1, N = 1403] = 33.72, p < .001), or Sadistic (2 [df = 1, N = 1400] =
26.31, p < .001). Females were more likely to qualify for a diagnosis of
Histrionic (2 [df = 1, N = 1403] = 5.35, p < .05).
Looked at another way, consistent gender differences emerged on
several personality-disorder scales. Males scored higher on the following
scales: Schizoid (Mmales = 1.47, Mfemales = 1.19, t [1254] = 4.48, p < .001),
Schizotypal (Mmales = 2.95, Mfemales = 2.68, t [1202] = 2.70, p < .01),
Passive-Aggressive (Mmales = 2.97, Mfemales = 2.66, t [1207] = 3.19, p <
.01), Narcissistic (Mmales = 3.18, Mfemales = 2.90, t [1170] = 3.15, p < .01),
Antisocial (Mmales = 1.79, Mfemales = .98, t [1027] = 11.03, p < .001), and
Sadistic (Mmales = 1.85, Mfemales = 1.30, t [1060] = 7.82, p < .001). Females
scored higher on the Dependent (Mmales = 2.06, Mfemales = 2.32, t [1270]
= 2.65, p < .01) and Histrionic (Mmales = 2.36, Mfemales = 2.72, t [1288] =
3.85, p < .001) scales. Interestingly, there were no gender differences on
the Borderline personality disorder scale, although this disorder is often
thought to be more characteristic of women than men (e.g., Ekselius et
al., 1996; but see also Golomb, Fava, Abraham, & Rosenbaum, 1995).
Intercorrelations of the scales. Intercorrelations of the personality-dis-

order scales are shown in Table 2. The moderately high correlations
indicate considerable comorbidity: Individuals scoring high on one
personality-disorder scale tend to score high on others as well. As the
sections below explain, however, the correlations are not uniform; and
their varying sizes reflect an underlying structure.
disorder (.9%) was quite low. No one was classified as having a Schizoid or an Antisocial
disorder, so these scales could not be associated with attachment style. Lastly, Johnson
et al. (1996) did not report scoring criteria for the Sadistic scale, so we could not use their
alternative scoring methods with this scale.
Table 2
Intercorrelations of 13 Personality Disorders Scales
Scale Name 1 2 3 4 5 6 7 8 9 10 11 12 13
1. Schizoid
2. Schizotypal .25
3. Paranoid .12 .54
4. Avoidant .22 .51 .44
5. Dependent .13 .29 .25 .42
6. Obsessive .08 .40 .36 .37 .27
7. Passive-Agg. .06 .34 .32 .31 .34 .37
8. Self-Def. .16 .53 .42 .42 .37 .38 .35
9. Histrionic .13 .32 .28 .26 .43 .34 .36 .38
10. Narcissistic .05 .39 .38 .34 .32 .37 .43 .37 .45
11. Borderline .07 .46 .43 .36 .37 .35 .40 .48 .44 .44
12. Antisocial .12 .24 .16 .12 .07 .07 .30 .23 .16 .28 .35
13. Sadistic .09 .26 .24 .13 .14 .22 .29 .24 .20 .31 .28 .36
Note. Because of the large N (13941406, depending on missing values), all correlations larger than .09 are significant beyond the .001 level.
Connections Between Attachment Styles and

Personality Disorders
In this section, we will test our first hypothesis, that there would be
associations between attachment styles and personality disorders, and
that the two kinds of variables would share most of their underlying
structure.
Tests using personality-disorder categories. The association of cate-

gorical attachment style with the categorical personality-disorder vari-
able (present vs. absent) reached significance (2 [df = 3, N = 1384]
= 148.85, p < .001). Secures were nearly twice as likely not to have a
personality disorder as to have one (75.0% vs. 38.8%). Fearfuls were
about four times as likely to have at least one personality disorder as not
(26.0% vs. 6.3%). Preoccupieds were about three times as likely to have
a disorder as not (18.3% vs. 5.7%). Dismissing individuals were nearly
equally distributed across the two categories of at least one disorder
and no disorders (16.9% vs. 12.9%). Said another way, for individuals
classified as secure, fearful, preoccupied, and dismissing, the percent-
age with at least one disorder was 60.6%, 92.4%, 90.5%, and 79.5%,
respectively.
For each personality disorder, we cross-tabulated the presence versus
absence of the disorder with attachment-style category, and performed a
chi-square test on each cross-tabulation. The frequency of participants
in each attachment-style category with a particular disorder is shown in
Table 3, along with the percentages of the four attachment styles within
each disorder category. In every case, the distribution of attachment styles
deviated from that of the total sample, as evidenced by the significant
chi-square tests. The percentage of secures with each personality disor-
der was lower than the overall percentage of secures in the total sample.
The relative proportion of each insecure group was usually greater than
its frequency in the total sample. For example, both kinds of avoidants
were much more likely to earn a diagnosis of Schizoid; fearfuls were
particularly likely to earn a diagnosis of Avoidant personality disorder;
preoccupied individuals were particularly likely to earn a diagnosis of
Dependent personality disorder.
Tests using personality-disorder dimensions Personality-disorder rat-

ings were then analyzed as a function of attachment-style category and
Table 3
Frequency of Disorders Within Each Attachment Group (and % of Each
Attachment Group Within Each Disorder Category)
Frequencies of Each Personality Disorder

Secure Fearful Preoccupied Dismissing 2
Schizoid 6 (8.5%) 29 (40.8%) 3 (4.2%) 33 (46.5%) 88.49
Schizotypal 41 (16.2%) 99 (39.1%) 69 (27.3%) 44 (17.4%) 143.90
Paranoid 174 (29.7%) 199 (34.0%) 107 (18.3%) 106 (18.1%) 162.02
Avoidant 49 (20.3%) 92 (38.2%) 61 (25.3%) 39 (16.2%) 106.90
Dependent 49 (27.4%) 46 (25.7%) 71 (39.7%) 13 (7.3%) 110.97
Obsessive-Compulsive 118 (31.7%) 116 (31.2%) 79 (21.2%) 59 (15.9%) 64.63
Passive-Aggressive 69 (27.8%) 76 (30.6%) 62 (25.0%) 41 (16.5%) 58.07
Self-Defeating 13 (12.9%) 41 (40.6%) 36 (35.6%) 11 (10.9%) 79.72
Histrionic 172 (40.7%) 102 (24.1%) 101 (23.9%) 48 (11.3%) 47.38
Narcissistic 70 (30.4%) 79 (34.3%) 45 (19.6%) 36 (15.7%) 44.21
Borderline 125 (29.8%) 135 (32.2%) 103 (24.6%) 56 (13.4%) 113.30
Antisocial 24 (23.5%) 33 (32.4%) 19 (18.6%) 26 (25.5%) 27.85
Sadistic 26 (23.2%) 33 (29.5%) 27 (24.1%) 26 (23.2%) 30.20
Overall Proportionsa 663 (47.9%) 291 (21.0%) 210 (15.2%) 220 (15.9%) 1407 (100%)
Note. N = 1407 for most tests. (In some cases, missing values on either attachment or personality-disorder measures resulted in minor discrepancies
between frequency of the total sample with a particular disorder and sum of the frequencies of each of the four attachment groups with the
disorder.) All tests were significant at p < .001.
aFrequencies (and %s) of the four attachment-style groups in the total sample.
gender in a 2 4 multivariate analysis of variance (MANOVA). The

resulting interaction term was not significant, so the data were analyzed
further without respect to gender. A MANOVA conducted on the personality-
disorder ratings as a function of attachment-style category was signifi-
cant, Hotellings F (df = 39, 4022) = 20.57, p < .001. The attachment-
group means for each of the personality-disorder scales are displayed in
Table 4 along with the univariate Fs and effect sizes (2s).
With two exceptions, securely attached participants scored lower on
every personality-disorder scale compared to insecurely attached partici-
pants. Secures scored higher than dismissing individuals on the Depen-
dent scale. Secures did not differ from dismissing individuals on the
Histrionic scale (both groups scored low). Fearful, dismissing, and
preoccupied participants generally scored similarly, except that preoccu-
pieds outscored fearfuls on the Histrionic and Dependent scales, but
scored lower than fearfuls on the Paranoid and Schizoid scales. Dis-
missings, for the most part, scored intermediately between secures, on
the one hand, and fearful and preoccupied individuals, on the other. There
were a few exceptions to this rule: The dismissing group scored lowest
on the Dependent scale. Dismissing individuals scored lower than fearful
and preoccupied individuals on the Histrionic scale. Dismissings scored
higher than fearfuls on the Schizoid scale. Dismissings scored similarly
to preoccupieds (but lower than fearfuls) on the Paranoid scale. Dis-
missings scored similarly to fearful and preoccupied individuals on the
Antisocial and Sadistic scales.
Discriminating among attachment-style categories using personality-

