DEPRESSIVE PERSONALITY DISORDER: RATES OF COMORBIDITY WITH PERSONALITY DISORDERS AND RELATIONS TO THE FIVE FACTOR MODEL OF PERSONALITY

Transcription

1 BAGBY Depressive ET AL. Personality Disorder Journal of Personality Disorders, 18(6), , The Guilford Press DEPRESSIVE PERSONALITY DISORDER: RATES OF COMORBIDITY WITH PERSONALITY DISORDERS AND RELATIONS TO THE FIVE FACTOR MODEL OF PERSONALITY R. Michael Bagby, PhD, Deborah R. Schuller, MD, Margarita B. Marshall, BSc, and Andrew G. Ryder, PhD Depressive personality disorder (DPD) is listed in the DSM IV as one of the Disorders for Further Study. In this investigation we examined (1) the rates of comorbidity of DPD with the 10 personality disorders (PDs) in the main text of DSM IV, and (2) the convergent and discriminant validity of DPD in its relation to the 30 facet traits of the Five Factor Model of personality (FFM). One hundred and sixty nine participants with psychiatric diagnoses were interviewed with the Structured Clinical Interview for DSM IV Personality Disorders Questionnaire (SCID II) and completed the Revised NEO Personality Inventory (NEO PI R). A total of 26 (15%) of the participants met diagnostic criteria for at least one of the 10 main text PDs, and 15 (9%) met criteria for DPD. Of those who met criteria for DPD, 10 (59%) of the participants also met criteria for one or more of the 10 main text PDs. Regression analyses indicated a four facet trait set derived from the NEO PI R thought to be uniquely associated with DPD accounted for a significant amount of variance in DPD SCID II PD scores and was significantly larger for DPD than it was for the 9 of the 10 main text PDs; the sole exception was for avoidant PD. Diagnostically, DPD overlaps significantly with other PDs but is distinguishable in its unique relation with traits from the FFM. Depressive personality disorder (DPD) has been the focus of ongoing debate and empirical investigation addressing its viability for inclusion in future editions of the DSM (Bagby, Ryder, & Schuller, 2003; Huprich, 2001a; McDermut, Zimmerman, & Chelminski, 2003; Ryder, Bagby, & Schuller, 2002). One area of controversy centers on the convergent and discriminant From the Centre for Addiction and Mental Health, and Department of Psychiatry, University of Toronto (R.M.B., D.R.S.), the Department of Psychology, McGill University in Montreal (M.B.M.), and the Department of Psychology, Concordia University in Montreal (A.G.R.). Preparation of this paper was facilitated by an Ontario Mental Health Foundation Senior Research Fellowship awarded to the first author. Margarita B. Marshall is now at Department of Psychology, McGill University, Montréal, Canada. Andrew G. Ryder is now at Department of Psychology, Concordia University, Montréal, Canada. Address correspondence to Dr. R. Michael Bagby, Section on Personality and Psychopathology, Centre for Addiction and Mental Health, Clarke Site, 250 College Street, Toronto, Ontario, M5T 1R8, Canada; E mail: michael_bagby@camh.net 542

2 BAGBY ET AL. 543 TABLE 1. Diagnostic Overlap among Personality Disorders PD % with DPD who have comorbid main text PD % without DPD who have a main text PD PAR SZT SZD HIS NAR BDL ASL AVD DPN OCD Any PD Note. All figures based on interview. PD = Personality Disorder; DPD = Depressive; PAR = Paranoid; SZT = Schizotypal; SZD = Schizoid; HIS = Histrionic; NAR = Narcissistic; BDL = Borderline; ASL = Antisocial; AVD = Avoidant; DPN = Dependent; OCD = Obsessive Compulsive. validity of the DPD diagnostic criteria in relation to the Axis I condition dysthymic disorder (DD) and the 10 Axis II personality disorders (PDs) in the main text of DSM IV (APA, 1994). Although the relation between DPD and DD has been extensively examined (for reviews see Huprich, 2001a; Ryder et al., 2002), the relation of DPD to other PDs has not been so thoroughly considered. One clear problem is the high degree of overlap between the frequency of DPD and main text PD diagnoses (see Table 1). For example, McDermut et al. (2003) reported that of those individuals who met criteria for DPD according to the Structured Interview for DSM IV Personality Disorders (Pfohl, Blum, & Zimmerman, 1997), 66% also met criteria for another PD. The highest rates of comorbidity