Depression and Physical Illnes PDF
Depression and Physical Illnes PDF
Depression and Physical Illnes PDF
Depression and physical illness are related intimately. Depressed mood is thought to contribute
to the development and progression of some illnesses, while physical illness can, in turn,
increase the risk of depression. This book provides a critical overview of the evidence linking
depression with several major health conditions, including coronary heart disease, diabetes,
cancer, chronic pain, disability, chronic fatigue and obesity. The book also explores the
biological and behavioural processes underlying the association, discussing the role of
neuroendocrine, immunological and inflammatory pathways and the relationship between
depression and health behaviours such as smoking, physical activity and adherence to medical
advice. It combines a thorough analysis of the clinical, biological and epidemiological data with
guidance to health professionals and patients on how to manage depression in people suffering
from physical illness, pointing the way to an integrated approach to healthcare.
Edited by
Andrew Steptoe
Department of Epidemiology and Public Health
University College London
London, UK
cambridge university press
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo
Cambridge University Press has no responsibility for the persistence or accuracy of urls
for external or third-party internet websites referred to in this publication, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
Every effort has been made in preparing this publication to provide accurate and
up-to-date information that is in accord with accepted standards and practice at the time
of publication. Although any case histories are drawn from actual cases, every effort has
been made to disguise the identities of the individuals involved. The authors, editors
and publishers can make no warranties that the information contained herein is totally
free fromerror, not least because clinical standards are constantly changing through
research and regulation. The authors, editors and publishers therefore disclaim all
liability for direct or consequential damages resulting from the use of material
contained in this publication. Readers are strongly advised to pay careful attention to
information provided by the manufacturers of any drugs or equipment that they plan to
use.
Contents
Part 4 Conclusions
Index 409
Contributors
The past few years have witnessed an upsurge in work on depression and physical
illness. This has been coupled with renewed interest in the biological processes
underpinning depression and exhortation of physicians to recognise and treat
depression in their patients. There are many reasons why the study of depression
and physical illness is important. First, there is growing evidence that depression
and depressive symptoms are determinants of some types of physical pathology.
The investigation of depression therefore contributes to knowledge about factors
promoting disease development and illness progression in people with existing dis-
orders. Second, depression is crucial to the everyday functioning and healthcare
utilisation of people suffering from physical illnesses. Severe depression is a good
indicator of whether daily functioning is likely to be impaired and whether there
is an increased risk of suicide. Third, self-management is a central feature of many
clinical conditions and disabilities. If depression impairs people’s engagement in
appropriate self-care, then the burden of distress and disability will be increased and
the effectiveness of medical management may be compromised. Fourth, treating
comorbid depression is likely to improve the wellbeing and quality of life of patients
with physical illnesses, and this may have an impact on the severity and progression
of underlying pathology. Bearing in mind the global burden of disease and pre-
dictions concerning the future high demands on healthcare systems attributable to
depression [1], there is a pressing need to understand the relationship of depression
with physical illness and disability.
This book addresses the issue of depression and physical illness from a num-
ber of perspectives, including disease aetiology, patient care, adaptation to illness,
underlying biology, and behavioural and lifestyle factors. The wealth of scholarship
presented attests to the vigour with which the topic is being addressed by researchers
and clinicians. The book is divided into four parts. Part 1 sets the scene, with dis-
cussion of the occurrence of depression in medically ill patients and the influence of
sociodemographic and psychosocial factors on clinical depression and subclinical
depressive symptoms. Part 2 is devoted to chapters that assess the relationship
xii Preface
Andrew Steptoe
REFERENCE
Introduction
Psychiatric disorders of all types are more common in people with physical illness
compared with the general population. Depression is the most common disorder,
accounting for approximately 50% of psychopathology in the medically ill, with
the remainder made up of various anxiety disorders and mixed subsyndromal
symptoms of anxiety and depression. The importance of depression in the medically
ill lies in its adverse effect on outcome, most notably health-related quality of life,
combined with the fact that it is rarely detected and treated adequately in people
who have physical illness.
The prevalence of depression in medically ill populations varies greatly according
to the definition of depression and the type of measure used [1–4]. Variation in the
definition and measuring instrument are the main reasons for the large variation
in the prevalence figures quoted in the literature [5]. A higher prevalence of depres-
sion has been reported by studies that have used a self-administered questionnaire
compared with those that used standardised research interviews administered by
a trained interviewer [6]. The prevalence of depression also varies according to
sociodemographic characteristics of the sample and the location of the survey
(out-patient, in-patient, community) [3]. Only after all of these factors have been
taken into account is it possible to assess whether the prevalence of depression varies
according to type of medical disorder, its chronicity or severity [2,3].
In this chapter we indicate the ways in which different definitions and differ-
ent modes of measurement used in previous research can affect the prevalence
of depression. We examine the prevalence of depression in different groups (in-
patients, out-patients, population-based samples) and review briefly the few stud-
ies that have examined the incidence of depression in the medically ill. Finally, we
mention the effect of depression on outcome – assessed in terms of health-related
quality of life and healthcare costs. This topic is discussed more fully in later
chapters.
different from the measurement of blood pressure – although there is an upper limit
of normal blood pressure, above which treatment is recommended, not everyone
with raised blood pressure will develop complications.
In general, the two methods of assessment can reach broad agreement as long as
the cut-off score on self-administered questionnaires is adjusted for the physically
ill population being screened (see below) [10,11]. In practice, the choice of method
may reflect the purpose for the study. A self-administered questionnaire is required
to screen a large population of physically ill patients [12] and may be used as the
first stage of a two-phase survey, which includes research interviews to determine
the actual cases of depressive disorder [13]. The latter method is also required to
identify cases for inclusion in a treatment trial.
Self-administered questionnaires
The alternative approach views depression as a continuum without any clear cut-off
separating depressed and non-depressed populations. This approach is probably a
more accurate representation of the true picture [12].
Most studies using this approach have employed self-rated questionnaires that
were originally designed for use in the general population, or in psychiatric popu-
lations, but not for use in the medically ill. These assessments (e.g. Beck Depression
6 Depression and physical illness
Depression Scale (GDS) [24,25]. The HADS is probably the most commonly used, as
it excludes many of the physical symptoms of depression (loss of appetite and weight,
sleep disturbance, fatigue, decreased sexual drive, poor concentration, psychomotor
changes) included in more conventional depression questionnaires [26–28]. Cut-
off scores were determined that best identified possible cases (scores of 8 or above)
or probable cases (scores of 11 or above) on the depression and anxiety subscales,
although these may need to be revalidated for the population in question [27].
Finally, it is possible to use a well-known self-administered questionnaire but with
a modified cut-off score. When the General Health Questionnaire (GHQ) (28-item
version) score was used in neurological patients, the cut-off score had to be raised
from the usual 4/5 to 11/12 [10].
Associated features of
Study Instrument No. of patients Prevalence of depressive disorder depressive disorder Comment
Feldman et al. [31] PSE 453 All depressive disorders = 14% Younger women
Unrelated to severity of medical
illness
Silverstone et al. [34] SCAN/DSM-IV 186 Depressive disorder = 9.7% Young age, female Nurses (33%) than doctors (22%) better
Unrelated severity of illness or at recognising depression
medical diagnosis
Silverstone [32] SCAN/DSM-IV 313 Major depressive disorder = Young age Endicott criteria used Low prevalence, as
5.1%; adjustment Unrelated to sex, severity of depression present before admission
disorder = 13.7% illness or medical diagnosis often resolved after hospital admission
Arolt et al. [30] CIDI/ICD-10 250 Depressive episode = 4.8%; Divorced/widowed status Worst social functioning associated with
dysthymia = 4%; depressive dysthymia, young age, depression
adjustment disorder = 7.6%; preceding medical illness and chronic
Total depression = 16.4% disease
Martucci et al. [7] CIDI/ICD-10 1039 (298 Depressive disorder = 12.8% Female sex Disability days increased in cases and
diagnoses interviewed) Recent life events, low family subthreshold cases
support
Creed et al. [9] SCAN 263 Depressive disorder = 20% Female sex, severity and Impaired functioning associated in
number of medical illnesses dose–response relationship with
cases/subthreshold cases and controls.
Hansen et al. [42] SCAN 294 Depressive disorders (including Young age, female sex and life- Of depressed patients only 18%
dysthymia) = 8.3% threatening illness but not recognised by physicians
chronic illness
Nair and Pillay [33] SCID 230 Depressive disorders = 7% Females
Unrelated to severity of medical
illness
CIDI, Composite International Diagnostic Interview; DSM-IV, Diagnostic and Statistical Manual of Mental Disorders, 4th edn; ICD-10, International Classification
of Diseases, 10th revision; PSE, Present State Examination; SCAN, Schedule for Clinical Assessment in Neuropsychiatry; SCID, Structured Clinical Interview for the
revised version of the Diagnostic and Statistical Manual of Mental Disorders.
9 Depression in the medically ill
CIDI, Composite International Diagnostic Interview; DSM-IV, Diagnostic and Statistical Manual of Mental Disorders, 4th edn; HADS, Hospital Anxiety and Depression Scale;
ICD-10, International Classification of Diseases, 10th revisions; ICDL, International Diagnostic Checklist for ICD-10; MDD, Major depressive disorder; MUS, medically unexplained
symptoms; PHQ, Patient Health Questionnaire; PSE, Present State Examination; WBI-S, Well-Being Index, version 5.
11 Depression in the medically ill
Population-based studies
Cross-sectional studies demonstrate a close association between depressive disor-
ders and physical illness in population-based studies. In the first large study, people
with cancer, heart disease, neurological disorder or physical handicap were found to
have a significantly higher prevalence of anxiety and depressive disorders (30.3%–
37.5%) than people without a chronic physical illness (17.5%) [16]. In another
population-based study, chronic disabling physical illness emerged together with
lack of a close confidant as the two predictors of depression and anxiety [40].
The nature of the association between physical and psychiatric disorders cannot
be discerned from cross-sectional studies, but evidence from a large (n = 7076)
population-based study suggests that some common generic factors (low educa-
tional attainment, high neuroticism) may be associated independently with psy-
chiatric and physical disorders [41].
The traditional view has tended towards seeing depression in physically ill people
as an understandable reaction to the symptoms and disabling effects of physical
illness. There is only limited evidence to support this notion, however, and some
studies show little or no association between depression and the severity of physical
illness [31,35]. Hansen and colleagues [42] found that life-threatening, but not
chronic, physical illness was associated with increased risk of depression.
Prospective studies are required to assess causality. Several studies have shown,
especially in older people, that physical illness is a risk factor for later development
of depression [43–45] and for continuing depression [46]; a long-term physical
illness can double the chance of subsequent depression [47]. There is evidence,
however, that psychiatric disorders, notably depressive disorder, are a predictor of
later physical illness [48,49]. This is especially so for coronary heart disease [50].
Thus, there are probably numerous ways in which physical and psychiatric disorders
may coexist.
Prospective studies in clinical samples have shown the importance of separating
depressive disorders that precede and follow the onset of the physical illness. For
12 Depression and physical illness
example, when depressive disorder precedes a myocardial infarction, the usual risk
factors for depression apply: younger age, female sex, social isolation, past history
of depression, separation from parent during childhood and experience of marked
social difficulties [51,52]. When depression occurs after heart attack, however, these
risk factors do not apply – the presence of one or more serious concurrent phys-
ical illness (e.g. arthritis, chest disease) was associated with later depression [51].
Depression that precedes the onset of physical illness has the greatest effect on social
functioning [29].
8.8 days. For people who had both one of the physical illnesses and depression, the
mean figure was 13.5 days [54].
In another study, the effect of depressive or anxiety disorders on subjective health-
related quality of life in the physical domain was considerable. The presence of a
chronic physical illness led to a mean reduction of four points on the physical
component summary score of the Short Form 36 (SF36) compared with healthy
controls [55–57]. The presence of a mood disorder alone led to a similar reduction
of four points, but if both a chronic physical illness and a mood disorder were
present then the SF36 physical component score was reduced by 11 points com-
pared with healthy controls [58]. Similar findings have been reported in clinical
samples. For example, in gastroenterology patients, whether the disorder is organic
(inflammatory bowel disease) or functional (irritable bowel syndrome), the SF36
physical component score is reduced greatly when a concurrent depressive disorder
is present [59,60].
Four causal processes have been suggested to account for the marked deteri-
oration of role daily-functioning when depressive disorders accompany physical
illness: depressive disorders may (a) amplify subjective reactions to somatic symp-
toms, (b) reduce motivation to care for physical illness, (c) lead to maladaptive direct
physiological effects on bodily symptoms and (d) reduce the capacity to cope with
physical illnesses through limitation of energy, cognitive capacity, affect regulation,
perception of shame or social stigma [53].
As well as leading to poor health-related quality of life, depressive disorder that
accompanies physical illness may lead to a number of other poor outcomes, includ-
ing increased mortality and increased healthcare use. In a large sample of medical
in-patients followed up over two years, and after adjustment for a number of con-
founders, depressive symptoms were related to subsequent mortality. The adjusted
relative risk of death for depressed versus non-depressed patients was 1.9 (95% con-
fidence interval [CI] 1.2–3.1) [61]. This effect has been shown in coronary artery
disease (see Chapter 4) and stroke [11,62].
In medical in-patients, subsequent high utilisation and costs have been shown
to be associated with psychiatric disorder (mostly depressive disorders and sub-
threshold disorders) [9]. In medical out-patients, the best predictors of outpa-
tient utilisation were number of medical diagnoses, number of somatic symptoms,
hypochondriacal attitudes and depressive disorder [63]. In primary care, healthcare
costs are higher in people who have depression, even after controlling for the effect
of physical illness [64].
The most important implication of these results concerning the effect of depres-
sion on outcome is the fact that physicians caring for people with these chronic
physical disorders often fail to detect and treat the co-occurring depression and
14 Depression and physical illness
REFERENCES
1. C. J. Meakin, Screening for depression in the medically ill: the future of paper and pencil
tests. Br. J. Psychiatry 160 (1992), 212–16.
2. G. Rodin, K. Voshart, Depression in the medically ill: an overview. Am. J. Psychiatry 143
(1986), 696–705.
3. R. Mayou, K. Hawton, Psychiatric disorder in the general hospital. Br. J. Psychiatry 149 (1986),
172–90.
4. J. S. McDaniel, D. L. Musselman, M. R. Porter, D. A. Reed, C. B. Nemeroff, Depression in
patients with cancer: diagnosis, biology, and treatment. Arch. Gen. Psychiatry 52 (1995),
89–99.
5. F. H. Creed, Assessing depression in the context of physical illness. In M. M. Robinson,
C. L. E. Katon, Depression and Physical Illness, ed. (Chichester: John Wiley & Sons, 1997),
pp. 3–19.
6. F. Creed, Psychological disorders in rheumatoid arthritis: a growing consensus? Ann. Rheum.
Dis. 49 (1990), 808–12.
7. M. Martucci, M. Balestrieri, G. Bisoffi, et al., Evaluating psychiatric morbidity in a general
hospital: a two-phase epidemiological survey. Psychol. Med. 29 (1999), 823–32.
8. G. Gainotti, A. Azzoni, C. Marra, Frequency, phenomenology and anatomical-clinical cor-
relates of major post-stroke depression. Br. J. Psychiatry 175 (1999), 163–7.
9. F. Creed, R. Morgan, M. Fiddler, et al., Depression and anxiety impair health-related quality
of life and are associated with increased costs in general medical inpatients. Psychosomatics
43 (2002), 302–9.
10. K. W. Bridges, D. P. Goldberg, The validation of the GHQ-28 and the use of the MMSE in
neurological in-patients. Br. J. Psychiatry 148 (1986), 548–53.
11. A. House, P. Knapp, J. Bamford, A. Vail, Mortality at 12 and 24 months after stroke may be
associated with depressive symptoms at 1 month. Stroke 32 (2001), 696–701.
12. A. House, Mood disorders in the physically ill: problems of definition and measurement. J.
Psychosom. Res 32 (1988), 345–53.
13. G. Dunn, A. Pickles, M. Tansella, J. L. Vazquez-Barquero, Two-phase epidemiological surveys
in psychiatric research. Br. J. Psychiatry 174 (1999), 95–100.
14. A. Solomon, D. A. Haaga, B. A. Arnow, Is clinical depression distinct from subthreshold
depressive symptoms? A review of the continuity issue in depression research. J. Nerv. Ment.
Dis. 189 (2001), 498–506.
15. P. Rucci, S. Gherardi, M. Tansella, et al., Subthreshold psychiatric disorders in primary care:
prevalence and associated characteristics. J. Affect. Dis. 76 (2003), 171–81.
15 Depression in the medically ill
34. P. H. Silverstone, T. Lemay, J. Elliott, V. Hsu, R. Starko, The prevalence of major depressive
disorder and low self-esteem in medical inpatients. Can. J. Psychiatry 41 (1996), 67–74.
35. W. Katon, P. Ciechanowski, Impact of major depression on chronic medical illness. J. Psy-
chosom. Res. 53 (2002), 859–63.
36. A. J. Carson, B. Ringbauer, L. MacKenzie, C. Warlow, M. Sharpe, Neurological disease,
emotional disorder, and disability: they are related – a study of 300 consecutive new refer-
rals to a neurology outpatient department. J. Neurol. Neurosurg. Psychiatry 68 (2000),
202–6.
37. B. Lowe, K. Grafe, S. Zipfel, et al., Diagnosing ICD-10 depressive episodes: superior criterion
validity of the Patient Health Questionnaire. Psychother. Psychosom. 73 (2004), 386–90.
38. S. Kisely, D. Goldberg, G. Simon, A comparison between somatic symptoms with and without
clear organic cause: results of an international study. Psychol. Med. 27 (1997), 1011–19.
39. T. B. Üstün, N. Sartorius, Mental Illness in General Health Care. An International Study.
Chichester: John Wiley & Sons 1995.
40. J. Harrison, S. Barrow, L. Gask, F. Creed, Social determinants of GHQ score by postal survey.
J. Public. Health. Med. 21 (1999), 283–8.
41. J. Neeleman, J. Ormel, R. V. Bijl, The distribution of psychiatric and somatic III health:
associations with personality and socioeconomic status. Psychosom. Med. 63 (2001),
239–47.
42. M. S. Hansen, P. Fink, M. Frydenberg, et al., Mental disorders among internal medical
inpatients: prevalence, detection, and treatment status. J. Psychosom. Res. 50 (2001), 199–
204.
43. G. Livingston, V. Watkin, B. Milne, M. V. Manela, C. Katona, Who becomes depressed? The
Islington community study of older people. J. Affect. Dis. 58 (2000), 125–33.
44. R. Finlay-Jones, G. W. Brown, Types of stressful life event and the onset of anxiety and
depressive disorders. Psychol. Med. 11 (1981), 803–15.
45. E. Murphy, Social origins of depression in old age. Br. J. Psychiatry 141 (1982), 135–42.
46. M. G. Cole, F. Bellavance, A. Mansour, Prognosis of depression in elderly community and
primary care populations: a systematic review and meta-analysis. Am. J. Psychiatry 156 (1999),
1182–9.
47. S. B. Patten, Long-term medical conditions and major depression in a Canadian population
study at waves 1 and 2. J. Affect. Dis. 63 (2001), 35–41.
48. M. Hotopf, R. Mayou, M. Wadsworth, S. Wessely, Temporal relationships between physical
symptoms and psychiatric disorder: results from a national birth cohort. Br. J. Psychiatry 173
(1998), 255–61.
49. E. Murphy, G. W. Brown, Life events, psychiatric disturbance and physical illness. Br. J.
Psychiatry 136 (1980), 326–38.
50. H. Hemingway, M. Marmot, Evidence based cardiology: psychosocial factors in the aetiology
and prognosis of coronary heart disease. Systematic review of prospective cohort studies. Br.
Med. J. 318 (1999), 1460–67.
51. C. M. Dickens, C. Percival, L. McGowan, et al., The risk factors for depression in first myocar-
dial infarction patients. Psychol. Med. 34 (2004), 1083–92.
17 Depression in the medically ill
69. J. J. Strik, A. Honig, R. Lousberg, J. Denollet, Sensitivity and specificity of observer and
self-report questionnaires in major and minor depression following myocardial infarction.
Psychosomatics 42 (2001), 423–8.
70. J. I. Escobar, M. Gara, R. C. Silver, H. Waitzkin, A. Holman, W. Compton, Somatisation
disorder in primary care. Br. J. Psychiatry 173 (1998), 262–6.
71. J. I. Escobar, H. Waitzkin, R. C. Silver, M. Gara, A. Holman, Abridged somatization: a study
in primary care. Psychosom. Med. 60 (1998), 466–72.
2
2
Psychosocial factors, depression and illness
Stephen Stansfeld and Farhat Rasul
Introduction
Depression, including the full range of minor to major depressive disorders, is
a common and disabling condition. As with most psychological disorders, the
aetiology of depression is multifactorial, but social and psychological factors play a
more prominent role in its aetiology than in most other conditions in medicine. This
has importance both in understanding the aetiology and in finding opportunities
for prevention and devising psychological and social treatments.
This chapter describes the evidence for social and psychological factors playing
a role in the aetiology of depression. Social factors include general factors at the
level of human society concerned with social structure and social processes that
impinge on the individual. Psychological factors include individual-level processes
and meanings that influence mental states. Sometimes these words are combined
as ‘psychosocial’. This is a shorthand term for the combination of psychological and
social, but it also implies that the effect of social processes is sometimes mediated
through psychological understanding. Hence, it has been said that the impact of
social inequality on health may be mediated partly through perceptions of shame.
Much of the data on psychosocial factors and mental health deal with broader
categories than depression, such as common mental disorder, of which depression
is a prominent component. This chapter focuses on depressive disorders but also
includes landmark studies that have used the broader categories that illustrate the
role of psychosocial variables.
Sociodemographic factors
in these differences than do biological or genetic factors [3]. Females seem more
sensitive than males to adverse experiences in childhood and are at greater risk of
sexual abuse, which may have long-term mental health consequences [4]. Women
have onset of depression at an earlier age than men; early-onset depression predicts
more and longer episodes of depression in adulthood [5]. Anxiety disorders, which
are more common in women, often predict early-onset depression and, hence, partly
explain the gender difference in rates of depression [6]. In adulthood, women do
not consistently experience more life events than men, but the reaction of women
to life events, particularly salient events related to children and relationships, may
have more meaning and subsequently more impact for women than men [7].
There are several hypotheses about the impact of different social roles on health.
Multiple roles, including work, domestic roles and leisure activities, are generally
thought to be beneficial for health as these spread an individual’s commitments and
sources of support and have different domains of satisfaction, resources and esteem
[8]. Conversely, the role overload hypothesis suggests that coping with too many
roles may be a risk factor for depression and that conflict and spill-over between
roles, for instance between home and work, may be a risk factor for mental ill-
health [9,10]. What seems to be crucial is the quality of each of the roles [11].
Undervalued roles, such as that of the housewife or lowly paid employment, may
contribute to higher rates of depression in women [12,13], although differences in
roles are only a partial explanation for gender differences [14]. Indeed, in studies
where psychosocial differences are controlled, such as in Jenkin’s [15] study of
executive officers in the civil service, the rates of mental ill-health were the same
in both sexes. There are probably multiple causes for gender differences in rates of
depression; suffice to say these causes are largely psychosocial.
for persistent depression (OR = 2.06) [23]. Poverty, financial difficulties and low
income may prevent an individual getting better once he or she is depressed [26].
disadvantage by contrast was associated with both onset and maintenance of com-
mon mental disorder.
Resolution events
Life events are not always negative in their impact; for instance, so-called ‘resolution
events’ that remove prolonged negative circumstances, such as re-employment after
prolonged unemployment, may lead to resolution of depression [51,52], and even
when difficulties remain in one domain a fresh start in another domain can begin
the process of remission [53].
disruptive behaviour on the part of a child, lover or close relative. Part of the impact
of full-time work might be through reduced contact with the child, although neg-
ative interaction with the child was found to be related highly to marked financial
hardship but not to full-time work. At the time, a major reason for choosing to
work full time rather than part time was in order to avoid poverty, and thus these
women were especially at risk [57].
but vary according to social circumstances. Low self-esteem has not only been
shown to be a significant risk factor for developing clinical depression [62] but
has also been related to other psychosocial risk factors, such as the quality of close
relationships and early adverse experience [63]. Over time, negative evaluation
of self improved for about half of those who had negative evaluation of self at
baseline. Although negative evaluation of self was associated with current depressive
symptoms, the improvement could not always be associated with recovery from
depression [64].
Social support
Social support has been defined as ‘resources provided by other persons’ [75]. It
has been seen as ‘information leading the subject to believe that he is cared for and
loved, is esteemed and valued and belongs to a social network of communication
and mutual obligation’ [76]. One of the most important distinctions is between
social networks and the functional aspects of support, i.e. the quality and type
of support that is provided by the network member. Social networks refer to the
social contacts of a group of individuals. Such contacts can be described in terms
of number of contacts and frequency of contacts.
Two psychological mechanisms have been postulated for the beneficial effect of
social relations on mental health: (i) support has a direct effect on wellbeing, and
(ii) the buffering hypothesis, whereby support moderates the impact of stressors
on the risk of depression [77]. Kessler and McLeod [78] proposed that network
size has a direct effect on mental wellbeing, such that larger networks are associated
with better health. In contrast, they suggest that emotional support has a buffering
effect in the face of stressors. They cited three cross-sectional studies [45,79,80]
and two panel studies [81,82] in support of a pervasive buffering effect on mental
health across all types of event. However, some authors dispute that these vul-
nerability factors really increase the risk of depression in a multiplicative fashion.
Although it seems plausible that the opportunities for sharing and diffusing nega-
tive emotions, provision of reassurance and active emotional support alleviate the
negative impact of a life event, many studies suggest that the effects of life events
and social support on the risk of depressive illness are merely additive [83]. In a
large panel study of adult female twins, Kessler et al. [84] found a buffering effect
for perceived support on major depression. They found no evidence to support
an underlying genetic factor influencing perception of support and accommoda-
tion to stress. Neither did they find evidence that buffering was attributable to any
improvement in coping promoted by social support. They also rejected the possi-
bility that stress-moderating effects of perceived support are mediated by received
support. Kessler and colleagues suggested the possibility that perceived availabil-
ity of support may lead to cognitive appraisals that are less threatening to mental
health.
There have been a few longitudinal studies of social support where questions of
causation can be better addressed. Some community studies have been completely
prospective, identifying deficiencies in social support before the onset of depression
and relating this to the onset of depression. Two studies found a buffering or
interactive effect [62,85]. The study by Brown et al. [62] found little predictive
effect on mental health of 17 measures of emotional support measured at baseline
in an inner-city sample of married mothers. However, Brown and colleagues did
29 Psychosocial factors, depression and illness
find a greater risk of depression in women who received little crisis support, i.e. little
support when it was needed in order to cope with a life event. A negative response
from a partner in a crisis was also associated with a subsequent risk of depression.
Among single mothers, the report of a close relationship at baseline was protective
against the development of depression following a subsequent life event. The onset
of postnatal depression in 507 women was predicted by lack of social support from
the primary group and lack of support in relation to becoming pregnant, adjusting
for antenatal depression, neuroticism, family and personal psychiatric history, and
adversity [86].
A longitudinal study of British middle-aged civil servants (the Whitehall II
study) showed in relation to psychological distress rather than depressive illness
a protective effect on mental health of emotional support from the closest per-
son in men [87] and from the primary group in women [56]. These were not
abolished by adjusting for either hostility as a measure of personality or psychi-
atric disorder at baseline. This is relevant because measures of personality often
confound associations between social support and mental illness. Moreover, this
study showed prospectively that negative aspects of close relationships were asso-
ciated with greater risk of future psychiatric disorder up to five years later. Nega-
tive aspects of close relationships were associated consistently with worse mental
health in both men and women. Negative aspects of close relationships have a
negative effect on mental health both directly [88–90] and in the presence of life
events.
Emotional support is what the respondent receives from the close person and
is associated with better mental health [62]. Perceived emotional support has a
larger effect than tangible or practical aspects of support, as has also been found in
elderly people [91]. Emotional support is distinct from confiding, which requires
the respondent to actively disclose information to the close person, although in
adaptive relationships these occur reciprocally. Some studies suggest that ill-advised
confiding, asking for support from sources unable to provide it, and confiding
without active emotional support may be risk factors for depression, and this may
explain why confiding in and receiving emotional support from a spouse does not
show a direct effect in women [92].
Social support in adulthood may help to buffer the persistent effects of early
childhood abuse and neglect. A study of women who had been in care during
childhood shows how social support in adulthood may exert a beneficial effect on
parenting problems, marital difficulties and psychiatric disorder [93]. Many of these
women returned from care to a discordant home environment, from which they
then tried to escape by early marriage. However, the women’s marital relationships
often turned out badly and resulted in the women becoming more vulnerable to
further difficulties. Nevertheless, a third of those women showed good parenting
30 Depression and physical illness
group members seek more affiliation as a way of coping with greater hardship than
white people. Ethnic minority group members coming from a collectivist cultural
heritage or having been exposed to greater hardship may have a greater tendency
than white people to develop stronger social support networks. This in turn may
afford protection against psychiatric disorder.
In a study of 4700 participants, Plant and Sachs-Ericsson [117] examined whether
there was any qualitative difference in social support between ethnic minorities and
white people and whether the quality of social support protected minority group
members from depression. The sample consisted of 85% white, 10% Hispanic, 4%
black and 1% Native American people of both sexes. The quality of social support
was measured using an interpersonal functioning scale. Overall, ethnic minority
groups tended to report better interpersonal functioning than white participants.
Black and Hispanic participants had lower scores on interpersonal functioning
than white participants; Native American participants’ scores fell between the two.
Further analysis indicated that more problems in interpersonal functioning were
related to both more depressive symptoms and prevalence of depression. Taken
together, these results pointed to the possibility that the quality of social support
may moderate the relationship between ethnicity and depression. When interper-
sonal functioning was controlled for in the prediction of depressive symptoms
and major depression, differences between ethnic minority groups and the white
group increased. People from ethnic minority groups reported even higher levels of
depressive symptoms and significantly higher risk of major depression compared
with white people. These findings suggested that social support, as measured by
interpersonal functioning, suppresses group differences in depressive symptoms
and prevalence of major depression [117].
Can this be taken as evidence that social support moderates the relationship
between ethnicity and depression? There are some problems with this study. First,
although physical functioning was measured and was clearly associated with depres-
sion, it is not clear from the analyses whether this was controlled for. Second, ethnic
groups were combined into one group, labelled ‘minority group members’. There
are similarities among ethnic minority groups (e.g. low SEP), but there are also many
differences between them. Third, this was a cross-sectional study and therefore the
temporal relationship between these factors and depression cannot be determined
definitely. Only a longitudinal approach could track changes in risk and protective
factors for depression in ethnic minority and other groups.
There is some evidence that black and Hispanic ethnic minority people in the USA
have more supportive family and social networks, tend to have larger households
and are more likely to have extended family members living together compared
with white people [110,118–120]. Such close supportive networks may moderate
the effects of adverse circumstances and protect members of ethnic minority groups
35 Psychosocial factors, depression and illness
from stress and depression. Holahan and Moos [121] showed that having high
social resources lessened the effects of stress through their association with active
coping strategies. Ennis et al. [122] showed that black women with acute economic
problems had lower levels of depressive symptoms if they had a strong social support
system than if they did not have such a strong support system.
Other social factors, such as religion, may also play an important role in men-
tal health for both minority group and white people. Religious activity may be
especially protective for members of ethnic minority groups; it is thought that
this protects against suicide in black and Hispanic people. Increased opportunities
for social support through engaging in religious activities may be the mechanism
by which religious activities confer a protective effect on mental health of ethnic
minorities [123–125].
Many factors may be associated with ethnic minority mental health variation,
including pre- and post-migratory factors (e.g. biological factors, forced migration,
culture conflict, acculturation) and factors associated with the process of migra-
tion itself. People from ethnic minorities also tend to experience more difficulties
in terms of racism, discrimination, economic difficulties and decreased opportu-
nities than people from non-ethnic minorities. Some of these factors may be causal
or simply associated with poor mental health in ethnic groups [126]. An addi-
tional difficulty is that tools developed for Western psychiatric practice may be
inappropriate for use in psychiatric surveys of people from ethnic minority popu-
lations, who experience and express mental illness within culturally bound norms
[127]. Nevertheless, as the vast literature on social capital and wellbeing shows, it is
not unreasonable to suppose that the protective effect of clustering on depression
might in part be mediated by the potency of close supportive networks in cohesive
minority groups.
have not relied on self-reports of work but have used assessments of the work
environment, external to the person reporting psychological distress, have not
found associations between work and mental health. Although this would tend
to support the response bias argument, it is possible that the subjective percep-
tions of work are a necessary mediating step between the work environment and
psychological distress.
then it is unlikely that new depressive onsets will occur for the first time in old
age [157].
Having discussed the differences between younger and older adults, there is also
a lot of continuity across the lifespan in the response to stressors [158,159], even
though not all studies find increased risk of depression associated with life events
[160]. Strong associations have been found between the onset of physical illness
and the risk of depression, with an increased risk of up to three times [95]. This
risk is maintained even after adjusting for confounding factors [161]. The onset
of new physical illness in elderly people symbolises the advent of disability and
intimations of mortality, which may increase the risk for depression. The onset of
disability itself is a powerful risk factor for depression and is often accompanied by
loss of independence, perception of loss of control over the environment, loss of
productivity and a shaming inability to perform everyday role activities [95,157].
Methodologically, it is often difficult to separate out the direction of causation,
because although disability is a risk factor for depression, depression may also
increase levels of disability (see Chapter 6). Death of a spouse, which is more likely
to occur in elderly people, is an evident risk factor for the onset of depression
[157,162]; but there is variability of response to death of a spouse, depending on
whether the death was expected and whether it was preceded by illness and disability
and caregiving demands.
There have been several prospective studies of social support and depression
in elderly people [91,95,151,161]. Lack of a relative or friend in whom to confide
predicted late-life suicide in a North American community survey [163]. Lack of
contact with friends predicted depression and modified the association between
depression and disability, a powerful predictor of depression [95]. In a similar
way, having a marital partner or, if unmarried, having social support significantly
reduced the impact of functional disabilities on depression [161]. Hence, in a
situation where provision of practical and emotional support may have profound
effects on an elderly person’s quality of life and functioning, a clear protective effect
of social support was found. Lack of social support has not been associated with
increased risk of depression in elderly people in all studies [161]. Some of this
variation may relate to the changing pattern of social relationships in old age as
friends and relatives die, and with increasing disability, the support from children
and statutory services becomes more important.
stressors are not specific to depression but are also implicated in the causation of
physical illness. Job strain and social isolation have been identified as risk factors for
coronary heart disease [164,165]. Some psychosocial risk factors seem more spe-
cific to physical illness, and others to mental illness, while some carry a common
cause (e.g. work) [166]. Thus, psychosocial risk factors may play a direct role in the
aetiology of physical illness or may exert their effects indirectly, mediated through
the development of depression. As other chapters in this book suggest, depression
may subsequently exert a direct influence on the development of physical illness.
REFERENCES
15. R. Jenkins, Sex differences in minor psychiatric morbidity. Psychol. Med. Monogr. Suppl. 7
(1985), 1–53.
16. S. Wyke, G. Ford, Competing explanations for associations between marital status and
health. Soc. Sci. Med. 34 (1992), 523–32.
17. P. E. Bebbington, Sex and depression. Psychol. Med. 28 (1998), 1–8.
18. I. Waldron, M. E. Hughes, T. L. Brooks, Marriage protection and marriage selection:
prospective evidence for reciprocal effects of marital status and health. Soc. Sci. Med. 43
(1996), 113–23.
19. C. A. Mathews, V. I. Reus, Assortative mating in the affective disorders: a systematic review
and meta-analysis. Compr. Psychiatry 42 (2001), 257–62.
20. D. Melzer, T. Fryers, R. Jenkins, T. Brugha, B. McWilliams, Social position and the common
mental disorders with disability: estimates from the National Psychiatric Survey of Great
Britain. Soc. Psychiatry Psychiatr. Epidemiol. 38 (2003), 238–43.
21. D. Melzer, T. Fryers, R. Jenkins, Social Inequalities and the Distribution of the Common Mental
Disorders (Hove: Psychology Press, 2004).
22. J. M. Murphy, D. C. Olivier, R. R. Monson, et al. Depression and anxiety in relation to social
status: a prospective epidemiologic study. Arch. Gen. Psychiatry 48 (1991), 223–9.
23. V. Lorant, D. Deliege, W. Eaton, et al., Socioeconomic inequalities in depression: a meta-
analysis. Am. J. Epidemiol. 157 (2003), 98–112.
24. S. Aro, H. Aro, I. Keskimaki, Socio-economic mobility among patients with schizophrenia
or major affective disorder: a 17-year retrospective follow-up. Br. J. Psychiatry 166 (1995),
759–67.
25. S. Weich, G. Lewis, Material standard of living, social class, and the prevalence of the common
mental disorders in Great Britain. J. Epidemiol. Community Health 52 (1998), 8–14.
26. S. Weich, G. Lewis, Poverty, unemployment, and common mental disorders: population
based cohort study. Br. Med. J. 317 (1998), 115–19.
27. T. Fryers, D. Melzer, R. Jenkins, Social inequalities and the common mental disorders: a
systematic review of the evidence. Soc. Psychiatry Psychiatr. Epidemiol. 38 (2003), 229–37.
28. B. P. Dohrenwend, I. Levav, P. E. Shrout, et al., Socioeconomic status and psychiatric
disorders: the causation-selection issue. Science 255 (1992), 946–52.
29. J. K. Myers, J. J. Lindenthal, M. P. Pepper, D. R. Ostrander, Social class, life events and
psychiatric symptoms: a longitudinal study. In Stressful Life Events: Their Nature and Effects,
ed. B. S. Dohrenwend and B. P. Dohrenwend. (New York: John Wiley & Sons, 1974), pp.
91–206.
30. R. C. Kessler, K. A. McGonagle, S. Zhao, et al., Lifetime and 12-month prevalence of DSM-
III-R psychiatric disorders in the United States: results from the National Comorbidity
Survey. Arch. Gen. Psychiatry 51 (1994), 8–19.
31. R. S. Kahn, P. H. Wise, B. P. Kennedy, I. Kawachi, State income inequality, household income,
and maternal mental and physical health: cross sectional national survey. Br. Med. J. 321
(2000), 1311–15.
32. S. A. Everson, S. C. Maty, J. W. Lynch, G. A. Kaplan, Epidemiologic evidence for the relation
between socioeconomic status and depression, obesity, and diabetes. J. Psychosom. Res. 53
(2002), 891–5.
42 Depression and physical illness
33. M. Broese, I. van Groenou, [Unequal chances for reaching ‘a good old age’: socio-economic
health differences among older adults from a life course perspective.] Tijdschr. Gerontol.
Geriatr. 34 (2003), 196–207.
34. M. Benzeval, K. Judge, Income and health: the time dimension. Soc. Sci. Med. 52 (2001),
1371–90.
35. S. E. Gilman, I. Kawachi, G. M. Fitzmaurice, S. L. Buka, Socioeconomic status in childhood
and the lifetime risk of major depression. Int. J. Epidemiol. 31 (2002), 359–67.
36. M. Marmot, M. Shipley, E. Brunner, H. Hemingway, Relative contribution of early life
and adult socioeconomic factors to adult morbidity in the Whitehall II study. J. Epidemiol.
Community Health 55 (2001), 301–7.
37. E. Goodman, The role of socioeconomic status gradients in explaining differences in US
adolescents’ health. Am. J. Public Health 89 (1999), 1522–8.
38. J. E. Ritsher, V. Warner, J. G. Johnson, B. P. Dohrenwend, Inter-generational longitudinal
study of social class and depression: a test of social causation and social selection models.
Br. J. Psychiatry Suppl. 40 (2001), 584–90.
39. R. J. Turner, F. Marino, Social support and social structure: a descriptive epidemiology.
J. Health Soc. Behav. 35 (1994), 193–212.
40. B. H. Brummett, J. C. Barefoot, P. P. Vitaliano, I. C. Siegler, Associations among social
support, income, and symptoms of depression in an educated sample: the UNC Alumni
Heart Study. Int. J. Behav. Med. 10 (2003), 239–50.
41. R. J. Turner, D. A. Lloyd, P. Roszell, Personal resources and the social distribution of depres-
sion. Am. J. Community Psychol. 27 (1999), 643–72.
42. S. A. Stansfeld, J. Head, M. G. Marmot, Explaining social class differences in depression and
well-being. Soc. Psychiatry Psychiatr. Epidemiol. 33 (1998), 1–9.
43. B. P. Dohrenwend, The role of adversity and stress in psychopathology: some evidence and
its implications for theory and research. J. Health Soc. Behav. 41 (2000), 1–19.
44. R. Finlay-Jones, G. W. Brown, Types of stressful life event and the onset of anxiety and
depressive disorders. Psychol. Med. 11 (1981), 803–15.
45. G. W. Brown, T. O. Harris, Social Origins of Depression: A Study of Psychiatric Disorder in
Women (London: Tavistock, 1978).
46. E. S. Paykel, Contribution of life events to causation of psychiatric illness. Psychol. Med. 8
(1978), 245–53.
47. E. S. Paykel, Life events and affective disorder. Acta Psychiatr. Scand. Suppl. 418 (2003), 61–6.
48. E. Frank, B. Anderson, C. F. Reynolds, III, A. Ritenour, D. J. Kupfer, Life events and the
research diagnostic criteria endogenous subtype: a confirmation of the distinction using
the Bedford College methods. Arch. Gen. Psychiatry 51 (1994), 519–24.
49. G. W. Brown, T. O. Harris, C. Hepworth, Loss, humiliation and entrapment among women
developing depression: a patient and non-patient comparison. Psychol. Med. 25 (1995),
7–21.
50. M. E. P. Seligman, Helplesssness. (San Francisco, CA: Freeman, 1975).
51. G. W. Brown, L. Lemyre, A. Bifulco, Social factors and recovery from anxiety and depressive
disorders: a test of specificity. Br. J. Psychiatry 161 (1992), 44–54.
43 Psychosocial factors, depression and illness
52. A. S. Leenstra, J. Ormel, R. Giel, Positive life change and recovery from depression and
anxiety: a three-stage longitudinal study of primary care attenders. Br. J. Psychiatry 166
(1995), 333–43.
53. T. Harris, Recent developments in understanding the psychosocial aspects of depression.
Br. Med. Bull. 57 (2001), 17–32.
54. M. L. Bruce, K. M. Kim, Differences in the effects of divorce on major depression in men
and women. Am. J. Psychiatry 149 (1992), 914–17.
55. K. S. Kendler, L. M. Thornton, C. A. Prescott, Gender differences in the rates of exposure
to stressful life events and sensitivity to their depressogenic effects. Am. J. Psychiatry 158
(2001), 587–93.
56. R. Fuhrer, S. A. Stansfeld, J. Chemali, M. J. Shipley, Gender, social relations and mental
health: prospective findings from an occupational cohort (Whitehall II study). Soc. Sci.
Med. 48 (1999), 77–87.
57. G. W. Brown, P. M. Moran, Single mothers, poverty and depression. Psychol. Med. 27 (1997),
21–33.
58. R. Plomin, P. Lichtenstein, N. L. Pedersen, G. E. McClearn, J. R. Nesselroade, Genetic
influence on life events during the last half of the life span. Psychol. Aging 5 (1990),
25–30.
59. K. L. Harkness, S. M. Monroe, A. D. Simons, M. Thase, The generation of life events in
recurrent and non-recurrent depression. Psychol. Med. 29 (1999), 135–44.
60. L. A. Champion, G. Goodall, M. Rutter, Behaviour problems in childhood and stressors in
early adult life: I. A 20 year follow-up of London school children. Psychol. Med. 25 (1995),
231–46.
61. A. Caspi, K. Sugden, T. E. Moffitt, et al., Influence of life stress on depression: moderation
by a polymorphism in the 5-HTT gene. Science 301 (2003), 386–9.
62. G. W. Brown, B. Andrews, T. Harris, Z. Adler, L. Bridge, Social support, self-esteem and
depression. Psychol. Med. 16 (1986), 813–31.
63. B. Andrews, G. W. Brown, Self-esteem and vulnerability to depression: the concur-
rent validity of interview and questionnaire measures. J. Abnorm. Psychol. 102 (1993),
565–72.
64. B. Andrews, G. W. Brown, Stability and change in low self-esteem: the role of psychosocial
factors. Psychol. Med. 25 (1995), 23–31.
65. G. Parker, Parental ‘affectionless control’ as an antecedent to adult depression: a risk factor
delineated. Arch. Gen. Psychiatry 40 (1983), 956–60.
66. C. Tennant, P. Bebbington, J. Hurry, Parental death in childhood and risk of adult depressive
disorders: a review. Psychol. Med. 10 (1980), 289–99.
67. G. W. Brown, T. O. Harris, M. J. Eales, Aetiology of anxiety and depressive disorders in an
inner-city population: 2. Comorbidity and adversity. Psychol. Med. 23 (1993), 155–65.
68. A. Bifulco, G. W. Brown, T. O. Harris, Childhood Experience of Care and Abuse (CECA): a
retrospective interview measure. J. Child. Psychol. Psychiatry 35 (1994), 1419–35.
69. J. Hill, R. Davis, M. Byatt, et al., Childhood sexual abuse and affective symptoms in women:
a general population study. Psychol. Med. 30 (2000), 1283–91.
44 Depression and physical illness
89. K. S. Rook, The negative side of social interaction: impact on psychological well-being. J.
Pers. Soc. Psychol. 46 (1984), 1097–108.
90. M. M. Burg, T. E. Seeman, Families and health: the negative side of social ties. Ann. Behav.
Med. 16 (1994), 109–15.
91. T. E. Oxman, L. F. Berkman, S. Kasl, D. H. Freeman, Jr, J. Barrett, Social support and
depressive symptoms in the elderly. Am. J. Epidemiol. 135 (1992), 356–68.
92. B. Andrews, G. W. Brown, Social support, onset of depression and personality: an
exploratory analysis. Soc. Psychiatry Psychiatr. Epidemiol. 23 (1988), 99–108.
93. D. Quinton, M. Rutter, C. Liddle, Institutional rearing, parenting difficulties and marital
support. Psychol. Med. 14 (1984), 107–24.
94. T. S. Brugha, P. E. Bebbington, B. MacCarthy, et al., Gender, social support and recov-
ery from depressive disorders: a prospective clinical study. Psychol. Med. 20 (1990), 147–
56.
95. M. J. Prince, R. H. Harwood, A. Thomas, A. H. Mann, A prospective population-based
cohort study of the effects of disablement and social milieu on the onset and maintenance
of late-life depression: the Gospel Oak Project VII. Psychol. Med. 28 (1998), 337–50.
96. H. O. Veiel, C. Kuhner, Relatives and depressive relapse: the critical period after discharge
from in-patient treatment. Psychol. Med. 20 (1990), 977–84.
97. G. Fadden, P. Bebbington, L. Kuipers, Caring and its burdens: a study of the spouses of
depressed patients. Br. J. Psychiatry 151 (1987), 660–6.
98. J. Y. Nazroo, Ethnicity and Mental Health (London: Policy Studies Institute, 1997).
99. R. Cochrane, S. S. Bal, Ethnic density is unrelated to incidence of schizophrenia. Br. J.
Psychiatry 153 (1988), 363–6.
100. D. Bhugra, Setting up psychiatric services: cross-cultural issues in planning and delivery.
Int. J. Soc. Psychiatry 43 (1997), 16–28.
101. K. Sproston J. Y. Nazroo, Ethnic Minority Psychiatric Illness Rates in the Community
(EMPIRIC): Quantitative Report (London, The Stationery Office, 2002).
102. D. R. Williams, Race, stress and mental health. In Minority Health in America: Findings
and Policy Implications from the Commonwealth Fund Minority Health Survey, ed. C. J. R.
Hogue, M. A. Hargreaves, K. S. Collins (Baltimore, MD: John Hopkins Press, 2000), pp. 194–
208.
103. D. Halpern, Minorities and mental health. Soc. Sci. Med. 36 (1993), 597–607.
104. R. E. Faris, H. W. Dunham, Mental Disorders in Urban Areas (Chicago, IL: University of
Chicago Press, 1939).
105. L. Levy, L. Rowitz, The Ecology of Mental Disorder (New York: Behavioral Publications,
1973).
106. J. G. Rabkin, Ethnic density and psychiatric hospitalisation: hazards of minority status. Am.
J. Psychiatry 136 (1979), 1562–6.
107. G. L. Mulhin, Mental hospitalisation of the foreign born and the role of cultural isolation.
Int. J. Soc. Psychiatry 25 (1979), 258–66.
108. D. Halpern, J. Y. Nazroo, The ethnic density effect: results from a national community survey
of England and Wales. Int. J. Soc. Psychiatry 46 (1999), 34–46.
46 Depression and physical illness
109. M. Stopes-Roe, R. Cochrane, Support networks of Asian and British families: comparisons
between ethnicities and between generations. Soc. Behav. 5 (1990), 71–85.
110. S. K. Hoppe, R. L. Leon, J. P. Realini, Depression and anxiety among Mexican Americans in
a family health center. Soc. Psychiatry Psychiatr. Epidemiol. 24 (1989), 63–8.
111. T. Joiner, J. C. Coyne, The Interactive Nature of Depression: Advances in Interpersonal
Approaches (Washington, DC: American Psychological Association, 1999).
112. C. J. Holahan, R. H. Moos, C. K. Holahan, R. C. Cronkite, Resource loss, resource gain, and
depressive symptoms: a 10-year model. J. Pers. Soc. Psychol. 77 (1999), 620–9.
113. C. J. Holahan, R. H. Moos, C. K. Holahan, R. C. Cronkite, Long-term posttreatment func-
tioning among patients with unipolar depression: an integrative model. J. Consult. Clin.
Psychol. 68 (2000), 226–32.
114. S. E. Hobfoll, R. J. Johnson, N. Ennis, A. P. Jackson, Resource loss, resource gain, and
emotional outcomes among inner city women. J. Pers. Soc. Psychol. 84 (2003), 632–43.
115. H. R. Markus, S. Kitayama, Culture and the self: implications for cognition, emotion and
motivation. Psychol. Rev. 98 (1991), 224–53.
116. G. Marin, H. C. Triandis, Allocentrism as an important characteristic of the behavior of Latin
Americans and Hispanics. In ed. Cross-cultural and National Studies in Social Psychology,
ed. R. Diaz-Guerrero (Amsterdam: North-Holland, 1985), pp. 85–104.
117. E. A. Plant, N. Sachs-Ericsson, Racial and ethnic differences in depression: the roles of social
support and meeting basic needs. J. Consult. Clin. Psychol. 72 (2004), 41–52.
118. D. R. Williams, D. T. Takeuchi, R. K. Adair, Marital status and psychiatric disorders among
blacks and whites. J. Health Soc. Behav. 33 (1992), 140–57.
119. R. J. Taylor, C. B. Hardison, L. M. Chatters, Kin and non-kin as sources of informal assistance.
In Mental Health in Black America, ed. H. W. Neighbours, J. S. Jackson (Thousand Oaks,
CA: Sage, 1996), pp. 130–45.
120. US Department of Health and Human Sciences. Mental Health: Culture, Race and Ethnicity.
A Supplement to Mental Health: A Report of the Surgeon General (Rockville, MD: Substance
Abuse and Mental Health Services Administration, Centre for Mental Health Services, 2001).
121. C. J. Holahan, R. H. Moos, Risk, resistance, and psychological distress: a longitudinal analysis
with adults and children. J. Abnorm. Psychol. 96 (1987), 3–13.
122. N. E. Ennis, S. E. Hobfoll, K. E. Schroder, Money doesn’t talk, it swears: how economic stress
and resistance resources impact inner-city women’s depressive mood. Am. J. Community
Psychol. 28 (2000), 149–73.
123. J. D. Hovey, Religion and suicidal ideation in a sample of Latin American immigrants.
Psychol. Rep. 85 (1999), 171–7.
124. W. J. Strawbridge, S. J. Shema, R. D. Cohen, G. A. Kaplan, Religious attendance increases
survival by improving and maintaining good health behaviors, mental health, and social
relationships. Ann. Behav. Med. 23 (2001), 68–74.
125. K. S. Kendler, X. Q. Liu, C. O. Gardner, M. E. McCullough, D. Larson, C. A. Prescott,
Dimensions of religiosity and their relationship to lifetime psychiatric and substance use
disorders. Am. J. Psychiatry 160 (2003), 496–503.
126. D. Bhugra, O. Ayonrinde, Depression in migrants and ethnic minorities. Adv. Psychiatr.
Treat. 10 (2004), 13–17.
47 Psychosocial factors, depression and illness
127. A. Kleinman, Anthropology and psychiatry: the role of culture in cross-cultural research
on illness. Br. J. Psychiatry 151 (1987), 447–54.
128. N. Beale, S. Nethercott, The nature of unemployment morbidity. 2. Description. J. R. Coll.
Gen. Pract. 38 (1988), 200–202.
129. S. Platt, N. Kreitman, Long term trends in parasuicide and unemployment in Edinburgh,
1968–87. Soc. Psychiatry Psychiatr. Epidemiol. 25 (1990), 56–61.
130. R. C. Kessler, J. S. House, J. B. Turner, Unemployment and health in a community sample.
J. Health Soc. Behav. 28 (1987), 51–9.
131. R. A. Karasek, Job demands, job decision latitude and mental strain: implications for job
redesign. Admin. Sci. Q. 24 (1979), 285–306.
132. R. Karasek, T. Theorell, Healthy Work: Stress, Productivity and the Reconstruction of Working
Life (New York: Basic Books, 1990).
133. J. Siegrist, Adverse health effects of high-effort/low-reward conditions. J. Occup. Health
Psychol. 1 (1996), 27–41.
134. D. E. Broadbent, The clinical impact of job design. Br. J. Clin. Psychol. 24 (1985), 33–44.
135. M. Estryn-Behar, M. Kaminski, E. Peigne, et al., Stress at work and mental health status
among female hospital workers. Br. J. Ind. Med. 47 (1990), 20–28.
136. E. J. Bromet, M. A. Dew, D. K. Parkinson, S. Cohen, J. E. Schwartz, Effects of occupational
stress on the physical and psychological health of women in a microelectronics plant. Soc.
Sci. Med 34 (1992), 1377–83.
137. N. Kawakami, T. Haratani, S. Araki, Effects of perceived job stress on depressive symptoms
in blue-collar workers of an electrical factory in Japan. Scand. J. Work Environ. Health 18
(1992), 195–200.
138. K. R. Parkes, C. A. Mendham, C. Von Rabenau, Social support and the demand–discretion
model of job stress: tests of additive and interactive effects in two samples. J. Voc. Behav. 44
(1994), 91–113.
139. S. A. Stansfeld, R. Fuhrer, J. Head, J. Ferrie, M. Shipley, Work and psychiatric disorder in
the Whitehall II study. J. Psychosom. Res. 43 (1997), 73–81.
140. I. Niedhammer, M. Goldberg, A. Leclerc, I. Bugel, S. David, Psychosocial factors at work
and subsequent depressive symptoms in the Gazel cohort. Scand. J. Work Environ. Health
24 (1998), 197–205.
141. Y. Mino, J. Shigemi, T. Tsuda, N. Yasuda, P. Bebbington, Perceived job stress and mental
health in precision machine workers of Japan: a 2 year cohort study. Occup. Environ. Med.
56 (1999), 41–5.
142. S. A. Stansfeld, R. Fuhrer, M. J. Shipley, M. G. Marmot, Work characteristics predict psy-
chiatric disorder: prospective results from the Whitehall II study. Occup. Environ. Med. 56
(1999), 302–7.
143. A. Weinberg, F. Creed, Stress and psychiatric disorder in healthcare professionals and hos-
pital staff. Lancet 355 (2000), 533–7.
144. B. Hesketh, G. Shouksmith, Job and non-job activities, job satisfaction and mental health
among veterinarians. J. Occup. Behav. 7 (1986), 325–9.
145. P. B. Warr, Decision latitude, job demands and employee well being. Work Stress 4 (1990),
285–94.
48 Depression and physical illness
165. S. A. Stansfeld, R. Fuhrer, Depression and coronary heart disease. In Stress and the Heart:
Psychosocial Pathways to Coronary Heart Disease, ed. S. A. Stansfeld, M. G. Marmot (London:
BMJ Books, 2002), pp. 101–23.
166. S. A. Stansfeld, J. Head, R. Fuhrer, J. Wardle, V. Cattell, Social inequalities in depressive
symptoms and physical functioning in the Whitehall II study: exploring a common cause
explanation. J. Epidemiol. Community Health 57 (2003), 361–7.
Part 2
Introduction
This chapter is concerned with the role of depression in the development of coronary
atherosclerosis and in the aetiology of coronary heart disease (CHD). Chapter 4
discusses the association between depression and prognosis in patients with existing
CHD. The relationship of depression with cardiovascular disease has been a topic of
intense research interest over the past 15 years and has driven much contemporary
thinking about how physical illness and depression are linked. It has stimulated
work on a variety of biological processes that potentially mediate the relationship,
including platelet activation [1], vascular inflammation [2], endothelial dysfunc-
tion [3], reduced baroreceptor reflex sensitivity [4], sympathovagal imbalance [5]
and neuroendocrine dysfunction [6].
The purpose of this chapter is to evaluate the strength and consistency of the
association between depression and future CHD, to evaluate the specificity of the
relationship, and to describe the biological processes that are probably involved.
The chapter is divided into five sections. The first section provides a critical review
of existing evidence from longitudinal observational studies that depression and
depressive symptoms are associated prospectively with CHD in initially healthy
adults. Studies of this topic have generated rather variable results, and so it is
worth considering what accounts for the discrepancies. The chapter then ques-
tions whether there is a specific link between depression and CHD, or whether
comparable associations are present for anxiety and more general psychological
distress. It could be that depression promotes the clinical manifestation of CHD,
rather than the long-term development of atherosclerosis, and so the relationship
of depression with subclinical atherosclerosis is the topic of the third section. Next,
the chapter discusses the time course of relationships between depressive symp-
toms and CHD, and whether depression in the days or even hours preceding acute
cardiac events is relevant. The final section of the chapter outlines the different
pathways that may translate depressive emotional experience into CHD. It touches
on several of the biological and behavioural processes discussed in detail in later
chapters, but it addresses them specifically in relation to coronary atherosclerosis
and CHD. Throughout the chapter, methodological problems in this literature are
highlighted, some of which arise from the difficulty of establishing associations
between a subjective experience such as depression and a physical illness such as
CHD, while others concern issues such as confounding and biological plausibility
that apply to studies of many risk factors.
whether it overlaps with other known risk factors. In the studies described here,
the association between depression and CHD was typically evaluated after taking
into account a variety of factors, such as age, sex, blood pressure, smoking, choles-
terol level, body mass, diabetes and socioeconomic status. However, a difficulty that
arises is that some confounding factors may actually be mediators of the association
between depression and CHD. Later chapters in this book document relationships
between depression, smoking and physical activity. Depression may also be linked
positively with risk factors such as high blood pressure [8]. If these factors are sta-
tistically co-varied, then there is a danger that associations between depression and
CHD could be attenuated.
The third difficulty arises from the nature of CHD. Clinical cardiac events are
the end product of the progressive development of coronary atherosclerosis. Post-
mortem studies of victims of accidental or violent death have shown significant
coronary atherosclerosis in adolescents and young adults, related to risk factors such
as smoking and high serum cholesterol [9]. No longitudinal studies of depression
that began even in early adult life have yet been able to rule out the possibility that
subclinical atherosclerosis was already more advanced in depressed individuals. The
impact of depression at different stages of CHD aetiology is discussed more fully
later in this chapter.
Table 3.1 summarises results from 27 longitudinal observational studies pub-
lished between 1964 and 2005. Studies were identified through systematic searches
on PubMed and PsychInfo and scrutiny of existing reviews [10–12]. Studies were
excluded if fatal CHD and myocardial infarction were not used as outcome meas-
ures. Many of these studies had several outcomes adjusted for different covariates.
The relative risks shown in Table 3.1 are those that were calculated for analyses
adjusting for the standard demographic, clinical and behavioural risk factors that
were available. Some studies included individuals with suspected or definite car-
diac disease at baseline, but the present results exclude such cases. Outcomes are
presented as relative risks with 95% confidence intervals (CI).
In purely numerical terms, the findings are weighted towards positive associa-
tions being present. Out of the 27 studies reviewed, 16 showed at least 1 significant
association between depression and future CHD, while 11 did not. It should also
be noted that some of the effect sizes in the null studies were quite substantial,
even though the confidence intervals crossed 1 [13–15]. None of these prospective
studies has shown that depressive symptoms were protective against CHD.
One representative study to find a positive effect was the analysis of the National
Health Examination Follow-up study described by Anda et al. [16]: 11.1% of the
cohort of 2832 men and women aged 45–77 years had depressed affect at baseline,
as assessed with the General Wellbeing Schedule. Depressed affect was associated
with female gender, African American ethnicity, low education, unmarried status,
Table 3.1 Depression as a predictor of future coronary heart disease (CHD)
BP, blood pressure; CI, confidence interval; CVD, cardiovascular disease; ECG, electrocardiogram; ECA, Ecological Catchment Area; EPESE, Established Populations for the Epidemiologic
Studies of the Elderly; MRFIT, Multiple Risk Factor Intervention Trial; NHANES, National Health and Nutrition Examination Survey; MMPI, Minnesota Multiphasic Personality Inventory;
DIS, Diagnositc Interview Schedule; CES-D, Center for Epidemiological Studies depression scale; GHQ, General Health Questionnaire; NF CHD, non-fatal coronary heart disease; NF MI,
non-fatal myocardial infarction; Standard risk factors, blood pressure, smoking, cholesterol; marital, marital status; SES, socioeconomic status; BMI, body mass index; PA, physical activity; NS,
not significant; Effect: + = positive association with 95% CI not crossing 1 after adjustement; 0 = association either not significant or with CI crossing 1 after adjustment.
60 Depression and physical illness
smoking and physical inactivity. Over the follow-up period of more than 12 years,
6.7% died from CHD. The rate of fatal CHD was 7.1 per 1000 person-years in
depressed people, compared with 5.2 in the non-depressed participants, with a
relative risk (RR) adjusted for other factors of 1.5 (95% CI 1.0 to 2.3).
One prospective study that has not been included in Table 3.1 because some
participants had CHD at baseline is the eight-year follow-up of a representative
sample of people aged 70 years or older in north Finland [17]. Elevated scores on
the Zung depression rating scale predicted sudden cardiac death, but not cardiac
mortality in general, or non-fatal myocardial infarction. Although a proportion of
participants had a history of myocardial infarction, chest pain and heart failure at
baseline, the multivariate hazard ratio was significant for those with high depression
scores (1.70, 95% CI 1.37 to 2.10) after these factors had been taken into account
along with gender, age, diabetes and body mass index (BMI). Similarly, in the Italian
Longitudinal Study on Aging, depressive symptoms were associated independently
with future cardiovascular mortality in men after adjusting for risk factors and
pre-existing clinical conditions [18].
Another study not included in Table 3.1 because it presented results as a case–
control analysis instead of a prospective cohort analysis is the Prospective Epidemio-
logical Study of Myocardial Infarction (PRIME) carried out in France and Northern
Ireland [19]. Depressive mood, assessed with a subscale from the Minnesota Mul-
tiphasic Personality Inventory (MMPI), was assessed in some 8600 initially healthy
men aged 50–59 years who were followed up for five to six years. A total of 335 men
with a first CHD event (angina, non-fatal myocardial infarction or cardiac death)
were compared with 670 controls matched by age and date of baseline assessment.
The odds ratio for the association of CHD with depressive mood was 1.50 (95% CI
1.04 to 2.15) after adjusting for smoking, cholesterol, BMI, education and marital
status.
There are no clear distinctions between the studies that showed positive and null
effects in terms of age of the samples, gender or length of follow-up. The same
measures of depression, such as the Center for Epidemiological Studies Depres-
sion (CES-D) scale generated both positive and null findings [14,20–25]. However,
two analyses from the same British sample involving the less established measure
of depression from the Crown–Crisp Experiential Index generated null findings
[13,26]. One factor that might be important is the intensity of depression. It is
interesting that most studies that involved clinical diagnoses or treatment-based
criteria as opposed to measures on questionnaires reported positive associations
[27–29]. Pratt et al. [29] showed a larger relative risk for major depression than
dysphoria, and an intensity effect was also found in the Amsterdam Longitudi-
nal Aging Study [22]. Similarly, Everson et al. [30] analysed the related construct
61 Depression and the development of coronary heart disease
of hopelessness in the Kuopio study from Finland and observed a higher risk of
cardiac death over six years in participants with severe compared with moderate
hopelessness (relative risk 3.66 and 2.52, respectively).
A related construct to depression is vital exhaustion, a syndrome of fatigue,
tiredness and hopelessness. Vital exhaustion has been associated with CHD in cross-
sectional and longitudinal studies [31]. For example, Prescott et al. [32] tracked
9563 Danish adults for up to six years. The risk of a first myocardial infarction
requiring hospitalisation in apparently healthy individuals at baseline was 2.20
(95% CI 1.53 to 3.17) for those with high vital exhaustion ratings, after adjustment
for age, blood pressure, cholesterol, smoking, BMI, waist/hip ratio, diabetes, family
history of CHD, physical activity, alcohol consumption, education and household
income.
A positive though rather mixed picture for the relationship between depression
and future CHD emerges from this survey. Other reviews that have been more select-
ive in their inclusion criteria have perhaps come to even more positive conclusions
[10–12]. Nonetheless, the inconsistencies in this literature are substantial.
Age
Country/ range Follow-up Distress Relative risk
Ref. study n (years) (years) measure Outcome Adjustments, co-variates (95% CI) Effect Issues
Haines et al. Northwick 1457 40–64 10 Phobic anxiety Fatal CHD, NF Age, sex, BMI, standard Fatal CHD: 3.77 +
[13] Park Heart MI risk factors, (1.64–8.64)
Study, UK fibrinogen, SES NF MI: 1.26 0
(0.62–2.54)
Eaker et al. Framingham, 749 45–64 20 (max.) Anxiety Fatal CHD, NF Age, standard risk 7.8 (1.9–32.3) + Only women
[159] USA MI factors, BMI, diabetes studied; effect in
homemakers
but not
employed
women
Kawachi Health 33 999 42–77 2 Phobic anxiety Fatal CHD, NF Age, smoking, BMI, Fatal CHD: 2.45 + Short follow-up
et al. [46] Professionals, MI diabetes, (1.0–5.96)
USA hypertension, alcohol, NF MI: 0.89 0
hypercholestero- (0.45–1.79)
laemia, exercise,
family history
Kubzansky Normative 1759 21–80 20 (max.) Worry about Fatal CHD, NF Age, standard risk Fatal CHD: 0.81 0 Only men studied
et al. Aging Study, social MI factors, BMI, alcohol, (0.45–1.44)
[160] USA conditions family history NF MI: 2.41 +
(1.40–4.13)
Haines et al. Northwick 1408 40–64 20.9 Phobic anxiety Fatal CHD Age, standard risk 1-point increase on 0 Only men studied
[26] Park Heart factors, BMI, alcohol, scale: 1.07
Study, UK fibrinogen, SES (0.99–1.15)
Stansfeld Whitehall II 7081 35–55 5 Distress ECG ischaemia Age, smoking, BMI, Men: 1.47 (1.0–2.1) + Effects seen only in
et al. [44] study, UK diet, alcohol, physical Women: 0.68 0 men
activity, SES (0.4–1.3)
Robinson UK 4501 18–75 8 (max.) Distress Fatal CHD Age, gender, smoking, 1.90 (1.08–3.35) +
et al. [45] deprivation, pain
Rasul et al. Midspan 4941 45–64 5 Distress Fatal CHD Age, standard risk Men:1.29 0 Small number of
[47] study, UK factors, marital status,(0.6–2.98) deaths
SES, BMI, glucose Women: 0.62 0
(0.4–2.36)
Albert et al. Nurses 72 359 30–55 12 Phobic anxiety Fatal CHD, NF Age, smoking, BMI, PA, Fatal CHD: 1.30 + Only women
[48] Health study, MI hypertension, family (1.02–1.67) NF studied
USA history, hypercholes- MI: 0.88 0
terolaemia, alcohol, (0.76–1.01)
menopausal status,
aspirin
Rasul et al. Midspan 6575 45–64 5 Distress Hospital Age, standard risk Men: 1.61 +
[49] study, UK admission for factors, marital status, (1.02–2.55) 0
CHD SES, BMI, glucose, Women: 1.37
physical illness (0.59–3.19)
BP, blood pressure; CI, confidence interval; CVD, cardiovascular disease; ECG, electrocardiogram; ECA, Ecological Catchment Area; EPESE, Established Populations for the
Epidemiologic Studies of the Elderly; MRFIT, Multiple Risk Factor Intervention Trial; NHANES, National Health and Nutrition Examination Survey; MMPI, Minnesota
Multiphasic Personality Inventory; DIS, Diagnositc Interview Schedule; CES-D, Center for Epidemiological Studies depression scale; GHQ, General Health Questionnaire;
NF CHD, non-fatal coronary heart disease; NF MI, non-fatal myocardial infarction; Standard risk factors, blood pressure, smoking, cholesterol; marital, marital status; SES,
socioeconomic status; BMI, body mass index; PA, physical activity; NS, not significant; Effect: + = positive association with 95% CI not crossing 1 after adjustement; 0 =
association either not significant or with CI crossing 1 after adjustment.; Phobic anxiety, Crown–Crisp Experiential Index; anxiety, Framingham measure; worry about social
conditions, Normative Aging Study measure; distress, General Health Questionnaire.
