European Psychiatry 27 (2012) 582–590
Original article
Psychometric properties of the Spanish version of the Diagnostic Interview for Depressive Personality L.J. Irastorza a,*, P. Rojano b, T. Gonzalez-Salvador b, J. Cotobal c, M. Leira a, C. Rojas d, G. Rubio c, C. Rodrı´guez-Rieiro e, J.M. Bellon e, M. Alvarez c, C. Rodrı´guez c, C. Arango f a
Mental Health Centre, Arganda del Rey, Hospital Virgen de la Torre, 28500 Madrid, Spain Mental Health Centre, Colmenar Viejo, Madrid, Spain c Mental Health Centre, Retiro, Madrid, Spain d Madrid, Spain e Service of Preventive Medicine-Statistics, Hospital Universitario Gregorio Maran˜on, Madrid, Spain f Department of Psychiatry, Hospital Universitario Gregorio Maran˜on, Centro de Investigacio´n Biome´dica en Red de Salud Mental, CIBERSAM, Madrid, Spain b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 1 October 2010 Received in revised form 15 November 2010 Accepted 16 November 2010 Available online 5 February 2011
The aim of this study was to evaluate the reliability and validity of the Spanish-language version of the Diagnostic Interview for Depressive Personality (DIDP). The DIDP was administered to 328 consecutive outpatients and the test–retest and inter-rater reliability were assessed. Factor analysis was used in search of factors capable of explaining the scale and a cutoff point was established. The DIDP scales showed adequate Cronbach’s a values and acceptable test–retest and inter-rater reliability coefficients. Convergent and discriminant validity were explored, the latter with respect to avoidant and borderline personality disorders. The results of the factor analysis were consistent with the four-factor structure of the DIDP scales. The receiver operating characteristic (ROC) analysis revealed the area under the curve to be 0.848. We found 30 to be a good cutoff point, with a sensitivity of 74.5% and a specificity of 78.5%. The DIDP proved to be a reliable and valid instrument for assessing depressive personality disorder, at least among our outpatients. The psychometric properties of the DIDP support its clinical usefulness in assessing depressive personality. ß 2010 Elsevier Masson SAS. All rights reserved.
Keywords: Depressive personality disorder Dimensions Interview Validation Factor analysis
1. Introduction The diagnostic classification of depressive personality disorder (DPD) has varied in recent years. In the DSM-IV-TR, it appears in Appendix B, where it is described as worthy of further study. The DSM-IV-TR indicates that DPD is characterised by a configuration of depressive traits and symptoms that are not better accounted for by dysthymic disorder and do not occur exclusively during major depressive episodes. In this classification, the definition of the condition is based on five or more criteria, which, when combined with others from the literature on depressive personality, form the four dimensions of the Diagnostic Interview for Depressive Personality (DIDP), i.e. negativistic, introverted/tense, passive/ unassertive, and self-sacrificing [46]. In the International Classification of Diseases, Ninth Revision (ICD-9), depressive personality is considered an independent diagnostic entity among the personality disorders and is included as an affective personality subtype.