were for avoidant (43%), borderline (22%), obsessive compulsive (21%), and paranoid (16%). All other disorders had overlap rates of less than 10%. Similarly, using DSM III R (APA, 1987) criteria for PDs, assessed with the Personality Disorder Examination (PDE; Loranger, 1988) and Akiskal s (1983) criteria for DPD, Klein and Shih (1998) reported that 58% of individuals who met this criteria for DPD also met criteria for another PD. Borderline (26%), avoidant (20%), histrionic (17%), and paranoid (16%) had the highest rates of comorbidity. All other PDs had overlap rates less than 6%. Although the individual overlap rates of DSM main text PDs with DPD are not excessive, the overall rate of overlap is large and suggests that the addition of DPD may exacerbate existing difficulties associated with the high comorbidity rates among main text PDs. It has been argued, however, that high rates of comorbidity of DPD with other PDs is not a valid criterion to evaluate its potential inclusion in the main text of the DSM, as many of the existing PDs in DSM already have such high rates (Clark & Watson, 1999). An alternative approach to establishing the validity of DPD is to examine its unique association to dimensional per-

3 544 DEPRESSIVE PERSONALITY DISORDER sonality traits relative to other PDs (Reynolds & Clark, 2001). Several authors contend that dimensional models of personality offer a potentially useful alternative conceptualization of PDs and that these models can better accommodate comorbidity (Lynam & Widiger, 2001). Consistent with this view, Widiger, Trull, Clarkin, Sanderson, and Costa (2002) suggested that PDs can be understood within the framework of the Five-Factor Model of personality (FFM). Costa and Widiger (2002) transposed main text PD symptom criteria onto FFM traits. The five personality domains of the FFM Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness are each composed of six lower order facet traits, which provide details about specific aspects of the higher order domains (see Table 2). Costa and Widiger proposed that a constellation of four facet traits (high levels of Anxiety, Depression, and Self Consciousness and a low level of Tendermindedness) characterize DPD. Huprich (2003b) recently examined these associations using bivariate correlations and regression analyses on data generated from a sample of 67 psychiatric outpatients. Results indicated that these four facets were significantly correlated with three separate measures of DPD symptoms an interview based instrument, the Diagnostic Interview for Depressive Disorder (DIDD; Gunderson, Phillips, Triebwasser, & Hirschfeld, 1994); and two self report measures, the Depressive Personality Disorder Inventory (DPDI; Huprich, Margrett, Barthelemy, & Fine, 1996) and the Structured Clinical Interview for DSM IV Axis II Disorders Self Report) (SCID II; (First, Gibbon, Spitzer, Williams, & Benjamin, 1997). Specifically, Self Consciousness and (low) Tendermindedness correlated significantly with all three DPD measures. Moreover, individual Revised NEO Personality Inventory (NEO P R) trait facets that were not anticipated as linked to DPD also correlated significantly with DPD scores. The goal of the current study was to extend these previous efforts in order to explore the distinctiveness of DPD in relationship to other main text PDs and personality dimensions of the FFM. We used interview based diagnoses to examine comorbidity rates and correlation and regression analyses to explore the association of these PDs with domain and facet traits from the FFM. Neither Klein and Shih (1998) nor McDermut et al. (2003) compared rates of comorbidity of DPD and main text PDs to comorbidity of main text PDs with one another. Examining DPD comorbidity rates relative to those of main text PDs provides a referent point from which one can begin to evaluate whether DPD adds meaningfully to our current PD nosology or simply worsens concerns with diagnostic overlap. In a similar vein, Huprich (2003b) used regression analysis to examine the association of the DPD four facet trait set with DPD, but did not similarly analyze the relation of that set to the main text PDs. Thus, although convergent validity of DPD has been examined, the discriminant validity of the DPD diagnostic criteria has not been demonstrated. Given the high rate of comorbidity between DPD and the main text PDs, demonstrating the discriminant validity of the DPD diagnos-