66 Depression and physical illness
findings were reported from a study of 726 French men and women aged 59–71
years [65]. Participants had no history of CHD, either at baseline or two years later.
But men with high anxiety scores at both time points showed greater progression
in carotid IMT over a four-year period than did less anxious individuals, with a
similar but weaker effect among women. Cross-sectionally, psychological strain
was associated with carotid plaque in a large German sample independently of
standard risk factors, body mass, alcohol consumption and family history [66]. By
contrast, hostile attitudes but not anxiety were associated with carotid IMT in a
smaller study of postmenopausal women [67]. Thus, although the evidence is not
yet comprehensive, there is reason to suppose that depression and allied negative
mood states are associated not only with manifest CHD but also with subclinical
disease.
previous 6 months was 4.33 (95% CI 3.39 to 6.11). These findings suggest that
depression might act as an acute precipitant of cardiac pathology in vulnerable
individuals.
position, which can also influence inflammatory markers. Small studies of selected
groups of healthy adults have shown associations between depressive symptoms and
pro-inflammatory cytokines, including IL-6, IL-1 and TNF- [87,88]. Population
studies have been much less consistent, with positive associations being observed
in some studies [89–91] but not others [92–94]. There are some indications that
associations may be more pronounced in older cohorts [89,90,92]. However, if this
is the case, then it may cast doubt on inflammatory processes being mediators of
the relationship between depression and early atherosclerosis. An analysis from the
PRIME study also suggests that inflammation is only part of the story. This study
found an association between baseline CRP, IL-6 and soluble intercellular adhesion
molecule 1 (ICAM-1) and future CHD, and between these inflammatory markers
and depressive mood [19]. After adjustment for sociodemographic and cardiovas-
cular risk factors, the 50% increase in odds ratio for depression being associated
with CHD was reduced to 39% by including CRP, IL-6 and ICAM-1 in the model.
This suggests that much of the association between CHD and depressive mood
remained unaccounted for.
also play a key role, since they rapidly accumulate at the site of vascular injury, and
the extent to which platelets are prone to activation may be critical [100].
It is known that psychological stress provokes acute haemostatic responses in
healthy individuals and in patients with CHD [101]. Our group, for example, has
shown that men and women without any history of CHD show increases in fib-
rinogen, von Willebrand factor, factor VIII, plasma viscosity and platelet activation
following challenging behavioural tasks [102,103]. One study has explored the
impact of depression and anxiety on haemostatic stress responses [104]. In a small
sample of men and women with an average age of 71 years, it was found that anxiety
and depression were associated positively with D-dimer responses but not with von
Willebrand factor, PAI or tissue plasminogen activator responses. A substantial pro-
portion of participants in this study had a history of cardiovascular disease, but the
extent of coronary atherosclerosis is not known. A population study in Greece has
documented a positive relationship between depressive symptoms and fibrinogen
concentration that was independent of age, blood pressure, cholesterol, smoking,
BMI, physical activity, medication and dietary habit [91].
There is a much larger literature relating platelet function with depression. Many
of these studies were not carried out in order to assess whether depression is asso-
ciated with platelet activation but were stimulated by the use of the serotonergic
receptor on platelets as a model of brain-receptor functioning. Nevertheless, some
researchers have tested whether platelet hyperactivity is characteristic of depres-
sion. This work has been reviewed systematically by von Kanel et al. [105]. Their
conclusion is that the evidence is inconsistent: while 16 studies have demonstrated
platelet hyperactivity in at least one measure, 18 have not. This may be due in part
to the wide range of measures of platelet activity used. Modern techniques based on
flow cytometry may elicit more consistent results than in vitro aggregation meth-
ods. There is certainly evidence that antidepressants affect platelet behaviour, but
the implications of such studies for the involvement of platelets in mediating links
with CHD is not yet clear.
sympathovagal balance and baroreceptor reflex sensitivity. All have been impli-
cated in CHD. Sympathoadrenal activation causes peripheral vasoconstriction
and increased haemodynamic load, which may promote disturbed blood flow
and endothelial sheer stress [124], which in turn increase local inflammation and
endothelial gene expression [125]. Elevated heart rate has been shown to predict
future CHD independently of standard cardiovascular risk factors in a number
of studies [126,127]. Autonomic cardiac control can be assessed both with simple
temporal measures, such as heart-rate variability, and with more elaborate power
spectrum analysis. Using the latter, high-frequency power is thought to indicate
parasympathetic or vagal tone, while low-frequency power probably reflects sym-
pathovagal balance and baroreceptor reflex modulation of heart rate [128]. Low
levels of heart-rate variability and low parasympathetic tone predict future CHD in
apparently healthy individuals [129] and are strongly predictive of future morbidity
in patients following acute myocardial infarction [130].
The relationship between depression and sympathetic nervous system activity
indexed by measures such as plasma or urinary catecholamines has been studied
extensively in clinical cohorts but less extensively in population studies. Compar-
isons between clinically depressed patients and healthy individuals have shown ele-
vated sympathetic nervous system activity [131,132], but such investigations have
difficulty controlling for factors such as smoking, body mass and physical activity,
which influence neuroendocrine function. Nevertheless, a recent investigation of
a population cohort of non-smoking healthy middle-aged women demonstrated
positive correlations between 24-hour urinary excretion of noradrenaline and both
depression and anxiety symptoms, after controlling for age, race and body mass
[133]. Sympathetic activation and reduced vagal tone are also known to reduce the
threshold for cardiac ventricular fibrillation and to stimulate ventricular arrhyth-
mias [134]. However, such effects are unlikely to be involved in the association
between depression and the early aetiology and CHD, since they typically occur
among individuals who already have advanced coronary artery disease.
There is strong evidence from clinical studies that resting heart rate is higher in
depressed than non-depressed individuals without CHD [135–137]. Data concern-
ing heart-rate variability and sympathovagal balance are more mixed. No associ-
ation between heart-rate variability and depression has been observed in several
studies of depressed patients [137,138], while marginal differences have been shown
in others [135,136]. One group has found reduced cardiac baroreceptor reflex sen-
sitivity in patients with treated depression [139], while another group showed that
cardiac baroreflex sensitivity was reduced only in depressed patients who were also
anxious [140]. Horsten et al. [141] described a population study of 300 healthy
women in whom the ratio of low- to high-frequency power in the 24-hour ECG
power spectrum was related negatively to depressive symptoms, but no differences
75 Depression and the development of coronary heart disease
were found in four other spectral analytic measures. By contrast, a larger population
study of 2627 postmenopausal women showed that on several different measures,
heart-rate variability was reduced in those with depressive symptoms [142]. In
CHD patients, depression has been associated with reduced heart-rate variability
and baroreceptor reflex sensitivity in some studies [4,5] but not others [143]. The
evidence relating sympathovagal balance, depression and CHD is, therefore, quite
mixed. This mechanism has become increasingly difficult to study in clinical sam-
ples because of the widespread use of beta-blockade, and so the inconsistencies in
the literature may be difficult to reconcile in the future.
Conclusions
The consensus of longitudinal observational studies discussed in this chapter indi-
cates that depression predicts future CHD in apparently healthy individuals, with
some evidence of a relationship with symptom severity. It must be acknowledged
that the results from different studies are quite variable. Considering that depres-
sion is often assessed on only a single occasion and that CHD is influenced by so
many other factors, it is remarkable that positive associations have been recorded
with any consistency at all. There is also reasonable evidence that other indicators of
psychological distress are predictive of future disease, although again findings have
been mixed. Given the present state of the evidence, it is prudent to conclude that
depression does not have an exclusive association with CHD and that other mani-
festations of psychological distress may also be significant. It should be borne in
mind that several other aspects of psychosocial adversity, including chronic work
stress and social isolation, have been related prospectively with CHD [42,144].
76 Depression and physical illness
The extent to which such effects are mediated through symptoms of psychological
distress or are independent of affective state has not been clarified fully.
The time course of associations between depression or psychological distress and
future CHD is understood poorly, due primarily to the use of single snapshots of
psychological state. It could be that depression operates at several levels on aetiology,
from the long-term development of atherosclerosis, through acceleration of disease
development in more advanced cases, to acute triggering of plaque rupture and the
development of acute coronary syndromes.
Depression and psychological distress appear to be associated with subclinical
atherosclerosis, and not only with the clinical manifestations of disease. This is
important, since it has implications for where the search for mediating mechanisms
might be focused. However, the evidence to date is largely cross-sectional, and
longitudinal studies of subclinical atherosclerosis are urgently required.
There is no dearth of candidate mechanisms that might translate depression into
CHD risk, but convincing evidence is sparse. Studies are needed that decompose
the depression–CHD link and compute the proportion of variance accounted for
by different factors. A similar exercise has been conducted fruitfully in trying to
understand the factors that contribute to the socioeconomic gradient in CHD. In
a number of studies, between 45% and 60% of the social gradient in CHD has
been accounted for by smoking, alcohol consumption, physical activity and body
mass [145–147]. It is likely that both behavioural and psychobiological pathways
are involved in the link with depression. But a striking feature of the studies of
potential biological mediators that have been carried out to date is that evidence is
stronger for their involvement in advanced CHD than in the early development of
coronary atherosclerosis. Findings linking autonomic dysregulation or heightened
vascular inflammatory markers with depression in younger individuals without
apparent CHD are sparse. Impaired endothelial function may be a stronger candi-
date for involvement in the long-term influence of depression on disease risk, since
it is an early event in atherosclerosis and can modify the association between risk
factors and disease development [148]. Heightened sympathoadrenal activity may
also be relevant from an early stage. However, perhaps we should not expect the
association with any single potentially mediating mechanism to be very strong,
since a combination of factors is most likely to be involved.
Acknowledgements
I am grateful to Emily Williams for helping to assemble the literature reviewed in this
chapter, and to Amanda Nicholson for her helpful comments on the manuscript.
This research was supported by the British Heart Foundation.
77 Depression and the development of coronary heart disease
REFERENCES
1. C. B. Nemeroff, D. L. Musselman, Are platelets the link between depression and ischemic
heart disease? Am. Heart J. 140 (2000), 57–62.
2. P.H. Black, L. D. Garbutt, Stress, inflammation and cardiovascular disease. J. Psychosom.
Res. 52 (2002), 1–23.
3. S. Rajagopalan, R. Brook, M. Rubenfire, et al., Abnormal brachial artery flow-mediated
vasodilation in young adults with major depression. Am. J. Cardiol. 88 (2001), 196–8, A7.
4. L. L. Watkins, P. Grossman, Association of depressive symptoms with reduced baroreflex
cardiac control in coronary artery disease. Am. Heart J. 137 (1999), 453–7.
5. R. M. Carney, J. A. Blumenthal, P. K. Stein, et al., Depression, heart rate variability, and
acute myocardial infarction. Circulation 104 (2001), 2024–8.
6. E. S. Brown, F. P. Varghese, B. S. McEwen, Association of depression with medical illness:
does cortisol play a role? Biol. Psychiatry 55 (2004), 1–9.
7. A. Rosengren, S. Hawken, S. Ounpuu, et al., Association of psychosocial risk factors with
risk of acute myocardial infarction in 11119 cases and 13648 controls from 52 countries
(the INTERHEART study): case–control study. Lancet 364 (2004), 953–62.
8. K. Davidson, B. S. Jonas, K. E. Dixon, J. H. Markovitz, Do depression symptoms predict
early hypertension incidence in young adults in the CARDIA study? Coronary Artery Risk
Development in Young Adults. Arch. Intern. Med. 160 (2000), 1495–500.
9. A. W. Zieske, G. T. Malcom, J. P. Strong, Natural history and risk factors of atherosclerosis
in children and youth: the PDAY study. Pediatr. Pathol. Mol. Med. 21 (2002), 213–37.
10. R. Rugulies, Depression as a predictor for coronary heart disease: a review and meta-analysis.
Am. J. Prev. Med. 23 (2002), 51–61.
11. H. S. Lett, J. A. Blumenthal, M. A. Babyak, et al., Depression as a risk factor for coronary
artery disease: evidence, mechanisms, and treatment. Psychosom. Med. 66 (2004), 305–15.
12. H. Hemingway, H. Kuper, M. Marmot, Psychosocial factors in the primary and secondary
prevention of coronary heart disease: an updated systematic review of prospective cohort
studies. In Evidence-Based Cardiology, 2nd edn, ed. S. Yusuf, J. A. Cairns, A. J. Camm, E. L.
Fallen, B. J. Gersh (London: BMJ Books, 2003), pp. 181–218.
13. A. P. Haines, J. D. Imeson, T. W. Meade, Phobic anxiety and ischaemic heart disease. Br.
Med. J. 295 (1987), 297–9.
14. C. F. Mendes de Leon, H. M. Krumholz, T. S. Seeman, et al., Depression and risk of coro-
nary heart disease in elderly men and women: New Haven EPESE, 1982–1991. Established
Populations for the Epidemiologic Studies of the Elderly. Arch. Intern. Med. 158 (1998),
2341–8.
15. H. D. Sesso, I. Kawachi, P. S. Vokonas, D. Sparrow, Depression and the risk of coronary
heart disease in the Normative Aging Study. Am. J. Cardiol. 82 (1998), 851–6.
16. R. Anda, D. Williamson, D. Jones, et al., Depressed affect, hopelessness, and the risk of
ischemic heart disease in a cohort of U.S. adults. Epidemiology 4 (1993), 285–94.
17. H. Luukinen, P. Laippala, H. V. Huikuri, Depressive symptoms and the risk of sudden cardiac
death among the elderly. Eur. Heart J. 24 (2003), 2021–6.
78 Depression and physical illness
18. C. Marzari, S. Maggi, E. Manzato, et al., Depressive symptoms and development of coronary
heart disease events: the Italian longitudinal study on aging. J. Gerontol. A Biol. Sci. Med.
Sci. 60 (2005), 85–92.
19. J. P. Empana, D. H. Sykes, G. Luc, et al., Contributions of depressive mood and circulating
inflammatory markers to coronary heart disease in healthy European men: the Prospective
Epidemiological Study of Myocardial Infarction (PRIME). Circulation 111 (2005), 2299–
305.
20. A. A. Ariyo, M. Haan, C. M. Tangen, et al., Depressive symptoms and risks of coronary
heart disease and mortality in elderly Americans: Cardiovascular Health Study Collaborative
Research Group. Circulation 102 (2000), 1773–9.
21. B. W. Penninx, J. M. Guralnik, C. F. Mendes, de Leon, et al., Cardiovascular events and
mortality in newly and chronically depressed persons > 70 years of age. Am. J. Cardiol. 81
(1998), 988–94.
22. B. W. Penninx, A. T. Beekman, A. Honig, et al., Depression and cardiac mortality: results
from a community-based longitudinal study. Arch. Gen. Psychiatry 58 (2001), 221–7.
23. S. Wassertheil-Smoller, W. B. Applegate, K. Berge, et al., Change in depression as a precursor
of cardiovascular events. SHEP Cooperative Research Group (Systolic Hypertension in the
Elderly). Arch. Intern. Med. 156 (1996), 553–61.
24. S. Wassertheil-Smoller, S. Shumaker, J. Ockene, et al., Depression and cardiovascular seque-
lae in postmenopausal women: the Women’s Health Initiative (WHI). Arch. Intern. Med.
164 (2004), 289–98.
25. P. J. Rowan, D. Haas, J. A. Campbell, D. R. Maclean, K. W. Davidson, Depressive symp-
toms have an independent, gradient risk for coronary heart disease incidence in a random,
population-based sample. Ann. Epidemiol. 15 (2005), 316–20.
26. A. Haines, J. Cooper, T. W. Meade, Psychological characteristics and fatal ischaemic heart
disease. Heart 85 (2001), 385–9.
27. A. Aromaa, R. Raitasalo, A. Reunanen, et al., Depression and cardiovascular diseases. Acta
Psychiatr. Scand. Suppl. 377 (1994), 77–82.
28. D. E. Ford, L. A. Mead, P. P. Chang, et al., Depression is a risk factor for coronary artery
disease in men: the precursors study. Arch Intern. Med. 158 (1998), 1422–6.
29. L. A. Pratt, D. E. Ford, R. M. Crum, H. K. Armenian, J. J. Gallo, W. W. Eaton, Depres-
sion, psychotropic medication, and risk of myocardial infarction: prospective data from the
Baltimore ECA follow-up. Circulation 94 (1996), 3123–9.
30. S. A. Everson, D. E. Goldberg, G. A. Kaplan, et al., Hopelessness and risk of mortality and
incidence of myocardial infarction and cancer. Psychosom. Med. 58 (1996), 113–21.
31. A. Appels, P. Mulder, Excess fatigue as a precursor of myocardial infarction. Eur. Heart J. 9
(1988), 758–64.
32. E. Prescott, C. Holst, M. Gronbaek, et al., Vital exhaustion as a risk factor for ischaemic
heart disease and all-cause mortality in a community sample: a prospective study of 4084
men and 5479 women in the Copenhagen City Heart Study. Int. J. Epidemiol. 32 (2003),
990–97.
33. B. Norton, L. J. Whalley, Mortality of a lithium-treated population. Br. J. Psychiatry 145
(1984), 277–82.
79 Depression and the development of coronary heart disease
68. F. Lesperance, N. Frasure-Smith, M. Talajic, Major depression before and after myocardial
infarction: its nature and consequences. Psychosom. Med. 58 (1996), 99–110.
69. C. M. Dickens, L. McGowan, C. Percival, et al., Lack of a close confidant, but not depression,
predicts further cardiac events after myocardial infarction. Heart 90 (2004), 518–22.
70. P. C. Strike, A. Steptoe, Behavioral and emotional triggers of acute coronary syndromes: a
systematic review and critique. Psychosom. Med. 67 (2005), 179–86.
71. M. Naghavi, P. Libby, E. Falk, et al., From vulnerable plaque to vulnerable patient: a call for
new definitions and risk assessment strategies: part I. Circulation 108 (2003), 1664–72.
72. M. Maclure, M. A. Mittleman, Should we use a case-crossover design? Ann. Rev. Public
Health 21 (2000), 193–221.
73. M. A. Mittleman, M. Maclure, J. B. Sherwood, et al., Triggering of acute myocardial infarction
onset by episodes of anger. Circulation 92 (1995), 1720–25.
74. S. N. Willich, M. Lewis, H. Lowel, et al., Physical exertion as a trigger of acute myocardial
infarction. N. Engl. J. Med. 329 (1993), 1684–90.
75. A. Steptoe, D. L. Whitehead, Depression, stress and coronary heart disease: the need for
more complex models. Heart 91 (2005), 419–20.
76. A. K. Ferketich, J. A. Schwartzbaum, D. J. Frid, M. L. Moeschberger, Depression as an
antecedent to heart disease among women and men in the NHANES I study. Arch. Intern.
Med. 160 (2000), 1261–8.
77. S. A. Everson, G. A. Kaplan, D. E. Goldberg, J. T. Salonen, Hypertension incidence is predicted
by high levels of hopelessness in Finnish men. Hypertension 35 (2000), 561–7.
78. R. Ross, Atherosclerosis: an inflammatory disease. N. Engl. J. Med. 340 (1999), 115–26.
79. M. J. Jarvisalo, A. Harmoinen, M. Hakanen, et al., Elevated serum C-reactive protein levels
and early arterial changes in healthy children. Arterioscler. Thromb. Vasc. Biol. 22 (2002),
1323–8.
80. O. T. Raitakari, M. Juonala, M. Kahonen, et al., Cardiovascular risk factors in childhood
and carotid artery intima-media thickness in adulthood: the Cardiovascular Risk in Young
Finns Study. J. Am. Med. Assoc. 290 (2003), 2277–83.
81. W. E. Chrapko, P. Jurasz, M. W. Radomski, et al., Decreased platelet nitric oxide synthase
activity and plasma nitric oxide metabolites in major depressive disorder. Biol. Psychiatry
56 (2004), 129–34.
82. A. J. Broadley, A. Korszun, C. J. Jones, M. P. Frenneaux, Arterial endothelial function is
impaired in treated depression. Heart 88 (2002), 521–3.
83. L. Ghiadoni, A. Donald, M. Cropley, et al., Mental stress induces transient endothelial
dysfunction in humans. Circulation 102 (2000), 2473–8.
84. M. Cesari, B. W. Penninx, A. B. Newman, et al., Inflammatory markers and onset of
cardiovascular events: results from the Health ABC study. Circulation 108 (2003), 2317–
22.
85. L. Lind, Circulating markers of inflammation and atherosclerosis. Atherosclerosis 169 (2003),
203–14.
86. P. M. Ridker, C. H. Hennekens, J. E. Buring, N. Rifai, C-reactive protein and other markers
of inflammation in the prediction of cardiovascular disease in women. N. Engl. J. Med. 342
(2000), 836–43.
82 Depression and physical illness
104. R. von Kanel, Platelet hyperactivity in clinical depression and the beneficial effect of antide-
pressant drug treatment: how strong is the evidence? Acta Psychiatr. Scand. 110 (2004),
163–77.
105. R. von Kanel, J. E. Dimsdale, K. A. Adler, et al., Effects of depressive symptoms and anxiety
on hemostatic responses to acute mental stress and recovery in the elderly. Psychiatry Res.
126 (2004), 253–64.
106. T. Ljung, B. Andersson, B. A. Bengtsson, P. Bjorntorp, P. Marin, Inhibition of cortisol
secretion by dexamethosone in relation to body fat distribution. Obes. Res. 4 (1996), 277–
82.
107. R. Pasquali, B. Ambrosi, D. Armanini, et al., Cortisol and ACTH response to oral dexametha-
sone in obesity and effects of sex, body fat distribution, and dexamethasone concentrations:
a dose–response study. J. Clin. Endocrinol. Metab. 87 (2002), 166–75.
108. V. Vicennati, R. Pasquali, Abnormalities of the hypothalamic-pituitary-adrenal axis in
nondepressed women with abdominal obesity and relations with insulin resistance: evi-
dence for a central and a peripheral alteration. J. Clin. Endocrinol. Metab. 85 (2000),
4093–8.
109. J. R. Seckl, N. M. Morton, K. E. Chapman, B. R. Walker, Glucocorticoids and 11beta-
hydroxysteroid dehydrogenase in adipose tissue. Recent Prog. Horm. Res. 59 (2004),
359–93.
110. E. S. Epel, B. McEwen, T. Seeman, et al., Stress and body shape: stress-induced cortisol
secretion is consistently greater among women with central fat. Psychosom. Med. 62 (2000),
623–32.
111. A. Steptoe, S. R. Kunz-Ebrecht, L. Brydon, J. Wardle, Central adiposity and cortisol responses
to waking in middle-aged men and women. Int. J. Obes. Relat. Metab. Disord. 28 (2004),
1168–73.
112. E. J. Brunner, H. Hemingway, B. R. Walker, et al., Adrenocortical, autonomic, and inflamma-
tory causes of the metabolic syndrome: nested case–control study. Circulation 106 (2002),
2659–65.
113. J. I. Hudson, M. S. Hudson, A. J. Rothschild, et al., Abnormal results of dexamethasone
suppression tests in nondepressed patients with diabetes mellitus. Arch. Gen. Psychiatry 41
(1984), 1086–9.
114. M. Roy, B. Collier, A. Roy, Hypothalamic-pituitary-adrenal axis dysregulation among dia-
betic outpatients. Psychiatry Res. 31 (1990), 31–7.
115. R. Rosmond, S. Wallerius, P. Wanger, et al., A 5-year follow-up study of disease incidence
in men with an abnormal hormone pattern. J. Intern. Med. 254 (2003), 386–90.
116. D. I. Phillips, D. J. Barker, C. H. Fall, et al., Elevated plasma cortisol concentrations: a link
between low birth weight and the insulin resistance syndrome? J. Clin. Endocrinol. Metab.
83 (1998), 757–60.
117. H. W. Cohen, S. Madhavan, M. H. Alderman, History of treatment for depression: risk factor
for myocardial infarction in hypertensive patients. Psychosom. Med. 63 (2001), 203–9.
118. D. A. Lawlor, G. D. Smith, S. Ebrahim, Association of insulin resistance with depression:
cross sectional findings from the British Women’s Heart and Health Study. Br. Med. J. 327
(2003), 1383–4.
84 Depression and physical illness
136. M. W. Agelink, C. Boz, H. Ullrich, J. Andrich, Relationship between major depression and
heart rate variability: clinical consequences and implications for antidepressive treatment.
Psychiatry Res. 113 (2002), 139–49.
137. A. C. Volkers, J. H. Tulen, W. W. van den Broek, et al., Motor activity and autonomic cardiac
functioning in major depressive disorder. J. Affect. Disord. 76 (2003), 23–30.
138. V. K. Yeragani, R. Pohl, R. Balon, et al., Heart rate variability in patients with major depres-
sion. Psychiatry Res. 37 (1991), 35–46.
139. A. J. Broadley, M. P. Frenneaux, V. Moskvina, C. J. H. Jones, A. Korszun, Baroreflex sensitivity
is reduced in depression. Psychosom. Med. 67 (2005), 648–51.
140. L. L. Watkins, P. Grossman, R. Krishnan, J. A. Blumenthal, Anxiety reduces baroreflex
cardiac control in older adults with major depression. Psychosom. Med. 61 (1999), 334–40.
141. M. Horsten, M. Ericson, A. Perski, et al., Psychosocial factors and heart rate variability in
healthy women. Psychosom. Med. 61 (1999), 49–57.
142. C. K. Kim, S. P. McGorray, B. A. Bartholomew, et al., Depressive symptoms and heart rate
variability in postmenopausal women. Arch. Intern. Med. 165 (2005), 1239–44.
143. A. Gehi, D. Mangano, S. Pipkin, W. S. Browner, M. A. Whooley, Depression and heart rate
variability in patients with stable coronary heart disease: findings from the Heart and Soul
Study. Arch. Gen. Psychiatry 62 (2005), 661–6.
144. H. Hemingway, M. Marmot, Evidence based cardiology: psychosocial factors in the aetiology
and prognosis of coronary heart disease – systematic review of prospective cohort studies.
Br. Med. J. 318 (1999), 1460–67.
145. J. W. Lynch, G. A. Kaplan, R. D. Cohen, J. Tuomilehto, J. Salonen, Do cardiovascular risk
factors explain the relation between socio-economic status, risk of all-cause mortality, car-
diovascular mortality, and acute myocardial infarction? Am. J. Epidemiol. 144 (1996), 934–
42.
146. F. J. Van Lenthe, E. Gevers, I. M. Joung, H. Bosma, J. P. Mackenbach, Material and behav-
ioral factors in the explanation of educational differences in incidence of acute myocardial
infarction: the Globe study. Ann. Epidemiol. 12 (2002), 535–42.
147. S. P. Wamala, M. A. Mittleman, K. Schenck-Gustafsson, K. Orth-Gomer, Potential explana-
tions for the educational gradient in coronary heart disease: a population-based case–control
study of Swedish women. Am. J. Public Health 89 (1999), 315–21.
148. M. Juonala, J. S. Viikari, T. Laitinen, et al., Interrelations between brachial endothelial
function and carotid intima-media thickness in young adults: the Cardiovascular Risk in
Young Finns Study. Circulation 110 (2004), 2918–23.
149. A. M. Ostfeld, B. Z. Lebovits, R. B. Shekelle, O. Paul, A prospective study of the relationship
between personality and coronary heart disease. J. Chronic Dis. 17 (1964), 265–76.
150. T. Hallstrom, L. Lapidus, C. Bengtsson, K. Edstrom, Psychosocial factors and risk of
ischaemic heart disease and death in women: a twelve-year follow-up of participants in
the population study of women in Gothenburg, Sweden. J. Psychosom. Res. 30 (1986), 451–
9.
151. J. C. Barefoot, M. Schroll, Symptoms of depression, acute myocardial infarction, and total
mortality in a community sample. Circulation 93 (1996), 1976–80.
86 Depression and physical illness
Coronary heart disease (CHD) is the leading cause of death in the USA and Europe
[1,2]. In roughly half the cases, the first clinical manifestations of CHD – myocardial
infarction (MI) or sudden death – are catastrophic. These events are sudden, unex-
pected and unpredictable. The economic cost of CHD is growing. For example, in
the USA over $130 billion is spent on CHD each year in direct medical costs, dis-
ability payments and lost productivity [2]. Moreover, traditional risk factors such
as cigarette smoking, hyperlipidaemia and hypertension do not account fully for
the timing and occurrence of these events.
Depression is also a major health problem. It is associated with significant impair-
ment of function, which may, at times, be worse than that of chronic medical dis-
orders [3]. Depressive symptoms have been correlated with the presence of one or
more chronic diseases [4,5], as well as inability to work [6], days in bed or days away
from normal activities [4], increased mortality risk [7], increased use of medical
services [8], and decreased wellbeing and lowered functioning [3]. Major depres-
sive disorder (MDD) is the most prevalent of all psychiatric disorders, affecting
up to 25% of women and 12% of men during their lifetime [9]. Since 1950, the
prevalence of depression has increased significantly [10].
Depression is disproportionately prevalent among cardiac patients, with esti-
mates of MDD of about 15% in patients following acute myocardial infarction
(AMI) or coronary artery bypass graft (CABG), and an additional 20% with
either minor depression or elevated levels of depressive symptoms as measured
by questionnaires such as the Beck Depression Inventory (BDI) [11–17]. In addi-
tion, evidence has suggested that depression is a significant and independent risk
factor for patients with CHD and may also be associated with increased ‘cardio-
vascular vulnerability’, a term that has been used to describe patients susceptible
to acute coronary events based upon plaque, blood or myocardial characteristics
[18,19]. Based on this evidence, there has been a great deal of interest into potential
Depression and Physical Illness, ed. A. Steptoe.
Published by Cambridge University Press. C Cambridge University Press 2006.
Table 4.1 Studies assessing the effect of depression on outcomes in coronary heart disease (CHD) Samples
Carney et al. [27] 52/22 cardiac events CHD 12 months Cardiac event RR = 2.2
Frasure-Smith et al. [145] 222/12 deaths MI 6 months Cardiac mortality DIS : HR = 4.29
Frasure-Smith et al. [105] 222/21 deaths 18 months DIS NS; BDI OR = 6.64
Frasure-Smith et al. [20] 896/39 deaths MI 1 year Cardiac mortality Men OR = 3.05, women OR = 3.29
Lesperance et al. [146] 896/155 deaths 5 year HR = 3.13–3.17
Lane et al. [22] 288/25 deaths MI 4 months Mortality NS
Lane et al. [23] 288/28 deaths 1 year Mortality NS
Lane et al. [24] 288/31 deaths CHD events NS
Lane et al. [25] 288/38 deaths 3 years Mortality NS
Ahern et al. [147] 265/NA MI Varying Mortality, cardiac RR = 1.38
arrest
Jenkinson et al. [148] 1376/247 deaths MI 3 years Mortality NS
Ladwig et al. [149] 560/12 deaths, 17 MI 6 months Cardiac mortality NS
arrythmic events Arrhythmic event
Welin et al. [49] 275/167 deaths MI 10 years Cardiac mortality RR = 3.16
Bush et al. [150] 144/17 deaths MI 4 months Mortality RR = 3.5
Kaufmann et al. [151] 331/15 deaths MI 6 months Mortality NS p = 0.04 (effect size not
331/33 deaths 12 months reported)
Irvine et al. [152] 318/number of events MI 2 years Sudden cardiac death RR = 2.45
not reported
Strik et al. [153] 318/25 cardiac events MI Mean 3.4 years Mortality or HR = 2.32
reinfarction
Strik et al. [154] 206/16 events MI 3 years Mortality or AMI NS
Horsten et al. [155] 292/81 deaths Acute CHD 5 years Mortality or cardiac RR = 1.9
event (MI or event
angina)
Lesperance et al. [156] 430/16 deaths, 28 Unstable 1 year Cardiac mortality OR = 6.73
events angina or AMI
Connerney et al. [29] 309/8 deaths, 42 events CABG 1 year Cardiac event, Cardiac events, RR = 2.3;
mortality mortality NS; BDI = NS
Baker et al. [30] 158/6 deaths CABG Median 24 months Mortality OR = 6.24
Saur et al. [31] 416/NA CABG 1 year Mortality NS
Blumenthal et al. [37] 817/122 deaths CABG Mean 5.2 years Mortality Moderate to severe depression,
HR = 2.84; persistent depression,
HR = 2.33
Burg et al. [36] 89/7 deaths CABG 2 years Cardiac mortality OR = 23.16
BDI, Beck Depression Inventory; DIS, Diagnostic Interview Schedule; HR, hazard ratio; MI, myocardial infarction; NS, non-significant; OR, odds ratio;
RR, relative risk.
Reprinted from Psychosom. Med. 66 (2004), 305–15, with permission from Lippincott Williams & Wilkins.
90 Depression and physical illness
treatments for depression in cardiac patients, with the hope that successful mental
health treatment might also have a favourable impact on cardiovascular and physical
health outcomes (see Chapter 5). An understanding of the mechanisms underlying
the relationship between depression and CHD will likely help to guide the most
effective interventions. This chapter describes the evidence that depression is a
risk factor in patients with established CHD and suggests potential mechanisms
underlying the relationship between depression and adverse outcomes.
Although there have been some null findings for MI patients, most have been
from small studies with methodological limitations such as limited follow-up or
inadequate assessment of depression [14]. For example, in a series of studies assess-
ing the relation of depression to outcome in 288 AMI patients at varying follow-up
times, Lane et al. [22–25] reported that depression (assessed by the BDI) was not
related to cardiac or all-cause mortality at 4 months’, 1 year’s or 3 years’ follow-up;
nor was it related to cardiac events at one year. However, a notable limitation of
this series of studies is the small sample size and event rate, which can yield highly
unstable estimates [26].
Studies of patients with stable CHD have reported significant associations of
depression and clinical outcomes. Carney et al. [27] followed 52 patients for 12
months after catheterisation and found that a diagnosis of MDD was associated
with more than twice the risk of having a cardiac event, even after controlling
for other risk factors. Barefoot et al. [28] assessed 1250 patients with documented
CHD using the Zung self-report depression scale at the time of diagnostic coronary
angiography and followed patients for up to 19.4 years. Results showed that patients
with moderate to severe depression were at 69% greater risk for cardiac death and
78% greater risk for all-cause death [28].
CABG surgery is a common surgical intervention for CHD patients. Depression
rates are known to be particularly high in CABG patients, both before and imme-
diately after surgery [29–35]. There has been an increase in studies assessing the
possibility that CABG patients with depression incur an increased risk; however, to
date there are relatively few prospective studies of this type [29–31,36,37]. Conner-
ney et al. [29] followed 309 CABG patients for one year after surgery. Compared with
non-depressed patients, depressed patients, as assessed by the Diagnostic Interview
Schedule (DIS), were more than twice as likely to have a cardiac event within 12
months after surgery but were not at higher risk for mortality within the first year.
In a more recent report from Duke University Medical Center [37], the effect of
depression on mortality after CABG surgery was assessed in 817 patients followed
for up to 12 years (mean = 5.2 years). On the day before surgery, the Center for Epi-
demiologic Studies Depression (CES-D) questionnaire [38] was used to categorise
patients as having no depression (CES-D < 16), mild depression (CES-D = 16–26)
or moderate to severe depression (CES-D ≥ 27). Results indicated that moderate to
severe depression was independently associated with a two- to three-fold increased
risk of mortality even after statistically controlling for age, gender, number of grafts,
diabetes, smoking, left ventricular ejection fraction and history of AMI. Figure 4.1
depicts the unadjusted Kaplan–Meier survival curves for patients with no depres-
sion, mild depression and moderate to severe depression. Furthermore, as displayed
in Figure 4.2, it was observed that even mild depression (CES-D > 16) that persisted
for at least six months following CABG was associated with increased risk.
92 Depression and physical illness
1.0
CES-D < 16
0.8
CES-D 16–26
CES-D ≥ 27
Survival function
0.6
0.4
0.2
0.0
0 2 4 6 8 10 12
Years after surgery
Figure 4.1 Kaplan–Meier survival curves for all-cause mortality among three categories of pre-surgery
(baseline) depressive symptoms. Compared with the absence of depressive symptoms,
the presence of moderate to severe symptoms was associated with a relative hazard of
2.4 (95% confidence interval [CI] 1.40 to 4.00, p = 0.001). The risk for mild symp-
toms was not different from the absence of symptoms (hazard ratio [HR] = 1.08, 95%
CI 0.70 to 1.67, p = .723). CES-D Center for Epidemiological Studies depression scale.
Reprinted from Lancet 362 (2003), 604–9. with permission from Elsevier.
Women are at increased risk after AMI [39] and CABG surgery [29,40,41] and
are more likely to experience depression [35,42–44]. Therefore, researchers have
theorised that elevated rates of depression may be responsible for the increased risk
in women. However, evidence thus far has not provided support for this hypothesis.
It is more likely that increased medical comorbidities might place women at greater
risk for depression and adverse clinical outcomes, such as decreased functionality,
older age at the time of surgery [44], increased angina, more severe hypertension
and diabetes [45,46] and smaller coronary artery diameter [41,47]. Frasure-Smith
et al. [20] investigated the possibility that the increased risk for women is due to
higher rates of depression or gender differences in the impact of depression on
clinical events, but they failed to find any evidence in support of these possibilities.
Depression is associated with several related psychosocial risk factors that have
also been related to CHD outcomes such as negative affect, vital exhaustion [48],
decreased social support [49,50], personality factors [51,52], anger expression [53],
hostility [54], negative emotions [55,56] and anxiety [57]. The possibility that
93 Depression and prognosis in cardiac patients
1.0
Remitted at 6 months
New-onset depression
0.4
0.2
0.0
0 2 4 6 8 10
Years after surgery
Figure 4.2 Kaplan–Meier survival curves for all-cause mortality among four categories of six-month
course of depression. Using patients who were never depressed as the reference, the hazard
ratio (HR) of persistent depression was 2.2 (95% confidence interval [CI] = 1.2 to 4.2,
p = 0.015), the HR for new-onset depression was 2.2 (95% CI 0.8 to 5.8, p = 0.121) and
the HR for remitted depression was 1.0 (95% CI 0.5 to 2.0, p = 0.954). Reprinted from
Lancet 362 (2003), 604–9, with permission from Elsevier.
depression is also predictive of outcome via its association with these risk factors
has been considered and is a source of controversy in the field today. For example,
Denollet et al. [51,58] considered type D personality, defined as high negative
affectivity and high social inhibition and marked by symptoms of depression, low
emotional expression and low perceived social support. In a representative study
[58], type D personality was assessed in patients (268 men, 35 women) with a
recent coronary event. At five-year follow-up, type D personality was a significant
predictor of all-cause mortality. Depression was a significant univariate predictor of
mortality, but it was no longer associated with mortality when type D was added to
the model. In a similar manner, Dickens et al. [50] provided preliminary evidence
that social isolation, but not depression, uniquely predicts outcome for post-AMI
patients. They reported that lack of a close confidant, but not a composite measure of
depression and anxiety retrospectively reported for the week before AMI, was related
to outcome. On the other hand, Frasure-Smith and Lesperance [56] compared
measures of depression, anxiety, self-reported health, anger, stress, perceived social
support and number of close friends and relatives in predicting mortality after
94 Depression and physical illness
AMI. Exploratory factors analysis suggested that the various questionnaires tapped
the underlying constructs of overt anger, social support and negative affect. Of
these factors, only negative affect was uniquely predictive of outcome. Although
few studies have conducted a comprehensive comparison of related psychosocial
constructs, the majority that are available suggest that depression and negative
affect are the factors that emerge as most predictive of increased risk [55].
Another controversy centres around the possibility that depression is merely a
marker for disease severity, which ultimately is responsible for increased risk. In the
studies presented in Table 4.1, the association between depression and measures
of disease severity of risk is statistically controlled for in order to account for this
possibility. However, as noted by Stewart et al. [59] and Lane et al. [60], it is almost
impossible to statistically control for all measures of disease severity, risking the
possibility of inflated estimates of association between depression and outcome.
On the other hand, one must strike a balance with the possibility of overadjusting
by including variables in the model that may mediate the relationship between
depression and outcome. As shown in Table 4.1, the results of the majority of
the well-designed studies suggest that depression confers a unique risk above and
beyond any association with disease severity. In addition, studies thus far indicate
that the cognitive and affective components of depression, rather than the somatic
components, are associated most with increased risk for morbidity and mortality
[56,61]. Thus, there is ample evidence that clinical depression is associated with
risk for mortality and morbidity for patients with varying manifestations of CHD.
Additionally, in a comprehensive review, Davidson et al. [12] noted that even minor
elevations in depressive symptoms significantly increase the risk of CHD in healthy
individuals and worsen the prognosis in patients with established CHD.
Biobehavioural mechanisms
A number of biobehavioural mechanisms have been hypothesised to underlie the
relationship between depression and CHD. Thus far, most evidence for mechanisms
comes from cross-sectional studies and prospective studies that track depression,
the hypothesised mechanism, and CHD outcomes over time are needed to provide
more conclusive support. Nonetheless, there is evidence that depression is associ-
ated with traditional risk factors for CHD such as hypertension, diabetes and insulin
resistance [62,63] and markers of cardiac risk such as platelet activity [64], dysregu-
lation of the autonomic nervous system [65,66], endothelial dysfunction, and alter-
ations in the immune response/inflammation [67]. Depression is also associated
with behavioural factors that are in turn associated with CHD risk, such as treatment
adherence [68], smoking [69], heavy alcohol use and physical inactivity [70].
95 Depression and prognosis in cardiac patients
Platelet activity
Numerous pathological and clinical studies have shown that increased platelet acti-
vation plays a key role in the pathogenesis of CHD and acute coronary syndromes
[74]. Furthermore, prospective data document a direct relationship between blood
platelet concentration and aggregability and the long-term incidence of fatal CHD
events in apparently healthy men [75]. Thus, it is no surprise that antiplatelet ther-
apy with aspirin and thienopyridines has become the foundation for the treatment
and prevention of CHD [76].
Several studies have demonstrated that platelet activation is increased in patients
suffering from MDD [64,67,77], and it has been proposed that this phenomenon
may be responsible for the increased risk of cardiovascular morbidity and mortality
in depressed populations [78,79]. The role of serotonin in both platelet function
[80] and depression [81] also provides suggestive evidence linking platelet activity
and depression. Patients with MDD have been shown to exhibit alterations in
multiple platelet parameters, including increases in serotonin receptor binding sites,
glycoprotein 2B/3A and P-selectin receptors, as well as heightened levels of beta-
thromboglobulin and platelet factor 4, which correspond with increased platelet
activity [82]. Indeed, serotonin itself has been implicated as a significant contribu-
tor to the pathogenesis of acute coronary events and the conversion of chronic
stable angina into an unstable coronary syndrome [83]. In animal models of CHD
and endothelial injury, serotonin receptor antagonists provided potent protection
against repetitive platelet aggregation even in the face of high levels of circulating
catecholamines [83]. Thus, evidence suggests that platelet activity is likely to be a
mechanism linking depression and CHD outcomes, a possibility warranting further
research.
Heart-rate variability
Reduced 24-hour HRV independently predicts mortality in patients with stable
CHD [91] with a recent AMI [92] or with heart failure [93]. In each of these patient
populations, depression is prevalent and is associated with abnormally low HRV,
suggesting that low HRV may be involved in the increased risk of mortality in
depressed cardiac patients [94–98]. In support of this notion, Carney et al. [99]
found that including HRV as a co-variate in a model predicting depression-related
risk for all-cause mortality significantly reduces the hazard ratio in AMI patients.
This finding provides further evidence that low 24-hour HRV is involved in the
relationship between depression and increased risk of mortality in cardiac patients.
Endothelial dysfunction
Endothelial dysfunction plays a vital role in the development, progression and clin-
ical manifestations of atherosclerosis [106,107]. It has been related to a wide range of
cardiovascular risk factors in both adults and children [108,109]. The clinical utility
97 Depression and prognosis in cardiac patients
Inflammation
A large body of experimental and clinical evidence suggests that inflammation plays
a central role in the development and progression of atherosclerotic heart disease. C-
reactive protein (CRP), an acute-phase reactant produced primarily in hepatocytes,
is a highly sensitive marker of underlying systemic inflammation [119]. An elevated
level of CRP is an independent risk factor for AMI and stroke, with some research
showing that it may be a better predictor of a first cardiovascular event than low-
density lipoprotein (LDL) cholesterol [120]. Several theories have been advanced to
explain this relationship between elevated CRP levels and cardiovascular morbidity.
Increased levels of CRP reflect low-grade inflammation in visceral adipose tissue and
may simply be a marker for the dyslipidaemia and insulin resistance associated with
98 Depression and physical illness
the metabolic syndrome. Alternatively, increased levels of CRP may reflect a state
of vascular or endothelial inflammation, which may predispose to atherogenesis
and plaque rupture [121].
Several cross-sectional studies have demonstrated increased levels of CRP and
other inflammatory markers in patients with depression [67,122,123] and with
other CHD risk factors, such as the metabolic syndrome [124,125]. In an analysis
of data describing 6914 men and women between the ages of 18 and 39 years from
the National Health and Nutrition Examination Survey (NHANES) III, a history
of MDD was associated strongly with elevated CRP. This association was present
even after adjusting for potential confounders; it was strongest in men with recent
or recurrent episodes of depression. However, there was no significant association
between CRP and depression in women [126]. Depressed mood has also been asso-
ciated with elevated levels of inflammatory markers in cohorts of elderly subjects. In
the Health, Aging, and Body Composition study of well-functioning older people
of 70–79 years of age, interleukin (IL)-6, tumor necrosis factor alpha (TNF) and
CRP were higher in individuals with depressed mood than in non-depressed indi-
viduals [127]. Similarly, in a subset of individuals aged 60 years or older from the
Rotterdam study, IL-6 and CRP were higher in depressed people than in controls,
although the findings for CRP were not significant after adjusting for confounding
factors [128]. When considered in aggregate, these observational studies suggest
that depression is characterised by elevated levels of CRP and other inflamma-
tory markers. Whether depression causes elevated CRP or an inflammatory state
causes the depression, or whether both depression and elevated CRP are due to an
independent disease process, is unclear.
potentiate other risk factors. Panagiotakos et al. [139] demonstrated that depression,
especially when accompanied by alcohol use, physical inactivity and/or smoking,
was associated with increased risk for the development of CHD.
Conclusions
There is considerable evidence that depression in CHD patients is associated with
increased risk for cardiac morbidity and morbidity. However, as summarised by
Carney and Freedland in Chapter 5, treatments thus far have been only moderately
successful in improving depression in this population and have been largely unsuc-
cessful in improving cardiac outcomes. Therefore, identifying effective treatments
has become a priority.
A greater understanding of mechanisms linking depression and increased risk
should guide selection of treatments and will provide the best chance of improving
both depression and cardiac risk. Mechanisms for which promising evidence exists
include traditional risk factors such as obesity, hypertension, diabetes and insulin
resistance, markers of cardiac risk such as platelet activity, dysregulation of the auto-
nomic nervous system, endothelial dysfunction and inflammation, and behavioural
and lifestyle factors. Research on patient-treatment match is also likely to improve
treatment efficacy. Ultimately, it may be possible to identify patients based on demo-
graphic, disease and psychological characteristics that are most likely to respond
to particular treatments. Although it is not possible at this point to recommend
any treatment for depression over another to reduce cardiac risk, depression is
common in this population and certainly warrants treatment to improve quality of
life.
REFERENCES
Coronary Artery Bypass Graft [CABG] Biobehavioral Study). Am. J. Cardiol. 79 (1997),
1017–24.
45. E. D. Eaker, Psychosocial risk factors for coronary heart disease in women. Cardiol. Clin. 16
(1998), 103–11.
46. J. Z. Ayanian, E. Guadagnoli, P. D. Cleary, Physical and psychosocial functioning of women
and men after coronary artery bypass surgery. J. Am. Med. Assoc. 274 (1995), 1767–70.
47. N. J. O’Connor, J. R. Morton, J. D. Birkmeyer, et al., Effect of coronary artery diameter
in patients undergoing coronary bypass surgery: Northern New England Cardiovascular
Disease Study Group. Circulation 93 (1996), 652–5.
48. W. J. Kop, A. P. Appels, C. F. Mendes de Leon, H. B. de Swart, F. W. Bar, Vital exhaustion
predicts new cardiac events after successful coronary angioplasty. Psychosom. Med. 56 (1994),
281–7.
49. C. Welin, G. Lappas, L. Wilhelmsen, Independent importance of psychosocial factors for
prognosis after myocardial infarction. J. Intern. Med. 247 (2000), 629–39.
50. C. M. Dickens, L. McGowan, C. Percival, et al., Lack of a close confidant, but not depression,
predicts further cardiac events after myocardial infarction. Heart 90 (2004), 518–22.
51. J. Denollet, D. L. Brutsaert, Personality, disease severity, and the risk of long-term cardiac
events in patients with a decreased ejection fraction after myocardial infarction. Circulation
97 (1998), 167–73.
52. I. Kawachi, D. Sparrow, L. D. Kubzansky, et al., Prospective study of a self-report type A
scale and risk of coronary heart disease: test of the MMPI-2 type A scale. Circulation 98
(1998), 405–12.
53. P. Angerer, U. Siebert, W. Kothny, et al., Impact of social support, cynical hostility and
anger expression on progression of coronary atherosclerosis. J. Am. Coll. Cardiol. 36 (2000),
1781–8.
54. L. A. Chaput, S. H. Adams, J. A. Simon, et al., Hostility predicts recurrent events among
postmenopausal women with coronary heart disease. Am. J. Epidemiol. 156 (2002), 1092–9.
55. L. D. Kubzansky, I. Kawachi, Going to the heart of the matter: do negative emotions cause
coronary heart disease? J. Psychosom. Res. 48 (2000), 323–37.
56. N. Frasure-Smith, F. Lesperance, Depression and other psychological risks following
myocardial infarction. Arch. Gen. Psychiatry 60 (2003), 627–36.
57. N. Frasure-Smith, F. Lesperance, M. Talajic, The impact of negative emotions on prognosis
following myocardial infarction: is it more than depression? Health Psychol. 14 (1995),
388–98.
58. J. Denollet, J. Vaes, D. L. Brutsaert, Inadequate response to treatment in coronary heart
disease: adverse effects of type D personality and younger age on 5-year prognosis and
quality of life. Circulation 102 (2000), 630–35.
59. R. A. Stewart, F. M. North, T. M. West, et al., Depression and cardiovascular morbidity and
mortality: cause or consequence? Eur. Heart J. 24 (2003), 2027–37.
60. D. Lane, D. Carroll, G. Y. Lip, Anxiety, depression, and prognosis after myocardial infarction:
is there a causal association? J. Am. Coll. Cardiol. 42 (2003), 1808–10.
61. J. C. Barefoot, B. H. Brummett, M. J. Helms, et al., Depressive symptoms and survival of
patients with coronary artery disease. Psychosom. Med. 62 (2000), 790–95.
103 Depression and prognosis in cardiac patients
94. R. M. Carney, M. W. Rich, A. Tevelde, et al., Major depressive disorder in coronary artery
disease. Am. J. Cardiol. 60 (1987), 1273–5.
95. R. M. Carney, R. D. Saunders, K. E. Freedland, et al., Association of depression with reduced
heart rate variability in coronary artery disease. Am. J. Cardiol. 76 (1995), 562–4.
96. R. M. Carney, J. A. Blumenthal, P. K. Stein, et al., Depression, heart rate variability, and
acute myocardial infarction. Circulation 104 (2001), 2024–8.
97. G. Casolo, E. Balli, T. Taddei, J. Amuhasi, C. Gori, Decreased spontaneous heart rate vari-
ability in congestive heart failure. Am. J. Cardiol. 64 (1989), 1162–7.
98. E. P. Havranek, M. G. Ware, B. D. Lowes, Prevalence of depression in congestive heart failure.
Am. J. Cardiol. 84 (1999), 348–50.
99. R. A. Carney, J. A. Blumenthal, K. E. Freedland, et al., Low heart rate variability and the
effect of depression on post-myocardial infarction mortality. Arch. Intern. Med. 165 (2005),
1486–91.
100. M. T. La Rovere, C. Bersano, M. Gnemmi, G. Specchia, P. J. Schwartz, Exercise-induced
increase in baroreflex sensitivity predicts improved prognosis after myocardial infarction.
Circulation 106 (2002), 945–9.
101. S. H. Hohnloser, T. Klingenheben, L. A. van de, et al., Reflex versus tonic vagal activity
as a prognostic parameter in patients with sustained ventricular tachycardia or ventricular
fibrillation. Circulation 89 (1994), 1068–73.
102. G. M. De Ferrari, M. Landolina, M. Mantica, et al., Baroreflex sensitivity, but not heart rate
variability, is reduced in patients with life-threatening ventricular arrhythmias long after
myocardial infarction. Am. Heart J. 130 (1995), 473-80.
103. G. E. Billman, P. J. Schwartz, H. L. Stone, Baroreceptor reflex control of heart rate: a predictor
of sudden cardiac death. Circulation 66 (1982), 874–80.
104. G. J. Kennedy, M. A. Hofer, D. Cohen, R. Shindledecker, J. D. Fisher, Significance of depres-
sion and cognitive impairment in patients undergoing programed stimulation of cardiac
arrhythmias. Psychosom. Med. 49 (1987), 410–21.
105. N. Frasure-Smith, F. Lesperance, M. Talajic, Depression and 18-month prognosis after
myocardial infarction. Circulation 91 (1995), 999–1005.
106. G. Brevetti, A. Silvestro, V. Schiano, M. Chiariello, Endothelial dysfunction and cardiovas-
cular risk prediction in peripheral arterial disease: additive value of flow-mediated dilation
to ankle-brachial pressure index. Circulation 108 (2003), 2093–8.
107. P. O. Bonetti, L. O. Lerman, A. Lerman, Endothelial dysfunction: a marker of atherosclerotic
risk. Arterioscler., Thromb. Vasc. Biol. 23 (2003), 168–75.
108. D. S. Celermajer, K. E. Sorensen, V. M. Gooch, et al., Non-invasive detection of endothelial
dysfunction in children and adults at risk of atherosclerosis. Lancet 340 (1992), 1111–15.
109. D. S. Celermajer, K. E. Sorensen, C. Bull, J. Robinson, J. E. Deanfield, Endothelium-
dependent dilation in the systemic arteries of asymptomatic subjects relates to coronary
risk factors and their interaction. J. Am. Coll. Cardiol. 24 (1994), 1468–74.
110. T. Neunteufl, S. Heher, R. Katzenschlager, et al., Late prognostic value of flow-mediated
dilation in the brachial artery of patients with chest pain. Am. J. Cardiol. 86 (2000), 207–10.
111. T. Heitzer, T. Schlinzig, K. Krohn, T. Meinertz, T. Munzel, Endothelial dysfunction, oxidative
stress, and risk of cardiovascular events in patients with coronary artery disease. Circulation
104 (2001), 2673–8.
106 Depression and physical illness
Introduction
Patients who survive an acute myocardial infarction (MI) often face a difficult period
of psychological and social adjustment. During the weeks and months following an
MI, survivors are confronted with the possibility of being physically incapacitated,
of having another heart attack and of dying. Their usual roles and daily routines
may be disrupted, their self-esteem may be injured, and some of their hopes and
plans may be jeopardised. Even patients with stable coronary heart disease (CHD)
who have never had an MI or other major cardiac event must live with the possibility
that they might eventually have a heart attack, and many of them have to cope with
angina pectoris, fatigue and other debilitating symptoms.
Given the stressors with which patients with CHD are confronted, it is not
surprising that depression and anxiety are very common in these individuals. As
many as 65% of post-MI patients are at least mildly anxious or depressed [1–3].
Although some patients return to their premorbid mood state within a few days or
weeks after their MI, many have, or will develop, a more serious or persistent form
of clinical depression.
Depressive disorders
The Diagnostic and Statistical Manual of Mental Disorders, 4th edn (DSM-IV-TR)
of the American Psychiatric Association [4] defines the most widely used criteria
for diagnosing psychiatric disorders. As described in DSM-IV-TR, major depression
is a serious psychiatric disorder that tends to follow a chronic or recurrent course. A
major depressive episode is present when five or more depressive symptoms, includ-
ing dysphoric mood and/or loss of interest in usual activities, persist for at least two
weeks and cause clinically significant distress or functional impairment. Approxi-
mately 16–22% of post-MI patients have a major depressive episode within a few
Depression and Physical Illness, ed. A. Steptoe.
Published by Cambridge University Press. C Cambridge University Press 2006.
110 Depression and physical illness
weeks after the acute event [5–7], and about a third have an episode within a year.
Many post-MI patients who meet the criteria for major depression were depressed
before their MI [8].
Minor depression, a less severe depressive disorder, may be present if there are
between two and four depressive symptoms. As with major depression, one of the
symptoms must be either dysphoric mood or loss of interest in usual activities,
the symptoms must be present for at least two weeks and the symptoms must
cause distress or functional impairment. Dysthymia is a relatively mild depressive
disorder. Unlike minor depression, dysthymia tends to be quite persistent and is
often associated with chronically low self-esteem. Dysthymia is not diagnosed unless
dysphoric mood and two or more other depressive symptoms have been present for
at least two years [4]. About 25% of post-MI patients have either minor depression
or dysthymia [5]. In some cases, a major depressive episode is superimposed on
dysthymia, a condition known as ‘double depression’.
Both major and minor depression are also common in patients who have not
recently had an acute MI but who do have angiographically proven coronary artery
disease (CAD). The prevalence of major depression is estimated to be between 17%
and 23% in these patients [9–11], and about 17% have minor depression [11].
Patients with minor depression are at risk for developing the more serious major
depressive disorder. In a study of patients with documented CAD but without
a recent cardiac event, half of the patients who initially had minor depression
developed major depression within the next 12 months [11]. Thus, although some
cardiac patients with minor depression are experiencing nothing more than a mild
transient emotional reaction that may not require any intervention other than
support from family, friends and a trusted physician, many others will progress
from minor to major depression, a condition that usually warrants treatment [11].
Minor depression is often treated for this reason, particularly in patients who have
a prior history of major depressive episodes.
Depression has also been found to increase the risk of further medical morbidity
and even mortality in CHD patients. Depression is associated with an increased
risk for cardiac events, including cardiac mortality, in patients with recently diag-
nosed coronary disease [17,18] and following coronary artery bypass graft surgery
[19–21]. The risk for mortality is especially high following an acute MI [7,22–27].
For example, Frasure-Smith et al. [7] found that depression was associated with
a more than four-fold increased risk of mortality during the first six months fol-
lowing an acute MI, after adjusting for established prognostic variables including
left ventricular dysfunction. Moreover, its prognostic significance was equivalent
to that of left ventricular dysfunction and of a prior history of MI (see Chapter 4).
Pharmacotherapy
There are four major classes of antidepressant medications: tricyclic antidepres-
sants (TCAs), monoamine oxidase inhibitors (MAOIs), selective serotonin reup-
take inhibitors (SSRIs) and second-generation heterocyclic antidepressants. All
have been shown to relieve depression in psychiatric patients. Although there is
no reason to believe that antidepressants are less effective for depression in cardiac
patients than in psychiatric patients, there has been relatively little efficacy research
on antidepressants in cardiac populations. Furthermore, some antidepressants are
contraindicated for many cardiac patients [28–30].
The TCAs and MAOIs are known to affect cardiac conduction, contractility, rate
and rhythm, and may cause orthostatic hypotension [28–30]. These side effects are
of special concern in patients with unstable angina, conduction disorders, heart
failure or other complications of coronary disease. Orthostatic hypotension is of
particular concern when treating older patients who are vulnerable to fall-related
injuries, including life-threatening hip fractures. For these reasons, most experts
caution against using any of the older antidepressants to treat cardiac patients.
SSRIs, on the other hand, are relatively free of cardiac side effects [31], but
they pose a significant risk for drug–drug interactions [31]. Most cardiac patients
take multiple medications for their disease and comorbid medical conditions, and
so drug–drug interactions must always be considered when prescribing SSRIs for
these patients. Nevertheless, SSRIs and other carefully selected antidepressants are
believed to be reasonably safe when administered to cardiac patients with appro-
priate precautions [31,32]. Due to their relative safety, SSRIs are frequently selected
to treat depression in patients with CHD.
However, there have been very few randomised controlled efficacy studies of any
antidepressant in cardiac patients. Most of the existing studies have, understand-
ably, focused on safety rather than efficacy. It is possible that depression in these
112 Depression and physical illness
patients may be qualitatively different from psychiatric depression and that it may
not respond to the same kinds of treatment. Nevertheless, most clinical trials tar-
geting depression in CHD patients have found that their depression can be treated
successfully with conventional antidepressants [32], although few of these studies
included a placebo control group. Shores and colleagues [32] published a compre-
hensive review of all published trials of pharmacological treatments for depression
in patients with CHD. They concluded that more clinical trials of SSRIs are needed,
as well as trials evaluating efficacy and safety of other types of antidepressant, such
as nefazodone, venlafaxine and bupropion, in depressed patients with CHD.
The Sertraline Antidepressant Heart Attack Randomized Trial (SADHART) [33]
was designed to evaluate the safety and efficacy of sertraline (an SSRI) in depressed
patients hospitalised for an acute MI or unstable angina. SADHART was the first
study to show that any antidepressant is safe for use early after an acute cardiac event.
However, sertraline proved to be only modestly efficacious, and only in patients
with relatively severe recurrent depression. Nevertheless, SADHART represents an
important step towards improving the care of depressed cardiac patients following
an acute cardiac event.
In summary, although there have been few trials of pharmacological treatment
of depression in patients with stable CHD, and only one randomised controlled
multicentre clinical trial in patients with a recent acute cardiac event, the existing
evidence suggests that at least some of the available antidepressants are safe and
effective in patients with CHD. Clearly, however, more clinical trials of antidepres-
sants for depressed patients with CHD are needed.
Psychotherapy
Despite the availability of effective antidepressant medications, psychotherapy con-
tinues to play an important role in the treatment of depressive disorders. Some
patients are unwilling to take psychotropic medications or do not comply with the
prescribed regimen; also, as discussed in the preceding section, some medications
are contraindicated for many cardiac patients. Regardless of whether they are taking
psychotropic medications, many depressed cardiac patients need more emotional
support and more intensive help with their personal adjustment problems than
their physicians have the time or training to provide. Also, there is evidence that
persistent depression may respond better to a combination of pharmacotherapy
and psychotherapy than to either modality alone, at least for patients who are able
to tolerate and cooperate with both forms of treatment [34].
A bewildering variety of psychotherapeutic interventions are used to treat depres-
sion in psychiatric patients, but they are not interchangeable. There is evidence that
some psychotherapeutic interventions are more efficacious than others for these
conditions. Of all the existing forms of psychotherapy for unipolar depression,
113 The management of depression in patients with coronary heart disease
Cognitive–behavioural therapy
CBT is a structured short-term treatment that focuses on modifying depressogenic
thoughts and beliefs, solving current problems and increasing the frequency of
productive and pleasurable activities [34,35]. Patients undergoing CBT learn that
their own thoughts govern their mood and behaviour, and that they can feel better by
challenging and changing distorted, erroneous or otherwise maladaptive thoughts
and then changing their behaviour accordingly. Most patients are not fully aware
of their distressing thoughts and beliefs when they present for treatment.
Cognitive–behavioural therapists use a variety of techniques to help their patients
identify and challenge maladaptive thoughts and replace them with more rational
and adaptive thoughts. For example, a patient may be asked to use a ‘dysfunctional
thought record’ (DTR) whenever he or she is feeling particularly discouraged,
anxious or upset [35]. Depressed patients’ DTR entries often reflect excessively
pessimistic views of themselves, their future or the world in which they live. When
feeling depressed, they may have such thoughts as ‘My whole life is ruined’, ‘Nobody
will ever want me again’, or ‘I can’t take it any more’.
CBT helps patients to replace these ideas with more adaptive thoughts and to
reinforce them with confirmatory behaviour change. For example, a depressed,
socially isolated patient may progress from thinking ‘I’ll never meet anyone’ and
‘People will reject me if I try to be friendly’ to ‘It’s not impossible to meet people,
and it won’t kill me even if some people don’t respond to me the way in which
I’d like them to.’ To validate the patient’s new way of thinking, the patient is given
behavioural homework assignments such as attending social gatherings and initi-
ating conversations and confirming that nothing terrible happens as a result. The
behavioural assignments are made more challenging as the patient’s confidence
and skills increase, to ensure that most of his or her experiences are successful
and encouraging. The patient’s mood typically improves along with these develop-
ments. Furthermore, the patient can use the skills acquired in the process to cope
more effectively with new problems that arise after the termination of therapy and
to help prevent recurrent episodes of depression or anxiety.
Some of the cognitive aspects of CBT such as examining core beliefs may be
difficult for medically ill, elderly or poorly educated patients, as well as for those with
severe depression. In such cases, the behavioural components of CBT usually take
precedence over the cognitive components, at least early in the course of treatment.
114 Depression and physical illness
Behavioural activation techniques, for example, can be used to help the patient
to resume normal activities without necessarily having to change the distorted
pessimistic ideas that may be maintaining his or her inactivity (e.g. ‘Now that I’ve
had a heart attack, I’ll never be able to play a round of golf again’). Resumption
of activities (or, if the patient has acquired a physical disability, finding satisfying
alternative activities) disconfirms such thoughts and usually has potent beneficial
effects on the patient’s mood.
Interpersonal psychotherapy
IPT was originally developed as a manualised short-term treatment for major
depression for use in clinical research [36]. Its efficacy for major depression has
been demonstrated in psychiatric patients, including in the National Institute of
Mental Health (NIMH) Treatment of Depression Collaborative Research Program
(TDCRP) [37]. Several studies have also shown that IPT can be used to treat both
major and minor depression in medically ill patients [40–42].
IPT focuses on solving interpersonal problems as a way to overcome depres-
sion. During the initial evaluation, the therapist identifies social and interpersonal
problems that appear to have played a role in the onset of the patient’s depres-
sion. Typical problems include interpersonal losses (e.g. death of a close friend),
role transitions (e.g. loss of employment due to medical illness) and role disputes
(e.g. marital conflict). Social skills deficits that interfere with the ability to form
relationships or to gain emotional support are also addressed in some cases.
Most of the subsequent sessions focus on strategies for overcoming the patient’s
most troubling social and interpersonal problems. For example, an elderly patient
might become depressed after relocating in order to live closer to his or her children
and grandchildren after undergoing coronary bypass surgery. Despite the advan-
tages of living closer to the family, a patient in this situation might suffer from the
loss of frequent contact with long-time friends. Consequently, the patient’s inter-
personal therapist might concentrate on helping the patient to mourn this loss and
to develop friendships in his or her new home. Towards the end of successful treat-
ment, the emphasis would shift toward consolidating and maintaining the gains
that have been achieved and developing relapse prevention skills.