* Corresponding author. Tel.: +34 918701125; fax: +34 918703937. E-mail address:
[email protected] (L.J. Irastorza). 0924-9338/$ – see front matter ß 2010 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.eurpsy.2010.11.003
In the latest edition of the ICD (ICD-10), DPD has disappeared as an independent entity and is included with dysthymia. Since the introduction of the structured DIDP by Gunderson et al. [17], additional inventories or interviews have been used for the diagnosis of this disorder. These include the Structured Clinical Interview for DSM-IV Axis II Disorders (SCID-II) [15], the Schneider criteria for DPD [53,34,35,37], the Depressive Personality Disorder Inventory (DPDI) [27,28], the Temperament and Character Inventory (TCI) [12,58], the NEO–Personality Inventory Revised (NEO-PI-R) [13], and the Millon Clinical Multiaxial Inventory-III (MCMI-III). Research has centered mostly on whether DPD exists as a personality disorder [35,20,22,31,32,43,48], whether it should be included as DPD dimensionally and as an affective disorder categorically [52], or whether it can be classified with dysthymia or as an affective disorder [34,20,2,36,49]. The most widely used dimensional classifications are the TCI [12] and the five-factor model [13]. The major impediment to the official recognition of DPD on Axis II has been the problem of comorbidity, particularly with the dependent, avoidant, and borderline personality disorders on Axis II, and with dysthymic disorder on Axis I. Nevertheless, several studies support a distinct DPD construct, with a distinct family history (e.g. mood disorders and substance use disorders)
L.J. Irastorza et al. / European Psychiatry 27 (2012) 582–590
and developmental history correlates (e.g. interpersonal loss and negative parental perceptions) [23,42,47]. The aim of the present study was to validate the Spanishlanguage version of an interviewer-administered scale for diagnosing DPD, the DIPD [17]. In addition to examining its psychometric properties, we performed a factor analysis to determine whether the dimensions of the initial scale or those of the previously adapted Spanish version should be maintained (Irastorza LJ. Personalidad depresiva: concepto y diferenciacio´n. Unpublished doctoral dissertation. Universidad Complutense de Madrid. 2000). 2. Patients and methods We performed a cross-sectional observational study (November 2005 to March 2008). The clinical sample comprised 328 consecutive outpatients (80.2% women) with a mean age of 45.5 11.6 years (range: 20–74) attended at three mental health centres (Arganda del Rey, Retiro, and Colmenar Viejo) in Madrid, Spain. The patients were selected from among those habitually attended by each professional and they all met the inclusion criteria: consecutively attended at each mental health centre, over 18 years of age, and suffering from depressive disorders, anxiety disorders, eating disorders, or personality disorders according to the DSM-IV-TR. The exclusion criteria were other DSM-IV-TR mental disorders and an educational level lower than elementary school. The Institutional Review Board of Hospital General Universi˜ o´n (Madrid, Spain) approved the research tario Gregorio Maran protocol and all participants gave their informed consent before entering the study. Mental disorders were assessed according to DSM-IV-TR criteria by six experienced psychiatrists and a psychologist who received three weeks of training in evaluation of the DIDP. Prior to the study, a discussion was held with the participating health professionals about possible difficulties that could arise at each mental health centre. All interviewers and patients were blind to all scores during the test–retest and inter-rater examination. Throughout the interview, participants were instructed to respond as they would normally, in order to minimise the confounding effect of current mood state. The main clinical diagnoses (DSM-IV Axis I) were depressive disorder (66.7%), anxiety disorder (16.8%), and other diagnoses (15.5%). 2.1. Instruments 2.1.1. Diagnostic Interview for Depressive Personality The DIDP [17] is a 26-item, semi-structured interview based on past theoretical descriptions of DPD and potential DPD criteria that were to be considered for inclusion in DSM-IV. The interview is divided into four major dimensions or subscales: negativistic, introverted/tense, passive/unassertive, and self-sacrificing. Examiners score each question as follows: 0, trait not present; 1, trait possibly (moderately, sometimes) present; 2, trait present.
