- Email: [email protected]
Depressive personality in the relatives of outpatients with dysthymic disorder and episodic major depressive disorder and normal controls
Journal of Affective Disorders 55 (1999) 19–27 www.elsevier.com / locate / jad
Research report
Depressive personality in the relatives of outpatients with dysthymic disorder and episodic major depressive disorder and normal controls Daniel N. Klein Departments of Psychology and Psychiatry and Behavioral Science, State University of New York at Stony Brook, Stony Brook, NY 11794 -2500, USA Received 20 March 1998; received in revised form 22 June 1998; accepted 29 June 1998
Abstract Background: This study examined whether there is a familial relationship between depressive personality and the mood disorders. Method: Rates of depressive personality were compared in 161 relatives of outpatients with dysthymic disorder (DD), 75 relatives of outpatients with non-chronic major depressive disorder (MDD), and 90 relatives of normal controls. All probands and relatives were evaluated using structured diagnostic interviews for Axis I disorders and depressive personality traits. Results: The relatives of patients with DD exhibited a significantly higher rate of depressive personality than the relatives of normal controls, while the relatives of patients with MDD fell in between, and did not differ from, the other two groups. These results held after controlling for a lifetime history of mood disorder in the relatives, and could not be explained by an increased rate of depressive personality in the DD probands. Limitations: The sample size was modest, comorbid non-mood Axis I and II disorders in the relatives were not controlled, and DSM-IV criteria for depressive personality disorder were not yet available at the time the study was undertaken. Conclusion: These findings are consistent with the view that depressive personality is part of a spectrum of mood disorders with a shared familial liability, but suggest that this link is strongest with chronic forms of depression such as DD and double depression. 1999 Elsevier Science B.V. All rights reserved. Keywords: Depressive personality; Dysthymia; Affective spectrum
In recent years, there has been growing interest in the classical clinical construct of depressive personality (Akiskal, 1989; Klein and Vocisano, in press; Phillips et al., 1990), which was recently included in the Appendix of DSM-IV as a condition needing further study. While no epidemiological data are available, depressive personality may be a fairly common condition, particularly in outpatient prac-
tice. For example, in a recent national survey of psychiatrists and psychologists, Westen (1997) found that 77% of clinicians reported treating one or more patients with depressive personality, a rate that was higher than for any Axis II condition other than borderline personality disorder. The longest-standing controversy in the literature on depressive personality concerns its relationship to
0165-0327 / 99 / $ – see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S0165-0327( 98 )00195-5
20
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
the mood disorders. Schneider (1958) believed that depressive personality represented the extreme end of a continuum of normal personality variation, and was independent from the mood disorders. In contrast, Kraepelin (1921); Kretschmer (1925) proposed that depressive personality represented a temperamental substrate of the mood disorders, predisposing to major affective episodes and aggregating in the families of patients with mood disorders. The leading contemporary advocate of the Kraepelinian position is Akiskal (1989, 1995, 1997), who views depressive personality as lying on a spectrum with the Axis I mood disorders. Akiskal has argued that depressive personality is particularly closely related to dysthymic disorder (DD), and that the two conditions represent alternative manifestations of a common constitutional diathesis. Family studies provide an important means of adjudicating between these competing positions. Two studies have reported an increased rate of mood disorders in the first-degree relatives of probands with depressive personality (Klein, 1990; Klein and Miller, 1993). Importantly, in one of these studies the increased rate of mood disorders in relatives was evident even in probands with depressive personality who had no history of diagnosable mood disorder themselves (Klein and Miller, 1993). However, both studies employed the family history method, relying on probands’ reports for both index diagnoses of depressive personality and mood disorders in relatives. There is also some evidence for an increased rate of depressive personality in the relatives of patients with mood disorders. Klein et al. (1988) reported that the adolescent and young adult offspring of parents with major depressive disorder (MDD) exhibited a significantly higher rate of depressive personality than the offspring of medical and normal controls. However, many of the offspring with depressive personality also had a history of DD and / or MDD, hence it is possible that diagnoses of depressive personality were confounded with current or partially remitted mood disorders. Finally, Weissman et al. (1984) reported a higher rate of depressive personality, as defined by the Research Diagnostic Criteria (Spitzer et al., 1978), in the first-degree relatives of outpatients with MDD than the relatives of normal controls. Importantly, this increase re-
mained even after imposing a diagnostic hierarchy that excluded relatives with major mood disorders. However, they did not report whether either of these comparisons was statistically significant. The present study presents additional data on the question of whether there is a familial relationship between depressive personality and MDD and DD. We compared the rates of depressive personality in the relatives of outpatients with DD (both with and without superimposed major depressive episodes), relatives of outpatients with non-chronic (episodic) MDD, and relatives of controls with no lifetime history of Axis I psychopathology. Two specific questions were addressed: (1) Is there an increased rate of depressive personality in the relatives of patients with Axis I depressive disorders compared to the relatives of normal controls?; and (2) do the rates of depressive personality in relatives differ between patients with chronic forms of depression, such as DD, and acute depressions, such as episodic MDD?
1. Method This paper is based on data from a family study of early-onset DD and episodic MDD. The method has been described in detail in previous publications (Klein et al., 1994, 1995; Riso et al., 1996); hence, it is only briefly summarized here.
1.1. Subjects Probands included 97 outpatients meeting DSMIII-R criteria for primary early-onset DD, 45 outpatients with DSM-III-R non-chronic MDD, and 45 subjects from the community with no lifetime history of Axis I disorder. Fifty-six (58%) of the DD patients were in a concurrent major depressive episode, and 75 (77%) had a lifetime history of MDD. The patients with episodic MDD were required to have an onset prior to age 35 in order to provide a closer match to the DD patients on demographic and clinical variables. All probands were between the ages of 18 and 60, English-speaking, and had knowledge of at least one first-degree relative. Patients were recruited by screening consecutive
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
admissions to the SUNY Stony Brook Outpatient Psychiatry Clinic and Psychological Center. A small number of patients were also drawn from the SUNY Stony Brook Counseling Center and a community mental health center. Normal control probands were recruited by randomly selecting telephone numbers from local telephone directories. Potential controls were screened by phone, and those who appeared eligible were given face-to-face diagnostic interviews. We attempted to group-match the normal probands to the patients by age and sex. All probands were administered the Structured Clinical Interview for DSM-III-R (SCID; Spitzer et al., 1990) and the Personality Disorder Examination (PDE, Loranger, 1988). The PDE was supplemented with a section assessing Akiskal’s (Akiskal, 1983) criteria for depressive personality, which are based on Schneider (1958). Depressive personality was considered present if subjects were rated as having five or more of the following groups of traits: (1) quiet, passive, and non-assertive; (2) gloomy, pessimistic, and incapable of fun; (3) self-critical, selfreproaching, and self-derogatory; (4) sceptical, hypercritical, and complaining; (5) conscientious and self-disciplining; (6) brooding and given to worry; and (7) preoccupied with inadequacy, failure, and negative events. These criteria overlap with, but are not identical to, those in DSM-IV, which was not available at the time the study was initiated. As discussed in Klein and Shih (in press), depressive personality traits were assessed and rated using the guidelines established for the PDE (Loranger, 1988). Information on lifetime psychopathology was obtained on all first-degree relatives over the age of 14 using direct and family history interviews. As there are no data on the reliability and validity of informant reports of depressive personality, the analyses were limited to relatives with direct interviews. Direct interviews were available for 177 first-degree relatives of probands with DD, 80 relatives of probands with episodic MDD, and 95 relatives of normal control probands. This represents 40% of all first-degree relatives, and 70% of all living relatives who spoke English, had known addresses in the United States, and whom the proband gave us permission to contact. Data on depressive personality were incomplete or missing for sixteen relatives of patients with DD, five relatives of patients with
21
episodic MDD, and five relatives of normal controls. The three groups of probands did not differ significantly on the proportion of relatives with direct interviews. In addition, there were few differences between relatives who were, and were not, available for direct interviews based on family history data (Norden et al., 1995). Relatives were administered the SCID, PDE and depressive personality supplement. Interviews were conducted blind to the clinical status of the probands. Interviewers included a licensed doctoral-level clinical psychologist, several doctoral students in clinical psychology, and an experienced master’s level psychiatric social worker. All interviewers received formal training in the use of the SCID and PDE by the developers of these instruments, were closely supervised throughout the study, and participated in weekly meetings to discuss difficult ratings and review taped interviews. The interrater reliability of our Axis I and II diagnoses was generally good–excellent, as reported elsewhere (Klein et al., 1994). In order to examine the interrater reliability of diagnoses of depressive personality, videotapes of fifteen randomly selected PDE interviews were independently rated by another diagnostician. Diagnostic concordance, as indexed by k, was 0.70. The intraclass correlation (Shrout and Fleiss, 1979, Case 1) for number of depressive personality traits was 0.81. All probands and relatives gave written informed consent.
