Research report
Personality and depressive symptoms: a multi-dimensional analysis

https://doi.org/10.1016/S0165-0327(02)00303-8Get rights and content

Abstract

Background: The relationship of temperamental aspects of personality to symptoms of depression in a community-based sample of 804 individuals was examined using a multi-dimensional approach to account for heterogeneity in symptom patterns. Method: The Temperament and Character Inventory (TCI) was used to assess personality and the Center for Epidemiologic Studies Depression scale (CES-D) was used to measure depressive symptoms. Canonical correlation analysis was used to relate CES-D item combinations to temperament traits in multiple dimensions. The relationships between temperament and various conditions correlated with depression were examined using logistic regression. Results: Temperamental aspects of personality are related not only to total CES-D score, but also to the patterns of CES-D items endorsed by subjects. High Harm Avoidance is related to total CES-D score; high Reward Dependence combined with high Persistence is associated with restless sleep and subjective symptoms; high Reward Dependence combined with low Persistence is negatively associated with appetite loss and low energy; high Novelty Seeking is related to maintenance of positive affect and inability to concentrate. High Novelty Seeking is also associated with past suicide attempts, after adjusting for total CES-D score. Limitations: Cross-sectional data prevent analysis of causation; the severest cases of clinical depression may not be represented in a general population sample. Depressive symptoms are self-reported. Conclusion: Substantial differences in level of symptoms and in symptom patterns exist among individuals in a continuum of depressed states and those differences are partially explained by temperament traits.

Introduction

The impact of major depression on public health has become more thoroughly understood in recent years (Olfson et al., 2002, Kessler et al., 1994). Yet, sub-clinical variants of depression and depressed mood states as assessed by self-report checklists also tend to be associated with significant impairment and elevated risk for adverse health outcomes (Wells et al., 1989, Angst and Merikangas, 1997). Hence, a detailed understanding of depressed mood and sub-clinical depressive symptoms in their own right is merited. One potentially productive avenue of research involves the relationships between personality and depression. Depressive symptoms are risk factors for adverse medical and psychological outcomes. However, correlations between personality and mood-state are strong enough that it might be difficult to disentangle whether transient mood-states, enduring personality traits, or both are relevant in determining a particular outcome. Hence, a detailed understanding of the interrelationships between mood and personality traits is desirable.

A great deal is known about personality traits and dispositions of individuals who experience depressed mood with greater frequency and severity than average (Enns and Cox, 1997) but it might be simplistic to regard depressed mood as a unitary phenomenon. Differences in course, symptoms and familiality led to the distinction between unipolar and bipolar depression and even more fine-grained distinctions have been suggested based on these and other criteria (e.g. Akiskal et al., 1989, Chen et al., 2000). Hence, it is reasonable to view the full spectrum of depressive symptoms, both clinical depression and less severe syndromes of depressed mood, as possessing a severity dimension as well as other dimensions related to patterns of expression of symptoms. If dispositional differences influence the rate at which particular individuals experience negative affective states, perhaps aspects of personality also influence how these states are experienced. In other words, depressed mood can be viewed as a multi-dimensional phenomenon in which a ‘quantitative’ dimension describes the severity of depressed mood and other vectors describe the relative prominence of particular patterns of symptoms. We will refer to the latter variables as qualitative dimensions.

Many studies are available that relate overall depression scores or diagnosis to personality dimensions (reviewed in Enns and Cox, 1997); however, we hypothesize that multi-dimensional relationships between aspects of personality and depressive symptoms should be observable. In other words, while personality traits are predictive of scores on depression rating scales, the qualitative experience of depressive symptoms is also expected to be related to an individual’s personality configuration. This is conceptually analogous to the hypothesis that temperamental differences could influence the symptomatology, course or self-reporting of symptoms in clinical depression (e.g. Cassano et al., 1992, Akiskal, 1984), although our analysis focuses on self-reported personality and symptoms of depression in a community sample. In this situation, most mood disturbances will be sub-clinical. Additionally, we expect that personality will influence other adverse outcomes often associated with depressed mood even after level or severity of depressed mood is taken into account. In this paper, we test these hypotheses through a detailed examination of the relationships between the Center for Epidemiologic Studies Depression Scale (CES-D: Radloff, 1977), a widely used self-assessment for depressed mood, and the temperament scales from Cloninger’s Temperament and Character Inventory, using data from a large community-based sample. We choose to focus on temperament because these scales exhibit low inter-correlations and can be reliably rated by observers (e.g. Cloninger et al., 1988), thereby making them relevant to a clinical setting.

The CES-D is commonly used to assess symptoms of depression in epidemiologic studies. It can be used to screen for major depression, although high scores may also be a result of generalized anxiety disorder, depressive personality, or other conditions (Breslau, 1985, Roberts and Vernon, 1983, Myers and Weissman, 1980). Hence, the CES-D might lack diagnostic specificity but it is a valid measure of depressive symptoms associated with both psychological distress and more serious depression, and it is particularly useful in a general population sample where a relatively low prevalence of current major depression is expected. We therefore use the CES-D in this work as a measure of ‘depressive symptoms,’ recognizing that such symptoms are not necessarily specific to clinical depression.