Unfortunately, the specific CBT or IPT interventions for depression and anxiety
that have strong empirical support in psychiatric populations have not been system-
atically tested on cardiac patients. Other types of psychotherapeutic intervention
have been tested instead, and many of these studies suffer from methodological
shortcomings, including small samples, inadequate control groups, compromised
randomisation procedures and inadequate assessment of psychological outcomes.
Most of these studies have evaluated the addition of psychosocial interventions to
usual care or to a cardiac rehabilitation programme. A wide variety of psychothera-
peutic and behavioural interventions have been tested, including traditional group
psychotherapy and relaxation training. The principal targets of treatment have been
equally diverse: depending upon the study, the intervention may have been intended
to reduce anxiety or depression, to modify type A behaviour, or to otherwise pro-
mote psychosocial adjustment.
Although the patients in these studies generally become less depressed as a result
of treatment, most of the randomised trials have failed to show a significantly
greater reduction in depression in the intervention than in the control groups, e.g.
[46–48]. However, as noted above, the majority of these studies have employed
psychotherapeutic interventions that are not considered optimal treatments for
depression. Thus, there is currently little evidence for the efficacy of the two standard
depression interventions, or indeed for any psychotherapeutic intervention, for the
treatment of depression in patients with CHD.
Other treatments
Exercise
Numerous controlled and uncontrolled studies have found certain types of exercise
to effectively reduce depression [49]. Much of this work has been marred, how-
ever, by the lack of proper controls. In a meta-analysis of 14 randomised controlled
clinical trials of exercise training as a treatment for depression, Lawlor and Hopker
[49] reported that most of the studies suffer from significant methodological weak-
nesses, including non-blinded assessments, outcome analyses that included only
protocol completers and inadequate follow-ups. They concluded that the efficacy
of exercise in treating depression has not yet been firmly established.
Blumenthal and colleagues [50] enrolled 156 elderly patients with major depres-
sive disorder in a randomised controlled clinical trial of exercise training or sertra-
line or sertraline plus exercise training. After 16 weeks of treatment, the groups did
not differ significantly on any measure of depression. Without an untreated con-
trol group, there is no way to determine how many of the depressed patients would
have improved without any form of treatment. Furthermore, although some of
these patients had stable heart disease, having CHD was not an inclusion criterion.
116 Depression and physical illness
Pharmacotherapy
Although there is at least limited evidence for the safety and efficacy of antide-
pressants for the treatment of depression in patients with CHD, little is known
about the effects of treating depression on subsequent cardiac events. In one of
the few relevant studies published to date, Avery and Winokur [53] found that
non-suicidal deaths, especially deaths that were cardiac-related, occurred more fre-
quently during a three-year follow-up among depressed patients who had received
inadequate treatment for depression than among those whose treatment was con-
sidered adequate. Although this is an intriguing finding, alternative interpretations
are possible. For example, it is possible that the patients who received inadequate
treatment for their depression also received less than adequate care for other med-
ical conditions. Additionally, the study was based on a small number of endpoints
and did not focus on patients with established CHD.
The SADHART sample was not large enough to establish definitively whether
treatment with sertraline reduces cardiac events in depressed patients. The authors
did report, however, a non-significant trend towards fewer cardiac events among
patients receiving sertraline compared with those in the placebo arm. The finding
of a possible benefit from the SSRI is consistent with the results of epidemiological
studies [54,55], and of laboratory studies showing that SSRIs act to inhibit platelet
activation [56,57]. Depressed patients have been shown to have higher levels of
platelet activation [58,59], which has been suggested as a possible mechanism for
the increased risk of cardiac events in depressed patients [60]. The encouraging
trend reported in the SADHART study suggests the need for a future study with a
larger sample of patients with major depression after hospitalisation for an acute
MI or unstable angina.
Psychotherapy
In a quantitative review of 16 randomised, controlled trials, Linden et al. [61] found
that interventions for CHD patients that include a psychotherapeutic component
117 The management of depression in patients with coronary heart disease
are associated with significantly lower two-year rates of morbidity (odds ratio [OR]
1.84) and mortality (OR 1.70), compared with patients randomised to control
groups. Although this is very encouraging, many of the studies included in the
review had serious methodological limitations, and some had negative results.
Furthermore, none of these studies specifically targeted depression or attempted to
evaluate the efficacy of standard depression interventions.
The two largest trials accounted for 77% and 37% of the treated subjects included
in the meta-analyses of mortality and morbidity outcomes, respectively, and both
trials have been criticised for flawed randomisation procedures [61,62]. Many of the
studies included in the review of Linden et al. [61] also combined psychotherapeutic
with educational interventions without controlling for the possible interactions
between them. If, for example, a study compared exercise training plus supportive
psychotherapy versus exercise alone, and the former group was found to have a lower
mortality rate, then this could have occurred because the supportive component
of the psychotherapy improved adherence to the exercise protocol. Thus, the effect
may have been due to differences in the level of exercise rather than to improved
psychosocial functioning.
Four large randomised studies have been published since the Linden et al. [61]
meta-analysis. The first was an attempt by Frasure-Smith et al. [63] to replicate their
successful study, the Ischemic Heart Disease Life Stress Monitoring Program, using
a more conventional random assignment procedure than was used in the first study.
As in the first study, the ‘distress levels’ of those in the intervention arm were assessed
by telephone beginning one week after discharge, and then monthly for one year,
using the 20-item General Health Questionnaire (GHQ). Patients scoring 6 or more
on this questionnaire and patients who were hospitalised were visited by a nurse,
who offered an individually tailored intervention that could include emotional
support, reassurance, education or referral to a mental health professional.
Although the patients were selected for ‘non-specific distress’ rather than for
anxiety or depression, many of the participants were at least mildly depressed.
However, the intervention had little effect on depression. Furthermore, unlike the
original study, there was no difference in survival between the intervention and
usual-care control group.
In a trial evaluating a psychological intervention for cardiac rehabilitation
patients, Jones and West [64] randomised 1168 patients with a recent MI to receive
a psychosocial intervention and 1160 patients to receive only usual care. The treat-
ment consisted of seven two-hour out-patient sessions of a stress-management
intervention, which included relaxation training, supportive counselling and edu-
cation. The sessions were led by clinical psychologists and other medical personnel.
Again, like the study of Frasure-Smith et al. [63], patients were not selected specif-
ically for depression, and the intervention was not designed specifically to treat
depression.
118 Depression and physical illness
At six months, there were no significant differences between the treated and
control patients in self-reported depression. Furthermore, by 12 months, 76 of the
treated patients and 75 of the controls had died. Thus, no difference in mortality
was found between the treated and control patients.
In a third study, Blumenthal and colleagues [65] randomly assigned 107 patients
with documented CAD who developed myocardial ischaemia during a laboratory-
based mental stress test either to exercise training or to a cognitive–behavioural
stress-management group. Patients who lived too far away from the medical centre
to conveniently return for follow-up visits were invited to participate as a usual-care
comparison group.
Patients assigned to the stress-management intervention had significantly less
severe wall-motion abnormalities on post-treatment mental stress testing, signifi-
cantly fewer ischaemic episodes during ambulatory monitoring and fewer cardiac
events compared with the usual-care participants. There were no significant dif-
ferences between the stress-management and exercise groups at the end of the
treatment. Unfortunately, a geographic criterion rather than random assignment
was used to form the usual-care comparison group. Thus, although this was an oth-
erwise excellent study with a very promising intervention, it cannot be concluded
with confidence that the treatment was responsible for the outcome. Blumenthal
et al. [65] argue against such interpretations, but it is possible that the usual-care
patients may not have received the same quality of medical care as the patients
who lived nearer to the medical centre. Blumenthal and colleagues are currently
attempting to replicate their study, enrolling a larger sample and including a ran-
domised usual-care control group. However, like the original study, this study does
not focus exclusively on depressed patients.
this represents a modest difference in depression between the intervention and the
usual-care group, the trial was not designed to test the efficacy of the intervention
for depression. Instead, the trial was designed to determine whether the interven-
tion could reduce the rate of recurrent infarction or all-cause mortality. Thus, there
was no attempt to control for non-specific treatment factors. Moreover, patients in
the usual-care group were free to seek outside treatment for depression. About a
fifth of the usual-care patients did so during the course of the trial.
Reinfarction-free survival during a mean follow-up of 29 months did not differ
between the groups [67]. One of the explanations that the investigators offered for
this null finding was that there was only a small difference between the groups in
the depression outcomes. If their conclusion is correct, then more effective inter-
ventions for depression may be needed in order to affect medical outcomes.
The ENRICHD results are still being evaluated, and it is not yet certain why the
trial failed or whether any form of depression treatment can improve survival. The
depressed participants who were treated with antidepressants, with or without
the addition of CBT, tended to survive longer than the patients who were not
receiving any antidepressant. However, antidepressant therapy was not allocated
by random assignment, and so this finding should be interpreted cautiously.
Another secondary analysis of the ENRICHD study sought to determine whether
there was a relationship between improvement in depression and survival after the
six-month intervention in patients who had a BDI score ≥ 10 and a past history
of major depression and who completed the six-month post-treatment assessment
[68]. Out of the 858 patients (409 usual care, 449 intervention) who met these
criteria, the intervention-group patients whose depression did not improve were at
higher risk for late mortality than were patients who responded to treatment. These
results suggest that patients whose depression is refractory to CBT and sertraline,
two standard treatments for depression, may be at higher risk for late mortality
after an acute MI. This analysis was not planned before the trial, and so the results
must be interpreted with caution. However, if the results are eventually replicated
in a prospective study, then patients who are resistant to standard depression treat-
ment should be followed more closely. More aggressive cardiological care may be
warranted, including efforts to treat all other modifiable risk factors and all comor-
bid medical disorders to the fullest extent possible. More aggressive treatment for
depression may also be warranted, but further research is needed to develop more
effective ways to treat depression in patients with CHD who do not respond to
first-line interventions.
Acknowledgements
The preparation of this chapter was supported in part by grant no. RO-1HL58946
from the National Heart, Lung, and Blood Institute, and by the Lewis and Jean
Sachs Charitable Lead Trust.
REFERENCES
4. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th
edn, text revision (Washington, DC: American Psychiatric Association, 2000).
5. S. J. Schleifer, M. M. Macari-Hinson, D. A. Coyle, et al., The nature and course of depression
following myocardial infarction. Arch. Intern. Med. 149 (1989), 1785–9.
6. R. M. Carney, K. E. Freedland, A. S. Jaffe, Insomnia and depression prior to myocardial
infarction. Psychosom. Med. 52 (1990), 603–9.
7. N. Frasure-Smith, F. Lespérance, M. Talajic, Depression following myocardial infarction:
impact on 6 month survival. J. Am. Med. Assoc. 270 (1993), 1819–25.
8. F. Lespérance, N. Frasure-Smith, M, Taljic, Major depression before and after myocardial
infarction: its nature and consequences. Psychosom. Med. 58 (1996), 99–110.
9. R. M. Carney, M. W. Rich, A. J. Tevelde, et al., Major depressive disorder in coronary artery
disease. Am. J. Cardiol. 60 (1987), 1273–5.
10. M. B. Gonzalez, T. B. Snyderman, J. T. Colket, et al., Depression in patients with coronary
artery disease. Depression 4 (1997), 57–62.
11. M. Hance, R. M. Carney, K. E. Freedland, J. Skala, Depression in patients with coronary heart
disease: a twelve month follow-up. Gen. Hosp. Psychiatry 18 (1995), 61–5.
12. M. J. Stern, L. Pascale, J. B. McLoone, Psychosocial adaption following an acute myocardial
infarction. J. Chronic. Dis. 29 (1976), 213–26.
13. R. Mayou, A. Foster, B. Williamson, Psychosocial adjustment in patients one year after
myocardial infarction. J. Psychosom. Res. 22 (1978), 447–53.
14. G. G. Lloyd, R. H. Cawley, Distress or illness? A study of psychosocial symptoms after myocar-
dial infarction. Br. J. Psychiatry 142 (1983), 120–25.
15. A. L. Stewart, S. Greenfield, R. D. Hays, et al., Functional status and well-being of patients
with chronic conditions: results from the Medical Outcomes Study. J. Am. Med. Assoc. 262
(1989), 907–13.
16. K. B. Wells, A. Stewart, R. Hays, et al., The functioning and well-being of depressed patients:
results from the Medical Outcomes Study. J. Am. Med. Assoc. 262 (1989), 914–19.
17. R. M. Carney, M. W. Rich, K. E. Freedland, et al., Major depressive disorder predicts cardiac
events in patients with coronary artery disease. Psychosom. Med. 50 (1988), 627–33.
18. J. C. Barefoot, M. J. Helms, D. B. Mark, et al., Depression and long-term mortality risk in
patients with coronary artery disease. Am. J. Cardiol. 78 (1996), 613–17.
19. I. Connerney, P. Shapiro, J. S. McLaughlin, R. P. Sloan, In hospital depression after CABG
surgery predicts 12-month outcome. Psychosom. Med. 62 (2000), 106.
20. M. M. Burg, C. M. Benedetto, R. Rosenberg, R. Soufer, Depression prior to CABG predicts
6-month and 2-year morbidity and mortality. Psychosom. Med. 63 (2001), 103.
21. J. A. Blumenthal, H. S. Lett, M. A. Babyak, et al., Depression as a risk factor for mortality
after coronary artery bypass surgery. Lancet 362 (2003), 604–9.
22. J. Irvine, A. Basinski, B. Baker, et al., Depression and risk of sudden cardiac death after acute
myocardial infarction: testing for the confounding effects of fatigue. Psychosom. Med. 61
(1999), 729–37.
23. D. K. Ahern, L. Gorkin, J. L. Anderson, et al., Biobehavioral variables and mortality or
cardiac arrest in the Cardiac Arrhythmia Pilot Study (CAPS). Am. J. Cardiol. 66 (1990), 59–
62.
122 Depression and physical illness
24. K. H. Ladwig, M. Kieser, J. Konig, G. Breithardt, M. Borggrefe, Affective disorders and survival
after acute myocardial infarction. Eur. Heart. J. 12 (1991), 959–64.
25. R. M. Carney, J. A. Blumenthal, D. Catellier, et al., Depression as a risk factor for mortality
following acute myocardial infarction. Am. J. Cardiol. 92 (2003), 1277–81.
26. D. E. Bush, R. C. Ziegelstein, M. Tayback, et al., Even minimal symptoms of depression
increase mortality risk after acute myocardial infarction. Am. J. Cardiol. 88 (2001), 337–41.
27. M. W. Kaufmann, J. P. Fitzgibbons, E. J. Sussman, et al., Relation between myocardial infarc-
tion, depression, hostility, and death. Am. Heart. J. 138 (1999), 549–54.
28. R. P. Pary, C. R. Tobias, S. Lippmann, Antidepressants and the cardiac patient: selecting an
appropriate medication. Postgrad. Med. 85 (1989), 267–9.
29. A. H. Glassman, S. P. Roose, J. T. Bigger, The safety of tricyclic antidepressants in cardiac
patients: risk/benefit reconsidered. J. Am. Med. Assoc. 269 (1993), 2673–5.
30. S. J. Warrington, C. Padgham, M. Lader, The cardiovascular effects of antidepressants. Psychol.
Med. Monogr. Suppl. 16 (1989), 1–40.
31. Y. Sheline, K. E. Freedland, R. M. Carney, How safe are serotonin reuptake inhibitors for
depression in patients with coronary heart disease? Am. J. Med. 102 (1997), 54–9.
32. M. M. Shores, M. Pascualy, R. C. Veith, Major depression and heart disease: treatment trials.
Semin. Clin. Neuropsychiatry 3 (1998), 87–101.
33. A. H. Glassman, C. M. O’Connor, R. M. Califf, et al., Sertraline treatment of major depression
in patients with acute MI or unstable angina. J. Am. Med. Assoc. 288 (2002), 701–9.
34. M. B. Keller, J. P. McCullough, D. N. Klein, et al., A comparison of nefazodone, the cognitive
behavioral-analysis system of psychotherapy, and their combination for the treatment of
chronic depression. N. Engl. J. Med. 342 (2000), 1462–9.
35. A. T. Beck, A. J. Rush, B. F. Shaw, G. Emery, Cognitive Therapy of Depression (New York:
Guilford Press, 1979).
36. G. L. Klerman, M. M. Weissman, B. J. Rounsaville, E. S. Chevron, Interpersonal Psychotherapy
of Depression (New York: Basic Books, 1984).
37. I. Elkin, M. T. Shea, J. T. Watkins, et al., National Institute of Mental Health Treatment of
Depression Collaborative Research Program: general effectiveness of treatments. Arch. Gen.
Psychiatry 46 (1989), 971–82.
38. G. E. Murphy, A. D. Simons, R. D. Wetzel, P. J. Lustman, Cognitive therapy and pharma-
cotherapy, singly and together in the treatment of depression. Arch. Gen. Psychiatry 41 (1984),
33–41.
39. G. E. Murphy, R. M. Carney, R. D. Wetzel, M. Knesevitch, P. Whitworth, Cognitive ther-
apy and depression: a controlled treatment outcome study. Psychol. Rep. 77 (1995), 403–
20.
40. J. C. Markowitz, G, L. Klerman, S. W. Perry, K. F. Clougherty, A. Mayers, Interpersonal
therapy of depressed HIV-seropositive patients. Hosp. Community Psychiatry 43 (1992), 885–
90.
41. H. C. Schulberg, C. P. Scott, M. J. Madonia, S. D. Imber, Applications of interpersonal
psychotherapy to depression in primary care practice. In New Applications of Interpersonal
Psychotherapy, ed. G. L. Klerman, M. M. Weisman (Washington, DC: American Psychiatric
Press, 1993), pp. 265–91.
123 The management of depression in patients with coronary heart disease
61. W. Linden, C. Stossel, J. Maurice, Psychosocial interventions for patients with coronary artery
disease: a meta-analysis. Arch. Intern. Med. 156 (1999), 745–52.
62. L. H. Powell, Unanswered questions in the Ischemic Heart Disease Life Stress Monitoring
Program. Psychosom. Med. 51 (1989), 479–84.
63. N. Frasure-Smith, F. Lespérance, R. Prince, et al., Randomized trial of home based psy-
chosocial nursing intervention for patients recovering from myocardial infarction. Lancet
350 (1997), 473–9.
64. D. A. Jones, R. R. West, Psychological rehabilitation after myocardial infarction: multicentre
randomized controlled trial. Br. Med. J. 313 (1996), 1517–21.
65. J. A. Blumenthal, W. Jiang, M. A. Babyak, et al., Stress management and exercise training in
cardiac patients with myocardial ischemia. Arch. Intern. Med. 157 (1997), 2213–23.
66. ENRICHD investigators. Enhancing Recovery in Coronary Heart Disease Patients
(ENRICHD): study design and methods. Am. Heart J. 139 (2000), 1–9.
67. ENRICHD investigators. Effects of treating depression and low perceived social support
on clinical events after a myocardial infarction: the Enhancing recovery in Coronary Heart
Disease Patients (ENRICHD) randomized trial. J. Am. Med. Assoc. 289 (2003), 3106–16.
68. R. M. Carney, J. A. Blumenthal, K. E. Freedland, et al., Depression and late mortality after
myocardial infarction in the Enhancing Recovery in Coronary Heart Disease (ENRICHD)
study. Psychosom. Med. 66 (2004), 466–74.
69. K. E. Freedland, M. W. Rich, J. A. Skala, et al., Depression in hospitalized patients with
congestive heart failure. Psychosom. Med. 63 (2001), 177.
6
6
Depression and physical disability
Brenda W. J. H. Penninx
This chapter describes the link between depression and disability. Disability can
be defined as a restriction in or lack of ability to perform an activity because of
impairment. These activities can include interpersonal relationships, work and
school activities and physical activities; the latter is defined as ‘physical disability’.
In this chapter, the main focus is on physical disability, since this is the type of
disability that has often been examined in relation to depression, especially in
old age. However, when appropriate the chapter also elaborates on other types of
disability.
The first section of this chapter describes the concept of disability in more detail.
The next section provides an overview of research that examines the link between
depression and disability. It then goes on to discuss underlying mechanisms that
could explain the link between depression and disability. Subsequently, results of
intervention studies that try to break the link between depression and disability are
described. The chapter ends with some concluding remarks.
Functional
Pathology Impairment Disability
limitations
although women live longer than men, women live with more physical disability
at an older age. This is caused in large part by the prevalence of non-lethal but
disabling chronic diseases generally being higher in older women than in older
men. For instance, arthritis and joint symptoms such as pain and stiffness are
typically higher in older women than older men; this is also the case for hearing
and vision impairment, osteoporosis (with resulting falls and fractures) and obesity,
which all have a marked adverse impact on disability levels.
Physical disability status has been demonstrated in epidemiological studies to
be one of the most potent of all health status indicators in predicting adverse
outcomes such as mortality, hospitalisation and nursing-home admission [4–6].
This is because disability measures are able to capture the impact of the presence
and severity of multiple pathologies, including physical, cognitive and psychological
conditions, and the potential synergistic effects of these conditions on overall health
status. In line with this, older adults report that they often worry about their risk
of disability more than about the disease itself, because function decline changes
the scope of their daily life and threatens their ability to live independently. Since
the adverse impact of disability on outcomes such as hospitalisation, healthcare
utilisation and nursing-home admission is large, it is not surprising that physical
disability has been associated with greatly increased health care costs. In a US study
of 843 people aged 72 years or older, those with ADL disability were found to
spend $10 000 more on 2-year costs for hospital, out-patient, nursing-home and
home-care services than those without ADL disability [7].
take away the significant link. In a non-depressed sample of 680 older people,
the presence of functional limitations (as assessed by vision, hearing or mobility
problems) predicted the development of major depressive disorder [23]. Of the
people without functional limitations, 11% developed depression during 4 years
of follow-up compared with 21% of those in the high-limitations group. Among
889 older residents of London, the presence of ADL disability increased the risk of
onset of significant depressive symptoms one year later by a factor of three [24]. In a
Dutch study, depression and physical disability were measured at eight consecutive
waves during three years of follow-up [25]. Using these data, the researchers found
that physical disability was a predictor of both the onset and the persistence of
depression.
This finding that disability increases the risk of depression is not surprising,
since the onset of disability is a major stressor that leads to loss of perceived control,
restriction of valued social or leisure activities, isolation and reduced quality of
social support, all of which are psychosocial risk factors for depression. In addition,
people with acquired physical disabilities inevitably encounter additional losses,
such as the loss of function, role and body image, and may experience a greater
dependency on others and a more negative view of themselves, their future and their
world. Depression can occur as an outcome of certain somatic illnesses or medi-
cation, reflecting a biologically mediated process. For example, the structural and
neurochemical changes involved in stroke and parkinsonism can lead to depression.
Several factors that often co-occur with disability, such as arthritis, pain and cogni-
tive and sensory impairment, can put older people at risk for late-life depression. It
must be mentioned that certain symptoms of depression, especially somatic symp-
toms such as low energy levels and sleeping problems, may partly be a manifestation
of disease [26].
It is interesting that several studies among older people have shown that symp-
toms of depression are predicted more strongly by the level of physical disability
than by number or specific type of chronic conditions [23,27,28]. This indicates
that the effect of chronic conditions on depressive symptomatology is mediated
partly through the level of existing physical disability.
60
50
Incidence (%)
40
30
20
10
0
No Yes No Yes
Depressed Depressed
Figure 6.2 Six-year incidence of disability in activities of daily living (ADL) and mobility disability (disabil-
ity in walking a quater of a mile or climbing stairs) among initially non-disabled people aged
65 years or older. Findings from the Establised Populations of the Epidemiologic Studies of
the Elderly (EPESE) study [34].
consequently gives us a clearer picture of the causal role that depression appears
to play in the development of disability [17,32–36]. The majority of these longi-
tudinal studies have confirmed that people with depression have an increased risk
of developing new disability in subsequent years. In general, the results of these
studies indicate that in initially non-disabled populations, the existence of depres-
sion at baseline increases the risk for disability by 60% or more [37]. However, the
estimated increased disability risks in the various studies range quite considerably.
This wide range may be explained by the different types of disability outcome used
(e.g. disability in less severe IADL versus disability in severe basic ADL), the type of
depression assessment used (e.g. mild depressive symptoms versus psychiatric diag-
nosis of MDD) and the follow-up duration considered (one year versus six years).
Although the effect of depression on disability has been confirmed for depressive
symptoms and for a psychiatric diagnosis of MDD, it seems that the effect is some-
what larger for the most severe type of depression, which suggests the existence of
a dose–response relationship between depression and subsequent disability.
In the Established Populations of the Epidemiologic Studies of the Elderly
(EPESE) study, we explored the effect of significant depressive symptoms (as defined
by a high score on the Center of Epidemiologic Studies Depression [CES-D] scale)
on the onset of disability among more than 6000 initially non-disabled people aged
65 years or older [34]. We followed these people for six years and asked them every
year about their disability levels. As shown in Figure 6.2, of the non-depressed
131 Depression and physical disability
people, 24% developed disability in ADL and 36% developed difficulty with walk-
ing a quarter of a mile or climbing stairs (mobility limitations). These percentages
were significantly higher among depressed people (48% and 67%, respectively).
The depressed people were 1.67 times more likely to develop new ADL disability
and 1.73 times more likely to develop new mobility disability. When these risk
estimates were adjusted for other variables that differed between depressed and
non-depressed people, such as age, gender, education, income and medical con-
ditions, the risk estimates decreased to 1.45 and 1.37, respectively. However, these
risks were still considerable and remained highly significant. In various other stud-
ies, factors such as age, gender, education and medical conditions could not explain
the link between depression and incident disability risk.
We may still wonder whether the interpretation of these disability studies is ham-
pered by the possibility that depressed people may give overly pessimistic appraisals
of their functioning and disability. Even when their functional status remains the
same, people who are depressed tend to become more depressed over time and,
therefore, may report more physical disabilities. However, the link between depres-
sion and functional decline is also observed for objectively assessed physical impair-
ments, identifying an earlier stage of the disablement process (see Figure 6.1).
Impairment assessments, e.g. walking speed or strength assessments, appear to be
less influenced by personality, cognition and mood than self-reported measures of
disability. In the EPESE study, we examined the link between depression at base-
line (assessed by the CES-D score) and the four-year change in performance of
standardised tasks, including a walking test, a standing balance test and a repeated
chair–stand test. Depression was found to cause greater physical decline over four
years [38]. Thus, the link between depression and incident disability is already
present at earlier stages of the disablement process. More recently, we demonstrated
with longitudinal data from the Dutch Longitudinal Aging Study Amsterdam that
depression not only is associated with physical impairments but also accelerates
the transition from physical impairments to disability [39]. In other words, these
findings indicate that depression appears to accelerate the disablement process in
older people.
disability levels then return to normal? A few studies have examined depression
and disability patterns at multiple periods over time and examined these questions.
These studies were able to distinguish transient states of depression from more
chronic depressive states.
Our data from the Longitudinal Aging Study Amsterdam conducted among
community-dwelling older people provide strong evidence that the course of
depression plays a differential role in the relationship of depression and physical
decline over time [40]. About half (49.1%) of the people depressed at baseline (as
defined by a CES-D score ≥ 16) were still depressed after 3 years. This percentage is
similar to that in other studies [41], and illustrates the chronic nature of depression
in old age. In this group of chronically depressed people, we found an increased
risk for three-year physical decline. Subjects who remitted from their depression
in three years did not show greater physical decline compared with people who
had never been depressed. So, especially when chronically present, depression was
found to have a substantial impact on physical decline over a longer period of
time.
Synchrony of change over time between depression and disability levels has been
described in other studies. For instance, in a study of 1994 primary-care patients and
a study of 371 people with major depression, social disability levels improved when a
depressed person became asymptomatic [42–44]. In a Dutch study, Ormel et al. [45]
examined the extent to which psychosocial disability after remission from a major
depressive episode was due to residual symptoms (state effect), the continuation of
premorbid disability (trait effect) and disability that developed during the major
depression episode and persisted beyond recovery (scar effect). Results showed
that post-morbid psychosocial disability largely reflects continuation of premorbid
psychosocial disability. Scarring did not occur routinely but was found in some
people with a severe episode of depression. Again, these findings confirm synchrony
of change between depressive symptoms and disability.
The finding of synchronous change of depression and disability in longitudinal
studies indicates that, especially when chronically present, depression has a detri-
mental effect on functional status, which in turn could result in increased depressive
symptoms. Such a situation could result in a process whereby depressive symptoms
and physical dysfunctions interact to cause a progressive downward spiral in the
health status of older people. It appears to be essential to break this downward spiral,
e.g. through depression treatment, since observational longitudinal studies indi-
cate that when mood is improved in depressed people this is likely associated with
improved functional status. Later, this chapter presents results from some inter-
vention studies that have tried to break the unfavourable link between depression
and disability.
133 Depression and physical disability
more likely to smoke more of each cigarette than non-depressed smokers [46]. In
addition, the food intake of depressed people may be less adequate and nutritious
than that of non-depressed people. It has been shown that some depressed peo-
ple have a higher 24-hour caloric intake than non-depressed people [47]. On the
other hand, certain vitamin deficiencies, such as vitamin B12 and folate deficien-
cies, are more prevalent in depressed older people [48], which illustrates that some
depressed individuals may not get adequate nutrition.
Depressed people also engage less in physical activities such as walking, gardening
and vigorous exercise activities such as sport. Physical inactivity thus is common
among depressed people [34,49], partly because their attitudes towards exercise and
exercise self-efficacy may be more negative. This can partly explain why depressed
older individuals are more at risk for adverse health outcomes, since a sedentary
lifestyle is one of the most important risk factors for the onset of disability in old
age. This is an especially important observation, since the level of physical activity is
potentially modifiable through an exercise regimen. Several clinical trials illustrate
that when depressed older people are randomised to an exercise intervention, their
depressed mood improves significantly.
Finally, depressed mood has shown to impede recovery processes by discouraging
people from obtaining adequate medical attention and rehabilitation and following
treatment regimens. It has been suggested that depressed people are generally less
compliant in taking medications or following up on certain lifestyle regimens pro-
vided by healthcare professionals. This lower compliance could be due in part to lack
of a supportive social network, which has been observed more often in depressed
than in non-depressed people [50]. One study demonstrated that depressed
cardiac patients received lower quality of care than their non-depressed peers, and
this contributed to their higher mortality risk [51].
Depression is also associated with various psychological factors that have been
related to poor health outcomes. For instance, a lower sense of control, lower feelings
of (physical) self-efficacy and higher feelings of neuroticism have been observed
consistently among depressed people [50]. In old age, feelings of losing control
over one’s life and low expectations regarding one’s own physical abilities may have
important consequences. Some researchers have proposed that depressed older
people may experience motivational depletion and that they give up holding on to
a healthy, active life, which ultimately leads to a worse health status.
on disability status. As yet, no studies have directly examined the link between
HPA axis function and disability, and so this explanatory mechanism needs to be
investigated in future research.
Inflammation, a chronic elevated immune response characterised by high levels
of inflammatory cytokines, may also play a role in the pathophysiology of depres-
sion and disability. Cytokines are intracellular signalling polypeptides that deter-
mine host defence and regulate immune responses, the acute-phase reaction and
haematopoiesis [59]. Depression can cause excess interleukin 6 (IL-6) production
[60] through chronic glucocorticoid elevations [61], increased activity of the HPA
axis [62], or central and peripheral catecholaminergic systems independent of the
HPA axis [63]. Increased levels of inflammatory markers, e.g. IL-6, tumour necro-
sis factor alpha (TNF-) and C-reactive protein (CRP), have been demonstrated
among depressed psychiatric patients [64–67]. In our own work in a cohort of
older people, a positive association was found between depressive symptomatology
and plasma levels of IL-6 and TNF- [68]. However, the association of depressive
symptomatology with inflammation remains equivocal, since there have been some
studies that could not confirm the link [69,70]. These inconsistent findings could
indicate that the link between depression and inflammation is present only or more
strongly for more severe depression and for older people, since inflammation lev-
els are generally higher in older compared with younger people. Probably because
inflammation is involved in many pathophysiological processes, including cardio-
vascular conditions, diabetes, osteoarthritis and lung disease, inflammation has
been shown to be one of the strongest biological risk factors for physical disability
in old age [71,72]. This indicates that inflammation could be a pathophysiological
process that links depression to disability.
Pharmacotherapy
The majority of intervention studies conducted in the area of depression involve
pharmacotherapy, in which the primary aim has been to examine effects on
depressed mood. The effect of pharmacotherapy on physical function has been
much less studied. However, some trials also included an assessment of physical
functioning, in most cases a rather broad measure, such as the Medical Outcomes
Study 36-item short-form (SF-36) scale, which not only assesses aspects of disability
but also incorporates concepts such as pain and social isolation. In a study by Lin
et al. [73] conducted among 228 persistently depressed patients in primary care,
those who received a collaborative care intervention that included psychiatric con-
sultation, patient education and pharmacotherapy experienced significantly less
limitations in their family, work and social activities compared with patients receiv-
ing usual primary care. Similar findings were observed in a study comparing phar-
macotherapy and psychotherapy with usual care among 276 primary-care patients
with major depression [74]. A few controlled trials have examined the effect of phar-
macotherapy on a specific physical disability outcome measure. Overall, these stud-
ies indicate that the active treatment group showed greater improvement in physical
functioning than did the placebo group [75–77]. For instance, in a six-week placebo-
controlled study of nortriptyline in ambulatory depressed patients, subjects in the
intervention group had significantly greater improved physical disability scores than
did those in the placebo group [75]. This study also assessed objective measures
of physical impairments, such as walking endurance. Although self-reported dis-
ability measures improved, the objective measures of physical impairments did not
improve, which may suggest that the improvement in disability is relatively inde-
pendent of improvement in physical impairment. In a randomised trial, 1801 older
people were assigned to a care-management programme for depression (including
antidepressant management) or to a usual-care control group. After 12 months of
follow-up, those in the intervention group showed significantly greater improve-
ment in functional impairment than those in the control group. This is one of
the only studies to confirm the effectiveness of depression care management on a
functional outcome over a long period of time [78].
Exercise interventions
Observational studies suggest that regular physical exercise may be one of the most
important preventive factors for onset of late-life disability [79]. Clinical trials
among older people have shown that exercise programmes improve self-reported
functional scores and can prevent the onset of severe physical disability (ADL
disability) [80,81]. These studies provide strong evidence for the benefits of physical
exercise. However, benefits are not restricted to physical health alone but extend to
the psychological domain of health as well. Longitudinal epidemiological studies
138 Depression and physical illness
have shown that a high level of physical activity reduces the risk of developing
high depressive symptomatology over time [82], and that this effect is not limited
to individuals who have been active throughout their adult life. In short, physical
activity seems to be an effective form of prevention for depressive symptomatology
among older adults, even when adopted later in life [83]. Physiological, biological
and psychological mechanisms have been suggested to explain the antidepressive
effects of exercise [84].
Evidence for a beneficial effect of exercise on psychological health is confirmed in
several experimental studies, most of which have been conducted among clinically
depressed people. For example, Greist et al. [85], McCann and Holmes [86] and
Martinsen et al. [87] have shown that clinically depressed people randomised to an
exercise programme were more likely to improve their depression status than those
not receiving the exercise programme. Other experimental studies have shown that
exercise programmes appear to be at least as effective in reducing clinical depression
as more conventional treatment regimens such as antidepressant medication [88–
90]. In clinically depressed patients, evidence for an antidepressive effect of exercise
is provided by controlled trials involving resistance exercise [91] or aerobic exercise
programmes [88,89].
In the Fitness, Arthritis and Senior Trial (FAST) involving 438 older partic-
ipants with knee osteoarthritis, Penninx et al. [92] showed that aerobic exer-
cise, but not resistance exercise, significantly lowered depression scores during an
18-month follow-up. The antidepressive effect of aerobic exercise was found for
both people with initially high and with initially low depressive symptomatology
and was the strongest for those who were the most compliant. It appeared that the
antidepressive effect of exercise induced improvements in all subjects and did not
normalise depressed mood only in those who had high initial depression levels.
Both in initially depressed and non-depressed people, assignment to either aerobic
or resistance exercise was associated with a significant improvement of physical
disability levels during the 18-month follow-up. Overall, the results from this ran-
domised controlled trial indicated that an exercise intervention among (mildly)
depressed older people improved both mood and physical function.
REFERENCES
1. L. M. Verbrugge, A. M. Jette, The disablement process. Soc. Sci. Med. 38 (1994), 1–14.
2. World Health Organization. International Classification of Functioning, Disability and Health.
(Geneva: WHO, 2001).
3. E. Kramarow, H. Lentzner, R. Rooks, J. Weeks, S. Saydah, Health, United States, 1999: Health
and Aging Chartbook (Hyattsville, MD: National Center for Health Statistics, 1999).
4. J. M. Guralnik, L. Ferrucci, E. M. Simonsick, M. E. Salive, R. B. Wallace, Lower-extremity
function in persons over the age of 70 years as a predictor of subsequent disability. N. Engl.
J. Med. 332 (1995), 556–61.
5. B. W. Penninx, L. Ferrucci, S. G. Leveille, et al., Lower extremity performance in nondisabled
older persons as a predictor of subsequent hospitalization. J. Gerontol. A Biol. Sci. Med. Sci.
55 (2000), M691–7.
6. J. M. Guralnik, L. P. Fried, M. E. Salive, Disability as a public health outcome in the aging
population. Annu. Rev. Public Health 17 (1996), 25–46.
7. T. R. Fried, E. H. Bradley, C. S. Williams, M. E. Tinetti, Functional disability and health care
expenditures for older persons. Arch. Intern. Med. 161 (2001), 2602–7.
8. A. T. Beekman, D. J. Deeg, A. W. Braam, J. H. Smit, W. van Tilburg, Consequences of major
and minor depression in later life: a study of disability, well-being and service utilization.
Psychol. Med. 27 (1997), 1397–409.
9. Y. Forsell, A. F. Jorm, B. Winblad, Association of age, sex, cognitive dysfunction, and disability
with major depressive symptoms in an elderly sample. Am. J. Psychiatry 151 (1994), 1600–604.
10. H. G. Koenig, L. K. George, Depression and physical disability outcomes in depressed med-
ically ill hospitalized older adults. Am. J. Geriatr. Psychiatry 6 (1998), 230–47.
11. J. M. Lyness, D. A. King, C. Cox, Z. Yoediono, E. D. Caine, The importance of subsyndromal
depression in older primary care patients: prevalence and associated functional disability. J.
Am. Geriatr. Soc. 47 (2001), 647–52.
12. P. A. Sinclair, J. M. Lyness, D. A. King, C. Cox, E. D. Caine, Depression and self-reported
functional status in older primary care patients. Am. J. Psychiatry 158 (2001), 416–19.
13. K. B. Wells, A. Stewart, R. D. Hays, et al., The functioning and well-being of depressed
patients: results from the Medical Outcomes Study. J. Am. Med. Assoc. 262 (1989), 914–
19.
140 Depression and physical illness
30. J. Ormel, T. Oldehinkel, E. Brilman, W. vanden Brink, Outcome of depression and anxiety
in primary care: a three-wave 3 1/2-year study of psychopathology and disability. Arch. Gen.
Psychiatry 50 (1993), 759–66.
31. M. von Korff, J. Ormel, W. Katon, E. H. Lin, Disability and depression among high utilizers
of health care: a longitudinal analysis. Arch. Gen. Psychiatry 49 (1992), 91–100.
32. M. L. Bruce, T. E. Seeman, S. S. Merrill, D. G. Blazer, The impact of depressive symptomatology
on physical disability: MacArthur Studies of Successful Aging. Am. J. Public Health 84 (1994),
1796–9.
33. J. J. Gallo, P. V. Rabins, C. G. Lyketsos, A. Y. Tien, J. C. Anthony, Depression without sadness:
functional outcomes of nondysphoric depression in later life. J. Am. Geriatr. Soc. 45 (1997),
570–78.
34. B. W. Penninx, S. Leveille, L. Ferrucci, J. T. van Eijk, J. M. Guralnik, Exploring the effect of
depression on physical disability: longitudinal evidence from the established populations for
epidemiologic studies of the elderly. Am. J. Public Health 89 (1999), 1346–52.
35. D. Cronin-Stubbs, C. F. de Leon, L. A. Beckett, et al., Six-year effect of depressive symptoms
on the course of physical disability in community-living older adults. Arch. Intern. Med. 160
(2000), 3074–80.
36. H. K. Armenian, L. A. Pratt, J. Gallo, W. W. Eaton, Psychopathology as a predictor of disability:
a population-based follow-up study in Baltimore, Maryland. Am. J. Epidemiol. 148 (1998),
269–75.
37. E. J. Lenze, J. C. Rogers, L. M. Martire, et al., The association of late-life depression and
anxiety with physical disability: a review of the literature and prospectus for future research.
Am. J. Geriatr. Psychiatry 9 (2001), 113–35.
38. B. W. Penninx, J. M. Guralnik, L. Ferrucci, et al., Depressive symptoms and physical decline
in community-dwelling older persons. J. Am. Med. Assoc. 279 (1998), 1720–26.
39. C. H. van Gool, G. I. Kempen, B. W. Penninx, et al., Impact of depression on disablement in
late middle aged and older persons: results from the Longitudinal Aging Study Amsterdam.
Soc. Sci. Med. 60 (2005), 25–36.
40. B. W. Penninx, D. J. Deeg, J. T. van Eijk, A. T. Beekman, J. M. Guralnik, Changes in depression
and physical decline in older adults: a longitudinal perspective. J. Affect. Disord. 61 (2000),
1–12.
41. G. J. Kennedy, H. R. Kelman, C. Thomas, Persistence and remission of depressive symptoms
in late life. Am. J. Psychiatry 148 (1991), 174–8.
42. J. Ormel, M. Von Korff, B. W. van den, et al., Depression, anxiety, and social disability show
synchrony of change in primary care patients. Am. J. Public Health 83 (1993), 385–90.
43. G. I. Kempen, N. Steverink, J. Ormel, D. J. Deeg, The assessment of ADL among frail elderly
in an interview survey: self-report versus performance-based tests and determinants of dis-
crepancies. J. Gerontol. B Psychol. Sci. Soc. Sci. 51 (1996), 254–60.
44. L. L. Judd, H. S. Akiskal, P. J. Zeller, et al., Psychosocial disability during the long-term course
of unipolar major depressive disorder. Arch. Gen. Psychiatry 57 (2000), 375–80.
45. J. Ormel, A. J. Oldehinkel, W. A. Nolen, W. Vollebergh, Psychosocial disability before, during,
and after a major depressive episode: a 3-wave population-based study of state, scar, and trait
effects. Arch. Gen. Psychiatry 61 (2004), 387–92.
142 Depression and physical illness
46. R. F. Anda, D. F. Williamson, L. G. Escobedo, et al., Depression and the dynamics of smoking:
a national perspective. J. Am. Med. Assoc. 264 (1990), 1541–5.
47. C. U. Onyike, R. M. Crum, H. B. Lee, C. G. Lyketsos, W. W. Eaton, Is obesity associated
with major depression? Results from the Third National Health and Nutrition Examination
Survey. Am. J. Epidemiol. 158 (2003), 1139–47.
48. B. W. Penninx, J. M. Guralnik, L. Ferrucci, et al., Vitamin B(12) deficiency and depression
in physically disabled older women: epidemiologic evidence from the Women’s Health and
Aging Study. Am. J. Psychiatry 157 (2000), 715–21.
49. T. Stephens, Physical activity and mental health in the United States and Canada: evidence
from four population surveys. Prev. Med. 17 (1988), 35–47.
50. B. W. Penninx, T. van Tilburg, A. J. Boeke, et al., Effects of social support and personal coping
resources on depressive symptoms: different for various chronic diseases? Health Psychol. 17
(1998), 551–8.
51. B. G. Druss, W. D. Bradford, R. A. Rosenheck, M. J. Radford, H. M. Krumholz, Quality
of medical care and excess mortality in older patients with mental disorders. Arch. Gen.
Psychiatry 58 (2001), 565–72.
52. B. W. Penninx, T. G. van Tilburg, D. M. Kriegsman, et al., Effects of social support and personal
coping resources on mortality in older age: the Longitudinal Aging Study Amsterdam. Am.
J. Epidemiol. 146 (1997), 510–19.
53. C. B. Nemeroff, D. L. Musselman, D. L. Evans, Depression and cardiac disease. Depress.
Anxiety 8 Suppl 1 (1998), 71–9.
54. D. L. Musselman, D. L. Evans, C. B. Nemeroff, The relationship of depression to cardio-
vascular disease: epidemiology, biology, and treatment. Arch. Gen. Psychiatry 55 (1998),
580–92.
55. M. T. La Rovere, J. T. Bigger, Jr, F. I. Marcus, A. Mortara, P. J. Schwartz, Baroreflex sensitivity
and heart-rate variability in prediction of total cardiac mortality after myocardial infarction:
ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) investigators. Lancet
351 (1998), 478–84.
56. T. G. Vrijkotte, L. J. van Doornen, E. J. de Geus, Effects of work stress on ambulatory blood
pressure, heart rate, and heart rate variability. Hypertension 35 (2000), 880–86.
57. R. Pasquali, V. Vicennati, Activity of the hypothalamic-pituitary-adrenal axis in different
obesity phenotypes. Int. J. Obes. Relat. Metab. Disord. 24 Suppl 2 (2000), 47–9.
58. J. L. Abelson, G. C. Curtis, Hypothalamic-pituitary-adrenal axis activity in panic disorder:
24-hour secretion of corticotropin and cortisol. Arch. Gen. Psychiatry 53 (1996), 323–31.
59. J. Van Snick, Interleukin-6: an overview. Annu. Rev. Immunol. 8 (1990), 253–78.
60. M. Maes, A. H. Lin, L. Delmeire, et al., Elevated serum interleukin-6 (IL-6) and IL-6 receptor
concentrations in posttraumatic stress disorder following accidental man-made traumatic
events. Biol. Psychiatry 45 (1999), 833–9.
61. R. M. Sapolsky, L. C. Krey, B. S. McEwen, The neuroendocrinology of stress and aging: the
glucocorticoid cascade hypothesis. Endocrinol. Rev. 7 (1986), 284–301.
62. A. Arimura, A. Takaki, G. Komaki, Interactions between cytokines and the hypothalamic-
pituitary-adrenal axis during stress. Ann. N. Y. Acad. Sci. 739 (1994), 270–81.
63. S. Reichlin, Neuroendocrine-immune interactions. N. Engl. J. Med. 329 (1993), 1246–53.
143 Depression and physical disability
81. W. H. Ettinger, Jr, R. Burns, S. P. Messier, et al., A randomized trial comparing aerobic
exercise and resistance exercise with a health education program in older adults with knee
osteoarthritis: the Fitness Arthritis and Seniors Trial (FAST). J. Am. Med. Assoc. 277 (1997),
25–31.
82. M. E. Farmer, B. Z. Locke, E. K. Moscicki, et al., Physical activity and depressive symptoms:
the NHANES I Epidemiologic Follow-up Study. Am. J. Epidemiol. 128 (1988), 1340–51.
83. T. C. Camacho, R. E. Roberts, N. B. Lazarus, G. A. Kaplan, R. D. Cohen, Physical activity and
depression: evidence from the Alameda County Study. Am. J. Epidemiol. 134 (1991), 220–31.
84. T. C. North, P. McCullagh, Z. V. Tran, Effect of exercise on depression. Exerc. Sport. Sci. Rev.
18 (1990), 379–415.
85. J. H. Greist, M. H. Klein, R. R. Eischens, et al., Running as treatment for depression. Compr.
Psychiatry 20 (1979), 41–54.
86. I. L. McCann, D. S. Holmes, Influence of aerobic exercise on depression. J. Pers. Soc. Psychol.
46 (1984), 1142–7.
87. E. W. Martinsen, A. Medhus, L. Sandvik, Effects of aerobic exercise on depression: a controlled
study. Br. Med. J. (Clin. Res. Ed.) 291 (1985), 109.
88. J. A. Blumenthal, M. A. Babyak, K. A. Moore, et al., Effects of exercise training on older
patients with major depression. Arch. Intern. Med. 159 (1999), 2349–56.
89. M. Babyak, J. A. Blumenthal, S. Herman, et al., Exercise treatment for major depression:
maintenance of therapeutic benefit at 10 months. Psychosom. Med. 62 (2000), 633–8.
90. J. K. McNeil, E. M. LeBlanc, M. Joyner, The effect of exercise on depressive symptoms in the
moderately depressed elderly. Psychol. Aging 6 (1991), 487–8.
91. N. A. Singh, K. M. Clements, M. A. Fiatarone, A randomized controlled trial of progressive
resistance training in depressed elders. J. Gerontol. A Biol. Sci. Med. Sci. 52 (1997), M27–35.
92. B. W. Penninx, W. J. Rejeski, J. Pandya, et al., Exercise and depressive symptoms: a comparison
of aerobic and resistance exercise effects on emotional and physical function in older persons
with high and low depressive symptomatology. J. Gerontol. B Psychol. Sci. Soc. Sci. 57 (2002),
124–32.
7
7
Chronic pain and depression: twin burdens
of adaptation
Christina M. Van Puymbroeck, Alex J. Zautra
and Peter-Panagioti Harakas
Introduction
It is estimated that chronic pain afflicts between 50 and 80 million people in the USA
[1]. Adding to this burden of pain, feelings of depression frequently accompany the
pain [2]. These depressive symptoms include feelings of sadness, loss of pleasure
and fatigue and range in severity from transient malaise to persistent and debili-
tating episodes. For many people, it is common sense that negative feelings would
follow painful experiences, but a number of researchers have noted that depressed
patients frequently report high levels of pain as well. Not only is pain a common
somatic complaint in individuals suffering from depressive disorders [3], but also,
according to some accounts, more than 50% of clinically depressed patients report
pain as a symptom [4]. Not all investigators use the same criteria to determine the
presence of depression, and so the exact prevalence of depression among patients
with chronic pain is not easy to estimate [5]. Banks and Kerns [6] reviewed only
studies that used standardised criteria to diagnose depressive disorders and esti-
mated that at any given point 30–54% of clinic-based patients suffer from major
depressive disorder (MDD), a rate substantially higher than that found in the gen-
eral population [7] and higher than in out-patients of other medical conditions1 .
Thus, there appears to be a strong association between depressive symptoms and
persistent pain, but the underlying causal mechanisms remain poorly understood.
Nevertheless, our conceptualisations of both pain and depression are evolving at a
rapid pace, offering the possibility of a full accounting of the complex relationships
between depression, pain, illness and immune functioning.
1 In an effort to compare rates of depressive symptomatology across different pain conditions, Hawley and
Wolfe [8] reported the results of a longitudinal study of 6153 pain patients: depression scores among various
chronic pain groups, such as rheumatoid arthritis, osteoarthritis and low back pain, were not significantly
different, but in fibromyalgia patients depression scores were elevated in comparison with other chronic
pain conditions.
For centuries, pain was understood as a sensation arising from underlying tissue
damage. In 1965, this bottom-up (stimulus-response) model of pain was challenged
by Melzack and Wall’s [9] gate-control theory of pain, which emphasised a top-
down multidimensional conceptualisation of pain. The gate-control theory posited
three dimensions of pain: a sensory-physiologic dimension, a motivational-affective
dimension and a cognitive-evaluative dimension. A number of psychosocial models
of the pain–depression relationship followed in the footsteps of gate-control theory
and further emphasised the importance of psychological processes in the experience
of chronic pain. Nonetheless, despite numerous studies conducted in this area over
the past decades, the causal relationship between pain and depression remains
controversial [10]. For this reason, we believe it is instructive to briefly review
the historically dominant hypotheses formulated about the nature of the pain–
depression relationship.
Antecedent hypothesis
The first proposed pathway for the relationship between depression and chronic
pain is that depression is responsible for the onset or maintenance of pain in indi-
viduals who suffer from both sets of symptoms. This hypothesis, often termed the
‘antecedent hypothesis’ [2], posits that depression precedes pain. Early studies used
psychogenic conceptualisations of pain to suggest that chronic pain was potentially
a variant of depressive disorder [11], a form of ‘masked’ depression characterised
by continuous pain, denial of emotional and interpersonal difficulties and inability
to tolerate success and happiness [12]. This research has been criticised widely on
both methodological [13] and theoretical [14] grounds. Despite the repudiation
of much of the early research, several more recent studies still suggest that depres-
sion plays a significant role in the aetiology of chronic pain and often precedes the
development of chronic pain [15,16].
Consequence hypothesis
The consequence hypothesis views depression as secondary to chronic pain. Accord-
ing to this view, depressive symptoms follow the onset of pain. This reactive depres-
sion is often seen as the result of an incapacitating physical condition that arises
from the sustained reduction in physical and social activities [17].
Common pathogenesis
The common pathogenesis model assumes that depression and pain, although
clearly distinct conditions, have a shared aetiology. The proposed mechanisms
include key neurotransmitters such as serotonin, noradrenaline, substance P and
corticotropin-releasing factor (CRF) [18]. In a similar fashion, other researchers
have proposed that chronic inescapable stress might be the link between chronic
147 Chronic pain and depression: twin burdens of adaptation
Cognitive–behavioural theories
The relationship between chronic pain and depression has often been explained
within a cognitive–behavioural framework. Here, coping beliefs and behaviours
are considered to play important roles in the patient’s adjustment. In this vein,
thoughts that sustain the ‘illness role’, or the belief that medications and solicitous
responses from others are necessary, have been shown to co-vary with depression.
One frequently discussed set of cognitions in patients with pain is referred to as
‘catastrophising’. Patients who catastrophise expect the worse outcome and worry
excessively about possible negative consequences of events in an effort to defend
against pain exacerbations. These cognitions have been found to be associated
with depression [23]. In the cognitive–behavioural mediation model of depression
[24], the direct relationship between pain and depressed mood is influenced by
cognitive appraisal variables such as perceived interference and lack of self-control.
Patients’ judgements of the extent to which pain affects their ability to participate in
social, recreational, vocational, family and domestic activities, and the satisfaction
they derive from such activities, is referred to as perceived interference [24]. The
cognitive–behavioural mediation model challenged the notion of pain as a variant
of depression and appeared to offer a parsimonious integration of earlier cognitive
and behavioural theories on the relationship of pain and depression [25–28].
Psychoneuroimmunological developments
A comprehensive review of current research provides strong evidence that depres-
sive symptoms can also be conceived of as affective, behavioural and cognitive
148 Depression and physical illness
cells migrate from the periphery into the central nervous system, leading directly
to central sensitisation.
A review of the clinical presentation of depressive disorder suggests that central
sensitisation processes may underlie depressed affect as well as pain. Depression
is persistent within episodes and typically recurrent throughout the lifespan. The
DSM-IV Mood Disorders field trials found that the most frequent course was
‘recurrent, with antecedent dysthymia, without full interepisode recovery’ [75].
Two related hypotheses have been offered to account for the chronicity of the
disorder: the kindling and scar hypotheses.
The scar hypothesis suggests that a depressive episode wears away personal
resources, leaving in its wake a relatively more vulnerable psyche to protect against
future depressions [76]. One area on which a scar is most evident is that of cognitive
attributions. Children who have been depressed show a deterioration of attribu-
tional styles that does not remit, even when the depressive episode has ended [77].
The kindling hypothesis proposes that changes in information processing poten-
tiates depressive processes so that where a stressor may have been present to evoke
the first depressive episode, each new episode is more and more autonomous and
less related to external stimuli [78]. Indeed, the neurochemical changes provoked
by stressors are typically fairly short-lived. However, these changes can be re-elicited
by mild stressor conditions that would have only minor impact on their own [79].
In 2000, Joiner [76] offered an integrative model in which he argued that depression
is characterised by both erosive processes, which corrode psychological resources,
and self-propagating processes, which serve to prolong or exacerbate symptoms
and leave the individual more vulnerable to recurrences.
There is now evidence that cytokines provoke a sensitisation response that can
exert a proactive influence on the development of depression and other forms of
psychopathology. IL-1 has been shown to elicit sensitisation effects in animal
studies, increasing the coexpression of the stress hormones CRF and arginine-
vasopressin (AVP). Upon initial administration of IL-1, increased levels of CRF
and AVP became evident after 4 days and peaked on day 11, although the phenotypic
change was present for several weeks following administration. If the rats were
subject to an additional stressor, in this case foot-shock, then the stress hormone
levels were significantly enhanced, providing support for the hypothesis that peptide
coexpression makes the HPA system more responsive to all sorts of challenges [80].
Administration of TNF- has also been shown to elicit sickness behaviours at
a much lower dosage than is typically required to evoke such behaviours if the
second administration follows the first by 14–28 days. This sensitisation has a
specific timeframe in which it can occur: sensitisation was not evident when the
second TNF- dosage was within seven days of the first [81]. Typically, psychological
stressors have been considered ‘processive’, as they involve the cognitive processing
154 Depression and physical illness
of a situation and require higher cortical functioning. The category of stressor has
now been broadened to include ‘systemic’ or metabolic insults, such as viral and
bacterial infections, which may evoke many of the same neurochemical changes
as the processive stressors. Interestingly, sensitisation occurs when the initial and
subsequent stressors are the same (e.g. instances of loss) and when the stressors
are of different classes (e.g. initial stressor, loss; second stressor, virus). Thus, cross-
sensitisation can occur between stressors and cytokine challenges [79]. In fact, when
systemic stressors occur on a backdrop of processive stress, a synergistic effect may
occur.
Thus, the following picture emerges: When a stressor occurs in sufficient strength,
regardless of whether it is a psychosocial or physiological threat, the organism
mounts a vigorous defence through the immune system, leading to high levels of
circulating cytokines, which can evoke both depressive symptoms and pain as part
of the array of sickness behaviours designed to protect and defend the individual.
In this view, the frequent comorbidity of depression and pain arises because each
symptom is a manifestation of the same homeostatic drive to conserve energy for
survival. This cascade of events may be highly adaptive following an acute stressor
but may become chronic and maladaptive. Central sensitisation processes may
sustain and reinstigate these sickness behaviours in a positive-feedback loop that,
over time, can give rise to depression and pain even without a precipitating threat.
Resilience
If pain and depressive symptoms both originate as processes of adaptation that
are vulnerable to becoming chronic and debilitating when dysregulated, then it
behoves us to consider what can be done to support and restore the self-regulation of
such processes. What do the aforementioned relationships suggest about potential
models of resilience and, relatedly, methods of prevention and intervention? Two
pathways seem particularly critical to the discussion of resilience in the face of most
types of pathology: the preservation of homeostatic boundaries and restoration of
equilibrium. The first pathway, preservation, can be thought of as a mechanism
of primary prevention: how can we preserve the self-regulation of these systems
in order to facilitate a response to threat of sufficient intensity and length to ward
off the danger while retaining the necessary homeostatic elements that bring our
physiology and psychology back to their baseline states? In particular, how do
we sustain the fine distinctions that individuals must make, particularly once the
context itself has become the cue for arousal? For instance, a child growing up in
an abusive environment shows resilience when, having few other options, he or she
can transport him- or herself out of the situation through fantasy and daydreaming.
However, when the child grows up, he or she may no longer be adaptive to resort
155 Chronic pain and depression: twin burdens of adaptation
to fantasy in the face of conflict. Here, the adult’s nervous and immune systems are
forced to make clear distinctions between past and present threats, between his or
her generalised learned fear of conflict and actual danger to the self. Charney [82]
makes the suggestion that resilience may in fact be characterised by an ability to
avoid overgeneralising conditioned stimuli to the larger context, having reversible
storage of emotional memories and being able to facilitate extinction of learned
responses. Psychophysiological flexibility built on complexity and a capacity for
variability in responding may hold the key.
The question of how to facilitate the extinction of learned responses leads us
directly to the second pathway, restorative processes, which allow a system to return
to normal functioning after a period of heightened responsiveness, sensitization,
and maladaption. McEwen and Stellar [83] identify allostatic load as the cumulative
impact that substantially raises health risk due to chronic dysregulations in multiple
systems. Considering how to reduce allostatic load is a preventive intervention as
well, but at a different stage of adaptation. Here we need to identify the ingredients
of recovery as well as the mechanisms for their appropriate utilisation.
Charney [82] offers one framework for the psychobiological mechanisms of
resilience and vulnerability in which he identifies 11 potential mediators of the
psychobiological response to extreme stress. Each of these 11 mediators offers the
possibility of a treatment target, either alone or in functional interactions. Char-
ney suggests that the psychobiological profile of a resilient individual is charac-
terised by high relative levels of dehydroepiandrasterone (DHEA), neuropeptide Y,
galanin, testosterone, and 5-hydroxytryptamine 1a (5-HT1a ) and benzodiazepine
receptor function, and low relative levels of HPA axis activation, CRF, and locus
coeruleus-noradrenaline activity. Based on the mounting evidence that there may be
an endophenotype for resistance to hopelessness and anhedonia in the face of stress,
Charney suggests the potential utility of a wide array of biochemical agents, includ-
ing psychostimulants, dopamine reuptake inhibitors, dopamine receptor agonists
and N-methyl-d-aspartate (NMDA) receptor antagonists to treat the symptoms of
anhedonia and hopelessness in the face of traumatic stress for individuals with a
more vulnerable endophenotype. Future research will continue to elucidate how
the restoration of balance in the hormonal and endocrine systems can alleviate the
negative consequences of stress-related systemic activation.
Emotion complexity
Work in emotional regulation is another promising avenue for interventions to
support and restore homeostatic functioning. At its foundation this work derives
from an understanding of emotions as complex motivational systems of approach
and avoidance that govern cognition and behaviour. Unlike Charney [82], here
the emphasis is on cognitive and affective systems of regulation rather than the
156 Depression and physical illness
among people who are ill, in pain or otherwise under chronic stress will lead to
greater flexibility in coping responses and better functional outcomes.
Negative emotions also have adaptive significance, as they narrow the thought–
action repertoire in response to threat, and thus allowing for a rapid corrective
response. However, positive emotions are necessary to rebound from negative
experiences and return to a more regulated state. In a study demonstrating
the relationship between physiological and psychological resilience, Tugade and
Frederickson [94] found that positive emotions and cognitive appraisals con-
tributed to the ability of resilient individuals to regulate their cardiovascular reac-
tivity quickly in response to negative emotional arousal. Furthermore, they found
that resilience can be taught to individuals who show greater stress reactivity for
a longer duration than people who return to homeostatic functioning more easily
after a threat. Tugade and Frederickson suggest that an intervention that promotes
positive appraisal styles might prove especially useful for building resilience. This is
particularly important in light of a recent study that provided evidence that people
who are dysphoric demonstrate a reduced ability to use mood-incongruent recall
to repair sad moods, even when instructed to do so [95]. Indeed, it has been shown
that people who report higher daily positive mood have more responsive immune
systems than those that report lower positive mood [96], and that people who
are able to regain and maintain positive emotional states are less likely to show
symptoms of ill-health or use medical services during stressful periods [97].
As predicted by the dynamic model of affect [86], the ability to focus inwards
and to identify complex emotions has been shown to increase the ability to regulate
mood [98]. Likewise, greater emotional knowledge, in particular the ability to
discriminate among negative emotions, was associated with larger repertoires of
emotional regulation strategies in a experience-sampling study [99].
Cognitive–behavioural therapy (CBT) is commonly used in pain-management
programmes to assist patients in changing maladaptive ways of thinking and feeling
in response to pain and illness. These therapies encompass a variety of techniques,
including biofeedback, autogenic training, relaxation training, cognitive restruc-
turing, distraction and activity pacing. An extensive literature on the use of CBT
for rheumatoid arthritis has confirmed the utility of CBT for increasing adaptive
pain-coping responses and self-efficacy expectations [100,101] as well as reducing
inflammatory processes [100] and joint pain and swelling [102]. However, a com-
prehensive review of studies of CBT for rheumatoid arthritis has demonstrated
one area of weakness, namely that CBT has not been shown to reduce depression
in patients with pain [103]. In fact, in a well-controlled study by Bradley et al.
[100], depression worsened while pain and disease indices improved as a result
of CBT. The current focus on pain management in these CBT programmes lim-
its their effectiveness by relative inattention to deficits in positive affect resources
158 Depression and physical illness
Conclusion
Depression and pain can be conceived of as two different, but closely related, sets
of symptoms that co-vary with immune activation in response to harm or threat of
harm. When these processes become dysregulated, the physiological, cognitive and
emotional changes engendered by neuroendocrine immune activation can become
chronic and systemic, leading to the maintenance of alarm responses long after
their utility has ended. Overactivation of the HPA axis and monoamine systems
as a result of these alarm responses is particularly implicated in the aetiology of
syndromes of both chronic pain and depression and may maintain these conditions
after the immune system itself has been quieted.
The mounting evidence for the implication of multiple systems in the experience
of and recovery from depression and pain provides a wide array of intervention
possibilities. Depression, pain and immune responses to a perceived threat may
initiate elevations in one another, potentially leading to the dysregulation in multi-
ple systems. Targets for intervention are diverse, including physiological, cognitive
and emotional regulation. Furthermore, a consideration for the interconnectivity
of the systems in the human body urges the adoption of a multifaceted approach
to restoration of homeostasis. Emotional complexity is one such approach that has
shown promise in regulating cognitive, affective and behavioural manifestations of
allostatic load.
REFERENCES
1. B. Fox, Pain and its magnitude. In Pain Management: A Practical Guide for Clinicians, ed.
R. S. Weiner (New York: CRS Press, 2002), pp. 3–8.
159 Chronic pain and depression: twin burdens of adaptation
22. R. M. Sapolsky, Why stress is bad for your brain. Science 273 (1996), 749–50.
23. F. G. Keefe, G. K. Brown, K. S. Wallston, D. S. Caldwell, Coping with rheumatoid arthritis
pain: catastrophizing as a maladaptive strategy. Pain 37 (1989), 51–6.
24. T. E. Rudy, R. D. Kerns, D. C. Turk, Chronic pain and depression: toward a cognitive-
behavioral mediation model. Pain 35 (1988), 129–40.
25. A. T. Beck, Cognitive Therapy and the Emotional Disorders (New York: Harper & Row, 1967).
26. L. P. Rehm, A self-control model of depression. Behav. Ther. 8 (1977), 787–804.
27. C. B. Ferster, A functional analysis of depression. Am. Psychol. 28 (1973), 857–70.
28. P. M. Lewinsohn, Clinical and theoretical aspects of depression. In Innovative Treatment
Methods in Psychopathology, ed. K. S. Calhoun, H. E. Adams, K. M. Mitchell (New York:
John Wiley & Sons, 1974), pp. 121–43.
29. E. P. Zorrilla, L. Luborsky, J. R. McKay, et al., The relationship of depression and stressors
to immunological assays: a meta-analytic review. Brain Behav. Immun. 15 (2001), 199–226.
30. L. Capuron, R. Dantzer, Cytokines and depression: the need for a new paradigm. Brain
Behav. Immun. 17 (2003), S119–24.
31. S. Kent, R. M. Bluthe, K. W. Kelley, R. Dantzer, Sickness behavior as a new target for drug
development. Trends Pharmacol. Sci. 13 (1992), 24–8.
32. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders,
4th edn, text revision (Washington DC: American Psychiatric Association, 2000).
33. B. L. Hart, Biological basis of the behavior of sick animals. Neurosci. Biobehav. Rev. 12 (1988),
123–7.
34. K. W. Kelley, R. M. Bluthe, R. Dantzer, et al., Cytokine-induced sickness behavior. Brain
Behav. Immun. 17: Suppl 1 (2003), 112–18.
35. H. Anisman, L. Kokkinidis, Z. Merali, Further evidence for the depressive effects of cytokines:
anhedonia and neurochemical changes. Brain Behav. Immun. 16 (2002), 544–56.
36. R. M. Bluthé, S. Layé, B. Michaud, et al., Role of interleukin-1 and tumor necrosis factor-
in lipopolysaccharide-induced sickness behaviour: a study with interleukin-1 type I receptor
deficient mice. Eur. J. Neurosci. 12 (2000), 4447–56.
37. S. F. Maier, L. R. Watkins, Intracerebroventricular interleukin-1 receptor antagonist blocks
the enhancement of fear conditioning and interference with escape produced by inescapable
shock. Brain Res. 695 (1995), 279–82.
38. M. Maes, H. Y. Meltzer, E. Bosmans, et al., Increased plasma concentrations of interleukin-6,
soluble interleukin-6, soluble interleukin-2, and transferrin receptor in major depression.