1.2. Data analysis Groups of probands and relatives were compared on demographic and clinical characteristics using x 2 and Fisher’s Exact tests for categorical variables, and one-way analysis of variance and t-tests for continuous variables. Pairwise comparisons were conducted with x 2 and Fisher’s Exact tests for categorical variables and Tukey’s Honestly Significant Difference procedure for continuous variables. Rates of depressive personality were compared between groups of relatives using multiple logistic regression analysis, using the relatives’ age, sex, and, in some cases, lifetime history of mood disorder, as covariates. Number of depressive personality traits in relatives were compared across groups using hierarchical multiple linear regression analysis, with rela-
22
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
tives’ age, sex, and, in some cases, lifetime history of mood disorder, as covariates. The rate of depressive personality (15.2%) and number of depressive personality traits (M 5 2.58; S.D. 5 1.71) in the 125 relatives of DD probands with a lifetime history of MDD did not differ significantly from the rate of depressive personality (11.1%) and number of depressive personality traits (M 5 2.47; S.D. 5 1.71) in the 36 relatives of DD probands without a history of MDD ( p . 0.50 for both comparisons), hence these groups were combined for all analyses.
2. Results
2.1. Descriptive characteristics Descriptive characteristics of the probands are presented in Table 1. The three groups of probands
did not differ significantly on age, sex, race and education. However, the groups did differ on socioeconomic status (Hollingshead, 1975) and marital status, with the normal control probands being significantly more likely to be married, and having a significantly higher social class distribution than the two groups of patients. Patients with episodic MDD had significantly higher scores on the 24-item Hamilton Rating Scale for Depression (Miller et al., 1985) than patients with early-onset DD, and both groups of patients had significantly higher levels of depression and poorer global functioning than the normal controls. The majority of patients in both groups had a history of recurrent major depressive episodes, and did not differ in that respect. Patients with DD had a significantly higher lifetime rate of substance abuse and / or dependence, and significantly higher rates of any DSM-III-R personality disorder and depressive personality than patients with episodic MDD, and both groups of patients had significantly higher rates
Table 1 Descriptive characteristics of probands Variable
Age M (S.D.) Sex (% female) Race (% white) Education M (S.D.) Social class (%) I II III IV V Marital status % Single % Married % Separated / divorced % Widowed HAM-D M (S.D.) GAF M (S.D.) Recurrent MDD (%) Lifetime anxiety disorder (%) Lifetime substance Abuse / dependence (%) Personality disorder (%) Depressive personality (%)
Proband group DD
EMDD
NC
Test statistic a
32.2 (9.8) 75.3 90.7 13.2 (2.2)
31.6 (9.2) 66.7 86.7 14.0 (2.2)
33.4 (10.1) 82.2 93.3 13.6 (1.9)
NS NS NS NS x 2 (8) 5 23.7*
9.1 28.4 25.0 17.0 20.5
17.1 29.3 24.4 24.4 4.9
9.3 51.2 32.6 7.0 0.0
48.5 29.9 19.6 2.1 26.3 (11.3) 56.4 (9.6) 62.9 43.3
42.2 31.1 26.7 0.0 30.4 (8.5) 56.9 (12.6) 62.2 28.9
26.7 73.3 0.0 0.0 4.6 (3.9) 87.5 (3.8) – –
49.5 59.8 73.2
24.4 17.8 22.2
x 2 (6) 5 31.7**
– 0.0 2.2
F(2,146) 5 97.09** F(2,183) 5 178.9** NS NS
x 2 (1) 5 7.94* x 2 (1) 5 20.2** x 2 (2) 5 74.14**
DD, Dysthymic Disorder (N 5 97); EMDD, Episodic Major Depressive Disorder (N 5 45); NC, Normal Controls (N 5 45); HAM-D, 24-item Hamilton Rating Scale for Depression; GAF, Global Assessment of Functioning Scale; MDD, Major Depressive Disorder. a Only significant findings are reported. * p , 0.01; ** p , 0.001.
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
of any personality disorder and depressive personality than normal probands. The interviewed relatives of the three groups of probands did not differ significantly on sex, age, marital status, or relationship to the proband. Overall, 62.6% of the relatives were female; their mean age was 38.4 (S.D. 5 16.4); 56.4% were married, 10.3% were separated or divorced, 4.4% were widowed, and 29.0% had never married; and 33.7% were parents, 51.2% were siblings, and 15.0% were offspring of the probands. The three groups of relatives differed significantly on education, F 5 4.83, df 5 2,310, p , 0.01. Pairwise comparisons indicated that the relatives of patients with episodic MDD had a significantly greater number of years of formal education (M 5 13.8; S.D. 5 2.4) than the relatives of normal controls (M 5 12.6; S.D. 5 2.2). The relatives of patients with DD (M 5 13.1; S.D. 5 2.6) did not differ significantly from either of these groups. Controlling for proband SES and marital status and relative education did not change the results. Hence, these variables were not included as covariates in the analyses reported below.
2.2. Depressive personality in relatives The rates of depressive personality and the mean number of depressive personality traits in the relatives of patients with early-onset DD and episodic MDD and normal controls are presented in Table 2. After controlling for the relatives’ age and sex, the rate of depressive personality was significantly higher in the relatives of patients with DD than normal controls, Odds Ratio (OR) 5 3.47 (CI 5 1.16– 10.42), p 5 0.026. However, there were no significant differences in the rates of depressive personality
23
between the relatives of patients with DD and episodic MDD, OR 5 1.81, p 5 0.22, or between the relatives of patients with episodic MDD and normal controls, OR 5 1.92, p 5 0.33. After controlling for the effects of age and sex, the groups of relatives also differed significantly on number of depressive personality traits, F change 5 6.37, df 5 4,321, p 5 0.002. The relatives of patients with DD exhibited a significantly higher number of depressive personality traits than the relatives of patients with episodic MDD, t 5 2.79, p , 0.006, and the relatives of normal controls, t 5 3.04, p , 0.003. The contrast between the relatives of patients with episodic MDD and normal controls was not significant. It is conceivable that these differences are due to the overlap between depressive personality and DD, as we have previously reported that there is a higher rate of DD in the relatives of probands with DD than the relatives of probands with episodic MDD and normal controls (Klein et al., 1995). Indeed, in this sample, 38% of the relatives with DD also met criteria for depressive personality disorder, and 36% of relatives with depressive personality met criteria for DD (k 5 0.30). In addition, it is also possible that some cases of depressive personality represent complications or residual effects of prior episodes of mood disorder, or that current affective disorder might distort subjects’ reports of depressive personality traits. Hence, we repeated these analyses, adding lifetime history of any form of Axis I mood disorder (including bipolar disorder, bipolar disorder not otherwise specified, MDD and DD) on the SCID as a third covariate. The rate of depressive personality continued to be significantly higher among the relatives of patients with early-onset DD than the relatives of normal controls even after controlling for
Table 2 Rates of depressive personality and mean number of depressive personality traits in the relatives of outpatients with dysthymia and episodic major depression and normal controls Proband group
Depressive personality diagnoses a % (N)
Depressive personality traits b M (S.D.)
