Clinicians and researchers alike have long noted the prevalence. Personality and Depression. In Review

Transcription

1 In Review Personality and Depression R Michael Bagby, PhD, C Psych 1 ; Lena C Quilty, PhD 2 ; Andrew C Ryder, PhD 3 Objective: To examine the implications of the association between personality and depression for the understanding, assessment, and treatment of major depression. Method: A broad range of peer-reviewed manuscripts relevant to personality and depression was reviewed. Particular emphasis was placed on etiology, stability, diagnosis, and treatment implications. Results: Personality features in depressed samples reliably differ from those of healthy samples. The associations between personality and depression are consistent with a variety of causal models; these models can best be compared through longitudinal research. Research demonstrates that attention to personality features can be useful in diagnosis and treatment. Indeed, personality information has been on the forefront of recent efforts to advance the current diagnostic classification system. Moreover, personality dimensions have shown recent promise in the prediction of differential treatment outcome. For example, neuroticism is associated with preferential response to pharmacotherapy rather than psychotherapy. Conclusions: Consideration of personality features is crucial to the understanding and management of major depression. Can J Psychiatry 2008;53(1):14 25 Information on funding and support and author affiliations appears at the end of the article. Clinical Implications Personality features can play a role in the diagnosis of depression. Personality assessment can play a role in treatment selection. Matching patient personality with treatment type maximizes response. Limitations To date, research has utilized widely diverse designs, resulting in different estimates of the relation between personality and depression. Results from treatment outcome studies are frequently limited by low levels of experimental control and statistical power. Results from treatment outcome studies may not generalize to everyday practice. Key Words: depression, personality disorders, personality, etiology, diagnosis, treatment selection Clinicians and researchers alike have long noted the prevalence of personality pathology in individuals with MDD. Indeed, the clarification of the personality features associated with major depression, and the implications of these associations for the understanding and care of major depression, have been the focus of much empirical work. Which personality features are common to individuals with MDD? What is the causal significance of this cooccurrence? Can clinicians make use of personality information during the assessment or treatment of depressed individuals? This review aims to shed some light upon these questions, through a review of personality models and their associations with major depression, the etiological connections that may underlie these associations, and the implications that personality may have for the diagnosis 14 La Revue canadienne de psychiatrie, vol 53, no 1, janvier 2008

2 Personality and Depression and treatment of MDD. It will quickly become apparent that previous research has generated more uncertainty than resolution. One key point, nonetheless, emerges clearly from this literature: consideration of personality features is critical to the understanding of depression and potentially of considerable utility in the optimization of its treatment. Models of Personality Personality Disorders With the introduction of DSM-III and the advent of the multiaxial system, personality features of clinical relevance were coded categorically as PDs on Axis II. 1 This system, with some modification, continues to be used in DSM-IV. 2 General criteria for PD describe a stable, pervasive, and inflexible pattern of atypical individual experience, manifested in cognition, affect, social functioning, and (or) impulse control. Individual criteria exist for specific PDs, which are divided into 3 clusters: cluster A includes paranoid, schizoid, and schizotypal PDs; cluster B includes antisocial, borderline, histrionic, and narcissistic PDs; and cluster C includes avoidant, dependent, and obsessive compulsive PDs. Each PD is defined by a set of pathological personality features, with a specific number of features required for a diagnosis. Individuals with MDD are characterized by higher rates of PDs, as compared with nondepressed samples. Methodological differences between studies, however, including method of assessment and sample characteristics, have resulted in widely varying estimates of this comorbidity, ranging from about 15% to 95%, with little consensus on the specific PDs most commonly exhibited by this population. In an early review, Corruble et al 3 argued that antisocial, borderline, histrionic, and schizotypal PDs are particularly common among depressed samples. In contrast, Skodol and colleagues 4 proposed that the most rigorous studies found elevated levels of borderline, avoidant, and dependent PDs in depressed individuals. A more recent investigation suggested that while all PDs are elevated among depressed samples, paranoid and obsessive compulsive PDs have the highest prevalence. 5 The relation between depression and PDs may further vary with the specific type of depression studied. For example, there is evidence that individuals with early-onset depression are more likely to have a comorbid PD, compared with those with late onset, as are those with atypical features, compared with those without. 6,7 Thus, whereas there is a general consensus Abbreviations used in this article DSM MDD PD Diagnostic and Statistical Manual of Mental Disorders major depressive disorder personality disorder that PDs are elevated among those with MDD, the extent and pattern of this comorbidity is not yet clear. Using an alternative approach to the DSM system, Akiskal 8 proposed that mood and personality pathology combine to produce patterns of chronic symptoms, termed subaffective personalities. He developed a system of affective temperaments corresponding to Axis I mood disorder categories (such as depressive and cyclothymic) that have been identified in clinical and nonclinical samples in several cultural groups and in cross-sectional and longitudinal research. 