Patients' and informants' reports of personality traits during and after major depression

Conceptualization and assessment of personality factors in depression

Research that relates personality to depression is one of the dominant themes in the clinical literature. The current paper examines this research from a critical perspective. It is argued that existing research is limited by (i) a failure to adopt a broad conceptual approach to the study of personality and depression; and (ii) the use of personality measures with questionable psychometric properties. Our observations lead us to suggest that greater adherence to established methodology and conceptual developments in the personality field will result in substantial improvements in research on personality and depression, and may ultimately provide a more accurate appraisal of the role of personality factors in depression. In addition to examining important issues, key directions for future research are discussed.

Agreement Between Self- and Informant-Reported Ratings of Personality Traits: The Moderating Effects of Major Depressive and/or Panic Disorder

Several personality traits are risk factors for psychopathology. As symptoms of psychopathology may influence self-rated personality, informant reports of personality are also sometimes collected. However, little is known about self-informant agreement in individuals with anxiety and/or depression. We investigated whether self-informant agreement on positive and negative affectivity (PA and NA) and anxiety sensitivity differs for individuals with major depressive disorder (MDD) and/or panic disorder (PD; total n = 117). Informant- and self-reported PA was correlated among those with MDD, but not among those without MDD. Informant- and self-reported anxiety sensitivity was correlated among those with PD, but not among those without PD. Informant- and self-reported NA was correlated irrespective of diagnosis. Results indicate that the agreement of self- and informant-reported personality may vary as a function of depression and/or anxiety disorders.

Reliability of personality disorder diagnosis during depression: the contribution of collateral informant reports

OBJECTIVE

Research has found low concordance of personality disorder diagnoses made during depression versus after remission and made using patient versus collateral informants, but little is known about the reliability of personality disorder (PD) diagnoses made during depression using patient and collateral reports.

METHOD

A total of 168 patients were evaluated for PDs during depression and following response using patient and close informant reports. kappa coefficients of inter-informant and test-retest reliability were calculated.

RESULTS

After depression response, the proportion diagnosed with cluster A and C PDs fell by both patient and close informant report, and overall inter-informant reliability declined. Overall test-retest reliability did not differ between patients and informants.

CONCLUSION

Collateral informants do not improve the reliability of PD diagnoses made during depressive episodes.

Cluster C personality disorders in major depressive episodes: the relationship between hostility and suicidal behavior

There is some evidence for an association between Cluster C Personality Disorders (CCPD) and suicidal behavior. We compared depressed inpatients with and without CCPD in terms of suicidal behavior and associated psychopathology. Cluster A or B personality disorder co-morbidity were exclusion criteria for both groups (cases and controls). Depressed inpatients with "pure" CCPD had higher levels of suicidal ideation but not more previous suicide attempts compared with patients without CCPD. Greater suicidal ideation in depressed patients with CCPD in our study was associated with more hostility. Future studies examining the relationship between suicidal ideation and hostility in CCPD may clarify whether treatment focused on hostility might be of use for decreasing suicidal ideation in depressed patients with CCPD (Spitzer, Williams, Gibbon et al., 1990).

Aggression and substance abuse in bipolar disorder

OBJECTIVES

The goal of this retrospective study was to examine factors differentiating persons with bipolar disorder who did or did not have comorbid lifetime substance use disorders (SUD) at an index assessment. We also explored the chronology of onset of mood and SUD.

METHODS

We studied 146 subjects with DSM-defined bipolar disorder. Subgroups with and without lifetime SUD were compared on demographic and clinical measures.

RESULTS

Substance abuse disorders in this bipolar sample were associated with male sex, impulsive-aggressive traits, comorbid conduct and Cluster B personality disorders, number of suicide attempts and earlier age at onset of a first mood episode. In a multivariable logistic regression analysis, male sex and aggression and possibly earlier age at mood disorder onset were associated with SUD. In those with or without SUD, the first mood episode tended to be depressive and to precede the onset of SUD.

CONCLUSIONS

In persons with bipolar disorder, an earlier age of onset and aggressive traits appear to be factors associated with later development of comorbid SUD.

Assessment of Adult Psychopathology: Meta-Analyses and Implications of Cross-Informant Correlations

Assessment of adult psychopathology relies heavily on self-reports. To determine how well self-reports agree with reports by "informants" who know the person being assessed, the authors examined 51,000 articles published over 10 years in 52 peer-reviewed journals for correlations between self-reports and "informants" reports. Qualifying correlations were found in 108 (0.2%) of the articles. When self-reports and informant reports were obtained with parallel instruments, mean cross-informant correlations were .681 for substance use, .428 for internalizing, and .438 for externalizing problems. When based on different instruments, the mean cross-informant correlation was .304. The moderate sizes of the correlations argue for systematically obtaining multi-informant data. National survey findings were used to illustrate practical ways to obtain and use such data.