Dysthymia Episodic major depression Normal controls
14.3% (23) 8.0% (6) 4.4% (4)
2.55 (1.71) 1.87 (1.46) 1.91 (1.47)
Dysthymia (DD) N 5 161; Episodic Major Depression (MDD) N 5 75; Normal Control (NOR) N 5 90. a DD vs. NOR OR 5 3.47 (CI 5 1.16–10.42), p 5 0.026, adjusting for relative age and sex. b DD vs. MDD t 5 2.79, p , 0.006; DD vs. NOR t 5 3.04, p , 0.003, adjusting for relative age and sex.
24
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
age, sex, and lifetime mood disorder, OR 5 3.01 (CI 5 0.98–9.22), p 5 0.05. In addition, after controlling for age, sex, and lifetime mood disorder, the groups of relatives differed significantly on number of depressive personality traits, F change 5 4.01, df 5 5,315, p , 0.02. Again, the number of depressive personality traits was significantly higher in the relatives of patients with DD than patients with episodic MDD, t 5 2.03, p 5 0.04, and normal controls, t 5 2.56, p 5 0.01. Finally, we examined whether the increased rate of depressive personality in the relatives of probands with DD could be due to the higher rate of depressive personality in the DD probands themselves. Of the relatives of the DD probands with depressive personality, 14.7% (17 / 116) were diagnosed as having depressive personality. In contrast, 13.3% (6 / 45) of the relatives of DD probands without depressive personality were diagnosed as having depressive personality. The difference between these rates did not approach significance. Similarly, the number of depressive personality traits in the relatives of DD probands with (M 5 2.56; S.D. 5 1.71), and without (M 5 2.56; S.D. 5 1.73), depressive personality did not differ. We also compared the rate of depressive personality in the relatives of DD probands without depressive personality to the relatives of the normal control probands. For these analyses, we excluded the one relative of the one normal control proband with depressive personality, although the results were the same when this subject (who did not have depressive personality) was included. After controlling for age and sex, there was a trend for a higher rate of depressive personality in the relatives of DD probands without depressive personality than the relatives of normal controls, OR 5 3.27, p 5 0.08. In addition, after controlling for age and sex, the relatives of DD probands without depressive personality exhibited a significantly higher number of depressive personality traits than the relatives of normal controls, F change 5 4.63, df 5 3,130, p 5 0.03.
3. Discussion The present findings support Akiskal’s (Akiskal, 1995, 1997) hypothesis that depressive personality
stems from the same processes that underlie DD, and can be conceptualized as a milder form, or alternative expression, of the same condition. However, the data also suggest that the view that depressive personality lies on a spectrum with the major mood disorders may require qualification, as the familial relationship between depressive personality and MDD was limited to those forms of MDD that were superimposed on a pre-existing DD, and did not appear to be linked to acute, remitting forms of MDD. The major finding in the study was that the firstdegree relatives of patients with early-onset DD (most of whom had a history of superimposed major depressive episodes) exhibited a significantly increased rate of depressive personality, and a significantly greater number of depressive personality traits, than the relatives of normal controls. Importantly, these findings held even after controlling for a lifetime history of dysthymia and other mood disorders in the relatives. In addition, the differing rates of depressive personality between the relatives of patients with DD and the relatives of normal controls were not due to the high rate of depressive personality in the DD probands. DD probands with and without depressive personality had similar rates of depressive personality disorder and identical mean numbers of depressive personality traits in their relatives. Moreover, the relatives of DD probands without depressive personality exhibited a significantly higher number of depressive personality traits, and a trend for a higher rate of depressive personality disorder, than the relatives of normal controls. Thus, depressive personality appears to aggregate in the relatives of probands with DD regardless of whether or not the probands themselves have depressive personality, indicating that depressive personality and DD derive from a shared familial liability (Klein and Riso, 1993). In a previous study using the family history method, I found that patients with DD reported a significantly higher rate of depressive disorders in their first-degree relatives than patients with depressive personality (Klein, 1990). Together with the present findings, these data suggest that while depressive personality and DD share a common familial liability, the degree of the liability appears to be greater in DD.
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
Approximately 4% of the relatives of normal control probands met criteria for depressive personality disorder. This figure is very close to Placidi et al.’s (Placidi et al., 1998) report that the prevalence of depressive temperament in a community sample of over 1000 adolescents and young adults was 3.6%. Taken together, these findings suggest that depressive personality may be fairly common in the community. However, it appears that many of these individuals do not seek treatment until they have developed an Axis I mood disorder (Klein and Miller, 1993). One of the major controversies concerning depressive personality involves whether it is distinct enough from DD to warrant a separate category (Akiskal, 1989; Klein and Vocisano, in press; Phillips et al., 1990). In five studies of the overlap between depressive personality and DD, a median 58% (range 5 20–81%) of individuals meeting criteria for depressive personality did not meet criteria for DD (Hirschfeld and Holzer, 1994; Klein, 1990; Klein and Miller, 1993; Klein and Shih, in press; Phillips et al., 1995). Thus, consistent with the present findings, it appears that the category of depressive personality captures a large number of persons with close phenomenological and familial links to DD who do not qualify for a DD diagnosis. It may be possible to broaden the criteria for DD to include these individuals. One of the major reasons that persons with depressive personality fail to meet criteria for DD is that they do not experience (or at least report) persistent depressed mood (Klein and Vocisano, in press). If the criteria for DD were modified to allow lack of interest or pleasure to substitute for depressed mood, as in the current criteria for MDD, these individuals might be captured. On the other hand, this would also increase the heterogeneity of the DD category, as depressive personality appears to be a milder condition than DD both in terms of symptomatology and familial aggregation (Klein, 1990). Thus, the increased parsimony of a lumping approach to classification must be weighed against the greater homogeneity and preservation of information associated with a splitting approach. The results of the current study did not support a familial relationship between depressive personality and episodic MDD. The rate of depressive personali-
25
ty in the relatives of probands with episodic MDD was midway between the rates in the relatives of the DD and normal probands, and did not differ significantly from either group. In addition, the relatives of patients with episodic MDD exhibited significantly fewer depressive personality traits than the relatives of DD patients, but did not differ from the relatives of normal controls. These findings should be viewed cautiously. The number of relatives of patients with episodic MDD was modest, hence replication with a larger sample is necessary before drawing strong conclusions. In addition, the episodic MDD group may be heterogeneous, and it is possible that a subgroup of these patients may share a familial link with depressive personality. Given the small number of subjects, we could only explore this in a very preliminary manner. There was no evidence that an early age of onset or recurrent subtype in the probands with episodic MDD was associated with depressive personality in their relatives. However, there was a non-significant tendency for a higher rate of depressive personality in the relatives of episodic MDD probands who reported being depressed for at least half the time during the past two years (12.2%) compared to the relatives of probands who were depressed for less than half the time during the past two years (2.9%), p 5 0.14. This raises the possibility that depressive personality is related to a broad spectrum of chronic, subchronic, and intermittent forms of depression, rather than to DD and double depression specifically. The present findings are consistent with previous studies in indicating that there is an elevated rate of depressive personality in the relatives of patients with mood disorders (Klein et al., 1988; Weissman et al., 1984), and are also consistent with previous findings of elevated rates of mood disorders in the relatives of subjects with depressive personality (Klein, 1990; Klein and Miller, 1993). However, none of these studies attempted to distinguish between chronic and episodic forms of depression. Thus, it is unclear whether the findings were limited to particular subtypes, such as double depression. The present study addressed only the familial aggregation aspect of the affective spectrum hypothesis (Akiskal, 1989; Kraepelin, 1921). Twin and adoption studies and developmental studies of infants and children are needed to address the question of
26
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
whether depressive personality is best conceptualized as a form of temperament, and longitudinal studies are necessary to determine whether depressive personality predisposes to the development of major mood disorders. This study also has implications for studying the role of personality in the mood disorders. In both the current study and our previous offspring study (Klein et al., 1988), semi-structured interviews assessing depressive personality traits successfully distinguished the relatives of depressed and control probands, whereas comprehensive batteries of self-report personality inventories did not (Ouimette et al., 1992, 1996). This suggests that the construct of depressive personality may better capture the personality traits that are associated with familial risk for mood disorders than constructs from the normal personality literature, and / or that semi-structured interviews are more sensitive than self-report questionnaires in assessing these traits. Kendler et al. (1996) recently reported similar findings for schizotypal personality disorder. In weighing the results of this study, several limitations should be considered. First, the sample was modest in size, and limited to several specific forms of depression. It will be important to conduct further studies using larger samples which are representative of the full range of depressive conditions, including late-onset DD and chronic MDD. Second, multiple relatives from the same family were included in the analyses. While this is common practice in family studies, it does violate the assumption of independent observations and could inflate significance levels. Third, comorbidity with non-mood Axis I and II disorders in relatives was not taken into account. Thus, we cannot rule out the possibility that the elevated rate of depressive personality in the relatives of the DD probands is due to comorbid anxiety, substance use, or other personality disorders. Finally, depressive personality was assessed using Akiskal’s (Akiskal, 1983) criteria. These criteria exhibit moderate overlap with the DSM-IV criteria, but are somewhat more restrictive (Hirschfeld and Holzer, 1994). Future studies should employ multiple criteria sets, including DSM-IV, and Akiskal’s (Akiskal, 1983) and Gunderson et al.’s (Gunderson et al., 1994) criteria.