9,10 An adaptation of his original criteria for depressive personality is included in Appendix B of DSM-IV as depressive PD, which describes an early-onset and long-lasting set of traits resembling descriptions of chronic, mild depression. As currently defined, a depressive PD diagnosis is suggested if individuals meet the general criteria for a PD while also presenting with 5 of the following 7 traits: usual mood is dominated by dejection, gloominess, cheerlessness, joylessness, and unhappiness; self-concept centres around beliefs of inadequacy, worthlessness, and low self-esteem; critical, blaming, and derogatory toward self; brooding and given to worry; negativistic, critical, and judgmental toward others; pessimistic; and prone to feeling guilty or remorseful. 2 The criteria for depressive PD also specify that a diagnosis is precluded if dysthymic disorder is present. A major criticism of PDs is their extensive diagnostic overlap and comorbidity with each other. Depressive PD has been further faulted for its overlap not only with other PDs but also with dysthymic disorder In addition to the descriptive similarities, a wide range of overlap rates were reported, ranging from a low of 19%, to a high of 95%, with an average of around 50%. 4,13 21 Statistical modelling techniques have suggested that while the criterion sets for depressive PD and dysthymic disorder can be empirically separated, the best results are obtained when the 2 psychological symptoms of dysthymic disorder are considered to be part of both categories. 13,20,21 These 2 symptoms low self-esteem and feelings of hopelessness are present in the overwhelming majority of overlapping cases. 22 Indeed, researchers are increasingly arguing that the PDs are inherently flawed categories. 23 Concerns include difficulties with differentiating the traits of the various disorders, low levels of interrater agreement, unacceptable overlap rates, a lack of utility and comprehensiveness in the system as a whole, as well as the fundamental assumption that PDs are discrete categories Many of these problems are the result of imposing a categorical system on dimensional phenomena. 30 A dimensional model of personality, then, appears to be more compatible with phenomenological, biological, and genetic evidence. Within such a model, a flawed category such as depressive PD could be reconceptualized as a clinically relevant personality The Canadian Journal of Psychiatry, Vol 53, No 1, January

3 In Review Table 1 Dimensional approaches to personality: universal and pathological models Models Neuroticism, negative affectivity, emotional dysregulation Extraversion, positive affectivity Dissocial, antagonistic behaviour Constraint, compulsivity, conscientiousness Openness, absorption, psychoticism Universal personality models Eyesenck Personality Questionnaire Sixteen Personality Questionnaire NEO Personality Inventory-Revised Temperament and Character Inventory Multidimensional Personality Questionnaire Pathological personality models Dimensional Assessment of Personality Pathology Schedule for Nonadaptive and Adaptive Personality Personality Psychopathology Five Affective Temperament Schedule of Memphis, Pisa, Paris and San Diego Neuroticism Extraversion Psychoticism Psychoticism a Psychoticism Anxiety Extraversion Independence Self-control Toughmindedness a Neuroticism Extraversion Agreeableness a Conscientiousness Openness Harm avoidance Negative emotionality Emotional dysregulation Novelty seeking, reward dependence Positive emotionality Inhibition a Cooperativeness a Persistence, self-directedness Self-transcendence Constraint a Constraint Absorption Dissocial behaviour Constraint Negative affectivity Positive affectivity Disinhibition Disinhibition a Negative emotionality Depressive, cyclothymic Low positive Aggression Disconstraint Psychoticism emotionality a Hyperthymic, cyclothymic Irritable a Reversed or a low score dimension, with descriptive, clinical, and perhaps even biogenetic links with depression. 22 The question remains, however, as to the configuration of abnormal traits in an overarching system and the positions of individual constructs within such a system. It is to this issue that we now turn. Personality Dimensions A variety of dimensional models have been developed to characterize personality features of clinical relevance. 31 These models are frequently hierarchical, composed of higher-order, broad trait domains (for example, neuroticism) and lower-order, narrow traits known most commonly as facets (such as angry hostility). Models of pathological personality have largely been developed through the identification of dimensional personality traits underlying the PDs, for example, through the factor analysis of PD criteria and associated features. Three of the most popular models resulting from this strategy are assessed by the Dimensional Assessment of Personality Pathology, 32 the Schedule for Nonadaptive and Adaptive Personality, 33 and the Personality Psychopathology Five. 34 Akiskal proposed a more specific model that focuses primarily on the border between personality and affective disorder and includes Axis I and II PD features (this model is principally assessed by the Affective Temperament Schedule of Memphis, Pisa, Paris and San Diego). 35 In contrast, models of universal normal-range personality traits were developed independently from the current diagnostic system. These models have originated from varied paradigms, such as neuropharmacological and neuroanatomical investigations (for example, the Seven-Factor Psychobiological Model of Temperament and Character) 36 and factor analyses of trait descriptors found within language (for example, the Five-Factor Model). 37 The traits of such universal models are considered pathological when they are extreme and associated with significant distress and (or) impairment. 38 See Table 1 for a list of the current, most widely used measures of such dimensional models of personality. Compared with the PDs, the power of dimensional models of personality lies not only in their improved reliability and validity but also in their capacity for integration. Although these models have originated across fields of psychological 16 La Revue canadienne de psychiatrie, vol 53, no 1, janvier 2008