disorder scales. To explore further the association between personality-
disorder scales and attachment styles, a discriminant-function analysis
was conducted using the 13 personality-disorder scales to predict sub-
jects self-reported attachment-style category. Table 5 displays the correla-
tions of the personality-disorder scales with the discriminant functions.
Table 6 displays the canonical discriminant functions evaluated at each
attachment-style groups mean. Function 1 distinguished insecures (es-
pecially fearfuls) from secures (2 [df = 39] = 701.44, p < .001). Function
2 distinguished preoccupied and dismissing groups (2 [df = 24] =
250.59, p < .001). Function 3 was only marginally significant, weakly
discriminating dismissing and preoccupied from fearful individuals (2
[df = 11] = 19.55, p < .10). Functions 1 and 2 accounted for 66.17% and
Table 4
Mean Personality-Disorder Ratings as a Function of Four Attachment-Style Categories
Personality-Disorders Attachment-Style Category

Ratings Secure Fearful Preoccupied Dismissing Univariate F 2
Schizoid .98a 1.69 c 1.04b 1.98c 64.12 .12
Schizotypal 2.05a 3.81 c 3.49c 2.98b 89.23 .16
Paranoid 2.50a 4.33 c 3.67b 3.45b 95.19 .17
Avoidant 1.34a 2.76 c 2.66c 1.98b 78.42 .15
Dependent 1.89b 2.56 c 3.56d 1.45a 70.41 .13
Obsessive-Compulsive 3.02a 4.03 c 3.99c 3.55b 34.33 .07
Passive-Aggressive 1.35a 2.58 c 2.76c 1.83b 78.17 .08
Self-Defeating 2.18a 3.14 c 3.38c 2.54b 65.90 .15
Histrionic 2.35a 2.81 b 3.42c 2.13a 27.56 .06
Narcissistic 2.61a 3.61 c 3.53c 2.91b 38.14 .08
Borderline 2.93a 4.31 c 4.32c 3.40b 54.12 .11
Antisocial 1.02a 1.63 b 1.43b 1.61b 21.79 .05
Sadistic 1.32a 1.71 b 1.72b 1.73b 11.20 .02
Note. dfs ranged from 3 and 1372 to 1380. All tests were significant at p < .001. Within each row, means whose superscripts differ are significantly
different at p < .05.
Table 5
Rotated Correlations Between 13 Personality Disorders Scales and
Three Discriminant Functions Predicting Attachment-Style Category
Personality Disorder Rating Function 1 Function 2 Function 3

Paranoid .72* .03 .39
Schizotypal .69* .15 .08
Avoidant .63* .23 .05
Self-Defeating .61* .38 .20
Borderline .51* .29 .04
Narcissistic .43* .22 .18
Obsessive-Compulsive .43* .14 .14
Dependent .31 .77* .16

Schizoid .44 .60* .31
Histrionic .22 .48* .15
Passive-Aggressive .43 .24 .44*

Sadistic .24 .01 .40*
Antisocial .35 .08 .36*
Note. N = 1358. Function 1, 2 (39) = 701.44, p < .001; Function 2, 2 (24) = 250.59,
p < .001; Function 3, 2 (11) = 19.55, p < .10. Asterisks denote each personality disorders
scales highest correlation with one of the three discriminant functions.
Table 6
Discriminant Functions Evaluated at Attachment-Style
Group Centroids (Means)
Attachment Style Function 1 Function 2 Function 3

Secure .63 .02 .03
Fearful .86 .06 .16
Preoccupied .45 .80 .15
Dismissing .33 .74 .17
31.27% of the variance, respectively; Function 3 accounted for the

remainder (2.56%).
The fearful group appeared the most troubled, scoring highest on the
discriminant function that correlated with personality disorders involving
some sort of distortion of reality along with negativity about others

(Paranoid, Schizotypal, Avoidant, Self-Defeating, Borderline, Narcissis-
tic, and Obsessive-Compulsive). The preoccupied and dismissing groups
appeared to suffer from diametrically opposed personality problems as
evidenced by the pattern of scores on the second function, which was
about Dependent and Histrionic versus Schizoid personality charac-
teristics. This pattern fits with Bartholomew and Horowitzs model:
Preoccupied persons appear overly dependent, whereas Dismissing per-
sons appear counter-dependent. The third personality-disorder func-
tion, which appeared akin to psychopathy (Passive-Aggressive, Sadistic,
and Antisocial), accounted for negligible variance among attachment
categories.
Underlying structure of the attachment-style ratings. To d e t e r m i n e

whether the four attachment-style ratings conformed to the two-
dimensional structure postulated by Bartholomew (1990), a principal
components analysis was conducted. Two principal components emerged
and were rotated, using an oblique procedure. (The same structure
appeared regardless of rotation method.) Factor 1 was defined by the
difference between secure and fearful attachment (with loadings of .88
and .82, respectively). Factor 2 was defined by the difference between
dismissing and preoccupied attachment (with loadings of .82 and .72,
respectively). These results are similar to those obtained by Brennan et
al. (1991). The first attachment factor can be conceptualized as Insecurity;
the other attachment factor can be conceptualized as Defensive Emotional
Style. The two factors accounted for 72.8% of the variance and were
correlated .03. (That is, they were essentially orthogonal.)
Underlying structure of the personality-disorder dimensions. To deter-

mine the structure of the personality-disorder scales, a principal compo-
nents analysis was conducted on all 13 personality-disorder scales.5
Three principal components emerged and were rotated using an oblique
procedure (again, the same structure appeared regardless of rotation
procedure). The results mirrored those of the discriminant function
5. Principal components analyses were also conducted separately for males and females.
The resulting three-factor structure in both subsamples was virtually identical to the one
obtained with the total sample. Correlations of each of the three total-sample factors with
the three corresponding factors obtained within each subsample all exceeded .99.
analysis. Avoidant, Schizotypal, Paranoid, Self-Defeating, Obsessive-

Compulsive, and Borderline personality-disorder scales loaded highly
on the first factor (the respective loadings were .84, .78, .69, .66, .60, and
.44). Schizoid, Histrionic, and Dependent scales loaded highly on the
second factor (the respective loadings were .79, .59, .55). The Anti-
social, Sadistic, Passive-Aggressive, and Narcissistic scales loaded
highly on the third factor (the respective loadings were .86, .75, .41, and
.37). (The Borderline, Passive-Aggressive, and Narcissistic scales loaded
nearly equally on the first and third factors. Schizoid and Dependent
scales also loaded on Factor 1.)
Given the pattern of factor-loadings, it is not surprising that the three
factors were somewhat intercorrelated: Factors 1 and 2 correlated .21;
Factors 1 and 3 correlated .37; Factors 2 and 3 correlated .12.6 Together
the three factors accounted for 56.3% of the variance. Notice that the
three functions appear strikingly similar to those that emerged from the
discriminant function analysis predicting attachment-category member-
6. Due to the moderate intercorrelations among the disorders scales, we decided to

conduct a second, item-level principal components analysis on the 135 individual
personality-disorder items. Three principal components emerged and were rotated using
the SPSS oblimin procedure (again, the same structure appeared regardless of rotation
procedure). The first factor, Neuroticism, was composed of items indicative of intense
negative affect, especially anxiety, oversensitivity to criticism or rejection, and other
self-defeating behaviors. The second factor, Psychopathy, derived mainly (95%) from
the Antisocial, Sadistic, and Borderline scales, and included items indicative of aggres-
sive, self-destructive, or impulsive behaviors. The third factor, Extraversion, was com-
posed of items indicative of sociability, or preference for and ease of being around others.
These three factors are similar to Eysencks (1970) Neuroticism, Psychoticism, and
Extraversion scales.
When four factors were selected, Factor 2 split into self-destructive (mainly Borderline
items) and other-destructive items (mainly Sadistic and Antisocial items). When five
factors were selected, the fifth factor contained a handful of items indicative of endorse-
ment of unusual ideas (e.g., belief in ESP, premonitions, receiving special messages
from others). Such a five-factor solution, although insufficiently strong in this nonclinical
sample, is reminiscent of other researchers findings revealing the presence of the
five-factor model in assessments of personality disorders (e.g., Wiggins & Pincus,
1994). The results of our item-level analysis (whether 3, 4, or 5 factors were selected)
contrasts with those of other researchers, who found anywhere from 12 to 23 factors
using the PDQ-R (Hyler et al., 1990; Livesley et al., 1992) or clinical interviews (Ekselius,
Lindstrom, von Knorring, Bodlund, & Kullgren, 1994; Torgersen, Skre, Onstad, Edvardsen,
& Kringlen, 1993).
ship. The first factor seemed to be a general pathology factor that taps
low sociability, distrust, low self-esteem, and disordered thinking pat-
terns; we therefore named it Generalized Pathology. The second factor
appeared to be about excessive self-reliance (vs. dependency); we named
it Counter-Dependence. The third factor appeared to be about interper-
sonal aggression and callousness; we named it Psychopathy.
Associations between the attachment and personality-disorder factors.