J. Affect. Disord. 34 (1995), 301–9.
39. A. Sluzewska, J. Rybakowski, E. Bosmans, et al., Indicators of immune activation in major
depression. Psychiatry Res. 64 (1996), 161–7.
40. M. Malaguarnera, I. Di Fazio, S. Restuccia, et al., Interferon alpha-induced depression
in chronic hepatitis C patients: comparison between different types of interferon alpha.
Neuropsychobiology 37 (1998), 93–7.
41. M. Malaguarnera, I. Di Fazio, B. A. Trovato, G. Pistone, G. Mazzoleni, Alpha-interferon
(IFN-) treatment of chronic hepatitis C: analysis of some predictive factors for the response.
Int. J. Clin. Pharmacol. Ther. 39 (2001), 239–45.
161 Chronic pain and depression: twin burdens of adaptation
42. S. Bonaccorso, V. Marino, A. Puzella, et al., Increased depressive ratings in patients with
hepatitis C receiving interferon-alpha based immunotherapy are related to interferon-alpha-
induced changes in the serotonergic system. J. Clin. Psychopharmacol. 22 (2002), 86–90.
43. R. de Beaurepaire. Questions raised by the cytokine hypothesis of depression. Brain Behav.
Immun. 16 (2002), 610–17.
44. M. G. Cavallo, P. Pozzilli, R. Thorpe, Cytokines and autoimmunity. Clin. Exp. Immunol. 96
(1994), 1–7.
45. G. A. Hutchinson, J. E. Nehall, D. T. Simeon, Psychiatric disorders in systemic lupus ery-
thematosus. West Indian Med. J. 45 (1996), 48–50.
46. T. Pincus, J. Griffith, S. Pearce, D. Isenberg, Prevalence of self-reported depression in patients
with rheumatoid arthritis. Br. J. Rheumatol. 35 (1996), 879–83.
47. K. G. Kahl, N. Kruse, H. Faller, H. Weiss, P. Rieckmann, Expression of tumor necrosis factor-
and interferon- mRNA in blood cells correlates with depression scores during an acute
attack in patients with multiple sclerosis. Psychoneuroendocrinology 27 (2002), 671–81.
48. Y. Pollak, E. Orion, I. Goshen, H. Ovadia, R. Yirmiya, Experimental autoimmune
encephalomyelitis-associated behavioral syndrome as a model of ‘depression due to multiple
sclerosis.’ Brain Behav. Immun. 16 (2002), 533–43.
49. Y. Pollak, H. Ovadia, E. Orion, R. Yirmiya. The EAE-associated behavioral syndrome II.
Modulation by anti-inflammatory treatments. J. Neuroimmunol. 137 (2003), 100–108.
50. P. D. Wall, Dorsal horn electrophysiology. In Handbook of Sensory Physiology: Somatosensory
System ed. A. Igg. (Berlin: Springer-Verlag, 1973), pp. 253–70.
51. D. D. Price, Psychological and Neural Mechanisms of Pain (New York, Raven, 1988).
52. A. D. Craig, The functional anatomy of lamina I and its role in post-stroke central pain.
In Nervous System Plasticity and Chronic Pain, ed. J. Sandkuhler, B. Bromm, G. F. Gebhart
(Amsterdam: Elsevier, 2000), pp. 137–51.
53. A. D. Craig, Pain mechanisms: labeled lines versus convergence in central processing. Annu.
Rev. Neurosci. 26 (2003), 1–30.
54. S. F. Maier, E. P. Wiertelak, D. Martin, L. R. Watkins, Interleukin-1 mediates behavioral
hyperalgesia produced by lithium chloride and endotoxin. Brain Res. 623 (1993), 321–4.
55. L. R. Watkins, S. F. Maier, Beyond neurons: evidence that immune and glial cells contribute
to pathological pain states. Physiol. Rev. 82 (2002), 981–1011.
56. Y. B. Abramov, A. Y. Kozlocv, O. S. Sinel’shchikova, G. V. Torgovanova, Nociceptive reactions
during stimulation of immunity in rats with different individual sensitivities to stress.
Neurosci. Behav. Physiol. 33 (2003), 821–6.
57. A. J. Zautra, D. C. Yocum, I. Villanueva, et al., Immune activation and depression in women
with rheumatoid arthritis. J. Rheumatol. 31 (2004), 457–63.
58. A. J. Zautra, B. W. Smith, Depression and reactivity to stress in older women with rheumatoid
arthritis and osteoarthritis. Psychosom. Med. 63 (2001), 687–96.
59. L. R. Watkins, S. F. Maier, L. E. Goehler, Cytokine-to-brain communication: a review and
analysis of alternative mechanisms. Life Sci. 57 (1995), 1011–26.
60. L. R. Watkins, S. F. Maier, Implications of immune-to-brain communication for sickness
and pain. Proc. Natl. Acad. Sci. U. S. A. 96 (1999), 7710–13.
162 Depression and physical illness
61. N. Castanon, R. M. Bluthe, R. Dantzer, Chronic treatment with the atypical antidepressant
tianeptine attenuates sickness behavior induced by peripheral but not central lipopolysac-
charide and interleukin-1 in the rat. Psychopharmacology 154 (2001), 50–60.
62. R. Yirmiya, Y. Pollak, O. Barak, et al., Effects of antidepressant drugs on the behavioral and
physiological responses to lipopolysaccharide (LPS) in rodents. Neuropsychopharmacology
24 (2001), 531–44.
63. D. L. Musselman, D. H. Lawson, J. F. Gumnick, et al., Paroxetine for the prevention of
depression induced by high-dose interferon-. N. Engl. J. Med. 344 (2001), 961–6.
64. Z. Xia, J. W. DePierre, L. Nässberger, Tricyclic antidepressants inhibit IL-6, IL-1 and TNF-
release in human blood monocytes and IL-2 and interferon- in T-cells. Immunopharma-
cology 34 (1996), 27–37.
65. M. Maes, C. Song, A. H. Lin, et al., Negative immunoregulatory effects of antidepressants
inhibition of interferon-gamma and stimulation of interleukin-10 secretion. Neuropsy-
chopharmacology 20 (1999), 370–79.
66. J. Sawynok, Antidepressants as analgesics: an introduction. J. Psychiatry. Neurosci. 26 (2001),
20.
67. J. Sawynok, M. J. Esser, A. R. Reid, Antidepressants as analgesics an overview of central and
peripheral mechanisms of action. J. Psychiatry Neurosci 26 (2001), 21–9.
68. R. Lee, P. S. Spencer, Antidepressants and pain: a review of the pharmacological data sup-
porting the use of certain tricyclics in chronic pain. J. Int. Med. Res. 5: Suppl 1 (1977),
146–56.
69. A. E. Panerai, M. Bianchi, P. Sacerdote, et al., Antidepressants in cancer care. J. Palliat. Care
7 (1991), 42–4.
70. J. S. McDaniel, D. L. Musselman, M. R. Porter, D. A. Reed, C. B. Nemeroff, Depression
in patients with cancer diagnosis, biology, and treatment. Arch. Gen. Psychiatry 52 (1995),
88–99.
71. M. E. Lynch, Antidepressants as analgesics a review of randomized controlled trials. J.
Psychiatry Neurosci. 26 (2001), 30–36.
72. A. J. Cook, C. J. Woolf, P. D. Wall, S. B. McMahon, Dynamic receptive field plasticity
in rat spinal cord dorsal horn following C-primary afferent input. Nature 325 (1987),
151–3.
73. J. Ru-Rong, T. Kohno, K. A. Moore, C. J. Woolf , Central sensitization and LTP do pain and
memory share similar mechanisms? Trends Neurosci. 26 (2003), 696–705.
74. B. A. Winkelstein, Mechanisms of central sensitization, neuroimmunology and injury
biomechanics in persistent pain: implications for musculoskeletal disorders. J. Electromyogr.
Kinesiol. 14 (2004), 87–93.
75. M. B. Keller, D. N. Klein, R. M. A. Hirshfeld, et al., Results of the DSM-IV mood disorders
field trial. Am. J. Psychiatry 152 (1995), 843–9.
76. T. E. Joiner, Depression’s vicious scar self-propagating and erosive processes in depression
chronicity. Clin. Psychol. Sci. Pract. 7 (2000), 203–18.
77. S. Nolen-Hoeksema, J. S. Girgus, M. E. P. Seligman, Learned helplessness in children: a
longitudinal study of depression, achievement, and explanatory style. J. Pers. Soc. Psychol.
51 (1986), 435–42.
163 Chronic pain and depression: twin burdens of adaptation
78. K. S. Kendler, L. M. Thornton, C. O. Gardner, Stressful life events and previous episodes in
the etiology of major depression in women: an evaluation of the ‘kindling’ hypothesis. Am.
J. Psychiatry 157 (2000), 1243–51.
79. H. Anisman, Z. Merali, S. Hayley, Sensitization associated with stressors and cytokine treat-
ments. Brain Behav. Immun. 17 (2003), 86–93.
80. F. J. H. Tilders, E. D. Schmidt, Cross-sensitization between immune and non-immune
stressors: a role in the etiology of depression? Adv. Exp. Med. Biol. 461 (1999), 179–97.
81. S. Hayley, K. Brebner, S. Lacosta, Z. Merali, H. Anisman, Sensitization to the effects of
tumor necrosis factor-: neuroendocrine, central monoamine and behavioral variations. J.
Neurosci. 19 (1999), 5654–65.
82. D. S. Charney, Psychobiological mechanisms of resilience and vulnerability: implications
for successful adaptation to extreme stress. Am. J. Psychiatry 161 (2004), 195–216.
83. B. S. McEwen, E. Stellar, Stress and the individual: mechanisms leading to disease. Arch.
Intern. Med. 153 (1993), 2093–101.
84. E. Diener, R. A. Emmons, The independence of positive and negative affect. J. Pers. Soc.
Psychol. 47 (1984), 1105–17.
85. D. Watson Mood and Temperment (New York: Guilford Press, 2000).
86. A. Zautra, B. Smith, G. Affleck, H. Tennen, Examinations of chronic pain and affect
relationships: applications of a dynamic model of affect. J. Consult. Clin. Psychol. 69 (2001),
786–95.
87. A. J. Zautra Emotions, Stress, and Health (New York: Oxford University Press, 2003).
88. G. Affleck, H. Tennen, S. Urrows, P. Higgins, Individual differences in the day-to-day expe-
rience of chronic pain: a prospective daily study of rheumatoid arthritis patients. Health
Psychol. 10 (1991), 419–26.
89. A. J. Zautra, M. H. Burleson, C. A. Smith, et al., Arthritis and perceptions of quality of
life: an examination of positive and negative affect in rheumatoid arthritis patients. Health
Psychol. 14 (1995), 399–408.
90. A. J. Zautra, J. J. Marbach, K. G. Raphael, et al., The examination of myofascial face pain and
its relationship to psychological distress among women. Health Psychol. 14 (1995), 223–31.
91. M. C. Davis, A. J. Zautra, B. W. Smith, Chronic pain, stress and the dynamics of affective
differentiation. J. Person. 72 (2004), 1133–59.
92. J. W. Reich, A. J. Zautra, M. C. Davis, Dimensions of affect relationships: models and their
integrative implications. Rev. Gen. Psychol. 7 (2003), 66–83.
93. A. J. Zautra, L. Johnson, M. C. Davis, Positive affect as a source of resilience for women in
chronic pain. J. Consult. Clin. Psychol. 73 (2005), 212–20.
94. M. M. Tugade, B. L. Frederickson, Resilient individuals use positive emotions to bounce
back from negative emotional experiences. J. Pers. Soc. Psychol. 86 (2004), 320–33.
95. J. Joorman, M. Siemer, Memory accessibility, mood regulation, and dysphoria: difficulties
in repairing sad mood with happy memories? J. Abnorm. Psychol. 113 (2004), 179–88.
96. A. A. Stone, J. M. Neale, D. S. Cox, et al., Daily events are associated with a secretory immune
response to an oral antigen in men. Health Psychol. 13 (1994), 440–46.
97. S. L. Goldman, D. T. Kraemer, P. Salovey, Beliefs about mood moderate the relationship of
stress to illness and symptom reporting. J. Psychosom. Res. 41 (1996), 115–28.
164 Depression and physical illness
98. C. McFarlan, R. Buehler, The impact of negative affect on autobiographical memory: the
role of self-focused attention to moods. J. Pers. Soc. Psychol. 75 (1998), 1424–40.
99. L. Felman Barrett, J. Gross, T. C. Christensen, M. Benvenuto, Knowing what you’re feeling
and knowing what to do about it: mapping the relation between emotion differentiation
and emotion regulation. Cogn. Emotion 15 (2001), 713–24.
100. L. A. Bradley, L. D. Young, K. O. Anderson, et al., Effects of psychological therapy on pain
behavior or rheumatoid arthritis patients. Arthritis Rheum. 30 (1987), 1105–14.
101. J. C. Parker, K. L. Smarr, S. P. Buckelew, et al., Effects of stress management on clinical
outcomes in rheumatoid arthritis. Arthritis Rheum. 38 (1995), 1807–18.
102. V. Radojevic, P. Nicassio, M. Weisman, Behavior therapy with and without family support
for rheumatoid arthritis. Behavi. Ther. 2 (1992), 13–30.
103. P. M. Nicassio, M. A. Greenberg, Psychosocial management of chronic pain in arthritis.
In Treatment of Rheumatic Diseases, 2nd edn, ed. M. H. Weisman, M. Weinblatt, J. Louie
(Orlando, FL: William Saunders, 2001), pp. 147–61.
8
8
The interrelationship of depression
and diabetes
Dominique L. Musselman, Angela Bowling, Natalie Gilles, Hannah
Larsen, Ephi Betan and Lawrence S. Phillips
. . . in the pre-war literature, diminished glucose tolerance was frequently reported to be statistically
demonstrable in patients with melancholia. H. M. Von Praag (1965)
Introduction: diabetes
The worldwide prevalence of diabetes is increasing at an alarming rate. According to
the World Health Organization (WHO), approximately 150 million people world-
wide have diabetes mellitus. Due to the combined impact of sedentary lifestyles,
increased prevalence of obesity, and the rising age of the population, this number
may double by 2025. In developed countries, people aged 65 or older will be at
highest risk; however, most new cases in developing countries will be individuals
between 45 and 64 years of age. Countries reporting the largest number of cases are
India, followed (in order) by China, the USA, Indonesia, Japan, Pakistan, Russia,
Brazil, Italy and Bangladesh [1]
Diabetes mellitus is a heterogeneous metabolic disease in which hyperglycaemia
is a central feature. Diabetes is the world’s leading cause of non-traumatic limb
amputation, new cases of end-stage renal disease [2] and blindness in adults [1,3].
Nearly 50% of the 18 million diabetes patients in the USA are unaware of their
diabetes [4]; in some countries, that number may be as high as 80% [5]. Other
debilitating consequences of diabetes include diabetic neuropathy and foot ulcers.
Diabetes is the fourteenth leading cause of death worldwide and also contributes to
cardiac and stroke-related morbidity and mortality. The actual number of deaths
per year attributed to diabetes is difficult to ascertain but is estimated to be approxi-
mately four million [1].
Type 2 diabetes dwarfs the prevalence of type 1 and other types of diabetes.
In addition to hyperglycaemia, diabetes is also associated with abnormalities in
Epidemiology
Prevalence of depression in
Ref. No. and type of patients Methods individuals with diabetes
∗
Aggregate mean = weighted by number of subjects per study.
patients results in symptom amplification; that is, even when accounting for the
severity of diabetes, diabetic patients with comorbid depression experience more
symptoms associated with their diabetes than do their non-depressed counterparts
[14]. Nevertheless, the increasing recognition of the negative impact of depression
upon the quality of life of those who suffer from diabetes has led to the development
of organisations such as the European Depression in Diabetes Research Consortium
(EDID), which conducts research regarding the most prevalent and costly mental
health problem associated with diabetes – depression [15].
studies, likely due to methodological differences. Some studies include both type 1
and type 2 diabetes patients without distinguishing between the types. Other studies
do not specify the duration of depression, the extent of diabetes-related compli-
cations or the time since the most recent diabetes-related hospitalisation. Most of
the studies relied on self-report questionnaires to ascertain the dimensional sever-
ity of depressive symptoms, while some of the studies used observer-administered
structured clinical interviews to provide the categorical diagnosis of major depres-
sion. The earlier meta-analysis by Gavard et al. [16] of studies of diabetic patients
with comorbid depression, and the more recent one by Anderson et al. [17] (which
included the studies analysed by Gavard) reveal a mean of 14% (range 9–27%), or
an aggregate mean (weighted by the number of subjects in a study) of 9%, of stud-
ies using diagnostic interviews. The studies that utilised self-reporting depressive
symptom scales reported a higher prevalence rate of moderate to severe depressive
symptoms ranging from 22 to 60% (mean 32%) [16] and an aggregate mean of
26% [17]. This almost two-fold difference in prevalence rates of depression between
formal diagnostic assessment and self-reporting of diabetic patients underscores
the importance of structured psychiatric diagnostic interviews in obtaining a con-
servative estimate of the prevalence of depression in diabetic patients.
Another major limitation in these studies of the rates of depression in diabetic
patients is the lack of information with regard to ethnic minority populations.
Despite the fact that prevalence of diabetes in certain minority populations is
nearly twice that in white populations, there is relatively little information regarding
the prevalence of depression in these minority groups [18–20], including Native
Americans [21] and African Americans.
Duration of
follow-up
Study Number of subjects Diagnostic method (years) Findings Comments
East Baltimore site of the 1715 men and women DIS [40] 13 RR: 2.23 (95% CI 0.90 to 5.55)
Epidemiologic Catchment
Area Survey [39]
Kawakami et al. [44] 2764 male employees Zung Depression 8 HR = 2.31 (95% CI 1.03 to
Rating Scale [41] 5.20)
Study of Women’s Health 2254 premenopausal, CES-D [43] 2 OR = 2.81 (95% CI 1.22 to Too few cases of diabetes
Across the Nation middle-aged women from 6.45) for depressed African developed in the other ethnic
(SWAN) [42] ethnically diverse American women groups to reliably examine
backgrounds depression in other ethnic
groups
CES-D, Center for Epidemiological Studies Depression Scale; CI, Confidence interval; DIS = Diagnostic Interview Schedule; OR, odds ratio; RR, relative risk.
171 The interrelationship of depression and diabetes
Anabolic Catabolic
Glucose Glucose
utilisation production
Insulin
Cortisol
Adrenaline and noradrenaline
Growth hormone
Muscle: protein synthesis Glucagon Proteolysis
Fat: triglyceride synthesis Lipolysis
Liver: glycogen storage Glycogenolysis
Gluconeogenesis
Brain
Figure 8.1 Depression and diabetes.
studies by de Groot and colleagues [34] has confirmed that patients with diabetes
and comorbid depression exhibit poorer glycaemic control and greater prevalence
of multiple diabetes complications (sexual dysfunction, neuropathy, retinopathy,
nephropathy, macrovascular disease). In a ten-year study of women (n = 76) with
type 1 or type 2 diabetes, development of coronary heart disease (CHD) was sig-
nificantly more rapid in the women with depression, even after adjusting for other
risk factors for CHD (age, duration of diabetes, body mass index, hypertension,
hyperlipidaemia) [48]. Whether depression acts as a major risk factor or a predictor
for diabetes-related mortality will likely be determined in the near future [49].
Biology
The central nervous system syndrome of major depression is associated with
myriad pathophysiological abnormalities, including increased release and action
of counter-regulatory hormones, perturbations in glucose transport function
and increased secretion of pro-inflammatory cytokines. Such abnormalities may
contribute to insulin resistance and/or pancreatic beta-islet cell dysfunction
(Figure 8.1, Table 8.3).
Brain (hypothalamus)
Corticotropin-releasing factor
Anterior pituitary
Adrenocorticotropin
Growth hormone
Adrenal gland
Adrenal cortex: cortisol
Adrenal medulla: adrenaline and noradrenaline
Autonomic nervous system
adrenaline and noradrenaline
Acetylcholine
Gut
Cholecystokinin
Glucagon-like peptide 1
Pancreas
Alpha-islet cell: glucagon
Beta-islet cell: insulin
Adipose organ
Leptin
Adiponectin
concentrations in fat (brown or white), heart and muscle (red or white). Under usual
conditions, glucose consumption in the brain does not affect glucose transport;
however, glucose transport in the periphery is affected by plasma concentrations
of glucose [70]. Only relatively recently has 13 C-magnetic resonance spectroscopy
(MRS) revealed that the reduced sensitivity to insulin of diabetic patients includes
defects in glucose transport across plasma membranes [71] and an element of
abnormal glycogen synthesis in muscle [72]. Whether such glucoregulatory per-
turbations are due to abnormalities in the GLUT4 transporter, and whether similar
abnormalities of glucose handling and glycogen storage exist in depressed patients
with and without diabetes, remain to be investigated.
Table 8.4 Treatment of depression in diabetes: results of double-blind controlled randomised trials
BGM, blood glucose monitoring; CBT, cognitive–behavioural therapy; ↓d, decrease; HbA1c, haemoglobin
A1c.
moreover, less glucose monitoring was reported by the depressed patients who
received cognitive–behavioural therapy (CBT) than those patients who received
diabetes education alone [31]. However, improved compliance with diabetic treat-
ment and enhanced glycaemic control has been shown to be largely dependent on
appropriate intensification of therapy by providers when patients present for their
follow-up appointments [98,99]. Thus, attendance at scheduled diabetes appoint-
ments (with associated increased intervention by the healthcare team) may be the
critical factor in determining whether effective depression treatment improves gly-
caemic control [100].
Table 8.5 Criteria for testing diabetes in asymptomatic, undiagnosed individuals [120]
1. Testing for diabetes should be considered in all individuals at age 45 years and above and, if normal,
repeated at 3-year intervals.
2. Testing should be considered at a younger age or be carried out more frequently in individuals who:
r are overweight (BMI ≥ 25 kg/m2 )
r have a first-degree relative with diabetes
r are members of a high-risk ethnic population (e.g. African American, Pacific Islander)
r have delivered a baby weighing > 9 lb or have been diagnosed with GDM
r are hypertensive ( ≥ 140/90mmHg)
r have an HDL cholesterol level ≤ 35 mg/dl (0.90 mmol/l) and/or a triglyceride level ≥ 250 mg/dl
(2.82 mmol/l)
r on previous testing had IGT or IFG*
The OGTT or FPG test may be used to diagnose diabetes. The FPG test is easier to implement, less expensive
and more convenient for patients but will miss the diagnosis in many patients who can be recognised with a
2-h OGTT.
∗
The terms IGT and IFG refer to a metabolic stage intermediate between normal glucose function and diabetes:
r IGT = plasma glucose levels during an OGTT are above normal but below those defined as diabetes: 2-h
plasma glucose levels 140–199 mg/dl (7.8–11.1 mmol/l). Plasma glucose < 140 mg/dl (7.8 mmol/l) is the
upper limit of normal glucose levels. A 2-h 75g OGTT plasma glucose level ≥ 200 mg/dl (11.1 mmol/l) is
indicative of diabetes (confirmatory testing must be performed).
r IFG = individuals with fasting glucose levels between 100 mg/dl (5.6 mmol/l) and 125 mg/dl
(6.9 mmol/l) but < 140 mg/dl (7/8 mg/dl). A fasting glucose concentration of < 100 mg/dl (5.6 mmol/l) is
the upper limit of normal glucose levels. A fasting glucose ≥ 126 mg/dl (7 mmol/l) is indicative of diabetes
(confirmatory testing must be performed).
FPG, fasting plasma glucose; GDM, gestational diabetes mellitus; HDL, high-density lipoprotein; IFG, impaired
fasting glucose; IGT, impaired glucose tolerance; OGTT, oral glucose tolerance test.
of this dosage, fluoxetine has been associated with clinically significant reductions
in HbA1c in patients with type 2 diabetes and weight loss [121,122], but not by
12 months [121].
and skill deficits, and less than half of people with diabetes achieve ideal glycaemic
control (HbA1c < 7.0%). To ascertain the efficacy of diabetes self-management edu-
cation in adults with type 2 diabetes, Norris et al. [125] systematically reviewed a
series of randomised controlled trials using diabetes education. Their meta-analysis
revealed that self-management education does improve HbA1c levels in the short
term (i.e. within six months) and that increased contact time with diabetes health-
care providers enhances this beneficial effect. However, glycaemic deterioration
reappears one to three months after the intervention ceases. As the course, severity
and outcome of diabetes are likely to be influenced heavily by psychological factors,
psychological distress is thought to be an important contributor to the lack of long-
term healthy diabetes self-management and subsequent glycaemic control [126],
perhaps due in part to interference with patient recall of the diabetes information
they received.
Psychological distress often can be reduced through the use of CBT. CBT
encompasses a wide range of psychological techniques designed to bring
about change in thinking patterns and behaviours, i.e. stress-management pro-
grammes. In overview, the studies reviewed here implement CBT and/or stress-
management/relaxation interventions (Table 8.6). The primary outcome variables
have typically included measures of psychological symptoms and glycaemic control,
most commonly HbA1c.
Nearly all of the studies administered individual therapies, but Surwit and col-
leagues’ [130] investigation supports the efficacy of out-patient stress-management
training administered in a group format for the improvement of glycaemic con-
trol in patients with type 2 diabetes. The impact of stress management did not
become evident until the 12-month data-collection point. Control and experimen-
tal patients both showed decreased HbA1c up to six months. After that point, the
average HbA1c of control subjects deteriorated, while that of stress-management
subjects continued to improve.
Future randomised, prospective controlled trials of patients with diabetes would
benefit from administration of structured diagnostic interviews in order to ascertain
the presence of major mental disorders. The promise of psychological treatments
is their potential for salutary effects upon neurophysiological mechanisms, e.g.
reduction in stress-induced counter-regulatory hormones and alterations of neu-
ropeptides related to appetite and satiety. In this regard, CBT, the most well-known
structured manual-driven treatment for depression, may contribute to glycaemic
control, although data supporting this conclusion are sparse. In the ten-week con-
trolled study administering weekly CBT for diabetic patients with major depression
(see Tables 8.4 and 8.6), both depressive symptoms and HbA1c improved [94].
Compared with patients receiving only diabetic education (the control group),
patients receiving CBT and diabetic education exhibited significantly greater clini-
cal improvement on the BDI (−19 vs. −7) at the end of the 10-week intervention;
Table 8.6 Controlled trials of psychological interventions in patients with diabetes
Lane et al. Type 2 n = 38 (26 To examine the 8 weekly 50-minute Eysenck By week 48, Both groups Biofeedback
[127] male, 12 female) benefits of biofeedback Personality individuals who achieved training did not
patients with biofeedback- training sessions Inventory, Nowicki responded best to significant improve heart rate
poorly controlled assisted relaxation (with 4 monthly Strickland Locus of relaxation training reductions in and blood pressure
diabetes training and to additional session) Control, STAI, with HbA1c, but there of intervention
investigate patient + diabetes HbA1c, glucose improvements in was no difference group; prevalence
characteristics that education vs. tolerance glucose tolerance between the groups of major
could predict a diabetes education were those with depression
positive response alone higher trait anxiety unknown in study
to relaxation and more population
training emotional lability
Spiess et al. Type 2 n = 23 To assess whether a Distress-reduction DSM-III-R By 3 months No differences in Elevated depressive
[128] (14 male, 9 female) distress-reduction programme (25 interview, BDI, follow-up, HbA1c levels symptoms in study
patients with programme weekly 90-minute LES, STAI, HbA1c improved between the groups subjects (mean
poorly controlled improves group sessions) + depressive and at 3 and 9 months BDI > 26), but no
diabetes psychological diabetes education anxiety symptoms follow-up, but study subject
variables and vs. diabetes in treatment group HbA1c of both fulfilled DSM-III-R
whether better education compared with groups improved criteria for major
coping strategies control group. By 9 significantly from depression
may be associated months’ follow-up, baseline
with improved depressive and
metabolic control anxiety symptoms
of two groups
similar to
pretreatment levels
Lustman Type 2 n = 51 (26 To assess the 10 weeks of DIS, BDI, HbA1c By 6 months At immediate All subjects
et al. [94] male, 25 female) efficacy of CBT for individual CBT + levels, self- follow-up, more post-treatment fulfilled categorical
with major depression in diabetes education monitoring patients in CBT follow-up, no diagnosis of major
depression; 41 patients with vs. diabetes compliance (58%) achieve significant depression
complete diabetes education alone remission of difference in
depressive HbA1c between
symptoms vs. treatment and
control (26%) control groups
(p = 0.03) By 6 months
follow-up, HbA1c
levels decrease by
0.7% in CBT group
and increase by
0.9% in control
group (p = 0.04)
Pouwer Types 1 and 2 To investigate During 12-month WBQ-12 Monitored patients No differences for Prevalence of
et al. [129] n = 345 (168 whether period, subjects reported improved HbA1c levels major depression
male, 177 female) monitoring of completed general wellbeing between the two unknown in study
psychological computerised and groups population
wellbeing improves self-report
mood or glycaemic psychological
control questionnaires at
initial, interim and
12-month visit.
Psychological
results discussed
with patient by
diabetes nurse +
standard care vs.
standard care alone
(cont.)
Table 8.6 (cont.)
Surwit et al. Type 2 n = 108 To determine 5 sessions of STAI, Perceived By 1 year Difference between Group therapy
[130] (63 male, 45 whether a weekly group Stress Scale, follow-up, no groups in HbA1c improves
female), 72 stress-management stress-management General Health improvement in appeared at 1 year glycaemic control
complete programme can therapy + diabetes Questionnaire, anxiety even follow-up, i.e. over time, but may
improve glucose education vs. HbA1c levels among highly stress- take several
metabolism diabetes education anxious patients at management- months to appear,
alone baseline trained patients is modest in effect
exhibited and is not due to
approximately a changes in exercise,
0.5% reduction in diet or BMI.
HbA1c compared Prevalence of
with control major depression
patients unknown in study
population
Williams Type 1 and 2 n = Multicentre study Depression care DSM-IV By 12 month By 12 months Study sample
et al. [131] 417 (194 male; 223 to determine management structured follow-up, follow-up, exhibited relatively
female) with major whether depression (educational interview, treatment group intervention did good glycaemic
depression or treatment materials + 6–8 Symptom exhibited not improve control at baseline
dysthymia; 232 improves mood brief sessions of Checklist-20, significantly glycaemic control (HbA1c level: 7.3
complete and diabetic depression HbA1c levels improved +/− 1.4%).
outcomes in adults psychotherapy) depressive Diabetes status
with diabetes and over 12-month symptoms and based on patient
depression period vs. usual overall function self-report
care for depression Intervention group
significantly more
likely to utilise
antidepressant or
psychotherapy
BDI, Beck Depression Inventory [10]; BDQ, Barriers in Diabetes Questionnaire; CBT, cognitive–behavioural therapy; DIS, National Institute of Mental Health
(NIMH) Diagnostic Interview Schedule [132]; DSCI, Diabetes Self-care Inventory; DSM-III-R, Diagnostic and Statistical Manual, 3rd edition, revised [133]; HbA1c,
haemoglobin A1c; HFS, Fear of Hypoglycemia Survey; LES, Life Events Scale [134]; PAID, Problem Areas in Diabetes; SCL-90, Symptom Checklist 90 [135]; STAI,
State-Trait Anxiety Inventory [136]; WBQ-12, Well-Being Questionnaire [8].
184 Depression and physical illness
Acknowledgements
This research was supported by grants RR-00039 from the National Insti-
tutes of Health, Bethesda, MD, and HS-07922 (Dr Phillips), and MH-069254
186 Depression and physical illness
(Dr Musselman). We are grateful for the assistance of Ms Erica Bruce, BSA, in the
preparation of the manuscript, and the nursing and laboratory staff of the Emory
General Clinical Research Center, Emory University Hospital, and the Diabetes
Clinic at Grady Memorial Hospital.
REFERENCES
1. World Health Organization. Diabetes: The Cost of Diabetes (Geneva: World Health Organi-
zation, 2002).
2. Diabetes Control and Complications Trial Research Group. The effect of intensive treatment
of diabetes on the development and progression of long-term complications in insulin-
dependent diabetes mellitus. N. Engl. J. Med. 329 (1993), 977–86.
3. R. Klein, B. E. K. Klein, S. E. Moss, M. D. Davis, D. L. DeMets, The Wisconsin Epidemiologic
Study of Diabetes Retinopathy III: prevalence and risk of diabetic retinopathy when age at
diagnosis is 30 or more years. Arch. Opthalmol. 102 (1984), 527–32.
4. Associated Press. Estimate of Americans with pre-diabetes doubles. Wall Street Journal, April
29, 2004.
5. World Health Organization. Laboratory Diagnosis and Monitoring of Diabetes Mellitus
(Geneva: World Health Organization, 2002).
6. Guidelines Subcommittee of the WHO/ISH Mild Hypertension Liaison Committee. Guide-
lines for the management of mild hypertension. Hypertension 22 (1993), 392–403.
7. J. S. McDaniel, D. L. Musselman, M. R. Porter, D. A. Reed, C. B. Nemeroff, Depression in the
cancer patient: a commentary on diagnosis, biology, and treatment. Arch. Gen. Psychiatry
52 (1995), 89–99.
8. C. Bradley, The well-being questionnaire. In Handbook of Psychology and Diabetes, ed.
C. Bradley (chur, Switzerland: Harwood Academic Publishers, 1994), pp. 89–109.
9. P. J. Lustman, H. Amado, R. D. Wetzel, Depression in diabetics: a critical appraisal. Comp.
Psychiatry 1 (1983), 65–74.
10. A. T. Beck, C. H. Ward, M. Mendelson, J. Mock, J. Erbaugh, An inventory for measuring
depression. Arch. Gen. Psychiatry 4 (1961), 561–71.
11. P. J. Lustman, L. S. Griffith, J. A. Gavard, R. E. Clouse, Depression in adults with diabetes.
Diabetes Care 15 (1992), 1631–9.
12. P. J. Lustman, K. E. Freedland, R. M. Carney, B. A. Hong, R. E. Clouse, Similarity of depression
in diabetic and psychiatric patients. Psychosom. Med. 54 (1992), 602–11.
13. P. J Lustman, L. S. Griffith, R. E. Clouse, et al., Effects of nortriptyline on depression and
glucose regulation in diabetes: results of a double-blind, placebo-controlled trial. Psychosom.
Med. 59 (1997), 241–50.
14. P. Ciechanowski, W. Katon, J. Russo, The relationship of depressive symptoms to symptom
reporting, self-care and glucose control in diabetes. Diabetes Care 25 (2002), 731–6.
15. L. E. Edge, D. Zheng, K. Simpson, Comorbid depression is associated with increased health
care use and expenditures in individuals with diabetes. Diabetes Care 25 (2002), 464–70.
187 The interrelationship of depression and diabetes
107. S. P. Roose, G. W. Dalack, Treating the depressed patient with cardiovascular problems.
J. Clin. Psychiatry 53 (1992), 25–31.
108. S. H. Sindrup, L. F. Gram, K. Brosen, O. Eshoj, E. F. Mogensen, The selective serotonin
reuptake inhibitor paroxetine is effective in the treatment of diabetic neuropathy symptoms.
Pain 42 (1990), 135–44.
109. C. L. DeVane, J. S. Markowitz, Psychoactive drug interactions with pharmacotherapy for
diabetes. Psychopharmacol. Bull. 36 (2002), 40–52.
110. L. M. Tremaine, K. D. Wilner, S. H. Preskorn, A study of the potential effect of sertraline on
the pharmacokinetics and protein binding of tolbutamide. Clin. Pharmacokinet. 32: Suppl
1 (1992), 31–6.
111. N. Sussman, D. L. Ginsberg, J. Bikoff, Effects of nefazodone on body weight: a pooled
analysis of selective serotonin reuptake inhibitor-and imipramine-controlled trials. J. Clin.
Psychiatry 62 (2001), 256–60.
112. M. Fava, Weight gain and antidepressants. J. Clin. Psychiatry 61: Suppl 11 (2000), 37–41.
113. M. D. Marcus, R. R. Wing, J. Guare, E. H. Blaire, A. Jawad, Lifetime prevalence of major
depression and its effect on treatment outcome in obese type II diabetic patients. Diabetes
Care 15 (1992), 253–5.
114. D. L. Rowland, L. Myers, A. Culver, J. M. Davidson, Bupropion and sexual function: a
placebo-controlled prospective study on diabetic men with erectile dysfunction. J. Clin.
Psychopharmacol. 17 (1997), 350–57.
115. C. E. Koro, D. O. Fedder, G. J. L’Italien, et al., Assessment of independent effect of olanzapine
and risperidone on risk of diabetes among patients with schizophrenia: population based
nested case–control study. Br. Med. J. 325 (2002), 243–7.
116. P. S. Masand, Atypical antipsychotics and DKA: issues and controversies. Thomson Profes-
sional Postgraduate Services 2 (2002), 1–3.
117. M. J. Serynak, D. L. Leslie, R. D. Alarcon, M. F. Losonczy, R. Rosenheck, Association of
diabetes mellitus with use of atypical neuroleptics in the treatment of schizophrenia. Am. J.
Psychiatry 159 (2002), 561–6.
118. J. I. Isojarvi, J. Rattya, V. V. Myllyla, et al., Valproate, lamotrigine, and insulin-mediated
risks in women with epilepsy. Ann. Neurol. 43 (1998), 446–51.
119. H. Jin, J. M. Meyer, D. V. Jeste, Phenomenology of and risk factors for new-onset diabetes
mellitus and diabetic ketoacidosis associated with atypical antipsychotics: an analysis of 45
published cases. Ann. Clin. Psychiatry 14 (2002), 59–64.
120. Report of the expert committee on the diagnosis and classification of diabetes mellitus.
Diabetes Care 26: Suppl 1 (2003), 5–20.
121. M. O’Kane, P. G. Wiles, J. K. Wales, Fluoxetine in the treatment of obese type 2 diabetic
patients. Diabetes Med. 11 (1994), 105–10.
122. L. Breum, U. Bjerre, J. F. Bak, S. Jacobsen, A. Astrup, Long-term effects of fluoxetine on
glycemic control in obese patients with non-insulin-dependent diabetes mellitus or glucose
intolerance: influence on muscle glycogen synthase and insulin receptor kinase activity.
Metabolism 44 (1995), 1570–76.
123. P. Mahuex, F. Ducros, J. Bourque, J. Garon, J.-L. Chiasson, Fluoxetine improves insulin
sensitivity in obese patients with non-insulin-dependent diabetes mellitus independently
of weight loss. Int. J. Obes. 21 (1997), 97–102.
193 The interrelationship of depression and diabetes
124. B. J. P. van Loon, J. K. Radder, M. Frolich, Fluoxetine increases insulin action in obese
nondiabetic and in obese noninsulin-dependent diabetic individuals. Int. J. Obes. 16 (1992),
79–85.
125. S. L. Norris, J. Lau, S. J. Smith, C. H. Schmid, M. M. Engelgau, Self-management education
for adults with type 2 diabetes: a meta-analysis of the effect on glycemic control. Diabetes
Care 25 (2002), 1159–71.
126. B. A. Hamburg, G. E. Inoff, Coping with predictable crises of diabetes. Diabetes Care 6
(1983), 409–16.
127. J. D. Lane, C. C. McCaskill, S. L. Ross, M. N. Feinglos, R. S. Surwit, Relaxation training for
NIDDM: predicting who may benefit. Diabetes Care 16 (1993), 1087–94.
128. K. Spiess, G. Sachs, P. Pietschmann, R. Prager, A program to reduce onset distress in unse-
lected type I diabetic patients: effects on psychological variables and metabolic control. Eur.
J. Endocrinol. 132 (1995), 580–86.
129. F. Pouwer, F. J. Snoek, H. M. van der Ploeg, H. J. Ader, R. J. Heine, Monitoring of psycho-
logical well-being in outpatients with diabetes: effects on mood, HbA(1c), and the patient’s
evaluation of the quality of diabetes care: a randomized controlled trial. Diabetes Care 24
(2001), 1929–35.
130. R. S. Surwit, M. A. L. Van Tilburg, N. Zucker, et al., Stress management improves long-term
glycemic control in type 2 diabetes. Diabetes Care 25 (2002), 30–34.
131. J. W. Williams, Jr, W. Katon, E. H. Lin, et al., The effectiveness of depression care management
on diabetes-related outcomes in older patients. Ann. Intern. Med. 140 (2004), 1015–24.
132. L. N. Robins, J. E. Helzer, J. Croughnan, J. B. W. Williams, R. L. Spitzer, NIMH Diagnostic
Interview Schedule: Version III-R (Rockville, MD: National Institutes of Mental Health,
1987).
133. American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders,
4th edn. (Washington, DC: American Psychiatric Association, 1994).
134. E. S. Paykel, Methodology of live events research. Adv. Psychosom. Med. 17 (1987), 13–
29.
135. W. A. Arrundell, SCL-90: Manual for a Multi-dimensional Indicator of Psychopathology (Lisse:
Swets and Zeitinger, 1986).
136. C. D. Spielberger, R. L. Gorsuch, R. Luschene, Test Manual for the State-Trait Anxiety Inven-
tory (Palo Alto, CA: Consulting Psychologists Press, 1970).
137. M. B. Keller, J. P. McCullough, D. N. Klein, et al., A comparison of nefazodone, the cognitive
behavioral-analysis system of psychotherapy, and their combination for the treatment of
chronic depression. N. Engl. J. Med. 342 (2000), 1462–70.
138. H. C. Schulberg, W. Katon, G. E. Simon, J. Rush, Treating major depression in primary care
practice: an update of the Agency for Health Care Policy and Research Practice Guidelines.
Arch. Gen. Psychiatry 55 (1998), 1121–7.
139. R. S. Surwit, R. B. Williams, I. C. Siegler, J. D. Lane, M. Helms, Hostility, race, and glucose
metabolism in nondiabetic individuals. Diabetes Care 25 (2002), 835–9.
140. V. W. Persky, J. Kempthorne-Rawson, R. B. Shekelle, Personality and risk of cancer: 20-year
follow-up of the Western Electric study. Psychosom. Med. 49 (1987), 435–49.
141. P. J. Lustman, L. S. Griffith, R. E. Clouse, et al., Improvement in depression is associated
with improvement in glycemic control. Diabetes 4: Suppl 1 (1995), 27A.
194 Depression and physical illness
142. R. Sinha, G. Fisch, B. Teague, et al., Prevalence of impaired glucose tolerance among children
and adolescents with marked obesity. N. Engl. J. Med. 346 (2002), 802–10.
143. B. S. McEwen, A. M. Magarinos, L. P. Reagan, Studies of hormone action in the hippocampal
formation: possible relevance to depression and diabetes. J. Psychosom. Res. 53 (2002), 883–
90.
144. W. H. Gispen, G.-J. Biessels, Cognition and synaptic plasticity in diabetes mellitus. Trends
Neurosci. 23 (2000), 542–9.
145. B. E. Levin, A. A. Dunn-Meynell, V. H. Routh, Brain glucose sensing and body energy
homeostasis: role in obesity and diabetes. Am. J. Physiol. Endocrinol. Metab. 276 (1999),
R1223–31.
146. B. E. Levin, Glucosensing neurons do more than just sense glucose. Int. J. Obes. 25: Suppl
5 (2001), 68–72.
147. A. A. Dunn-Meyell, N. E. Rawson, B. E. Levin, Distribution and phenotype of neurons
containing the ATP-sensitive K+ channel. Brain Res. 814 (1998), 41–54.
148. P. J. Lustman, K. E. Freedland, L. S. Griffith, R. E. Clouse, Fluoxetine for depression in
diabetes: a randomized double-blind placebo-controlled trial. Diabetes Care 23 (2000),
618–23.
9
9
Depression and chronic fatigue
Peter D. White
Introduction
This chapter reviews fatigue as a symptom and some of its syndromes, includ-
ing chronic fatigue syndrome and vital exhaustion. The chapter also reviews the
similarities and differences between fatigue and depression. The links include com-
mon symptoms, sleep disturbance, physical inactivity and common treatments
with both cognitive – behavioural therapy (CBT) and graded exercise therapy. The
differences involve aetiology, pathophysiology, nosology and response to antide-
pressants. Depression and fatigue should be considered as separate phenomena
with common presentations and associations.
Epidemiology of fatigue
Fatigue is a common symptom in both the community and primary care. Between
10% and 20% of people in the community, if asked, will admit to feeling abnormally
tired at any one time [1]. At the same time, fatigue is distributed continuously within
the community, with no point of rarity [2]. Therefore, any cut-off is arbitrary, and
the prevalence will vary by definition of fatigue, how the question is asked or the
symptom volunteered, and its context; for instance, fewer people volunteer fatigue
to their general practitioner (GP). Studies of fatigue reporting to primary-care
doctors from Ireland and Holland calculated annual incidences of 6.5 and 5.3
reports per 100 patients, respectively [3,4]. One study showed that the incidence of
the complaint of fatigue and its synonyms (e.g. tired, worn out, exhausted) recorded
by GPs in the UK did not change in the 12 years between 1990 and 2001, being
reported by 1.5% of GP attenders per annum [5]. The differences in incidence were
probably due to different denominator populations, particularly in age ranges, with
the later study involving patients of all ages rather than only adults.
Depression and Physical Illness, ed. A. Steptoe.
Published by Cambridge University Press. C Cambridge University Press 2006.
196 Depression and physical illness
Women report fatigue more commonly than men. Fatigue is uncommon before
puberty [5]. The commonest aetiological associations with the complaint of fatigue
are depression, anxiety and reported ‘stress’ [6]. Physical diseases or states are
uncommon causes, but thyroid disease, pregnancy, breastfeeding and certain infec-
tions, such as infectious mononucleosis, are well described causes of both acute and
persistent fatigue [1,7].
failed to help them. The other possibility is that the process of being a member
and, for example, hearing stories of non-recovery reduces motivation and hope for
recovery. This may particularly be the case if one of the aims of the support group
is public advocacy to legitimise the illness as a real and chronic disabling disease
that cannot be cured [23–25].
The role of abnormal illness behaviour in predisposing to or maintaining CFS
is uncertain. Patients with CFS have higher scores on measures of hypochondriasis
and disease conviction when compared with general practice controls [26] but
not when compared with chronically ill controls suffering from multiple sclerosis
[27]. Hall and colleagues [28] reported a case – control study of the pre-claim
medical reports of 133 claimants of income protection, diagnosed as suffering
from CFS. They compared these CFS claimants with multiple sclerosis claimants
and non-claimants, controlling for age but not for gender. They found that CFS
claimants had reported significantly more illnesses at the time of proposal for
insurance compared with the two comparison groups. The illnesses were many
and various, but lethargy and infections (glandular fever, upper-respiratory-tract
infections, chest infections) were reported most commonly. This group also made
significantly more claims between their accepted proposal and the diagnosis of
CFS. There were several methodological weaknesses in this study: There was a
significantly greater proportion of women in the CFS group. It is well known that
women report more symptoms in general [29], and fatigue in particular [30], than
men. Without a comparison group of sufferers with another ‘functional’ disorder,
we cannot know whether the differences were specific to CFS or to a more general
problem. The same group went on to study GP attenders with CFS or ME, and
replicated their findings of excess consultations for those with CFS/ME up to 15
years before diagnosis compared with patients with multiple sclerosis [31].
CFS is thus an ‘orphan’ illness, belonging entirely to neither neurological nor psy-
chiatric classifications [43].
Of relevance to this, multivariate analytical approaches to symptoms and other
clinical features support the apparent independence of CFS from mood disorders.
Hickie and colleagues [45] showed that both the individuals and clinical presenta-
tions of attenders at a chronic fatigue clinic could be divided into two main groups:
three-quarters of patients had symptoms similar to descriptions of neurasthenia,
with mainly fatigue-related symptoms; a quarter of patients had a larger number of
symptoms, with greater chronicity, whom the authors thought had presentations
more consistent with a somatoform disorder. The authors replicated this work with
patients from an international cross-section of mainly fatigue clinics [46]. Other
studies have supported the independence and reliability over time of a chronic
or persistent fatigue syndrome in primary care or the community using similar
methodologies [38,39,41,47]. These studies all suggest that the nosology of fatigue
syndromes is different from that of depressive illness.
Another approach to testing the independence of CFS from mood disorders is to
study the development of CFS in high-risk populations. The next section describes
such an approach.
than depressive disorder [53,54]. Symptoms were consistent with acute sickness
behaviour [55], suggesting that cytokines might be involved.
Most importantly, in differentiating CFS from depressive disorder, the authors
showed that an empirically derived fatigue syndrome was predicted by different
risk factors than depressive illness [56]. The fatigue syndrome was predicted most
consistently by the immune response close to onset and afterwards by objective
evidence of physical deconditioning, the latter finding being replicated by others
[57]. In contrast, depressive illness was predicted by life events and difficulties, a
past psychiatric history and an emotional personality [56].
Buchwald and colleagues have reported a US cohort of 150 patients studied after
acute IM also in primary care [58]. They found that 12% had not recovered by
six months [59,60]. Fatigue and excessive sleep were also the most prominent con-
valescent symptoms. Buchwald and colleagues went on to examine the symptoms
associated with fatigue in the same sample, followed up to four years after onset [61].
They used principal components analyses to separate the symptoms into factors
and latent class analyses to separate out the patients. They found that patients with
IM went on to have two fatigue syndromes – one with hypersomnia and no pain,
the other with insomnia and pain. The latter was still discernable and independent
four years after onset. Both the syndromes were independent of mood disorder.
Depressive illnesses are reported following infections, but much less commonly
than fatigue, and they are usually transient [7,60] and not as clearly related to the
infection [48]. It is clear that although infections are aetiological agents in fatigue,
this is unlikely to be the case for depression.
three scales combined. This methodologically robust method showed that vital
exhaustion and depression were not independent in patients with heart disease.
This finding has been replicated more recently. McGowan and colleagues [66]
examined 305 consecutive inpatients about four days after their first myocardial
infarction. They confirmed that vital exhaustion, measured by the Maastricht ques-
tionnaire, was correlated highly with the Hospital Anxiety and Depression Scale
(HADS) depression subscale. A factor analysis of the Maastricht questionnaire
showed four independent factors: fatigue, depression, poor concentration and sleep
difficulties. The fatigue factor was still correlated highly with the HADS depression
subscale score. The authors recommended that future studies should study fatigue
separately from vital exhaustion and depression. Prescott and colleagues [67] found
that both exhaustion and depression items of a vital exhaustion measure were related
to later deaths from IHD in a large community sample.
The influence of depression on IHD is reviewed in Chapters 3 and 4. Rele-
vant and shared pathophysiological mechanisms, found to be associated with vital
exhaustion, may include altered haemostasis and cytokine expression [68–70].
The independence of vital exhaustion and CFS has not been examined directly,
but the close correlation between vital exhaustion and depression, fatigue as an
independent factor subsumed in vital exhaustion, and demoralisation being a cen-
tral feature of vital exhaustion all suggest some separation between CFS and vital
exhaustion, even though they share features of both fatigue and sleep disturbance
[71].
latency does not occur in CFS [35]. Sleep in CFS is consistent with psychophysio-
logical disturbance, rather than any specific sleep architecture [35], which may also
occur in less severe depressive illness. Hypersomnia is particularly associated with
post-infectious fatigue [52,53] but also occurs in atypical depression [80].
The role of the immune system in both CFS and depressive illness is being studied
more now than previously [81,82]. Immunological abnormalities are being found
in both disorders, particularly in the innate and cell-mediated immune systems
[81,83]. Cytokines can induce both depression and fatigue [84,85]. However, no
researchers have demonstrated that these immunological abnormalities are func-
tionally important, with regard to more infections or abnormal responses to anti-
gens, or that they are associated with features of either illness [86]. These immune
changes may be caused by the sleep disturbance and inactivity, themselves caused
by depression and CFS [82,87].
of efficacy [88,89]. One large RCT found that fluoxetine did not even help the
depression associated with CFS [98]. Certain antidepressants may be useful in CFS
when there is associated insomnia or pain to the extent of also meeting criteria for
fibromyalgia [99].
Secondary fatigue
Fatigue is commonly associated with chronic medical disorders, but it should be
differentiated from fatiguability. Fatiguability is the onset of a physical sensation of
fatigue and weakness after exertion and is commonly reported with neurological
diseases such as multiple sclerosis and myopathies. Particularly when fatiguability
is associated with post-exertional weakness, neurological diseases are important
differential diagnoses.
The diseases with high risks for fatigue (found in the majority of patients) are
hormonal (hypothyroidism), metabolic (diabetes mellitus), neoplastic (haema-
tological malignancies), infectious (human immunodeficiency virus [HIV]) and
inflammatory (systemic lupus erythematosus). There have been a considerable
number of studies of the associations of ‘secondary’ fatigue [100–105]. Apart from
measures of disease activity, other associations repeatedly found include sleep dis-
turbance, mood disorders, inactivity and physical deconditioning [100–102,106].
Studies of fatigue associated with multiple sclerosis are instructive and exemplary.
As in all studies of secondary fatigue, measures of the severity or pathophysiology
of the disease itself are associated with fatigue [107]. Some cytokines are associated,
but others are not [108]. Associations vary depending on the fatigue measure,
confirming the multidimensional nature of fatigue, but all measures were associated
with depression [109]. Fatigue is associated independently with both depression
and disability [110]. Objectively confirmed sleep disturbance is associated with
fatigue [111].
There have been a number of studies of various treatments aimed at reversing
the associations of secondary fatigue in general, in the hope they would help fatigue
directly, with variable results. The most consistent evidence of efficacy has been with
graded exercise programmes and CBT. As is the case with CFS, antidepressants seem
to help depression but not fatigue [112], with the one possible exception of fatigue
associated with multiple sclerosis [113].
Conclusions
This chapter has reviewed the similarities and differences between fatigue and
depression. Our current lack of understanding of these symptoms or conditions is
exacerbated by the likelihood that both are heterogeneous in either form or causes.
204 Depression and physical illness
Setting that aside, we can conclude that fatigue and depression can be differentiated
from each other in several ways. The syndromes of fatigue and depression share
symptoms, which leads to easy misdiagnosis and inflated comorbidity. The concept
of vital exhaustion represents aspects of both depression and fatigue, which should
be measured separately.
The practical implications are important. Clinicians should assess their patients
carefully and differentiate those with one or other condition as their sole problem,
while remembering that there will be those who seem to have both at the same
time. Treatments will be both different and similar, but should be tailored to the
individual needs of the patient.
REFERENCES
1. S. C. Wessely, M. Hotopf, M. Sharpe., Chronic Fatigue and its Syndromes (Oxford: Oxford
University Press, 1998).
2. T. Pawlikowska, T. Chalder, S. R. Hirsch, et al., Population based study of fatigue and
psychological distress. Br. Med. J. 308 (1994), 763–6.
3. W. Cullen, Y. Kearney, G. Bury, Prevalence of fatigue in general practice. Ir. J. Med. Sci. 171
(2002), 10–12.
4. E. Kenter, I. Okkes, S. Oskam, H. Lamberts, Tiredness in Dutch family practice: data on
patients complaining of and/or diagnosed with ‘tiredness’. Fam. Pract. 20 (2003), 434–40.
5. A. M. Gallagher, J. M. Thomas, W. T. Hamilton, P. D. White, The incidence of fatigue
symptoms and diagnoses presenting in UK primary care from 1990 to 2001. J. R. Soc. Med.
97 (2004), 571–5.
6. M. K. Chen, The epidemiology of self-perceived fatigue among adults. Prev. Med. 15 (1986),
74–81.
7. P. D. White, J. M. Thomas, J. Amess, et al., Incidence, risk and prognosis of acute and
chronic fatigue syndromes and psychiatric disorders after glandular fever. Br. J. Psychiatry
173 (1998), 475–81.
8. L. Darbishire, L. Ridsdale, P. T. Seed, Distinguishing patients with chronic fatigue from
those with chronic fatigue syndrome: a diagnostic study in UK primary care. Br. J. Gen.
Pract. 53 (2003), 441–5.
9. Royal College of Psychiatrists. Royal College of Physicians and Royal College of General
Practitioners, Chronic Fatigue Syndrome: Report of a Joint Working Group of the Royal Colleges
of Physicians, Psychiatrists and General Practitioners (London, Royal College of Physicians,
1996)
10. CFS/ME working group. Report of the CFS/ME Working Group to the Chief Medical Officer,
2001. www.dh.gov.uk/assetroot/04/05/95/06/04059506.pdf
11. W. C. Reeves, A. Lloyd, S. D. Vernon, et al., Identification of ambiguities in the 1994 chronic
fatigue syndrome research case definition and recommendations for resolution. BMC Health
Serv Res 3 (2003), 25.
205 Depression and chronic fatigue
66. L. McGowan, C. Dickens, C. Percival, et al., The relationship between vital exhaustion,
depression and comorbid illnesses in patients following first myocardial infarction. J. Psy-
chosom. Res. 57 (2004), 183–8.
67. E. Prescott, C. Holst, M. Grønbaek, et al., Vital exhaustion as a risk factor for ischaemic
heart disease and all-cause mortality in a community sample: a prospective study of 4084
men and 5479 women in the Copenhagen City Heart Study. Int. J. Epidemiol. 32 (2003),
990–97.
68. W. J. Kop, K. Hamulyák, C. Pernot, A. Appels, Relationship of blood coagulation and
fibrinolysis to vital exhaustion. Psychosom. Med. 60 (1998), 352–8.
69. A. Van der Ven, R. van Diest, K. Hamulyák, et al., Herpes viruses, cytokines, and altered
hemostasis in vital exhaustion. Psychosom. Med. 65 (2003), 194–200.
70. P. Jeanmonod, R. von Känel, F. E. Maly, J. E. Fischer, Elevated plasma C-reactive protein in
chronically distressed subjects who carry the A allele of the TNF-alpha-308 G/A polymor-
phism. Psychosom. Med. 66 (2004), 501–6.
71. R. Van Diest, W. P. Appels, Sleep physiological characteristics of exhausted men. Psychosom.
Med. 56 (1994), 28–35.
72. A. J. Cleare, D. Blair, S. Chambers, S. Wessely, Urinary free cortisol in chronic fatigue
syndrome. Am. J. Psychiatry 158 (2001), 641–3.
73. A. J. Cleare, The neuroendocrinology of chronic fatigue syndrome. Endocr. Rev. 24 (2003),
236–52.
74. J. W. Kasckow, D. Baker, T. D. Geracioti, Jr, Corticotropin-releasing hormone in depression
and post-traumatic stress disorder. Peptides 22 (2001), 845–51.
75. P. E. Mullen, C. R. Linsell, D. Parker, Influence of sleep disruption and calorie restriction
on biological markers for depression. Lancet 2 (1986), 1051–5.
76. G. Leese, P. Chattington, W. Fraser, et al., Short-term night-shift working mimics the
pituitary-adrenocortical dysfunction in chronic fatigue syndrome. J. Clin. Endocrinol.
Metabol. 81 (1996), 1867–70.
77. A. J. Cleare, The HPA axis and the genesis of chronic fatigue syndrome. Trends Endocrinol.
Metabol. 15 (2004), 55–9.
78. J. Gaab, V. Engert, V. Heitz, et al., Associations between neuroendocrine responses to the
insulin tolerance test and patient characteristics in chronic fatigue syndrome. J. Psychosom.
Res. 56 (2004), 419–24.
79. D. E. Giles, D. J. Kupfer, A. J. Rush, H. P. Roffwarg, Controlled comparison of electrophysio-
logical sleep in families of probands with unipolar depression. Am. J. Psychiatry 155 (1998),
192–9.
80. A. Korszun, V. Moskvina, S. Brewster, et al., Familiality of symptom dimensions in depres-
sion. Arch. Gen. Psychiatry 61 (2004), 468–74.
81. M. Irwin, Psychoneuroimmunology of depression: clinical implications. Brain Behav.
Immun. 16 (2002), 1–16.
82. M. Lyall, M. Peakman, S. Wessely, A systematic review and critical evaluation of the
immunology of chronic fatigue syndrome. J. Psychosom. Res. 55 (2003), 79–90.
83. A. Skowera, A. Cleare, D. Blair, et al., High levels of type 2 cytokine-producing cells in
chronic fatigue syndrome. Clin. Exp. Immunol. 135 (2004), 294–302.
209 Depression and chronic fatigue
Introduction
Recent advances in the detection [1] and treatment [2] of cancer have led to longer
survival times [3]; for example, the five-year survival rate of localised breast cancer is
now 97% [4]. As a result of this lengthened life expectancy, greater attention is being
paid to quality of life and psychosocial care for cancer patients. In the short term,
this means ensuring that patients maintain their quality of life during diagnostic
and treatment phases. In the longer term, the aim is to ensure that psychological
problems are prevented or ameliorated so that cancer patients can rehabilitate and
resume functioning at the level they maintained before their cancer.
Cancer remains a life-threatening illness linked by many authors with fears about
incapacity, disfigurement and death [5,6]. Unsurprisingly, many individuals diag-
nosed with cancer experience at least transient psychological distress. This recog-
nition, however, can lead to a dismissal of depression as a normative response to
cancer and missed opportunities to address a highly impairing, but readily treatable,
psychiatric disorder when depression does occur. On the other hand, overestima-
tion of the extent to which cancer results in depression can misguide allocation of
resources, leading to an emphasis on strategies for aggressively detecting psychi-
atric disorders at the expense of follow-up care for patients who have already been
identified, as well as attention to more common problems and basic supportive
needs that affect quality of life among cancer patients more generally.
In addition to concern about depression as an added disease burden, there is
considerable speculation about depression as a cause of cancer and predictor of
disease progression and survival [7]. If depression can be shown to affect progression
and survival, then ensuring that it is detected and treated effectively becomes all
the more important. Yet, tying the value of treatment of depression to its effect on
disease course and an extension of life can lead to a discrediting of quality of life as
‘borderline’ cases of depression and anxiety, for which the efficacy of intervention
has not been established. Burgess and colleagues note that a year after diagnosis,
the prevalence of ‘clinically significant’ depression and anxiety was only 15%, and
thereafter the risk of anxiety and depression was no greater than in the general
population, suggesting that the depression and anxiety being discussed are largely
self-limiting and not enduring clinical disorders.
There are a number of factors that contribute to the wide variation in reporting of
depression in cancer patients. As mentioned already, levels of depression are likely
to vary according to the length of time that has elapsed since diagnosis of cancer, and
this may be particularly true where subclinical cases are concerned, e.g. [14]. Higher
estimates may also reflect use of self-report questionnaires of ‘depressive symptoms’
rather than diagnostic interviews to ascertain formal clinical diagnoses [15] or the
sampling of in-patients with more severe somatic disturbance or advanced stage of
disease [16,17].
In older adults (age 60–74 years) in the UK, the group most likely to experience
a diagnosis of cancer, the one-month prevalence is 7% for mixed anxiety and
depressive disorder, 3% for general anxiety disorder and 1% for depressive episodes
[18]. Using an index of depressive symptoms and a wider age range (16–74 years),
the one-month prevalence is 10% for all adults in the UK [19]. In the USA, Kessler
et al. [20] found a 16.6% lifetime prevalence of depression and a 6.6% 12-month
prevalence. Some studies of the prevalence of depression among cancer patients
report levels comparable to or lower than these, suggesting that cancer itself may
provide no additional risk for depression. This hypothesis was tested systematically
in a meta-analysis of 58 studies comparing levels of anxiety and depression in
cancer patients with the general population. This showed that levels of depression
were higher in cancer patients than in the general population, but levels of anxiety
and general psychological distress were not [21]. Other reviews also conclude that
depression is more common in cancer patients than among the general population
[8,15,22]. However, differences in sampling and assessment strategies, definitions
of depression and diagnostic criteria, and handling of overlap between somatic
symptoms of disease/treatment and depression make it difficult to draw definitive
conclusions.
Another useful comparison is between levels of depression in cancer patients
and other patient groups. Rates in patients with neurological disorders appear to
be more consistently higher (30–50%) than in those with cancer or other medical
illnesses [9]. Analyses of a longitudinal study of depressive symptoms in a sample of
8387 adults (aged 51–61 years) suggested that cancer patients have more depressive
symptoms than those with hypertension, arthritis, heart disease, diabetes or stroke
in the first two years after diagnosis. Levels of depressive symptoms in cancer
patients were comparable to levels found in patients with chronic lung disease [23].
214 Depression and physical illness
However, the age range for this study is lower than the mean age of diagnosis of most
cancers, and among cancer patients lower age is associated with greater distress.
More generally, estimates of the prevalence of major depressive disorder among
cancer patients based on semi-structured diagnostic interviews with representative
samples overlap with the 5–13% range found in primary-care patients [24–28].
The best estimates of the prevalence of depression in cancer patients suggest that
the disorder is more common than in the general population, but perhaps only as
common as or a little more common than in general medical patients, including
primary-care patients. Even accepting the higher estimates of prevalence in cancer
patients, the majority of individuals with cancer will not develop major depression.
As Raison and Miller’s [22] review concludes, ‘cancer is a risk factor, rather than a
mandate, for depression’ (p. 283). This raises an interesting but overlooked issue.
If cancer does not raise the level of current depression as once thought, then what
effect does it have on lifetime rates of depression? Depression is a recurrent episodic
condition with a mean first onset in the early twenties, long before the mean age
at which most cancers are diagnosed. If most cases of depression during cancer
treatment are recurrences, then this would have important implications for efforts
to detect and treat this depression, and perhaps even prevent it among cancer
patients vulnerable on the basis of past history.
interviewers required to accept patient reports at face value without probing, e.g.
as in [44], produce dubious diagnoses, rendering the conclusions of some research
studies suspect [45].
Disease site
Patients with cancer are not, of course, a homogeneous group in terms of the cancers
they have. For this reason alone, differences exist in the levels of distress experienced
by patients with cancer. Zabora et al. [46] studied 5000 cancer patients, of whom
58% had received a diagnosis within the past 90 days. The highest prevalence of
distress was found in those with lung cancer, where 43.4% of the sample showed
high levels of distress. Similarly high levels of distress were experienced by patients
with brain, liver, pancreatic and head and neck cancers. Significantly lower levels of
distress were experienced in patients with gynaecological, breast, colon and prostate
cancers. This may be because of the better prognosis for the latter cancer sites. Other
studies report similar findings; for example, Stommel et al. [47] found higher levels
of depression in patients with lung cancer compared with those with breast, colon
or prostate cancers in their sample of 860 older (≥ 65 years) patients.
Disease stage
Stage of disease at diagnosis is another factor that affects reporting of psychological
adjustment. Being diagnosed with more advanced disease implies a poorer prog-
nosis and, understandably, can be expected to pose a greater threat to psychological
wellbeing. A number of studies show that more advanced disease is associated with
increased psychological distress, e.g. [48–50], although other studies find little or
no relationship between stage and psychological outcomes [51–53]. Studies that
fail to find an effect often do not include patients with the most advanced disease
stages [14,52], or make a comparison between early- and advanced-stage patients
in widely different timeframes since diagnosis, e.g. [51]. These varied responses
may also reflect the fact that patients are notoriously poor at understanding the
precise implication of their diagnoses [54–57].
Treatments of cancer
Some cancer treatments are also associated with increased levels of depression.
Immunotherapeutic agents, such as interferon alfa and interleukin, used to treat
some cancers (e.g. kidney, melanoma) induce depressed mood as a direct side
effect [58–60]. Other more extreme treatments such as bone-marrow and stem-
cell transplants appear to result in higher levels of distress than cancers with other
treatment modalities indirectly because of the more severe trauma to the body and
217 Cancer and depression
signs of distress such as crying but miss symptoms such as suicidal thoughts and
hopelessness, which would require more involved interviewing or direct enquiry to
elicit [73,74]. Undetected depression in cancer patients represents a missed oppor-
tunity to reduce suffering and impairment. This has led to a growing chorus of calls
for routine screening of cancer patients for psychological distress and depression
[75–77].
positive for depression, despite many of them being long-term survivors report-
ing for a follow-up visit. This figure could be cited as evidence of the enduring
distress associated with breast cancer, but it is actually almost identical to what
would be obtained in a primary-care waiting room [87]. Furthermore, only 8%
of the waiting-room sample was subsequently found to have MDD, and the odds
of an individual screening positively having MDD were only 21%. Many of the
depressed women had a recent or current prescription for an antidepressant, and
so the probability that an interview with a woman who screened positive would
yield an untreated case of MDD was only 7%. In short, two-stage screening and
interviewing in this sample could provide little in terms of improved detection of
untreated depression for a lot of effort.