Acknowledgements The author acknowledges the assistance of Rochelle L. Anderson, Thomas A. Aronson, Humberto Lizardi, Shauna K. Donaldson, Tova Ferro, Rollin M. Gallagher, Robert M.A. Hirschfeld, Karen L. Kasch, Helen S. Kelly, Laura M. Klein, Armand W. Loranger, Kimberly A. Norden, Paige Crosby Ouimette, Carolyn M. Pepper, Charles L. Rich, Lawrence P. Riso, and Joseph E. Schwartz in the design of the study, data collection and analysis, and the recruitment of subjects. The author was supported by NIMH Research Grant No. RO1 MH45757.
References Akiskal, H.S., 1983. Dysthymic disorder: Psychopathology of proposed chronic depressive subtypes. Am. J. Psychiatry 140, 11–20. Akiskal, H.S., 1989. Validating affective personality types. In: Robins, L., Barrett, J. (Eds.), The Validity of Psychiatric Diagnosis, Raven Press, New York, pp. 217–227. Akiskal, H.S., 1995. Toward a temperament based approach to depression: Implications for neurobiologic research. Adv. Biochem. Psychopharmacol. 49, 99–112. Akiskal, H.S., 1997. Overview of chronic depressions and their clinical management. In: Akiskal, H.S., Cassano, G.B. (Eds.), Dysthymia and the Spectrum of Chronic Depressions, Guilford Press, New York, pp. 1–34. Gunderson, J.G., Phillips, K.A., Triebwasser, J.T., Hirschfeld, R.M.A., 1994. The diagnostic interview for depressive personality. Am. J. Psychiatry 151, 1300–1304. Hirschfeld, R.M.A., Holzer III, C.E., 1994. Depressive personality disorder: Clinical implications. J. Clin. Psychiatry 55, 10–17. Hollingshead, A.B., 1975. Four Factor Index of Social Position, Yale University, Department of Sociology, New Haven, CT. Kendler, K.S., Thacker, L., Walsh, E., 1996. Self-report measures of schizotypy as indices of familial vulnerability to schizophrenia. Schiz. Bull. 22, 511–520. Klein, D.N., 1990. Depressive personality: Reliability, validity, and relation to dysthymia. J. Abnorm. Psychol. 99, 412–421. Klein, D.N., Clark, D.C., Dansky, L., Margolis, E.T., 1988. Dysthymia in the offspring of parents with primary unipolar affective disorder. J. Abnorm. Psychol. 97, 265–274. Klein, D.N., Miller, G.A., 1993. Depressive personality in nonclinical subjects. Am. J. Psychiatry 150, 1718–1724. Klein, D.N., Ouimette, P.C., Kelly, H.S., Ferro, T., Riso, L.P., 1994. Test–retest reliability of team consensus best-estimate diagnoses of Axis I and II disorders in a family study. Am. J. Psychiatry 151, 1043–1047.
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27 Klein, D.N., Riso, L.P., 1993. Psychiatric disorders: Problems of boundaries and comorbidity. In: Costello, C.G. (Ed.), Basic Issues in Psychopathology, Guilford Press, New York, pp. 19–66. Klein, D.N., Riso, L.P., Donaldson, S.K., Schwartz, J.E., Anderson, R.L., Ouimette, P.C., Lizardi, H., Aronson, T.A., 1995. Family study of early-onset dysthymia: Mood and personality disorders in relatives of outpatients with dysthymia and episodic major depression and normal controls. Arch. Gen. Psychiatry 52, 487–496. Klein, D.N., Shih, J.H., 1998. 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. 107, 319–327. Klein, D.N., Vocisano, C. Depressive and self-defeating (masochistic) personality disorders. In: Millon, T., Blaney, P.H., Davis, R.D. (Eds.), Oxford Textbook of Psychopathology, Oxford University Press, New York, in press. Kraepelin, E., 1921. Manic Depressive Insanity and Paranoia, E&S Livingstone, Edinburgh. Kretschmer, E., 1925. Physique and Character, Harcourt Brace, New York. Loranger, A.W., 1988. Personality Disorder Examination (PDE) Manual, DV Communications, Yonkers, New York. Miller, I.W., Bishop, S., Norman, W.H., Maddever, H., 1985. The modified Hamilton rating scale for depression: Reliability and validity. Psychiatry Res. 14, 131–142. Norden, K.A., Klein, D.N., Ferro, T., Kasch, K., 1995. Who participates in a family study? Compr. Psychiatry 36, 199–206. Ouimette, P.C., Klein, D.N., Clark, D.C., Margolis, E.T., 1992. Personality traits in offspring of patients with unipolar affective disorder: An exploratory study. J. Personality Disord. 6, 91–98. Ouimette, P.C., Klein, D.N., Pepper, C.M., 1996. Personality traits in the first-degree relatives of outpatients with depressive disorders. J. Affective Disord. 39, 43–53.
27
Phillips, K.A., Gunderson, J.G., Hirschfeld, R.M.A., Smith, L.E., 1990. A review of the depressive personality. Am. J. Psychiatry 147, 830–837. Phillips, K.A., Hirschfeld, R.M.A., Shea, M.T., Gunderson, J.G., 1995. Depressive personality disorder: Perspectives for DSMIV. In: Livesley, W.S. (Ed.), DSM-IV Personality Disorders, Guilford Press, New York, pp. 287–302. Placidi, G.F., Signoretta, S., Liguori, A., Gervasi, R., Maremmani, I., Akiskal, H.S., 1998. The semi-structured affective temperament interview (TEMPS-I): Reliability and psychometric properties in 1010 14–26 year-old students. J. Affective Disord. 47, 1–10. Riso, L.P., Klein, D.N., Ferro, T., Kasch, K.L., Pepper, C.M., Schwartz, J.E., Aronson, T.A., 1996. Understanding the comorbidity between early-onset dysthymia and cluster B personality disorders: A family study. Am. J. Psychiatry 153, 900–906. Schneider, K., 1958. Psychopathic Personalities, Cassell, London. Shrout, P.E., Fleiss, J.L., 1979. Intraclass correlations: Uses in assessing rater reliability. Psychol. Bull. 86, 420–428. Spitzer, R.L., Endicott, J., Robins, E., 1978. Research diagnostic criteria: Rationale and reliability. Arch. Gen. Psychiatry 35, 773–782. Spitzer, R.L., Williams, J.B.W., Gibbon, M., First, M.B., 1990. Structured Clinical Interview for DSM-III-R, American Psychiatric Press, Washington, DC. Weissman, M.M., Gershon, E.S., Kidd, K.K., Prusoff, B.A., Leckman, J.F., Dibble, E., Hamovit, J., Thompson, W.D., Pauls, D.L., Guroff, J.J., 1984. Psychiatric disorders in the relatives of probands with affective disorders: The Yale University–National Institute of Mental Health collaborative study. Arch. Gen. Psychiatry 41, 13–21. Westen, D., 1997. Divergences between clinical and research methods for assessing personality disorders: Implications for research and the evolution of Axis II. Am. J. Psychiatry 154, 895–903.