A higher-order principal components analysis was conducted on the
two attachment factors plus the three personality-disorder factors.
Two principal components emerged and were rotated using an oblique
procedure. The first higher-order factor was defined by the Insecurity
attachment factor (with a loading of .83), the Generalized Pathology
personality-disorder factor (with a loading of .82), and the Psycho-
pathy personality-disorder factor (with a loading of .66). The second
factor comprised the Counter-Dependence personality-disorder factor
(with a loading of .83) and the Defensive Emotional Style attachment
factor (with a loading of .82). Together, the two factors accounted for
64.5% of the variance. The two factors correlated .11; that is, they
were essentially orthogonal.
From these results we can say that the attachment styles and 10 of
the 13 personality disorders (all but the scales making up the Psycho-
pathy factor) are located in a two-dimensional conceptual space
corresponding to Bartholomews (1990) attachment-style typology.
As mentioned, the first attachment-related factor can be conceptual-
ized as Insecurity; the other attachment-related factor can be concep-
tualized as Defensive Emotional Style. The first personality-disorder
factor, General Pathology, is made up of 6 of the 13 personality-
disorder scales and is similar to the first attachment factor. The second
personality-disorder factor, Counter-Dependence, is made up of 3
additional personality-disorder scales (Schizoid, Histrionic, and De-
pendent) and is similar to the second attachment factor. Four remaining
personality-disorder scales (Antisocial, Sadistic, Passive-Aggressive,
and Narcissistic) formed a third factor, Psychopathy, which was mod-
estly correlated with the first attachment factor (r = .28) but not with
the second (r = .002).
Given that the two kinds of variables share much of their variance, it
seems reasonable to look for common origins. In the next section, we
examine the association of attachment and personality disorders with
three family-background variables: parental death, parental divorce, and

perceptions of parental caregiving.
Association of Personality Disorders and

Attachment Styles with Family Variables
In this section, we test our second hypothesis. We predicted that parental
divorce and negative representations of early relationships with parents
would be related to both attachment insecurity and personality disorders.
The association of parental death with either attachment insecurity or
personality disorders was less clear, and our results on that issue will be
considered exploratory.
Parental divorce. The associations of parental marital status (divorced

vs. nondivorced) with both attachment style and personality disorders
were explored. A chi-square test was computed on the cross-tabulation
of attachment-style category and (dichotomous) parental marital status:
2 (df = 3, N = 1312) = 8.69, p < .05. A series of two-tailed tests of
independent proportions revealed that individuals with divorced parents
were more likely to be fearful (z = 2.87, p < .01) and less likely to be
secure (z = 1.91, p < .05) than individuals from intact families. Consid-
ering that secure and fearful scales load on a single Insecurity factor, the
latter results suggest that, on average, subjects who had experienced the
divorce of their parents were less secure (in the attachment sense).
Personality-disorder ratings were analyzed as a function of parental
marital status and gender in a 2 2 MANOVA. The overall interaction
term was significant (Hotellings F [df = 13, 1294] = 1.98, p < .05).
Compared to females from intact families, females from divorced fami-
lies scored higher on the following three personality disorder scales:
Schizotypal (Mintact = 2.55, Mdivorced = 2.98, t = 3.11, p < .01), Self-
Defeating (Mintact = 1.78, Mdivorced = 2.19, t = 3.25, p < .01), and Borderline
(Mintact = 3.32, Mdivorced = 3.90, t = 3.96, p < .001). The main effect for
parental marital status was also significant, Hotellings F (df = 13, 1296)
= 2.12, p < .05. Compared to college-aged children whose parents were
still married, the children with divorced parents scored significantly higher
on two personality-disorder scales: Paranoid (Mintact = 3.10, Mdivorced =
3.44, t = 3.26, p < .01) and Narcissistic (Mintact = 2.95, Mdivorced = 3.14,
t = 2.16, p < .05). (Significant main effects for the Borderline and
Schizotypal scales emerged but were qualified by the interaction of
gender and parental marital status.) These results should be treated with
caution, as parental marital status accounted for less than 1% of the
variance in any particular personality-disorder scale.
The five higher-order factors were similarly examined as a function of
parental divorce. Two of the five factors were associated with parental
divorce: Attachment Insecurity (Mintact = .06, Mdivorced = .12, t = 3.00,
p < .01) and Generalized Pathology (Mintact = .05, Mdivorced = .10, t = 2.51,
p < .05) were higher for those from divorced families.
Parental death. The associations of parental mortality status (one or

both parents died vs. neither parent died) with both attachment style and
personality disorders were explored. A chi-square test was computed on
the cross-tabulation of attachment-style category and (dichotomous)
parental mortality status: 2 (df = 3, N = 1381) = 11.36, p < .01. A
two-tailed test of independent proportions revealed that individuals with
deceased parents were more likely to be dismissing than were individuals
whose parents were both still living (z = 2.92, p < .01).
A MANOVA conducted on the set of personality disorders as a
function of parental mortality status and participants gender was not
significant. Nor was the MANOVA significant when gender was ignored,
Hotellings F (df = 13, 1363) = 1.25, p = .23. The absence of significant
effects might have been due partly to the fact that only a small fraction
of the sample had experienced the death of one or both parents. Still, the
effect on dismissing attachment was discernible.
These findings were mirrored by those obtained when examining
the five higher-order factors as a function of parental mortality.
Scores on the second attachment factor, Defensive Emotional Style,
were elevated for those who had experienced the death of one or both
parents (M living = .01, M deceased = .34, t = 2.51, p < .05). No other
higher-order factors were associated with parent mortality status.
Perceptions of childhood relationships with parents. Mean scores on

the scales measuring perceptions of childhood relationships with parents
were analyzed as a function of gender and attachment-style grouping in
a 2 4 MANOVA. The overall interaction term was not significant, so
the data were analyzed further without respect to gender. A MANOVA
conducted on the mother-father scales as a function of attachment-style
category was significant, Hotellings F (df = 18, 4046) = 8.50, p < .001.
(The attachment-group means for each of the six mother-father scales