4 TABLE 2. Pearson Correlations among SCID II Personality Disorder Self Report Scales and FFM Facet Trait Scales Facet Scale DPD PAR SZT SZD HIS NAR BDL ASL AVD DPN OCD N1 Anxiety.62**.42**.30** **.09.54**.35**.27 N2 Anger Hostility.49**.44** **.49**.19.32**.28**.36** N3 Depression.72**.44**.36** **.06.63**.41**.30** N4 Self Consciousness.63**.44**.36** **.08.72**.37**.28 N5 Impulsiveness.30** ** N6 Vulnerability.62**.41**.33** **.14.54**.45**.17 E1 Warmth.38**.38**.29**.19.36** ** E2 Gregariousness.30**.23.32**.31**.30** ** E3 Assertiveness.34** ** **.14**.03 E4 Activity ** ** E5 Excitement Seeking ** E6 Positive Emotions.51**.32**.30**.31**.36** ** O1 Fantasy.31** ** O2 Aesthetics O3 Feelings O4 Actions O5 Ideas O6 Values A1 Trust.45**.51**.36** **.38**.15.41** A2 Straightforwardness ** A3 Altruism A4 Compliance **.32**.28** A5 Modesty ** A6 Tender Mindedness C1 Competence.46** **.28.44** C2 Order **.28**.19 C3 Dutifulness ** C4 Achievement Striving.34** ** ** C5 Self Discipline.41** **.09.43**.35**.08 C6 Deliberation.25** **.39** Note. N = 169. **Correlation is significant at p <.0002 (Bonferonni corrected for p <.05). 545

5 546 DEPRESSIVE PERSONALITY DISORDER tic criteria will be an important step toward determining whether DPD should be included in the main text of future editions of DSM. 1 METHOD PARTICIPANTS The sample was composed of 169 participants (89 women, 80 men), all of whom were outpatients assessed and screened for treatment and/or inclusion in various research protocols in a clinical research department of a large university affiliated teaching and research hospital. The mean age of participants was years (SD = 12.58). Forty six percent of the sample were single and never married; 33% married; 16% either divorced or separated; and 3% widowed. The sample was predominantly of European descent. The mean Blishen value (a Canadian socioeconomic status index) was (SD = 15.69), indicating that participants in the present study were generally of middle class socioeconomic status, with (SD = 3.42) mean years of education. This sample was diagnostically heterogeneous with most participants meeting diagnostic criteria for anxiety, mood, or substance disorders. 2 MEASURES SCID II (First et al., 1997) is designed to assess the main text PDs and is the companion to the Structured Clinical Interview for DSM IV Axis I Disorders Patient edition (SCID I/P). The administration of SCID II follows a two tiered procedure. First, respondents complete a 119 item self report questionnaire using a Yes/No response format. Each of the questions corresponds to a diagnostic criterion for either one of the main text PDs or the two additional PDs listed in Appendix B of DSM IV (i.e., Passive Aggressive and DPD). Only scores for the main text PDs and DPD were used in this study. After respondents have completed the questionnaire, the interviewer identifies those personality disorders for which respondents endorsed sufficient criteria for a particular PD diagnosis. Persons meeting self report criteria for any given PD are then administered those portions of the SCID II clinical interview that correspond to those PDs endorsed by participants in order to assign a formal diagnosis. 1. Huprich (2003a) did address the issue of discriminant validity indirectly by removing symptom variance in hierarchical regressions using a PD symptom composite score as a covariate. We employ a similar strategy in this study, but also examine directly the discriminant validity of the four trait facet set separately with each individual PD. 2. The Structured Clinical Interview for DSM IV, Axis I Disorders (Version 2.0/Patient Form) (SCID I/P; DSM IV First, Spitzer, Gibbon, & Williams, 1995) was used to assess for the presence of Axis I disorders for these participants. A detailed breakdown of diagnoses for this sample is available from the first author upon request. Given that most participants met criteria for either a mood or anxiety disorder, an Axis I mood or anxiety disorder diagnosis did not preclude conferring of a PD diagnosis.