There is some evidence that the introduction of screening with adequate feedback
to medical staff can increase the number of cases of MDD that are identified [13]
and the number of patients who are referred for psychiatric care [88]. This should,
presumably, result in a reduction in the overall level of depression in the screened
sample [89]. However, there has been no demonstration that routine screening for
depression reduces the rate of the disorder or level of depressive symptoms among
cancer patients on a clinic or population basis. The consistent finding in the general
medical literature is that screening does not improve patient outcomes without the
introduction of considerable resources, and it is exceedingly difficult to identify
any enduring benefits after the enriched resources of a demonstration project are
withdrawn [90]. We know of no demonstration that screening by itself reduces
depression on a population basis, despite an extensive review of the literature [90].
The result of one pilot study with patients with prostate cancer raises some issues
that deserve more attention. Roth and colleagues [91] were able to get 77% (93/121)
of a waiting-room sample to complete questionnaires; 29 scored above a cut-off
point on one or both screening instruments. These 29 patients were referred for
further evaluation, but 12 missed or refused an appointment. One recurrent issue in
dealing with these patients was the difficulty of getting such an evaluation without
requiring a return visit by the patient. Of the 17 evaluated, three patients not
currently receiving treatment were given a psychiatric diagnosis. A re-evaluation
was recommended to them, but there is no report of how many patients completed
this. Consistent with some of the difficulties that Roth and colleagues encountered,
Shimizu and colleagues [88] found that only 28% (19/67) of cancer patients who
scored positive on a screening test for depression accepted a referral to a psychiatrist.
Screening is clearly not a panacea in the improvement of the outcome of depres-
sion on a cancer setting or population basis. We would urge caution before any
commitment to screening as the sole or primary means to address depression in a
cancer-care setting and would urge careful consideration of its likely costs, benefits
and alternatives. Maguire [92] apparently shares our concerns and recommends
220 Depression and physical illness
the comfort zone of oncologists for prescribing antidepressants for some patients’
(p. 1942). There were numerous methodological problems with this study, starting
with difficulties accruing patients despite a US statewide recruitment effort. More-
over, the differences between their intervention and placebo groups were greater at
baseline than at follow-up, and in the follow-up assessment on which they base their
claim of effectiveness data were available for only 14 of the 81 intervention patients
and 19 of the 78 placebo patients. Our major concerns with this study are three-fold
(see also [102,103]): first, there is more generally a lack of evidence that antide-
pressants are effective for individuals who have subthreshold depression, and the
investigators’ encouragement of such use is counter to available evidence. Second,
it is already difficult to get non-psychiatric physicians such as oncologists to adhere
to diagnostic criteria and provide adequate patient education and follow-up; Fisch
and colleagues imply that neither is needed. Finally, from the point of view of pol-
icy, it would be unfortunate if inaccurate claims about the effectiveness of mailing
antidepressants to patients who are not even depressed were used to argue against
the provision of appropriate psychosocial services. Substantial rates of prescription
of antidepressants to cancer patients who are not depressed is already an important
issue in quality and economics of mental healthcare for cancer patients [102].
In summary, the case for the efficacy of intervention for depression among cancer
patients depends mainly on evidence derived from other populations. Demonstra-
tions that such claims can be validly extended to cancer patients are sorely overdue.
However, our earlier review of the prevalence of depression among cancer patients
suggests a challenge in the mounting of methodologically adequate studies: if only
9% of patients with breast cancer suffer from major depression and a substan-
tial proportion are already receiving treatment, then a multisite trial with massive
amounts of screening may be necessary to accrue a sufficient sample.
and ‘usual care’. Sollner and colleagues [106] found that only 42% of patients with
breast cancer who screened positive on a measure of distress endorsed an interest
in counselling, a proportion no greater than for non-distressed patients with breast
cancer. In the study by Roth and colleagues [91], one of the patients with prostate
cancer refusing screening outright stated: ‘This is a psychiatric evaluation? Right
now I have my own problems and I don’t want to get involved.’
The data are mixed as to the extent to whether cancer-care professionals influence
patient uptake of mental health and psychosocial treatment. Although most (68%)
of a sample of patients with prostate cancer reported that they would prefer not
to take medication for depression, 75% indicated that they would do so if advised
by their physician [107]. However, Eakin and Strycker [108] found that although
clinicians reported referring 70% of cancer patients to support services, only 24%
of these patients recalled having discussed the issue.
The problems of low uptake and patient resistance to mental health and psy-
chosocial services is often framed in terms of stigma, with the requisite solution
being a destigmatisation of help-seeking [109]. However, there is consistent evi-
dence that offered a choice, patients with cancer prefer more support and commu-
nication from oncologists and oncological nurses than interactions with mental
health professionals, and interventions designed to increase their access to quality,
understandable information about their condition and its treatment to counselling
and psychotherapy [107,110–113]. Perhaps, not unreasonably, patients prefer to
deal directly with the perceived source of their distress and depression [106]. It
remains an important but unaddressed question as to whether improved access to
care specialists and medical information is a more acceptable and effective means
of preventing and ameliorating depression than formal mental health services, or, if
not, whether such improved access and information might function as an effective
first-stage response, such that patients who do not benefit sufficiently are particu-
larly appropriate for specialty mental health interventions.
Even when depressed cancer patients are interested in empirically supported
treatments, they may have difficulty accessing such services. Cancer care tends to
be provided in tertiary-care settings to which patients travel many miles [10,107].
It may be logistically difficult or undesirable to coordinate mental healthcare with
visits for chemotherapy or radiation and unrealistic to expect that patients will
make regular return visits expressly for care for depression.
Then there is the issue of the availability in cancer-care settings of professionals
trained in the effective treatment of depression. It has been argued that quality
mental health and psychosocial care is an essential component of comprehensive
cancer care, but psycho-oncology units, if present at all in routine cancer care, tend to
be understaffed [76]. Pamphlets for the Memorial Sloane Kettering Cancer Center,
which has been instrumental in the development of psycho-oncology, state flatly
224 Depression and physical illness
that specialty psychosocial services are not available in their affiliated cancer-care
settings. Furthermore, the patient base needed for interventions targeting focused
groups of cancer patients, such as those with early breast cancer or metastatic
breast cancer, has not usually been appreciated. Goodwin et al.’s [114] presentation
of lessons learned in a randomised trial of supportive expressive therapy for patients
with metastatic breast cancer deserves careful consideration.
The unfortunate state of affairs for treatment of depression in routine care for
cancer is that if depression is addressed at all, then it is likely to be with peer- or pro-
fessionally led support groups with patients of varying cancer sites, stages of cancer
and time since diagnosis; non-specific individual counselling and psychotherapy;
or, more predominately, prescription of antidepressants by non-psychiatric physi-
cians and nurses that is both non-specific and followed up poorly. The gap in
outcome is likely to be great between what occurs with these treatments and what
is obtained with empirically supported treatments in the context of a clinical trial.
We have learned that improving the care and outcomes of depression in primary care requires
some or all of the following: a systematic approach to the recognition and assessment of
depression; evidence-based decision support; patient education and activation; ongoing mon-
itoring and feedback regarding patient adherence and outcomes; integration of mental health
specialists for patients who are not improving as expected; and physician education.
Brody [121] notes, however, that serious difficulties have been encountered
in attempting to sustain these innovations beyond well-funded demonstration
projects.
Depression-management specialists, who are usually Master’s level nurses or
social workers who take responsibility for assessment and follow-up, are becom-
ing a basic component of cost-effective strategies for improving the outcome of
depression in general medical care [122,123]. These managers can be trained to
meet patient preferences by offering either medication management or brief struc-
tured psychotherapies. One promising extension of this work is the utilisation of
centralised systems of telephone management of depressed patients, with regular
direct contact of depression-management specialists with patients and support and
feedback to treating professionals [124,125].
The challenges of improving the effectiveness with which depression is man-
aged in cancer care are undoubtedly even more overwhelming than what has been
encountered in primary medical care. Yet, innovations developed in primary care
are now being disseminated into cancer care, at least on a pilot and demonstra-
tion project basis. Strong et al. [126] delivered a nurse-led intervention to cancer
patients with a diagnosis of major depression in which a cancer nurse was trained
to deliver a problem-solving therapy and encourage the patient to consult with
a primary-care physician concerning antidepressants. Patients receiving the inter-
vention showed significant reductions in depression compared with the control
group. Strong et al. [126] noted a number of difficulties associated with patient
rejection of participation in the programme and primary-care physicians rejecting
the advice of the depression specialists, but this pilot project nonetheless shows the
promise of depression-management specialists in the context of cancer care. In a
study in which the depression-management specialist had responsibility for pro-
moting adherence to antidepressants, Dwight-Johnson et al. [127] randomised 55
low-income Hispanic American patients from a public-sector hospital with breast
or cervical cancer and comorbid depression to either collaborative care delivered by
a depression-management specialist or usual care. The usual-care arm of the study
226 Depression and physical illness
involved informing the patient of her diagnosis of depression and encouraging her
to seek treatment. The patients receiving the intervention experienced significantly
greater reduction in depressive symptoms, which the investigators attributed to
both the effectiveness of the intervention and the difficulties the control women
experienced in obtaining care for their depression.
early-stage breast cancer patients. This group was followed for 15 years and the effect
persisted over that time. However, the sample size was relatively small (n = 69).
Watson et al.’s ongoing study of survival in breast cancer patients (n = 578) also
shows that helplessness/hopelessness is related to decreased survival at both five
[137] and ten [138] years’ follow-up. Watson et al.’s study fails to replicate the
findings for fighting spirit found in the study by Greer et al. [136]. Watson et al.
reported a very limited effect for depression at the five-year follow-up, but this was
no longer present at the ten-year follow-up [137,138].
With regard to the possibility that psychological interventions can improve sur-
vival, Spiegel’s group [7,139,140] have argued that there is evidence to support this
possibility. They cite [7,140] five studies that support this hypothesis [139,141–144]
and five studies that do not [133,145–148]. Since Spiegel and colleagues’ reviews,
an additional negative finding has been obtained [51]. Palmer and Coyne [149]
note that three of the studies cited by Spiegel and colleagues ([142–144]) involved
a confounding of improved medical surveillance and more intensive medical care
with psychological intervention, so that any improvement in survival cannot be
attributed unambiguously to psychological intervention. The remaining two stud-
ies ([139,141]) have serious statistical problems. Survival data for cancer patients
is highly skewed, with a small number of patients typically living much longer than
the rest, making medians better summary statistics than means. When the Spiegel
et al. [139] data are summarised in terms of median survival time, the advantage for
the patients receiving the intervention is non-significant, being only a few months.
Fawzy et al. [141] dropped a number of patients from their analyses, including one
patient because he was clinically depressed and so did not provide the intention-
to-treat analyses that are standard in evaluating medical interventions. Not only
do the available studies not support an effect of psychological intervention on sur-
vival, but also their relevance to the question of a role for depression in cancer
progression is unclear. Spiegel and Giese-Davis’s [7] question ‘What are the aspects
of psychosocial intervention that seem to be most effective in reducing depression
. . . ?’ (p. 275) seemed quite unanswerable. As noted, Fawzy et al. [141] dropped
the one depressed patient in their study; other studies do not target depression or
assess it as an outcome, and in general these studies have recruited patients that are
not notably depressed, even when suffering metastatic disease.
In summary, the evidence that depression causes cancer or accelerates its pro-
gression or that intervention for depression improves survival is largely absent or
negative. This stands in sharp contrast to the robust prognostic value for depres-
sion in cardiovascular disease [150]. In the absence of credible evidence, continued
claims for a role for depression in cancer and claims that psychological interventions
promote survival can prove damaging to the field. Lesperance and Frasure-Smith
[151] stated: ‘Although the prevention of death is a powerful tool to influence many
228 Depression and physical illness
of our medical colleagues who use it to justify the allocation of resources, death is
not everything’ (p. 19). When strong claims about the role of depression in cancer
ultimately need to be abandoned, it will seem to be an undignified retreat to claim
that the usefulness of detecting and treating depression based on the ‘mere’ benefits
for wellbeing or quality of life. An unwarranted strong claim thus could rob the
credibility of what had always been a reasonable claim.
Conclusions
Depression is unlikely to cause or accelerate cancer progression. This is in contrast
to the apparent role of depression in other diseases, such as cardiac morbidity and
mortality discussed in Chapters 3 and 4 of this book. The prevalence of depression
in cancer patients is relatively low but is an additional, and often unrecognised,
problem for some patients. Difficulties assessing prevalence and identifying cases of
depression in cancer patients have hampered efforts to provide appropriate clinical
services. There are a number of effective treatments for depression, but there is
often limited availability of trained staff to deliver them, and many cancer patients
continue to reject interventions for psychological problems. Depression could be
reduced in cancer patients by providing more effective symptom-management or
behavioural interventions. These forms of treatment could be more acceptable to
patients and delivered more easily by existing oncology staff, but data concerning
the potential of these methods are still limited.
Acknowledgements
Preparation of this chapter was supported in part by National Cancer Center (NCI)
grant 5R01MH063172 and by Cancer Research, UK.
REFERENCES
22. C. L. Raison, A. H. Miller, Depression in cancer: new developments regarding diagnosis and
treatment. Biol. Psychiatry 54 (2003), 283–94.
23. D. Polsky, J. A. Doshi, S. Marcus, et al., Long-term risk for depressive symptoms after a
medical diagnosis. Arch. Intern. Med. 165 (2005), 1260–66.
24. J. E. Barrett, J. A. Barrett, T. E. Oxman, P. D. Gerber, The prevalence of psychiatric disorders
in a primary care practice. Arch. Gen. Psychiatry 45 (1988), 1100–106.
25. J. C. Coyne. Self-reported distress: analog or Ersatz depression? Psychol. Bull. 116 (1994),
29–45.
26. L. G. Kessler, P. D. Cleary, J. D. Burke, Jr, Psychiatric disorders in primary care. Results of a
follow-up study. Arch. Gen. Psychiatry 42 (1985), 583–7.
27. H. C. Schulberg, M. McClelland, B. J. Burns, Depression and physical illness: the prevalence,
causation and diagnosis of co-morbidity. Clin. Psychol. Rev. 7 (1987), 145–67.
28. W. Katon, The epidemiology of depression in medical care. Int. J. Psychiatry Med. 17 (1987),
93–112.
29. K. Reuter, M. Harter, The concepts of fatigue and depression in cancer. Eur. J. Cancer Care
13 (2004), 127–34.
30. J. C. Coyne, S. C. Palmer, P. A. Sullivan, Screening for depression in adults. Ann. Intern.
Med. 138 (2003), 767–8.
31. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders,
4th edn (Washington, DC: American Psychiatric Association, 1994).
32. W. Katon, Depression: somatic symptoms and medical disorders in primary care. Compr.
Psychiatry 23 (1982), 274–87.
33. S. Cavanaugh, D. C. Clark, R. D. Gibbons, Diagnosing depression in the hospitalized med-
ically ill. Psychosomatics 24 (1983), 809–15.
34. D. A. Clark, A. Cook, D. Snow, Depressive symptom differences in hospitalized, medically ill,
depressed psychiatric inpatients and nonmedical controls. J. Abnorm. Psychol. 107 (1998),
38–48.
35. D. C. Clark, C. S. vonAmmon, R. D. Gibbons, The core symptoms of depression in medical
and psychiatric patients. J. Nerv. Ment. Dis. 171 (1983), 705–13.
36. S. A. Cohen-Cole, A. Stoudemire, Major depression and physical illness. Special consider-
ations in diagnosis and biologic treatment. Psychiatr. Clin. North Am. 10 (1987), 1–17.
37. M. K. Popkin, A. L. Callies, E. A. Colon, A framework for the study of medical depression.
Psychosomatics 28 (1987), 27–33.
38. P. C. Trask, Assessment of depression in cancer patients. J. Natl. Cancer Inst. Monogr. (2004),
80–92.
39. R. G. Kathol, A. Mutgi, J. Williams, G. Clamon, R. Noyes, Jr, Diagnosis of major depression
in cancer patients according to four sets of criteria. Am. J. Psychiatry 147 (1990), 1021–4.
40. J. Endicott, Measurement of depression in patients with cancer. Cancer 53 (1984), 2243–9.
41. H. G. Koenig, L. K. George, B. L. Peterson, C. F. Pieper, Depression in medically ill hospi-
talized older adults: prevalence, characteristics, and course of symptoms according to six
diagnostic schemes. Am. J. Psychiatry 154 (1997), 1376–83.
42. J. C. Coyne, C. M. Pepper, H. Flynn, Significance of prior episodes of depression in two
patient populations. J. Consult. Clin. Psychol. 67 (1999), 76–81.
231 Cancer and depression
79. L. S. Radloff, The CES-D scale: a self-report depression scale for research in the general
population. Appl. Psychol. Meas. 1 (1977), 385–401.
80. L. R. Derogatis, N. Melisaratos, The Brief Symptom Inventory: an introductory report.
Psychol. Med. 13 (1983), 595–605.
81. A. Zigmond, R. Snaith, The Hospital Anxiety and Depression Scale. Acta Psychiatr. Scand.
67 (1983), 361–70.
82. B. Lowe, K. Kroenke, K. Grafe, Detecting and monitoring depression with a two-item
questionnaire (PHQ-2). J. Psychosom. Res. 58 (2005), 163–71.
83. B. Lowe, J. Unutzer, C. M. Callahan, A. J. Perkins, K. Kroenke, Monitoring depression
treatment outcomes with the patient health questionnaire-9. Med. Care 42 (2004), 1194–
201.
84. V. Henkel, R. Mergl, J. C. Coyne, et al., Screening for depression in primary care: will one
or two items suffice? Eur. Arch. Psychiatry Clin. Neurosci. 254 (2004), 215–23.
85. R. L. Spitzer, J. B. Williams, K. Kroenke, et al., Utility of a new procedure for diagnosing
mental disorders in primary care: the PRIME-MD 1000 study. J. Am. Med. Assoc. 272 (1994),
1749–56.
86. A. C. Leon, J. E. Kelsey, A. Pleil, et al., An evaluation of a computer assisted telephone
interview for screening for mental disorders among primary care patients. J. Nerv. Ment.
Dis. 187 (1999), 308–11.
87. S. Fechner-Bates, J. C. Coyne, T. L. Schwenk, The relationship of self-reported distress
to depressive disorders and other psychopathology. J. Consult. Clin. Psychol. 62 (1994),
550–9.
88. K. Shimizu, T. Akechi, M. Okamura, et al., Usefulness of the nurse-assisted screening and
psychiatric referral program. Cancer 103 (2005), 1949–56.
89. S. A. McLachlan, A. Allenby, J. Matthews, et al., Randomized trial of coordinated psychoso-
cial interventions based on patient self-assessments versus standard care to improve the
psychosocial functioning of patients with cancer. J. Clin. Oncol. 19 (2001), 4117–25.
90. S. C. Palmer, J. C. Coyne, Screening for depression in medical care: pitfalls, alternatives, and
revised priorities. J. Psychosom. Res. 54 (2003), 279–87.
91. A. J. Roth, A. B. Kornblith, L. Batel-Copel, et al., Rapid screening for psychologic distress
in men with prostate carcinoma: a pilot study. Cancer 82 (1998), 1904–8.
92. P. Maguire, Improving communication with cancer patients. Eur. J. Cancer 35 (1999),
1415–22.
93. C. Dean, Psychiatric morbidity following mastectomy: preoperative predictors and types of
illness. J. Psychosom. Res. 31 (1987), 385–92.
94. J. Harrison, P. Maguire, Predictors of psychiatric morbidity in cancer patients. Br. J. Psychi-
atry 165 (1994), 593–8.
95. J. S. McDaniel, D. L. Musselman, C. B. Nemerof, Cancer and depression: Theory and treat-
ment. Psychiatr. Ann. 27 (1997), 360–64.
96. D. L. Evans, D. S. Charney, L. Lewis, et al., Mood disorders in the medically ill: scientific
review and recommendations. Biol. Psychiaty 58 (2005), 175–89.
97. D. Razavi, J. F. Allilaire, M. Smith, et al., The effect of fluoxetine on anxiety and depression
symptoms in cancer patients. Acta Psychiatr. Scand. 94 (1996), 205–10.
234 Depression and physical illness
98. A. K. Koike, J. Unutzer, K. B. Wells, Improving the care for depression in patients with
comorbid medical illness. Am. J. Psychiatry 159 (2002), 1738–45.
99. T. Sheard, P. Maguire, The effect of psychological interventions on anxiety and depression
in cancer patients results of two meta-analyses. Br. J. Cancer 80 (1999), 1770–80.
100. A. M. Nezu, C. M. Nezu, S. H. Felgoise, K. S. McClure, P. S. Houts, Project Genesis: assessing
the efficacy of problem-solving therapy for distressed adult cancer patients. J. Consult. Clin.
Psychol. 71 (2003), 1036–48.
101. M. J. Fisch, P. J. Loehrer, J. Kristeller, et al., Fluoxetine versus placebo in advanced cancer
outpatients: a double-blinded trial of the Hoosier Oncology Group. J. Clin. Oncol. 21 (2003),
1937–43.
102. J. C. Coyne, S. C. Palmer, P. J. Shapiro, Prescribing antidepressants to advanced cancer
patients with mild depressive symptoms is not justified. J. Clin. Oncol. 22 (2004), 205–6.
103. B. A. Erwin, P. A. Sullivan, T. R. ten Have, Trial of antidepressants for mildly depressed
cancer patients should have been reported in a manner allowing independent evaluation of
investigators’ claims. J. Clin. Oncol. 22 (2004), 753–4.
104. C. Moynihan, J. M. Bliss, J. Davidson, L. Burchell, A. Horwich, Evaluation of adjuvant
psychological therapy in patients with testicular cancer: randomised controlled trial. Br.
Med. J. 316 (1998), 429–35.
105. M. Sharpe, V. Strong, K. Allen, et al., Management of major depression in outpatients
attending a cancer centre: a preliminary evaluation of a multicomponent cancer nurse-
delivered intervention. Br. J. Cancer 90 (2004), 310–13.
106. W. Sollner, S. Maislinger, A. Konig, A. Devries, P. Lukas, Providing psychosocial support for
breast cancer patients based on screening for distress within a consultation-liaison service.
Psychooncology. 13 (2004), 893–7.
107. P. J. Shapiro, J. C. Coyne, L. K. Kruus, et al., Interest in services among prostate cancer
patients receiving androgen deprivation therapy. Psychooncology 13 (2004), 512–25.
108. E. G. Eakin, L. A. Strycker, Awareness and barriers to use of cancer support and information
resources by HMO patients with breast, prostate, or colon cancer: patient and provider
perspectives. Psychooncology 10 (2001), 103–13.
109. J. C. Holland, History of psycho-oncology overcoming attitudinal and conceptual barriers.
Psychosom. Med. 64 (2002), 206–21.
110. M. Beutel, G. Henrich, A. Sellschopp, M. Keller, W. Adermayer, Needs and utilization of
ambulatory psychosocial services by cancer patients – exemplified by the oncologic day
clinic. Psychother. Psychosom. Med. Psychol. 46 (1996), 304–11.
111. R. E. Gray, M. Fitch, C. Davis, C. Phillips, Interviews with men with prostate cancer about
their self-help group experience. J. Palliat. Care 13 (1997), 15–21.
112. K. N. Moore, A. Estey, The early post-operative concerns of men after radical prostatectomy.
J. Adv. Nurs. 29 (1999), 1121–9.
113. F. Wong, D. E. Stewart, J. Dancey, et al., Men with prostate cancer: influence of psychological
factors on informational needs and decision making. J. Psychosom. Res. 49 (2000), 13–9.
114. P. J. Goodwin, M. Leszcz, G. Quirt, et al., Lessons learned from enrollment in the BEST
study: a multicenter, randomized trial of group psychosocial support in metastatic breast
cancer. J. Clin. Epidemiol. 53 (2000), 47–55.
235 Cancer and depression
115. D. A. Regier, W. E. Narrow, D. S. Rae, et al., The de facto US mental and addictive disor-
ders service system: Epidemiologic Catchment Area prospective 1-year prevalence rates of
disorders and services. Arch. Gen. Psychiatry 50 (1993), 85–94.
116. P. A. Sullivan, J. C. Coyne, S. C. Palmer, J. M. Metz, Antidepressant use in breast cancer
patients: an Oncolink study. Presented at the 27th annual meeting of the American Society
of Preventive Oncology, Philadelphia, PA, 3 March 2003.
117. W. J. Katon, J. Unutzer, G. Simon, Treatment of depression in primary care: where we are,
where we can go. Med. Care 42 (2004), 1153–7.
118. J. W. Williams, Jr, C. D. Mulrow, E. Chiquette, et al., A systematic review of newer phar-
macotherapies for depression in adults: evidence report summary. Ann. Intern. Med. 132
(2000), 743–56.
119. W. Katon, K. M. Von, E. Lin, et al., Collaborative management to achieve treatment
guidelines: impact on depression in primary care. J. Am. Med. Assoc. 273 (1995),
1026–31.
120. J. C. Coyne, S. C. Palmer, P. A. Sullivan, Is case-finding an inefficient way of address-
ing depression as a public health problem? Pharmacoepidemiol. Drug Saf. 11 (2002),
545–7.
121. D. S. Brody, Improving the management of depression in primary care: recent accomplish-
ments and ongoing challenges. Dis. Manage Health Outcomes 11 (2003), 21–31.
122. T. E. Oxman, A. J. Dietrich, H. C. Schulberg, The depression care manager and mental
health specialist as collaborators within primary care. Am. J. Geriatr. Psychiatry 11 (2003),
507–16.
123. H. C. Schulberg, C. Bryce, K. Chism, et al., Managing late-life depression in primary care
practice: a case study of the health specialist’s role. Int. J. Geriatr. Psychiatry 16 (2001),
577–84.
124. E. M. Hunkeler, J. F. Meresman, W. A. Hargreaves, et al., Efficacy of nurse telehealth care
and peer support in augmenting treatment of depression in primary care. Arch. Fam. Med.
9 (2000), 700–708.
125. G. E. Simon, M. VonKorff, C. Rutter, E. Wagner, Randomised trial of monitoring, feedback,
and management of care by telephone to improve treatment of depression in primary care.
Br. Med. J. 320 (2000), 550–54.
126. V. Strong, M. Sharpe, A. Cull, et al., Can oncology nurses treat depression? A pilot project.
J. Adv. Nurs. 46 (2004), 542–8.
127. M. Dwight-Johnson, K. Ell, P. J. Lee, Can collaborative care address the needs of low-
income Latinas with comorbid depression and cancer? Results from a randomized pilot
study. Psychosomatics 46 (2005), 224–32.
128. J. Lemon, S. Edelman, Perceptions of the ‘mind-cancer’ relationship among the public,
cancer patients, and oncologists. J. Psychosoc. Oncol. 21 (2004), 43–58.
129. B. D. Doan, R. E. Gray, C. S. Davis, Belief in psychological effects on cancer. Psychooncology
2 (1993), 139–50.
130. B. L. Andersen, W. B. Farrar, D. M. Golden-Kreutz, et al., Psychological, behavioral, and
immune changes after a psychological intervention: a clinical trial. J. Clin. Oncol. 22 (2004),
3570–80.
236 Depression and physical illness
131. H. Faller, M. Schmidt, Prognostic value of depressive coping and depression in survival of
lung cancer patients. Psychooncology 13 (2004), 359–63.
132. F. Efficace, L. Biganzoli, M. Piccart, et al., Baseline health-related quality-of-life data as
prognostic factors in a phase III multicentre study of women with metastatic breast cancer.
Eur. J. Cancer 40 (2004), 1021–30.
133. P. J. Goodwin, M. Leszcz, M. Ennis, et al., The effect of group psychosocial support on
survival in metastatic breast cancer. N. Engl. J. Med. 345 (2001), 1719–26.
134. F. Efficace, P. Therasse, M. J. Piccart, et al., Health-related quality of life parameters as
prognostic factors in a nonmetastatic breast cancer population an international multicenter
study. J. Clin. Oncol. 22 (2004), 3381–8.
135. B. Garssen, Psychological factors and cancer development: evidence after 30 years of research.
Clin. Psychol. Rev. 24 (2004), 315–38.
136. S. Greer, T. Morris, K. W. Pettingale, Psychological response to breast cancer: effect on
outcome. Lancet 2 (1979), 785–7.
137. M. Watson, J. S. Haviland, S. Greer, J. Davidson, J. M. Bliss, Influence of psychological
response on survival in breast cancer: a population-based cohort study. Lancet 354 (1999),
1331–6.
138. M. Watson, J. Homewood, J. Haviland, J. M. Bliss, Influence of psychological response on
breast cancer survival: 10-year follow-up of a population-based cohort. Eur. J. Cancer 41
(2005), 1665–6.
139. D. Spiegel, J. R. Bloom, H. C. Kraemer, E. Gottheil, Effect of psychosocial treatment on
survival of patients with metastatic breast cancer. Lancet 2 (1989), 888–91.
140. D. Spiegel, Effects of psychotherapy on cancer survival. Nat. Rev. Cancer 2 (2002), 383–9.
141. F. I. Fawzy, N. W. Fawzy, C. S. Hyun, et al., Malignant melanoma: effects of an early structured
psychiatric intervention, coping, and affective state on recurrence and survival 6 years later.
Arch. Gen. Psychiatry 50 (1993), 681–9.
142. J. L. Richardson, D. R. Shelton, M. Krailo, A. M. Levine, The effect of compliance with
treatment on survival among patients with hematologic malignancies. J. Clin. Oncol. 8
(1990), 356–64.
143. T. Kuchler, D. Henne-Bruns, S. Rappat, et al., Impact of psychotherapeutic support on
gastrointestinal cancer patients undergoing surgery: survival results of a trial. Hepatogas-
troenterology 46 (1999), 322–35.
144. R. McCorkle, N. E. Strumpf, I. F. Nuamah, et al., A specialized home care intervention
improves survival among older post-surgical cancer patients. J. Am. Geriatr. Soc. 48 (2000),
1707–13.
145. M. W. Linn, B. S. Linn, R. Harris, Effects of counseling for late stage cancer patients. Cancer
49 (1982), 1048–55.
146. A. Ilnyckyj, J. Farber, M. Cheang, B. Weinerman, A randomized controlled trial of psy-
chotherapeutic intervention in cancer patients. Ann. R. Coll. Physicians Surg. Can. 27 (1994),
93–6.
147. A. J. Cunningham, C. V. Edmonds, G. P. Jenkins, et al., A randomized controlled trial of the
effects of group psychological therapy on survival in women with metastatic breast cancer.
Psychooncology 7 (1998), 508–17.
237 Cancer and depression
148. S. Edelman, J. Lemon, D. R. Bell, A. D. Kidman, Effects of group CBT on the survival time
of patients with metastatic breast cancer. Psychooncology 8 (1999), 474–81.
149. S. C. Palmer, J. C. Coyne, Examining the evidence that psychotherapy improves the survival
of cancer patients. Biol. Psychiatry 56 (2004), 61–2.
150. N. Frasure-Smith, F. Lesperance, Reflections on depression as a cardiac risk factor. Psycho-
som. Med. 67: Suppl 1 (2005), 19–25.
151. F. Lesperance, N. Frasure-Smith, The seduction of death. Psychosom. Med. 61 (1999), 18–20.
11
11
Depression and obesity
Lucy Cooke and Jane Wardle
Introduction
It is estimated that at least 300 million people worldwide are obese and two to three
times more are overweight [1]. Rates vary enormously from country to country,
but the situation in the USA is particularly alarming: current figures suggest that
over 65% of US adults are overweight, of whom over 30% are obese (body mass
index [BMI] ≥ 30) and almost 5% are severely obese (BMI ≥ 40) [2]. Although
associations between obesity and physical health consequences such as diabetes,
heart disease and cancer, as well as all-cause mortality, are well documented [3–6],
the psychological impact of obesity remains poorly characterised.
In the heyday of psychosomatics, obesity was believed to be a consequence of
either misinterpreting emotional arousal as hunger or using food as a form of
self-medication in order to cope with distress [7]. Obese people were assumed to
have extensive psychopathology, and the treatment of choice was psychotherapy.
However, when larger-scale epidemiological studies were carried out, it became
clear that there were no systematic differences in either personality or rates of
psychiatric illness between obese and normal-weight adults [8]. With time, the
psychosomatic theory lost ground as a basis for management of obesity, but interest
in the link with depression has lived on. Recent years have seen evidence that stress is
associated with adiposity [9,10], and both cortisol and leptin have been implicated
in linking the hypothalamic-pituitary-adrenocortical (HPA) axis and adipose tissue
[11,12]. Along with new ideas on the role of ‘comfort foods’ as moderators of HPA
function [13], interest has revived in the idea that distress could contribute to
overeating and overweight.
The major emphasis in contemporary research is on the emotional consequences
of obesity, reflecting the range of difficulties faced by obese people living in Western
countries, where fatness is widely regarded as abhorrent. Negative attitudes towards
Carpenter n = 40 086 BMI (self-report) AUDADIS (interview) No significant effect on Significant gender by weight No significant interaction with race
et al. [26] Mean age 44 years Obese = BMI ≥ 30 depression overall status interaction – lower
(African American and risk of depression in obese
white) men and higher risk in
obese women
Dong et al. n = 482 families BMI (measured) Structured Significant association Not tested Effect still significant after controlling
[29] (n = 2103 individuals) Obese = BMI ≥ 30 questionnaire on between obesity and for chronic disease and other
(African and European depression, anxiety and depression potential confounders, although
Americans) suicide (interview) measure of chronic illness was
somewhat basic
Stronger effect in BMI > 40
Jorm et al. n = 2280 age 20–24 BMI (self-report) Goldberg Depression Significant association Interaction not tested After controlling for confounders the
[30] years Obese = BMI ≥ 30 Scale (Questionnaire) between obesity and Effects seen in men and effect was reversed – underweight
n = 1334 age 40–44 depression women women were most depressed;
years overweight and obese women were
n = 2305 age 60–64 least depressed
years (Australian)
Johnston n = 2482 adults BMI (measured) CES-D (questionnaire) More obese participants Not tested Tables giving CES-D mean scores by
et al. [27] (Canadian) Obese = BMI ≥ 30 met the clinical cut-off age, gender and weight status show a
point for depression somewhat variable effect
Kaplan et al. n = 12 823 BMI (self-report) Psychological Distress Significant association Psychological distress and Unadjusted analyses would probably
[33] Over 65 years old Obese = BMI ≥ 30 (6 items) (telephone between psychological obesity significantly have shown stronger effects but are
(Canadian) interview) distress and obesity associated in women; in not presented here
men the same association
was not quite significant
Onyike et al. N = 84 1015 39 years BMI (measured) DIS (interview) Significant association with No evidence of interaction Adjusted odds ratios for obesity
[28] old (North American) Obese = BMI ≥ 30 past month depression between gender and obesity (BMI > 30) were rendered
Significant effect in women; non-significant, but effect in severely
same effect size was not obese people (BMI > 40) remained
quite significant in men strong and significant
∗Roberts N = 2298 BMI (self-report) DSM-12D Significant association Not tested in cross-sectional Effect still significant in adjusted
et al. [45] Alameda County Study Obese = > 85th (questionnaire) between obesity and analyses analyses
participants percentile depression Using BMI > 30 as definition of
Over 50 years old obese, effect was significant in
unadjusted but non-significant after
adjusting for confounders
∗ This study included a one-year follow-up. AUDADIS, Alcohol Use Disorders and Associated Disabilities Schedule; BMI, body mass index; CES-D, Center for Epidemiologic Studies Depression
Scale; DIS, Diagnostic Interview Schedule; DSM-12D, Diagnostic and Statistical Manual 12-Item Depression Scale.
242 Depression and physical illness
that were stronger than associations with BMI and that remained significant after
adjusting for BMI. Cross-sectional analyses of data from the Alameda County
Study revealed that obese participants (BMI ≥ 30) aged 50 years or older were
more than twice as likely to be depressed than their non-obese counterparts [32].
Kaplan et al. [33] examined data from 12 823 Canadian community dwellers aged
65 and over. Obesity was related to high psychological distress overall, although in
men the association did not reach significance. The presence of comorbid physical
conditions, as well as a number of other potential confounds, was controlled for
in these analyses, and it is likely therefore that unadjusted odds ratios would have
been considerably higher. In contrast, a study of 56 167 elderly Chinese men and
women enrolled as members of ‘elderly health centres’ reported data showing that
obese individuals of both sexes were approximately 20% less likely to suffer from
depressive symptoms than those of normal weight [34]. The authors explain that
in Chinese culture, there is an association between happiness or good fortune and
body fat, and this may be protective against depressive symptoms.
Taken together, these studies suggest that at least in Western countries, obesity,
particularly severe obesity, is associated with depression, especially among women.
Overall, the size of the association is modest. In Onyike and colleagues’ [28] data,
for example, past-month depression rates were increased from 2.8% to 5.1%. This
is important in public health terms but highlights the fact that the majority of obese
people in community samples are not depressed. The few studies that have examined
physical health consequences have raised the possibility that poor physical health
might play an important role in explaining why some obese people are depressed,
but results have been mixed [30,33] and more research is needed.
Baseline obesity
associated with Baseline depression
Measure of depression at associated with obesity at
Ref. Sample details Measure of obesity depression follow-up? follow-up? Gender differences Controlled for Comments
Bardone n = 459 girls BMI (measured) DISC-C Not tested Adolescent depression did Not tested SES, age at menarche,
et al. [38] Age 15 years at (interview) not predict BMI at absence of father figure,
baseline (New follow-up at age 21 years parental smoking,
Zealanders) childhood health, maternal
health and maternal BMI
Barefoot n = 3560 (mostly BMI (measured at MMPI (OBD Not tested No main effect of Gender by depression Gender, baseline BMI, Lack of gender effects
et al. [42] men) baseline, self-report subscale) at depression, but those with interaction was not gender by BMI interaction, might have been due
<25 years old at at follow-up) baseline; NEOPI high baseline depression significant exercise, smoking to small number of
baseline (depression gained less weight than women in this
(North American) subscale) at non-depressed if they were sample
follow-up initially lean but more if
(questionnaires) they were initially heavy;
trend strongest in those who
were more depressed at
baseline and follow-up
Hasler n = 591 BMI (self-report) SPIKE (interview) Being overweight Not tested No Substance use, levels of Association not
et al. [43] Age 19–40 years Overweight = and increased physical activity, stronger if higher
(Swiss) BMI ≥ 25 weight gain demographic variables, thresholds for
associated with family history of weight defining overweight
atypical depression problems (BMI > 27.5 or
in men and women BMI > 30.0) used
but not with major,
recurrent brief, or
minor depression
Pine et al. n = 644 BMI (self-report) DIS-C (interview) Not tested. Obesity in young adulthood Yes: obesity associated with Age, gender, ethnicity,
[39] Aged 9–18 years at Obese = BMI > 80th related to adolescent adult depression in both parental social class,
baseline percentile depression in the whole genders, but negative parental sociopathy, physical
(North American) sample association in males and health status, childhood IQ,
positive in females smoking and alcohol
Pine et al. n = 90 with major BMI (measured at Schedule for Not tested Depression in childhood No gender by depression Age, gender social class, Associations
[40] depression baseline, self-report Affective Disorders associated positively with interaction ethnicity, physical health persisted after
n = 87 with no at follow-up) (interview) BMI in adulthood; duration status, number of controlling for
psychiatric disorder of depression also emerged pregnancies, smoking, potential
Aged 6–17 years at as predictor of adult BMI alcohol/medication use confounding
baseline (USA) variables
Richardson n = 1037 BMI (measured) DIS-C (interview) Not tested Adjusted for baseline BMI, Yes: girls only; no Childhood BMI, gender, Dose–response
et al. [41] Aged 11 years at Obese = BMI ≥ 30 depressed late adolescent relationship among boys, parental obesity, SES, relationship
baseline (New girls were more than twice nor in either gender for maternal depression observed with
Zealanders) as likely to be obese at depression in early number of episodes
26 years than non-depressed adolescence of depression (N.B.
same cohort as
Bardone et al. [38])
Roberts Participants in the BMI (self-report) DSM-12D Greater relative risk Relationship did not operate No significant Age, gender, education,
et al. [32, Alameda County Obese = BMI ≥ 30 (questionnaire) for depression at in reverse direction – obesity–gender interaction marital status, mental health
44] Study (2002 and 2003) 5-year follow-up depression at baseline did problems at baseline, social
n = 1739 (2002) among those who not predict obesity at isolation, social support,
n = 2123 (2003) were obese at follow-up (2003) chronic medical problems,
Mean age 63 years baseline functional impairment, life
(waves of data events, financial strain,
1994–95 and 1999) physical activity
BMI, body mass index; DIS-C, Diagnostic Interview Schedule for Children; DISC-C, Diagnostic Interview Schedule for Children; DSM-12D, Diagnostic and Statistical Manual 12-Item Depression Scale; MMPI, Minnesota
Multiphasic Personality Inventory; NEOPI, NEO Personality Inventory; OBD, Obvious Depression Scale; SES, socioeconomic status; SPIKE, Structured Psychopathological Interview and Rating of the Social Consequences
for Epidemiology.
246 Depression and physical illness
The possibility that depression in early life might have even more adverse conse-
quences on future weight trajectories has been addressed in several studies looking
at adolescent depression as a predictor of adult obesity, with mixed results. The
Dunedin Multidisciplinary Health and Development study failed to find a greater
risk of early adult obesity among depressed adolescents in New Zealand [38]. Sim-
ilarly, a study of over 700 North American adolescents found that an association
between depression in childhood and adult BMI was rendered non-significant when
conduct disorder symptoms were controlled for [39]. However, two more recent
studies have reported positive results. In the first of these, a small sample (n = 90)
of children aged between 6 and 17 years with major depression were followed up
after 10–15 years [40]. At baseline, depressed children did not differ in BMI from
healthy controls, but at follow-up they were on average almost two BMI points
above the healthy controls (26.1 vs. 24.2), a difference that appeared not to be
explained by a number of confounding factors, including medication history. In
addition, duration of depression emerged as a predictor of adult BMI.
The second study assessed whether major depression in early or late-adolescence
increased risk of obesity at 26 years of age in the Dunedin cohort [41]. After adjusting
for baseline BMI, depressed late-adolescent girls were more than twice as likely to
be obese as adults compared with their non-depressed peers, with a dose–response
relationship with number of episodes of depression, although there was no rela-
tionship in boys, or in either sex in relation to depression in early adolescence.
One possibility, rarely explored in the epidemiological literature, is that depres-
sion may exacerbate pre-existing weight-change tendencies. In their sample of col-
lege students, Barefoot and colleagues [42] found no main effect of depression but
found a significant interaction between depression status and baseline BMI, such
that individuals with high baseline depression scores gained less weight than their
non-depressed peers if they were lean at baseline, but gained more weight if they
were overweight at baseline.
The alternative direction of causation is that obesity is a risk factor for devel-
opment of depression. In a community sample of young adults, Hasler et al. [43]
found that being consistently overweight in early adulthood was associated with
atypical but not typical depression, but the fact that overeating is part of the clinical
diagnosis of atypical depression suggests that the mechanism is unlikely to be a true
effect of weight on depression. Again using data from the Alameda County Study,
Roberts and colleagues [44,45] found that obese older adults (> 50 years of age at
baseline) were at increased risk of subsequent depression. In an attempt to clarify
the temporal relation between the two factors, the researchers looked at two waves
of data collected five years apart [32]. Individuals who were obese at baseline were
at increased risk of depression five years later, but no evidence was found to support
the reverse hypothesis.
247 Depression and obesity
Obese treatment
seekers more Obese treatment
depressed than seekers more
Measure of Measure of non-obese controls depressed than
Ref. Sample details obesity depression Main findings Other results or population norms obese controls
Britz n = 47 obese BMI M-CIDI Clinical group had In most, psychiatric Yes Yes
et al. [53] in-patients (interview) significantly higher rates disorders emerged after
n = 47 obese controls of all disorders; no onset of obesity;
n = 1608 controls differences between obese population-based
Aged 15–21 years controls and population obese had significantly
(German) controls lower BMI than clinical
group
Elfhag n = 120 obese patients BMI BDI Depression not related to Yes Not tested
et al. [48] (Swedish) (questionnaire) degree of obesity, but high
rates overall
Higgs n = 20 obese surgery BMI SCL-90-R Surgery candidates Surgery candidates had Not tested Not tested
et al. [58] candidates Obese = (questionnaire significantly more higher BMI, but this
Mean age 39.95 years BMI ≥ 30 and interview) distressed than those was controlled in all
n = 18 seeking dietary seeking support for analyses
intervention dietary restriction
Mean age 46.8 years
(Australian)
Kolotkin n = 3353 varying in BMI IWQOL-Lite Obesity-specific QOL Treatment seekers Not tested Yes
et al. [57] treatment-seeking (questionnaire) most impaired in those more impaired than
status and intensity of with highest BMI, white non-treatment seekers
treatment sought people and women QOL varied with
Aged 18–90 years intensity of treatment
(North American) sought Most
intense = most
impairment
Kolotkin n = 339 surgical cases BMI IWQOL-Lite QOL significantly lower BMI, gender, Not tested Yes
et al. [56] n = 87 obese controls Diagnosis of and depression rates treatment-seeking
(North American) depression by higher in surgery status and depression
primary care candidates accounted for most
provider variance in QOL
(cont.)
Table 11.3 (cont.)
Obese treatment
seekers more Obese treatment
depressed than seekers more
Measure of Measure of non-obese controls depressed than
Ref. Sample details obesity depression Main findings Other results or population norms obese controls
Laferrere n = 145 obese women BMI BSI, LDI BSI depression slightly No racial differences Yes Not tested
et al. [50] Mean age 46.3 years (questionnaire) above non-obese control
(African American and population
white American)
Marchesini n = 207 obese BMI BDI 43% had BDI scores Significantly related to Yes Not tested
et al. [51] treatment seekers (questionnaire) indicative of depression poor QOL
(Italian)
Musante n = 1184 patients of BMI BDI BDI scores above general Gender differences in Yes Not tested
et al. [61] residential weight-loss Obese = (questionnaire) population norms; those predictors of comorbid
programme BMI ≥ 30 with moderate to severe depression
Mean age 47 years depression had
(North American) significantly higher BMI
than those with none,
mild or moderate
depression
Ryden n = 2510 BMI Measure of Surgical candidates Not tested Not tested
et al. [62] Aged 37–57 years distress devised reported significantly
(Swedish) for the SOS study higher levels of distress
(questionnaire) than conventional
treatment seekers
BDI, Beck Depression Inventory; BMI, body mass index; BSI, Brief Symptom Inventory; GBP, gastric bypass; IWQOL, Impact of Weight on Quality of Life; LDI, Life
Distress Inventory; M-CIDI, Munich Composite International Diagnostic Interview; MMPI-2, Minnesota Multiphasic Personality Interview 2; N/A, not applicable; QOL,
quality of life; SCL-90-R, Symptom Checklist 90, revised; SF-36, Medical Outcomes Study Short Form Health Survey, SOS, Symptom Onset in Schizophrenia.
251 Depression and obesity
There is evidence to show that individuals who seek treatment for their obesity
may differ in important ways from those who do not [54–57]. Because of potential
problems of interpretation, Fitzgibbon et al. [54] matched their three groups demo-
graphically. Treatment seekers had higher levels of psychopathology and more binge
eating than the obese or normal-weight controls, but both obese groups reported
more symptoms of distress than the normal-weight group. Similarly, Kolotkin et al.
[56] reported higher rates of diagnosed depression in surgery cases than in matched
non-treatment-seeking controls. However, Fontaine et al. [55] found that although
obese individuals who had sought treatment were heavier and had a higher preva-
lence of chronic disease and pain than those who were not trying to lose weight,
they were not more depressed.
Even within groups of treatment seekers there is variation. For example, Musante
et al. [61] documented increased severity of depression with increasing BMI in
1184 self-admitted patients, although no such effect was observed by Elfhag and
colleagues [48], albeit in a much smaller sample (n = 120). Most of the research
in this area has looked at candidates for gastric surgery, but one study attempted to
compare the health-related quality of life of overweight/obese individuals between
groups differing in treatment-seeking status and treatment intensity [57]. The study
found that the greater the intensity of treatment sought, the greater the impairment
in quality of life experienced, although depression itself was not assessed. Two
other studies also suggest that those seeking more intensive forms of treatment
are significantly more distressed than those enrolled in more conventional weight-
management programmes [58,62].
Intervention and treatment studies provide more powerful methods with which
to examine causal pathways between obesity and depression, although, with some
exceptions, most studies in this area concern outcomes after gastric surgery, which
limits the possibility of assessing the impact of more typical (10%) weight losses. In
addition, control groups are rarely included, and cases are simply compared with
population norms, reducing the usefulness of the findings. Overall there is little
doubt that, at least in the immediate post-surgery period, significant improvements
in mood accompany substantial weight loss [63–66]. In a two-year follow-up of
the Swedish Obese Subjects study, Ryden et al. [67] found that the pattern and
magnitude of change in distress for individuals who had undergone surgery was the
same as for those who had undergone more conventional weight-loss treatment;
those who lost weight by whatever means experienced a reduction in distress,
whereas weight gainers’ distress levels remained the same. Positive long-term effects
were documented by Dixon and colleagues [68], whose participants completed the
BDI before surgery and annually thereafter. After one year, weight loss was associated
with significant reductions in BDI score that were sustained up to four years post-
surgery. The greatest reductions were seen in those who lost most weight, were
female and were younger.
252 Depression and physical illness
Moderating variables
The most common research strategy in the literature has been to examine sim-
ple associations between obesity and depression. However, both Friedman and
Brownell [16] and Faith et al. [17] recommended further investigation of moder-
ator/mediator models in which causal pathways within subgroups of individuals
can be tested. The search for moderating variables addresses the possibility that
some population subgroups might experience greater psychological consequences
of obesity than others; for example, obesity might pose more of a psychological
hazard for women than for men. Unfortunately, even in the more recent literature,
few studies have been designed to test for moderation, and where they do power
for these subgroup analyses is often limited. Nevertheless, a number of putative
moderators are emerging.
Gender is one such moderator, with some studies reporting stronger associations
between obesity and depression in women than in men, e.g. [26,30], although
253 Depression and obesity
others have found no gender differences [45]. Some studies have found similar
effect sizes in both sexes, but with wider confidence intervals (CI) for the male
sample, reflecting the lower risk in men of both obesity and depression, e.g. [28].
Overall, the pattern of results is not sufficiently clear to conclude that gender is an
important moderating variable.
Age has been examined as a moderator in some studies, with an assumption
that obesity in adolescence might be particularly stressful because of the salience of
appearance in this age group [74–76]. Surprisingly, there has been little evidence
for a moderating effect; neither has there been any sign that the associations tend
to be stronger in samples of younger adults. In fact, the effects are probably more
consistently negative among adolescents than older adults [77].
Race and socioeconomic status (SES) have often been linked with the social
costs of obesity and were identified by Friedman and Brownell [16] as plausible
mediators. Obesity appears to carry more adverse social implications in higher-SES
and white groups, perhaps because meeting appearance norms is more valued [78]
or appearance norms are for smaller body sizes. Ross [79] found that being over-
weight increased depression among well-educated individuals (those with a college
degree) but had an almost non-existent effect among poorly educated individuals;
however, others have failed to find an effect of education [20,28]. Carpenter and
colleagues [26] are one of the few groups to look at interactions with ethnicity; they
found no evidence that obesity had a differential association with depression in
African Americans compared with whites. Another study to address demographic
moderators directly is an investigation of the psychological correlates of obesity in
the Health and Behaviour in Teenagers (HABITS study) [80]. Weights and heights
were measured in almost 5000 British adolescents aged 11–12 years and depressed
mood was assessed using Goodman’s Strengths and Difficulties Scale [81]. In this
sample, neither the overall association between obesity and depression, nor any
interactions with gender, ethnicity or SES were significant, giving no support for
the role of demographic moderating variables in this population.
The severity of obesity has emerged as a moderator in most studies in which
it has been addressed [28,29,61]. This could be because recent secular trends in
obesity have resulted in greater acceptance of ‘normal’ obesity, with stigmatisa-
tion now limited to the ‘super obese’ (BMI > 40), although Myers and Rosen
[82] found that the experience of stigmatisation increases in proportion to body
mass up to a BMI of 40 and levels off thereafter. Possibly more salient is the
fact that severe obesity is associated with dramatically higher rates of ill-health,
which of itself is known to be linked with higher rates of depression [30,79]. If
ill-health is an important mediating mechanism in the obesity-depression associ-
ation, then there is a need for studies that compare levels of depression in obese
patients with other clinical groups suffering comparable functional limitations
254 Depression and physical illness
before concluding that the relationship between obesity and depression has a unique
status.
Binge eating
One of the most consistent markers of psychological distress – including
depression – in obese patients is binge eating. According to a thorough review of the
literature, there is considerable evidence that obese individuals with binge-eating
disorder (BED) have higher rates of general psychiatric symptomatology (espe-
cially depression) than obese non-binge eaters [83]. More recent research tends to
support this conclusion in younger adults [84], patients undergoing gastric bypass
surgery [85] and weight-loss treatment seekers [86]. One study has failed to confirm
the effect [87], and the authors suggest that this is because they diagnosed BED
using strict DSM-IV criteria, unlike most other studies, which have overestimated
its prevalence. However, with this one exception, there is general agreement that
obese binge eaters are a subgroup who have more psychological difficulties.
Once the association between obesity and depression is established, the search for
mediators becomes important. These are variables that lie on the causal pathway
from obesity to depression; for example, experiences of social exclusion associ-
ated with obesity might be part of the causal process leading to a greater risk of
depression.
The pathway through experiences of social rejection or humiliation is one
important route, and a significant number of studies have focused on teasing as the
marker of adverse social experiences. The literature on teasing is limited by being
concerned almost exclusively with children, although some studies have looked at
its impact on body dissatisfaction in adulthood [88,89]. Jackson and colleagues
[89] found that a history of teasing about physical appearance was associated with
both depression and body dissatisfaction. In another study examining the impact of
stigmatisation and discrimination on the psychological functioning of obese indi-
viduals, more frequent exposure to stigmatisation was associated with greater psy-
chological distress [82]. The problem with reports of teasing as the marker of social
adversity is the possibility of reporting bias arising from a general negative affectivity.
overweight and depression for both men and women and was also related directly to
negative affect. Likewise, Sarwer et al. [91] found an increased risk of depression in
treatment-seeking obese individuals with more negative body image. Interestingly,
overall their patients did not differ from normal-weight population controls in
depression. There is evidence from population samples that body dissatisfaction –
at all weight levels – is associated with depression [92], making body dissatisfaction
a plausible mechanism for depression in the obese. But again there is difficulty
in determining the direction of effect: does body dissatisfaction raise the risk of
depression, or does depression promote dissatisfaction, or could the association be
another reflection of negative affectivity?
Dieting status has also been implicated in the association between obesity and
depression. Ross [79] used data from a telephone interview of a random sample
of US adults aged 18–90 years and found that although overweight was related to
depression after controlling for sociodemographic variables, dieting was found to
increase depression and rendered its relationship with obesity non-significant. A
related issue may be that of weight cycling. It is difficult to imagine that repeated
failures to maintain weight loss would not have adverse psychological consequences,
but the data do not provide strong support. Some cross-sectional studies have
reported negative psychological effects of weight fluctuation [69,93] and others
no effect [94,95]. A rare prospective study [96] failed to support an association
between weight cycling and long-term adverse psychological effects, and a review
by the same research group [97] suggests that conclusions that weight cycling is
psychologically damaging are premature.
Conclusions
Large-scale cross-sectional studies of community samples of adults suggest that
obesity, and particularly severe obesity, is associated with depressed mood or a
diagnosis of depression, with rates of obesity being approximately doubled in obese
compared with normal-weight groups. However, the majority of obese individuals
are not depressed, and depression in obese individuals should be treated as seriously
as in others and not presumed to be a consequence of obesity. Prospective studies are
scarce, but support a link between obesity and incident depression. Studies in clini-
cally obese populations generally document higher rates of depression than among
either non-treatment-seeking obese or normal-weight control groups. However,
those who seek treatment may differ in other important ways from those who do
not, and depression itself may be a stimulus for seeking professional help. More-
over, even within clinical groups there is variation, and it cannot be assumed that
depression is an inevitable consequence of severe overweight. As two reviews have
concluded, the most important area for future research concerns the investigation
of the mechanisms underlying depression–obesity covariations [16,17]. Potential
mediators are suggested by the literature and, where modifiable, should be targeted
for intervention.
REFERENCES
1. World Health Organization, Obesity and Overweight (Geneva: World Health Organization,
2003).
2. K. M. Flegal M. D. Carroll, C. L. Ogden, C. L. Johnson, Prevalence and trends in obesity
among US adults, 1999–2000. J. Am. Med. Assoc. 288 (2002), 1723–7.
3. G. A. Bray, Medical consequences of obesity. J. Clin. Endocrinol. Metab. 89 (2004), 2583–9.
4. L. Khaodhiar, K. C. McCowen, G. L. Blackburn, Obesity and its comorbid conditions. Clin.
Cornerstone 2 (1999), 17–31.
5. F. X. Pi-Sunyer, Medical hazards of obesity. Ann. Intern. Med. 119 (1993), 655–60.
257 Depression and obesity
6. T. L. Visscher, J. C. Seidell, The public health impact of obesity. Annu. Rev. Public Health 22
(2001), 355–75.
7. H. I. Kaplan, H. S. Kaplan, The psychosomatic concept of obesity. J. Nerv. Ment. Dis. 125
(1957), 181–201.
8. P. B. Goldblatt, M. E. Moore, A. J. Stunkard, Social factors in obesity. J. Am. Med. Assoc. 192
(1965), 1039–44.
9. P. Bjorntorp, Endocrine abnormalities of obesity. Metabolism 44 (1995), 21–3.
10. G. Oliver, J. Wardle, E. L. Gibson, Stress and food choice: a laboratory study. Psychosom. Med.
62 (2000), 853–65.
11. R. B. Harris, Leptin: much more than a satiety signal. Annu. Rev. Nutr. 20 (2000), 45–75.
12. A. Steptoe, S. R. Kunz-Ebrecht, L. Brydon, J. Wardle, Central adiposity and cortisol responses
to waking in middle-aged men and women. Int. J. Obes. Relat. Metab. Disord. 28 (2004),
1168–73.
13. M. F. Dallman, N. Pecoraro, S. F. Akana, et al., Chronic stress and obesity: a new view of
‘comfort food’. Proc. Natl. Acad. Sci. U. S. A. 100 (2003), 11 696–701.
14. R. M. Puhl, K. D. Brownell, Psychosocial origins of obesity stigma: toward changing a powerful
and pervasive bias. Obes. Rev. 4 (2003), 213–27.
15. R. Puhl, K. D. Brownell, Bias, discrimination, and obesity. Obes. Res. 9 (2001), 788–805.
16. M. A. Friedman, K. D. Brownell, Psychological correlates of obesity: moving to the next
research generation. Psychol. Bull. 117 (1995), 3–20.
17. M. S. Faith, P. E. Matz, M. A. Jorge, Obesity-depression associations in the population. J.
Psychosom. Res. 53 (2002), 935–42.
18. A. J. Stunkard, M. S. Faith, K. C. Allison, Depression and obesity. Biol. Psychiatry 54 (2003),
330–37.
19. T. A. Wadden, D. B. Sarwer, L. G. Womble, et al., Psychosocial aspects of obesity and obesity
surgery. Surg. Clin. North Am. 81 (2001), 1001–24.
20. J. Istvan, K. Zavela, G. Weidner, Body weight and psychological distress in NHANES I. Int.
J. Obes. Relat. Metab. Disord. 16 (1992), 999–1003.
21. T. A. Wadden, G. D. Foster, A. J. Stunkard, J. R. Linowitz, Dissatisfaction with weight and
figure in obese girls: discontent but not depression. Int. J. Obes. 13 (1989), 89–97.
22. A. L. Stewart, R. H. Brook, Effects of being overweight. Am. J. Public Health 73 (1983),
171–8.
23. M. J. Seegers, C. Mertens, Psychological and bioclinical CHD risk factors: quantitative dif-
ferences between obese, normal and thin subjects. J. Psychosom. Res. 18 (1974), 403–11.
24. M. Sullivan, J. Karlsson J. Sjostrom, et al., Swedish obese subjects (SOS): an intervention
study of obesity – baseline evaluation of health and psychosocial functioning in the first 1743
subjects examined. Int. J. Obes. Relat. Metab. Disord. 17 (1993), 503–12.
25. R. R. Wing, K. A. Matthews, L. H. Kuller, E. N. Meilahn, P. Plantinga, Waist to hip ratio in
middle-aged women: associations with behavioral and psychosocial factors and with changes
in cardiovascular risk factors. Arterioscler. Thromb. 11 (1991), 1250–57.
26. K. M. Carpenter, D. S. Hasin, D. B. Allison, M. S. Faith, Relationships between obesity and
DSM-IV major depressive disorder, suicide ideation, and suicide attempts: results from a
general population study. Am. J. Public Health 90 (2000), 251–7.
258 Depression and physical illness
27. E. Johnston, S. Johnson, P. McLeod, M. Johnston, The relation of body mass index to depres-
sive symptoms. Can. J. Public Health 95 (2004), 179–83.
28. C. U. Onyike, R. M. Crum, H. B. Lee, C. G. Lyketsos, W. W. Eaton, Is obesity associated
with major depression? Results from the Third National Health and Nutrition Examination
Survey. Am. J. Epidemiol. 158 (2003), 1139–47.
29. C. Dong, L. E. Sanchez, R. A. Price, Relationship of obesity to depression: a family-based
study. Int. J. Obes. Relat. Metab. Disord. 28 (2004), 790–95.
30. A. F. Jorm, A. E. Korten, H. Christensen, et al., Association of obesity with anxiety, depression
and emotional well-being: a community survey. Aust. N. Z. J. Public Health 27 (2003), 434–
40.
31. A. C. Ahlberg, T. Ljung, R. Rosmond, et al., Depression and anxiety symptoms in relation to
anthropometry and metabolism in men. Psychiatry Res. 112 (2002), 101–10.
32. R. E. Roberts, S. Deleger, W. J. Strawbridge, G. A. Kaplan, Prospective association between
obesity and depression: evidence from the Alameda County Study. Int. J. Obes. Relat. Metab.
Disord. 27 (2003), 514–21.
33. M. S. Kaplan, N. Huguet, J. T. Newsom, B. H. McFarland, J. Lindsay, Prevalence and cor-
relates of overweight and obesity among older adults: findings from the Canadian National
Population Health Survey. J. Gerontol. A Biol. Sci. Med. Sci. 58 (2003), 1018–30.
34. L. Z. Bin, H. S. Yin, C. W. Man, et al., Obesity and depressive symptoms in Chinese elderly.
Int. J. Geriatr. Psychiatry 19 (2004), 68–74.
35. H. Noppa, T. Hallstrom, Weight gain in adulthood in relation to socioeconomic factors,
mental illness and personality traits: a prospective study of middle-aged women. J. Psychosom.
Res. 25 (1981), 83–9.
36. L. DiPietro, R. F. Anda, D. F. Williamson, A. J. Stunkard, Depressive symptoms and weight
change in a national cohort of adults. Int. J. Obes. Relat. Metab. Disord. 16 (1992), 745–53.
37. G. I. Papakostas, T. Petersen, D. V. Iosifescu, et al., Obesity among outpatients with major
depressive disorder. Int. J. Neuropsychopharmacol. 8 (2005), 59–63.
38. A. M. Bardone, T. E. Moffitt, A. Caspi, et al., Adult physical health outcomes of adolescent
girls with conduct disorder, depression, and anxiety. J. Am. Acad. Child Adolesc. Psychiatry
37 (1998), 594–601.
39. D. S. Pine, P. Cohen, J. Brook, J. D. Coplan, Psychiatric symptoms in adolescence as predictors
of obesity in early adulthood: a longitudinal study. Am. J. Public Health 87 (1997), 1303–10.
40. D. S. Pine, R. B. Goldstein, S. Wolk, M. M. Weissman, The association between childhood
depression and adulthood body mass index. Pediatrics 107 (2001), 1049–56.
41. L. P. Richardson, R. Davis, R. Poulton, et al., A longitudinal evaluation of adolescent depres-
sion and adult obesity. Arch. Pediatr. Adolesc. Med. 157 (2003), 739–45.
42. J. C. Barefoot, B. L. Heitmann, M. J. Helms, et al., Symptoms of depression and changes
in body weight from adolescence to mid-life. Int. J. Obes. Relat. Metab. Disord. 22 (1998),
688–94.
43. G. Hasler, D. S. Pine, A. Gamma, et al., The associations between psychopathology and being
overweight: a 20-year prospective study. Psychol. Med. 34 (2004), 1047–57.
44. R. E. Roberts, W. J. Strawbridge, S. Deleger, G. A. Kaplan, Are the fat more jolly? Ann. Behav.
Med. 24 (2002), 169–80.
259 Depression and obesity
45. R. E. Roberts, G. A. Kaplan, S. J. Shema, W. J. Strawbridge, Are the obese at greater risk for
depression? Am. J. Epidemiol. 152 (2000), 163–70.
46. I. Berlin, F. Lavergne, Relationship between body-mass index and depressive symptoms in
patients with major depression. Eur. Psychiatry 18 (2003), 85–8.
47. T. L. Schwartz, N. Nihalani, S. Jindal, S. Virk, N. Jones, Psychiatric medication-induced
obesity: a review. Obes. Rev. 5 (2004), 115–21.
48. K. Elfhag, S. Rossner, A. M. Carlsson, Degree of body weight in obesity and Rorschach
personality aspects of mental distress. Eat. Weight Disord. 9 (2004), 35–43.
49. J. Glinski, S. Wetzler, E. Goodman, The psychology of gastric bypass surgery. Obes. Surg. 11
(2001), 581–8.
50. B. Laferrere, S. Zhu, J. R. Clarkson, et al., Race, menopause, health-related quality of life, and
psychological well-being in obese women. Obes. Res. 10 (2002), 1270–75.
51. G. Marchesini, M. Bellini, S. Natale, et al., Psychiatric distress and health-related quality of
life in obesity. Diabetes Nutr. Metab. 16 (2003), 145–54.
52. D. W. Black, R. B. Goldstein, E. E. Mason, Prevalence of mental disorder in 88 morbidly obese
bariatric clinic patients. Am. J. Psychiatry 149 (1992), 227–34.
53. B. Britz, W. Siegfried, A. Ziegler, et al., Rates of psychiatric disorders in a clinical study group
of adolescents with extreme obesity and in obese adolescents ascertained via a population
based study. Int. J. Obes. Relat. Metab. Disord. 24 (2000), 1707–14.
54. M. L. Fitzgibbon, M. R. Stolley, D. S. Kirschenbaum, Obese people who seek treatment have
different characteristics than those who do not seek treatment. Health Psychol. 12 (1993),
342–5.
55. K. R. Fontaine, S. J. Bartlett, I. Barofsky, Health-related quality of life among obese persons
seeking and not currently seeking treatment. Int. J. Eat. Disord. 27 (2000), 101–5.
56. R. L. Kolotkin, R. D. Crosby, R. Pendleton, et al., Health-related quality of life in patients
seeking gastric bypass surgery vs non-treatment-seeking controls. Obes. Surg. 13 (2003),
371–7.
57. R. L. Kolotkin, R. D. Crosby, G. R. Williams, Health-related quality of life varies among obese
subgroups. Obes. Res. 10 (2002), 748–56.
58. M. L. Higgs, T. Wade, M. Cescato, et al., Differences between treatment seekers in an obese
population: medical intervention vs. dietary restriction. J. Behav. Med. 20 (1997), 391–405.
59. R. J. Hafner, J. M. Watts, J. Rogers, Psychological status of morbidly obese women before
gastric restriction surgery. J. Psychosom. Res. 31 (1987), 607–12.
60. E. V. Leckie, R. F. Withers, Obesity and depression. J. Psychosom. Res. 11 (1967), 107–15.
61. G. J. Musante, P. R. Costanzo, K. E. Friedman, The comorbidity of depression and eating
dysregulation processes in a diet-seeking obese population: a matter of gender specificity.