are listed in Table 7, along with the univariate Fs and effect sizes [2s].)
As shown in Table 7, all six parent scales differed significantly as a
function of attachment style. Relative to insecure groups, secures recalled
more accepting relationships with mothers and fathers. Insecure groups
appeared to perceive their relationships with fathers as equally negative,
but fearful individuals described their mothers as even more rejecting
than did preoccupied and dismissing individuals (who did not differ from
each other). Both secure and dismissing individuals recalled parents as
having fostered their independence, relative to fearful and preoccupied
individuals (who did not differ from each other). The fearful group
reported less mother idealization than any other group. Preoccupieds
reported less mother idealization than did secures, and dismissings
scored between the two. Both secures and dismissings reported more
idealistic childhood relationships with fathers than did fearfuls and
preoccupieds (who did not differ from each other). Interestingly, despite
dismissing individuals perceptions of parental rejection, they did not
differ from secures in degree of idealization of either parent. These
findings partly replicate those found by Carnelley and Janoff-Bulman
(1992), but provide added information about the dismissings. This pat-
tern also fits with the idea in the AAI scoring system (Goldwyn & Main,
1994) that parental idealization (in combination with evidence that one
or both parents were rejecting) is a sign of dismissing attachment.7
We then correlated the six parental-perception scales with the 13
personality-disorder scales. (These correlations are shown in Table 8.) In
general, all three positive parental features were correlated significantly
negatively, but weakly, with personality disorders.
7. At first glance, it may appear odd that both secures and dismissings report similar
levels of parental idealization. It is problematic to use Epsteins idealization scale to
single-handedly attempt to discriminate persons with positive parental caregiving histo-
ries (and hence accurately positive descriptions of parents) from persons with negative
parental caregiving histories (and hence inaccurately positive, or idealistic, descriptions
of parents). Logically, scores on the idealization scale must be evaluated in conjunction
with scores on the acceptance scale. That is, secures idealization, in the context of their
reports of parental acceptance, may accurately reflect a history of positive parenting.
Dismissings idealization, in the context of their reports of parental rejection, is not at all
consistent with a history of positive parenting. Thus, only the dismissing groups
idealization is likely to represent true idealization (cf. Shaver, Belsky, & Brennan,
1997).
Table 7
Mean Ratings of Perceptions of Childhood Relationships with Parents
as a Function of Four Attachment-Style Categories
Attachment-Style Category
Secure Fearful Preoccupied Dismissing Univariate F 2
Relationship with Mother
Acceptance 4.49c 4.11a 4.29b 4.32b 23.44 .05
Fostered Independence 3.73b 3.47a 3.47a 3.62b 14.65 .03
Idealization of Mother 3.48c 3.00a 3.25b 3.37bc 16.20 .03
Relationship with Father

Acceptance 4.25b 3.79a 3.84a 3.97a 27.14 .06
Fostered Independence 3.88b 3.61a 3.63a 3.84b 16.76 .04
Idealization of Father 3.19b 2.72a 2.75a 3.06b 14.30 .03
Note. df = 3, 1355. All tests were significant at p < .001. Within each row, means whose superscripts differ are significantly different at p < .05.
Table 8
Correlations Between Personality-Disorder Ratings and Perceptions of Childhood Relationships with Parents
Perceptions of Childhood Relationships with Parents

Mother Scales Father Scales
Personality- Fostered Warm Was the Fostered Warm Was the
Disorder rating independence acceptance ideal parent independence acceptance ideal parent
Schizoid .08 .17 .15 .10 .19 .12
Schizotypal .18 .24 .14 .21 .23 .14
Paranoid .19 .21 .12 .24 .22 .13
Avoidant .23 .19 .12 .19 .21 .12
Dependent .21 .15 .07 .15 .11 .04
Obsessive-Compulsive .22 .20 .10 .22 .17 .08
Passive-Aggressive .22 .21 .18 .18 .17 .12
Self-Defeating .20 .23 .14 .20 .24 .12
Histrionic .13 .15 .07 .14 .17 .07
Narcissistic .20 .22 .15 .19 .19 .12
Borderline .20 .25 .20 .23 .26 .17
Antisocial .10 .15 .12 .09 .12 .04
Sadistic .17 .16 .05 .14 .11 .00
Note. All correlations larger than .09 are significant at p < .001. Gender was partialed from all correlations, which were computed based on
N = 1371 to 1401.
Finally, we correlated the six parental-perception scales with the five

higher-order factors underlying the attachment and personality-disorder
measures. (These correlations are shown in Table 9.) The use of the
higher-order factors makes it easy to see at a glance which aspects of
attachment and personality disorders are associated with recollections of
nonoptimal parental caregiving. Maternal and paternal caregiving reports
were associated with the first (Insecurity) attachment factor and with
the first (Generalized Pathology) and third (Psychopathy) personality-
disorder factors. Defensive Emotional Style (arising from the attachment
measures) and Counter-Dependence (arising from the personality-disorder
measures) were weakly associated with reports of a parental caregiving
style that emphasized independence, and were also associated with a
tendency to idealize fathers.
DISCUSSION
The primary purpose of this study was to examine possible associations
between attachment styles and personality disorders. We hypothesized
that attachment style would be associated with all of the personality
disorders and share most of their underlying structure. Results revealed
that, first, attachment styles are related to personality disorders (in a
sample that is, presumably, mostly within the normal spectrum). Evi-
dence for convergence emerged whether categorical or dimensional
measures of personality disorders were used.
Second, information about the underlying common structure of the
two kinds of measures was obtained when we conducted a discriminant
analysis predicting attachment-style categories from all 13 personality
disorders. The set of personality-disorder scales formed a first function
that appeared to assess generalized pathology (disordered thoughts and
negative attitudes about others), and this function discriminated secures
from insecures with low self-esteem, particularly fearful individuals. The
second personality-disorder function appeared to assess overdepen-
dency, and distinguished preoccupied from dismissing individuals. The
third function, which was only marginally significant, appeared to assess
a tendency toward interpersonal aggression and callousness.
Third, consistent with previous theory (Bartholomew, 1990), attach-
ment patterns are arrayed in a two-dimensional space. When we concep-
tualize the dimensions in terms of relations among the measures, the
dimensions are secure versus fearful (or maximally insecure) and dismissing
Table 9
Correlations Between Higher-Order Factors and Perceptions of Childhood Relationships with Parents
Perceptions of Childhood Relationships with Parents

Mother Scales Father Scales
Higher-order Fostered Warm Was the Fostered Warm Was the
factor independence acceptance ideal parent independence acceptance ideal parent
Insecurity .20*** .26*** .20*** .22*** .27*** .18***
Defensive Emot. Style .09** .01 .04 .07* .05 .06*
Generalized Pathology .29*** .31*** .19*** .29*** .31*** .18***
Counter-Dependence .10*** .03 .01 .06* .01 .01
Psychopathy .19*** .24*** .16*** .17*** .19*** .08**
Note. Gender was partialed from all correlations, which were computed based on N = 1344 to 1401.
* p < .05. ** p < .01. *** p < .001.
versus preoccupied. We have named the first attachment-related factor

Insecurity and the second Defensive Emotional Style. These data repli-
cate the findings of other researchers who have uncovered a two- dimen-
sional structure underlying attachment measures (e.g., Brennan et al.,
1998).
Fourth, our data show that personality disorders are empirically ar-
rayed in a three-dimensional space. The first personality-disorder factor,
General Pathology, consisted of six scales assessing thought disorders
and troubled interactions with others. The second personality-disorder
factor, Counter-Dependence, consisted of the Schizoid, Dependent, and
Histrionic scales (the latter two correlated negatively with this factor).
The third personality-disorder factor, Psychopathy, includes personality-
disorder scales that tap aggressive behavior (Antisocial, Sadistic,
Passive-Aggressive, and Narcissistic). These three factors are very simi-
lar to the three discriminant functions that emerged in the analysis in
which the personality-disorder scales were used to discriminate attach-
ment styles.
Fifth, two of the three personality-disorder dimensions appear to
overlap very clearly with the attachment dimensions. The first personality-
disorder factor, General Pathology, coincides well with the first attach-
ment factor, Insecurity. The second personality-disorder factor,
Counter-Dependence, coincides well with the second attachment factor,
Defensive Emotional Style. The third personality-disorder factor, Psy-
chopathy, correlates only modestly with the first attachment factor, and
not at all with the second. Thus, most of the personality-disorder scales
appear to map onto the two-dimensional space represented by the attach-
ment measures.
The presence of the Psychopathy personality-disorder factor indicates,
however, that some kinds of personality disorders are not so highly
related to the two dimensions underlying the attachment measures. This
factor included the Antisocial, Sadistic, Passive-Aggressive, and Narcis-
sistic scales, which share items related to impulsive, remorseless, de-
structive behaviors. Psychopathy researchers have agreed on a two-factor
definition of psychopathy: The first factor is impulsive, antisocial, or
criminal behaviors; the second factor is manipulativeness, deceitfulness,
egocentrism, and lack of empathy for others suffering (Hare, 1993).
Some theorists point to social or environmental conditions that give rise
to psychopathy (Olwens, Block, & Radke-Yarrow, 1986, as cited in
Dutton, 1995). Others have demonstrated at least some evidence of