Gunderson et al. state that a score of 37 or higher is suggestive of a depressive personality. These authors recorded an inter-rater reliability (k) of 0.67 and a test–retest reliability (k) of 0.41, initially with a cutoff score of 42 [17], but later with a cutoff score of 37 [23]. The first interview comprised 32 items [17], although this was subsequently reduced to 30 [48], and in the end we used the 26item Spanish version. After 2 years, this version, which was used with the authors’ permission [17], had a test–retest reliability
583
(with one-third of the sample [10 patients]) of 0.67 and an intraclass correlation coefficient (ICC) of 0.53. All interviewers were blind to the DIDP scores of the participants. 2.1.2. Structured Clinical Interview for DSM-IV Axis II Disorders – SelfReport (SCID-II SR) SCID-II SR [16] is a 119-item, yes/no format questionnaire assessing the diagnostic criteria of the 11 DSM-IV personality disorders. Individuals obtain a score for each of the 11 personality disorder scales. Few reliability and validity data are available for the DSM-IV version of SCID-II SR (all ICC values indicated fair to good agreement; range: 0.49 to 0.86). However, Dreessen et al. [14] cite several studies with the DSM-III-R version of the instrument that demonstrated its reliability and validity and imply that similar reliability and validity might be expected with DSM-IV [15]. DPD was added to SCID-II, although no reliability or validity data are available. SCID-II SR was designed to assess the personality disorders presented in the main text of DSM-IV. Administration of the scale follows a two-tiered procedure. First, respondents complete a 119item self-reporting questionnaire using a yes/no response format. Each of the questions corresponds to a diagnostic criterion for each of the personality disorders or the two additional personality disorders listed in Appendix B of DSM-IV (i.e. passive–aggressive and DPD). After the respondents have completed the questionnaire, the interviewer identifies those personality disorders for which the respondents fulfilled sufficient criteria for diagnosis of a particular personality disorder. Respondents meeting self-reporting criteria for any given personality disorder are then administered those portions of the SCID-II clinical interview that correspond to the personality disorders affecting the patients. A formal diagnosis is then assigned. The Spanish version of the SCID-II used in this study was adapted in 1999 (SCID-II-1999) [54]. 2.1.3. Hamilton Depression Scale (17 items) (HAMD-17) This scale was used to evaluate the severity of depressive symptoms [18]. No exclusion criteria were applied to patients in a depressive state, since personality was evaluated over several years and outside the depressive state. 2.1.4. Global Assessment of Functioning (GAF) This is a commonly used clinician-rated single-item instrument ranging from 1 to 100 and indicating symptom severity and level of functioning. 2.1.5. Sociodemographic and clinical questionnaire This questionnaire was used to obtain information on the following clinical variables: family history of mental disorders, personal history of physical and mental disorders, psychiatric admissions, current psychiatric diagnosis, and suicide attempts. 2.2. Translation and back-translation of the Diagnostic Interview for Depressive Personality The DIDP was translated into Spanish by two independent translators, who later reached agreement on a harmonised version. This version was back-translated by a psychiatrist under blind conditions. The English back-translation was then reviewed by the author (Gunderson) and the appropriate modifications were made by the Spanish team for the final version. The version of the DIDP that we finally used was similar to the previous Spanish adaptation of the questionnaire. It comprised 26 major dimensions and 63
584
L.J. Irastorza et al. / European Psychiatry 27 (2012) 582–590
items, with a cutoff point of 32 for the diagnosis of DPD. The differences in the Spanish version that we used compared with Gunderson’s initial DIDP [17] are as follows: negativistic: the asthenic dimension disappears and the selfcriticism dimension is changed to self-sacrificing. The number of items in this subscale decreases from 11 to 9; introverted: the unsociable or poorly sociable item disappears, leaving six items instead of the original seven items; passive/unassertive: the original eight items are reduced to six by eliminating psychomotor inertia and low sex drive; self-sacrificing in place of masochistic: the original four items increase by one item (self-criticism) that was in the negativistic subscale.