Research report
Depressive personality in the relatives of outpatients with dysthymic disorder and episodic major depressive disorder and normal controls Daniel N. Klein Departments of Psychology and Psychiatry and Behavioral Science, State University of New York at Stony Brook, Stony Brook, NY 11794 -2500, USA Received 20 March 1998; received in revised form 22 June 1998; accepted 29 June 1998
Abstract Background: This study examined whether there is a familial relationship between depressive personality and the mood disorders. Method: Rates of depressive personality were compared in 161 relatives of outpatients with dysthymic disorder (DD), 75 relatives of outpatients with non-chronic major depressive disorder (MDD), and 90 relatives of normal controls. All probands and relatives were evaluated using structured diagnostic interviews for Axis I disorders and depressive personality traits. Results: The relatives of patients with DD exhibited a significantly higher rate of depressive personality than the relatives of normal controls, while the relatives of patients with MDD fell in between, and did not differ from, the other two groups. These results held after controlling for a lifetime history of mood disorder in the relatives, and could not be explained by an increased rate of depressive personality in the DD probands. Limitations: The sample size was modest, comorbid non-mood Axis I and II disorders in the relatives were not controlled, and DSM-IV criteria for depressive personality disorder were not yet available at the time the study was undertaken. Conclusion: These findings are consistent with the view that depressive personality is part of a spectrum of mood disorders with a shared familial liability, but suggest that this link is strongest with chronic forms of depression such as DD and double depression. 1999 Elsevier Science B.V. All rights reserved. Keywords: Depressive personality; Dysthymia; Affective spectrum
In recent years, there has been growing interest in the classical clinical construct of depressive personality (Akiskal, 1989; Klein and Vocisano, in press; Phillips et al., 1990), which was recently included in the Appendix of DSM-IV as a condition needing further study. While no epidemiological data are available, depressive personality may be a fairly common condition, particularly in outpatient prac-
tice. For example, in a recent national survey of psychiatrists and psychologists, Westen (1997) found that 77% of clinicians reported treating one or more patients with depressive personality, a rate that was higher than for any Axis II condition other than borderline personality disorder. The longest-standing controversy in the literature on depressive personality concerns its relationship to
0165-0327 / 99 / $ – see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S0165-0327( 98 )00195-5
20
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
the mood disorders. Schneider (1958) believed that depressive personality represented the extreme end of a continuum of normal personality variation, and was independent from the mood disorders. In contrast, Kraepelin (1921); Kretschmer (1925) proposed that depressive personality represented a temperamental substrate of the mood disorders, predisposing to major affective episodes and aggregating in the families of patients with mood disorders. The leading contemporary advocate of the Kraepelinian position is Akiskal (1989, 1995, 1997), who views depressive personality as lying on a spectrum with the Axis I mood disorders. Akiskal has argued that depressive personality is particularly closely related to dysthymic disorder (DD), and that the two conditions represent alternative manifestations of a common constitutional diathesis. Family studies provide an important means of adjudicating between these competing positions. Two studies have reported an increased rate of mood disorders in the first-degree relatives of probands with depressive personality (Klein, 1990; Klein and Miller, 1993). Importantly, in one of these studies the increased rate of mood disorders in relatives was evident even in probands with depressive personality who had no history of diagnosable mood disorder themselves (Klein and Miller, 1993). However, both studies employed the family history method, relying on probands’ reports for both index diagnoses of depressive personality and mood disorders in relatives. There is also some evidence for an increased rate of depressive personality in the relatives of patients with mood disorders. Klein et al. (1988) reported that the adolescent and young adult offspring of parents with major depressive disorder (MDD) exhibited a significantly higher rate of depressive personality than the offspring of medical and normal controls. However, many of the offspring with depressive personality also had a history of DD and / or MDD, hence it is possible that diagnoses of depressive personality were confounded with current or partially remitted mood disorders. Finally, Weissman et al. (1984) reported a higher rate of depressive personality, as defined by the Research Diagnostic Criteria (Spitzer et al., 1978), in the first-degree relatives of outpatients with MDD than the relatives of normal controls. Importantly, this increase re-
mained even after imposing a diagnostic hierarchy that excluded relatives with major mood disorders. However, they did not report whether either of these comparisons was statistically significant. The present study presents additional data on the question of whether there is a familial relationship between depressive personality and MDD and DD. We compared the rates of depressive personality in the relatives of outpatients with DD (both with and without superimposed major depressive episodes), relatives of outpatients with non-chronic (episodic) MDD, and relatives of controls with no lifetime history of Axis I psychopathology. Two specific questions were addressed: (1) Is there an increased rate of depressive personality in the relatives of patients with Axis I depressive disorders compared to the relatives of normal controls?; and (2) do the rates of depressive personality in relatives differ between patients with chronic forms of depression, such as DD, and acute depressions, such as episodic MDD?
1. Method This paper is based on data from a family study of early-onset DD and episodic MDD. The method has been described in detail in previous publications (Klein et al., 1994, 1995; Riso et al., 1996); hence, it is only briefly summarized here.
1.1. Subjects Probands included 97 outpatients meeting DSMIII-R criteria for primary early-onset DD, 45 outpatients with DSM-III-R non-chronic MDD, and 45 subjects from the community with no lifetime history of Axis I disorder. Fifty-six (58%) of the DD patients were in a concurrent major depressive episode, and 75 (77%) had a lifetime history of MDD. The patients with episodic MDD were required to have an onset prior to age 35 in order to provide a closer match to the DD patients on demographic and clinical variables. All probands were between the ages of 18 and 60, English-speaking, and had knowledge of at least one first-degree relative. Patients were recruited by screening consecutive
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
admissions to the SUNY Stony Brook Outpatient Psychiatry Clinic and Psychological Center. A small number of patients were also drawn from the SUNY Stony Brook Counseling Center and a community mental health center. Normal control probands were recruited by randomly selecting telephone numbers from local telephone directories. Potential controls were screened by phone, and those who appeared eligible were given face-to-face diagnostic interviews. We attempted to group-match the normal probands to the patients by age and sex. All probands were administered the Structured Clinical Interview for DSM-III-R (SCID; Spitzer et al., 1990) and the Personality Disorder Examination (PDE, Loranger, 1988). The PDE was supplemented with a section assessing Akiskal’s (Akiskal, 1983) criteria for depressive personality, which are based on Schneider (1958). Depressive personality was considered present if subjects were rated as having five or more of the following groups of traits: (1) quiet, passive, and non-assertive; (2) gloomy, pessimistic, and incapable of fun; (3) self-critical, selfreproaching, and self-derogatory; (4) sceptical, hypercritical, and complaining; (5) conscientious and self-disciplining; (6) brooding and given to worry; and (7) preoccupied with inadequacy, failure, and negative events. These criteria overlap with, but are not identical to, those in DSM-IV, which was not available at the time the study was initiated. As discussed in Klein and Shih (in press), depressive personality traits were assessed and rated using the guidelines established for the PDE (Loranger, 1988). Information on lifetime psychopathology was obtained on all first-degree relatives over the age of 14 using direct and family history interviews. As there are no data on the reliability and validity of informant reports of depressive personality, the analyses were limited to relatives with direct interviews. Direct interviews were available for 177 first-degree relatives of probands with DD, 80 relatives of probands with episodic MDD, and 95 relatives of normal control probands. This represents 40% of all first-degree relatives, and 70% of all living relatives who spoke English, had known addresses in the United States, and whom the proband gave us permission to contact. Data on depressive personality were incomplete or missing for sixteen relatives of patients with DD, five relatives of patients with
21
episodic MDD, and five relatives of normal controls. The three groups of probands did not differ significantly on the proportion of relatives with direct interviews. In addition, there were few differences between relatives who were, and were not, available for direct interviews based on family history data (Norden et al., 1995). Relatives were administered the SCID, PDE and depressive personality supplement. Interviews were conducted blind to the clinical status of the probands. Interviewers included a licensed doctoral-level clinical psychologist, several doctoral students in clinical psychology, and an experienced master’s level psychiatric social worker. All interviewers received formal training in the use of the SCID and PDE by the developers of these instruments, were closely supervised throughout the study, and participated in weekly meetings to discuss difficult ratings and review taped interviews. The interrater reliability of our Axis I and II diagnoses was generally good–excellent, as reported elsewhere (Klein et al., 1994). In order to examine the interrater reliability of diagnoses of depressive personality, videotapes of fifteen randomly selected PDE interviews were independently rated by another diagnostician. Diagnostic concordance, as indexed by k, was 0.70. The intraclass correlation (Shrout and Fleiss, 1979, Case 1) for number of depressive personality traits was 0.81. All probands and relatives gave written informed consent.