genetic transmission for psychopathy (Livesley et al., 1993).
Our three higher-order personality-disorder factors bear scant resem-
blance to the three DSM-III-R clusters. Two of the scales we included,
Sadistic and Self-Defeating, are not part of the three clusters. We there-
fore removed those two scales to make a cleaner comparison. The first
personality-disorder factor (Avoidant, Schizotypal, Paranoid, Obsessive-
Compulsive) correlated highly with two of the three Cluster A disorders
(Paranoid, Schizoid, Schizotypal). The third personality-disorder factor
(Antisocial, Borderline, Passive-Aggressive, and Narcissistic) correlated
highly with three of the four Cluster B disorders (Antisocial, Borderline,
Histrionic, Narcissistic). But the second personality-disorder factor
(Schizoid, Histrionic, and Dependent) correlated substantially with only
one of the disorders in Cluster C (Avoidant, Dependent, Obsessive-
Compulsive, Passive-Aggressive).
A secondary purpose of the present study was to explore connections
between attachment, personality disorders, and three family-background
factors: parental death, parental divorce, and current representations of
early childhood relationships with parents. Both attachment style and
personality disorders are associated with parental divorce, but only
attachment style is associated with parental death (i.e., those with dis-
missive, or emotionally defensive styles, more often reported the loss of
one or both parents). Although we cannot know how our participants
were actually treated in childhood, attachment insecurity and two of the
three personality-disorder factors were related to current representations
of childhood relationships with parents. The two personality-disorder
factors related to parental treatment were most closely related to attach-
ment insecurity (Generalized Pathology and Psychopathy), rather than
to the other attachment dimension, Defensive Emotional Style.
Do Attachment Styles and Personality Disorders

Share Common Origins?
Based on our findings, we suggest that patterns of insecure attachment
overlap with patterns of disordered personality, many of which are
moderately related to each other. Thus, it would be parsimonious to
consider the idea that the same environmental conditions that give rise to
insecure attachment also contribute to at least some personality disorders.
Of course, it is likely that such environmental conditions combine with,
or interact with, inborn temperaments (e.g., behavioral inhibition). In

addition, cultural variations in the extent to which particular traits (e.g.,
independence, eccentricity) are valued are also likely to result in cross-
cultural differences in the expression of personality. Neither option,
however, is incompatible with the idea that the quality of ones early
attachment to caregivers accounts for some of the variance in abnormal,
or maladaptive, personality functioning. In fact, at first glance, it is
astonishing that such connections have been largely overlooked. Such
oversight may be due to discrepancies in method and choice of research
subjects in the domains of attachment, on the one hand, and personality
disorders, on the other. Ainsworths and others attachment research was
grounded in orthodox developmental methods: behavioral observations
of young children and their parents. In contrast, the study of personality
disorders has traditionally relied on the retrospective reports of disturbed
adults in therapy.
That the two domains would eventually converge on the same disor-
dered relationship patterns is therefore not surprising and, we think,
would not have surprised John Bowlby. Throughout his career, Bowlby
theorized about the mechanisms likely to account for adolescent and
adult psychopathology (1944, 1973, 1977, 1980, 1988). Particularly
noteworthy is Bowlbys distinction between feeling alarmed and feel-
ing anxious. Drawing on a military analogy, Bowlby (1973) described
how an armys safety depends:
not only on its defending itself against direct attack but also on its
maintaining open communication with its base. Any military com-
mander who fails to give as much attention to his base and lines of
communication as to his main front soon finds himself defeated. . .
. [A similar situation] holds between an individual and his attach-
ment figure. Each party is autonomous. Given basic trust the ar-
rangement can work well. But any possibility of defection by the
attachment figure can give rise to acute anxiety in the attached. And
should he be experiencing alarm from another source at the same
time, it is evident that he is likely to feel the most intense fear. (pp.
9394)
Theoretically, environmental threats should provoke withdrawal, freez-
ing, or escape behavior. In a well-functioning goal-corrected partnership
between child and parent, such escape usually entails escaping to an
attachment figure. The failure, however, to establish a well-functioning
partnership with a caregiver means that an infant may develop the chronic
perception that help is unavailable when needed, or, at best, is inconsis-
tently available. In other words, distress may result from either perceived
threats from the environment or perceived lack of support from ones
caregivers. According to Bowlby (1973), the combination of both, expe-
rienced often enough throughout the course of development, is likely to
produce the most acute forms of psychopathology.
It is possible to delineate the four adult attachment patterns identified
by Bartholomew (1990) in terms of Bowlbys distinction between per-
ceived support from attachment figures and perceived threat from the
environment. Secure individuals trust their attachment figures and per-
ceive little environmental threat; as a result, confident of their attachment
figures reliability, secures can defend themselves against environmental
threats, and hence are able to process emotions in a fluid and nondefen-
sive manner and remain the least troubled by personality disorders.
Dismissing individuals, chronically lacking support from attachment
figures, may, as a result, habitually deny or dismiss environmental
threats. They may therefore have a higher threshold for experiencing
negative emotions or perceiving attachment needs, and thus exhibit what
Bowlby called compulsive self-reliance. Preoccupied individuals, wary
following a history of inconsistent support from caregivers, are likely to
have a lower threshold for perceiving environmental threat. Such a low
threshold for experiencing distress, in combination with mixed percep-
tions of support from caregivers, is likely to contribute to frequent
activation of the attachment system, with all the concomitant distress and
anger such activation provokes. Hence, they are likely to suffer from
compulsive care-seeking, dependency, and Histrionic personality disor-
der. Lastly, fearful individuals, who may be the adult versions of disor-
ganized/disoriented infants, appear to be the most troubled in terms of
personality disorders. New evidence has emerged that disorganized/dis-
oriented infants experience frightening behavior from attachment fig-
ures; their attachment figures, in turn, are more likely to reveal
themselves in the Adult Attachment Interview as Unresolved with
respect to loss or trauma (Lyons-Ruth & Jacobvitz, in press).
Limitations and Directions for Future Research

One obvious limitation of our research is its use of a nonclinical sample.
To some extent, this limitation turns on whether one accepts a purely
categorical (vs. dimensional) understanding of personality disorders. If,