4 Personality and Depression inquiry, they have more similarities than differences. 39 Models of pathological personality appear to be best understood in terms of 4 of the 5 domains of the Five-Factor Model. 39, 40 Further, models of pathological and normal personality, as assessed in clinical and healthy populations, appear to share a hierarchical factor structure 41 that can be modelled as a hierarchy branching from 2 overarching supertraits (for example, externalizing and internalizing trait domains) to 5 broad domains resembling those of the Five-Factor Model (such as extraversion and neuroticism), and finally to numerous, more specific traits (such as gregariousness and selfconsciousness). 42 Thus, the Five-Factor Model in particular appears to demonstrate considerable promise as an integrating framework, as it not only adequately captures the content of pathological and normal models of personality but also has demonstrated psychometric soundness in clinical populations and clinical utility in the discrimination between psychiatric conditions Table 1 is organized with reference to these 5 broad domains. Figure 1 Etiological models of the association between personality and depression Individuals with MDD differ from nondepressed individuals upon measures of dimensional personality traits. Depressed individuals demonstrate elevated levels of broad traits such as neuroticism, negative emotionality, negative affectivity, and harm avoidance, as well as more specific vulnerability factors such as self-criticism, dependency, and perfectionism They further display reduced extraversion, positive emotionality, positive affectivity, and novelty seeking, as well as conscientiousness (see Bagby and Ryder 49 and Enns and Cox 50 for reviews). These and related traits reliably differ by 1 to 2 standard deviations in depressed individuals, compared with healthy individuals. Etiology of Relations Regardless of whether PDs or dimensions are used to characterize personality features of depressed individuals, there are 4 principal etiological models to account for observed relations between personality and MDD These models differ in the proposed relation between personality and depression, particularly the factors judged to be causally responsible for the comorbidity between personality features and MDD. These models are neither exhaustive nor mutually exclusive; indeed, a single model is unlikely to fully account for the complex association between personality and depression. See Figure 1 for a depiction of the following etiological models. Vulnerability Model. This model posits that personality features predispose individuals to develop MDD. By this view, personality constructs increase the likelihood that MDD will develop, and thus act as risk factors for this disorder. To use neuroticism as an example, this model would propose that individuals with high levels of neuroticism have a higher The Canadian Journal of Psychiatry, Vol 53, No 1, January

5 In Review likelihood of developing MDD than those with lower levels of this trait. Pathoplasty Model. This model proposes that personality features impact the way in which MDD is expressed. More specifically, personality affects the onset, severity or course of MDD, including response to treatment. Within this model, while neuroticism may not necessarily increase the likelihood that MDD will develop, high levels of neuroticism result in greater levels of severity and chronicity of this disorder, as well as a more negative prognosis. Complication/Scar Model. This model suggests that MDD leads to personality change. This model maintains that MDD influences personality, either transiently (the Complication Model) or permanently (the Scar Model). The former posits that personality trait change is a concomitant of MDD and will resolve with remission. In contrast, the latter posits that personality trait change due to MDD is fundamental and long-lasting. By this account, high levels of neuroticism are the result of MDD, which resolve with the disorder (complication) or remain elevated following remission (scar). Spectrum Model. This model asserts that an underlying dimension or continuum exists, extending from normal processes to mild, moderate, and severe pathological processes. Normative personality features represent normal processes, whereas pathological personality features represent a subclinical manifestation of psychopathology. Within this model, a dimension spans from normative levels of neuroticism, to elevated levels often associated with subclinical depressive symptoms and severe levels typically associated with clinically significant pathology known as MDD. Thus, this dimension accounts for normal and pathological personality as well as MDD, which are related to each other by degree rather than kind. Common Cause Model. Also known as the Shared Cause, Liability, or Factor Model, this model suggests that a shared etiological factor gives rise to personality features and MDD. For example, a genetic diathesis might give way not only to the personality features commonly manifest in MDD but also to MDD itself. By this account, genetic or biological bases such as serotonin functioning might result in personality features and MDD symptomatology. The longitudinal association between each personality construct related to MDD must be investigated to determine the propriety of each of these models to characterize that relation. However, the stability of each personality construct must be initially established before the stability of personality in the context of depression can be explored to elucidate these causal links. Stability and Change Stability of Personality Disorders As outlined by the DSM-IV, 2 PDs involve maladaptive patterns of thoughts, moods, and behaviours present from adolescence or early adulthood and stable throughout adult life. However, the DSM-IV provides little research evidence regarding such a course within the PDs, noting only that several disorders (that is, antisocial, borderline, and avoidant PDs) may remit with advancing age. Nevertheless, the putative rationale for the inclusion of PDs upon a separate axis within the current diagnostic system is an effort to ensure that such long-standing personality features are not overlooked in the presence of more florid, episodic Axis I disorders, clearly suggesting that the PDs are characterized by high, long-term stability. In fact, estimates of the stability of PDs have varied widely Most research conducted in this area is characterized by various methodological flaws, including limited sample size, high attrition rates, short follow-up periods, inadequate assessment methods (including unstandardized assessment instruments and limited attention to interrater reliability), inadequate characterization of comorbid Axis I and II conditions, and the absence of appropriate control groups. Further, very little research attention has been dedicated to PDs beyond borderline or antisocial. These limitations clearly compromise the ability of this research to establish the stability of the PDs; keeping this caveat in mind, early empirical work demonstrated only modest to moderate stability of the PDs, and substantial improvement and remission over time. More recent longitudinal research has bolstered these early findings 58 and has suggested that stability does not provide a 59, p 379 meaningful distinction between Axis I and II disorders. Test retest studies of PD diagnoses have been increasingly augmented with larger-scale longitudinal studies with multiple assessments including dimensional as well as categorical measures of personality disorder features. Preliminary research from the Longitudinal Study of Personality Disorders (LSPD) suggests considerable stability in PD features. 