1.2. Data analysis Groups of probands and relatives were compared on demographic and clinical characteristics using x 2 and Fisher’s Exact tests for categorical variables, and one-way analysis of variance and t-tests for continuous variables. Pairwise comparisons were conducted with x 2 and Fisher’s Exact tests for categorical variables and Tukey’s Honestly Significant Difference procedure for continuous variables. Rates of depressive personality were compared between groups of relatives using multiple logistic regression analysis, using the relatives’ age, sex, and, in some cases, lifetime history of mood disorder, as covariates. Number of depressive personality traits in relatives were compared across groups using hierarchical multiple linear regression analysis, with rela-
22
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
tives’ age, sex, and, in some cases, lifetime history of mood disorder, as covariates. The rate of depressive personality (15.2%) and number of depressive personality traits (M 5 2.58; S.D. 5 1.71) in the 125 relatives of DD probands with a lifetime history of MDD did not differ significantly from the rate of depressive personality (11.1%) and number of depressive personality traits (M 5 2.47; S.D. 5 1.71) in the 36 relatives of DD probands without a history of MDD ( p . 0.50 for both comparisons), hence these groups were combined for all analyses.
2. Results
2.1. Descriptive characteristics Descriptive characteristics of the probands are presented in Table 1. The three groups of probands
did not differ significantly on age, sex, race and education. However, the groups did differ on socioeconomic status (Hollingshead, 1975) and marital status, with the normal control probands being significantly more likely to be married, and having a significantly higher social class distribution than the two groups of patients. Patients with episodic MDD had significantly higher scores on the 24-item Hamilton Rating Scale for Depression (Miller et al., 1985) than patients with early-onset DD, and both groups of patients had significantly higher levels of depression and poorer global functioning than the normal controls. The majority of patients in both groups had a history of recurrent major depressive episodes, and did not differ in that respect. Patients with DD had a significantly higher lifetime rate of substance abuse and / or dependence, and significantly higher rates of any DSM-III-R personality disorder and depressive personality than patients with episodic MDD, and both groups of patients had significantly higher rates
Table 1 Descriptive characteristics of probands Variable
Age M (S.D.) Sex (% female) Race (% white) Education M (S.D.) Social class (%) I II III IV V Marital status % Single % Married % Separated / divorced % Widowed HAM-D M (S.D.) GAF M (S.D.) Recurrent MDD (%) Lifetime anxiety disorder (%) Lifetime substance Abuse / dependence (%) Personality disorder (%) Depressive personality (%)
Proband group DD
EMDD
NC
Test statistic a
32.2 (9.8) 75.3 90.7 13.2 (2.2)
31.6 (9.2) 66.7 86.7 14.0 (2.2)
33.4 (10.1) 82.2 93.3 13.6 (1.9)
NS NS NS NS x 2 (8) 5 23.7*
9.1 28.4 25.0 17.0 20.5
17.1 29.3 24.4 24.4 4.9
9.3 51.2 32.6 7.0 0.0
48.5 29.9 19.6 2.1 26.3 (11.3) 56.4 (9.6) 62.9 43.3
42.2 31.1 26.7 0.0 30.4 (8.5) 56.9 (12.6) 62.2 28.9
26.7 73.3 0.0 0.0 4.6 (3.9) 87.5 (3.8) – –
49.5 59.8 73.2
24.4 17.8 22.2
x 2 (6) 5 31.7**
– 0.0 2.2
F(2,146) 5 97.09** F(2,183) 5 178.9** NS NS
x 2 (1) 5 7.94* x 2 (1) 5 20.2** x 2 (2) 5 74.14**
DD, Dysthymic Disorder (N 5 97); EMDD, Episodic Major Depressive Disorder (N 5 45); NC, Normal Controls (N 5 45); HAM-D, 24-item Hamilton Rating Scale for Depression; GAF, Global Assessment of Functioning Scale; MDD, Major Depressive Disorder. a Only significant findings are reported. * p , 0.01; ** p , 0.001.
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
of any personality disorder and depressive personality than normal probands. The interviewed relatives of the three groups of probands did not differ significantly on sex, age, marital status, or relationship to the proband. Overall, 62.6% of the relatives were female; their mean age was 38.4 (S.D. 5 16.4); 56.4% were married, 10.3% were separated or divorced, 4.4% were widowed, and 29.0% had never married; and 33.7% were parents, 51.2% were siblings, and 15.0% were offspring of the probands. The three groups of relatives differed significantly on education, F 5 4.83, df 5 2,310, p , 0.01. Pairwise comparisons indicated that the relatives of patients with episodic MDD had a significantly greater number of years of formal education (M 5 13.8; S.D. 5 2.4) than the relatives of normal controls (M 5 12.6; S.D. 5 2.2). The relatives of patients with DD (M 5 13.1; S.D. 5 2.6) did not differ significantly from either of these groups. Controlling for proband SES and marital status and relative education did not change the results. Hence, these variables were not included as covariates in the analyses reported below.
2.2. Depressive personality in relatives The rates of depressive personality and the mean number of depressive personality traits in the relatives of patients with early-onset DD and episodic MDD and normal controls are presented in Table 2. After controlling for the relatives’ age and sex, the rate of depressive personality was significantly higher in the relatives of patients with DD than normal controls, Odds Ratio (OR) 5 3.47 (CI 5 1.16– 10.42), p 5 0.026. However, there were no significant differences in the rates of depressive personality
23
between the relatives of patients with DD and episodic MDD, OR 5 1.81, p 5 0.22, or between the relatives of patients with episodic MDD and normal controls, OR 5 1.92, p 5 0.33. After controlling for the effects of age and sex, the groups of relatives also differed significantly on number of depressive personality traits, F change 5 6.37, df 5 4,321, p 5 0.002. The relatives of patients with DD exhibited a significantly higher number of depressive personality traits than the relatives of patients with episodic MDD, t 5 2.79, p , 0.006, and the relatives of normal controls, t 5 3.04, p , 0.003. The contrast between the relatives of patients with episodic MDD and normal controls was not significant. It is conceivable that these differences are due to the overlap between depressive personality and DD, as we have previously reported that there is a higher rate of DD in the relatives of probands with DD than the relatives of probands with episodic MDD and normal controls (Klein et al., 1995). Indeed, in this sample, 38% of the relatives with DD also met criteria for depressive personality disorder, and 36% of relatives with depressive personality met criteria for DD (k 5 0.30). In addition, it is also possible that some cases of depressive personality represent complications or residual effects of prior episodes of mood disorder, or that current affective disorder might distort subjects’ reports of depressive personality traits. Hence, we repeated these analyses, adding lifetime history of any form of Axis I mood disorder (including bipolar disorder, bipolar disorder not otherwise specified, MDD and DD) on the SCID as a third covariate. The rate of depressive personality continued to be significantly higher among the relatives of patients with early-onset DD than the relatives of normal controls even after controlling for
Table 2 Rates of depressive personality and mean number of depressive personality traits in the relatives of outpatients with dysthymia and episodic major depression and normal controls Proband group
Depressive personality diagnoses a % (N)
Depressive personality traits b M (S.D.)
Dysthymia Episodic major depression Normal controls
14.3% (23) 8.0% (6) 4.4% (4)
2.55 (1.71) 1.87 (1.46) 1.91 (1.47)
Dysthymia (DD) N 5 161; Episodic Major Depression (MDD) N 5 75; Normal Control (NOR) N 5 90. a DD vs. NOR OR 5 3.47 (CI 5 1.16–10.42), p 5 0.026, adjusting for relative age and sex. b DD vs. MDD t 5 2.79, p , 0.006; DD vs. NOR t 5 3.04, p , 0.003, adjusting for relative age and sex.