2.3. Visit procedure A pilot test was performed with the resulting version. Using the draft of the questionnaire, we evaluated 10 individuals to ensure their similarity with the individuals in the study sample. This pilot test allowed us to identify the most appropriate types of questions and ensure that the phrasing was correct and understandable and that the questions were of an adequate length. At the baseline visit, the HAMD-17 was administered, along with the sociodemographic and clinical questionnaire, SCID-II, DIDP, and GAF. The test–retest evaluation was done in 36 patients, 15–20 days after the baseline visit by administering the DIDP. Inter-rater reliability was evaluated in turn by administering the DIDP to a total of 20 patients, 30 days after the baseline visit. 2.4. Statistical analysis A descriptive analysis was done of sociodemographic variables and history of interest – as well as of the results of the items on the diagnostic scale – for quantitative variables, or using percentages in the case of categorical variables. The associations between quantitative and qualitative variables in the personality disorder interviews (DIDP and SCID-II, respectively) were determined using the Chi-square test for a linear trend. The postulates for principal components analysis and confirmatory factor analysis were assessed (Kaiser-Meyer-Olkin [KMO] sampling adequacy and Bartlett’s sphericity test) to analyse the validity of the measurement instrument construct. An oblimin rotation was performed on the items to increase discrimination of the resulting dimensions, with an eigenvalue > 1 and inspection of the screen plot of eigenvalues. Reliability was evaluated using the k statistic. As a rule of thumb, a k of 0.75 or greater is considered to reflect excellent agreement, 0.60 to 0.74 indicates good agreement, 0.40 to 0.59 reflects fair agreement, and under 0.40 indicates poor agreement [10]. The internal consistency of the questionnaire was evaluated using Cronbach’s a for the questionnaire as a whole and for each of the factors within the different subscales. In all cases, the correlation between item and total widely exceeded the minimum value of 0.30, and the elimination of an item did not increase the total Cronbach’s a value of the DIDP scale. Inter-rater reliability and test–retest variability was evaluated using the k statistic and the ICC. In order to obtain the validity of the criterion, we evaluated the relationship between the score for each individual and a gold standard – the SCID-II for DSM-IV and DPD – to ensure that we were measuring what we wanted to measure. The area under the ROC curve is the best indicator of the predictive capacity of the DIDP, independently of the prevalence of
DPD in the reference population, and can be used to establish comparisons between different diagnostic tests (ROC, 0.848 [95% CI, 0.807–0.889]). It enables the largest number of patients to be classified correctly, according to the adopted cutoff point. Descriptive statistics for the DIDP were obtained for both samples and compared using a Student’s t test. We also calculated the sensitivity and specificity of different cutoff values of the DIDP in detecting DPD. An overall analysis was performed using SPSS version 15 with a 95% confidence interval. A two-tailed P value of 0.05 was considered to be statistically significant. 3. Results Of the 328 patients enrolled in the study, 212 patients were married (65%), 172 were employed (52%), 68 were homemakers (21%), 52 were on sick leave (16%), 25 lived with family (7.62%), and 188 had no medical history of interest (57%). Current and past clinical diagnoses are reported in Table 1. All personality disorders were represented, with a mean of 3.1 2.4 disorders per participant: 77.1% had group C, 56.3% had group B, and 38.2% had group A personality disorders. A good correlation was observed, with a test–retest reliability of 0.810 for 36 patients (Table 2). The dimensions with low item-total scale reliability and no significance were melancholic, negative reactivity, overly dependent, tendency to worry, hypersensitive to rejection, and self-sacrificing. A good intraclass inter-rater correlation was also recorded (ICC = 0.821; range: 0.602–0.925) in 20 patients. The internal consistency of the subscales yielded an internal reliability (Cronbach’s a) for the DIDP of 0.854. The first three showed good reliability (Table 3), as follows: negativistic (0.796), introverted (0.783), and self-sacrificing (0.567). Cronbach’s a was lower for the passive subscale (0.409), with a low item-total scale reliability for the passive (0.297), difficulty in becoming angry
Table 1 Axis I and II diagnoses for the 328 patients.