Our objectives were to establish relationships between temperament dimensions assessed by the TCI and the 20 items measured by the CES-D. We hypothesized that meaningful relationships could be found between temperament dimensions and combinations of CES-D items and that some of these combinations are unrelated to severity of depressed mood, i.e. qualitative dimensions that are orthogonal to total CES-D score. Multidimensional relationships were dissected using canonical correlation analysis of the temperament scales and the set of 20 CES-D items. Subsequently, we examined the practical implications of these statistical analyses by assessing the relationships between temperament scales and variables associated with complications of depression, namely, history of suicide attempts, gastro-intestinal distress, panic attacks and symptoms of alcoholism, while adjusting for severity of depressive symptoms (total CES-D score). We believe that these exploratory analyses provide a useful methodological framework for understanding individual differences in the experience of negative affective states.

Section snippets

Subjects

In 1994, a survey of personality and mental health was administered to a stratified random sample representing the adult population of the greater metropolitan area of St Louis, MO. Subjects were selected from standard telephone lists. Participants were asked to complete a self-report booklet that included, among other measures, the Temperament and Character Inventory (TCI: Cloninger et al., 1994), the short Michigan Alcohol Screening test (MAST: Selzer et al., 1975), the NIMH Center for

Distribution of CES-D scores and bivariate correlations with temperament

The distribution of CES-D scores was highly skewed to the left (Skewness=1.4) with a mean value of 11.9 (S.D.=10.0) and a median score of 9; 25% of the sample met the cut-off score for depression of 16 or higher. The cut-off score of 16 corresponds to endorsement of at least six items (Radloff, 1977). Although this cut-off may lack specificity for a clinical diagnosis, it is noteworthy that this is the optimal cut-off for longitudinal prediction of future coronary heart disease in males and is

Discussion

The primary purpose of these analyses was to determine whether depressed mood and its relationships to personality could be viewed from a multidimensional perspective. In other words, we sought to determine whether or not personality could predict more about depressive syndromes than their severity, i.e. whether personality is related to qualitative aspects of depressed mood. The results of the canonical correlation analysis, shown in Table 1, Table 2, illustrate a systematic analysis of these

Conclusion

Through the analyses of the relationships between personality and CES-D responses, we propose that personality traits not only affect an individual’s susceptibility to symptoms of depression, but also influence the types of symptoms and complications that are experienced by a particular person. Several promising studies have suggested the utility of personality assessment in predicting differential response of depression to pharmacotherapy and the advantages of psychotherapy informed by

Acknowledgments

This work was supported in part by NIH grant MH31302. RAG is an NIMH trainee in Psychiatric Epidemiology and Biostatistics, NIH grant MH17104. We are grateful to Dr Barry Hong for numerous helpful suggestions, to Madeleine Grucza for editorial review, and to the anonymous reviewers for their comments.

References (30)

  • C.R. Cloninger

    The genetics and psychobiology of the seven-factor model of personality

  • C.R. Cloninger et al.

    The Temperament and Character Inventory (TCI): A guide to its development and use

    (1994)
  • C.R. Cloninger et al.

    Childhood personality predicts alcohol abuse in young adults

    Alcohol Clin. Exp. Res.

    (1988)
  • C.R. Cloninger et al.

    Integrative psychobiological approach to psychiatric assessment and treatment

    Psychiatry

    (1997)
  • F.L. Collins et al.

    The integration of empirically derived personality assessment data into a behavioral conceptualization and treatment plan. Rationale, guidelines, and caveats

    Behav. Modif.

    (1993)
  • Cited by (86)

    • Temperament and character profiles associated with depression and treatment response in patients with or without comorbid substance abuse

      2016, Psychiatry Research
      Citation Excerpt :

      However, in the AUP group it explained over one fourth of the variance in ΔMADRS, and the increase in RD contributed independently to the antidepressive treatment response. This could mean that the depressive symptoms in patients with alcohol use problems are more related with loss of energy, appetite loss, vegetative symptoms or subjective dysphoria (Grucza et al., 2003) and could be explained with differences in the function of brain anti-reward mechanisms associated with noradrenergic circuits and the hypothalamic-pituitary-adrenal (HPA) axis, with potential moderation by endogenous oxytocin (Bell et al., 2006; BuismanPijlman et al., 2014; Cloninger, 1986; Koob and Le Moal, 2008). There may be more reactive depression state dependent changes in these functions reflected in altered RD scores, which are a marked predictor of antidepressive treatment response in AUP patients.

    • Do spirituality and religiousness differ with regard to personality and recovery from depression? A follow-up study

      2016, Comprehensive Psychiatry
      Citation Excerpt :

      Personality is defined as the characteristics of a person responsible for consistency in feelings and thoughts and is determined genetically and by environmental factors which is also believed, to a certain extent, for the concepts of spirituality and religiousness [20]. Personality characteristics have also been associated with depression in community samples as well in other populations [21,22,23]. The personality model especially relevant for the research of associations between personality and spirituality/religiousness is psychobiological model by Cloninger and his collegues.

    View all citing articles on Scopus
    View full text