Int. J. Eat. Disord. 23 (1998), 65–75.
62. A. Ryden, J. Karlsson, L. O. Persson, et al., Obesity-related coping and distress and relationship
to treatment preference. Br. J. Clin. Psychol. 40 (2001), 177–88.
63. M. P. Dymek, D. le Grange, K. Neven, J. Alverdy, Quality of life and psychosocial adjustment
in patients after Roux-en-Y gastric bypass: a brief report. Obes. Surg. 11 (2001), 32–9.
64. J. Karlsson, L. Sjostrom, M. Sullivan, Swedish obese subjects (SOS): an intervention study
of obesity – two-year follow-up of health-related quality of life (HRQL) and eating behavior
260 Depression and physical illness
after gastric surgery for severe obesity. Int. J. Obes. Relat. Metab. Disord. 22 (1998), 113–
26.
65. S. R. Maddi, S. R. Fox, D. M. Khoshaba, et al., Reduction in psychopathology following
bariatric surgery for morbid obesity. Obes. Surg. 11 (2001), 680–85.
66. G. S. Waters, W. J. Pories, M. S. Swanson, et al., Long-term studies of mental health after the
Greenville gastric bypass operation for morbid obesity. Am. J. Surg. 161 (1991), 154–7.
67. A. Ryden, J. Karlsson, M. Sullivan, J. S. Torgerson, C. Taft, Coping and distress: what hap-
pens after intervention? A 2-year follow-up from the Swedish Obese Subjects (SOS) study.
Psychosom. Med. 65 (2003), 435–42.
68. J. B. Dixon, M. E. Dixon, P. E. O’Brien, Depression in association with severe obesity: changes
with weight loss. Arch. Intern. Med. 163 (2003), 2058–65.
69. T. A. Wadden, A. J. Stunkard, J. Liebschutz, Three-year follow-up of the treatment of obesity
by very low calorie diet, behavior therapy, and their combination. J. Consult. Clin. Psychol.
56 (1988), 925–8.
70. O. Ryden, S. A. Olsson, A. Danielsson, P. Nilsson-Ehle, Weight loss after gastroplasty: psy-
chological sequelae in relation to clinical and metabolic observations. J. Am. Coll. Nutr. 8
(1989), 15–23.
71. W. T. Tsushima, M. P. Bridenstine, J. F. Balfour, MMPI-2 scores in the outcome prediction
of gastric bypass surgery. Obes. Surg. 14 (2004), 528–32.
72. M. M. Clark, B. M. Balsiger, C. D. Sletten, et al., Psychosocial factors and 2-year outcome
following bariatric surgery for weight loss. Obes. Surg. 13 (2003), 739–45.
73. E. M. Inelmen, E. D. Toffanello, G. Enzi, et al., Predictors of drop-out in overweight and
obese outpatients. Int. J. Obes. Relat. Metab. Disord. 29 (2005), 122–8.
74. P. J. Cooper, I. Goodyer, Prevalence and significance of weight and shape concerns in girls
aged 11–16 years. Br. J. Psychiatry 171 (1997), 542–4.
75. D. Neumark-Sztainer, J. Croll, M. Story, et al., Ethnic/racial differences in weight-related
concerns and behaviors among adolescent girls and boys: findings from Project EAT.
J. Psychosom. Res. 53 (2002), 963–74.
76. P. Packard, K. S. Krogstrand, Half of rural girls aged 8 to 17 years report weight concerns and
dietary changes, with both more prevalent with increased age. J. Am. Diet. Assoc. 102 (2002),
672–7.
77. J. Wardle, L. Cooke, The impact of obesity on psychological well-being in children. Best Pract.
Res. Clin. Endocrinol. Metab. 19 (2005), 421–40.
78. J. Wardle, K. A. Robb, F. Johnson, et al., Socioeconomic variation in attitudes to eating and
weight in female adolescents. Health Psychol. 23 (2004), 275–2.
79. C. E. Ross, Overweight and depression. J. Health Soc. Behav. 35 (1994), 63–79.
80. J. Wardle, S. Williamson, F. Johnson, C. Edwards, The psychological consequences of obesity
in adolescence: cultural moderators of the obesity-depression association. Int. J. Obes. (Lond.)
30 (2006), 634–42.
81. R. Goodman, The Strengths and Difficulties Questionnaire: a research note. J. Child Psychol.
Psychiatry 38 (1997), 581–6.
82. A. Myers, J. C. Rosen, Obesity stigmatization and coping: relation to mental health symptoms,
body image, and self-esteem. Int. J. Obes. Relat. Metab. Disord. 23 (1999), 221–30.
261 Depression and obesity
83. M. de Zwaan, Binge eating disorder and obesity. Int. J. Obes. Relat. Metab. Disord. 25: Suppl
1 (2001), 51–5.
84. T. Reichborn-Kjennerud, C. M. Bulik, P. F. Sullivan, K. Tambs, J. R. Harris, Psychiatric and
medical symptoms in binge eating in the absence of compensatory behaviors. Obes. Res. 12
(2004), 1445–54.
85. M. Malone, S. Alger-Mayer, Binge status and quality of life after gastric bypass surgery: a
one-year study. Obes. Res. 12 (2004), 473–81.
86. J. A. Linde, R. W. Jeffery, R. L. Levy, et al., Binge eating disorder, weight control self-efficacy,
and depression in overweight men and women. Int. J. Obes. Relat. Metab. Disord. 28 (2004),
418–25.
87. M. de Zwaan, J. E. Mitchell, L. M. Howell, et al., Characteristics of morbidly obese patients
before gastric bypass surgery. Compr. Psychiatry 44 (2003), 428–34.
88. C. M. Grilo, D. E. Wilfley, K. D. Brownell, J. Rodin, Teasing, body image, and self-esteem in
a clinical sample of obese women. Addict. Behav. 19 (1994), 443–50.
89. T. D. Jackson, C. M. Grilo, R. M. Masheb, Teasing history, onset of obesity, current eating
disorder psychopathology, body dissatisfaction, and psychological functioning in binge eating
disorder. Obes. Res. 8 (2000), 451–8.
90. K. E. Friedman, S. K. Reichmann, P. R. Costanzo, G. J. Musante, Body image partially mediates
the relationship between obesity and psychological distress. Obes. Res. 10 (2002), 33–41.
91. D. B. Sarwer, T. A. Wadden, G. D. Foster, Assessment of body image dissatisfaction in obese
women: specificity, severity, and clinical significance. J. Consult. Clin. Psychol. 66 (1998),
651–4.
92. F. Johnson, J. Wardle, Dietary restraint, body dissatisfaction, and psychological distress: a
prospective analysis. J. Abnorm. Psychol. 114 (2005), 119–25.
93. J. P. Foreyt, R. L. Brunner, G. K. Goodrick, et al., Psychological correlates of weight fluctuation.
Int. J. Eat. Disord. 17 (1995), 263–75.
94. S. J. Bartlett, T. A. Wadden, R. A. Vogt, Psychosocial consequences of weight cycling. J. Consult.
Clin. Psychol. 64 (1996), 587–92.
95. T. A. Wadden, S. Bartlett, K. A. Letizia, et al., Relationship of dieting history to resting
metabolic rate, body composition, eating behavior, and subsequent weight loss. Am. J. Clin.
Nutr. 56 (1992), 203–8S.
96. G. D. Foster, T. A. Wadden, P. C. Kendall, A. J. Stunkard, R. A. Vogt, Psychological effects of
weight loss and regain: a prospective evaluation. J. Consult. Clin. Psychol. 64 (1996), 752–7.
97. G. D. Foster, D. B. Sarwer, T. A. Wadden, Psychological effects of weight cycling in obese
persons: a review and research agenda. Obes. Res. 5 (1997), 474–88.
Part 3
Introduction
The prevalence of depression in medically ill patients is high, ranging from 5% to
20% versus 3–5% in the general population. The comorbidity of depressive disor-
ders in general hospital patients or in-patients with chronic pathologies who consult
general practitioners (GPs) is about 40% [1,2]. The rates are similar in women with
breast cancer [3]. Several studies have shown that patients with chronic diseases,
including arthritis and coronary heart disease (CHD), are more at risk than healthy
subjects for developing mood disorders [4–7]. In most of these studies, depression
was identified as a significant risk factor for mortality and health/treatment compli-
cations. Depression was associated with greater physical limitation, more frequent
subjective health complaints, lower treatment compliance and satisfaction, and
reduced perceived quality of life. In addition, the prevalence of major depression in
later life was also found to be more frequent among people with chronic illnesses
[8].
Psychological factors are usually put forward to explain the comorbidity of
depression with medical illnesses. Medically ill patients often suffer from physi-
cal limitations and have difficulties maintaining positive social relationships. They
frequently report feelings of hopelessness, since they cannot cope with the prob-
lems they are confronted with, especially in situations of poor social support. These
factors increase the risk for the occurrence of depressive disorders. Lack of com-
pliance with medical treatments, decreased social support and the consequences of
biological changes associated with depression (e.g. impaired cellular immune func-
tions) are believed to account for the association between depression and increased
morbidity and mortality.
The objective of this chapter is to provide an alternative perspective on the
intricate relationships that exist between depressive disorders and a large number
can act on distant targets), cytokines are usually active in the microenvironment
in which they are produced and function as autocrine and paracrine factors rather
than as hormones. An important finding is the observation that the production of
IL-1 and other pro-inflammatory cytokines by peripheral innate immune cells is
associated with the production of the same cytokines in the brain by macrophage-
like cells in the meninges and around blood vessels and by microglial cells within
the brain parenchyma [14,15]. In the brain, IL-1 acts as a pivotal cytokine, since it
regulates not only its own synthesis but also the synthesis of other pro-inflammatory
cytokines, including TNF- and IL-6 [16].
All these findings point to the existence of a brain compartment of cytokines
that reflects at the cellular and molecular levels the activation of peripheral innate
immune cells, the induction of which is responsible for the central component of the
acute-phase reaction. This cannot be possible without communication pathways
allowing peripheral immune messages to be transmitted from the periphery to the
brain. There are at least three communication pathways that transmit the peripheral
immune message to the brain. These pathways include blood transfer of cytokines
and pathogen-associated molecular patterns, specific transporters supporting pas-
sage of cytokines across the blood–brain barrier, and activation of afferent neural
pathways. The exact nature of these communication pathways varies with the phys-
iological endpoint under consideration, probably because of the spatial location
of the neuronal circuits that underlie its expression. Afferent neural pathways play
a major role in the development of cytokine-induced sickness behaviour [17]. In
contrast, the humoral route appears to play a more important role for the fever
response than for sickness behaviour [18].
The brain cytokine message originates in the choroid plexus and the circum-
ventricular organs, which are devoid of a functional blood–brain barrier. Subse-
quently, cytokines propagate into the brain parenchyma by volume diffusion and
act on brain structures that, for some of them, certainly need to be sensitised by
the afferent neural message [17]. It is not yet known whether cytokines ultimately
act on brain functions by targeting neuronal cells or whether they act indirectly
by modifying the neural environment that is formed by glial and endothelial cells.
The generation of molecular intermediates such as prostaglandins and nitric oxide
(NO) may be involved in these actions, but the detailed molecular mechanisms of
cytokine actions in the brain still need to be elucidated [19]. These mechanisms
involve complex relationships between endothelial cells of blood vessels and glial
cells, including macrophages surrounding the blood vessels and macrophages in the
meninges. Within the brain parenchyma, microglial cells are also involved. Astro-
cytes interposed between blood vessels and neurons play an important role in the
regulation of glutamate metabolism and ionic equilibrium and function as both
targets and sources of cytokines.
269 Inflammation, sickness behaviour and depression
In summary, when the innate immune system is activated, the brain forms a
molecular and cellular image of the peripheral inflammatory response. It is this
neural representation of the peripheral immune events that organises the host’s
response in its subjective, behavioural and metabolic components.
and even suicidal ideation [30]. The onset of depressive symptoms depends on the
cytokine and treatment modalities (e.g. dosage, administration route). Interest-
ingly, the occurrence of depression can be prevented by pretreatment with paroxe-
tine, a selective serotonin reuptake inhibitor (SSRI) with antidepressant properties.
Nevertheless, pretreatment with paroxetine has a minimal or null effect on the
development of neurovegetative symptoms of sickness, including fever, fatigue and
anorexia, confirming the dissociation between sickness behaviour and depression
[31].
These findings are important because they can be interpreted to suggest that
depressive disorders develop from cytokine-induced sickness behaviour only in vul-
nerable patients. Vulnerability, in the present context, refers to an innate or acquired
predisposition to develop a given pathology when the causal factors are present.
Dysfunction in genes controlling key proteins in serotoninergic neurotransmission,
e.g. activity of the serotonin transporter [32], and childhood trauma [33] have been
identified as vulnerability factors for depression. Vulnerability to cytokine-induced
depression can be revealed by psychological features. Indeed, patients with a rela-
tively high score on depression scales, including the Montgomery Asberg Depression
Rating Scale (MADRS) and the Hamilton Depression Rating Scale, at the start of
cytokine treatment are more likely to develop depressive syndrome in response to
cytokine treatment than patients who have a relatively low score at baseline [34,35].
Vulnerability can also be revealed by physiological features. Patients who respond to
the first injection of IFN- by an exaggerated pituitary-adrenal response are more
likely to become depressed in response to repeated administration of IFN- than
patients who display a lower pituitary-adrenal response [36]. These two different
characteristics are markers of vulnerability. They help us to identify patients who
are at risk, but they do not explain why patients having these characteristics are
more vulnerable than those who do not have them. In particular, it remains to
be elucidated whether the risk factors for developing depression during cytokine
therapy are the same as those accounting for depressive disorders in psychiatric
patients, or whether these risk factors are specific to cytokines and involve, for
instance, polymorphisms in genes involved in cytokine production.
The model of cytokine-induced depression has the advantage of providing clini-
cians with the possibility of observing the development of depressive symptoms in a
longitudinal manner in a relatively large number of patients who can be monitored
closely from the time they start receiving immunotherapy. Furthermore, patients
who develop depression can be compared transversally with patients who remain
free of any mood disorder. The model of cytokine-induced depression provides
valuable insights into the relationship between cytokines and depression. At the
clinical level, there is some evidence that symptoms of mood disorder are more
polymorphic than simply depression. A study on patients with hepatitis C treated
274 Depression and physical illness
with IFN- showed that dysphoria and mixed states dominate the clinical presenta-
tion of patients, with increases in irritability and anxiety as the main symptoms [37].
The reasons for the differences between patients with hepatitis C and those with
cancer, who present mainly with depressed mood, are not known. They could be due
to the medical context (cytokine immunotherapy is only palliative in cancer, but it
is usually curative in hepatitis C), to the immunological context (immune responses
of patients infected by a virus are different from those of patients with cancer) or
simply to variations in treatment modalities (high doses of IFN- administered
intravenously and daily to patients with malignant melanoma versus low doses of
pegylated IFN-. administered once a week to patients with chronic hepatitis C).
These variations may also be due to differences in affective/psychiatric background,
since a large number of patients with chronic hepatitis C have a history of substance
abuse.
At the pathophysiological level, an important insight into the chain of events
linking cytokines to mood alterations has emerged from the observation that
cancer patients treated with cytokines develop a drastic decrease in plasma tryp-
tophan levels that correlates with depression scores at four weeks of treatment
[38]. This decrease in plasma trytophan levels had already been noted, but in a
qualitative rather than a quantitative manner [39]. These findings are important,
since bioavailability of tryptophan is the limiting factor for the synthesis of sero-
tonin. The acute depletion of tryptophan produced by feeding excess amounts of
large neutral amino acids that compete with tryptophan for entry into the brain
results in the development of depressed mood in subjects at risk for depression.
A likely candidate for this decrease in plasma tryptophan in patients submitted
to cytokine immunotherapy is the enzyme indoleamine-2,3-dioxygenase, which
degrades tryptophan into kynurenine and quinolinic acid. This enzyme is present
in macrophages, monocytes, endothelial cells and brain glial cells. It is potently
activated by pro-inflammatory cytokines such as TNF- and interferon gamma
(IFN- ), both at the periphery and in the brain [40]. Its activation results in a
decrease in tryptophan bioavailability for the synthesis of serotonin and in the for-
mation of neuroactive compounds such as kynurenine and quinolinic acid that act
at the level of glutamate receptors. The interference of pro-inflammatory cytokines
with serotoninergic neurotransmission can certainly explain some of the clinical
signs that develop in vulnerable patients, such as impulsivity and depressed mood.
However, it does not account for the anhedonia, fatigue and psychomotor retar-
dation that are also observed in cytokine-treated patients [30]. These symptoms
probably reflect a decrease in dopaminergic neurotransmission. This hypothesis is
supported by neuroimaging studies showing alterations in the activity of the basal
ganglia during cytokine therapy [41,42]. A possible mechanism for this effect is
the alteration of tyrosine hydroxylase activity, which is associated with cytokine
275 Inflammation, sickness behaviour and depression
with the disease [49]. This view is in accordance with the postulated role of chronic
neuro-inflammation in the non-specific symptoms of sickness that can be observed
in individuals with Alzheimer’s disease providing one searches for them [50]. What
applies to the various medical conditions that are listed here as examples certainly
applies to other medical conditions that are associated with chronic inflammation
and in which a higher prevalence of mood disorders has already been noted. Based
on the data that have been collected in patients treated with cytokine immunother-
apy, it is now possible to go beyond an educated guess and submit this hypothesis
to test by measuring and relating to each other the inflammatory status, trypto-
phan metabolism, mood changes and other non-specific symptoms in depressed
and non-depressed patients, preferably at the same stage of the disease process.
Various methodologies can be used for this purpose. In view of the progress that
has been made since the mid 1990s, the time is certainly ripe for developing some
form of consensus on the selection of the best techniques available to engage in
this type of research. At the same time, it should be possible to look for possible
cytokine or neurotransmitter candidate genes to account for the vulnerability that
has been alluded to earlier and that represents a key element in the understanding
of cytokine-induced depression.
Acknowledgements
Supported by the Institute National de la Recherche Agronomique (INRA), the
Centre National de la Recherche Scientifique (CNRS), French Ministry of Research
(ACI Neurosciences intégrées et computationnelles to NC) and National Institutes
of Health (NIH) 1 R01 MH071349-01 to RD.
REFERENCES
26. A. Aubert, G. Goodall, R. Dantzer, Compared effects of cold ambient temperature and
cytokines on macronutrient intake in rats. Physiol. Behav. 57 (1995), 869–73.
27. A. Aubert, C. Vega, R. Dantzer, G. Goodall, Pyrogens specifically disrupt the acquisition of a
task involving cognitive processing in the rat. Brain Behav. Immun. 9 (1995), 129–48.
28. A. Aubert, G. Goodall, R. Dantzer, G. Gheusi, Differential effects of lipopolysaccharide on
pup retrieving and nest building in lactating mice. Brain Behav. Immun. 11 (1997), 107–18.
29. H. Anisman, Z. Merali, Cytokines, stress and depressive illness: brain–immune interactions.
Ann. Med. 35 (2003), 2–11.
30. L. Capuron, A. Ravaud, A. H. Miller, R. Dantzer, Baseline mood and psychosocial characteris-
tics of patients developing depressive symptoms during interleukin-2 and/or interferon-alpha
cancer therapy. Brain Behav. Immun. 18 (2004), 205–13.
31. L. Capuron, J. F. Gumnick, D. L. Musselman, et al., Neurobehavioural effects of interferon-
alpha in cancer patients: phenomenology and paroxetine responsiveness of symptom dimen-
sions. Neuropsychopharmacology 26 (2002), 643–52.
32. I. Jones, N. Craddock, Candidate gene studies of bipolar disorder. Ann. Med. 33 (2001),
248–56.
33. C. Heim, C. B. Nemeroff, The role of childhood trauma in the neurobiology of mood and
anxiety disorders: preclinical and clinical studies. Biol. Psychiatry 49 (2001), 1023–39.
34. L. Capuron, A. Ravaud, Prediction of the depressive effects of interferon alfa therapy by the
patient’s initial affective state. N. Engl. J. Med. 340 (1999), 1370.
35. H. Miyaoka, T. Otsubo, K. Kamijima, et al., Depression from interferon therapy in patients
with hepatitis C. Am. J. Psychiatry 156 (1999), 1120.
36. L. Capuron, G. Neurauter, D. L. Musselman, et al., Interferon-alpha-induced changes
in tryptophan metabolism: relationship to depression and paroxetine treatment. Biol.
Psychiatry 54 (2003), 906–14.
37. C. Constant, L. Castera, R. Dantzer, et al., Mood alterations during interferon-alpha therapy
in chronic hepatitis C: evidence for an overlap between manic/hypomanic and depressive
symptoms. J. Clin. Psychiatry 66 (2005), 1050–57.
38. L. Capuron, A. Ravaud, P. J. Neveu, et al., Association between decreased serum tryptophan
concentrations and depressive symptoms in cancer patients undergoing cytokine therapy.
Mol. Psychiatry 7 (2002), 468–73.
39. M. Maes, S. Bonaccorso, V. Marino, et al., Treatment with interferon-alpha (IFN alpha) of
hepatitis C patients induces lower serum dipeptidyl peptidase IV activity, which is related to
IFN alpha-induced depressive and anxiety symptoms and immune activation. Mol. Psychiatry
6 (2001), 475–80.
40. J. Lestage, D. Verrier, K. Palin, R. Dantzer, The enzyme indoleamine 2,3-dioxygenase is
induced in the mouse brain in response to peripheral administration of lipopolysaccharide
and superantigen. Brain Behav. Immun. 16 (2002), 596–601.
41. F. D. Juengling, D. Ebert, O. Gut, et al., Prefrontal cortical hypometabolism during low-dose
interferon alpha treatment. Psychopharmacology (Berl.) 152 (2000), 383–9.
42. L. Capuron, G. Pagnoni, D. H. Lawson, et al., Altered fronto-pallidal activity during high-
dose interferon-alpha treatment as determined by positron emission tomography. Abstr. Soc.
Neurosci. (2002), 498.5.
279 Inflammation, sickness behaviour and depression
Introduction
Steroids are an extensive family of chemical agents distributed widely in the brain.
They include the classical stress hormone cortisol, oestradiol, testosterone and
progesterone (collectively known as the sex hormones), aldosterone and dehy-
droepiandrosterone (DHEA). Cortisol and DHEA are the most implicated in the
response to demand. Both have a high density in the limbic system but are also
found in the cortex. Circulating levels of steroids can be measured relatively easily
in the periphery from blood, urine and saliva. These peripheral levels are corre-
lated with levels in the cerebrospinal and ventricular fluid in the brain [1]. There
is clear-cut evidence that certain steroids are manufactured in the brain and play a
key role in brain development and plasticity [2]. These steroids include DHEA and
its sulphate DHEAS. Within the brain, these neurosteroids modulate the effects of
other transmitters, including gamma-aminobutyric acid (GABA) and glutamate.
Neurosteroids can, therefore, alter neuronal excitability throughout the brain very
rapidly by binding to receptors for inhibitory or excitatory neurotransmitters at the
cell membrane. Alterations in levels of the adrenal steroids cortisol and DHEA have
important implications for general cognitive and emotional function. These effects
are brought about through altered sensitivity in receptors in steroid-sensitive areas
of the brain, notably the limbic system, and their related frontal regions.
Hippocampus
Brain Hypothalamus
Feedback
ACTH system
Anterior Endorphins
pituitary
Pregnenolone
P450c17 21-OH
Adrenal gland
DHEA Cortisol
it is receptor sensitivity that is most likely to determine the impact of high steroid
levels on the tissues in the brain and the subsequent physiological and psychologi-
cal events. It may be that allelic variation leads to differences in receptor sensitivity
and therefore variations in response to corticoids. Thus, the genetic structure of
the receptor itself may contribute to modulating the risk for individual differ-
ences in behavioural response and psychiatric disorder via its sensitivity to cir-
culating cortisol levels. Finally, allelic variations of CRF may be associated with a
wide variety of behavioural characteristics, including the psychological process of
behavioural inhibition (a measure of impulse control) [13] and high body mass
index (BMI), indicating excess weight gain [14]. This indicates the complex and
manifold functions of large peptide molecules in functional activity within an
individual.
few hours of the day, followed by a decline over the second half of the day to a low
nadir from early evening. This diurnal rhythm remains in this form throughout
life.
Investigations of diurnal rhythm of cortisol in twins show that the early-morning
levels are significantly more alike in monozygotic twins than in dizygotic twins,
whereas evening levels show no such similarity [16]. Other studies have shown that
genetic influences on circulating cortisol persist over the 24-hour cycle and may
affect the timing of the nadir and the apex [17,18]. This suggests that there is marked
genetic control over the switching on of the axis in the morning, but environmental
factors exert increasing effects on levels as the day proceeds, although genetic factors
remain an influence [16].
As well as diurnal rhythm, levels show reactivity to events. These episodic move-
ments in levels generally show a rapid rise in the presence of a stimulus, regardless
of its salience, indicating that this change is related more to surprise or novelty
than to personal meaning. Cortisol levels can remain quite high in relation to the
behavioural requirements demanded by the stimulus, for example to engage in
social conversation with a surprise visitor or to deal with an unpleasant event such
as a car crash. Cortisol levels lower gradually as the consequences of the event
pass. In general, levels return to baseline about 40 minutes after the cessation of
the behavioural response. This rise and fall can occur at any time of the day and
appears to be independent of the stage of the diurnal rhythm that the individual is
at.
1. Salivary collections taken at 8 a.m. (or within one hour of awakening), 12 p.m., 4 p.m. and
8 p.m.
2. Samples collected over at least 4 consecutive days
3. Consider negative chemical implications of using salivary aids to salivation
4. Validated assays to a research standard (sensitivity, accuracy, specificity)
5. Sufficient power (sample size) to avoid type 2 errors
6. Control and comparison groups appropriate to the research hypothesis
7. Careful choice of derived measure for analysis (means, peaks, maximum levels, diurnal
variation, ratios, interquartile range, etc.)
8. Monitor potential confounds, e.g. steroid therapies including inhalers, prescribed and illicit
drugs
9. Record lifestyle factors, including exercise, dieting and smoking
10. Data analytical procedures should take into account sex differences and multiple or
repeated comparisons
For DHEA, a second major adrenal steroid, the picture is somewhat different.
This steroid is also present in the saliva (and CSF) at about 5% of plasma levels, but
there is no known plasma binding protein for DHEA, so the source of this relation
remains obscure. The diurnal rhythm in saliva (8 a.m. to 8 p.m.) is also much less
(about two-fold) than for cortisol. In adolescents, the mean ratio of salivary DHEA
from 8 a.m. to 8 p.m. was 2.07 (coefficient of variation = 33.8%). Interestingly,
levels vary with demand, being somewhat higher early in depressive illness but
becoming increasingly reduced under conditions of chronicity.
The cortisol/DHEA ratio thus changes during the day from about 5–7 in the
morning (8 a.m.) to about 2 at 8 p.m. This ratio is interesting in the light of
interactions between the two steroids (see below). From the above, a minimum
standard protocol for studies of cortisol and DHEA is proposed (Table 13.1).
established that morning cortisol hypersecretion precedes and predicts the onset of
major depression in both adult women and adolescents of both sexes [29,30]. These
studies measured morning and evening cortisol in adolescents of both sexes and
adult women, respectively. Both samples were at high psychosocial risk for unipolar
major depression. The participants were prospectively followed and re-evaluated
at 12 months. Higher morning cortisol levels were associated with the subsequent
onset of major depressive episodes over the follow-up period. In contrast, there
was no difference in evening cortisol levels between those who became depressed
and those who did not. The association between morning cortisol level and sub-
sequent affective disorder was not a consequence of concurrent recent undesirable
events or difficulties, and so it does not appear to be a reactivity effect to imme-
diate or recent negative social experiences. Neither was the hormone–depression
relationship a consequence of level of depressive symptoms at the time of cortisol
measurement. In the adolescent study there were no sex differences in the liability
for subsequent major depression. It appears that higher morning cortisol is a risk
factor for both sexes. The findings suggest that this hormonal vulnerability arises
from more distal origins rather than recent life events and difficulties. Whether
this is a consequence of environmental programming of the HPA axis in early
life, genetic factors influencing sensitivity to circulating cortisol, or both, remains
unknown.
A further community study failed to establish a direct association between higher
cortisol levels and depression in adult women but noted a strong link between corti-
sol levels and an increase in negative life events [31]. Depressed patients with higher
cortisol levels are significantly more likely to experience further negative life events
than depressed individuals with normal cortisol levels [32]. These depression-
dependent negative life events increase the liability for persistent disorder. Thus,
the pathological process that arises from high cortisol levels in currently depressed
patients is one that disturbs some aspects of affective-cognitive function that disrupt
interpersonal behaviour.
There is increasing evidence that cortisol hypersecretion is also correlated with
disturbances in memory [33,34]. A range of factors may influence this cortisol–
memory relationship, including social adversity and a previous history of cortisol
hypersecretion [35–38]. It seems highly likely that a key cortisol function in the brain
of healthy humans is to modulate learning and memory and, perhaps, retrieval of
information. This modulation process may have its key focus in the limbic system
but is likely to exert an influence on general brain state involving many regions,
including the prefrontal and orbitofrontal cortex [39,40].
Overall, the neurochemical coding for responses to social events and the subse-
quent liability for anxiety and/or depression may be as follows:
287 The hypothalamic–pituitary–adrenal axis
DHEA
DHEA shows a very different developmental history to that of cortisol [47]. Unlike
cortisol, concentrations of DHEA and its sulphate DHEAS vary with age [48].
DHEA is made by the placenta, and so the fetus is exposed to its action: concen-
trations decline from the first few months of life until 5 years of age and then rise
rapidly from age 7 in girls and around 9 in boys (this is called adrenarche), until
288 Depression and physical illness
levels reach their peak between the ages of 20 and 30 years. Adrenarche is sep-
arable from puberty, since gonadotrophins and oestrogen have no effect on DHEA
levels and the two events are not linked across time. After age 20–30 years, levels
begin to decline in both sexes. By the age of 70–80 years, levels are approximately
10–20% that of a 20-year-old [49]. Unlike cortisol, there is no clear-cut notion of
how DHEA is regulated. It is not under the tight control of the HPA axis, and so
although levels vary somewhat with the rise and fall of cortisol as the adrenal is
stimulated to secrete glucocorticoids, there are clearly other factors involved. DHEA
has been shown to act as an antagonist to cortisol at the level of the glucocorticoid
receptor [5]. There is also evidence that DHEA promotes neurogenesis [4]. This
ability to promote new neuronal growth could, if replicated, be a key neural feature
of this hormone. Since high cortisol levels persisting over days are associated with
an increase in depression, DHEA may have neuroprotective effects in the brain,
diminishing the liability for cortisol to damage neurons and thereby decreasing the
risk for psychopathology.
There is also preliminary evidence that increasing DHEA levels lower cortisol in
the periphery, providing further support for an antiglucocorticoid effect of DHEA
[50]. If this is the action of DHEA, then subjects at risk for psychiatric disorders and
showing high cortisol levels would be expected to show high levels of DHEA; this
was demonstrated in a prospective study of adolescents. Those who subsequently
developed major depression over the next 12 months had significantly higher levels
than expected for their age and sex [21,29]. The possibility that this rise is an
attempt to offset the increase in cortisol activity is attractive but requires further
investigation before firm conclusions can be drawn. For example, there was no
clear-cut evidence from the adolescent study that higher DHEA decreased cortisol
in some at-risk youths and thereby lowered the liability for subsequent depression.
What is clear is that DHEA not only is a passive indicator of a maturational effect of
age but also is involved in some active process. The finding requires replication, but
DHEA is a promising candidate as a neuroprotective agent at times of acute social
adversity.
There is some evidence that although DHEA levels rise during the early phase of
a major depression, they may fall if the disorder persists. Two longitudinal studies
of depressed patients, one from clinically referred patients and the other from
a community-based study, have shown that higher cortisol/DHEA ratios predict
persistent disorder and occur due to declining DHEA levels during the illness
[32,51]. These findings support the notion that DHEA secretion is vulnerable to
chronic illness effects and that its decline is associated with a poor short-term
outcome. The reasons for this decline in a potentially helpful neurosteroid are not
clear. It may reflect the severity of the metabolic strain that a severe mental illness
can produce. Similar findings have been noted in critically medically ill patients,
289 The hypothalamic–pituitary–adrenal axis
where low DHEA levels predict a poor response to treatment and higher mortality
[52,53].
A further difference from cortisol is the absence of any clear-cut associations
with psychological processes. DHEA does not appear to be related to memory,
learning or self-evaluation measures [54]. By contrast there is quite good evidence
that DHEA enhances positive mood and may act as an antidepressant [55–58].
Interestingly, DHEA does not appear to exert effects via testosterone, even though
it is in the same metabolic pathway of androgen production. Rather, DHEA appears
to have direct effects on the brain, perhaps via its antiglucocorticoid actions in the
amygdala and hippocampus, thereby showing a bias for modulating neuroaffective
rather than neurocognitive systems. Although definitive studies of sufficient sample
size are required to confirm these observations, the evidence that DHEA has mood-
enhancing effects is compelling. Following the discovery of an endothelial receptor
for DHEA, it may not be too long now before we are able to describe the physiology
of this compound. From the psychiatric perspective, DHEA is potentially an impor-
tant modulator of disordered mood states and deserves to be studied in detail as
an adjunct to current treatments, particularly in persistent mood disorders. Much
greater precision is required in investigating cognitive–DHEA relations before con-
cluding that there is no association between memory, learning and information
retrieval and individual differences in DHEA.
Infant stress
Research has implicated a role for infant exposure to adverse early experiences
in the formation of HPA axis sensitivity. Animal studies have reported a sub-
stantial non-genetic effect of adverse maternal rearing practices on the develop-
ment of chemical coding systems for behaviour, including HPA axis, hypotha-
lamic and extra-hypothalamic corticotropin-releasing hormone, monoaminergic
and GABA/benzodiazepine systems. Loss of maternal care through separation leads
to a potential change in the chemical signalling processes between the limbic system
and the frontal and prefrontal cortex. Findings in rodents have outlined the neuro-
chemical mechanisms that occur via epigenetic programming as a consequence of
variations in maternal care style [59]. The evidence shows that there are major effects
on glucocorticoid receptor sensitivity in rodent pups dependent on the degree of
positive care obtained. This early environmental effect can be lifelong, with adverse
effects on behaviour persisting through to adult rodent life. The pathological mech-
anisms result in significantly fewer steroid receptors in the limbic system and lower
expression of neurotrophic genes involved in brain growth and function. These dys-
functions occur because in pups exposed to adverse rearing, there is methylation
290 Depression and physical illness
not clear whether low levels of cortisol, suggesting a suppressed HPA axis, are related
to a history of adversity, to a particular form of adversity or even to no adversity in
this group of behavioural disorders. Indeed, about 10% of the population at large
may have flat cortisol levels over the 24-hour period rather than the more common
diurnal variation [67]. The implications of a flat cortisol level in the population at
large, and the finding that this occurs in some individuals with markedly different
disorders, such as PTSD and severe conduct disorder, is a puzzle that requires some
considerable sorting out.
The presence of extreme levels (high or low) of cortisol suggests a loss of syn-
chrony or perhaps a blunting between the neurochemical signalling pathways within
the brain. For example, experimental studies on conduct-disordered children that
induced frustration showed no increase in cortisol levels relative to age-matched
controls, who exhibited the predicted rise at the time of stress stimulus [68]. Inter-
estingly, although cortisol levels remained flat during this emotionally charged chal-
lenge test, heart rate also remained flat, although the conduct-disordered children
reported feeling out of control and angry [69]. These findings showing dissoci-
ation between affective-cognitive and physiological responses following induced
frustration suggest a potential loss of synchrony at the neurochemical level.
Repeating these studies using neuroimaging techniques may allow us to test
whether conduct-disordered individuals do indeed show a different pattern of
neural response to that of controls. Lower heart rate in childhood is associ-
ated with higher risk for antisocial personality disorder in adult life [70], and
behaviourally disordered individuals report fearlessness, even when confronted
with fearful stimuli [71,72]. There are also reported reductions in the grey matter
of the brain in adults diagnosed with psychopathic disorders [73]. All of these
findings support a brain-based aetiology to explain the fact that responses to
social adversity are different in behaviourally disordered individuals over the life-
span. There are also changes in the serotonin system in severe psychopaths, who
show low serotonin function correlated with increased impulsivity compared with
controls [74].
What does this impaired serotonin and increased impulsivity response have to do
with low or flat cortisol levels at times of stress? Cortisol is part of the chemical cod-
ing pathway that accesses personally salient emotional related memories (episodic
memories) [35]. Low sensitivity to fearfulness may impair the mobilisation of a
fear response through loss of retrieval of fear-related memories. Such memories
may not even be kept in memory by behaviour-disordered individuals. This will
blunt any signalling processes to cognitive centres in the cortex, and serotonin-
vulnerable individuals will be at risk for disinhibited, impulsive non-socially adap-
tive behavioural responses following adverse experiences. Low or flat cortisol levels
during challenge may indicate a defective affectively driven information-processing
pathway.
293 The hypothalamic–pituitary–adrenal axis
Acknowledgements
The author is supported by the Wellcome Trust. He is a member of the Medi-
cal Research Council (MRC) Centre for Brain, Behaviour and Neuropsychiatry,
Cambridge University.
REFERENCES
37. S. J. Lupien, C. W. Wilkinson, S. Briere, et al., Acute modulation of aged human memory
by pharmacological manipulation of glucocorticoids. J. Clin. Endocrinol. Metab. 87 (2002),
3798–807.
38. S. J. Lupien, S. King, M. J. Meaney, B. S. McEwen, Child’s stress hormone levels correlate with
mother’s socioeconomic status and depressive state. Biol. Psychiatry 48 (2000), 976–80.
39. W. C. Drevets, Neuroimaging abnormalities in the amygdala in mood disorders. Ann. N. Y.
Acad. Sci. 985 (2003), 420–44.
40. D. A. Seminowicz, H. S. Mayberg, A. R. McIntosh, et al., Limbic-frontal circuitry in major
depression: a path modeling metanalysis. Neuroimage 22 (2004), 409–18.
41. M. Liotti, H. S. Mayberg, S. K. Brannan, et al., Differential limbic–cortical correlates of
sadness and anxiety in healthy subjects: implication for affective disorders. Biol. Psychiatry
48 (2000), 30–42.
42. J. H. Meyer, S. Kapur, B. Eisfeld, et al., The effect of paroxetine on 5-HT(2A) receptors in
depression: an [(18)F]setoperone PET imaging study. Am. J. Psychiatry 158 (2001), 78–85.
43. S. Wust, E. F. Van Rossum, I. S. Federenko, et al., Common polymorphisms in the gluco-
corticoid receptor gene are associated with adrenocortical responses to psychosocial stress.
J. Clin. Endocrinol. Metab. 89 (2004), 565–73.
44. A. Stevens, D. W. Ray, E. Zeggini, et al., Glucocorticoid sensitivity is determined by a specific
glucocorticoid receptor haplotype. J. Clin. Endocrinol. Metab. 89 (2004), 892–7.
45. E. F. Van Rossum, P. G. Voorhoeve, S. J. te Velde, et al., The ER22/23EK polymorphism in the
glucocorticoid receptor gene is associated with a beneficial body composition and muscle
strength in young adults. J. Clin. Endocrinol. Metab. 89 (2004), 4004–4009.
46. E. F. van Rossum, J. W. Koper, A. W. van den Beld, et al., Identification of the BclI polymor-
phism in the glucocorticoid receptor gene: association with sensitivity to glucocorticoids in
vivo and body mass index. Clin. Endocrinol. (Oxf.) 59 (2003), 585–92.
47. P. D. Kroboth, F. S. Salek, A. L. Pittenger, T. J. Fabian, R. F. Frye, DHEA and DHEA-S: a
review. J. Clin. Pharmacol. 39 (1999), 327–48.
48. C. R. Parker, Dehydroepiandrosterone and dehydroepiandrosterone sulfate production in
the human adrenal during development and aging. Steroids 64 (1999), 640–47.
49. F. Labrie, A. Belanger, L. Cusan, J. L. Gomez, B. Candas, Marked decline in serum concen-
trations of adrenal C19 sex steroid precursors and conjugated androgen metabolites during
aging. J. Clin. Endocrinol. Metab. 82 (1997), 2396–402.
50. P. D. Kroboth, J. A. Amico, R. A. Stone, et al., Influence of DHEA administration on 24-hour
cortisol concentrations. J. Clin. Psychopharmacol. 23 (2003), 96–9.
51. I. M. Goodyer, R. J. Park, J. Herbert, Psychosocial and endocrine features of chronic first-
episode major depression in 8–16 year olds. Biol. Psychiatry 50 (2001), 351–7.
52. A. Beishuizen, L. G. Thijs, I. Vermes, Decreased levels of dehydroepiandrosterone sulphate
in severe critical illness: a sign of exhausted adrenal reserve? Crit. Care 6 (2002), 434–8.
53. C. Marx, S. Petros, S. R. Bornstein, et al., Adrenocortical hormones in survivors
and nonsurvivors of severe sepsis: diverse time course of dehydroepiandrosterone,
dehydroepiandrosterone-sulfate, and cortisol. Crit. Care Med. 31 (2003), 1382–8.
54. F. A. Huppert, J. K. Van Niekerk, Dehydroepiandrosterone (DHEA) supplementation for
cognitive function. Cochrane Database Syst. Rev. 2 (2001), CD000304.
297 The hypothalamic–pituitary–adrenal axis
55. P. J. Hunt, E. M. Gurnell, F. A. Huppert, et al., Improvement in mood and fatigue after
dehydroepiandrosterone replacement in Addison’s disease in a randomized, double blind
trial. J. Clin. Endocrinol. Metab. 85 (2000), 4650–56.
56. J. K. van Niekerk, F. A. Huppert, J. Herbert, Salivary cortisol and DHEA: association with
measures of cognition and well-being in normal older men, and effects of three months of
DHEA supplementation. Psychoneuroendocrinology 26 (2001), 591–612.
57. O. M. Wolkowitz, V. I. Reus, A. Keebler, et al., Double-blind treatment of major depression
with dehydroepiandrosterone. Am. J. Psychiatry 156 (1999), 646–9.
58. R. D. Strous, R. Maayan, R. Lapidus, et al., Dehydroepiandrosterone augmentation in the
management of negative, depressive, and anxiety symptoms in schizophrenia. Arch. Gen.
Psychiatry 60 (2003), 133–41.
59. I. C. Weaver, N. Cervoni, F. A. Champagne, et al., Epigenetic programming by maternal
behavior. Nat. Neurosci. 7 (2004), 847–54.
60. S. L. Halligan, J. Herbert, I. M. Goodyer, L. Murray, Exposure to postnatal depression predicts
elevated cortisol in adolescent offspring. Biol. Psychiatry 55 (2004), 376–81.
61. M. Champoux, A. Bennett, C. Shannon, et al., Serotonin transporter gene polymorphism,
differential early rearing, and behavior in rhesus monkey neonates. Mol. Psychiatry 7 (2002),
1058–63.
62. A. Caspi, K. Sugden, T. E. Moffitt, et al., Influence of life stress on depression: moderation by
a polymorphism in the 5-HTT gene. Science 301 (2003), 386–9.
63. M. R. Gunnar, S. J. Morison, K. Chisholm, M. Schuder, Salivary cortisol levels in children
adopted from Romanian orphanages. Dev. Psychopathol. 13 (2001), 611–28.
64. R. Yehuda, S. L. Halligan, J. A. Golier, R. Grossman, L. M. Bierer, Effects of trauma exposure
on the cortisol response to dexamethasone administration in PTSD and major depressive
disorder. Psychoneuroendocrinology 29 (2004), 389–404.
65. R. Loeber, S. M. Green, B. B. Lahey, P. J. Frick, K. McBurnett, Findings on disruptive behavior
disorders from the first decade of the Developmental Trends Study. Clin. Child. Fam. Psychol.
Rev. 3 (2000), 37–60.
66. K. Pajer, W. Gardner, R. T. Rubin, J. Perel, S. Neal, Decreased cortisol levels in adolescent girls
with conduct disorder. Arch. Gen. Psychiatry 58 (2001), 297–302.
67. A. A. Stone, J. E. Schwartz, J. Smyth, et al., Individual differences in the diurnal cycle of salivary
free cortisol: a replication of flattened cycles for some individuals. Psychoneuroendocrinology
26 (2001), 295–306.
68. S. H. van Goozen, W. Matthys, P. T. Cohen-Kettenis, et al., Salivary cortisol and cardiovas-
cular activity during stress in oppositional-defiant disorder boys and normal controls. Biol.
Psychiatry 43 (1998), 531–9.
69. S. H. van Goozen, W. Matthys, P. T. Cohen-Kettenis, J. K. Buitelaar, H. van Engeland,
Hypothalamic-pituitary-adrenal axis and autonomic nervous system activity in disruptive
children and matched controls. J. Am. Acad. Child. Adolesc. Psychiatry 39 (2000), 1438–
45.
70. A. Raine, P. H. Venables, S. A. Mednick, Low resting heart rate at age 3 years predisposes to
aggression at age 11 years: evidence from the Mauritius Child Health Project. J. Am. Acad.
Child. Adolesc. Psychiatry 36 (1997), 1457–64.
298 Depression and physical illness
Introduction
Depression has a huge impact on individuals and society, with a lifetime prevalence
of over 15%. By 2020, depression will be the second leading illness in the world,
as projected by the World Health Organization (WHO). In addition to the emo-
tional consequences of depression, the disorder is increasingly implicated in a wide
range of medical conditions. Moreover, a growing body of evidence indicates that
depression, including even minor depression, has notable immunological conse-
quences. It is important therefore to consider possible immune mechanisms in the
detrimental effects of depression on health, particularly in vulnerable individuals
such as elderly people and patients with chronic disease.
This chapter provides a review of the research being conducted on the relation-
ship between depression and immunity, beginning with an overview of the clinical
importance of depressive disorders for mortality risk. The various immune alter-
ations that occur during depression are examined, with discussion of the role of
autonomic, neuroendocrine and behavioural (e.g., sleep) mechanisms. The chap-
ter concludes with consideration of the clinical implications of immune changes in
depression for several medical disorders.
Definition of depression
Depression can be defined as a condition that primarily entails a disturbance of
mood; this affective disturbance is often characterised by a mood that is sad, hope-
less, discouraged or simply depressed [1]. The diagnostic definition of a major
depressive disorder states that the depressed mood or loss of pleasure (i.e. anhe-
donia) must last for at least two weeks, be accompanied by a series of related
Depression and Physical Illness, ed. A. Steptoe.
Published by Cambridge University Press. C Cambridge University Press 2006.
300 Depression and physical illness
symptoms (e.g. vegetative signs, sleep disturbance) and should not be caused by
another psychiatric disorder, general medical condition or substance use/abuse.
The symptoms typically experienced during depression include significant weight
change (+/−5%), sleep disturbance (insomnia, hypersomnia), psychomotor retar-
dation or agitation, pervasive fatigue, feelings of worthlessness or irrational guilt,
mental concentration difficulties and recurrent suicidal ideation.
Although most studies of immune abnormalities have focused on major depres-
sion, there appears to be a dose–response relationship between severity of depres-
sion and changes of immunity. Thus, several studies have examined the effects of
minor depression on immunity, as discussed below. While minor depression is
noted as a potential diagnostic category that needs further empirical validation [2],
it is essentially characterised by a disturbance in mood or a loss of pleasure, with
the presence of two additional symptoms (rather than four for major depressive
disorder). Minor depression is also associated with functional impairments and is
often viewed as a prodromal state for the occurrence of major depression [2].
Clinical course
The clinical course of major depressive disorder varies markedly with age: its onset
is protracted in older adults, with prodromal and progressive elevation of depressive
symptoms, in contrast to the relatively short clinical onset in younger patients [3].
Furthermore, in older adults depressive symptoms and stress of physical disability
can initiate a continuous decline in physical and psychological health. Even with
ongoing clinical management, there are often lingering depressive symptoms and
an increased risk of relapse in older adults who are experiencing life stress and
chronic disease burden [4]. Together, these data are consistent with the view of
depressive disorders as chronic and recurrent with residual disability [5].
data show that behavioural changes that are commonly found in depression, such
as disordered sleep, also have independent effects on mortality risk. For example,
in older adults, difficulties with sleep initiation and/or maintenance prospectively
predicted increased risk of death after controlling for age, gender and medical status
[12–15].
pathogen. Additional studies indicate that sympathetic activation can also shunt
some immune system cells out of circulating blood and into the lymphoid organs
(e.g. spleen, lymph nodes, thymus) while recruiting other types of immune cell
into circulation (e.g. NK cells). It is thought that sympathetic activation reduces the
immune system’s ability to destroy pathogens that live inside cells (e.g. viruses) by
decreasing of the cellular immune response, while enhancing the humoral immune
response to pathogens that live outside cells (e.g. bacteria) [22].
Neuroendocrine axis
A hallmark of major depression is dysregulation of the hypothalamic-pituitary-
adrenal (HPA) axis and the overexpression of cortisol. Cortisol exerts diverse effects
on a wide variety of physiological systems and also coordinates the actions of vari-
ous cells involved in an immune response by altering the production of cytokines or
immune messengers [23]. Similar to sympathetic activation, cortisol can suppress
the cellular immune response critical to defending the body against viral infec-
tions. Cortisol can also prompt some immune cells to move out from circulating
blood into lymphoid organs or peripheral tissues such as the skin [24]. The role of
HPA dysregulation and abnormal glucocorticoid sensitivity in relation to immune
abnormalities in depression is discussed below.
Enumerative measures
One of the first immunological findings identified in depressed individuals was
an increase in the total number of white blood cells and an increase in the num-
bers and percentages of neutrophils and lymphocytes [25–28]. Inconsistent find-
ings regarding monocyte counts have been garnered, with increases found for
depressed individuals in some studies [27] and other depression samples show-
ing decreases or no differences in the absolute or relative numbers of monocytes
[25,29]. One population-based cohort study (n = 11 367) revealed an association
between recency of depression history and an incremental elevation of white blood
cell counts in men but not in women, even with adjustment for age and smoking
303 Depression and immunity: biological and behavioural mechanisms
status [30]. Similar to the findings in adults, depressed adolescents show increased
numbers of leukocytes, lymphocytes and monocytes [31].
Cellular enumeration of lymphocyte subsets by the quantification of phenotypic
specific cell-surface markers has shown that depression is related negatively to the
number and percentages of lymphocytes that are B-cells, T-cells, T-helper cells and
T-suppressor/cytotoxic cells [32]. A decrease in circulating number of cells that
express the NK cell phenotype has also been reported, which in part is moderated
by gender: a decline of NK cell numbers is found in male but not female depressed
individuals as compared with gender-matched controls [33]. In depressed adol-
escents, there is also evidence of a reduced number of NK cells [31]. However,
multiple discrepant findings have been reported and, in one of the largest study
samples of depressed subjects, no difference in the number of peripheral blood
lymphocytes or T-lymphocyte subsets was found between depressed patients and
controls [34]. Indeed, with the recent accumulation of studies, it is questionable as
to whether there are consistent changes in the number of circulating B-, T- or NK
cells in depression [35].
Functional measures
Function of the immune system has typically been evaluated in depressed individ-
uals by assay of non-specific mitogen-induced lymphocyte proliferation, mitogen-
stimulated cytokine production and NK cytotoxicity. Some of the first observa-
tions evaluating depression and mitogen responses showed reduced proliferation
in depressed individuals compared with controls [36,37]. However, subsequent
studies failed to replicate these observations, raising questions about the reli-
ability of this immune alteration in depression [34]. Nevertheless, with over a
dozen studies that have examined lymphocyte proliferation in depressed adults
adolescents [34,38–52), it appears that an impairment in the response of lym-
phocytes to all three non-specific mitogens predominates in studies of depressed
subjects [35].
A reduction of NK activity is now considered to be one of the most reliable
and reproducible alterations of ex vivo immune function in depression [35,53].
Irwin and colleagues [29] first reported a decline of NK activity in depressed
patients compared with age- and gender-matched comparison controls, and ten
subsequent independent samples have replicated this observation [54,55]. Several
caveats must be considered when reviewing the associations between NK activity,
NK numbers and depression. Compared with controls, decreases in NK counts and
NK activity were limited to male but not female depressed individuals [33], and
other studies have found no difference of NK activity in depression [34,56]. One
study of depressed adolescents found an increase of NK activity [31]. Nevertheless,
the meta-analyses of Zorrilla et al. [35] and Herbert and Cohen [32] found that
304 Depression and physical illness
adult populations with depression may be at greater risk for abnormal elevations
of pro-inflammatory cytokines. It has been speculated that immune activation is a
consequence of depression, although the acute induction of depression by trypto-
phan depletion of central monoamines failed to alter plasma levels of IL-6 or TNF
receptor, and decreases of the Th2 cytokine IL-4 were found [68].
As discussed below, immune activation and the expression of inflammatory
cytokines are thought to mediate the link between depression and inflammatory
disorders, including cardiovascular disease [69] and rheumatoid arthritis [70]. In
addition, pro-inflammatory cytokines are hypothesised to contribute to the occur-
rence of depressive symptoms in certain medical populations [71], although review
of this topic extends beyond the scope of this chapter.
Treatment of depression
Castanon et al. [86] reviewed animal research that focused on the immuno-
logical effects of antidepressant medications and concluded that antidepressants
decrease pro-inflammatory cytokine expression. Although the specific mecha-
nism(s) involved were not identified, antidepressants modify the expression of
glucocorticoids and their receptors [87,88], limit the synthesis of prostaglandin and
nitric oxide (NO), which contribute to pro-inflammatory effects [89,90], and act on
intracellular messenger pathways, with downstream effects on pro-inflammatory
cytokines [91]. However, there are also contradictory findings: in humans, in vitro
co-incubation of stimulated immune cells with imipramine, venlafaxine or fluox-
etine increased the production of IL-6 but did not affect TNF [92].
Several studies have now investigated the in vivo effects of antidepressant medica-
tion on immunity by examining the clinical course of depression along with changes
of immunity in relation to antidepressant medication treatment and symptom reso-
lution. In one longitudinal case–control study, Irwin and co-workers [93] found
that depressed patients showed a decrease of Hamilton depression scores and an
increase in NK activity from the start of treatment to follow-up. Changes of NK
activity were related to changes in depressive symptom severity but not treatment
status with antidepressant medications. In another longitudinal follow-up study
of young adults with unipolar depression involving six weeks of treatment with
nortriptyline and alprazolam, clinical improvements in the severity of depressive
307 Depression and immunity: biological and behavioural mechanisms
Biological mediators
Depression alters two pathways that have effects on immune system functioning:
sympathetic nervous system activity and HPA axis function. Some data are now
available to indicate that abnormalities in these efferent mechanisms contribute to
immune changes in depression.
Considerable data show that depression is associated with abnormal elevations
in sympathetic tone, as measured by increases in resting levels and exaggerated
responses to psychological stress. For example, Irwin et al. [21] found that plasma
concentrations of neuropeptide Y are elevated in depressed patients, aged people
and people undergoing severe Alzheimer’s disease caregiver stress. In support of the
hypothesis that elevated sympathetic activity is associated with immune alterations
in depression, further study found that activation of the sympathetic nervous system
and release of neuropeptide Y correlated with a reduction of natural cytotoxicity
in depression and life stress. Other studies show that excretion of 3-methoxy-4-
hydroxy-phenylglycol (MHPG), an index of total body noradrenergic turnover, or
sympathetic activity, is related inversely to lymphocyte proliferative responses in
depressed patients [48].
An important function of the HPA axis is to shape-immune responses, par-
ticularly inflammatory changes such as increases in the production of acute-phase
reactants, elevations in circulating levels of pro-inflammatory cytokines (e.g. IL-6 in
plasma) and increases in stimulated in vitro peripheral blood monocyte production
of proinflammatory cytokines [98]. For example, dexamethsone non-suppression
identifies a group of depressed individuals most likely to demonstrate immune
activation in which increased concentrations of plasma IL-6 correlate positively
with post-DST cortisol levels [99]. Likewise, in a study of 28 in-patients with
308 Depression and physical illness
major depression, Maes and colleagues [100] found a significant positive corre-
lation between mitogen-induced IL-1β production and post-DST cortisol values,
which suggested that IL-1β oversecretion may contribute to HPA axis overactivity
in depression. Finally, in treatment-resistant depression, cells of non-suppressors
produced significantly less TNF compared with suppressors [58].
Unlike the association between markers of immune activation and HPA dysreg-
ulation, several studies of patients with major depression have failed to show a rela-
tionship between altered T-cell responses or NK cell activity and plasma or urinary
concentrations of free cortisol, despite the recognised action of glucocorticoids
to suppress multiple aspects of lymphocyte function. For example, in depressed
patients, decreased lymphocyte responses to mitogens are not associated with dexa-
methasone non-suppression [36] or with increased excretion rates of urinary free
cortisol [41]. Furthermore, in bereavement complicated by depressive symptoms,
changes of NK activity are not associated with plasma cortisol levels [101]. Raison
and Miller [102] have suggested that the impairments in T-cell proliferation and
NK cell activity frequently reported in major depression might be accounted for
by either excessive or insufficient glucocorticoid signalling, which is due in part to
the action of pro-inflammatory cytokines to disrupt T-cell signalling pathways and
inhibit NK cell activity. One study to our knowledge has shown that low NK cell
activity is not associated with increases of circulating levels of IL-6 in the context
of major depression [65].
Behavioural mechanisms
In addition to the biological mediators of immune changes in depression, exam-
ination of health status and behavioural factors is needed in clinical psychoneu-
roimmunology. For example, alcohol and tobacco have well-recognised effects on
immunity, and yet there is limited empirical information on the processes by which
these substances alter immune function in depressed individuals, despite their high
rates of use in depressed patients. In the following section, four of the more perti-
nent behavioural factors linking depression and immune dysfunction are discussed,
namely tobacco smoking, alcohol/substance abuse, activity/exercise and sleep
disturbance.
history, and together these two factors produce greater declines of NK activity than
those found in depressed or smoking groups alone. Cigarette use alone, or in com-
bination with depression, might also contribute to the suppression of other non-
specific measures of immune function, such as mitogen-induced lymphocyte pro-
liferation. Given the effects of cigarette smoking on markers of immune activation
[107], it is also important to address whether smoking status alters the reported rela-
tionship between depression and increases in serum levels of IL-6 and acute-phase
proteins. Pike and Irwin [65] showed that elevations of IL-6 are found predomi-
nantly in depressed smokers, but not in depressed non-smokers, although others
have suggested that depression has an independent effect on circulating levels of IL-6
and CRP [64,69]. Importantly, depression appears to interact with cigarette smok-
ing to impact health, rather than there being a unitary link between depression and
cancer. In a 12-year follow-up of 2264 adult men and women, depressed mood was
found to interact with cigarette smoking, and together depressed mood and cigarette
smoking were associated with a marked increase in the relative risk of cancer [108]
as compared with the risk associated with smoking or depression status alone.
been shown to have potent salutary effects on immune measures and has even
been found to promote a remission of depressive symptoms in older adults. In the
meta-analysis of Herbert and Cohen [32], melancholic depression correlated with
greater impairments of cellular immunity, which may be due, at least in part, to an
increased predominance of neurovegetative symptoms [54]. Cover and Irwin [85]
found that severity of psychomotor retardation uniquely predicted declines of NK
activity, similar to the effects of insomnia.
sleep may be a critical behavioural factor that mediates the relationship between
depression and immune alterations.
Cardiovascular disease
Atherosclerosis is now thought to be an inflammatory process that involves a series
of steps [119], each of which appears to be impacted by stress and/or depression.
Activated macrophages within the vasculature secrete pro-inflammatory cytokines,
which in turn leads to expression of adhesion molecules. With recruitment of
immune cells to the vascular cell wall or endothelium, and the release of inflam-
matory cytokines, the vascular endothelium expresses adhesion molecules that
facilitate further binding of immune cells. Importantly, psychological and physi-
cal stressors increase the release of pro-inflammatory cytokines and expression of
adhesion molecules that tether (‘slow down’) and bind immune cells to the vas-
cular endothelium [120,121]. Moreover, it appears that depression is associated
with endothelium activation. Acute coronary patients who are depressed show an
increased expression of an adhesion molecule that is released following activation
of the vascular endothelium (i.e. soluble intracellular adhesion molecule) [69].
Importantly, this molecular marker of endothelial activation and IL-6 predict risk
of future myocardial infarction.
(including the stress of HIV infection itself) have all been shown to predict the rate
of immune-system decay in patients with HIV. Immune-system decline and HIV
replication are particularly rapid in patients living under chronic stress (e.g. gay
men who conceal their homosexuality) and in patients with high levels of sympa-
thetic nervous system (SNS) activity (e.g. socially inhibited introverts) [124,125].
Tissue culture studies have shown that SNS neurotransmitters and glucocorticoids
can accelerate HIV replication by rendering T-lymphocytes more vulnerable to
infection and by suppressing production of the antiviral cytokines that help cells
to limit viral replication [126]. Current research is focusing on pharmacological
strategies to block the effects of stress neuroendocrine hormones on chronic viral
infections such as HIV.
Summary
Substantial evidence generated over two decades of research has found that depres-
sion is related to many aspects of immune function, including enumerative changes,
313 Depression and immunity: biological and behavioural mechanisms
Acknowledgements
This work was supported in part by grants AA13239, DA16541, MH55253,
AG18367, T32-MH19925, AR/AG41867, 1 R01 AR 49840–01 and M01-RR00865
from the General Clinical Research Centers Program and the Cousins Center for
Psychoneuroimmunology.
REFERENCES
27. M. Maes, M. Van der Planken, W. J. Stevens, et al., Leukocytosis, monocytosis, and
neutrophilia: hallmarks of severe depression. J. Psychiatr. Res. 26 (1992), 125–34.
28. Z. Kronfol, R. Turner, H. Nasrallah, G. Winokur, Leukocyte regulation in depression and
schizophrenia. Psychiatry Res. 13 (1984), 13–18.
29. M. Irwin, T. L. Smith, J. C. Gillin, Low natural killer cytotoxicity in major depression. Life
Sci. 41 (1987), 2127–33.
30. P. Surtees, N. Wainwright, N. Day, et al., Association of depression with peripheral leukocyte
counts in EPIC-Norfolk-role of sex and cigarette smoking. J. Psychsom. Res. 54 (2003), 303–
6.
31. S. J. Schleifer, J. A. Bartlett, S. E. Keller, et al., Immunity in adolescents with major depression.
J. Am. Acad. Child Adolesc. Psychiatry 41 (2002), 1054–60.
32. T. B. Herbert, S. Cohen, Depression and immunity: a meta-analytic review. Psychol. Bull.
113 (1993), 472–86.
33. D. L. Evans, J. D. Folds, J. M. Petitto, et al., Circulating natural killer cell phenotypes in men
and women with major depression. Arch. Gen. Psychiatry 49 (1992), 388–95.
34. S. J. Schleifer, S. E. Keller, R. N. Bond, J. Cohen, M. Stein, Major depressive disorder and
immunity: role of age, sex, severity, and hospitalization. Arch. Gen. Psychiatry 46 (1989),
81–7.
35. E. P. Zorrilla, L. Luborsky, J. R. McKay, et al., The relationship of depression and stressors
to immunological assays: a meta-analytic review. Brain Behav. Immun. 15 (2001), 199–226.
36. Z. Kronfol, J. D. House, Depression, hypothalamic-pituitary adrenocortical activity and
lymphocyte function. Psychopharmacol. Bull. 21 (1985), 476–8.
37. Z. Kronfol, J. Silva, J. Greden, et al., Impaired lymphocyte function in depressive illness. Life
Sci. 33 (1983), 241–7.
38. J. Albrecht, J. Helderman, M. Schlesser, J. Rush, A controlled study of cellular immune
function in affective disorders before and during somatic therapy. Psychiatry Res. 15 (1985),
185–93.
39. E. Syvalahti, J. Eskola, O. Ruuskanen, T. Laine, Nonsuppression of cortisol and immune
function in depression. Prog. Neuropsychopharmacol. Biol. Psychiatry 9 (1985), 413–22.
40. S. J. Schleifer, S. E. Keller, S. G. Siris, K. L. Davis, M. Stein, Depression and immunity:
lymphocyte function in ambulatory depressed patients, hospitalized schizophrenic patients,
and patients hospitalized for herniorrhaphy. Arch. Gen. Psychiatry 42 (1985), 129–33.
41. Z. Kronfol, J. D. House, J. Silva, J. Greden, B. J. Carroll, Depression, urinary free cortisol
excretion, and lymphocyte function. Br. J. Psychiatry 148 (1986), 70–73.
42. J. R. Calabrese, R. G. Skwerer, B. Barna, et al., Depression, immunocompetence, and
prostaglandins of the E-series. Psychol. Res. 17 (1986), 41–7.
43. M. T. Lowy, A. T. Reder, G. J. Gormley, H. Y. Meltzer, Comparison of in vivo and in vitro
glucocorticoid sensitivity in depression: relationship to the dexamethasone suppression test.
Biol. Psychiatry 24 (1988), 619–30.
44. M. Maes, E. Bosmans, E. Suy, B. Minner, J. Raus, Impaired lymphocyte stimulation by
mitogens in severely depressed patients: a complex interface with HPA axis hyperfunction,
noradrenergic activity and the ageing process. Br. J. Psychiatry 155 (1989), 793–8.
45. L. L. Altshuler, S. Plaeger-Marshall, S. Richeimer, M. Daniels, L. R. Baxter, Jr, Lymphocyte
function in major depression. Acta Psychiatr. Scand. 80 (1989), 132–6.
316 Depression and physical illness
63. D. L. Musselman, A. H. Miller, M. R. Porter, et al., Higher than normal plasma interleukin-6
concentrations in cancer patients with depression: preliminary findings. Am. J. Psychiatry
158 (2001), 1252–7.
64. G. E. Miller, C. A. Stetler, R. M. Carney, K. E. Freedland, W. A. Banks, Clinical depression
and inflammatory risk markers for coronary heart disease. Am. J. Cardiol. 90 (2002), 1279–
83.
65. J. L. Pike, M. R. Irwin, Dissociation of inflammatory markers and natural killer cell activity
in major depressive disorder. Brain Behav. Immun. 20 (2006), 169–74.
66. B. W. Penninx, S. B. Kritchevsky, K. Yaffe, et al., Inflammatory markers and depressed mood
in older persons: results from the health, aging and body composition study. Biol. Psychiatry
54 (2003), 566–72.
67. F. Brambilla, P. Monteleone, M. Maj, Interleukin-1beta and tumor necrosis factor-alpha in
children with major depressive disorder or dysthymia. J. Affect. Disord. 78 (2004), 273–7.
68. J. Stastny, A. Konstantinidis, M. J. Schwarz, et al., Effects of tryptophan depletion and
catecholamine depletion on immune parameters in patients with seasonal affective disorder
in remission with light therapy. Biol. Psychiatry 53 (2003), 332–7.
69. F. Lesperance, N. Frasure-Smith, P. Theroux, M. Irwin, The association between major
depression and levels of soluble intercellular adhesion molecule 1, interleukin-6, and C-
reactive protein in patients with recent acute coronary syndromes. Am. J. Psychiatry 161
(2004), 271–7.
70. A. J. Zautra, D. C. Yocum, I. Villanueva, et al., Immune activation and depression in women
with rheumatoid arthritis. J. Rheumatol. 31 (2004) 457–63.
71. A. H. Miller, Cytokines and sickness behavior: implications for cancer care and control.
Brain Behav. Immun. 17: Suppl 1 (2003), 132–4.
72. G. E. Miller, S. Cohen, S. Pressman, et al., Psychological stress and antibody response to
influenza vaccination: when is the critical period for stress, and how does it get inside the
body? Psychosom. Med. 66 (2004), 215–23.
73. M. Irwin, C. Costlow, H. Williams, et al., Cellular immunity to varicella-zoster virus in
patients with major depression. J. Infect. Dis. 178 (1998), 104–8.
74. F. S. Dhabhar, Acute stress enhances while chronic stress suppresses skin immunity: the
role of stress hormones and leukocyte trafficking. Ann. N. Y. Acad. Sci. 917 (2000), 876–
93.
75. I. Hickie, C. Hickie, A. Lloyd, D. Silove, D. Wakefield, Impaired in vivo immune responses
in patients with melancholia. Br. J. Psychiatry 162 (1993), 651–7.