Axis I diagnosis No psychiatric disorders Affective disorders only Anxiety disorders only Affective and anxiety disorders Substance use disorders only Affective and/or anxiety and substance use disorders Other Axis I disorders Axis II diagnosis No PD Paranoid PD Schizoid PD Schizotypal PD Histrionic PD Narcissistic PD Antisocial PD Avoidant PD Dependent PD Obsessive–compulsive PD Passive–aggressive PD Depressive PD Non-specified PD
Current diagnosis
Lifetime diagnosis
0 125 55 94 4 10
97 105 46 49 6 7
(0) (38.4) (16.9) (28.9) (1.2) (3.1)
3 (0.9)
34 86 35 52 7 72 14 122 56 171 83 154 18
(29.7) (32) (14.1) (14.9) (1.8) (2.1)
6 (1.8)
(10.4) (26.2) (10.7) (15.9) (2.1) (22) (4.3) (37.2) (17.1) (52.1) (25.3) (47.1) (5.5)
Values are expressed as absolute frequencies (percentage). PD: personality disorder. The ‘‘Other Axis I disorders’’ category includes Axis I diagnoses of eating disorders, adaptive disorders, and other Axis I diagnoses not already covered. Diagnoses can total more than 100%, as some patients receive more than one diagnosis of PD. The PDs are evaluated using the DSM-IV.
L.J. Irastorza et al. / European Psychiatry 27 (2012) 582–590
ROC curve
Table 2 Test–retest reliability (kappa coefficient) for DIDP dimensions. Test–retest (n: 36)
Melancholic Pessimistic Negative reactivity Bitter Low self-esteem Overly dependent Tendency to worry Oral Hypersensitive to rejection Introverted Quiet Serious Restricted Tense Limited pleasure capacity Passive Difficulty in becoming angry Critical of others Underachiever Counter-dependent Moralistic Self-sacrificing Burdened Self-critical Remorseful Unassertive
0.191 0.465 0.080 0.361 0.477 0.178 0.193 0.600 0.154 0.391 0.594 0.441 0.251 0.602 0.650 0.289 0.239 0.485 0.457 0.438 0.381 0.206 0.355 0.257 0.570 0.324
DIDP: Diagnostic Interview for Depressive Personality. All kappa coefficients that appear in boldface are significant at P < 0.05.
Table 3 Cronbach’s a and item-total correlation of DIDP dimensions. Corrected item-total correlation
Cronbach’s a if item is eliminated
Negativistic scale (Cronbach’s a = 0.796) Pessimistic Melancholic Tendency to worry Low self-esteem Negative reactivity Hypersensitive to rejection Bitter Overly dependent Oral
0.631 0.596 0.539 0.535 0.521 0.427 0.396 0.392 0.303
0.755 0.760 0.769 0.769 0.771 0.783 0.787 0.788 0.796
Introverted scale (Cronbach’s a = 0.783) Quiet Serious Tense Introverted, inhibited Limited pleasure capacity Restricted
0.649 0.644 0.552 0.543 0.456 0.413
0.729 0.722 0.748 0.748 0.772 0.783
Self-sacrificing scale (Cronbach’s a = 0.567) Self-critical Burdened Remorseful Self-denying Counter-dependent
0.398 0.349 0.337 0.318 0.237
0.474 0.498 0.505 0.516 0.560
Passive scale (Cronbach’s a = 0.409) Unassertive Passive Underachiever Difficulty in becoming angry Moralistic Critical of others
0.377 0.297 0.223 0.142 0.120 0.038
0.235 0.304 0.348 0.396 0.410 0.462
Dimension
100% 80%
Sensitivity
Dimension
585
60% 40% 20% 0% 0%
20%
40%
60%
80%
100%
1 - Specificity Fig. 1. ROC: receiving operating characteristic; DIDP: Dignostic Interview for Depressive Personality.