60 Subsequent research, however, has used more fine-grained methods of analysis to investigate PD stability in individual patients, yielding a more complex picture. Individual growth curve analysis is a technique that allows the researcher to study changes in individual study participants over time. This approach to the LSPD data revealed considerable individual- level variability in PD features, directly contradicting the assumption that such features are enduring. More specifically, these changes were typically and uniformly in the direction of decreasing personality disorder features over time. The Collaborative Longitudinal 61, p 1021 Study of Personality Disorders, meanwhile, has suggested that PD features may be hybrids, including pervasive, 18 La Revue canadienne de psychiatrie, vol 53, no 1, janvier 2008

6 Personality and Depression enduring, trait-like criteria along with intermittent, transient, state-like criteria. 62 These 2 types of criteria differ in prevalence and stability, and may contribute to the understanding of the differential and mean-level stability demonstrated in previous studies. Stability of Personality Traits As discussed, limitations in the PDs, such as diagnostic overlap and heterogeneity, resulted in greater attention toward dimensional conceptualizations of personality pathology. However, there are various ways in which to conceptualize the consistency or stability of personality dimensions at the group and the individual level (see Figure 2 for an illustration of the various types of stability). First, there are 3 group-level types of consistency, the first 2 of which are most familiar to researchers. Absolute or mean-level stability is the extent to which a group changes upon a trait over time, typically assessed with group comparisons. For example, the change in a patient group s average neuroticism levels from pretreatment to posttreatment provides an estimate of how patients changed upon this trait on the whole. Differential or rank-order stability is the extent to which relative differences among individuals within a group are maintained over time, typically assessed with test retest correlations. In this case, the correlation between a patient group s pretreatment neuroticism and posttreatment neuroticism results in an estimate of the extent to which the patients changed upon this trait over time, relative to each other. Structural or factorial stability is the extent to which the covariation among traits within a group change over time, typically assessed with structural equation modelling. For example, the extent to which associations between neuroticism and other variables changed from pretreatment to posttreatment provides an estimate of structural stability. Second, there are 2 further indices of consistency that study the individual person rather than the overall sample. Intraindividual or individual-level stability is the extent to which any individual changes on a given trait over time, assessed with a variety of techniques including individual growth curve analysis techniques. 63,64 This type of stability provides information regarding the number of patients who remain stable or who change on neuroticism from pretreatment to posttreatment, and in the case of more advanced statistical techniques, the rate and extent of such change. Ipsative stability is the extent to which the ordering and organization of traits within an individual change over time, requiring the assessment of the elevation, scatter, and shape of individual trait profiles over time. This type of stability provides an estimate of how extremity, variability, and patterns of elevations upon multiple traits change from pretreatment to posttreatment within specific individuals. James, 65 and more recently McCrae and Costa, 66 proposed that personality traits are essentially fixed by age 30. Indeed, empirical work has demonstrated that personality traits are characterized by high group-level stability, as demonstrated by high absolute, differential, and structural stability In a recent metaanalysis, Roberts and DelVecchio 70 demonstrated that personality exhibits high levels of consistency across the lifespan, which increases linearly until about age 50, at a level allowing for continued, minor changes at advancing ages. It appears then that traits are mostly consistent in adulthood, with some indication that they retain a dynamic quality. 70 In an innovative examination of the longitudinal stability of the PDs and the personality dimensions of the Five-Factor Model, Warner and colleagues 71 demonstrated that PDs and dimensional traits displayed high levels of rank order stability. More importantly, changes in traits significantly predicted changes in PDs, suggesting that the stability of the former underlies that of the latter. Indeed, this causal relation suggests that personality traits serve as the causal mechanism by which PDs are expressed. As noted by Warner and colleagues, 71 this pattern of results highlights the importance of research aimed toward understanding personality trait change, particularly in the context of treatment. Personality Stability in the Context of Acute Symptom Change As previously discussed, empirical work has demonstrated that personality functioning differs in individuals with MDD from that of healthy populations. These differences in personality functioning are reduced albeit not normalized with treatment, 72,73 fuelling arguments that personality features in depressed populations are incidental to depression. 74 However, these changes in personality are only modestly associated with changes in depression, suggesting that the personality changes displayed over the course of treatment represent true personality change rather than biased self-descriptions or temporary behavioural changes Indeed, in a recent and comprehensive investigation of the stability of personality dimensions in individuals receiving treatment for MDD, De Fruyt and colleagues 78 demonstrated absolute change in the domains of the Five-Factor Model, again only explained to a small extent by change in depression. They further demonstrated differential and structural stability for all domains, as well as intraindividual and ipsative stability for 4 of the 5 domains. Diagnostic Implications Regardless of their causal connections, is it possible to utilize personality information in the diagnosis of individuals with MDD? Again, estimates of the cooccurrence of personality pathology and MDD vary considerably, owing in large part to differences such as patient population, assessment The Canadian Journal of Psychiatry, Vol 53, No 1, January