76. M. Shinkawa, K. Nakayama, H. Hirai, M. Monma, H. Sasaki, Depression and immunore-
activity in disabled older patients. J. Am. Geriatr. Soc. 50 (2002), 198–9.
77. R. Glaser, J. K. Kiecolt-Glaser, R. H. Bonneau, et al., Stress-induced modulation of the
immune response to recombinant hepatitis B vaccine. Psychosom. Med. 54 (1992), 22–9.
78. J. K. Kiecolt-Glaser, R. Glaser, S. Gravenstein, W. B. Malarkey, Chronic stress alters the
immune response to influenza virus vaccine in older adults. Proc. Natl. Acad. Sci. U. S. A.
93 (1996), 3043–7.
79. K. Vedhara, N. K. Cox, G. K. Wilcock, et al., Chronic stress in elderly carers of dementia
patients and antibody response to influenza vaccination. Lancet 353 (1999), 627–31.
318 Depression and physical illness
98. A. Sluzewska, Indicators of immune activation in depressed patients. Adv. Exp. Med. Biol.
461 (1999), 59–73.
99. M. Maes, S. Scharpe, H. Y. Meltzer, et al., Relationships between interleukin-6 activity,
acute phase proteins, and function of the hypothalamic-pituitary-adrenal axis in severe
depression. Psychiatry Res. 49 (1993), 11–27.
100. M. Maes, E. Bosmans, H. Y. Meltze, S. Scharpe, E. Suy, Interleukin-1 beta: a putative medi-
ator of HPA axis hyperactivity in major depression. Am. J. Psychiatry 150 (1993), 1189–
93.
101. M. Irwin, M. Daniels, S. C. Risch, E. Bloom, H. Weiner, Plasma cortisol and natural killer
cell activity during bereavement. Biol. Psychiatry 24 (1988), 173–8.
102. C. Raison, A. H. Miller, When not enough is too much: the role of insufficient glucocorticoid
signaling in the pathophysiology of stress-related disorders. Am. J. Psychiatry 160 (2003),
1554–65.
103. M. L. Sopori, W. Kozak, Immunomodulatory effects of cigarette smoke. J. Neuroimmunol.
83 (1998), 148–56.
104. J. A. Rosecrans, L. D. Karin, Effects of nicotine on the hypothalamic-pituitary-axis (HPA)
and immune function: introduction to the Sixth Nicotine Round Table Satellite, American
Society of Addiction Medicine Nicotine Dependence Meeting. Psychoneuroendocrinology
23 (1998), 95–102.
105. C. G. McAllister-Sistilli, A. R. Caggiula, S. Knopf, et al., The effects of nicotine on the
immune system. Psychoneuroendocrinology 23 (1998), 175–87.
106. W. Jung, M. Irwin, Reduction of natural killer cytotoxic activity in major depression:
interaction between depression and cigarette smoking. Psychosom. Med. 61 (1999), 263–
70.
107. M. A. Mendall, P. Patel, M. Asante, et al., Relation of serum cytokine concentrations to
cardiovascular risk factors and coronary heart disease. Heart 78 (1997), 273–7.
108. R. W. Linkins, G. W. Comstock, Depressed mood and development of cancer. Am. J. Epi-
demiol. 132 (1990), 962–72.
109. M. A. Schuckit, J. E. Tipp, M. Bergman, et al., Comparison of induced and independent
major depressive disorders in 2945 alcoholics. Am. J. Psychiatry 54 (1997), 948–57.
110. M. Irwin, C. Caldwell, T. L. Smith, et al., Major depressive disorder, alcoholism, and reduced
natural killer cell cytotoxicity: role of severity of depressive symptoms and alcohol consump-
tion. Arch. Gen. Psychiatry 47 (1990), 713–19.
111. M. Irwin, M. Schuckit, T. L. Smith, Clinical importance of age at onset in Type 1 and Type
2 primary alcoholics. Arch. Gen. Psychiatry 47 (1990), 320–24.
112. M. Irwin, G. Rinetti, L. Redwine, S. Motivala, C. Ehlers, Pro-inflammatory cytokines and
disordered sleep in alcohol dependence. Psychosom. Med. 65 (2003), A-4.
113. M. Hall, A. Baum, D. J. Buysse, et al., Sleep as a mediator of the stress-immune relationship.
Psychosom. Med. 60 (1998), 48–51.
114. M. Irwin, C. Clark, B. Kennedy, J. Christian Gillin, M. Ziegler, Nocturnal catecholamines
and immune function in insomniacs, depressed patients, and control subjects. Brain Behav.
Immun. 17 (2003), 365–72.
320 Depression and physical illness
115. J. Savard, L. Laroche, S. Simard, H. Ivers, C. M. Morin, Chronic insomnia and immune
functioning. Psychosom. Med. 65 (2003), 211–21.
116. M. Irwin, J. C. Gillin, J. Dang, et al., Sleep deprivation as a probe of homeostatic sleep
regulation in primary alcoholics. Biol. Psychiatry 51 (2002), 632–41.
117. M. Irwin, C. Miller, J. C. Gillin, A. Demodena, C. L. Ehlers, Polysomnographic and spectral
sleep EEG in primary alcoholics: an interaction between alcohol dependence and African-
American ethnicity. Alcohol Clin. Exp. Res. 24 (2000), 1376–84.
118. L. Redwine, J. Dang, M. Hall, M. Irwin, Disordered sleep, nocturnal cytokines, and immunity
in alcoholics. Psychosom. Med. 65 (2003), 75–85.
119. P. H. Black, L. D. Garbutt, Stress, inflammation and cardiovascular disease. J. Psychosom.
Res. 52 (2002), 1–23.
120. L. Redwine, S. Snow, P. Mills, M. Irwin, Acute psychological stress: effects on chemotaxis
and cellular adhesion molecule expression. Psychosom. Med. 65 (2003), 598–603.
121. M. U. Goebel, P. J. Mills, M. R. Irwin, M. G. Ziegler, Interleukin-6 and tumor necrosis
factor-alpha production after acute psychological stress, exercise, and infused isoproterenol:
differential effects and pathways. Psychosom. Med. 62 (2000), 591–8.
122. J. F. Sheridan, C. Dobbs, D. Brown, B. Zwilling, Psychoneuroimmunology: stress effects on
pathogenesis and immunity during infection. Clin. Microbiol. Rev. 7 (1994), 200–12.
123. S. Cohen, G. Miller, Stress, immunity and susceptibility to upper respiratory infection.
In Psychoneuroimmunology, 3rd edn, ed. R. Ader, D. Felten, N. Cohen (San Diego, CA:
Academic Press, 2000), pp. 499–509.
124. S. W. Cole, B. D. Naliboff, M. E. Kemeny, et al., Impaired response to HAART in HIV-
infected individuals with high autonomic nervous system activity. Proc. Natl. Acad. Sci. U.
S. A. 98 (2001), 12 695–700.
125. S. W. Cole, M. E. Kemeny, S. E. Taylor, Social identity and physical health: accelerated
HIV progression in rejection-sensitive gay men. J. Pers. Soc. Psychol. 72 (1997), 320–
35.
126. S. W. Cole, Y. D. Korin, J. L. Fahey, J. A. Zack, Norepinephrine accelerates HIV replication
via protein kinase A-dependent effects on cytokine production. J. Immunol. 161 (1998),
610–16.
127. E. M. Sternberg, G. P. Chrousos, R. L. Wilder, P. W. Gold, The stress response and the
regulation of inflammatory disease. Ann. Intern. Med. 117 (1992), 854–66.
128. L. J. Crofford, K. T. Kalogeras, G. Mastorakos, et al., Circadian relationships between inter-
leukin (IL)-6 and hypothalamic-pituitary-adrenal axis hormones: failure of IL-6 to cause
sustained hypercortisolism in patients with early untreated rheumatoid arthritis. J. Clin.
Endocrinol Metab. 82 (1997), 1279–83.
129. C. Dickens, L. McGowan, D. Clark-Carter, F. Creed, Depression in rheumatoid arthritis: a
systematic review of the literature with meta-analysis. Psychosom. Med. 64 (2002), 52–60.
130. A. J. Zautra, B. W. Smith, Depression and reactivity to stress in older women with rheumatoid
arthritis and osteoarthritis. Psychosom. Med. 63 (2001), 687–96.
131. J. M. Smyth, A. A. Stone, A. Hurewitz, A. Kaell, Effects of writing about stressful experiences
on symptom reduction in patients with asthma or rheumatoid arthritis: a randomized trial.
J. Am. Med. Assoc. 281 (1999), 1304–9.
15
15
Smoking and depression
Jon D. Kassel and Benjamin L. Hankin
Introduction
Cigarette smoking remains the most preventable cause of illness and death in society
today. Upwards of 440 000 smokers die in the USA each year [1,2], and worldwide
over 1 200 000 deaths a year are attributable to smoking-related causes [3]. Despite
these staggering statistics, smokers and non-smokers alike often do not appreciate
fully the health risks of tobacco use, particularly cigarette smoking. The latest epi-
demiological studies indicate that death rates for smokers are two to three times
higher than for non-smokers at all ages. This means that half of all smokers will
eventually die as a result of their smoking. If current smoking trends persist, then
about 500 million people currently alive – nearly 9% of the world’s population –
will die as a result of tobacco use [3].
Moreover, tobacco is arguably the most addictive substance known to
humankind. Whereas approximately 32% of people who initiate cigarette smoking
become nicotine-dependent, only 23%, 17% and 15% of individuals who experi-
ment with heroin, cocaine and alcohol, respectively, become dependent on these
drugs [4]. Interestingly, unlike with harder drugs, the subjective effects of smoking
are subtle. Rarely does a smoker describe that smoking provides them with any-
thing approaching the euphoria or high often attributed to other drugs. And yet, as
Shiffman [5] observes, ‘the addictive potential of nicotine is all the more impressive
for its ability to engender such compulsive use without impressive subjective effects’
(p. 15).
The pathways to becoming a smoker are no doubt complex [6,7]. However, of the
numerous factors believed to heighten vulnerability to both smoking initiation and
subsequent development of nicotine dependence, the role played by various forms
of psychopathology and emotional distress appears particularly critical: numer-
ous studies have reliably found high smoking rates among selected populations
of individuals with mental illness. For example, drawing upon a large, nationally
representative sample in the USA, Lasser et al. [8] found that individuals with a
lifetime history of any psychiatric disorder had higher rates of lifetime and current
smoking compared with those individuals who had never suffered from mental ill-
ness. Other investigations have reported similar findings, demonstrating strong and
reliable associations between psychiatric disorders and cigarette smoking among
adults [9–11] and, to a lesser extent, among adolescents [12,13], but see [14].
Whereas the largest associations between smoking initiation have typically been
found with substance abuse disorders [9,12,15] and various manifestations of
externalising disorders [12,16–20], the link between smoking status and depressive
symptomatology has also proven to be robust. Put simply, relative to non-smokers,
smokers are more likely to experience depression, and those individuals who experi-
ence depressive symptoms are more likely to smoke cigarettes relative to those who
are depression-free. Questions invariably arise from such associations, the answers
to which would likely have a bearing on smoking prevention and treatment. Thus, a
thorough understanding of the processes linking smoking and depression is clearly
needed.
The focus of this chapter is on exploring the links between smoking behaviour
and depressive symptoms. Rather than providing an exhaustive review, we high-
light some of the major findings from the literature and address several conceptual
and methodological issues that we believe are critical to gaining a better under-
standing of smoking–depression associations. As such, we begin by providing brief
overviews of the constructs of depression and smoking, noting that neither may
be as straightforward as many believe. Next, we present the case that delineating
the nature of smoking–depression relationships calls for research that goes beyond
simple description of cross-sectional correlational data. Indeed, as we have argued
previously [21], explorations of the links between smoking and any given affective
condition (including depression) call for both between- and within-subject levels
of analyses. Next, we review several conceptual models that may lend themselves
to further elucidation of the processes underlying associations between smoking
and depression. We then highlight several potentially important moderators of the
smoking-depression link. Finally, we offer thoughts on future research directions
for this important area of inquiry.
Some smokers are clearly nicotine-dependent, such that they experience nicotine
withdrawal when they go without nicotine for even short periods of time. But not
all smokers meet criteria for nicotine dependence [39]. Thinking about variability
in smoking behaviour from a developmental perspective, several investigators have
posited that all smokers proceed through a series of stages and transitions on the
path towards nicotine dependence. For example, Flay [40] suggests that smokers
progress through the following five discrete stages: (i) the preparatory stage, in which
attitudes towards nicotine and its perceived functions are formed; (ii) the initial
trying stage, which includes smoking the first two to three cigarettes, usually in
a social context, and the resulting physiological and psychosocial reinforcements
obtained; (iii) the experimentation stage, which includes situational-specific irregu-
lar use over an extended period of time; (iv) regular use, during which the individual
(typically an adolescent) smokes on a regular basis, e.g. at weekends or daily; and
(v) nicotine dependence or addiction, in which smoking is governed predominantly
by an internally regulated need for nicotine. Hence, one important implication of a
stage-modelling perspective is that the relationship between smoking and depres-
sion may differ as a function of where a smoker falls on this stage continuum.
Put another way, whereas smoking at all certainly heightens the risk of eventu-
ally becoming nicotine dependent, such an outcome must not be viewed as destiny.
Some smokers, for example, remain at an experimentation or intermittent smoking
stage for prolonged periods of time.
It is also important to note that some of the processes governing smoking–
depression associations are inherently between-person, whereas others are within-
person (see Kassel et al. [21] for an in-depth discussion of these issues). For example,
the question of whether smokers experience heightened depressive symptoms rel-
ative to non-smokers necessarily calls for a between-person level of analysis, as
groups of people are being compared. On the other hand, ascertaining whether
depressive episodes or states actually cue smoking requires a within-person level
of analysis, as it must be demonstrated that a given smoker smokes on occasions
defined by depressive symptoms. Finally, some questions combine both levels of
analysis. For example, the popular notion that some smokers smoke in order to
alleviate unpleasant affect – such as depression or dysphoria – suggests that cer-
tain between-person variables are involved in within-person attempts to regulate
depressive negative affect when it occurs. We believe that these distinctions have
been largely overlooked in previous analyses of smoking–depression relationships.
Thus, in order to understand thoroughly the relationship between cigarette
smoking and depression, several distinct but frequently blurred questions must
be asked: First, does depression promote smoking? Specifically, are there valid and
reliable associations between depression and (i) smoking status (smoker vs. non-
smoker) and (ii) actual cueing (prompting) of smoking? Second, even if it was
325 Smoking and depression
Do smokers and
non-smokers differ on levels Do depressive states cue Does smoking reduce
Smoking stage of depression? smoking? symptoms of depression?
established that depression is linked to smoking (at either or both of these levels of
analysis), this does not mean that smoking necessarily relieves depressive symptoms.
This often ignored point leads to another, very different, question: Does smoking
genuinely reduce depressive symptoms? Moving beyond the self-report of smokers,
what do experimental studies reveal regarding the influence of smoking on depres-
sive symptomatology? In sum, then, the following three questions are addressed
regarding smoking–depression relationships: (i) Do smokers and non-smokers dif-
fer on levels of depression? (ii) Do depressive symptoms genuinely cue smoking?
(iii) Does smoking alleviate symptoms of depression? Moreover, and as noted ear-
lier, these questions ideally should be asked across different stages of smoking
experience, as the answers may vary across smoking stages. Here, we address these
questions across the smoking initiation (irregular, experimental smoking) mainte-
nance (regular, daily smoking) and relapse (resumption of smoking after a period
of abstinence) stages (see Table 15.1 for a summary of our review).
326 Depression and physical illness
Maintenance stage
As noted with regard to smoking initiation, depressive symptomatology in particu-
lar emerges as a strong correlate of smoking status among those in the maintenance
stage [65]. In 1978, Waal-Manning and de Hamel [66] reported that smokers had
elevated depressive symptoms relative to non-smokers. Similar relationships have
been found in studies using large nationally representative samples [11,67]. In a
study of 3000 young adults, Breslau and colleagues [9] found that those who met
criteria for nicotine dependence were also more likely to meet criteria for major
depression. Breslau and colleagues [68] also reported that a history of major depres-
sive disorder produced a two-fold increased risk for progression to nicotine depend-
ence. Importantly, findings from Breslau et al. [69] indicated that the association
between major depression and smoking was specific to nicotine dependence; non-
dependent smokers did not differ from non-smokers in this respect. As observed in
the initiation stage, the presence of nicotine dependence also appears to heighten
the risk for subsequent development of major depression [68]. In our view, the best
studies suggest that the frequently observed link between nicotine dependence and
depression may reflect (i) bidirectional causal processes (e.g. smoking to alleviate
depressed mood and neuropharmacological effects of nicotine on neural substrates
linked to depression), (ii) common factors (e.g. neuroticism) that predispose to both
disorders [59,68,70] or (iii) the influence of non-shared (individual) environmental
factors [71].
Finally, with respect to whether depressive affect actually cues or prompts smok-
ing, the literature has little to say. Again, there is reason to believe that both smoking
status (smokers vs. non-smokers) and the amount smoked co-vary with various
indices of depression [71]. In an effort to establish the direction of causality, sev-
eral laboratory studies have demonstrated that during stressful situations, smoking
328 Depression and physical illness
intensity increases [72–74], smokers tend to smoke more [75–77] and self-reported
desire to smoke is heightened [78]. Taken together, these within-person studies
make a compelling case that stress – but not necessarily depression – increases
(cues) smoking among regular smokers.
Relapse stage
Relapse is the modal outcome among those attempting to quit smoking [79,80].
Whereas the best available treatments yield one-year abstinence rates of about 30%,
even among smokers who successfully quit for a full year as many as 40% eventually
return to regular smoking [81]. Smokers who attempt to quit on their own fare
even less well than those who seek formal treatment, with relapse rates ranging
from 90% to 97% [82,83].
With respect to the question of whether depression predisposes to or even
prompts smoking relapse, a number of studies point to a strong relationship. Here,
we find that the presence of clinically significant levels of negative affect (often
depressive symptoms) is frequently predictive of relapse [84,85]. For instance, one
study reported that the likelihood of quitting smoking was 40% lower among
depressed smokers compared with non-depressed smokers [67]. Glassman et al.
[84] reported a quitting rate of 14% for smokers meeting criteria for major depres-
sion, while 31% of participants with no psychiatric diagnosis successfully quit. In
the absence of current symptomatology, history of depression appears to heighten
risk for both relapse [86] and recurrence of depressive symptomatology subse-
quent to cessation [87]. These effects may be more pronounced among women
[88]. Importantly, Niaura et al. [89] demonstrated that even low levels of depres-
sive symptoms assessed at baseline for smokers enrolled in a cessation programme
were predictive of time to first cigarette smoked after attempted quitting.
Despite all of the studies demonstrating a link between depression and propensity
to relapse, a meta-analysis revealed that lifetime history of major depression does
not appear to be an independent risk factor for cessation failure in smoking cessation
treatment [90]. Given the corpus of data previously suggesting otherwise, this study
suggests that more work must be done in order to better understand the role of
depression in promoting relapse.
Initiation stage
Sadly, there has been very little empirical investigation with respect to the effect of
smoking on depressive affect or any other affective states among smoking initiates.
A small body of self-report data, however, does bear on this issue. McNeill and
colleagues [91] found that the most frequently cited motive for smoking among a
329 Smoking and depression
sample of female adolescent smokers was that smoking is calming. Dozois and col-
leagues [92] and Nichter and colleagues [93] similarly found that ‘smoking to relax’
and ‘stress reduction’ were commonly reported motives among adolescent smoking
initiates. Whereas adolescents often attribute their smoking to motives associated
with stress reduction, it is not clear that such motives necessarily generalise to states
of depressive affect.
One can also turn to the animal literature in order to gain some insight into
this question. For example, it has been reported that nicotine had greater activity-
stimulating (perhaps antidepressant) effects in adolescent male rats than adult male
rats [94]. A study further reported that although nicotine exerted anxiety-reducing
effects in adolescent male rats, it increased anxiety among adolescent female rats
and adult males and females [95]. Finally, Cheeta and colleagues [96] found that
female rats were more sensitive to the anxiety-reducing effects of nicotine than were
male rats. Taken together, these and other studies suggest that smoking (nicotine)
may yield differential effects on adolescents relative to adults, and that these effects
may be moderated by sex. Still unclear, however, is the extent to which smoking
offers genuine antidepressant effects among smoking initiates.
Maintenance stage
Once again, the literature reveals a dearth of information regarding the effects of
smoking on depressive affect among regular smokers. Given the reliable association
between depression and smoking discussed earlier, it is surprising that relatively
few studies have directly assessed the effects of smoking or nicotine on depression.
Smokers reliably report that they smoke more when they are stressed, angry, anxious
or sad [97–100] (but see also [101]) and hold the expectation that smoking will
alleviate these negative moods [102,103]. Moreover, a study found that, relative to
non-depressed smokers, depressed smokers perceive more benefits from smoking
and find cigarettes more appealing than alternative rewards [104]. At the same time,
the fact that smokers believe that smoking helps to reduce these manifestations of
negative affect does not, in and of itself, render this a valid conceptualisation.
Additionally, most of the experimental work to date examining nicotine effects on
emotion has assessed anxiety; other emotional states, including depression, have
been relatively ignored. As we have argued elsewhere [21], this represents a notable
gap in the literature that sorely needs to be addressed.
As observed with the smoking initiation stage, however, the animal literature
offers some clues regarding nicotine effects on depressive symptomatology. Sev-
eral studies have suggested that nicotine may exhibit antidepressant effects in rats
[105–107]. For example, Tizabi and colleagues [108] reported that acute and chronic
administration of nicotine significantly improved the performance of Flinders Sen-
sitive Line rats (bred for their hyperresponsiveness to cholinergic stimulation and,
330 Depression and physical illness
Relapse stage
The question of whether smoking a cigarette relieves symptoms of depression
among those in the throes of relapse has garnered little empirical attention. Brandon
et al. [112] asked smokers who smoked subsequent to participating in a cessa-
tion programme to describe their affective reactions to their initial lapse. Whereas
almost 50% described feeling depressed or hopeless, 16% anxious or tense, and 10%
angry or irritated, only 8% reported feeling relaxed and 6% happy, celebratory or
confident. Using palmtop computers to assess real-time affective antecedents and
consequences of smoking lapses, Shiffman et al. [113] reported that whereas lapses
resulted in increases in negative affect, temptation episodes did not. Correspond-
ingly, relative to temptations, lapses almost inevitably resulted in significant drops
in self-efficacy and increases in feelings of guilt and discouragement.
Thus, it appears that smoking a cigarette after having tried to quit smoking results
in an exacerbation of depressed mood. However, the finding that smoking lapses
appear to increase depressive symptoms likely reflect, at least in part, an abstinence
violation effect (AVE). The AVE refers to a frequently observed constellation of
negative emotions and disparaging self-evaluation that follows a transgression of
one’s commitment to abstinence [114].
Another interesting perspective on whether smoking reduces depressive symp-
toms comes from studies examining the time course of depressive symptomatology
among those who successfully quit smoking. Almost all of these studies report that
although there is usually an initial increase in dysphoria subsequent to cessation,
these symptoms diminish over time to levels lower than observed before quitting
[115–118]; but see [119]. Moreover, smokers who were unable to maintain absti-
nence generally continued to manifest high levels of stress and negative affect over
time [115,116]. By implication, it has been suggested that whereas smoking engen-
ders depressive symptomatology [120], quitting results in lower depressive levels
over time.
Summary
Based upon a review of the empirical literature, we observed consistent between-
person associations between smoking status and various indices of depression across
331 Smoking and depression
Despite the corpus of cross-sectional and prospective studies that have established
an association between smoking and depression, such associations may represent a
multitude of different relations, including the presence of other factors influencing
both smoking and depression. Highlighting this point, Dierker and colleagues [123]
asserted: ‘Although the association between depression and smoking has been con-
sistently established, little evidence regarding the mechanisms that influence this
association is currently available’ (p. 947). Because the theories and claims about
causal processes have bearing on smoking prevention, intervention and even policy,
determination of the underlying causal mechanisms represents a critically import-
ant next step for the tobacco-control research agenda. Hence, the primary focus of
this section is to consider a variety of conceptual models that may help to explain
the smoking–depression association more formally.
These models, although proposed and examined principally within the person-
ality and psychopathology literature [124], hold promise for advancing a deeper
understanding of the causal processes involved in the link between smoking and
depression. In general, we focus our empirical review of these models around rigor-
ous prospective longitudinal studies as opposed to cross-sectional research, because
the former approach affords the best available evidence with which to evaluate these
different models and because prospective studies clearly hold the potential to dis-
entangle some of the methodological problems inherent in purely cross-sectional
research, e.g. confusing cause, correlates and consequences [21].
Predisposition model
The predisposition model states explicitly that depression (symptoms or clinical
episode) precedes and is causally implicated in smoking initiation, maintenance
and relapse. As noted earlier, a wealth of prospective longitudinal studies provides
evidence for the predisposition model [12,13,50,55,59,63,122,125–128]. Many of
these investigations examined experimental or initial non-smokers who were then
followed prospectively to show that they progressed to become regular smokers, and
several of the studies prospectively examined elevations in future smoking levels
after controlling for initial levels of smoking and other confounds.
Consequence model
The consequence model posits the opposite causal direction: long-term smoking (or
even, as some argue, short-term smoking) leaves a direct, aetiological ‘scar’ that con-
tributes to the risk for developing depression. Longitudinal studies, many of which
also examined the predisposition model, provide burgeoning evidence for the con-
sequence model as well [12,50,51,59,63,121,122,125,127,129]. Overall, these studies
show that smoking predicts onset of depressive disorder or prospective increases
in depressive symptoms over time in both adolescents and adults. Several of these
333 Smoking and depression
studies tested both the predisposition model and the consequence model with the
same data and found evidence for bidirectional associations [12,13,50,59,63,125,
127], although other studies exploring the use-to-distress and distress-to-use
hypotheses found support for only one direction [51,129,130].
Spectrum model
The spectrum, or common-cause, model states that smoking and depression arise
from shared aetiological factors. In essence, the spectrum model posits that the sig-
nificant co-variation found between smoking and depression, even when examined
in prospective longitudinal studies with multiple waves of data, may be explained
by a single factor or set of variables that give rise to risk for both smoking and
depression. Evidence for this model comes from multiple sources and theoreti-
cally proposed vulnerability factors. We consider a few specific examples here; see
[121] for a sophisticated longitudinal examination of the common-cause hypoth-
esis. For example, psychosocial influences, such as a child’s attachment to his
or her parents, have been shown to predict smoking initiation in a four-year
prospective study [131]; likewise, parental attachment predicts depression in youth
[132]. Behavioural genetic research conducted by Kendler and colleagues [133]
found a shared genetic liability to major depression that overlaps with smok-
ing [71,134]. In addition to shared genetic influences, neurotransmitters, includ-
ing dopamine, serotonin, nonadrenaline, gamma-aminobutyric acid (GABA) and
glutamate, appear to be neural substrates that are implicated in the aetiology of
both depression and smoking [135]. Personality trait vulnerabilities, such as neur-
oticism, have also been identified as risks for both depression [136] and smoking
[137].
As there are many potential shared aetiological vulnerabilities that could be
examined, it should be emphasised that many of the common vulnerabilities giv-
ing rise to the co-variation of depression and smoking are likely not completely
independent risk factors. For example, a molecular genetic study found that the
promotor for the serotonin transporter gene interacted with neuroticism to pre-
dict smoking behaviour, such that neuroticism was associated positively with cur-
rent smoking and associated negatively with smoking cessation among those with
poorly transcribed serotonin genotypes [138]. Of interest, results from a prospect-
ive birth cohort study showed that this same serotonin transporter gene inter-
acted with stressful life events to predict increases in depression over time [139].
Thus, these two different studies with separate predicted outcomes (smoking [138]
and depression [139]) reveal that the same molecular genetic vulnerability (poorly
transcribed serotonin transporter genes) interacts with other risk factors, highlight-
ing the importance of considering common-cause models at multiple levels with
different outcomes.
334 Depression and physical illness
Pathoplasticity model
Finally, the pathoplasticity model posits that the influence of smoking and depres-
sion on the presentation of each other is bidirectional, such that the presence of
depression influences the course and presentation of smoking over time (e.g. ini-
tiation, maintenance, relapse) and smoking, or nicotine dependence, affects the
course of depression (e.g. severity, recurrence). Consistent with the pathoplastic-
ity perspective, we have already observed that smoking places the individual at
risk for subsequent development of depressive symptomatology, and that depres-
sion predisposes to smoking behaviour, particularly nicotine dependence. We have
also noted that according to a number of studies, a lifetime history of depres-
sion puts the individual at heightened risk for smoking relapse, but see [90]. Also
of relevance, a study reported that individuals with one of various mental disor-
ders and whose illness had remitted were not at increased risk for daily smoking,
in contrast to individuals with active disorders [140]. An important implica-
tion of these findings points to the possibility of previously unrecognised public-
health benefits of early treatment of mental disorders, i.e. prevention of smoking
initiation.
A burgeoning literature on smoking cessation outcome studies of depressed
smokers, utilising treatments aimed at reducing depressive symptomatology, also
bears on notions derived from the pathoplasticity model. For instance, the antide-
pressant medication bupropion has received empirical scrutiny within the realm
of smoking cessation. Bupropion has demonstrated efficacy for standard smoking
treatment [141–143] and in conjunction with nicotine-replacement therapy [144]
and for relapse prevention [145]. Importantly, several studies have reported that
bupropion was equally efficacious for individuals with a history of major depression
[146,147]. Lerman et al. [142] found that highly nicotine-dependent smokers who
received bupropion were more likely to experience decreases in depressive symp-
toms during active drug treatment but were also more likely to experience a rebound
in depressive symptoms when bupropion was discontinued. Taken together, these
findings support the notion that at least one mode of pharmacotherapy specifi-
cally designed to treat major depressive disorder also facilitates smoking cessation,
regardless of whether the smoker has a history of depression or even current depres-
sive symptoms.
Non-pharmacological approaches to treating depression have also been assessed
in the context of smoking cessation. Several studies have incorporated cognitive–
behavioural mood-management modules aimed specifically at facilitating cessation
among formerly depressed smokers. Brown et al. [148] found that although inclu-
sion of a mood-management component did not reduce depressive symptoms in
formerly depressed smokers trying to quit smoking, it was more effective than
335 Smoking and depression
standardised treatment (without mood management) for heavy smokers and for
those with a history of recurrent major depressive disorder. Interestingly, about
40% of participants showed a pattern of increasing depressive symptoms dur-
ing the two weeks subsequent to quitting smoking, a pattern that was associated
with poor smoking outcome [149]. Several studies by Hall and colleagues have
utilised similar designs and yielded mixed findings regarding the efficacy of cessation
programmes incorporating mood-management techniques [150,151].
Summary
Several conceptual models were described that offer potential explanation for the
observed associations between smoking and depressive symptomatology. These
explanations include (i) smoking predisposes to depression, (ii) depression predis-
poses to smoking, (iii) shared aetiological factors account for both smoking and
depression and (iv) the relationship between smoking and depression is reciprocal,
such that each influences the course and development of the other. Indeed, as we
have seen, there exists evidence in support of all four perspectives. Nonetheless,
much more work needs to be done to understand fully the complex nature of the
smoking–depression link.
than static conceptualisations. Hence, this observation suggests caution in not over-
interpreting evidence for or against one or another conceptual model because any
association between smoking and depression is likely to be modified, at least in
part, by social context, climate and time.
Second, we reviewed several studies highlighting the influence that sex may play
in the association between smoking and depression. In a two-cohort study of 14-
to 18-year olds [55], boys at baseline who did not have a smoking history but had
higher depression symptom scores were more likely to initiate smoking over the
three- to four-year prospective follow-up, whereas depression scores were not found
to predict smoking initiation among non-smoking girls. In another longitudinal
study with three prospective waves following youth from high school to age 24 years
[13], smoking initiation preceded onset of depression and, likewise, lifetime depres-
sion was associated with smoking initiation for both sexes, although early-onset
smokers had a family history of depression among women but not men. An in-
depth prospective electronic diary study with adolescents [128] showed that both
boys and girls with depressive symptoms exhibited elevated rates of actual smok-
ing and urges to smoke, but depression increased the risk of smoking among girls
with behavioural problems, whereas depression decreased the likelihood of smok-
ing among boys with externalising symptoms. Finally, in a four-year longitudinal
study of youths aged 12–19 years [153], smoking predicted greater depressive symp-
toms over time; this effect was stronger among girls than boys. In sum, these studies
show that it is important to consider sex when evaluating both the predisposition
and consequence models.
Third, and as we have alluded to throughout this chapter, it is important to
consider differences between initiation/onset and persistence/recurrence in under-
standing the unfolding of both smoking and depression over time. As highlighted
in the depression literature, there may be similar and/or different processes that
contribute to first onsets of depression compared with maintenance or recurrence
[154]. For example, Lewinsohn and colleagues [155] examined numerous psy-
chosocial risk factors for depression onset versus recurrence. They found that
whereas major life events predicted onset of first depression, dysfunctional atti-
tudes (rigid extreme thinking) were associated more with depression recurrence.
Likewise, a large behavioural genetic study from multiple countries found that the
genetic influences for smoking persistence were not exactly the same as those for
smoking initiation; indeed, less than 40% of the genetic variance in persistence was
the same as that for initiation [134].
Related to the distinction between initiation and persistence, researchers may
also have to consider the effect that severity of depression and/or smoking has on
causal processes across different etiological models. Severity may be an important
337 Smoking and depression
variable to take into account based on inconsistent results from two behavioural
genetic twin studies examining the genetic links between smoking and depres-
sion. In the first of these studies [133], genetic influences accounted for the
association between depression diagnosis and liability to smoking over the life-
time, whereas the second study [71] found that non-shared specific environmen-
tal influences accounted for the majority of the association between depressive
symptoms and lifetime smoking liability. Although the latent variable under-
lying depression is dimensional (as reviewed earlier), studies measuring clin-
ical levels of diagnosed depression tend to examine the more severe end of the
depression continuum compared with those studies assessing depressive symptom
variability.
The concept of multifinality suggests that any one factor may function differently
depending on the organisation of the system in which it operates and, hence, may
be indicative of interaction effects [158].
The bottom line is that in this field, we must move beyond a simple descrip-
tion of cross-sectional associations between smoking and depression towards more
sophisticated and theoretically informed methodologies. The smoking problem is
not going away soon, and neither is depression. As each disorder exacts its devas-
tating toll in terms of lives interrupted and lives lost, it becomes incumbent upon
both tobacco and depression researchers alike to elucidate the underlying nature of
the smoking–depression link more accurately.
Acknowledgements
The writing of this chapter was supported, in part, by grant P01 CA98262–01A2
from the National Cancer Institute.
REFERENCES
10. L. S. Covey, A. H. Glassman, F. Stetner, Cigarette smoking and major depression. J. Addict.
Dis. 17 (1998), 35–46.
11. L. Degenhardt, W. Hall, The relationship between tobacco use, substance-use disorders and
mental health: results from the National Survey of Mental Health and Well-being. Nicotine
Tob. Res. 3 (2001), 225–34.
12. R. A. Brown, P. M. Lewinsohn, J. R. Seeley, E. F. Wagner, Cigarette smoking, major depression,
and other psychiatric disorders among adolescents. J. Am. Acad. Child Adolesc. Psychiatry
35 (1996), 1602–10.
13. P. Rohde, P. M. Lewinsohn, R. A. Brown, J. M. Gau, C. W. Kahler, Psychiatric disorders,
familial factors and cigarette smoking: I. Associations with smoking initiation. Nicotine Tob.
Res. 5 (2003), 85–98.
14. H. P. Upadhyaya, D. Deas, K. T. Brady, M. Kruesi, Cigarette smoking and psychiatric comor-
bidity in children and adolescents. J. Am. Acad. Child Adolesc. Psychiatry 41 (2002), 1294–
305.
15. K. S. Kendler, M. C. Neale, P. Sullivan, et al., A population-based twin study in women of
smoking initiation and nicotine dependence. Psychol. Med. 29 (1999), 299–308.
16. D. B. Kandel, J. G. Johnson, H. R. Bird, et al., Psychiatric disorders associated with substance
use among children and adolescents: findings from the methods for the Epidemiology of
Child and Adolescent Mental disorders (MECA) Study. J. Abnorm. Child. Psychol. 25 (1997),
121–32.
17. S. G. Kellam, J. C. Anthony, Targeting early antecedents to prevent tobacco smoking: findings
from an epidemiologically based randomized field trial. Am. J. Public Health 88 (1998),
1490–94.
18. M. T. Lynskey, D. M. Ferguson, Childhood conduct problems, attention deficit behaviors,
and adolescent alcohol, tobacco, and illicit drug use. J. Abnorm. Child. Psychol. 23 (1995),
281–302.
19. R. J. McMahon, Child and adolescent psychopathology as risk factors for subsequent tobacco
use. Nicotine Tob. Res. 1 (1999), S45–50.
20. S. Miller-Johnson, J. E. Lochman, J. D. Coie, R. Terry, C. Hyman, Comorbidity of con-
duct and depressive problems at sixth grade: substance use outcomes across adolescence.
J. Abnorm. Child. Psychol. 26 (1998), 221–32.
21. J. D. Kassel, L. R. Stroud, C. A. Paronis, Smoking, stress, and negative affect: correlation,
causation, and context across stages of smoking. Psychol. Bull. 129 (2003), 270–304.
22. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders,
4th edn (Washington, DC: American Psychiatric Association, 1994).
23. J. C. Coyne, Self-reported distress: analog or ersatz depression. Psychol. Bull. 116 (1994),
29–45.
24. G. L. Flett, K. Vrendenburg, L. Krames, The continuity of depression in clinical and non-
clinical samples. Psychol. Bull. 121 (1997), 395–416.
25. A. M. Ruscio, J. Ruscio, The latent structure of analogue depression: should the Beck Depres-
sion Inventory be used to classify groups? Psychol. Assess. 14 (2002), 135–45.
26. J. Ruscio, A. M. Ruscio, Informing the continuity controversy: a taxometric analysis of
depression. J. Abnorm. Psychol. 109 (2000), 473–87.
340 Depression and physical illness
81. Department of Health and Human Services, The Health Benefits of Smoking Cessation: A
Report of the Surgeon General, DHHS publication no. CDC 90–8416. (Atlanta, GA: Public
Health Service, Centers for Disease Control, National Center for Chronic Disease Prevention
and Health Promotion, Office on Smoking and Health, 1990).
82. S. Cohen, E. Lichtenstein, J. O. Prochaska, et al., Debunking myths about self-quitting:
evidence from 10 prospective studies of persons who attempt to quit smoking by themselves.
Am. Psychol. 44 (1989), 1355–65.
83. J. R. Hughes, S. B. Gulliver, J. W. Fenwick, et al., Smoking cessation among self-quitters.
Health Psychol. 11 (1992), 331–4.
84. A. H. Glassman, J. E. Helzer, L. S. Covey, et al., Smoking, smoking cessation and major
depression. J. Am. Med. Assoc. 264 (1990), 1546–9.
85. S. M. Hall, R. F. Munoz, V. I. Reus, K. L. Sees, Nicotine, negative affect, and depression.
J. Consult. Clin. Psychol. 61 (1993), 761–7.
86. L. S. Covey, Tobacco cessation among patients with depression. Primary Care 26 (1999),
691–706.
87. L. S. Covey, A. H. Glassman, F. Stetner, Major depression following smoking cessation. Am.
J. Psychiatry 154 (1997), 263–5.
88. B. Borrelli, B. Bock, T. King, B. Pinto, B. H. Marcus, The impact of depression on smoking
cessation in women. Am. J. Prev. Med. 12 (1996), 378–87.
89. R. Niaura, D. M. Britt, W. G. Shadel, et al., Symptoms of depression and survival experience
among three samples of smokers trying to quit. Psychol. Addict. Behav. 15 (2001), 13–17.
90. B. Hitsman, B. Borrelli, D. E. McChargue, B. Spring, R. Niaura, History of depression and
smoking cessation outcome: a meta-analysis. J. Consult. Clin. Psychol. 71 (2003), 657–63.
91. A. D. McNeill, M. Jarvis, R. West, Subjective effects of cigarette smoking in adolescents.
Psychopharmacologia 92 (1987), 115–17.
92. D. N. Dozois, J. A. Farrow, A. Miser, Smoking patterns and cessation motivations during
adolescence. Int. J. Addict. 30 (1995), 1485–98.
93. M. Nichter, M. Nichter, N. Vuckovic, G. Quintero, C. Ritenbaugh, Smoking experimentation
and initiation among adolescent girls: qualitative and quantitative findings. Tobacco Control
6 (1997), 285–95.
94. M. M. Faraday, B. M. Elliott, N. E. Grunberg, Adult vs. adolescent rats differ in biobehavioral
responses to chronic nicotine administration. Pharmacol. Biochem. Behav. 70 (2001), 475–
89.
95. B. M. Elliott, M. M. Faraday, J. M. Phillips, N. E. Grunberg, Effects of nicotine on elevated
plus maze and locomotor activity in male and female adolescent and adult rats. Pharmacol.
Biochem. Behav. 77 (2004), 21–8.
96. S. Cheeta, E. E. Irvine, S. Tucci, J. Sandhu, S. E. File, In adolescence, female rats are more
sensitive to the anxiolytic effect of nicotine than are male rats. Neuropsychopharmacology
25 (2001), 601–7.
97. R. Coan, Personality variables associated with cigarette smoking. J. Pers. Soc. Psychol. 26
(1973), 86–104.
98. F. F. Ikard, P. E. Green, D. Horn, A scale to differentiate between types of smoking as related
to the management of affect. Int. J. Addict. 4 (1969), 649–59.
344 Depression and physical illness
99. McKennell, Smoking motivation factors. Br. J. Soc. Clin. Psychol. 9 (1970), 8–22.
100. M. A. H. Russell, J. Peto, U. A. Patel, The classification of smoking by factorial structure of
motives. J. R. Stat. Soc. Ser. A 137 (1974), 313–33.
101. S. Shiffman, Assessing smoking patterns and motives. J. Consult. Clin. Psychol. 61 (1993),
732–42.
102. T. H. Brandon, T. B. Baker, The smoking consequences questionnaire: the subjective expected
utility of smoking in college students. Psychol. Assess. 3 (1991), 484–91.
103. A. L. Copeland, T. H. Brandon, E. P. Quinn, The Smoking Consequences Questionnaire –
Adult: measurement of smoking outcome expectancies of experienced smokers. Psychol.
Assess. 7 (1995), 484–94.
104. B. Spring, R. Pingitore, D. E. McChargue, Reward value of cigarette smoking for comparably
heavy smoking schizophrenic, depressed, and nonpatient smokers. Am. J. Psychiatry 160
(2003), 316–22.
105. T. P. George, M. R. Picciotto, C. D. Verrico, R. H. Roth, Effects of nicotine pretreatment on
dopaminergic and behavioral responses to conditioned fear stress in rats: dissociation of
biochemical and behavioral effects. Biol. Psychiatry 49 (2001), 300–306.
106. T. P. George, C. D. Verrico, R. H. Roth, Effects of repeated nicotine pre-treatment on
mesoprefrontal dopaminergic and behavioral responses to acute footshock stress. Brain
Res. 801 (1998), 36–49.
107. J. Semba, C. Mataki, S. Yamada, M. Nankai, M. Toru, Antidepressantlike effects of chronic
nicotine on learned helplessness paradigm in rats. Biol. Psychiatry 43 (1998), 389–91.
108. Y. Tizabi, D. H. Overstreet, A. H. Rezvani, et al., Antidepressant effects of nicotine in an
animal model of depression. Psychopharmacology (Berl.) 142 (1999), 193–9.
109. D. J. Djuric, E. Dunn, D. H. Overstreet, A. Dragomir, M. Steiner, Antidepressant effect of
ingested nicotine in female rats of Flinders resistant and sensitive lines. Physiol. Behav. 67
(1999), 533–7.
110. D. J. K. Balfour, D. L. Ridley, The effects of nicotine on neural pathways implicated in
depression: a factor in nicotine addiction? Pharmacol. Biochem. Behav. 66 (2000), 79–85.
111. D. J. K. Balfour, The influence of stress on psychopharmacological responses to nicotine.
Br. J. Addiction 86 (1991), 489–93.
112. T. H. Brandon, S. T. Tiffany, K. M. Obremski, T. B. Baker, Postcessation cigarette use: the
process of relapse. Addict. Behav. 15 (1990), 105–14.
113. S. Shiffman, M. Hickcox, J. A. Paty, et al., The abstinence violation effect following smoking
lapses and temptations. Cogn. Ther. Res. 21 (1997), 497–523.
114. G. A. Marlatt, J. R. Gordon Relapse Prevention: Maintenance Strategies in the Treatment of
Addictive Behaviors (New York: Guilford Press, 1985).
115. M. P. Carey, D. L. Kalra, K. B. Carey, S. Halperin, C. S. Richards, Stress and unaided smoking
cessation: a prospective investigation. J. Consult. Clin. Psychol. 61 (1993), 831–8.
116. S. Cohen, E. Lichtenstein, Perceived stress, quitting smoking, and smoking relapse. Health
Psychol. 9 (1990), 466–78.
117. J. R. Hughes, Tobacco withdrawal in self-quitters. J. Consult. Clin. Psychol. 60 (1992), 689–97.
118. R. West, P. Hajek, What happens to anxiety levels on giving up smoking? Am. J. Psychiatry
154 (1997), 1589–92.
345 Smoking and depression
138. S. Hu, C. L. Brody, C. Fisher, et al., Interaction between the serotonin transporter gene and
neuroticism in cigarette smoking behavior. Mol. Psychiatry 5 (2000), 181–8.
139. A. Caspi, K. Sugden, T. E. Moffitt, et al., Influence of life stress on depression: moderation
by a polymorphism in the 5-HTT gene. Science 301 (2003), 386–9.
140. N. Breslau, S. P. Novak, R. C. Kessler, Daily smoking and the subsequent onset of psychiatric
disorders. Psychol. Med. 34 (2004), 323–33.
141. J. S. Ahluwalia, K. J. Harris, D. Catley, K. S. Okuyemi, M. S. Mayo, Sustained-release bupro-
pion for smoking cessation in African Americans: a randomized controlled trial. J. Am.
Med. Assoc. 288 (2002), 468–74.
142. C. Lerman, R. Niaura, B. N. Collins, et al., Effect of bupropion on depression symptoms in
a smoking cessation clinical trial. Psychol. Addict. Behav. 18 (2004), 362–6.
143. R. D. Hurt, D. P. Sachs, E. D. Glover, et al., A comparison of sustained-release bupropion
and placebo for smoking cessation. N. Engl. J. Med. 337 (1997), 1195–202.
144. D. E. Jorenby, S. J. Leischow, M. A. Nides, et al., A controlled trial of sustained-release
bupropion, a nicotine patch, or both for smoking cessation. N. Engl. J. Med. 340 (1999),
685–91.
145. J. T. Hays, R. D. Hurt, N. A. Rigotti, et al., Sustained-release bupropion for pharmacologic
relapse prevention after smoking cessation: a randomized, controlled trial. Ann. Intern. Med.
135 (2001), 423–33.
146. K. E. Hayford, C. A. Patten, T. A. Rummans, et al., Efficacy of bupropion for smoking ces-
sation in smokers with a former history of major depression or alcoholism. Br. J. Psychiatry
174 (1999), 173–8.
147. L. Sanderson, C. A. Patten, R. S. Niaura, et al., Efficacy of bupropion for relapse prevention
in smokers with and without a past history of major depression. J. Gen. Intern. Med. 19
(2004), 828–34.
148. R. A. Brown, C. W. Kahler, R. Niaura, et al., Cognitive-behavioral treatment for depression
in smoking cessation. J. Consult. Clin. Psychol. 69 (2001), 471–80.
149. E S. Burgess, R. A. Brown, C. W. Kahler, et al., Patterns of change in depressive symptoms
during smoking cessation: who’s at risk for relapse? J. Consult. Clin. Psychol. 70 (2002),
356–61.
150. S. M. Hall, R. F. Munoz, V. I. Reus, Cognitive-behavioral intervention increases abstinence
rates for depressive-history smokers. J. Consult. Clin. Psychol. 62 (1994), 141–6.
151. S. M. Hall, R. F. Munoz, V. I. Reus, et al., Mood management and nicotine gum in smoking
treatment: a therapeutic contact and placebo-controlled study. J. Consult. Clin. Psychol. 64
(1996), 1003–1009.
152. J. M. Murphy, N. J. Horton, R. R. Monson, et al., Cigarette smoking in relation to depression:
historical trends from the Stirling County Study. Am. J. Psychiatry 160 (2003), 1663–9.
153. N. L. Galambos, B. J. Leadbeater, E. T. Barker, Gender differences in and risk factors for
depression in adolescence: a 4-year longitudinal study. Int. J. Behav. Dev. 28 (2004), 16–25.
154. B. L. Hankin, J. R. Z. Abela, Depression from childhood through adolescence and adulthood:
a developmental vulnerability-stress perspective. In Development of Psychopathology: A
Vulnerability–Stress Perspective, ed. B. L. Hankin, J. R. Z. Abela (Thousand Oaks, CA, Sage,
2005), pp. 245–88.
347 Smoking and depression
Introduction
There is an extensive scientific literature dating back several decades linking lack of
physical exercise with depressed mood, limited coping skills and low levels of psy-
chological well-being [1,2]. Physical inactivity is also associated with increased risk
for many of the medical problems discussed in this book. In some illnesses, such as
coronary heart disease (CHD), physical inactivity is thought to have a causal role in
disease risk, mediated through metabolic and other pathways [3]. In other condi-
tions, such as chronic pain, inactivity does not cause the problem but nevertheless
it contributes to disability. Physical exercise is also important in the regulation and
management of chronic conditions, such as diabetes. The question therefore arises
of whether the associations between depression and medical problems are medi-
ated through low levels of physical activity and increased sedentary behaviour.
Such possibilities have important implications for prevention and patient
care.
This chapter discusses the evidence linking physical activity, depression and med-
ical problems and addresses two general questions: First, is reduced physical activity
responsible in part for associations between depression and the medical conditions
and disabilities discussed in this book? Second, can increases in physical exercise
improve the mood of patients and, therefore, be used in the management of these
medical problems? The chapter begins with a discussion of the associations between
physical activity, depression and depressed mood in clinical and population stud-
ies. There are many studies documenting more positive moods in active compared
with inactive people, but it is important to assess the strength of the evidence and
to rule out the contribution of underlying or unmeasured psychosocial or clinical
factors. The chapter goes on to evaluate the impact of changes in depression and
physical activity. Such changes can be evaluated in two ways: assessing the effects of
treating depression on physical activity levels, and assessing the influence of physical
training interventions on depression. This field has been beset with methodological
problems, some of which are outlined here. Evaluations of physical training inter-
ventions in physically healthy populations are followed by considering the evidence
that increased physical activity might alleviate depression in patients with medical
problems. Finally, the chapter outlines some of the psychobiological mechanisms
that might link physical activity, depression and pathophysiological processes in
human disease.
This chapter is not intended to review the evidence that physical activity and
exercise training are beneficial for health. There are many settings in which positive
effects have been established rather convincingly, such as in cardiac rehabilita-
tion, pain management and the postponement of disability [4,5]. Such findings are
important but may have nothing to do with depression; rather, they have direct
effects on musculoskeletal or aerobic function. It is probable that there are recipro-
cal relationships between physical activity and depression: depressed mood leads
to inactivity, while in some circumstances exercise can alleviate depression. It is
important to try to establish whether both of these pathways operate, but ulti-
mately it may be very difficult in many clinical settings to establish the primacy of
one factor over the other. This chapter suggests that understanding of the potential
role of exercise training in the management of depressed mood in medical patients
requires an evaluation of converging evidence from different types of study: popula-
tion and clinical observational studies, randomised trials and experimental studies.
The nature of associations varies with different clinical conditions, and so there
is no definitive study that can answer all the questions. It must also be borne in
mind that vigorous physical activity may have adverse as well as beneficial effects
in medical conditions and is not entirely risk-free. There is evidence, for exam-
ple, of a small but significant risk of acute cardiac events during vigorous exercise,
particularly in normally sedentary individuals [6]. Physical exercise can also cause
musculoskeletal problems. It is important therefore to establish the precise nature
and the limits of evidence for effectiveness.
Research concerning physical activity has been carried out both in relation to
depressive symptoms in the general population and with clinically depressed indi-
viduals. Studies of depressive symptoms have typically assessed mood disturbances
with questionnaires such as the Beck Depression Inventory (BDI) and the Center for
Epidemiologic Studies Depression Scale (CES-D), while clinical studies use diag-
nostic interviews and expert assessments to identify severe depression and major
depressive disorder. Both types of study are discussed in this chapter. The term
‘positive mood’ is used in this chapter to indicate states of subjective wellbeing and
vigour, while adverse states such as anxiety and depression are described as ‘negative
mood’.
350 Depression and physical illness
the Strengths and Difficulties Questionnaire [16] were associated inversely with
vigorous physical exercise in boys and related positively to sedentary behaviours
such as television watching in girls, independently of ethnicity, socioeconomic and
family background, self-rated health, body mass index (BMI) and environmen-
tal factors such as the availability of local sports facilities. Studies of adolescents
are interesting, since at these ages much physical activity is not a matter of per-
sonal choice but is determined by school policies and parental encouragement. The
selection biases caused by personal enjoyment of sports are thereby reduced still
further.
Longitudinal studies
The population studies described above have applied extensive statistical controls to
bolster the argument that associations between lower levels of depressive symptoms
and greater physical activity are not due to confounding factors and co-variates.
However, cross-sectional studies cannot tease out causal relationships, and it is
not clear whether greater activity reduces depressive symptoms or greater levels of
depression lead to inactivity, or whether the two factors are simply correlated. A
number of longitudinal population studies have therefore been conducted, assessing
whether physical activity at baseline predicts later depressive symptoms independ-
ently of current depression. The results for adults are summarised in Table 16.1.
Several studies have documented a positive association between exercise and later
depressive symptoms, but results have been quite mixed. Two analyses from the
Alameda County Study showed positive associations, although in one of these the
impact of physical activity on later depressive symptoms was no longer significant
after adjustment for all relevant co-variates [17,18]. An eight-year follow-up of the
National Health and Nutrition Examination Survey (NHANES) 1 showed positive
findings, but curiously not in men who had low levels of depressive symptoms at
baseline or in women with high depressive symptoms at baseline [19]. The studies
of cohorts aged 65 years or older have all found that physical activity predicts
later depressive symptoms over four to eight years independently of confounding
factors, although the magnitude of effects has been small in some cases [20–22].
Some other studies of healthy participants have not found associations between
baseline physical exercise and subsequent depression [23,24].
Particularly relevant in the context of this book is the analysis of the Medical
Outcomes Study [25]. The adult participants in this study had one or more of the
following conditions: diabetes, hypertension, congestive heart failure, recurrent
myocardial infarction, severe depressive symptoms or current depressive disorder.
Over the two-year observational period, baseline levels of physical activity were
associated inversely with later depressive symptoms across all medical conditions,
with particularly strong effects among diabetics.
Table 16.1 Longitudinal population studies of physical activity and depressive symptoms
Duration of
follow-up
Ref. Sample (years) Depression measure Physical activity measure Baseline PA predicts later depression?
Camacho 1947 men and women aged 9 Specially constructed General physical activity Yes, but not significant after adjustment
et al. [17] over 20 years, USA questionnaire for all co-variates
Farmer et al. [19] 1900 men and women aged 8 CES-D Recreational physical activity Yes, in women with low depression and
25–77 years, USA men with high depression at baseline
Kritz-Silverstein 944 men and women aged 8 BDI Regular strenuous exercise No
et al. [23] 50–89 years, USA
Lampinen 663 men and women aged 8 Short BDI General physical activity Yes, for walking but not strenuous
et al. [20] 65–84 years, Finland exercise
Mobily et al. [21] 2084 men and women aged 3 CES-D Walking Yes
65–101 years, USA
Morgan and Bath 1042 men and men aged 4 and 8 HAD Outdoor/leisure activity Yes
[22] 65+ years, UK
Stewart et al. [25] 1758 men and women 2 MOS measure General physical activity Yes
mean age 56.1 years, USA
Strawbridge 1947 men and women aged 5 Interview General physical activity Yes
et al. [18] 50–94 years, USA
Weyerer [24] 1536 men and women aged 5 CIS Exercise for sport No
15+ years, Germany
BDI, Beck Depression Inventory; CES-D, Center for Epidemiologic Studies Depression Scale; CIS, Clinical Interview Schedule; HAD, Hospital Anxiety and
Depression Scale; MOS, Medical Outcomes Survey; PA, physical activity.
353 Depression and physical activity
One longitudinal study that has not been included in Table 16.1 was a follow-up
of former medical students from Johns Hopkins’ University, since the depression
outcome was self-reported clinical depression [26]. The Reno Diet-Heart Study
has also been excluded, even though it described an association between changes
in physical activity and changes in depressive symptoms over five years, because
it involved a non-validated measure of activity [27]. Only a single study to date
has investigated these associations in adolescents. Motl et al. [28] assessed more
than 4000 adolescents who provided data three times over an 18-month period.
Increases in exercise levels were associated with reductions in depressed mood,
which were independent of gender, socioeconomic position, smoking and alcohol
consumption.
The longitudinal data from population studies therefore provide moderate sup-
port for the notion that regular exercise protects against the development of depres-
sive symptoms. It is unfortunate that none of these investigations has tested the
reverse proposition that depressive symptoms reduce subsequent physical activity.
As noted in Chapter 6, depressive symptoms predict incident disability and deteri-
oration in activities of daily living in older people, but relationships in younger
groups have not been studied extensively. There are also sufficient discrepancies in
observational studies to suggest that the effects of physical activity on depressive
symptoms are quite modest and may be swamped by other factors.
More precise types of investigation involving randomised control trials are there-
fore required to clarify the relationship between depressive symptoms and physical
activity. Two types of intervention trial are theoretically relevant: studies that exam-
ine the effects on physical activity of treating depression, and studies of the effects
on depression of modifying exercise levels.
treatment of depression, it is not certain whether this would indicate a specific asso-
ciation, since increased physical activity might be part of general re-engagement
with everyday life. Thus, despite the theoretical possibility that studies of antidepres-
sant treatment would throw light on associations with physical activity, in practice
the inferences that can be drawn are rather limited.
dropped out. Physical activity studies are notorious for high drop-out rates. This
introduces a selection bias, since the individuals who complete the protocol may be
those who have benefited psychologically, while those who drop out have probably
not experienced positive changes. There is good evidence that clinical effectiveness
is overestimated if analysis is restricted to people who have completed all aspects of
the protocol [36]. Additionally, if drop-outs are excluded from analyses, then the
allocation of participants to conditions is no longer random and a false-positive
impression will emerge. With few exceptions [5,37–40], most studies of physical
activity and depression have limited analysis to people who completed the study.
Another methodological problem from the clinical trials perspective is the use of
non-blinded assessment. Ideally, the researchers who measure outcomes should not
be aware of which treatment the participant has undergone, otherwise their own
biases and expectations might influence results. Studies that rely on self-reporting
assessment of depressive symptoms inevitably do not satisfy this criterion, and few
trials have involved measurements of depression from clinicians blind to group
allocation.
A third set of limitations relates to the representativeness of people who partici-
pate in randomised trials. Studies of physical activity involve substantial personal
commitment from volunteers, and only a proportion of individuals with appro-
priate levels of depressive symptoms will agree to participate or be randomised.
Even in the best trials, the results may therefore be relevant only to the types of
people who agree to take part in studies and not to clinical patients or depressed
individuals in the population more generally.
The challenges inherent in conducting methodologically rigorous studies are
illustrated by the Depression Outcomes Study of Exercise (DOSE), a trial designed
to assess dose–response antidepressant effects of exercise in young adults with mild
or moderate major depression [41]. The study involved a comparison of low and
high ‘doses’ of physical exercise three or five times per week for 24 weeks, with
an attention-placebo condition of very-low-intensity exercise. A rigorous selection
procedure was used in recruitment to ensure that participants were sedentary, in
the appropriate range for depression, not taking medication or undergoing psy-
chotherapy and not suffering from substance or alcohol abuse. Statistical power was
estimated with analysis on an intention-to-treat basis, and it was decided that each
of the four treatment groups required 18 participants completing the programme,
with 12 in the control group. In order to achieve this sample size, no fewer than
1664 patients were screened. More than 1500 were excluded because they were tak-
ing medication or undergoing psychotherapy (even though still depressed), were
too active, did not meet depression criteria, changed their minds about partici-
pating or some other reason. This represents an enormous expenditure of effort
and professional resources on a comparatively small number of participants. This
356 Depression and physical illness
experience is not unique. A study in Scotland of older adults who had responded
poorly to standard antidepressant treatment involved evaluation of 1885 patients
in order to recruit 86 study participants [38]. Clearly, a dedicated and well-funded
research programme is required in order to undertake such a project. The question
also arises as to how generalisable the results will be beyond the 5% of patients who
finally entered the study.
Literature on the effects of physical activity interventions on depressive symp-
toms has been reviewed a number of times, with conclusions ranging from the
cautious to the very positive, depending on how many of these methodological
factors are emphasised. Lawlor and Hopker [33] carried out a meta-analysis of
randomised trials published up to 2000, involving clinically depressed patients or
adults in the community with high levels of depressive symptoms. They noted
that trials were typically small, with only a single study involving more than
100 patients; most involved no follow-up beyond the active intervention period. The
standardised mean difference in effect sizes was −1.1 (95% confidence interval (CI)
−1.5 to −0.6), favouring physical activity compared with no treatment. This corres-
ponded to a mean difference of just over seven points on the BDI. The comparisons
of physical activity with established treatments such as cognitive or pharmaco-
therapy showed no differences between conditions. There was no convincing
evidence at that time of a dose–response relationship between physical activity and
depression [42]. However, newer studies suggest that this summary may be overly
pessimistic.
Illustrative studies
This chapter describes in more detail three methodologically sound studies to illus-
trate the impact of increasing physical activity on depressive symptoms. Singh et al.
[39,43] randomised 32 community-dwelling adults aged over 60 years who had
elevated BDI scores to exercise and control groups. Exercise training involved high-
intensity resistance training of upper- and lower-body muscle groups three days
per week for ten weeks, followed by a further ten weeks of unsupervised exercise
at home or at a health club. The control group attended a series of ten weekly
health-education lectures. Both groups received weekly telephone calls from the
investigators to provide matched attention levels, and depression was not men-
tioned to participants in either condition as an endpoint of the study. At the end of
the 20-week intervention, there was a significantly greater decrease in BDI scores
in the physical exercise group. Some 73% of those randomised to exercise training
had BDI scores below the threshold for probable clinical depression, compared with
36% of health-education controls. Participants were reassessed after 26 months, and
differences between groups were still present. Interestingly, the reduction in BDI
was marginally greater in those individuals who continued to be active during
357 Depression and physical activity
the follow-up period, compared with those who stopped exercising. This study is
notable because it involved an elderly population (aged 60–84 years) and tested
strength and resistance training rather than the more commonly studied aerobic
activity.
A second study with important findings involved 156 men and women with
major depressive disorder who were randomised to aerobic training or medication
with setraline (an SSRI), or a combination of the two [37]. Exercise training was car-
ried out in three supervised sessions per week for 16 weeks, and was based primarily
on walking and jogging. Intention-to-treat analyses were incorporated, and psy-
chological state was measured by assessors blind to treatment allocation. Although
pharmacologically treated patients showed more rapid therapeutic responses, there
were no differences between the conditions in the reductions in depression achieved
at the end of 16 weeks. However, a further evaluation six months after the end of
the trial showed different results [44]. Interview-based assessments indicated that
clinical depression was present in only 30% of participants in the exercise training
group, compared with 52% in the medication and 55% in the combination condi-
tion. Controlling for age, gender, baseline depression and the use of antidepressant
medication, the odds of major depressive disorder at six months post-treatment
were 0.49 (C.I 0.32 to 0.74) among individuals who were exercising for 50 minutes
or more per week compared with those exercising below this level.
The results of the DOSE study mentioned earlier were published in 2005 [40].
Data were analysed both on an intention-to-treat basis (including all enrolled
participants) and as an efficacy trial, including only those who completed the trial.
Major depressive disorder was assessed using the Hamilton Rating Scale for Depres-
sion (HRSD), completed by trained observers who were blind to the treatment con-
dition. As noted earlier, the study assessed the impact of lower- and higher-intensity
physical activity performed either three or five times per week. All physical activity
was carried out under supervision in the laboratory in order to ensure that the
correct intensity was applied. The results indicate that reductions in depression
were stimulated by exercise, with the intensity rather than frequency being more
important. Two of the outcomes are summarised in Figure 16.1. Figure 16.1a shows
the HRSD scores at the end of the 12-week programme. The groups did not differ
at baseline, and so it is apparent that greater average reductions were observed with
higher intensity exercise. Figure 16.1b shows the proportion of participants with a
decrease in HRSD score of at least 50%, where again the intensity effect is apparent.
Follow-up analyses have not yet been published.
The results of these and other studies suggest that physical activity training
does have beneficial effects on clinical depression and depressive symptoms in
otherwise healthy adults, although corroboration from additional well-controlled
trials is certainly needed. There is evidence in older adults at least that strength and
358 Depression and physical illness
20 50
HRSD post-treatment
16 40
8 20
4 10
0 0
Placebo Lower Higher Placebo Lower Higher
intensity intensity intensity intensity
(a) (b)
Figure 16.1 (a) Mean scores on the Hamilton Rating Scale for Depression (HRSD) following treatment
in the control, lower-intensity and higher-intensity training conditions of the Depression
Outcomes Study of Exercise (DOSE) study. Error bars are standard deviations. (b) Proportion
of participants in each condition who achieved at least a 50% reduction in HRSD scores from
pre- to post-treatment. For details, see Dunn et al. [40].
resistance training and aerobic exercise have positive effects, and the impact of all
these programmes appears to depend on maintaining regular exercise patterns.
were reported over the study, the results were inconclusive. Depressive symptoms
were assessed with three measures, but only one showed significant differences
between groups. Depressive symptoms on the Derogatis Symptom Checklist (SCL-
90) decreased for the combined exercise, education and stress-management con-
dition, and not in the education and stress-management group. However, SCL-90
depressive symptoms also fell in the waiting-list control group. In a subsequent
report, patients who continued to exercise one year later were compared with those
who did not exercise, and differences on the SCL-90 were again observed [50].
However, the randomised element was no longer present in these analyses.
Another condition in which the effects of exercise interventions have been studied
is human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome
(AIDS). Aerobic training has been advocated for its effects on immune function,
viral load and cardiorespiratory fitness, and a systematic review has been published
in the Cochrane Library [51]. Unfortunately, few of these studies have involved
systematic assessments of depressive symptoms, but some researchers have shown
favourable effects of aerobic exercise training on psychological wellbeing in this
group [52].
The largest, and one of the best conducted, studies of exercise training and depres-
sion in a medical condition is the trial of patients with osteoarthritis published by
Penninx et al. [5] (see Chapter 6). More than 400 patients were randomised to a
health-education control, or resistance training involving strengthening exercises,
or aerobic training with a walking-based programme. Intention-to-treat ana-
lysis of depressive symptoms assessed with the CES-D was carried out at 3, 9 and
18 months. Aerobic exercise led to significantly greater reductions in depressive
symptoms than did control, after adjustment for age, race, gender, education, base-
line disability and baseline depression. Resistance training was no more effective
than control. Effects were observed in patients with both high and low depres-
sion scores at baseline, although the absolute effects were larger in those who were
initially more depressed. Aerobic exercise also produced lower disability ratings
and pain at follow-up. Subsequent analyses suggested that the positive effects of
aerobic exercise on depression were mediated partly through changes in pain and
disability.
Exercise training is an important part of rehabilitation for many painful con-
ditions. However, the fact that aerobic exercise had positive effects on depression
in Penninx and colleagues’ study does not mean that it is necessarily helpful in
other painful conditions. Negative results have been published for problems such
as fibromyalgia [53] and chronic neck pain [54]. It is necessary, therefore, to evaluate
effects separately in different pain syndromes.