(0.142), critical of others (0.038), underachiever (0.223), and moralistic (0.120) dimensions. 3.1. Receiver operating characteristic curve The area under the curve was 0.848 (95% CI, 0.807, 0.889) (Fig. 1). The resulting sensitivity was 74.5%, with a specificity of 78.5%, positive predictive value of 75.5% (scoring at or above the threshold for the diagnosis of DPD and having it), negative predictive value of 77.6% (scoring below the threshold for being labeled with DPD and not having it), and efficiency of 76.6%, with a cutoff score of 30. Thirty-seven out of 151 patients passed the cutoff point and did not present DPD according to SCID-II, while 39 out of 154 patients presented DPD according to SCID-II but did not pass the cutoff point in the DIDP. We try to improve these results, for instance, by reducing this sample. We removed those depressive patients with a score higher than 20 in the Hamilton Scale depression. But the obtained results did not differ significantly from the initial simple: area under ROC curve 0.858, sensitivity 75.9, specificity 79.5, and a cutoff point of 28. So, we go on with the original sample. 3.2. Validity
DIDP: Diagnostic Interview for Depressive Personality.
The association between DPD according to the DIDP and DPD according to the SCID-II showed convergent validity in the broadest sense and concurrent validity in the narrowest sense (i.e. concurrent validity, because DPD in both DIDP and SCID-II purportedly measures the depressive personality construct). 3.2.1. Convergent-divergent validity The diagnostic convergency between the SCID-II for DPD and the DIDP illustrates their convergent validity. We observed that 47.1% of the sample had DPD according to the SCID-II and that 46.5% had DPD according to the DIDP. When both instruments were considered together, DPD affected 35.1% of the sample. We reported the association between DIDP scores and DPD symptoms according to the SCID-II and the association between DPD and symptoms of the avoidant and borderline personality disorders as assessed by the SCID-II (Table 4). This type of association matrix shows that the DIDP is more strongly related to
L.J. Irastorza et al. / European Psychiatry 27 (2012) 582–590
586
Table 4 Association between dimensions of positive DIDP and DPD (DIDP 30; SCID-II for DSM-IV), avoidant and borderline personality disorder (SCID-II). Dimension
DIDP
x Melancholic Pessimistic Negative reactivity Bitter Remorse Low self-esteem Worry Burdened Critical of others Self-critical Introverted Quiet Serious Restricted Tense Limited capacity for pleasure Unassertive Passive Overly dependent Difficulty in becoming angry Hypersensitive rejection Oral Counter-dependent Moralistic Self-sacrificing Underachiever
2
SCID n (%)
b
89.0 89.5b 74.8b 40.7b 37.2b 68.9b 65.0b 29.3b 17.0b 44.3b 47.7b 53.8b 54.6b 57.9b 67.2b 94.7b 76.3b 44.7b 24.8b 9.2a 43.9b 29.0b 20.3b 12.8a 57.2b 35.8b
114 (72.2) 98 (73.7) 108 (72.5) 36 (73.5) 86 (66.7) 113 (68.5) 133 (61.3) 98 (62.0) 71 (60.2) 125 (59.2) 70 (72.2) 18 (69.2) 55 (73.3) 76 (75.2) 19 (73.1) 77 (83.7) 87 (75.0) 86 (69.9) 65 (66.3) 51 (58.0) 124 (60.8) 14 (77.8) 101 (55.5) 57 (60.0) 107 (66.0) 64 (68.8)
x
2
Avoidant PD
x
n (%) b
56.8 71.0b 44.4b 34.5b 53.8b 71.7b 50.6b 28.8b 23.9b 46.8b 18.4b 13.0a 20.7b 17.9b 21.1b 31.9b 26.2b 24.2b 29.5b 5.3 33.8b 33.8b 8.5 2.0 20.7b 12.7b
109 99 98 38 92 114 130 99 76 129 59 12 49 64 19 61 75 79 67 48 121 9 97 49 92 54