7 In Review Figure 2 Types of stability 20 La Revue canadienne de psychiatrie, vol 53, no 1, janvier 2008

8 Personality and Depression Figure 3 Mediation, compared with moderation instruments, and other clinical and treatment characteristics. The consequence is to greatly increase the difficulty involved with integrating and understanding such information. 79 The specificity of relations between personality and MDD has been an area of some contention. 48 Evaluation of the links between personality and depression requires consideration of comorbidity between personality pathology and other Axis I disorders and Axis II PDs, and moreover, between MDD and other Axis I disorders The association of Axis I and II disorders with dimensional personality traits such as neuroticism has spurred many researchers to suggest a move toward a hierarchical dimensional model incorporating trait and state psychopathology. The emphasis upon the necessity of considering the comorbidities and continuities between Axis I and II 83 has resulted in the consideration of dimensions of behavioural and affective regulation that span traditional categorical boundaries. Siever and Davis 84 argued more than 15 years ago that phenomenological, genetic, and biological evidence suggests that 4 dimensions span all Axis I and II disorders: cognitive perceptual organization, impulsivity aggression, affective instability, and anxiety inhibition. At the same time, Clark, Watson, and colleagues, 85 focusing upon the affective disorders, proposed that one general negative affectivity dimension accounted for the covariation of anxiety and mood, while 2 specific dimensions of low positive affectivity and physiological hyperarousal differentiated them. The intervening years have seen numerous additions and refinements to these initial proposals. A recent article by Watson 86 proposed that sufficient information has accumulated regarding mood and anxiety disorders to allow for a grouping of disorders based on empirically demonstrated cooccurrence and covariation rather than rationally derived, shared, phenomenological features. His hierarchical model of Axis I affective disorders proposes a broad class of emotional disorders, within which bipolar, distress, and fear disorders are located. Within this model, MDD is described as a distress disorder, owing in part to its strong associations with negative and low positive affectivity. Watson emphasizes that the specificity of any relation between these dimensions and disorders must be viewed as relative, however, as very few would be specific to a single disorder of the DSM-IV. Indeed, low positive affectivity is associated with several conditions such as social phobia, generalized anxiety disorder, and schizophrenia. Treatment Implications Can personality information be of assistance in the treatment of individuals with MDD? Indeed, there are a variety of ways in which personality features may play a role in the treatment of MDD. Personality features may mediate the relation between interventions provided and depressive symptomatology, such that personality change can be considered the way in which treatments exert their effects on The Canadian Journal of Psychiatry, Vol 53, No 1, January

9 In Review outcome in MDD. Personality features may further moderate the relation between treatment and depressive symptomatology, such that particular personality features may be associated with treatment response in general or treatment response to specific types of treatment (see Figure 3 for an illustration of how these relations differ). Although the former model of influence has received little empirical attention to date, the latter idea has been explored using both PDs and dimensional traits. Personality Disorders and Treatment Outcome It has been long believed that PDs are associated with a poorer treatment outcome for individuals with MDD. 87 However, recent work has suggested that PDs might in fact not negatively impact prognosis as previously proposed. 88 In a recent review, Mulder 89 concluded that the impact of PD diagnosis upon treatment outcome in depressed samples varied widely with research design, and that the most methodologically robust studies demonstrated the least evidence of an association between PD and treatment outcome. This conclusion was bolstered by a subsequent metaanalysis of randomized control trials of pharmacotherapy by Kool et al, 90 but later contested by a metaanalysis of randomized control trials and case series by Newton-Howes et al. 91 Continuing empirical work, with sensitive statistical analyses are required 92 ; however, recent naturalistic 93 and controlled 94 studies suggest that a more optimistic view of the treatment outcome of MDD with comorbid PD may be justified. Personality Traits and Treatment Outcome It has similarly been long believed that certain personality traits are indicative of positive or negative prognosis in depression treatment. 37 Despite theorizing regarding what these types of associations might be, 95,96 there has been little systematic, empirical investigation of the ability of personality traits to predict treatment outcome in general or to specific treatments. Indeed, the investigations that were conducted to date have involved various and frequently low levels of methodological control and statistical power. Further systematic investigations of how personality traits can contribute to the prediction of general and specific treatment outcome and, ultimately, treatment selection and optimization are thus required. The empirical evidence for the associations between personality traits and treatment outcome is inconsistent at this time. There is support for the idea that neuroticism, negative affectivity, and related constructs are associated with negative outcome to treatment in general. 89 More recent evidence indicates that this trait is a negative prognostic indicator for psychotherapy but not for pharmacotherapy, suggesting that the latter might be the preferred treatment for individuals high in this trait. Bagby and colleagues 97 suggested that individuals with elevated neuroticism might be too emotionally dysregulated to recruit the psychological resources required to fully engage in psychotherapeutic efforts. 