There has been considerable interest in the impact of exercise training on the
subjective wellbeing of patients with cancer [55]. Unfortunately, many of the
360 Depression and physical illness
A third issue that has more practical treatment implications is whether increasing
physical activity has a positive effect on health because it leads to improvement in
depressive symptoms. The evidence summarised in previous sections indicates that
physical exercise interventions can impact favourably on psychological wellbeing
in patients with chronic health problems. It is less certain, however, whether this
antidepressive effect of physical activity is responsible for the more general health
benefits of exercise programmes. One study suggests that this mechanism is not
very important, at least as far as coronary heart disease and mortality are concerned.
Blumenthal et al. [61] followed up over 2000 men and women for up to four years
after acute myocardial infarction. Self-reports of physical activity were obtained six
months after the cardiac admission. It was found that patients who said that they
exercised regularly had a reduced risk of dying, with a hazard ratio of 0.69 (95%
CI 0.47 to 0.97) in comparison with non-exercisers, after adjusting for co-variates
that contribute to mortality. It was also found that people who exercised had lower
depression scores on the BDI at the time of infarction and greater reductions in
depression between infarction and six months. These results are consistent with
physical activity having effects on both cardiac health and depression. However,
adding depression to the statistical model did not alter the association between regu-
lar exercise and mortality, and so there was no evidence that depression mediated
the beneficial effects of exercise.
A further possibility is that shared biological processes underpin some of the
associations between physical exercise, depression and physical illness. Increasing
physical activity levels might have positive effects on biological responses that are
implicated in disease pathology, which are in turn also related to depression. If this
were the case, then depression, physical activity and illness might be linked through
common biological mechanisms, without depression playing either a causal or a
mediating role. Three possible pathways are outlined below.
Neuroendocrine function
The relationship between hormones such as cortisol and dehydroepiandrosterone
(DHEA) and depression was reviewed in Chapter 13. Acute exercise increases not
only circulating glucocorticoid levels but also tissue sensitivity to glucocorticoids
[72]. This is thought to be a protective response designed to limit muscle inflam-
matory responses and to reduce exercise-induced muscle damage. Exercise training
increases circulating levels of DHEA sulphate (DHEAS) at rest, while endurance-
trained individuals show a dampened cortisol response to physical activity [73,74].
[79]. Acutely, stress reactivity is reduced in the period following episodes of stren-
uous exercise [80]. This could be a further mechanism linking regular physical
exercise, depression and disease risk.
Conclusions
Discussion of physical exercise, depression and health generates strong feelings on
both sides of the argument. Many advocates of regular physical activity regard the
issue as settled, while sceptics are much more cautious about the interpretation
of observational and intervention studies. There is certainly evidence that increas-
ing physical exercise in people willing to become more active leads to reductions in
depression mood and that these effects may be present both in patients with medical
disorders and in healthy individuals. There are also plausible biological mechanisms
linking physical activity with pathways relevant to both depression and physical ill-
ness. The mechanisms reviewed here are those that might specifically relate physical
activity with pathophysiology. There is another literature concerned with the psy-
chological processes through which activity influences depressive states, involving
factors such as self-efficacy, self-esteem, distraction and coping; this has not been
discussed since it does not impinge directly on biological responses [34].
In many circumstances, increasing physical activity is likely to have benefits for
health, regardless of whether it influences mood. But there is not yet compelling
evidence in most illness groups that relief of depression through promotion of
physical activity will have direct effects on disease states. Nevertheless, research in
this field is continually developing, so if and when such effects are demonstrated
they will undoubtedly have important implications for the management of many
physical illnesses.
Acknowledgements
I am grateful to Mark Hamer for his comments on an earlier draft of this chapter.
This research was supported by the British Heart Foundation.
REFERENCES
1. V. B. Hammett, Psychological changes with physical fitness training. Can. Med. Assoc. J. 96
(1967), 764–9.
2. C. H. Folkins, W. E. Sime, Physical fitness training and mental health. Am. Psychol. 36 (1981),
373–89.
3. P. D. Thompson, D. Buchner, I. L. Pina, et al., Exercise and physical activity in the preven-
tion and treatment of atherosclerotic cardiovascular disease a statement from the Council
on Clinical Cardiology (Subcommittee on Exercise, Rehabilitation, and Prevention) and the
364 Depression and physical illness
55. K. S. Courneya, Exercise in cancer survivors: an overview of research. Med. Sci. Sports Exerc.
35 (2003), 1846–52.
56. L. M. Oldervoll, S. Kaasa, M. J. Hjermstad, J. A. Lund, J. H. Loge, Physical exercise results
in the improved subjective well-being of a few or is effective rehabilitation for all cancer
patients? Eur. J. Cancer 40 (2004), 951–62.
57. K. S. Courneya, C. M. Friedenreich, R. A. Sela, et al., The group psychotherapy and home-
based physical exercise (group-hope) trial in cancer survivors physical fitness and quality of
life outcomes. Psychooncology 12 (2003), 357–74.
58. A. J. Daley, N. Mutrie, H. Crank, R. Coleman, J. Saxton, Exercise therapy in women who have
had breast cancer: design of the Sheffield women’s exercise and well-being project. Health
Educ. Res. 19 (2004), 686–97.
59. R. Belardinelli, D. Georgiou, G. Cianci, A. Purcaro, Randomized, controlled trial of long-term
moderate exercise training in chronic heart failure: effects on functional capacity, quality of
life, and clinical outcome. Circulation 99 (1999), 1173–82.
60. R. van den Berg-Emons, A. Balk, H. Bussmann, H. Stam, Does aerobic training lead to a
more active lifestyle and improved quality of life in patients with chronic heart failure? Eur.
J. Heart Fail. 6 (2004), 95–100.
61. J. A. Blumenthal, M. A. Babyak, R. M. Carney, et al., Exercise, depression, and mortality after
myocardial infarction in the ENRICHD trial. Med. Sci. Sports Exerc. 36 (2004), 746–55.
62. R. J. Shephard, Cytokine responses to physical activity, with particular reference to IL-6:
sources, actions, and clinical implications. Crit. Rev. Immunol. 22 (2002), 165–82.
63. R. J. Shephard, Adhesion molecules, catecholamines and leucocyte redistribution during and
following exercise. Sports Med. 33 (2003), 261–84.
64. E. S. Ford, Does exercise reduce inflammation? Physical activity and C-reactive protein among
US adults. Epidemiology 13 (2002), 561–8.
65. D. B. Panagiotakos, C. Pitsavos, C. Chrysohoou, S. Kavouras, C. Stefanadis, The associations
between leisure-time physical activity and inflammatory and coagulation markers related to
cardiovascular disease: the ATTICA Study. Prev. Med. 40 (2005), 432–7.
66. J. K. Smith, R. Dykes, J. E. Douglas, G. Krishnaswamy, S. Berk, Long-term exercise and
atherogenic activity of blood mononuclear cells in persons at risk of developing ischemic
heart disease. J. Am. Med. Assoc. 281 (1999), 1722–7.
67. J. P. LeMaitre, S. Harris, K. A. Fox, M. Denvir, Change in circulating cytokines after 2 forms
of exercise training in chronic stable heart failure. Am. Heart J. 147 (2004), 100–105.
68. A. J. Broadley, A. Korszun, C. J. Jones, M. P. Frenneaux, Arterial endothelial function is
impaired in treated depression. Heart 88 (2002), 521–3.
69. U. Laufs, S. Wassmann, T. Czech, et al., Physical inactivity increases oxidative stress, endothe-
lial dysfunction, and atherosclerosis. Arterioscler. Thromb. Vasc. Biol. 25 (2005), 809–14.
70. R. Hambrecht, A. Wolf, S. Gielen, et al., Effect of exercise on coronary endothelial function
in patients with coronary artery disease. N. Engl. J. Med. 342 (2000), 454–60.
71. D. G. Edwards, R. S. Schofield, S. L. Lennon, et al., Effect of exercise training on endothelial
function in men with coronary artery disease. Am. J. Cardiol. 93 (2004), 617–20.
72. M. Duclos, C. Gouarne, D. Bonnemaison, Acute and chronic effects of exercise on tissue
sensitivity to glucocorticoids. J. Appl. Physiol. 94 (2003), 869–75.
368 Depression and physical illness
73. K. Aizawa, T. Akimoto, H. Inoue, et al., Resting serum dehydroepiandrosterone sulfate level
increases after 8-week resistance training among young females. Eur. J. Appl. Physiol. 90
(2003), 575–80.
74. M. S. Tremblay, J. L. Copeland, W. Van Helder, Effect of training status and exercise mode
on endogenous steroid hormones in men. J. Appl. Physiol. 96 (2004), 531–9.
75. K. C. Light, R. V. Kothandapani, M. T. Allen, Enhanced cardiovascular and catecholamine
responses in women with depressive symptoms. Int. J. Psychophysiol. 28 (1998), 157–66.
76. J. R. Jennings, T. W. Kamarck, S. A. Everson-Rose, et al., Exaggerated blood pressure responses
during mental stress are prospectively related to enhanced carotid atherosclerosis in middle-
aged Finnish men. Circulation 110 (2004), 2198–203.
77. F. A. Treiber, T. Kamarck, N. Schneiderman, et al., Cardiovascular reactivity and development
of preclinical and clinical disease states. Psychosom. Med. 65 (2003), 46–62.
78. A. Steptoe, M. Marmot, Impaired cardiovascular recovery following stress predicts 3-year
increases in blood pressure. J. Hypertens. 23 (2005), 529–36.
79. T. W. Spalding, L. A. Lyon, D. H. Steel, B. D. Hatfield, Aerobic exercise training and cardio-
vascular reactivity to psychological stress in sedentary young normotensive men and women.
Psychophysiology 41 (2004), 552–62.
80. A. Steptoe, N. Kearsley, N. Walters, Cardiovascular activity during mental stress following
vigorous exercise in sportsmen and inactive men. Psychophysiology 30 (1993), 245–52.
17
17
Depression and adherence to medical advice
Douglas A. Raynor, Rena R. Wing and Suzanne Phelan
The coexistence of depression and medical illness has been the focus of a great deal
of attention in the past several years. Depression appears to occur in at least 25% of
medical patients, although estimates vary based on measurement criteria, type and
stage of medical disease, and level of depressive severity [1,2]. Depression has also
been associated with poorer health outcomes in these diseases. Depression may
influence the development and progression of diseases directly via physiological
pathways or indirectly via behavioural pathways [3].
Adherence to treatment regimens is one possible behavioural pathway by which
depression affects physical health. Adherence is the degree to which a person’s
behaviour (e.g. taking medications, attending treatment sessions, executing lifestyle
changes) coincides with medical or health advice [4]. Adherence is critical for the
prevention and treatment of medical diseases. Unfortunately, adherence is less than
optimal across a wide range of medical regimens [5]. Individuals who are depressed
may be less likely to adhere to various aspects of their treatment regimens, and this
lack of adherence may produce poorer health outcomes.
There is a large literature on predictors of adherence. A substantial number of
variables have been associated with adherence, but findings across studies have been
inconsistent [6]. Depression has emerged as one of the few promising predictors
of treatment adherence among medical patients [7]. There are many reasons to
hypothesise that depression may negatively influence an individual’s willingness
and capacity to adhere to a treatment regimen. For instance, depression tends to
reduce concentration, energy, motivation, social support and positive expectations,
which in turn may lead to poor treatment adherence. The purpose of this chapter is
to discuss the research evaluating the relationship between depression and adher-
ence to medical regimens. We consider methodological issues in the assessment of
adherence and depression and then review the empirical studies that have examined
whether depressed individuals have poorer treatment adherence.
Depression and Physical Illness, ed. A. Steptoe.
Published by Cambridge University Press. C Cambridge University Press 2006.
370 Depression and physical illness
Figure 17.1 Model showing adherence as the mediator of the relationship between depression and
health outcome.
Although more advanced methods of measuring food intake are currently lim-
ited, several improved measures of exercise and smoking have been developed. For
instance, doubly labelled water, pedometers and accelerometers are available to
measure physical activity [20]. Expired carbon monoxide and salivary cotinine are
two relatively accurate measures of smoking cessation [21]. Each of these measures
has advantages and disadvantages in terms of accuracy and validity. However, due
to the high costs of these more advanced measures, self-reporting continues to be
the most commonly used measure in the adherence literature.
This state of affairs has important implications for evaluating the literature on the
effects of depression on adherence. Dunbar-Jacob and colleagues [7] compared a set
of predictors across three measurement methods: electronic event monitoring, 24-
hour recall and an interview reviewing the previous month. When using 24-hour
recall and the interview, psychosocial characteristics did not predict adherence;
when using electronic event monitoring, social support and pain significantly pre-
dicted adherence. Dunbar-Jacob et al. [7] suggested that ‘the electronic event mon-
itors cause us to reexamine what we know about predictors of adherence. Our data
suggest that many of the findings on adherence predictors may be related to the
measurement method rather than the actual behaviour’ (p. 110). Given that any sin-
gle adherence measure is limited in some way, the use of multiple measures within
studies (e.g. electronic monitoring, biological assay) would provide the most accu-
racy and facilitate greater understanding of predictors of adherence. Confidence in
the observed association between depression and treatment adherence may thus
depend on the type and number of adherence measures.
Another problem in this field has been the tendency to dichotomise adherence.
Patients who are taking 80% or more of their prescribed medication are often
considered ‘adherent’ whereas those taking below 80% are deemed ‘non-adherent’.
The decision of which criterion to select for the cut-off point is typically subject-
ive. Moreover, such a criterion does not allow investigators to examine temporal
distributions of adherence. An alternative is to define adherence to chronic disease
regimens along a continuum, which avoids the use of arbitrary cut-off scores,
maximises statistical power and facilitates the comparison of adherence rates across
a range of behaviours, diseases and studies. Still, since relatively complex non-
parametric strategies may be necessary in order to take advantage of the entire range
of non-normal adherence data [7], the use of dichotomisation is omnipresent in
the adherence literature.
clinical disorders and their symptoms [22]. Different definitions of depression have
been used throughout the literature on the relationship between depression and
adherence.
Mental health professionals have developed a taxonomic classification system
that is codified in the Diagnostic and Statistical Manual of Mental Disorders (DSM).
In the current version (DSM-IV) [23], major depressive disorder (MDD) is the most
severe unipolar affective disorder. Episodes of major depression are characterised
by the presence of depressed mood or anhedonia ‘most of the day, nearly every
day’, persisting for at least two weeks and accompanied by at least four additional
symptoms. Although the DSM-IV stipulates a minimum duration of two weeks,
patients often experience a depressive episode lasting for at least four to six months
[24]. Additionally, recurrence of episodes is common among individuals with
MDD.
Dimensional approaches to defining depression primarily rely on rating scales
of depressive symptoms, such as the Beck Depression Inventory (BDI) [25] and
the Center for Epidemiological Studies–Depression Scale (CES-D) [26]. In general,
these instruments are composed of a standardised series of statements or ques-
tions based upon typical symptoms of depression. These instruments are dimen-
sional, given that their scoring system consists of summing items into a total score
that reflects the overall severity of depressive symptoms along a continuum [27].
Advocates of the dimensional perspective argue that there is a linear relationship
between mild, moderate and severe forms of depression [28]. Although individuals
with MDD usually have high scores along the dimension of depressive symptoma-
tology, individuals with milder forms of depression may also have relatively high
scores due to a period of distress or negative mood. In contrast to major depressive
episodes, episodes of depression in non-clinical populations are thought to be more
short-lived, with periods lasting hours or days instead of months [22]. The differ-
ing lengths of depression that are derived from the taxonomic and dimensional
approaches may be an important factor in the interpretation of the depression-
treatment adherence literature.
There are two predominant approaches to measuring depression: interview and
self-report. Structured and semi-structured interviews have been developed for
diagnosing depressive disorders and other psychopathological conditions. Exam-
ples of structured diagnostic interviews include the Schedule for Affective Disorders
and Schizophrenia (SADS) [29], the Diagnostic Interview Schedule (DIS) [30] and
the Structured Clinical Interview for DSM-IV [31]. These interviews are composed
of standardised criteria that increase the reliability and generalisability of research
findings [32]. Another important attribute of the interview approach is that exam-
iners can differentially diagnose seemingly overlapping disorders. Furthermore,
reporting biases by patients are minimised with this approach. A drawback of the
interviews is their extended duration and expense of administration. In fact, despite
374 Depression and physical illness
the relative advantages of the interviews, the high costs result in their relatively
infrequent use, particularly in cases where psychopathology is not the primary
focus of research.
Patient rating scales are the most common approach for measuring depression in
the medically ill, largely due to their inexpensiveness and the ease of administration
and scoring. Their low cost permits administration to a wide range of patients. This
method is also useful for tracking changes of depressive symptoms over time. One
of the problems with these instruments is that for the most part they were originally
developed for use in specific non-medical populations. For instance, the CES-D was
developed for use in non-psychiatric populations in the community, and the BDI
was developed to measure depression severity in populations previously diagnosed
with major depression [33]. As a result, the instruments often possess adequate
sensitivity but relatively poor specificity. Additionally, given that all but a select few
of the patient rating scales are collected via patient self-reporting, these measures
are subject to accuracy problems. For instance, depressed patients often experience
cognitive dysfunction, which may interfere with the patient’s ability to understand
and respond accurately to scale items [34].
An inherent challenge in the measurement of depression among medical
patients is the confounding of somatic and neurovegetative symptoms across the
psychological and medical conditions. It is difficult to determine reliably whether
depressive symptoms in a medically ill patient are a function of the psychopatho-
logical condition or the disease itself, or both. Specific neurovegetative and somatic
symptoms that may overlap include fatigue, sleep disturbance, appetite/weight
change, loss of libido, and psychomotor retardation.
An additional complication in measuring depression is the problem of psychiatric
comorbidity. Depression often coexists with other psychiatric conditions. Among
individuals experiencing MDD in the community, 56% were diagnosed with one or
more other psychiatric disorders [35]. Anxiety symptoms and disorders are the most
commonly reported comorbidity. The fact that depression and anxiety share many
symptoms, such as sleep disturbance, worry, guilt and poor concentration, makes
differential diagnosis quite difficult. Without controlling for anxiety, an observed
association between depression and treatment adherence may actually be due to
the effects of anxiety on adherence.
Thus, in evaluating the literature on depression and adherence, it is important to
consider the approaches that have been used to measure each of these constructs.
Articles were identified through electronic database searches and the ancestry
approach. First, the PsychLit and MedLine electronic databases were searched from
1968 through 2004. Keywords included ‘patient adherence’ or ‘patient compliance’
and ‘medical’ or ‘disease’. Second, references from review or empirical articles were
reviewed for applicability. Studies were included in this review if the following cri-
teria were met: (i) adults diagnosed with a chronic medical disease; (ii) patients
prescribed medical or health treatment involving some type of behaviour, such as
pharmacotherapy, diet, exercise, smoking cessation, behavioural self-monitoring
(e.g. checking blood glucose levels) or attending a medical intervention (e.g.
chemotherapy); and (iii) measures of depression and treatment adherence were
included. Forty-one studies met these criteria and are described in Table 17.1
[36–76].
There were substantial differences across the 41 studies in terms of research loca-
tion, research design, medical diseases and measures of depression and treatment
adherence. As a result, there was no common denominator for computing effect
sizes that would facilitate quantitative comparison of results across studies. Thus,
Table 17.1 simply indicates whether there was a significant association between
depression and adherence. A negative correlation would support the hypothesis
that higher levels of depression are associated with poorer levels of adherence.
‘Mixed’ correlations occur when depression is associated negatively with some but
not all adherence measures.
Measure of
Study Sample/design information Measure of adherence depression Results
Ades et al. 226 patients with recent MI or coronary Participation in a cardiac Unspecified Participation in cardiac rehabilitation
[36] bypass surgery; 57% male; mean age 70 rehabilitation programme self-report questions associated with less depression before
years, cross-sectional design hospitalisation, but not with depression
while hospitalised
Avants et al. 43 patients with HIV, opioid dependence Weekly structured interview of BDI Depression associated negatively with
[37] and cocaine abuse; 69% male; mean age medication adherence medication adherence; a composite score of
41 years; 48% African American, 38% ‘emotional functioning’, including
white, 14% Hispanic; prospective design depression, negatively associated with
(4 weeks) medication adherence in multiple regression
analysis
Ayres et al. 74 women with breast cancer, 77% African Attendance at chemotherapy Affect Balance Scale Depression associated positively with
[38] American; prospective design (4–6 months) appointments chemotherapy adherence
Blumenthal 35 patients with MI; mean age 54 years; 91% Attendance at exercise sessions MMPI Depression Depression associated negatively with
et al. [39] male; prospective design (1 year) Scale exercise adherence
Bosley et al. 102 patients with asthma; age 18–70 years; Turbohaler inhalation HADS Depression but not anxiety associated
[40] prospective design (12 weeks) computers negatively with medication adherence
Bosley et al. 93 patients with COPD; age 45–77 years; Dataloggers HADS Depression and anxiety not associated with
[41] prospective design (4 weeks) medication adherence
Botelho and 59 elderly patients with ≥2 chronic diseases; Home-based pill counts by BDI Depression not associated with medication
Dudrak [42] age 65+ years; prospective design (8 weeks) staff adherence
Carney et al. 55 patients with CAD; mean age 69 years; Electronic blister medication Diagnostic Interview Depression associated negatively with
[43] 75% male; prospective design (3 weeks) packets Schedule with medication adherence
DSM-III-R
Carney et al. 92 patients with CAD and MI; mean age Electronic blister medication BDI Depression not associated with medication
[44] 59 years; 82% male; prospective design packets adherence
(8 weeks)
Carrieri et al. 96 patients with HIV; 69% male; age Self-reporting via CES-D Depression associated negatively with
[45] 18+ years; prospective design questionnaire and interview medication adherence
(18 months)
Catz et al. 72 patients with HIV; mean age 40 years; Self-reporting via interview CES-D Depression associated negatively with
[46] 56% white, 36% African American; medication adherence
cross-sectional design
Ciechanowski 367 patients with NIDDM and IDDM; mean Prescription refills, self-report Depression subscale Depression associated negatively with
et al. [47] age 61 years; 56% female; 86% white; exercise, diet, glucose testing of the Hopkins adherence to diet recommendations (diet
cross-sectional; retrospective and Symptom type and amount) and medication, but not
prospective (both 6 months) designs Checklist-90 exercise or glucose testing
Ciechanowski 407 patients with type 1 and 2 diabetes; Self-report diet (amount, Hopkins Symptoms Depression associated negatively with
et al. [48] mean age 49 years; 53% female; 91% white type), exercise, and blood Checklist-90 Revised adherence to diet (amount and type) and
cross-sectional design sugar monitoring exercise, but not blood sugar monitoring
Cluley and 103 patients with asthma; mean age 42 Electronic diskhalers HADS and Depression associated negatively with
Cochrane years; 66% female; prospective design Structured Clinical medication adherence
[49] (8 weeks) Interview for
DSM-III-R
De-Nour and 100 patients with ESRD; age 20–60+ years; Physician report of Psychiatric interview Depression associated negatively with diet
Czaczkes 61% male; prospective design (3 years) interdialytic weight gain (i.e. recommendations
[50] diet recommendations for
fluid intake)
(cont.)
Table 17.1 (cont.)
Measure of
Study Sample/design information Measure of adherence depression Results
Eaton et al. 127 patients with IDDM; age 14–50+ years; Self-report blood sugar CES-D Depression not associated with combined
[51] 59% female retrospective design (1 year) checking and chart review of measure of behavioural adherence
telephone appointment, office
appointment and
self-monitoring adherence
Everett et al. 42 patients with ESRD; mean age 47 years; Interdialytic weight gain CES-D Depression associated negatively with fluid
[52] 52% male; 69% African American; adherence
cross-sectional design
Frazier et al. 241 renal transplant patients; mean age Self-report of medication BDI Depression associated negatively with
[53] 42 years; 58% male; 91% white; adherence, self-report of medication and follow-up adherence,
cross-sectional design follow-up adherence (clinic although effect not significant in a
attendance, reporting multivariate regression model
problems and laboratory
values)
Gilbar and 53 cancer patients and 53 matched controls; Drop-out of chemotherapy Depression subscale Depression associated positively with
De-Nour [54] mean age 31 years; 62% female; of Brief Symptom treatment drop-out
retrospective design (<1 year) Inventory
Glazer et al. 46 patients with coronary heart disease; Attendance at exercise sessions BDI Depression associated negatively with
[55] mean age 58 years; 74% male (cardiac rehabilitation) adherence to exercise sessions
Gordillo et al. 366 HIV patients from Spain; 20–72 years; Self-report and pill count of BDI Depression associated negatively with
[56] median age 35 years; 76% male; medication (>90% vs. ≤90%) medication adherence, although effect not
cross-sectional design significant in a multivariate regression
model
Graveley and 249 male veterans with unspecified chronic Pill count at home visits Zung Self-Rating Depression associated negatively with
Oseasohn illnesses; age 65–87 years; 54% white, 36% (>80% vs. ≤80%) Depression Scale medication adherence, although effect not
[57] Hispanic, 7% African American, 3% Asian significant in multivariate regression model
American; prospective design (2 weeks)
Guiry et al. 264 post-MI patients; age ≤60 years; 79% Self-report smoking cessation, Unspecified Depression associated negatively with
[58] male; prospective (1 year) weight loss, exercise semi-structured smoking cessation and exercise adherence,
interview but not weight loss
Katz et al. 56 dialysis patients; mean age 61 years; 46% Interdialytic weight gain Self-report of distress Distress not associated with dietary
[59] female; cross-sectional design (IWG), levels of serum sodium symptoms in past adherence for any of the three measures
and potassium (all categorical), month,
e.g. IWG < or ≥3.5 kg unstandardised
Kiley et al. 105 kidney transplant patients; mean age Medical records of medication CES-D Depression associated negatively with
[60] 42 years; 51% African American, 26% white, levels, weight gain prevention, adherence to diet and medication measures,
20% Hispanic, 3% Asian American; appointment keeping (all but not clinic visits
retrospective design (mean duration categorical)
35 months)
Kimmel et al. 295 ESRD patients treated with Behavioural measures of Cognitive Depression Measure of depressive symptoms associated
[61] haemodialysis; 42% with diabetes; mean age dialysis adherence (i.e. time Index of BDI negatively with total dialysis adherence;
55 years; 71% male; 90% African American, dialysed, attendance) adherence predicted mortality risk, but
6% white; prospective design (mean controlling for depression did not attenuate
follow-up 26 months) association
Lebovits et al. 51 breast cancer patients; mean age 53 years; Patient interview and chart Research Diagnostic Depressive symptoms associated negatively
[62] 51% white; prospective design (26 weeks) review of medication Criteria via interview, with medication adherence at week 26, but
adherence (>90% vs. ≤90%) SCL-90 not at weeks 2, 4 or 13; depressive disorders
were not related to adherence; depression
not related to adherence in a multivariate
regression model
McDonough 30 patients with head or neck cancer; mean Self-report measure of diet, BDI, BAI, PANAS Positive affect associated positively with
et al. [63] age 62 years; 70% white; retrospective design medication use, physician overall adherence, but depressive symptoms,
(6 months) appointments, smoking, negative mood and anxiety symptoms not
alcohol consumption associated; only anxiety predicted dietary
adherence; nothing predicted adherence
(cont.)
Table 17.1 (cont.)
Measure of
Study Sample/design information Measure of adherence depression Results
Mohr et al. 85 patients with multiple sclerosis taking Self-report question: ‘Are you 3 self-report New or increased depression associated
[64] interferon beta-1b; 68%; female 85% white; still taking Betaseron?’ questions about negatively with medication adherence;
6% African American, 4% Hispanic; depressive feelings, antidepressant treatment (medication or
retrospective design (6 months) antidepressant counselling) associated with increased
medication and adherence to interferon beta-1b
antidepressant
counselling
Paterson et al. 81 patients with HIV; median age 40 years Medication Event Monitoring BDI and chart review Depression associated negatively with
[65] (range 21–62 years); prospective design System (electronic pill caps) for history of medication adherence; medication
(median 6 months, range 3–15 months) depression adherence associated positively with
virological failure, but depression not
associated with virological failure
Rodriguez 12 ‘ideally adherent’ and 12 ‘non-adherent’ Medical record review of Medical record Non-adherent patients significantly more
et al. [66] renal transplant patients in Puerto Rico; patient’s adherence to diet, review of patient’s likely to have history of depression
mean age 29 years; 63% male; retrospective appointments, medication and psychosocial history
design (time not reported) interdialytic weight gain conducted by
transplant team
social worker
Schneider 50 patients with ESRD; mean age 56 years; Interdialytic weight gain (<3.0 BDI, Spielberger Threshold depression or anxiety not
et al. [67] 74% male; 78% white; retrospective kg or ≥3.0 kg) from medical Trait Anxiety Scale associated with dietary adherence
(8 weeks) chart review of preceding 8
weeks (24 measures)
Singh et al. 46 male HIV patients; median age 40 years; Prescription refills; biological BDI; POMS Depressive symptoms from the BDI not
[68] age 23–68 years; 65% white, 33% African assays to confirm medication depression and associated with medication adherence but
American, 2% Hispanic; prospective design consumption anxiety factor scores POMS depression score associated
(1 year) negatively with adherence; POMS anxiety
not predictive
Spire et al. 445 patients with HIV; mean age 39 years; Self-report medication CES-D Increased depression over time associated
[69] 78% male; ethnicity not reported; adherence negatively with medication adherence in
prospective design (4 months) univariate (multivariate effect not reported)
Taal et al. [70] 86 patients with rheumatoid arthritis; mean Self-report index of problems Dutch version of the Measure of depressive symptoms not related
age 60 years; 71% female; cross-sectional adhering to treatment Arthritis Impact to problems with adhering to treatment
design recommendations derived via Measurement Scales recommendations
interview (i.e. joint protection, – depression subscale
medication, physical therapy)
Tucker et al. 1910 patients with HIV; 58% between ages Medication adherence via Short-form of the Depression but not dysthymia associated
[71] 35–49 year; 78% male; 32% African interview WHO Composite negatively with medication adherence
American, 15% Hispanic; cross-sectional International
design Diagnostic Interview
Turner et al. 5103 patients with HIV and substance Prescription refills for Prescription refill for Depression associated positively with
[72] dependence; 36% female; retrospective anti-retroviral drugs antidepressant medication adherence
design (at least 60 days) medication
Wang et al. 496 patients with hypertension; 60% > 65 Prescription refills Modified version of Depression associated negatively with
[73] years old; 67% male; 95% white; the Brief Symptom medication adherence
cross-sectional design Inventory
Williams 40 ESRD patients, mean age 60 years; 56% Self-report via interview of BDI Threshold depression not associated with
et al. [74] female; 68% African American, 32% white; adherence to exercise exercise adherence
prospective design (12 weeks)* programme
Wilson et al. 184 NIDDM patients; mean age 58 years; Self-report diet, exercise, Composite of BDI Composite measure of depressive symptoms
[75] 98% white; cross-sectional design glucose testing, medication and CES-D associated negatively with glucose testing
taking State-Trait Anxiety adherence, but unrelated to diet, exercise or
Inventory medication taking; anxiety associated
negatively with diet and glucose testing
(cont.)
Table 17.1 (cont.)
Measure of
Study Sample/design information Measure of adherence depression Results
Ziegelstein 204 post-MI patients; 30% with diabetes; Self-report via telephone BDI, SCID-III-R Threshold depression, MDD and dysthymia
et al. [76] mean age 64 years; 57% male; prospective interview with Medical associated negatively with diet, exercise,
design (4 months) Outcomes Study Specific stress reduction and social support
Adherence Scale (i.e. diet, adherence; MDD and dysthymia linked with
medication exercise, stress, poor medication adherence; among diabetic
blood glucose checking, patients, MDD and dysthymia linked with
increasing social support) poorer diet adherence
∗
Only 22/40 patients completed the BDI.
BAI, Beck Anxiety Inventory; BDI, Beck Depression Inventory; BP = blood pressure; CAD, cardiovascular disease; CES-D, Center for Epidemiological Studies –
Depression Scale; COPD, chronic obstructive pulmonary disease; DSM, Diagnostic and Statistical Manual; ESRD, end-stage renal disease; HADS, Hospital Anixiety and
Depression Scale; HAM-D, Hamilton Rating Scale for Depression; HDL, high-density lipoproteins; HIV, human immunodeficiency virus; IDDM, insulin-dependent
diabetes mellitus; LDL, low-density lipoproteins; MDD, major depressive disorder; MI, myocardial infarction; MMPI, Minnesota Multiphasic Personality Inventory;
NIDDM, non-insulin-dependent diabetes mellitus; PANAS, Positive and Negative Affect Scale; POMS, Profile of Mood States; SCID, Structured Clinical Interview
for DSM; SCL-90, Symptom Checklist-90; WHO, World Health Organization
383 Depression and adherence to medical advice
several of the stronger studies on this topic. These studies highlight the complexity
of this topic, especially as it relates to the ‘mediational hypothesis’.
Bosley et al. evaluated treatment adherence and depression among 102 patients
with asthma [40] and 93 with chronic obstructive pulmonary disease (COPD)
[41]. Strengths of these studies include the fact that electronic event monitoring
was used to assess medication adherence and the Hospital Anxiety and Depression
Scale (HADS) was used to assess depressive symptoms independent of anxiety.
Prospective designs were used in both studies, with 12-week and 4-week periods
in the asthma and COPD studies, respectively. The authors reported a negative
association between depression and medication adherence in the asthma study but
no association in the COPD study. They were not able to determine the cause for
the mixed findings.
Another relatively strong methodological study was conducted by Kimmel et al.
[61], who prospectively examined 295 end-stage renal disease (ESRD) patients
over a mean follow-up period of 26 months. Adherence to dialysis was measured
behaviourally, and depressive symptoms were measured via the BDI and the Cog-
nitive Depression Index (CDI), a non-somatic version of the BDI. The BDI and
the CDI were associated with poor adherence to dialysis treatment. Moreover, poor
dialysis adherence predicted mortality risk. However, depression did not predict
mortality. Thus, these data suggest that depression is associated negatively with
adherence, but the data do not support the mediational hypothesis that depression
affects mortality via treatment adherence.
Ciechanowski et al. [47] explored the impact of depressive symptoms on adher-
ence to medical regimens (i.e. medication, exercise, diet, glucose self-monitoring)
among 367 patients with diabetes who were seen in primary-care clinics of a large
health-maintenance organisation. Medication adherence was measured retrospect-
ively (previous six months) and prospectively (subsequent six months) by prescrip-
tion refills from a centralised pharmacy database. Other adherence behaviours were
measured cross-sectionally via self-report. Depressive symptoms were measured
with the depression subscale of the Hopkins Symptom Checklist-90. Higher levels
of depressive symptomatology were associated with poorer adherence to diet and
medication regimens, but depression was not related to adherence to exercise or
glucose self-monitoring. Importantly, depressive symptom severity was not asso-
ciated with glycosylated haemoglobin levels, thereby failing to support a necessary
condition of the mediational hypothesis.
Adherence to medical recommendations among 204 recent post-myocardial
infarction (MI) patients was examined by Ziegelstein et al. [76]. Telephone inter-
views were used to assess adherence to several behaviours, including diet, exercise,
medication, stress management, blood glucose checking for diabetics, and social
support. Patients with above-threshold depressive symptoms (BDI > 10), major
384 Depression and physical illness
Table 17.2 Association between depression and treatment adherence by disease category
Cancer 0 2 1 1
CAD 4 0 1 2
Diabetes 0 0 1 3
HIV 4 1 0 4
Pulmonary 1 0 1 1
Renal 4 0 3 2
Other 2 3 2 1
Intervention studies
Four studies have moved beyond simply examining the correlation between depres-
sion and adherence to intervention approaches testing whether treatment of depres-
sion will result in improved levels of adherence [77–80]. In addition, these studies
examined whether the changes in adherence were related to changes in health
outcomes. These intervention studies are presented in Table 17.3. Goodnick et al.
[77] conducted a 10-week open-label antidepressant intervention with 28 patients
diagnosed with non-insulin-dependent diabetes mellitus (NIDDM) and MDD.
All individuals were administered sertraline after a two-week single-blind placebo
washout period. After the ten-week period, Hamilton rating scale for depression
(HAM-D) and BDI scores were significantly reduced. Individuals with low baseline
dietary adherence manifested significantly improved adherence during the period.
Finally, 76% of patients with high baseline glycosylated haemoglobin levels showed
significant reductions. These findings are suggestive of a link between depression
and adherence, but the lack of a control group limits the conclusions that can be
drawn.
Lustman and colleagues conducted two randomised controlled trials with dia-
betic patients. In the first trial [78], 68 patients with NIDDM or insulin-dependent
diabetes mellitus (IDDM), 28 of whom had current MDD, took either nortriptyline
or placebo for 8 weeks. Adherence to the medication and glucose self-monitoring
regimen was measured with two electronic devices. Patients treated with nor-
triptyline experienced a significantly greater reduction in BDI scores than patients
treated with placebo. Path analysis indicated that depression improvement had an
independent beneficial effect on glycosylated haemoglobin levels; this association
Table 17.3 Intervention Studies on Depression and Adherence to Medical Regimens
Measure of
Study Sample/design information Measure of adherence depression results
Goodnick 28 patients with NIDDM and MDD; mean age Unspecified dietary HAM-D, BDI Individuals with low baseline dietary adherence and
et al. [77] 54 years; 57% male; uncontrolled treatment compliance treated with sertraline showed significantly improved
study (10 weeks) HAM-D and BDI scores and dietary adherence
Lustman 68 patients with NIDDM or IDDM; 30% Electronic blister pacs BDI Nortriptyline led to reductions of depressive
et al. [78] diagnosed with MDD; mean age 49 years; 51% (meds), electronic symptoms in depressed patients, but this
female; 88% white; prospective intervention glucometers (i.e. measures improvement not associated with improved
study (2 months) blood glucose monitoring) adherence in medication taking or glucose
monitoring
Lustman 51 patients with NIDDM and diagnosis of Electronic glucometer BDI CBT led to greater reductions in depressive
et al. [79] MDD; mean age 55 years∗ ; 60% female; 81% symptoms than in control group; blood glucose
white; prospective intervention study (6-month monitoring adherence declined in the CBT group
follow-up) compared with control group
Williams 417 patients with DM and MDD or dysthymia; Self-report measure of SCL-90 Intervention (education, problem-solving treatment,
et al. [80] mean age 71 years; 65% female; 77% white, 12% diet, exercise, glucose or support for antidepressant management) led to
African American, 8% Hispanic; prospective testing, foot care, reduction of depressive symptoms and to increase in
intervention study (12 months follow-up) medication adherence weekly exercise days at 12-months’ follow-up;
intervention not associated with improved
adherence in diet, glucose testing, medication taking
or foot care; intervention did not lead to
improvement in HA1c values
∗
All demographic data presented for study completers (42/51) only.
BDI, Beck Depression Inventory; CBT, cognitive–behavioural therapy; DM, diabetes mellitus, unspecified type; HA1c , glycosylated haemoglobin; HAM-D,
Hamilton Rating Scale for Depression; IDDM, insulin-dependent diabetes mellitus; MDD, major depressive disorder; NIDDM, non-insulin-dependent diabetes
mellitus; SCL-90, Symptom Checklist-90.
387 Depression and adherence to medical advice
Conclusions
In conclusion, this review suggests that there is a negative relationship between
depression and adherence, but this relationship appears less consistent than is
typically assumed. Frequently, depression was related to some aspects of self-care
behaviours but not to others. Moreover, the interventions studies suggest that treat-
ing depression does not consistently lead to improved adherence.
Research on depression and adherence is increasing rapidly. It is important that
future research on this topic be designed with greater methodological sophisti-
cation. For example, objective measures of adherence should be used whenever
possible. Self-reporting of adherence is particularly problematic in studies of the
388 Depression and physical illness
or weight control, which might impact on the levels of depression but should also
have a more direct effect on improving health outcomes.
REFERENCES
34. L. Smithline, Is depression a risk factor for cardiovascular disease? A critical review. Exami-
nation paper. Pittsburgh, PA: University of Pittsburgh, 1997.
35. D. G. Blazer, R. C. Kessler, K. A. McGonagle, M. S. Swartz, The prevalence and distribution
of major depression in a national comorbidity survey. Am. J. Psychiatry 151 (1994), 979–86.
36. P. A. Ades, M. L. Waldmann, W. J. McCann, S. O. Weaver, Predictors of cardiac rehabilitation
participation in older coronary patients. Arch. Intern. Med. 152 (1992), 1033–5.
37. S. K. Avants, A. Margolin, L. A. Warburton, K. A. Hawkins, J. Shi, Predictors of nonadher-
ence to HIV-related medication regimens during methadone stabilization. Am. J. Addict. 10
(2001), 69–78.
38. A. Ayres, P. W. Hoon, J. B. Franzoni, et al., Influence of mood and adjustment to cancer on
compliance with chemotherapy among breast cancer patients. J. Psychosom. Res. 38 (1994),
393–402.
39. J. A. Blumenthal, S. Williams, A. G. Wallace, R. B. Williams, T. L. Needles, Physiological
and psychological variables predict compliance to prescribed exercise therapy in patients
recovering from myocardial infarction. Psychosom. Med. 44 (1982), 519–27.
40. C. M. Bosley, J. A. Fosbury, G. M. Cochrane, The psychological factors associated with poor
compliance with treatment in asthma. Eur. Respir. J. 8 (1995), 899–904.
41. C. M. Bosley, Z. M. Corden, P. J. Rees, G. M. Cochrane, Psychological factors associated with
use of home nebulized therapy for COPD. Eur. Respir. J. 9 (1996), 2346–50.
42. R. J. Botelho, R. Dudrak, Home assessment of adherence to long-term medication in the
elderly. J. Fam. Pract. 35 (1992), 61–5.
43. R. M. Carney, K. E. Freedland, S. A. Eisen, M. W. Rich, A. S. Jaffe, Major depression and
medication adherence in elderly patients with coronary artery disease. Health Psychol. 14
(1995), 88–90.
44. R. M. Carney, K. E. Freedland, S. A. Eisen, et al., Adherence to prophylactic medication
regimen in patients with symptomatic versus asymptomatic ischemic heart disease. Behav.
Med. 24 (1998), 35–9.
45. M. P. Carrieri, M. A. Chesney, B. Spire, et al., Failure to maintain adherence to HAART
in a cohort of French HIV-positive injecting drug users. Int. J. Behav. Med. 10 (2003),
1–14.
46. S. L. Catz, J. A. Kelly, L. M. Bogart, E. G. Benotsch, T. L. McAuliffe, Patterns, correlates, and
barriers to medication adherence among persons prescribed new treatments for HIV disease.
Health Psychol. 19 (2000), 124–33.
47. P. S. Ciechanowski, W. J. Katon, J. O. Russo, Depression and diabetes: Impact of depres-
sive symptoms on adherence, function, and costs. Arch. Intern. Med. 160 (2000), 3278–
85.
48. P. S. Ciechanowski, W. J. Katon, J. E. Russo, I. B. Hirsch, The relationship of depressive symp-
toms to symptom reporting, self-care and glucose control in diabetes. Gen. Hosp. Psychiatry
25 (2003), 246–52.
49. S. Cluley, G. M. Cochrane, Psychological disorder in asthma is associated with poor control
and poor adherence to inhaled steriods. Respir. Med. 95 (2001), 37–9.
50. A. K. De-Nour, J. W. Czaczkes, The influence of patient’s personality on adjustment to chronic
dialysis. J. Nerv. Ment. Dis. 162 (1976), 323–33.
392 Depression and physical illness
69. B. Spire, S. Duran, M. Souville, et al., Adherence to highly active antiretroviral therapies
(HAART) in HIV-infected patients: from a predictive to a dynamic approach. Soc. Sci. Med.
54 (2002), 1481–96.
70. E. Taal, J. J. Rasker, E. R. Seydel, O. Wiegman, Health status, adherence with health recom-
mendations, self-efficacy and social support in patients with rheumatoid arthritis. Patient
Educ. Couns. 20 (1993), 63–76.
71. J. S. Tucker, A. Burnam, C. D. Sherbourne, F. Y. Kung, A. L. Gifford, Substance use and mental
health correlates of nonadherence to antiretroviral medications in a sample of patients with
human immunodeficiency virus infection. Am. J. Med. 114 (2003), 573–80.
72. B. J. Turner, C. Laine, L. Cosler, W. W. Hauck, Relationship of gender, depression, and health
care delivery with antiretroviral adherence in HIV-infected drug users. J. Gen. Intern. Med.
18 (2003), 248–57.
73. P. S. Wang, R. L. Bohn, E. Knight, et al., Noncompliance with antihypertensive medications:
the impact of depressive symptoms and psychosocial factors. J. Gen. Intern. Med. 17 (2002),
504–11.
74. A. Williams, R. Stephens, T. McKnight, S. Dodd, Factors affecting adherence of end-stage
renal disease patients to an exercise programme. Br. J. Sports Med. 25 (1991), 90–3.
75. W. Wilson, V. Ary, A. Biglan, et al., Psychosocial predictors of self-care behaviors (compliance)
and glycemic control in non-insulin-dependent diabetes mellitus. Diabetes Care 9 (1986),
614–22.
76. R. C. Ziegelstein, J. A. Fauerbach, S. S. Stevens, et al., Patients with depression are less likely to
follow recommendations to reduce cardiac risk during recovery from a myocardial infarction.
Arch. Intern. Med. 160 (2000), 1818–23.
77. P. J. Goodnick, A. Kumar, J. H. Henry, V. M. V. Buki, R. B. Goldberg, Sertraline in coexisting
major depression and diabetes mellitus. Psychopharmacol. Bull. 33 (1997), 261–4.
78. P. J. Lustman, L. S. Griffith, R. E. Clouse, et al., Effects of nortriptyline on depression and
glycemic control in diabetes: results of a double-blind, placebo-controlled trial. Psychosom.
Med. 59 (1997), 241–50.
79. P. J. Lustman, L. S. Griffith, K. E. Freedland, S. S. Kissel, R. E. Clouse, Cognitive behavior
therapy for depression in type 2 diabetes mellitus: a randomized, controlled trial. Ann. Intern.
Med. 129 (1998), 613–21.
80. J. W. Williams, W. Katon, E. Lin, et al., The effectiveness of depression care management on
diabetes-related outcomes in older patients. Ann. Intern. Med. 140 (2004), 1015–24.
81. R. M. Baron, D. A. Kenny, The moderator-mediator variable distinction in social psycho-
logical research: conceptual, strategic, and statistical considerations. J. Pers. Soc. Psychol. 51
(1986), 1173–82.
82. R. R. Wing, S. Phelan, D. F. Tate, The role of adherence in mediating the relationship between
depression and health outcomes. J. Psychosom. Res. 53 (2002), 877–81.
Part 4
Conclusions
18
18
Integrating clinical with biobehavioural
studies of depression and physical illness
Andrew Steptoe
adults in the USA, showing that the association between depression and increased
mortality was no longer significant after baseline self-reported health and func-
tional impairment were taken into account. Existing physical illness and disability
can increase depressive symptoms, as is evident from many of the chapters in this
book, and so the association between depression and future mortality may in some
cases be due to existing physical pathology. This issue is difficult to resolve, since it
is possible to overcontrol statistically in studies of depression and physical illness
[20]. If biological and lifestyle factors are part of the pathway linking depression
with health outcomes, then controlling for them eliminates part of what is being
studied.
Behavioural factors:
Physical • Smoking Depressed
illness • Physical activity mood
• Alcohol, diet
Emotional
distress
Figure 18.1 Outline of different models relating depression with physical illness.
and obesity (Chapter 11) are more equivocal, and the conclusion of the review of
cancer is negative (Chapter 10).
There is also good evidence that the biological and behavioural factors discussed
in Part 3 contribute to adverse health outcomes and are plausible mediators of
depression effects. What has been demonstrated less clearly is that these factors
mediate associations between depression and illness. For instance, Raynor and
colleagues in Chapter 17 have provided a systematic analysis of the hypothesis
401 Integrating clinical with biobehavioural studies of depression and physical illness
that depression leads to poor adherence to treatment and medical advice, which
in turn promotes disease progression and negative outcomes. The evidence for
this sequence is far less impressive than some behavioural scientists and clinicians
assume. Many studies have failed to show that depression predicts poor adherence,
and there is even less support from either observational or intervention studies that
adherence mediates associations between depression and health outcomes. Simi-
larly, there is little direct evidence that smoking or physical inactivity are mediators
(Chapters 15 and 16).
The second possibility outlined in Figure 18.1 is the consequence model, in which
depressed mood arises as a result of physical illness. It is interesting that many of
the same biological and behavioural factors could act as mediators of this process,
just as they might operate in the reverse direction in the antecedent model. The
fact that there are common mediators reflects either the reality of the situation or
the possibility that research has not advanced sufficiently beyond the correlational
to allow definitive conclusions about causal processes to be drawn.
Depression may arise in an individual with serious physical illness as an emo-
tional response to diagnosis, treatment and the destruction of future life prospects.
There are also more direct biological mediators. One biological pathway related
to the consequence model is detailed by Dantzer and colleagues in Chapter 12,
where they describe the impact of inflammation and cytokine release on depres-
sion, sleep disturbance and fatigue. The clinical manifestation of these processes
is described in relation to chronic pain and fatigue in Chapters 7 and 9. Carney
and Freedland in Chapter 5 point out that the depression experienced by patients
following myocardial infarction may be qualitatively different from other forms of
depression and may be stimulated partly by the acute inflammatory state that is
present during acute coronary syndromes. Certainly, the risk factors for depression
in cardiac patients differ from those observed generally [21]. One implication may
be that depression in physically ill people needs distinctive treatment, since existing
interventions may not target the relevant mechanisms. Different factors mediate
the impact of illness on depressed mood in specific conditions. In stroke patients,
for example, depression appears to arise from direct damage to certain regions of
the brain, and particularly from lesions in the left frontal area and the basal ganglia
[22].
The third model in Figure 18.3 is the shared determinants or common patho-
genesis model. This argues that there are factors underlying increased risk for both
depressed mood and illness, such that the relationship between the two has no causal
significance. Stansfeld and Rasul in Chapter 2 detail how low socioeconomic back-
ground, poor educational attainment and childhood adversity contribute to risk of
depression in later life; the same factors predict adverse physical health outcomes
and biological and behavioural risk profiles. In Chapter 13, Goodyer outlines how
402 Depression and physical illness
Methodological problems
Numerous methodological issues have been discussed in this book that are signifi-
cant in particular fields of clinical and basic research. But two overarching themes
have emerged that are relevant across domains and warrant reiterating here.
The first is the issue of whether depression is regarded as a categorical diag-
nostic entity or as a continuous dimensional variable. It is generally agreed that
there are gradations of depression. From the psychiatric perspective, there may
exist a critical threshold of intensity and duration above which treatment is war-
ranted, either because it is known to be efficacious or cost-effective with depression
above this level or because there is serious concern about self-harm. The issue for
studies of physical illness is whether relationships exist only with depression of
a severity that is recognised psychiatrically, or whether lesser degrees of depres-
sion are important. A related question is whether dose–response associations are
present.
There is good evidence from many of the contributions to this book that a
dimensional approach to depression is valuable, in that associations exist between
physical health problems and moderate or subclinical levels of depressed mood.
However, evidence for dose–response effects is more elusive, and such associations
have not been established definitively for many conditions. In the absence of this
information, it is possible that depressed mood has to exceed a threshold in order
to be relevant to physical illness, and that a categorical approach is still valid. Of
course, the threshold may be different from the one that is relevant for psychiatric
conditions and may not even be the same for different health outcomes.
The second general methodological issue raised in several chapters in this book
is the confounding of depression and illness at the symptomatic level. This is partic-
ularly relevant to chronic fatigue, as pointed out by White in Chapter 9, but it also
applies elsewhere. Many health problems are characterised by low energy levels and
tiredness, and these phenomena are parts of depressive syndromes. More specific
403 Integrating clinical with biobehavioural studies of depression and physical illness
features of depression may also be present. For example, Musselman and colleagues
in Chapter 8 detail the way in which diabetes is frequently accompanied by loss of
libido and psychomotor retardation, while immune dysfunction and sleep distur-
bance are linked closely in depression (Chapter 14). Additionally, depressed people
often exhibit reporting biases in relation to somatic symptoms, making it difficult
to disentangle depression from illness and disability.
These issues are particularly problematic when definitions of health states depend
on self-rated rather than objective indicators, and so provide an argument for the
use of objective clinical, behavioural and biological markers wherever possible. It is
arguable whether studies relating depressive symptoms with outcomes such as self-
rated health can ever avoid this measurement confound. Methodological difficulties
may also be exacerbated when depression is assessed by self-report questionnaire
rather than by interview or observer ratings. There is an important message for
researchers in this field to select study designs and assessments that eliminate these
confounds as far as possible.
wellbeing and quality of life that may result from depression treatment are desirable
aims in themselves.
Treatment issues
We are at an important stage in understanding how to treat depressive symptoms
in medical patients. The effects of conventional cognitive–behavioural and phar-
macological treatments have not been as great as expected in many conditions.
Van Puymbroeck and colleagues in Chapter 7 point to studies in which cognitive–
behavioural therapy (CBT) has had favourable effects on chronic pain without
influencing depression, while the impact of antidepressant medication on depres-
sive states in diabetic and post-myocardial infarction patients has been inconsis-
tent (Chapters 5 and 8). Carney and Freedland in Chapter 5 speculate that the
often elderly and medically ill people who survive an acute coronary event may
find it difficult to engage with standard CBT, with its focus on automatic negative
thoughts and core beliefs. Similar problems are apparent among cancer patients,
many of whom are not prepared to engage with mental health issues (Chapter 10).
At the same time, there is reason to suppose that mental healthcare methods can
be adapted for these different settings, and some newer antidepressant mediations
also look promising [22,36]. The knowledge that is accumulating about biological
processes (Chapters 12–14) will stimulate further developments in treatments that
may have effects on both psychological and biological outcomes. Other modali-
ties, such as physical exercise training, may be useful in helping at least moderately
depressed people, with the added benefit that they might also improve clinical risk
profiles and help reduce disability (Chapters 6 and 16). The challenge will be how
406 Depression and physical illness
to integrate such methods into the clinical care of patients who do not primarily
have psychological problems and who may consider attention to their mental state
to be a distraction from the treatment of their physical condition.
REFERENCES
1. J. A. Cook, D. Grey, J. Burke, et al., Depressive symptoms and AIDS-related mortality among
a multisite cohort of HIV-positive women. Am. J. Public Health 94 (2004), 1133–40.
2. A. A. Lopes, J. Bragg, E. Young, et al., Depression as a predictor of mortality and hospitalization
among hemodialysis patients in the United States and Europe. Kidney Int. 62 (2002), 199–207.
3. T. A. Hughes, H. F. Ross, R. H. Mindham, E. G. Spokes, Mortality in Parkinson’s disease and
its association with dementia and depression. Acta Neurol. Scand. 110 (2004), 118–23.
4. K. B. Stage, T. Middelboe, C. Pisinger, Depression and chronic obstructive pulmonary disease
(COPD): impact on survival. Acta Psychiatr. Scand. 111 (2005), 320–3.
5. B. A. Golomb, L. Tenkanen, T. Alikoski, et al., Insulin sensitivity markers: predictors of
accidents and suicides in Helsinki Heart Study screenees. J. Clin. Epidemiol. 55 (2002), 767–
73.
6. D. A. Lawlor, G. D. Smith, S. Ebrahim, Association of insulin resistance with depression:
cross sectional findings from the British Women’s Heart and Health Study. Br. Med. J. 327
(2003), 1383–4.
7. S. L. Larson, P. L. Owens, D. Ford, W. Eaton, Depressive disorder, dysthymia, and risk of stroke:
thirteen-year follow-up from the Baltimore epidemiologic catchment area study. Stroke 32
(2001), 1979–83.
8. M. May, P. McCarron, S. Stansfeld, et al., Does psychological distress predict the risk of
ischemic stroke and transient ischemic attack? The Caerphilly Study. Stroke 33 (2002), 7–12.
9. B. B. Gump, K. A. Matthews, L. E. Eberly, Y. F. Chang, Depressive symptoms and mortality
in men: results from the Multiple Risk Factor Intervention Trial. Stroke 36 (2005), 98–102.
10. S. Wassertheil-Smoller, S. Shumaker, J. Ockene, et al., Depression and cardiovascular sequelae
in postmenopausal women: the Women’s Health Initiative (WHI). Arch. Intern. Med. 164
(2004), 289–98.
11. W. Jiang, J. Alexander, E. Christopher, et al., Relationship of depression to increased risk of
mortality and rehospitalization in patients with congestive heart failure. Arch. Intern. Med.
161 (2001), 1849–56.
12. T. A. Murberg, G. Furze, Depressive symptoms and mortality in patients with congestive
heart failure: a six-year follow-up study. Med. Sci. Monit. 10 (2004), CR643–8.
13. J. Junger, D. Schellberg, T. Muller-Tasch, et al., Depression increasingly predicts mortality in
the course of congestive heart failure. Eur. J. Heart Fail. 7 (2005), 261–7.
14. R. Schulz, R. A. Drayer, B. L. Rollman, Depression as a risk factor for non-suicide mortality
in the elderly. Biol. Psychiatry 52 (2002), 205–25.
15. P. Cuijpers, F. Smit, Excess mortality in depression: a meta-analysis of community studies. J.
Affect. Disord. 72 (2002), 227–36.
407 Integrating clinical with biobehavioural studies of depression and physical illness
16. K. T. Ensinck, A. G. Schuurman, M. van den Akker, et al., Is there an increased risk of dying
after depression? Am. J. Epidemiol. 156 (2002), 1043–8.
17. K. Frojdh, A. Hakansson, I. Karlsson, A. Molarius, Deceased, disabled or depressed: a
population-based 6-year follow-up study of elderly people with depression. Soc. Psychia-
try Psychiatr. Epidemiol. 38 (2003), 557–62.
18. R. S. Wilson, J. L. Bienias, C. F. Mendes de Leon, D. A. Evans, D. A. Bennett, Negative affect
and mortality in older persons. Am. J. Epidemiol. 158 (2003), 827–35.
19. S. A. Everson-Rose, J. S. House, R. P. Mero, Depressive symptoms and mortality risk in
a national sample: confounding effects of health status. Psychosom. Med. 66 (2004), 823–
30.
20. A. Steptoe, D. L. Whitehead, Depression, stress and coronary heart disease: the need for more
complex models. Heart 91 (2005), 419–20.
21. C. M. Dickens, C. Percival, L. McGowan, et al., The risk factors for depression in first myocar-
dial infarction patients. Psychol. Med. 34 (2004), 1083–92.
22. R. G. Robinson, Poststroke depression: prevalence, diagnosis, treatment, and disease pro-
gression. Biol. Psychiatry 54 (2003), 376–87.
23. M. Dong, S. R. Dube, V. J. Felitti, W. H. Giles, R. F. Anda, Adverse childhood experiences
and self-reported liver disease: new insights into the causal pathway. Arch. Intern. Med. 163
(2003), 1949–56.
24. M. Dong, W. H. Giles, V. J. Felitti, et al., Insights into causal pathways for ischemic heart
disease: adverse childhood experiences study. Circulation 110 (2004), 1761–6.
25. H. S. Friedman, S. Booth-Kewley, The ‘disease-prone personality’: A meta-analytic view of
the construct. Am. Psychol. 42 (1987), 539–55.
26. J. Suls, J. Bunde, Anger, anxiety, and depression as risk factors for cardiovascular disease: the
problems and implications of overlapping affective dispositions. Psychol. Bull. 131 (2005),
260–300.
27. M. Friedman, C. E. Thoresen, J. J. Gill, et al., Feasibility of altering type A behavior pattern after
myocardial infarction: Recurrent Coronary Prevention Project Study – methods, baseline
results and preliminary findings. Circulation 66 (1982), 83–92.
28. D. Spiegel, J. R. Bloom, H. C. Kraemer, E. Gottheil, Effect of psychosocial treatment on
survival of patients with metastatic breast cancer. Lancet 2 (1989), 888–91.
29. N. Schneiderman, P. G. Saab, D. J. Catellier, et al., Psychosocial treatment within sex by
ethnicity subgroups in the Enhancing Recovery in Coronary Heart Disease clinical trial.
Psychosom. Med. 66 (2004), 475–83.
30. C. Herrmann, International experiences with the Hospital Anxiety and Depression Scale: a
review of validation data and clinical results. J. Psychosom. Res. 42 (1997), 17–41.
31. K. Kroenke, R. L. Spitzer, J. B. Williams, The PHQ-9: validity of a brief depression severity
measure. J. Gen. Intern. Med. 16 (2001), 606–13.
32. K. Kroenke, R. L. Spitzer, J. B. Williams, The Patient Health Questionnaire-2: validity of a
two-item depression screener. Med. Care 41 (2003), 1284–92.
33. K. E. Freedland, J. A. Skala, R. M. Carney, et al., The Depression Interview and Structured
Hamilton (DISH): rationale, development, characteristics, and clinical validity. Psychosom.
Med. 64 (2002), 897–905.
408 Depression and physical illness
34. G. Lewis, A. J. Pelosi, R. Araya, G. Dunn, Measuring psychiatric disorder in the community:
a standardized assessment for use by lay interviewers. Psychol. Med. 22 (1992), 465–86.
35. B. G. Druss, D. W. Bradford, R. A. Rosenheck, M. J. Radford, H. M. Krumholz, Mental
disorders and use of cardiovascular procedures after myocardial infarction. J. Am. Med.
Assoc. 283 (2000), 506–11.
36. C. L. Raison, A. H. Miller, Depression in cancer: new developments regarding diagnosis and
treatment. Biol. Psychiatry 54 (2003), 283–94.
Index
activities of daily living (ADL) 126–7, 129 dysregulation and coronary heart disease
activity see physical activity (CHD) development 73–5
adherence to treatment regimens dysregulation with depression 73–5, 135, 301–2,
association with depression 369–70 307
chronic disease regimens 370–72 link with coronary heart disease (CHD) and
and coronary heart disease (CHD) 98, 383 depression 95–6
defining and assessing adherence 371–2 and physical disability 135
defining and assessing depression 372–4 sympathetic activation in depression 301–2
and diabetes 383, 385–6 antidepressant treatments
intervention studies 385–6, 387 actions on the immune system 151–2
as mediator between depression and health analgesic properties 151–2
outcome 369–70, 388 contraindications for coronary heart disease
methodological issues 387–8 (CHD) 111
methods for identifying, including and for depression in coronary heart disease (CHD)
evaluating studies 374–5, 376–82 111–12, 116
non-adherence among depressed people 134 for diabetes and depression 176–8
overview of existing research 375–82, 384, 385 drug–drug interactions in coronary heart
predictors of 369, 372 disease (CHD) 111–12
adrenal steroids, measurement of circulating levels effects on chronic fatigue syndrome 202–3
283–5 effects on chronic pain 151–2
ageing see older people effects on disability 137
alcohol abuse effects on fibromyalgia 202–3
among depressed people 133–4 effects on immune function 306–7
effects on immune dysfunction in depression efficacy in coronary heart disease (CHD)
309 111–12
link with depression 98–9, 134 monitoring for impaired glucose function 177–8
risk factor for coronary heart disease (CHD) testing for diabetes 177–8
98–9 use in cancer patients 220–22
Alzheimer’s disease 275–6 anxiety
AMI (acute myocardial infarction) life events as risk factor 23–4
acute triggering 68–9 as predictor of coronary heart disease (CHD)
anger and cardiac events 68, 92 62–5, 66
depression and psychosocial adjustment 110–11 work-related risk factors 37
depression as a risk factor 54–61 atherogenesis, effects of physical activity 362–3
effects of depression on outcome 88–9, 90–92, atherosclerosis
94, 110–11 aetiology 70–71
elevated CRP as risk factor 70–71, 97–8 carotid IMT (intima-medial layer thickness)
prevalence of depression and anxiety 87–90, 66–7
109–10 carotid plaque changes at subclinical stage 66–7
psychosocial risk factors for outcome 92–4 and endothelial dysfunction 96–7
psychological and social re-adjustment 109 impact of immune dysfunction in depression
role of haemostatic responses 71–2 311
angina pectoris 54 and inflammation 97–8
ANS (autonomic nervous system) function non-invasive measurement 66–7
410 Index
type D personality and outcome 92–4 monoamine oxidase inhibitors (MAOIs) 111
unexplained by traditional risk factors 87 pharmacotherapeutic treatment studies 116
vital exhaustion as risk factor 60–61 pharmacotherapy 111–12
CHD (coronary heart disease) aetiology psychotherapeutic intervention studies 114–15,
autonomic dysregulation 73–5 116–19
baroreceptor reflex sensitivity 73–5 psychotherapy techniques 112–15
behavioural and lifestyle factors 69–70 quality-of-life benefits 120
cortisol metabolism abnormalities 56–9, 72–3, to reduce medical morbidity and mortality
96 116–19
endothelial dysfunction 70–71 serotonin reuptake inhibitors (SSRIs) 111–12
haemostatic processes and platelet activity 71–2 sertraline (SADHART) study 112, 116
heart-rate variability 73–5 side effects of drugs 111
HPA (hypothalamic-pituitary-adrenocortical) tricyclic antidepressants (TCAs) 111
dysregulation 56–9, 72–3 CHF (congestive heart failure) and depression 398
inflammatory markers 70–71 childhood abuse/adversity
limitations to studies of mechanisms 75 behavioural consequences 291
neuroendocrine function 56–9, 72–3 buffer effect of social support in adulthood
stages of atherosclerosis 70–72 29–30
sympathoadrenal activation 73–5 cortisol secretion 291
sympathovagal balance 73–5 and depression 27, 286, 291
vascular inflammation 70–71 see also infant exposure to adversity
CHD (coronary heart disease) link with depression chronic fatigue syndrome (CFS) 196–7
adherence to treatment 383–4 CBT (cognitive behavioural therapy) 202
autonomic nervous system functioning 95–6 comparison with depression 197–200, 201–3
baroreflex activity (baroreceptor-mediated definition 196, 198
heart-rate control) 96 effects of antidepressants 202–3
behavioural and lifestyle risk factors 98–9 exercise therapy 202–3
biobehavioural mechanisms 94–9 features of fatigue 197
biological processes that may mediate 53, 56–9, following infection 199–200
69–75 high rates of depressive illness 198–9
confounding factors in studies 54–5 HPA (hypothalamic–pituitary–adrenocortical)
diabetes 94, 95 axis dysregulation 201
endothelial dysfunction 70–71, 96–7 independence from mood disorders 198–9, 402
heart-rate variability 96, 74–5 management 202
high alcohol consumption 98–9 pathophysiological differences from depression
HPA (hypothalamic–pituitary–adrenocortical) 201–2
axis dysregulation 72–4, 96 relationship to vital exhaustion 200–201
inflammation 70–71, 97–8, 311, 401 risk markers 196–7
longitudinal studies 54–9, 61 treatment similarities and differences from
metabolic syndrome 95 depression 202–3
nature of relationship with depression 53–4 chronic heart failure, effects of physical activity
obesity 95 interventions 360, 361–2
physical inactivity 98–9 chronic obstructive pulmonary disease
platelet activity 71–2, 95 and adherence 383
risk factors 87, 94, 95 effects of physical activity interventions 358–9
serotonin activity 95 chronic pain and depression
smoking 98–9 actions of psychological stressors 153–4
timing of associations with depression 56–9, analgesic properties of antidepressants 151–2
67–9 antecedent hypothesis 146
underlying association with depression 56–9, and chronic stress 146–7
67–9, 75–6 catastrophising cognitions in pain patients 147
CHD (coronary heart disease)-linked depression, CBT (cognitive behavioural therapy) 147, 157–8
treatments 111–20 common pathogenesis model 146–7
antidepressant medications 111–12, 116 consequence hypothesis 146
cognitive behaviour therapy (CBT) 113–14 cytokine activity and sensitization 150–51,
depression treatment and quality of life 120 152–4
drug–drug interactions 111–12 depression and cytokine activity 148–9
ENRICHD clinical trial 118–19 depression and the immune system 148–9
exercise 115–16 dynamic affect model 156–7
interpersonal psychotherapy 114 effects of inflammatory processes 150–52
lack of treatment for depression with CHD 120 emotional regulation interventions 155–8
412 Index