24
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
age, sex, and lifetime mood disorder, OR 5 3.01 (CI 5 0.98–9.22), p 5 0.05. In addition, after controlling for age, sex, and lifetime mood disorder, the groups of relatives differed significantly on number of depressive personality traits, F change 5 4.01, df 5 5,315, p , 0.02. Again, the number of depressive personality traits was significantly higher in the relatives of patients with DD than patients with episodic MDD, t 5 2.03, p 5 0.04, and normal controls, t 5 2.56, p 5 0.01. Finally, we examined whether the increased rate of depressive personality in the relatives of probands with DD could be due to the higher rate of depressive personality in the DD probands themselves. Of the relatives of the DD probands with depressive personality, 14.7% (17 / 116) were diagnosed as having depressive personality. In contrast, 13.3% (6 / 45) of the relatives of DD probands without depressive personality were diagnosed as having depressive personality. The difference between these rates did not approach significance. Similarly, the number of depressive personality traits in the relatives of DD probands with (M 5 2.56; S.D. 5 1.71), and without (M 5 2.56; S.D. 5 1.73), depressive personality did not differ. We also compared the rate of depressive personality in the relatives of DD probands without depressive personality to the relatives of the normal control probands. For these analyses, we excluded the one relative of the one normal control proband with depressive personality, although the results were the same when this subject (who did not have depressive personality) was included. After controlling for age and sex, there was a trend for a higher rate of depressive personality in the relatives of DD probands without depressive personality than the relatives of normal controls, OR 5 3.27, p 5 0.08. In addition, after controlling for age and sex, the relatives of DD probands without depressive personality exhibited a significantly higher number of depressive personality traits than the relatives of normal controls, F change 5 4.63, df 5 3,130, p 5 0.03.
3. Discussion The present findings support Akiskal’s (Akiskal, 1995, 1997) hypothesis that depressive personality
stems from the same processes that underlie DD, and can be conceptualized as a milder form, or alternative expression, of the same condition. However, the data also suggest that the view that depressive personality lies on a spectrum with the major mood disorders may require qualification, as the familial relationship between depressive personality and MDD was limited to those forms of MDD that were superimposed on a pre-existing DD, and did not appear to be linked to acute, remitting forms of MDD. The major finding in the study was that the firstdegree relatives of patients with early-onset DD (most of whom had a history of superimposed major depressive episodes) exhibited a significantly increased rate of depressive personality, and a significantly greater number of depressive personality traits, than the relatives of normal controls. Importantly, these findings held even after controlling for a lifetime history of dysthymia and other mood disorders in the relatives. In addition, the differing rates of depressive personality between the relatives of patients with DD and the relatives of normal controls were not due to the high rate of depressive personality in the DD probands. DD probands with and without depressive personality had similar rates of depressive personality disorder and identical mean numbers of depressive personality traits in their relatives. Moreover, the relatives of DD probands without depressive personality exhibited a significantly higher number of depressive personality traits, and a trend for a higher rate of depressive personality disorder, than the relatives of normal controls. Thus, depressive personality appears to aggregate in the relatives of probands with DD regardless of whether or not the probands themselves have depressive personality, indicating that depressive personality and DD derive from a shared familial liability (Klein and Riso, 1993). In a previous study using the family history method, I found that patients with DD reported a significantly higher rate of depressive disorders in their first-degree relatives than patients with depressive personality (Klein, 1990). Together with the present findings, these data suggest that while depressive personality and DD share a common familial liability, the degree of the liability appears to be greater in DD.
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
Approximately 4% of the relatives of normal control probands met criteria for depressive personality disorder. This figure is very close to Placidi et al.’s (Placidi et al., 1998) report that the prevalence of depressive temperament in a community sample of over 1000 adolescents and young adults was 3.6%. Taken together, these findings suggest that depressive personality may be fairly common in the community. However, it appears that many of these individuals do not seek treatment until they have developed an Axis I mood disorder (Klein and Miller, 1993). One of the major controversies concerning depressive personality involves whether it is distinct enough from DD to warrant a separate category (Akiskal, 1989; Klein and Vocisano, in press; Phillips et al., 1990). In five studies of the overlap between depressive personality and DD, a median 58% (range 5 20–81%) of individuals meeting criteria for depressive personality did not meet criteria for DD (Hirschfeld and Holzer, 1994; Klein, 1990; Klein and Miller, 1993; Klein and Shih, in press; Phillips et al., 1995). Thus, consistent with the present findings, it appears that the category of depressive personality captures a large number of persons with close phenomenological and familial links to DD who do not qualify for a DD diagnosis. It may be possible to broaden the criteria for DD to include these individuals. One of the major reasons that persons with depressive personality fail to meet criteria for DD is that they do not experience (or at least report) persistent depressed mood (Klein and Vocisano, in press). If the criteria for DD were modified to allow lack of interest or pleasure to substitute for depressed mood, as in the current criteria for MDD, these individuals might be captured. On the other hand, this would also increase the heterogeneity of the DD category, as depressive personality appears to be a milder condition than DD both in terms of symptomatology and familial aggregation (Klein, 1990). Thus, the increased parsimony of a lumping approach to classification must be weighed against the greater homogeneity and preservation of information associated with a splitting approach. The results of the current study did not support a familial relationship between depressive personality and episodic MDD. The rate of depressive personali-
25
ty in the relatives of probands with episodic MDD was midway between the rates in the relatives of the DD and normal probands, and did not differ significantly from either group. In addition, the relatives of patients with episodic MDD exhibited significantly fewer depressive personality traits than the relatives of DD patients, but did not differ from the relatives of normal controls. These findings should be viewed cautiously. The number of relatives of patients with episodic MDD was modest, hence replication with a larger sample is necessary before drawing strong conclusions. In addition, the episodic MDD group may be heterogeneous, and it is possible that a subgroup of these patients may share a familial link with depressive personality. Given the small number of subjects, we could only explore this in a very preliminary manner. There was no evidence that an early age of onset or recurrent subtype in the probands with episodic MDD was associated with depressive personality in their relatives. However, there was a non-significant tendency for a higher rate of depressive personality in the relatives of episodic MDD probands who reported being depressed for at least half the time during the past two years (12.2%) compared to the relatives of probands who were depressed for less than half the time during the past two years (2.9%), p 5 0.14. This raises the possibility that depressive personality is related to a broad spectrum of chronic, subchronic, and intermittent forms of depression, rather than to DD and double depression specifically. The present findings are consistent with previous studies in indicating that there is an elevated rate of depressive personality in the relatives of patients with mood disorders (Klein et al., 1988; Weissman et al., 1984), and are also consistent with previous findings of elevated rates of mood disorders in the relatives of subjects with depressive personality (Klein, 1990; Klein and Miller, 1993). However, none of these studies attempted to distinguish between chronic and episodic forms of depression. Thus, it is unclear whether the findings were limited to particular subtypes, such as double depression. The present study addressed only the familial aggregation aspect of the affective spectrum hypothesis (Akiskal, 1989; Kraepelin, 1921). Twin and adoption studies and developmental studies of infants and children are needed to address the question of
26
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27
whether depressive personality is best conceptualized as a form of temperament, and longitudinal studies are necessary to determine whether depressive personality predisposes to the development of major mood disorders. This study also has implications for studying the role of personality in the mood disorders. In both the current study and our previous offspring study (Klein et al., 1988), semi-structured interviews assessing depressive personality traits successfully distinguished the relatives of depressed and control probands, whereas comprehensive batteries of self-report personality inventories did not (Ouimette et al., 1992, 1996). This suggests that the construct of depressive personality may better capture the personality traits that are associated with familial risk for mood disorders than constructs from the normal personality literature, and / or that semi-structured interviews are more sensitive than self-report questionnaires in assessing these traits. Kendler et al. (1996) recently reported similar findings for schizotypal personality disorder. In weighing the results of this study, several limitations should be considered. First, the sample was modest in size, and limited to several specific forms of depression. It will be important to conduct further studies using larger samples which are representative of the full range of depressive conditions, including late-onset DD and chronic MDD. Second, multiple relatives from the same family were included in the analyses. While this is common practice in family studies, it does violate the assumption of independent observations and could inflate significance levels. Third, comorbidity with non-mood Axis I and II disorders in relatives was not taken into account. Thus, we cannot rule out the possibility that the elevated rate of depressive personality in the relatives of the DD probands is due to comorbid anxiety, substance use, or other personality disorders. Finally, depressive personality was assessed using Akiskal’s (Akiskal, 1983) criteria. These criteria exhibit moderate overlap with the DSM-IV criteria, but are somewhat more restrictive (Hirschfeld and Holzer, 1994). Future studies should employ multiple criteria sets, including DSM-IV, and Akiskal’s (Akiskal, 1983) and Gunderson et al.’s (Gunderson et al., 1994) criteria.