97 Extraversion appears to be consistently associated with a more favourable outcome to treatment in general. 72,98 Agreeableness has similarly exhibited a relation with positive treatment outcome, and more recently, a preferential response to psychotherapy, which is interpretive rather than supportive, and interpersonal rather than cognitive in orientation Some evidence exists to support associations linking both openness to experience and conscientiousness to general positive outcome; however, the majority of studies have failed to find such an association. 96,100,101 Clinical assessment has been criticized for the lack of empirical evidence for its practical value, such as contributing to favourable treatment outcomes. 102 There has simultaneously been a call for treatment selection to incorporate individual difference information. 103 The examination of personality trait relations with differential treatment response promises to develop explicit guidelines regarding how to make practical use of individual difference information during treatment selection, and thus provide an evidence-based way in which to tailor treatment selection to a particular patient. Conclusions and Future Directions A consideration of the relation between personality and depression requires attention to an abundance of theoretical, empirical, and methodological complexities. That said, theoretical and empirical work conducted to date converges on several broad conclusions, in response to the queries we posed at the outset. Although empirical work does suggest that PDs are found at elevated levels within depressed samples, the prevalence of each specific PD within this population is indeterminate. This uncertainty may result from not only variations in research design but also the limited reliability and validity of the PD diagnoses themselves. Personality dimensions can be more reliably assessed, and appear to more validly reflect the variation found in phenomenological, biological, and genetic data. Empirical work consistently demonstrates that depressed individuals exhibit elevated levels of neuroticism and negative affectivity, and reduced levels of extraversion and conscientiousness. A variety of causal models can account for the relation between personality and MDD. Longitudinal investigations of personality and depression are required to establish which of these etiological models is most appropriate. Such longitudinal relations may be more difficult to investigate among PDs, owing to their limited stability; however, personality dimensions have demonstrated considerable evidence of group- and individual-level stability, even in the context of 22 La Revue canadienne de psychiatrie, vol 53, no 1, janvier 2008

10 Personality and Depression acute symptom change. Indeed, personality changes over the course of treatment are only minimally associated with symptom reductions, suggesting that treatment results in meaningful personality change. While additional research is most certainly required, existing data is broadly consistent with vulnerability, pathoplasty, spectrum, and common cause models. Personality features can be used to assist diagnosis; however, they are unlikely to provide definitive evidence for one diagnosis rather than another, owing to the range of specificity of the relations between personality and depression. For example, elevated neuroticism is found across a broad range of clinical disorders, while reduced extraversion is found across a more limited extent of pathology. Personality features have thus been used not only to assist diagnosis but also to inform potential modifications to the diagnostic system as a whole. The relation between personality and depression is complex, owing to the hierarchical relation between them, as well as other clinical conditions. Personality dimensions are on the forefront of discussions regarding how to improve diagnostic clarification, and may provide a useful way in which to understand and model the comorbidities among and between Axis I and II conditions. Personality features can be used to assist treatment. Various effective treatments for depression exist; however, a substantial proportion of individuals do not remit in response to the treatment they are offered. Increased attention is now paid to ways of optimizing the delivery of existing interventions through the determination of what works best for whom. For example, recent work has suggested that elevated levels of neuroticism may be a negative prognostic indicator for psychotherapy but not pharmacotherapy suggesting that the latter may be the treatment of choice for individuals high in this trait. In contrast, elevated levels of agreeableness appear to be a positive prognostic indicator for psychotherapy, particularly of an interpersonal orientation suggesting that this type of treatment may be preferred for those high in this trait. This line of research suggests that the assessment of personality characteristics may ultimately provide an avenue by which clinicians can provide empirically-based, patient-centred care care that is appropriately sensitive to patient diversity as manifested in personality features. Funding and Support An honorarium is available for the In Review series. Acknowledgements This work was supported, in part, by a Senior Research Fellowship awarded to Dr RM Bagby by the Ontario Mental Health Foundation, and a Postdoctoral Fellowship awarded to Dr LC Quilty by the Canadian Institutes of Health Research. References 1. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 3rd ed. Washington (DC): American Psychiatric Association; American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington (DC): American Psychiatric Association; Corruble E, Ginestet D, Guelfi JD. Comorbidity of personality disorders and unipolar major depression: a review. J Affect Disord. 1996;37(2 3): Skodol AE, Stout RL, McGlashan TH, et al. Co-occurrence of mood and personality disorders: a report from the Collaborative Longitudinal Personality Disorders Study (CLPS). Depress Anxiety. 1999;10: Grant BF, Hasin DS, Stinson FS, et al. Co-occurrence of 12-month mood and anxiety disorders and personality disorders in the US: results from the National Epidemiologic Survey on Alcohol and Related Conditions. J Psychiatr Res. 2005;39(1): Alpert JE, Uebelacker LA, McLean NE, et al. Social phobia, avoidant personality disorder and atypical depression: co-occurrence and clinical implications. Psychol Med. 1997;27(3): Fava M, Alpert JE, Borus JS, et al. Patterns of personality disorder comorbidity in early onset versus late-onset major depression. Am J Psychiatry. 