6 BAGBY ET AL. 547 Dimensionalized scores derived using the self report symptom count scores for each of the PDs by summing the endorsed symptoms for each disorder. A number of studies have shown these dimensional self report scales are both valid (Carey, 1994; Ekselius, Lindstrom, von Knorring, Bodlund, & Kullgren, 1994; Huprich, 2003b; Jacobsberg, Perry, & Frances, 1995; Neal, Fox, Carroll, Holden, & Barnes, 1997) and relatively stable (Ouimette & Klein, 1995). The Revised NEO Personality Inventory (Costa & McCrae, 1992) is composed of 240 self report items, with separate scales for each of the five domains of the FFM. Items are answered on a 5 point Likert format scale ranging from 0 = Strongly Disagree to 4 = Strongly Agree. Each domain scale also consists of six lower order facet scales, which in total constitute 30 trait facet scales (see Table 2). The same five factors and their corresponding 30 facets captured in nonclinical populations are represented in psychiatric samples (Bagby et al., 1999) and stability estimates are adequate in both clinical and nonclinical groups (Costa & McCrae, 1992; Harkness, Bagby, Levitt, & Joffe, 2002; Santor, Bagby, & Joffe, 1997; Trull, Useda, Costa, & McCrae, 1995). PROCEDURE All patients completed the SCID II self report questionnaire and the NEO PI R and were then interviewed using the SCID II by advanced research assistants, M.A.-level clinical psychologists, or a postdoctoral clinical fellow. Although inter rater agreement was not formally determined, all interviewers were trained extensively in the interview procedures. STATISTICAL ANALYSES Rates of comorbidity were determined using percentages. Measures of central tendency (Means and Medians) were calculated to characterize average comorbidity rates between DPD and the 10 main text PDs, as well as comorbidity among the main text PDs alone. These analyses were all based on interview derived SCID II PD diagnoses using the DSM IV categorical method. Associations between DPD, the 10 main text PDs, and the trait facets of the FFM were examined using Pearson (bivariate) correlations and linear as well as hierarchical regressions. In the linear regression analyses, dimensionalized PD sum scores based on the SCID II self report questionnaires served as the criterion variables and the trait facets from the NEO PI R served as the predictor variables. In the hierarchical regressions, corrected composite PD scores (see below) also served as a predictor variable. Z tests were used to examine differences between R 2 values derived from the regression analyses, with alpha Bonferonni corrected and set at p <.01.

7 548 DEPRESSIVE PERSONALITY DISORDER RESULTS COMORBIDITY RATES Based on the SCID II interview, 15% (n = 26) of the participants met diagnostic criteria for at least one of the 10 main text PDs, and 9% (n = 15) met criteria for DPD. 3 Comorbidity rates of DPD with the main text PDs are summarized in Table 1. The overlap rates of DPD with these other PDs ranged from 0% to 20%, (M = 8%; Mdn = 7%). The individual disorders that overlapped most highly with DPD were avoidant, borderline, and schizoid PDs. Overall, 60% of the individuals with DPD met diagnostic criteria for one or more of the 10 main text PDs. Of those respondents who met criteria for at least one other main text PD, 35% also met criteria for DPD (range 0 100%; M=43%; Mdn = 38%). Finally, the percentage of those in the sample who did not meet criteria for DPD but who did meet criteria for another PD ranged from 0% to 5% (M=2%; Mdn =1 %). The lowest frequency rates diagnoses within the 10 main text PDs alone occurred for histrionic, dependent, and schizoid PDs. Overall, fewer than 14% of those who did not meet criteria for DPD met criteria for a main text PD. RELATIONS TO PERSONALITY TRAITS Bivariate Correlations. Table 2 displays the bivariate correlations between the 30 trait facets of the NEO PI R and the SCID II self report scores, including DPD. Given the large number of correlations in this matrix, the significance level was Bonferonni corrected for p <.05. Three of the four facets from the DPD FFM facet set anxiety, depression and self consciousness correlated significantly with DPD SCID II scores (the exception was the Tendermindedness facet); however, as a group, these same three trait facets also correlated significantly with paranoid, schizotypal, narcissistic, borderline, avoidant, dependent, and obsessive compulsive PDs. Moreover, 19 other trait facets not identified as FFM trait descriptors also correlated significantly with DPD scores. Linear and Hierarchical Regressions. A series of regression analyses was performed to examine the convergent and discriminant validity of DPD using the FFM four facet DPD trait set. In the first series of analyses, we performed simple linear regressions with DPD and main text PDs serving as criterion variables in separate analyses. In each of these analyses the four trait facet scales served as the predictor variable and were entered as a block in a single step. If the DPD construct is descriptively distinguishable from the main text PDs, then the four facets thought to be, as a group, uniquely associated with this disorder should produce larger R 2 adj values for DPD than for the other PDs. Table 3 displays the re- 3. Rates of DPD diagnoses are based on SCID II interview.