as we do, one assumes that personality disorders can be arrayed on
continua, then our results ought to generalize to clinical populations. It
would be desirable to conduct a similar study in a large enough (i.e.,
general population) sample that would allow a credible test of the
categories versus dimensions question. If, on the other hand, one assumes
that personality disorders reflect true taxons, then our results may not
generalize well to clinical populations. In particular, our measure of
personality disorders suffers from an overinclusion bias. Examining all
13 personality disorders together, three-fourths of the sample had at least
one personality disorder. This fact is consistent with the PDQ-Rs posi-
tive bias (cf. Moldin et al., 1994).
The prevalence of individual personality disorders in our sample is
generally consistent with previous research, but four scales prevalences
were higher than is generally found in clinical samples: Paranoid, His-
trionic, Borderline, and Obsessive-Compulsive. Note that, when standard
scoring methods were used, these four disorders occurred relatively
frequently in the two college student studies by Johnson and colleagues,
and three of the four (i.e., all but Borderline) occurred relatively fre-
quently in the medical student study by Maffei et al. (1995). Why the
higher prevalence rates for these disorders? We draw the same conclu-
sions as did Johnson and Maffei and their colleagues. First, our samples
higher scores on the Paranoid and Obsessive-Compulsive scales may
reflect high intelligence and work standards in college student popula-
tions. Second, several items on the Borderline and Histrionic scales
appear related to patterns of thoughts, feelings, and behavior more typical
of adolescence than adulthood. Our young sample (median age = 18)
may be relatively more concerned than older adults with appearance and
gossip; may be more moody and impulsive, particularly with regard to
behaviors involving sex, drugs, and alcohol; and may experience en-
hanced difficulties stemming from exploring conflicting beliefs, social
networks, and occupational goals. Not surprisingly, researchers have
reported higher rates of personality disorders in adolescents than in adults
(Bernstein et al., 1993; Johnson et al., 1996; Maier, Lichtermann, Klingler,
Heun, & Hallmayer, 1992; Swartz, Blazer, George, & Winfield, 1990).
A second, related limitation of our research is its reliance on a
self-administered measure of personality disorders, which requires a fair
degree of insight on the part of respondents. Although the PDQ-R is
reliably associated with clinicians classifications (49% to 58%; Hyler et
al., 1990), the overlap is not perfect. When Johnson et al.s (1996) more
stringent scoring criteria are used, however, prevalence rates of person-
ality disorders assessed with the PDQ-R more closely resemble preva-
lence rates reported for nonclinical samples using structured clinical
interviews (Maier et al., 1992; Coryell & Zimmerman, 1989). Therefore,
it would be desirable to test the associations between attachment styles
and personality disorders assessed via structured clinical interviews and
self-administered measures in the same sample.
A third limitation of the current research is its cross-sectional nature.
Longitudinal research is needed to tease apart causal relations between
attachment and personality disorders, on the one hand, and family
variables, on the other. It is unlikely, however, that even with a longitu-
dinal design one could determine the ontological primacy of attachment
styles versus personality disorders. Although attachment styles are ob-
servable in the first year of life, only a handful of disorders resembling
adult personality disorders are detectable from behavioral observations
of young children. Two examples come to mind. First, Separation Anxi-
ety Disorder in childhood appears similar in many respects to Dependent
personality disorder in adulthood, and to anxious (or preoccupied) at-
tachment in both childhood and adulthood. Separation Anxiety Disorder
is characterized by excessive anxiety concerning separation from the
home or from those to whom the person is attached . . . beyond what is
expected for the individuals developmental level (APA, 1994, p. 110).
Second, childhood disruptive behavior disorders seem similar in many
ways to psychopathy in adulthood and may be associated with the
Disoriented/Disorganized infant attachment pattern (Lyons-Ruth,
1996). Two childhood disruptive behavior disorders are worth mention-
ing. Conduct Disorder is a repetitive and persistent pattern of behavior
in which the basic rights of others or major age-appropriate societal
norms or rules are violated (APA, 1994, p. 85). This disorder includes
aggression directed toward other people or animals, property damage
(e.g., fire-setting, window-smashing), theft, deceitfulness, or other seri-
ous violations of rules. Oppositional Defiant Disorder, which is often a
precursor to Conduct Disorder, includes negativistic, defiant, disobedi-
ent, and hostile behavior [directed] toward authority figures (APA,
1994, p. 91), but excludes the more serious aggression and rule violations
of Conduct Disorder. Other personality disorders cannot be assessed
behaviorally in very young children, perhaps because a significant cog-
nitive component is necessary (e.g., Paranoid, Obsessive-Compulsive, or
Schizotypal personality disorders). (See also Zeanah, 1996, for a review