1996;153(10): Akiskal HS. Characterologic manifestations of affective disorders: toward a new conceptualization. Integr Psychiatr. 1984;2(3): Akiskal HS, Kilzieh N, Maser JD, et al. The distinct temperament profiles of bipolar I, bipolar II and unipolar patients. J Affect Disord. 2006;92(1): Akiskal HS, Placidi GF, Maremmani I, et al. TEMPS-I: delineating the most discriminant traits of the cyclothymic, depressive, hyperthymic and irritable temperaments in a nonpatient population. J Affect Disord. 1998;51(1): Bagby RM, Ryder AG, Schuller DR. Depressive personality disorder. Curr Psychiatry Rep. 2003;5: McDermut W, Zimmerman M, Chelminski I. The construct validity of depressive personality disorder. J Abnorm Psychol. 2003;112(1): Ryder AG, Bagby RM. Diagnostic viability of depressive personality disorder: theoretical and conceptual issues. J Personal Disord. 1999;13(2): Bagby RM, Ryder AG. The diagnostic discriminability of dysthymia and depressive personality disorder. Depress Anxiety. 1999;10: Hirschfeld RMA, Holzer CE. Depressive personality disorder: clinical implications. J Clin Psychiatry. 1994;55(4 Suppl): Huprich SK. Object loss and object relations in depressive personality analogues. Bull Menninger Clin. 2001;65(4): Klein DN, Miller GA. Depressive personality in nonclinical subjects. Am J Psychiatry. 1993;150(11): Klein DN, Shih JH. Depressive personality: associations with DSM-III-R mood and personality disorders and negative and positive affectivity, 30-month stability, and prediction of course of Axis I depressive disorders. J Abnorm Psychol. 1998;107(2): Phillips KA, Gunderson JG, Triebwasser J, et al. Reliability and validity of depressive personality disorder. Am J Psychiatry. 1998;155(8): Ryder AG, Bagby RM, Dion KL. Chronic, low-grade depression in a nonclinical sample: depressive personality or dysthymia? J Personal Disord. 2001;15(1): Ryder AG, Schuller DR, Bagby RM. Depressive personality and dysthymia: evaluating symptom and syndrome overlap. J Affect Disord. 2006;91(2 3): Ryder AG, Bagby RM, Marshal MB, et al. The depressive personality: psychopathology, assessment, and treatment. In: Rosenbluth M, Kennedy SH, Bagby RM, editors. Depression and Personality. Washington (DC): American Psychiatric Publishing, Inc; p Ryder AG, Bagby RM, Schuller DR. The overlap of depressive personality disorder and dysthymia: a categorical problem with a dimensional solution. Harv Rev Psychiatry. 2002;10(6): Clark LA, McEwen JL, Collard LM, et al. Symptoms and traits of personality disorder: two new methods for their assessment. Psychol Assess. 1993;5(1): Clark LA, Livesley WJ, Morey L. Personality disorder assessment: the challenge of construct validity. J Personal Disord. 1997;11(3): Livesley WJ. The classification of personality disorder: I. The choice of category concept. Can J Psychiatry. 1985;30(5): Tyrer P. What s wrong with DSM-III personality disorders? J Personal Disord. 1988;2(4): Widiger TA. The DSM-III-R categorical personality disorder diagnoses: a critique and an alternative. Psychol Inq. 1993;4(2): Widiger TA, Frances A. The DSM-III personality disorders: perspectives from psychology. Arch Gen Psychiatry. 1985;42(6): Bagby RM, Schuller DR, Marshall MB, et al. Depressive personality disorder: rates of comorbidity with personality disorder and relations to the Five-Factor Model of personality. J Personal Disord. 2004;18(6): Trull TJ, Durrett CA. Categorical and dimensional models of personality disorder. Annual Review of Clinical Psychology. 2005;1(1): Livesley WJ, Jackson DW, Schroeder ML. Dimensions of personality pathology. Can J Psychiatry. 1991;36(8): The Canadian Journal of Psychiatry, Vol 53, No 1, January

11 In Review 33. Clark LA. Schedule for nonadaptive and adaptive personality (SNAP). Manual for administration, scoring and interpretation. Minneapolis (MN): University of Minnesota Press; Harkness AR, McNulty JL. The personality psychopathology five (PSY-5): issue from the pages of a diagnostic manual instead of a dictionary. In: Strack S, Lorr M, editors. Differentiating normal and abnormal personality. New York (NY): Springer; Placidi GF, Signoretta S, Liguori A, et al. The semi-structured affective temperament interview (TEMPS-I): reliability and psychometric properties in year-old students. J Affect Disord. 1998;47(1 3): Cloninger CR, Svrakic DM, Przybeck TR. A psychobiological model of temperament and character. Arch Gen Psychiatry. 1993;50(12): Costa PT, McCrae RR. Revised NEO Personality Inventory: professional manual. Odessa (FL): Psychological Assessment Resources; Lynam DR, Widiger TA. Using the Five-Factor Model to represent the DSM-IV personality disorders: an expert consensus approach. J Abnorm Psychol. 2001;110(3): Widiger TA. Four out of five ain t bad. Arch Gen Psychiatry. 1998;55(10): O Connor BP, Dyce JA. Rigid and extreme: a geometric representation of personality disorders in Five-Factor Model space. J Pers Soc Psychol. 2001;81(6): O Connor BP. The search for dimensional structure differences between normality and abnormality: a statistical review of published data on personality and psychopathology. J Pers Soc Psychol. 2002;83(4): Markon KE, Krueger RF, Watson D. Delineating the structure of normal and abnormal personality: an integrative hierarchical approach. J Pers Soc Psychol. 2005;88(1): Bagby RM, Costa PT, McCrae RR, et al. Replicating the Five-Factor Model of personality in a psychiatric sample. Pers Individ Dif. 1999;27(6): Bagby RM, Costa PT, Widiger TA, et al. DSM-IV personality disorders and the Five-Factor Model of personality: a multi-method examination of domain and facet level predictions. Eur J Pers. 2005;19(4): Reynolds SK, Clark LA. Predicting dimensions of personality disorder from domains and facets of the Five-Factor Model. J Pers. 2001;69(2): Beck AT. Cognitive therapy of depression: new perspectives. In: Clayton PJ, Barrett JE, editors. Treatment of depression: old controversies and new approaches. New York (NY): Raven; p Blatt SJ, D Afflitti JP, Quinlan DM. Experiences of depression in normal adults. J Abnorm Psychol. 1976;85(4): Cox BJ, McWilliams LA, Enns MW, et al. Broad and specific personality dimensions associated with major depression in a nationally representative sample. Comp Psychiatry. 2004;45(4): Bagby RM, Ryder AG. Personality and the affective disorders: past efforts, current models, and future directions. Curr Psychiatry Rep. 2000;2: Enns MW, Cox BJ. Personality dimensions and depression: review and commentary. Can J Psychiatry. 