8 BAGBY ET AL. 549 TABLE 3. Coefficients of Determination, Standardized Beta Weights, and Standard Error Values for Linear Regressions Using the FFM, DPD Four Facet Trait Set Predictor N1 N3 N4 A6 PD R 2 adj B (SE) B (SE) B (SE) B (SE) DPD.55* (.02).46*.08 (.02) (.02) (.01) PAR.24* (.02) (.02) (.02) (.01) SZT.13* (.02) (.02) (.02) (.01) SZD (.01) (.01) (.01) (.01) HIS (.01) (.01) (.01) (.01) NAR.10* (.03) (.03) (.03) (.02) BDL.32* (.03) (.03) (.03) (.02) ASL (.02) (.02) (.02) (.01) AVD.53* (.01) (.01).53*.09 (.01) (.01) DPN (.01) (.01) (.01) (.01) OCD.10* (.02) (.02) (.02) (.01) Note. *Regression model or facets is significant at p <.0045 (Bonferonni corrected for p <.05). Standard error; R 2 adj = Adjusted R squared value; β = Standardized beta coefficient; B = Unstandardized beta coefficient; SE = Standard error (statistic based on final model); N1 = Anxiety; N3 = Depression; N4 = Self Consciousness; A6 = Tendermindedness. sults from these linear regressions. The four facet set accounted for a significant amount of variance in DPD scores; however, this set was also a significant predictor for 6 of the 10 main text PD scores. The R 2 adj value for the four facet set was nonetheless significantly larger for DPD than for all other PDs (all z scores > 2.73, p <.05), with the exception of avoidant PD (z = 0.20, p = 0.841). In the next set of regression analyses hierarchical models were constructed. A corrected composite scale score consisting of self report SCID II items was created for each regression equation by deleting those items representing the criterion variable in each of the respective regression models. Each respective corrected composite scale was entered first into the model (Step 1), followed by the block entry of the DPD four facet set (Step 2). This modelling was designed to control for the possibility that the significance of the relation between the DPD four facet set and any given SCID II PD criterion score might be spuriously attributable to the co occurring personality psychopathology associated with the other PDs. Table 4 displays the results from these regression analyses. In Step 1, the corrected composite SCID II scale significantly predicted individual SCID II PD scores for DPD and all main text PDs, with the exception of schizoid PD. In Step 2, after accounting for variance contributed by the respective, corrected composite scores, the four facet set explained significant additional variance for DPD as well histrionic and avoidant PDs. The adjusted R 2 change value associated with the addition of the four facet set to the model predicting DPD symptom count scores was significantly larger than in the

9 TABLE 4. Coefficients of Determination, Coefficients of Change, Standardized Beta Weights, and Standard Error Values for Hierarchical Regressions Using the FFM, DPD Four Facet Trait Set Predictor Step 1 Step 2 SCID II N1 N3 N4 A6 R 2 adj R 2 PD B (SE) B (SE) B (SE) B (SE) B (SE) DPD.67*.16*.44*.08 (.01) (.01).36*.06 (.01) (.01) (.01) PAR.50*.01.67*.11 (.01) (.02) (.02) (.02) (.01) SZT.36*.01.62*.09 (.01) (.02) (.01) (.02) (.01) SZD (.01) (.01) (.01) (.01) (.01) HIS.11*.12*.42*.05 (.01) (.01) (.01) (.01) (.01) NAR.45*.05.80*.19 (.02) (.02) (.02) (.02) (.02) BDL.52*.04.57*.17 (.02) (.02) (.02) (.02) (.02) ASL.09* (.01) (.02) (.02) (.02).23*.04 (.01) AVD.57*.28*.23*.04 (.01) (.01) (.01).49*.08 (.01) (.01) DPN.29*.03.46*.05 (.01) (.01) (.01) (.01) (.01) OCD.15*.03.28*.04 (.01) (.02) (.02) (.02) (.01) Note. Change coefficient or facet is significant at p <.0045 (Bonferonni corrected p <.05). R 2 adj = Adjusted R squared value; R2 = R squared change value; β = Standardized beta coefficient (final hierarchical model); B = Unstandardized beta coefficient; SE = Standard error; SCID II = Overall SCID II PD symptom count score (not including criterion PD); N1 = Anxiety; N3 = Depression; N4 = Self Consciousness; A6 = Tendermindedness. 550