of infant attachment disorders.)
In summary, there is a fair degree of overlap between attachment styles
and personality disorders, and both sets of variables are associated with
family variables. Future research should focus on elucidating the devel-
opmental antecedents of personality disorders to determine which ones
are shared by patterns of insecure attachment, and which are separate.
Given the relative lack of understanding of the etiology of personality
disorders, we hope that clinicians and researchers studying personality
disorders will consider attachment theory as a framework for exploring
possible origins of personality disorders. In addition to investigating
developmental antecedents of personality disorders, future research
should compare attachment styles and personality disorders in their
ability to predict patterns of intrapersonal processes (e.g., affect-regulation
and coping with distress) as well as interpersonal functioning over time.
We hope that the present study, connecting attachment styles and person-
ality disorders with each other and with family variables, is a useful first
step.
REFERENCES
Ainsworth, M. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment:
A psychological study of the Strange Situation. Hillsdale, NJ: Erlbaum.
American Psychiatric Association. (1987). Diagnostic and statistical manual of mental
disorders (Rev. 3d ed.). Washington, DC: Author.
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental
disorders (4th ed.). Washington, DC: Author.
Aro, H. M., & Palosaari, U. K. (1992). Parental divorce, adolescence, and transition to
young adulthood: A follow-up study. American Journal of Orthopsychiatry, 62,
421429.
Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspective. Journal of
Social and Personal Relationships, 7, 147178.
Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A
test of a four-category model. Journal of Personality and Social Psychology, 61,
226244.
Belsky, J., & Nezworski, T. (1988). Clinical implications of attachment. In J. Belsky &
T. Nezworski (Eds.), Clinical implications of attachment (pp. 317). Hillsdale, NJ:
Erlbaum.
Ben-Amos, B. (1992). Depression and conduct disorders in children and adolescents: A
review of the literature. Bulletin of the Menninger Clinic, 56, 188208.
Benjamin, L. S. (19931996). Interpersonal diagnosis and treatment of personality
disorders (2d ed.). New York: Guilford Press.
Bernstein, D. P., et al. (1993). Prevalence and stability of the DSM-III-R personality
disorders in a community-based survey of adolescents. American Journal of Psychia-
try, 150, 12371243.
Bornstein, R. F. (1992). The dependent personality: Developmental, social, and clinical
perspectives. Psychological Bulletin, 112, 323.
Bowlby, J. (1944). Forty-four juvenile thieves: Their characters and home-life. Interna-
tional Journal of Psycho-Analysis, 25, 1953, 107127.
Bowlby, J. (1969/1982). Attachment., Vol. 1 of Attachment and loss (2nd ed.). New York:
Basic Books.
Bowlby, J. (1973). Separation. Vol. 2 of Attachment and loss. New York: Basic Books.
Bowlby, J. (1977). The making and breaking of affectional bonds, 1: Aetiology and
psychopathology in the light of attachment theory. British Journal of Psychiatry, 130,
201210.
Bowlby, J. (1980). Loss. Vol. 3 of Attachment and loss. New York: Basic Books.
Bowlby, J. (1988). A secure base: Parentchild attachment and healthy human develop-
ment. New York: Basic Books.
Brennan, K. A., & Bosson, J. K. (in press). Attachment-style differences in attitudes
toward and reactions to feedback from romantic partners: An exploration of the
relational bases of self-esteem. Personality and Social Psychology Bulletin.
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult
attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.),
Attachment theory and close relationships (pp. 4676). New York: Guilford Press.
Brennan, K. A., & Morris, K. A. (1997). Attachment styles, self-esteem, and patterns of
seeking feedback from romantic partners. Personality and Social Psychology Bulletin,
23, 2331.
Brennan, K. A., & Shaver, P. R. (1995). Dimensions of adult attachment, affect regulation,
and romantic relationship functioning. Personality and Social Psychology Bulletin,
21, 267283.
Brennan, K. A., Shaver, P. R., & Tobey, A. E. (1991). Attachment styles, gender, and
parental problem drinking. Journal of Social and Personal Relationships, 8, 451466.
Carnelley, K., & Janoff-Bulman, R. (1992). Optimism about love relationships: General
vs. specific lessons from ones personal experience. Journal of Social and Personal
Relationships, 9, 520.
Carver, C. S. (1997). Adult attachment and personality: Converging evidence and a new
measure. Personality and Social Psychology Bulletin, 23, 865883.
Collins, N. L., & Read., S. J. (1990). Adult attachment, working models, and relationship
quality in dating couples. Journal of Personality and Social Psychology, 58, 644663.
Coryell, W. H., & Zimmerman, M. (1989). Personality disorder in the families of
depressed, schizophrenic, and never-ill probands. American Journal of Psychiatry,
146, 496502.
Crittenden, P. M. (1988). Relationships at risk. In J. Belsky & T. Nezworski (Eds.),
Clinical implications of attachment (pp. 136174). Hillsdale, NJ: Erlbaum.
Crittenden, P. M., Parridge, M. F., & Clauseen, A. H. (1991). Family patterns of
relationship in normative and dysfunctional families. Development and Psychopathol-
ogy, 3, 491512.
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,
1, 118128.
Dahl, A. A. (1993). The personality disorders: A critical review of family, twin, and
adoption studies. Journal of Personality Disorders, 1, 8699.
Duggan, E. S., & Brennan, K. A. (1994). Social avoidance and its relation to
Bartholomews adult attachment typology. Journal of Social and Personal Relation-
ships, 11, 147153.
Dutton, D. G. (1994). The origin and structure of the abusive personality. Journal of
Personality Disorders, 8, 181191.
Dutton, D. G. (1995). The batterer: A psychological profile. New York: Basic Books.
Ekselius, L., Lindstrom, E., von Knorring, L., Bodlund, O., & Kullgren, G. (1994). A
principal component analysis of the DSM-III-R Axis II personality disorders. Journal
of Personality Disorders, 8, 140148.
Ekselius, L., et al. (1996). Sex differences in DSM-III-R, Axis II-Personality disorders.
Personality and Individual Differences, 20, 457461.
Elicker, J., Englund, M., & Sroufe, L. A. (1992). Predicting peer competence and peer
relationships in childhood from early parentchild relationships. In R. Parke & G.
Ladd (Eds.), Familypeer relationships: Modes of linkage. Hillsdale, NJ: Erlbaum.
Epstein, S. (1983). The mother-father-peer scale. Unpublished manuscript, University
of Massachusetts, Amherst.
Eysenck, H. J. (1970). The structure of human personality. London: Methuen.
Feeney, J. A., & Noller, P. (1990). Attachment style as a predictor of adult romantic
relationships. Journal of Personality and Social Psychology, 58, 281291.
Feeney, J. A., & Noller, P. (1996). Adult attachment. Thousand Oaks, CA: Sage.
Fuendeling, J. M. (in press). Affect regulation as a stylistic process within adult attach-
ment. Journal of Social and Personal Relationships.
Gacano, C. B., Meloy, J. R., & Berg, J. L. (1992). Object relations, defensive operations,
and affective states in narcissistic borderline and antisocial personality disorder.
Journal of Personality Assessment, 59, 3249.
George, C., Kaplan, N., & Main, M. (1984/1985/1996). Adult Attachment Interview.
Unpublished manuscript, Department of Psychology, University of California,
Berkeley.
Glickhauf-Hughes, C., & Wells, M. (1991). Differential diagnosis of the masochistic
personality. Psychoanalysis and Psychotherapy, 9, 167176.
Golomb, M., Fava, M., Abraham, M., & Rosenbaum, J. F. (1995). Gender differences in
personality disorders. American Journal of Psychiatry, 152, 579582.
Griffin, D., & Bartholomew, K. (1994). Models of the self and other: Fundamental
dimensions underlying measures of adult attachment. Journal of Personality and
Social Psychology, 67, 430445.
Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process.
Journal of Personality and Social Psychology, 52, 511524.
Hazan, C., & Shaver, P. R. (1990). Love and work: An attachment-theoretical perspective.
Journal of Personality and Social Psychology, 59, 270280.
Heard, D. H., & Lake, B. (1986). The attachment dynamic in adult life. British Journal
of Psychiatry, 149, 430438.
Hyler, S. E., Lyons, M., Rieder, R. O., Young, L., Williams, J.B.W., & Spitzer, R. L.
(1994). The factor structure of self-report DSM-III Axis II symptoms and their
relationship to clinicians ratings. American Journal of Psychiatry, 147, 751757.
Hyler, S. E., & Rieder, R. O. (1987). PDQ-R: Personality Questionnaire. Unpublished
manuscript, New York State Psychiatric Institute, New York City.
Hyler, S. E., Rieder, R. O., Spitzer, R. L., & Williams, J.B.W. (1982). The Personality
Diagnostic Questionnaire. New York: New York State Psychiatric Institute.
Hyler, S. E., Rieder, R. O., Williams, J.B.W., Spitzer, R. L., Hendler, J., & Lyons, M.
(1988). The Personality Diagnostic Questionnaire: Development and preliminary
results. Journal of Personality Disorders, 2, 229237.
Hyler, S. E., Rieder, R. O., Williams, J. B. W., Spitzer, R. L., Lyons, M., & Hendler, J.
(1989). A comparison of clinical and self-report diagnoses of DSM-III personality
disorders in 552 patients. Comprehensive Psychiatry, 30, 170178.
Hyler, S. E., Skodol, A. E., Kellman, H. D., Oldham, J. M., & Rosnick, L. (1990). Validity
of the Personality Diagnostic Questionnaire-Revised: Comparison with two struc-
tured interviews. American Journal of Psychiatry, 147, 10431048.
Hyler, S. E., Skodol, A. E., Oldham, J. M., Kellman, H. D., & Doidge, N. (1992). Validity
of the Personality Diagnostic Questionnaire-Revised: A replication in an outpatient
sample. Comprehensive Psychiatry, 33, 7377.
Johnson, J. G., Bornstein, R. F., & Sherman, M. F. (1996). A modified scoring algorithm
for the PDQ-R: Psychiatric symptomatology and substance use in adolescents with
personality disorders. Educational and Psychological Measurement, 56, 7689.
Kirkpatrick, L. A., & Davis, K. E. (1994). Attachment style, gender, and relationship
stability: A longitudinal analysis. Journal of Personality and Social Psychology, 66,
502512.
Kirkpatrick, L. A., & Hazan, C. (1994). Attachment styles and close relationships: A
four-year prospective study. Personal Relationships, 1, 123142.
Laport, L., & Guttman, H. (1996). Traumatic childhood experiences as risk factors for
borderline and other personality disorders. Journal of Personality Disorders, 10,
247259.
Livesley, W. J., Jackson, D. N., & Schroeder, M. L. (1992). Factorial structure of traits
delineating personality disorders in clinical and general population samples. Journal
of Abnormal Psychology, 101, 432440.
Livesley, W. J., Jang, K. L., Jackson, D. N., & Vernon, P. A. (1993). Genetic and
environmental contributions to dimensions of personality disorder. American Journal
of Psychiatry, 150, 18261831.
Livesley, W. J., Schroeder, M. L., & Jackson, D. N. (1990). Dependent personality
disorder and attachment problems. Journal of Personality Disorders, 4, 131140.
Lyons-Ruth, K. (1996). Attachment relationships among children with aggressive behav-
ior problems: The role of disorganized early attachment patterns. Journal of Consult-
ing and Clinical Psychology, 64, 6473.
Lyons-Ruth, K., & Jacobvitz, D. (in press). Attachment disorganization: Unresolved loss,
relational violence, and lapses in behavioral and attentional strategies. In J. Cassidy
& P. R. Shaver (Eds.), Handbook of attachment theory and research. New York:
Guilford Press.
Maffei, C., Fossati, A., Lingiardi, V., Madeddu, F., Borellini, C., & Petrachi, M. (1995).
Personality maladjustment, defenses, and psychopathological symptoms in nonclini-
cal subjects. Journal of Personality Disorders, 9, 330345.
Maier, W., Lichtermann, D., Klingler, T., Heun, R., & Hallmayer, J. (1992). Prevalences
of personality disorders (DSM-III-R) in the community. Journal of Personality
Disorders, 6, 187196.
Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adult-
hood: A move to the level of representation. In I. Bretherton & E. Waters (Eds.),
Growing points in attachment theory and research. Monographs of the Society for
Research in Child Development, 50 (12, Serial No. 209), 66106.
Main, M., & Solomon, J. (1990). Procedures for identifying insecure-disorganized/dis-
oriented infants: Procedures, findings, and implications for the classification of
behavior. In M. Greenberg, D. Cicchetti, & M. Cummings (Eds.), Attachment in the
preschool years: Theory, research, and intervention (pp. 121160). Chicago: Univer-
sity of Chicago Press.
Mattick, R. P., & Newman, C. R. (1991). Social phobia and avoidant personality disorder.
International Review of Psychiatry, 3, 163173.
Moldin, S. O., Rice, J. P., Erlenmeyer-Kimling, L., & Squires-Wheeler, E. (1994). Latent
structure of DSM-III-R Axis II psychopathology in a normal sample. Journal of
Abnormal Psychology, 103, 259266.
Norden, K. A., Klein, D. N., Donaldson, S. K., Pepper, C. M., & Klein, L. M. (1995).
Reports of the early home environment in DSM-III-R personality disorders. Journal
of Personality Disorders, 9, 213223.
Olwens, D., Block, J., & Radke-Yarrow, M. (1986). Development of antisocial behavior
and delinquency. Clinical Psychology Review, 10, 141.
Paris, J. (1993). Personality disorders: A biopsychosocial model. Journal of Personality
Patrick, M., Hobson, R. P., Castle, D., Howard, R., & Maughan, B. (1994). Personality
disorder and the mental representation of early social experience. Development and
Psychopathology, 4, 375388.
Ricks, M. H. (1985). The social transmission of parental behavior: Attachment across
generations. In I. Bretherton & E. Waters (Eds.), Growing points in attachment theory
and research. Monographs of the Society for Research in Child Development, 50
(12, Serial No. 209), 211227.
Riso, L. P., Klein, D. N., Ferro, T., Kasch, K. L., Pepper, C. M., Schwartz, J. E., &
Aronson, T. A. (1996). Understanding the comorbidity between early-onset dysthymia
and Cluster B personality disorders: A family study. American Journal of Psychiatry,
153, 900906.
Sack, A., Sperling, M. B., Fagen, G., & Foelsch, P. (1996). Attachment style, history, and
behavioral contrasts for a borderline and normal sample. Journal of Personality
Scharfe, E., & Bartholomew, K. (1994). Reliability and stability of adult attachment
patterns. Personal Relationships, 1, 2343.
Segal, B. M. (1988). A borderline style of functioning: The role of family, society, and
heredity: An overview. Child Psychiatry and Human Development, 18, 219238.
Shaver, P. R., Belsky, J., & Brennan, K. A. (1997). Comparing measures of adult
attachment: An examination of interview and self-report methods. Unpublished
manuscript, Department of Psychology, University of California, Davis.
Shaver, P. R., & Brennan, K. A. (1992). Attachment styles and the big five personality
traits: Their connections with each other and with romantic relationship outcomes.
Personality and Social Psychology Bulletin, 18, 536545.
Shaver, P. R., & Clark, C. L. (1994). The psychodynamics of adult romantic attachment.
In J. M. Masling & R. F. Bornstein (Eds.), Empirical perspectives on object relations
theory (Vol. 5, pp. 105156). Washington, DC: American Psychological Association.
Shaver, P. R., & Hazan, C. (1993). Adult romantic attachment: Theory and evidence. In
D. Perlman & W. Jones (Eds.), Advances in personal relationships (Vol. 4, pp. 2970).
London: Jessica Kingsley.
Shaver, P. R., Papalia, D., Clark, C. L., Koski, L. R., Tidwell, M. C., & Nalbone, D.
(1996). Androgyny and attachment security: Two related models of optimal person-
ality. Personality and Social Psychology Bulletin, 22, 582597.
Sheldon, A. E., & West, M. (1990). Attachment pathology and low social skills in avoidant
personality disorder: An exploratory study. Canadian Journal of Psychiatry, 35,
596599.
Simpson, J. A. (1990). Influence of attachment styles on romantic relationships. Journal
of Personality and Social Psychology, 59, 971980.
Simpson, J. A., & Rholes, W. S. (Eds.). (in press). Attachment and close relationships.
New York: Guilford Press.
Sroufe, L. A., & Waters, E. (1977). Attachment as an organizational construct. Child
Development, 48, 11841199.
Stalker, C. A., & Davies, F. (1995). Attachment organization and adaptation in sexually
abused women. Canadian Journal of Psychiatry, 40, 234240.
Swartz, M., Blazer, D., George, L., & Winfield, I. (1990). Estimating the prevalence of
borderline personality disorder in the community. Journal of Personality Disorders,
4, 257272.
Torgersen, S., & Alnaes, R. (1992). Differential perception of parental bonding in
schizotypal and borderline personality disorder patients. Comprehensive Psychiatry,
33, 3438.
Torgersen, S., Skre, I., Onstad, S., Edvardsen, J., & Kringlen, E. (1993). The psychometric-
genetic structure of DSM-III-R personality disorder criteria. Journal of Personality
Vaillant, G. E. (1987). A developmental view of old and new perspectives of personality
disorders. Journal of Personality Disorders, 1, 146156.
Van der Kolk, B. A., Perry, J. C., & Herman, J. L. (1991). Childhood origins of
self-destructive behavior. American Journal of Psychiatry, 148, 16651671.
Watson, D., & Clark, L. A. (1984). Negative affectivity: The disposition to experience
aversive emotional states. Psychological Bulletin, 96, 465498.
West, M., & Keller, A. (1994). Psychotherapy strategies for insecure attachment in
personality disorders. In M. R. Sperling & W. H. Berman (Eds.), Attachment in adults:
Clinical and developmental perspectives (pp. 313330). New York: Guilford Press.
West, M., Rose, S., & Sheldon-Keller, A. (1994). Assessment of patterns of insecure
attachment in adults and application to dependent and schizoid personality disorders.
Journal of Personality Disorders, 8, 249256.
West, M., & Sheldon, A. E. (1988). Classification of pathological attachment patterns in
adults. Journal of Personality Disorders, 2, 153159.
West, M., & Sheldon-Keller, A. E. (1994). Patterns of relating: An adult attachment
perspective. New York: Guilford Press.
Whitely, J. S. (1994). Attachment, loss, and the space between: Personality disorder in
the therapeutic community. Group Analysis, 27, 359387.
Widiger, T. A., & Frances, A. (1985). The DSM-III personality disorders: Perspectives
from psychology. Archives of General Psychiatry, 42, 615623.
Wiggins, J. S., & Pincus, A. L. (1994). Personality structure and the structure of
personality disorders. In P. T. Costa & H. A. Widiger (Eds.), Personality disorders and
the five-factor model of personality (pp. 7393). Washington, DC: American Psycho-
logical Association.
Williams, D., & Schill, T. (1993). Attachment histories for people with characteristics of
self-defeating personality disorder. Psychological Reports, 73(3, pt. 2), 12321234.
Zeanah, C. H. (1996). Beyond insecurity: A reconceptualization of attachment disorders
of infancy. Journal of Consulting and Clinical Psychology, 64, 4252.