1997;42(3): Farmer R, Nelson-Gray RO. Personality disorders and depression: hypothetical relations, empirical findings, and methodological considerations. Clin Psychol Rev. 1990;10: Widiger TA, Trull TJ. Personality and psychopathology: an application of the Five-Factor Model. J Pers. 1992;60: Klein MH, Wonderlich S, Shea MT. Models of relationships between personality and depression: toward a framework for theory and research. In: Klein MH, Kupfer DJ, Shea MT, editors. Personality and depression: a current view. New York (NY): The Guilford Press; p Shea MT, Yen S. Personality traits/disorders and depression: a summary of conceptual and empirical findings. In: Rosenbluth M, Kennedy SH, Bagby RM, editors. Personality and depression: conceptual and clinical challenges. Washington (DC): American Psychiatric Publishing, Inc: p Grilo CM, McGlashan TH, Oldham JM. Course and stability of personality disorders. Journal Practical Psychiatry and Behavioral Health. 1998;4: McDavid JD, Pilkonis PA. The stability of personality disorder diagnoses. J Personal Disord. 1996;10(1): Perry JC. Longitudinal studies of personality disorders. J Personal Disord. 1993;Suppl 1: Grilo CM, Sanislow CA, Gunderson JG, et al. Two-year stability and change of schizotypal, borderline, avoidant and obsessive-compulsive personality disorders. J Consult Clin Psychol. 2004;72(5): Shea MT, Yen S. Stability as a distinction between Axis I and Axis II disorders. J Personal Disord. 2003;17(5): Lenzenweger MF. Stability and change in personality disorder features: the longitudinal study of personality disorders. Arch Gen Psychiatry. 1999;56(11): Lenzenweger MF, Johnson MD, Willett JB. Individual growth curve analysis illuminates stability and change in personality disorder features: the longitudinal study of personality disorders. Arch Gen Psychiatry. 2004:61(10): McGlashan TH, Grilo CM, Sanislow CA, et al. Two-year prevalence and stability of individual DSM-IV criteria for schizotypal, borderline, avoidance and obsessive compulsive personality disorders: toward a hybrid model of Axis II disorders. Am J Psychiatry. 2005;162(5): Block J, Robins RW. A longitudinal study of consistency and change in self-esteem from early adolescence to early adulthood. Child Dev. 1993;64(3): Tate RL, Hokanson JE. Analyzing individual status and change with hierarchical models: illustraction with depression in college students. J Pers. 1993;61(2): James W. The principles of psychology. New York (NY): Henry Holt and Co; McCrae RR, Costa PT. Personality in Adulthood. New York (NY): The Guilford Press; Schuerger JM, Zarrella KL, Hotz AS. Factors that influence the temporal stability of personality by questionnaire. J Pers Soc Psychol. 1989;56(5): Funder DC, Parke RD, Tomlinson-Keasey C, et al. Studying lives through time: personality and development. Washington (DC): American Psychological Association; McCrae RR, Costa PT. Personality in adulthood: a Five-Factor Theory perspective. 2nd ed. New York (NY): The Guilford Press; Roberts BW, DelVecchio WF. The rank order consistency of personality traits from childhood to old age: a quantitative review of longitudinal studies. Psychol Bull. 2000;126(1): Warner MB, Morey LC, Finch JF, et al. The longitudinal relationship of personality traits and disorders. J Abnorm Psychol. 2005;113(2): Bagby RM, Joffe RT, Parker JDA, et al. Major depression and the Five-Factor Model of personality. J Personal Disord. 1995;9(3): Fava M, Farabaugh AH, Sickinger AH, et al. Personality disorders and depression. Psychol Med. 2002;32(6): Barnett PA, Gotlib IH. Psychosocial functioning and depression: distinguishing among antecedents, concomitants, and consequences. Psych Bull. 1988;104(1): Costa PT, Bagby RM, Herbst JH, et al. Personality self-reports are concurrently reliable and valid during acute depressive episodes. J Affect Disord. 2005;89(1 3): Santor DA, Bagby RM, Joffe RT. Evaluating stability and change in personality and depression. J Pers Soc Psychol. 1997;73(6): Trull TJ, Goodwin AH. Relationship between mood changes and the report of personality disorder symptoms. J Pers Assess. 1993;61(1): De Fruyt F, Van Leeuwen K, Bagby RM, et al. Assessing and interpreting personality change and continuity in patients treated for major depression. Psychol Assess. 2006;18(1): Farabaugh A, Mischoulon D, Fava M, et al. The overlap between personality disorders and major depressive disorder (MDD). Ann Clin Psychiatry. 2004;16(4): Hasin DS, Goodwin RD, Stinson FS, et al. Epidemiology of major depressive disorder: results from the National Epidemiologic Survey on Alcoholism and Related Conditions. Arch Gen Psychiatry. 2005;62(10): Kessler RC, Berglund R, Demler O, et al. National Comorbidity Survey Replication. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). JAMA. 2003;289(23): Mineka S, Watson D, Clark LA. Comorbidity of anxiety and unipolar mood disorders. Annual Review of Psychology. 1998;49: Grilo CM, McGlashan TH, Skodol AE. Stability and course of personality disorders: the need to consider comorbidities and continuities between Axis I psychiatric disorders and Axis II personality disorders. Psychiatr Q. 2000;71(4): Siever LJ, Davis KL. A psychobiological perspective on the personality disorders. Am J Psychiatry. 1991;148(12): Clark LA, Watson D. Tripartite Model of anxiety and depression: psychometric evidence and taxonomic implications. J Abnorm Psychol. 1991;100(3): Watson D. Rethinking the mood and anxiety disorders: a quantitative hierarchical model for DSM-V. J Abnormal Psychol. 2005;114(4): Shea MT, Widiger TA, Klein MH. Comorbidity of personality disorders and depression: implications for treatment. J Consult Clin Psychol. 1992;60(6): Russell JM, Kornstein SG, Shea MT, et al. Chronic depression and comorbid personality disorders: response to sertraline versus imipramine. J Clin Psychiatry. 2003;64(5): Mulder RT. Personality pathology and treatment outcome in major depression: a review. Am J Psychiatry. 2002;159(3): Kool S, Schoevers R, de Maat S, et al. Efficacy of pharmacotherapy in depressed patients with and without personality disorders: a systematic review and meta-analysis. J Affect Disord. 2005;88(3): Newton-Howes G, Tyrer P, Johnson T. Personality disorder and the outcome of depression: meta-analysis of published studies. Br J Psychiatry. 2006;188(1): Grilo CM, Sanislow CA, Shea MT, et al. Two-year prospective naturalistic study of remission from major depressive disorder as a function of personality disorder comorbidity. J Consult Clin Psychol. 2005;73(1): Brieger P, Ehrt U, Bloeink R. Consequences of comorbid personality disorders in major depression. J Nerv Ment Dis. 2002;190(5): La Revue canadienne de psychiatrie, vol 53, no 1, janvier 2008