10 BAGBY ET AL. 551 models predicting four of the main text PDs (all z scores > 2.40, p <.01); specifically, these were paranoid, schizotypal, schizoid, borderline, dependent, and obsessive compulsive PDs. There was no significant difference in adjusted R 2 change ( R 2 ) values associated with the prediction of histrionic, narcissistic, borderline, antisocial, avoidant, or dependent PDs versus DPD, suggesting that DPD is not generally distinguishable from these main text PDs after the shared symptom variance among them is removed. Although not statistically significant, it is notable that the R 2 value is nevertheless larger in the model predicting avoidant PD symptom count scores than for the model predicting DPD symptom count scores. Examination of the standardized β weights from the final model presented in Table 4, however, reveals a different relative contribution of each of the four facets in relation to each PD. Specifically, (low) Depression was associated with the greatest predictive capacity for predicting histrionic PD, (low) Self Consciousness for narcissistic PD, (low) Tendermindedness for antisocial PD, Self Consciousness for avoidant PD, and Depression for DPD. 4 DISCUSSION One means of establishing the diagnostic validity of DPD is to demonstrate its uniqueness relative to main text PDs. In this investigation we addressed this issue by first examining rates of comorbidity of DPD and main text PDs; then by defining the unique association of DPD with a set of FFM trait facets thought to characterize DPD. The comorbidity rate of DPD with main text PDs was approximately 60% a rate similar to that reported by Klein AND Shih (1998) and McDermut et al., 2003). It is thus reasonable to presume that if an individual met diagnostic criteria for one or more main text PDs, then he/she would also meet diagnostic criteria for DPD. Conversely, if an individual did not meet diagnostic criteria for DPD, then it is unlikely he or she meets criteria for another main text PD. Rates of comorbidity between DPD and individual main text PDs are not excessively high in that overlap rates for individual PDs with DPD never exceed 20%. The uniqueness of the DPD diagnostic criteria set is strongly supported by the results from the regression analyses. For example, the current results indicate that the four facet trait set from the FFM thought to translate into DPD actually shows stronger associations with dimensionalized DPD scores relative to the main text PDs. Furthermore, this effect is maintained for all but one disorder (avoidant PD), even after accounting for the entire variance associated with the main text PDs as captured by the corrected composite score. At the same time, it should be noted that many of these differences did not reach Bonferonni corrected levels of statistical significance. 4. The relationships between the Depression facet and Histrionic PD as well as the Self Consciousness facet and Narcissistic PD were not significant.

11 552 DEPRESSIVE PERSONALITY DISORDER Although the four facet trait set fails to differentiate DPD from avoidant PD, the pattern of association of the individual facets differs for the two disorders and is thematically consistent with the diagnostic coloration of each. For example, the highest weight for DPD is the Depression trait facet, whereas the highest weight for avoidant PD is the Self Consciousness facet. These results not only highlight the potential contribution of FFM facet traits to make important and meaningful diagnostic distinctions between disorders (Lynam & Widiger, 2001; Morey et al., 2002), but suggest also that similarities and differences between PDs, particularly those with high levels of comorbidity, are better understood as differences in dimensional personality traits rather than by categorically separate entities with high rates of comorbidity. The four facet trait set proposed by Widiger et al. (2002) was a unique and significant predictor of the DPD diagnostic criteria set; however, some of these facets contributed only weakly, and other facets from the FFM not included in this facet set correlated significantly and strongly with this disorder. Huprich (2003a) reported that Self Consciousness and Tender mindedness facets contributed significantly to the prediction of DPD, whereas in the current investigation, Anxiety and Depression facets were the strongest predictors of DPD. We believe that a slightly modified FFM facet set is required to define more accurately the descriptive diagnostic characteristics of DPD. One approach might be to develop a set of consensus prototypical ratings as recently employed by Lynam and Widiger (2001) to describe the main text PDs using facet traits from the FFM. These ratings could then be examined and validated