Attachment Styles and Personality Disord PDF

Uploaded by

Copyright:

Available Formats

Attachment Styles and Personality Disord 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

Attachment Styles and Personality Disord PDF

Uploaded by

Copyright:

Available Formats

Attachment Styles and Personality

Disorders: Their Connections to Each Other

ABSTRACT Attachment theory was explored as a means of understanding

Journal of Personality 66:5, October 1998.

According to the most recent Diagnostic and Statistical Manual of the

as socially avoidant, fearful avoidants scored lower in self-esteem

1. As predicted, Bartholomews secure and preoccupied types were closely akin to

the other attachment types to report having grown up with problem-

Research on attachment and personality disorders. In the clinical lit-

attachments with one or both parents and the Self-Defeating personality

likely to stress self-sufficiency and downplay attachment needs. As men-

The Present Study

Passive-Aggressive scale describes a person who uses passive means to accom-

prevalence of various personality-disorder diagnoses ranged from 5.2%

Frequencies (Percentages) of Each

According to Johnson et al., individuals cannot be classified as having a

Gender differences. The percentage of males with at least one person-

Intercorrelations of the scales. Intercorrelations of the personality-dis-

Connections Between Attachment Styles and

Tests using personality-disorder categories. The association of cate-

Tests using personality-disorder dimensions Personality-disorder rat-

Frequencies of Each Personality Disorder

gender in a 2 4 multivariate analysis of variance (MANOVA). The

Discriminating among attachment-style categories using personality-

Personality-Disorders Attachment-Style Category

Personality Disorder Rating Function 1 Function 2 Function 3

Dependent .31 .77* .16

Passive-Aggressive .43 .24 .44*

Attachment Style Function 1 Function 2 Function 3

31.27% of the variance, respectively; Function 3 accounted for the

some sort of distortion of reality along with negativity about others

Underlying structure of the attachment-style ratings. To d e t e r m i n e

Underlying structure of the personality-disorder dimensions. To deter-

analysis. Avoidant, Schizotypal, Paranoid, Self-Defeating, Obsessive-

6. Due to the moderate intercorrelations among the disorders scales, we decided to

Associations between the attachment and personality-disorder factors.

three family-background variables: parental death, parental divorce, and

Association of Personality Disorders and

Parental divorce. The associations of parental marital status (divorced

Parental death. The associations of parental mortality status (one or

Perceptions of childhood relationships with parents. Mean scores on

(The attachment-group means for each of the six mother-father scales

Relationship with Father

Perceptions of Childhood Relationships with Parents

Finally, we correlated the six parental-perception scales with the five

Perceptions of Childhood Relationships with Parents

versus preoccupied. We have named the first attachment-related factor

Dutton, 1995). Others have demonstrated at least some evidence of

Do Attachment Styles and Personality Disorders

or interact with, inborn temperaments (e.g., behavioral inhibition). In

Limitations and Directions for Future Research

categorical (vs. dimensional) understanding of personality disorders. If,

Schizotypal personality disorders). (See also Zeanah, 1996, for a review

You might also like