Acknowledgements The author acknowledges the assistance of Rochelle L. Anderson, Thomas A. Aronson, Humberto Lizardi, Shauna K. Donaldson, Tova Ferro, Rollin M. Gallagher, Robert M.A. Hirschfeld, Karen L. Kasch, Helen S. Kelly, Laura M. Klein, Armand W. Loranger, Kimberly A. Norden, Paige Crosby Ouimette, Carolyn M. Pepper, Charles L. Rich, Lawrence P. Riso, and Joseph E. Schwartz in the design of the study, data collection and analysis, and the recruitment of subjects. The author was supported by NIMH Research Grant No. RO1 MH45757.
References Akiskal, H.S., 1983. Dysthymic disorder: Psychopathology of proposed chronic depressive subtypes. Am. J. Psychiatry 140, 11–20. Akiskal, H.S., 1989. Validating affective personality types. In: Robins, L., Barrett, J. (Eds.), The Validity of Psychiatric Diagnosis, Raven Press, New York, pp. 217–227. Akiskal, H.S., 1995. Toward a temperament based approach to depression: Implications for neurobiologic research. Adv. Biochem. Psychopharmacol. 49, 99–112. Akiskal, H.S., 1997. Overview of chronic depressions and their clinical management. In: Akiskal, H.S., Cassano, G.B. (Eds.), Dysthymia and the Spectrum of Chronic Depressions, Guilford Press, New York, pp. 1–34. Gunderson, J.G., Phillips, K.A., Triebwasser, J.T., Hirschfeld, R.M.A., 1994. The diagnostic interview for depressive personality. Am. J. Psychiatry 151, 1300–1304. Hirschfeld, R.M.A., Holzer III, C.E., 1994. Depressive personality disorder: Clinical implications. J. Clin. Psychiatry 55, 10–17. Hollingshead, A.B., 1975. Four Factor Index of Social Position, Yale University, Department of Sociology, New Haven, CT. Kendler, K.S., Thacker, L., Walsh, E., 1996. Self-report measures of schizotypy as indices of familial vulnerability to schizophrenia. Schiz. Bull. 22, 511–520. Klein, D.N., 1990. Depressive personality: Reliability, validity, and relation to dysthymia. J. Abnorm. Psychol. 99, 412–421. Klein, D.N., Clark, D.C., Dansky, L., Margolis, E.T., 1988. Dysthymia in the offspring of parents with primary unipolar affective disorder. J. Abnorm. Psychol. 97, 265–274. Klein, D.N., Miller, G.A., 1993. Depressive personality in nonclinical subjects. Am. J. Psychiatry 150, 1718–1724. Klein, D.N., Ouimette, P.C., Kelly, H.S., Ferro, T., Riso, L.P., 1994. Test–retest reliability of team consensus best-estimate diagnoses of Axis I and II disorders in a family study. Am. J. Psychiatry 151, 1043–1047.
D.N. Klein / Journal of Affective Disorders 55 (1999) 19 – 27 Klein, D.N., Riso, L.P., 1993. Psychiatric disorders: Problems of boundaries and comorbidity. In: Costello, C.G. (Ed.), Basic Issues in Psychopathology, Guilford Press, New York, pp. 19–66. Klein, D.N., Riso, L.P., Donaldson, S.K., Schwartz, J.E., Anderson, R.L., Ouimette, P.C., Lizardi, H., Aronson, T.A., 1995. Family study of early-onset dysthymia: Mood and personality disorders in relatives of outpatients with dysthymia and episodic major depression and normal controls. Arch. Gen. Psychiatry 52, 487–496. Klein, D.N., Shih, J.H., 1998. 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. 107, 319–327. Klein, D.N., Vocisano, C. Depressive and self-defeating (masochistic) personality disorders. In: Millon, T., Blaney, P.H., Davis, R.D. (Eds.), Oxford Textbook of Psychopathology, Oxford University Press, New York, in press. Kraepelin, E., 1921. Manic Depressive Insanity and Paranoia, E&S Livingstone, Edinburgh. Kretschmer, E., 1925. Physique and Character, Harcourt Brace, New York. Loranger, A.W., 1988. Personality Disorder Examination (PDE) Manual, DV Communications, Yonkers, New York. Miller, I.W., Bishop, S., Norman, W.H., Maddever, H., 1985. The modified Hamilton rating scale for depression: Reliability and validity. Psychiatry Res. 14, 131–142. Norden, K.A., Klein, D.N., Ferro, T., Kasch, K., 1995. Who participates in a family study? Compr. Psychiatry 36, 199–206. Ouimette, P.C., Klein, D.N., Clark, D.C., Margolis, E.T., 1992. Personality traits in offspring of patients with unipolar affective disorder: An exploratory study. J. Personality Disord. 6, 91–98. Ouimette, P.C., Klein, D.N., Pepper, C.M., 1996. Personality traits in the first-degree relatives of outpatients with depressive disorders. J. Affective Disord. 39, 43–53.
27
Phillips, K.A., Gunderson, J.G., Hirschfeld, R.M.A., Smith, L.E., 1990. A review of the depressive personality. Am. J. Psychiatry 147, 830–837. Phillips, K.A., Hirschfeld, R.M.A., Shea, M.T., Gunderson, J.G., 1995. Depressive personality disorder: Perspectives for DSMIV. In: Livesley, W.S. (Ed.), DSM-IV Personality Disorders, Guilford Press, New York, pp. 287–302. Placidi, G.F., Signoretta, S., Liguori, A., Gervasi, R., Maremmani, I., Akiskal, H.S., 1998. The semi-structured affective temperament interview (TEMPS-I): Reliability and psychometric properties in 1010 14–26 year-old students. J. Affective Disord. 47, 1–10. Riso, L.P., Klein, D.N., Ferro, T., Kasch, K.L., Pepper, C.M., Schwartz, J.E., Aronson, T.A., 1996. Understanding the comorbidity between early-onset dysthymia and cluster B personality disorders: A family study. Am. J. Psychiatry 153, 900–906. Schneider, K., 1958. Psychopathic Personalities, Cassell, London. Shrout, P.E., Fleiss, J.L., 1979. Intraclass correlations: Uses in assessing rater reliability. Psychol. Bull. 86, 420–428. Spitzer, R.L., Endicott, J., Robins, E., 1978. Research diagnostic criteria: Rationale and reliability. Arch. Gen. Psychiatry 35, 773–782. Spitzer, R.L., Williams, J.B.W., Gibbon, M., First, M.B., 1990. Structured Clinical Interview for DSM-III-R, American Psychiatric Press, Washington, DC. Weissman, M.M., Gershon, E.S., Kidd, K.K., Prusoff, B.A., Leckman, J.F., Dibble, E., Hamovit, J., Thompson, W.D., Pauls, D.L., Guroff, J.J., 1984. Psychiatric disorders in the relatives of probands with affective disorders: The Yale University–National Institute of Mental Health collaborative study. Arch. Gen. Psychiatry 41, 13–21. Westen, D., 1997. Divergences between clinical and research methods for assessing personality disorders: Implications for research and the evolution of Axis II. Am. J. Psychiatry 154, 895–903.