empirically, with particular focus on identifying a set of traits that maximally distinguish DPD from the other PDs. Of course, the diagnostic overlap of the PDs makes this a challenging task. Another approach would be to derive a purely empirically based set of facets derived from the FFM 30 facet trait set by using methods to maximize distinctiveness, selecting sets of traits that both characterize but also distinguish DPD from main text PDs. Whatever approach is applied, we believe that using dimensional traits rather than categorical nosology to characterize the DPD has the potential to enhance the conceptualization and description of this disorder (Ryder, Bagby, Marshall, & Costa, in press). The use of the dimensional traits to describe other personality disorders, including antisocial, borderline, avoidant, obsessive compulsive, and schizotypal, has been employed by other investigators and provided a more precise description of these disorders and has clarified, to some extent, the high rates of comorbidity among some of these disorders (Lynam & Widiger, 2001; Miller, Lynam, Widiger, & Leukefeld, 2001; Morey, Gunderson, Quigley, & Lyons, 2000; Morey et al., 2002; Trull, Widiger, Lynam, Costa, & Lyons, 2003). Two limitations of the present study must acknowledged. First, in an effort to maximize statistical power, we opted to base our results on the SCID II questionnaire scores instead of the SCID II interview data. Although data suggests that the results of the questionnaire are comparable

12 BAGBY ET AL. 553 to those obtained using the interview, symptom count scores derived from the interview would provide a more stringent test of Costa and Widiger s proposed set of facets as characteristic of DPD. Second, only 15 participants met diagnostic criteria for DPD based on the SCID II interview. Thus it could be argued that the NEO PI R facet scales are predicting DPD symptoms that are secondary to main text PD symptoms or Axis I mood disorder symptoms. REFERENCES Akiskal, H.S. (1983). Dysthymic disorder: Psychopathology of proposed chronic depressive subtypes. American Journal of Psychiatry, 140, American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., Rev.) Washington, DC: Author. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders. (4th ed.). Washington, DC: Author. Bagby, R.M., Costa, P.T., McCrae, R.R., Livesley, W.J., Kennedy, S.H., Levitan, R.D., et al. (1999). Replicating the five factor model of personality in a psychiatric sample. Personality and Individual Differences, 27, Bagby, R.M., Ryder, A.G., & Schuller, D.R. (2003). Depressive personality disorder: A critical overview. Current Psychiatry Reports, 5, Carey, K.B. (1994). Use of the structured clinical interview for DSM III R personality questionnaire in the presence of severe axis I disorders: A cautionary note. Journal of Nervous and Mental Disease, 182, Clark, L.A., & Watson, D. (1999). Personality, disorder, and personality disorder: Towards a more rational conceptualization. Journal of Personality Disorders, 13, Costa, P.T., & McCrae, R.R. (1992). Revised NEO personality inventory: Professional manual. Odessa, FL: Psychological Assessment Resources. Costa, P.T., & Widiger, T.A. (2002). Personality disorders and the five factor model of personality (2nd ed.). Washington, DC: American Psychological Association Press. Ekselius, L., Lindstrom, E., von Knorring, L., Bodlund, O., & Kullgren, G. (1994). SCID II interviews and the SCID screen questionnaire as diagnostic tools for personality disorders in DSM III R. Acta Psychiatrica Scandinavica, 90, First, M.B., Gibbon, M., Spitzer, R.L., Williams, J.B.W., & Benjamin, L.S. (1997). User s guide for the structured clinical interview for DSM IV axis II personality disorders. Washington, DC: American Psychiatric Press. First, M.B., Spitzer, R.L., Gibbon, M., & Williams, J.B.W. (1995). Structured clinical interview for DSM IV Axis I disorders Patient edition, Version 2.0. New York: New York Biometrics Research Department. Gunderson, J.G., Phillips, K.A., Triebwasser, J., & Hirschfeld, R.M. (1994). The diagnostic interview for depressive personality. American Journal of Psychiatry, 151, Harkness, K.L., Bagby, R.M., Levitt, A.J., & Joffe, R.T. (2002). Major depression, minor depression and the five factor model of personality. European Journal of Personality, 16, Huprich, S.K. (2001a). The overlap of depressive personality disorder and dysthymia, reconsidered. Harvard Review of Psychiatry, 9, Huprich, S.K. (2001b). Object loss and object relations in depressive personality analogues. Bulletin of the Menninger Clinic, 65, Huprich, S.K. (2003a). Evaluating NEO Personality inventory revised profiles in veterans with personality disorders. Journal of Personality Disorders, 17, Huprich, S.K. (2003b). Evaluating facet level predictions and construct validity of depressive personality disorder. Journal of Personality Disorders, 17,