Five-factor model prototypes for personality disorders: the utility of self-reports and observer ratings

The current study examined the prototype-matching technique for using the five-factor model (FFM) of personality to assess personality disorders (PDs) and their correlates. The sample was composed of 69 psychiatric patients, most of whom suffered from affective or anxiety disorders. The participants were predominantly outpatients (78%), Caucasian (94%), and women (58%). NEO Personality Inventory (NEO-PI) data from these patients and informants were available for deriving PD prototype scores. The results supported four conclusions. First, the FFM prototype-matching method is useful for clinical samples. Second, agreement between self-reports and information from significant others (SOs) using this method is good compared to previous results. Third, this agreement varies systematically with the observability of the PD criteria. Fourth, the value of self- and other-reported personality information depends on the rating source of the outcome variables. Other-reported information provides incremental utility in predicting impairment and interpersonal distress rated by clinicians and significant others.

Concordance between self- and observer-ratings on Kasahara's Inventory for the Melancholic Type Personality

The concordance between self- and observer-ratings was investigated for items on Kasahara's Inventory for the Melancholic Type Personality (KIMTP). Subjects consisted of 44 patients diagnosed with an ICD-10 diagnostic criteria for research (DCR) depressive episode (F32) or recurrent depressive disorder (F33) unipolar affective disorder, and 44 observers. Thirty-one of the 44 patients were diagnosed with endogenous depression. The observers consisted of five parents, eight children, thre siblings, 25 spouses, and three partners. Concordance was tested with Student's paired t-test and one-way analysis of variance intraclass correlation coefficient (ANOVA ICC). There was no statistical difference between the mean total self-rating score and the mean total observer-rating score in the endogenous depression group, but the mean total self-rating score was significantly higher than the mean total observer-rating score in the non-endogenous depression group. When the self and observer tests in this non-endogenous depression group were compared on an item-by-item basis, most items tended to receive higher scores from the self-raters than from the observer-raters. The ICC had good concordance for mean total self- and observer-ratings in the endogenous depression group but not the non-endogenous depression group. On an item-by-item basis, concordance was high between self- and observer-raters for more items in the endogenous depression group than in the non-endogenous depression group. There were some important differences between self- and observer-ratings for certain items in the non-endogenous depression group. Patients in this group tended to give higher scores to items representing Typus melancholicus in relation to scores given by observers.

Diagnostic differences in social anhedonia: a longitudinal study of schizophrenia and major depressive disorder

This study examined the hypothesis that, in schizophrenia, elevated trait social anhedonia (SA) is a stable individual difference, whereas in depression, increased SA is a reflection of a current clinical state that will diminish with recovery. Differences in trait Negative Affect (NA) and Positive Affect (PA) were also examined. Individuals with schizophrenia (n = 55) and depression (n = 34) were evaluated at baseline during hospitalization and compared with nonpsychiatric control participants (n = 41). Participants were assessed again at a 1-year follow-up. At baseline, compared with control participants, individuals with schizophrenia and depression were both characterized by elevated SA, greater NA, and lower PA. In schizophrenic individuals, elevated SA remained stable over the follow-up. However, in recovered depressed patients, SA declined over the follow-up period. Group differences remained in NA and PA over the 1-year follow-up. These results support the view that elevated SA is enduring in schizophrenia but that elevated SA is transiently related to clinical status in depression.

Reliability of personality disorder symptoms and personality traits in substance-dependent inpatients

The authors compared the internal consistency, 1-year temporal stability, and self-informant agreement of ratings of personality trait (NEO Five-Factor Inventory; NEO-FFI; P. T. Costa & R. R. McCrae, 1992) and personality disorder symptom severity (Structured Clinical Interview for DSM-III-R Personality Disorders Questionnaire; SCID-II-Q; R. L. Spitzer, J. B. W. Williams, M. Gibbon, & M. First, 1990) in 131 substance-dependent inpatients. Internal consistency coefficients were acceptable to very good for most NEO-FFI and SCID-II-Q scales, and temporal stability correlations were significant for all measures. Agreement between patient and informant ratings was more modest. Substance abuse and depression symptom severity moderated the temporal stability and self-informant agreement of several personality trait and disorder ratings. The authors did not find that the five factors were more reliable than the Axis II symptoms. Issues related to the reliability of personality assessment in multiply diagnosed patients are discussed.

Peer assessment of personality traits and pathology in female college students

Assessment procedures for personality disorders (PDs) typically rely on self-reports, even though some people with PDs may be unable to view themselves realistically or are unwilling to report socially undesirable traits. Close associates may provide important information regarding the presence of PD traits. Peer nomination is a reliable and valid assessment procedure that can be adapted to the study of PDs for research purposes. This study focused on characteristic features that define narcissistic, dependent, and obsessive-compulsive PDs using information collected from both self and others in a nonclinical sample of women. It was designed to identify specific areas of agreement and discrepancy between self-report and peer assessment in the measurement of characteristic features of these disorders.