13 554 DEPRESSIVE PERSONALITY DISORDER Huprich, S.K., Margrett, J.E., Barthelemy, K.J., & Fine, M.A. (1996). The depressive personality disorder inventory: An initial investigation of its psychometric properties. Journal of Clinical Psychology, 52, Jacobsberg, L., Perry, S., & Frances, A. (1995). Diagnostic agreement between the SCID II screening questionnaire and the personality disorder examination. Journal of Personality Assessment, 65, 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. Journal of Abnormal Psychology, 107, Loranger, A.W. (1988). Personality disorder examination (PDE) manual. Yonkers, NY: DV Communications. Lynam, D.R., & Widiger, T.A. (2001). Using the five factor model to represent the DSM IV personality disorders: An expert consensus approach. Journal of Abnormal Psychology, 110, McDermut, W., Zimmerman, M., & Chelminski, I. (2003). The construct validity of depressive personality disorder. Journal of Abnormal Psychology, 112, Miller, J.D., Lynam, D.R., Widiger, T.A., & Leukefeld, C. (2001). Personality disorders as extreme variants of common personality dimensions: Can the five factor model adequately represent psychopathy? Journal of Personality, 69, Morey, L.C., Gunderson, J., Quigley, B.D., & Lyons, M. (2000). Dimensions and categories: The big five factors and the DSM personality disorders. Assessment, 7, Morey, L.C., Gunderson, J.G., Quigley, B.D., Shea, M.T., Skodol, A.E., McGlashan, T.H., et al. (2002). The representation of borderline, avoidant, obsessive compulsive, and schizotypal personality disorders by the five factor model. Journal of Personality Disorders, 16, Neal, L.A., Fox, C., Carroll, N., Holden, M., & Barnes, P. (1997). Development and validation of a computerized screening test for personality disorders in DSM III R. Acta Psychiatrica Scandinavica, 95, Ouimette, P.C., & Klein, D.N. (1995). Test retest reliability, mood state dependence, and informant subject concordance of the SCID Axis II questionnaire in a nonclinical sample. Journal of Personality Disorders, 9, Pfohl, B., Blum, N., & Zimmerman, M. (1997). Structured interview for DSM IV personality (SIDP IV). Washington, DC: American Psychiatric Association Press. Reynolds, S.K., & Clark, L.A. (2001). Predicting dimensions of personality disorder from domains and facets of the Five Factor Model. Journal of Personality, 69, Ryder, A.G., Bagby, R.M., Marshall, M.B., & Costa, P.T. (in press). Depressive Personality as a Category or Dimension: Theoretical and Clinical Implications. (M. Rosenbluth, S.H. Kennedy, & R.M. Bagby, Personality Dimensions, Disorders & Depression: Models & Implications for Treatment (Arlington VA: American Psychiatric Publishing Inc. Ryder, A.G., Bagby, R.M., & Schuller, D.R. (2002). The overlap of depressive personality disorder and dysthymia: a categorical problem with a dimensional solution. Harvard Review of Psychiatry, 10, Santor, D.A., Bagby, R.M., & Joffe, R.T. (1997). Evaluating stability and change in personality and depression. Journal of Personality and Social Psychology, 73, Trull, T.J., Useda, J.D., Costa, P.T., & Mc- Crae, R.R. (1995). Comparison of the MMPI 2 Personality Psychopathology Five (PSY 5), the NEO PI, and NEO PI R. Psychological Assessment, 7, Trull, T.J., Widiger, T.A., Lynam, D.R., & Costa, P.T., Jr. (2003). Borderline personality disorder from the perspective of general personality functioning. Journal of Abnormal Psychological, 112, Widiger, T.A., Trull, T.J., Clarkin, J.F., Sanderson, C., & Costa, P.T. (2002). A description of the DSM IV personality disorders with the five factor model or personality. In P.T. Costa, Jr. & T.A. Widiger, Personality disorders and the five factor model of personality (2nd ed. Pp ). Washington, DC: American Psychological Association.