(68.1) (73.9) (65.8) (77.6) (70.8) (68.7) (59.9) (62.3) (64.4) (60.8) (60.8) (48) (65.3) (63.4) (73.1) (66.3) (64.7) (64.8) (68.4) (55.2) (59.6) (50) (53) (52.1) (57.1) (57.4)
2 b
38.0 40.8b 17.7b 22.8b 18.7b 25.7b 17.9b 7.1 4.3 22.8b 52.4b 29.0b 35.9b 45.0b 35.6b 46.5b 39.3b 25.9b 7.9 16.1b 24.4b 18.6b 12.9a 0.4 20.5b 18.3b
Borderline PD n (%)
x2
n (%)
86 76 73 28 67 83 96 71 49 99 58 12 41 62 17 57 64 67 45 95 95 6 83 36 74 49
19.0b 26.3b 17.3b 9.9a 25.1b 33.4b 16.2b 8.3 1.0 17.1b 7.0a 4.6 3.0 4.0 4.8 6.0 2.7 10.7a 21.2b 4.6 50.4b 25.4b 3.0 3.4 14.1b 8.5
97 84 93 28 85 105 122 92 63 122 53 11 40 58 16 55 65 74 66 44 132 10 99 49 95 52
(53.8) (56.7) (49.0) (57.1) (51.5) (50.0) (44.2) (44.7) (41.5) (46.7) (59.8) (48.0) (54.7) (61.4) (65.4) (62.0) (55.2) (54.9) (45.9) (46.8) (46.8) (33.3) (45.4) (38.3) (46.0) (52.1)
(60.6) (62.7) (62.4) (57.1) (65.4) (63.3) (56.2) (57.9) (53.4) (57.5) (54.6) (44) (53.3) (57.4) (61.5) (59.8) (56) (60.7) (67.3) (50.6) (65) (55.6) (54.1) (52.1) (59) (55.3)
PD: personality disorder; DIDP: Diagnostic Interview for Depressive Personality; DPD: depressive personality disorder. a P is significant at 0.01 (two-tailed). b P is significant at 0.001 (two-tailed).
DPD symptoms than to symptoms of avoidant personality disorders (APD) and to borderline personality disorders (BPD) (i.e. convergent-divergent validity). The counter-dependent, difficulty in becoming angry, moralistic, and underachiever dimensions showed the lowest association between the DIDP and DPD. Discriminant validity was also established with APD. It could be postulated that APD is more introverted, with limited capacity for pleasure, restricted, and pessimistic, and less burdened, critical of others, moralistic, and overly dependent. As for BPD, we observed more dimensions in the negativistic subscale, fewer dimensions in the introverted subscale, and high values for hypersensitivity to rejection and overly dependent. 3.2.2. Discriminant validity Patients who had DPD according to the DIDP were compared with those who did not have DPD (Table 5), taking into account variables such as comorbidity with other personality disorders (as diagnosed by the SCID-II for DSM-IV), the HAMD-17 score, and the GAF functional capacity score. There were no differences in comorbidity with the histrionic, narcissistic, antisocial, and nonspecified personality disorders. APD and BPD are the personality disorders that are most closely associated with DPD. Functional capacity (GAF) decreased, HAMD-17 increased (without reaching the diagnostic threshold of 17), and an association was observed with a greater number of personality disorders in the positive group according to the DIDP. Because no statistically significant differences were recorded for sociodemographic and clinical variables (age, sex, marital status, profession, cohabitation, medical history, previous psychiatric admissions, current psychiatric diagnosis according to DSM-IV, or psychiatric history), these were not included in the scale. A tendency towards significant differences was observed for family psychiatric history (x2 = 14.71, P < 0.056) in 58.2% versus 41.8% of patients, particularly in depressive disorders (62% versus 38%).
3.3. Factor analysis Prior to the factor analysis, the Kaiser-Meyer-Olkin test was applied (KMO = 0.842), along with Bartlett’s sphericity test (x2 = 2226; P < 0.001), which indicated the adequacy of the data for applying this technique. The results of the confirmatory factor Table 5 Comparison between patients with depressive DIDP (Gunderson, 1994) ( 30) and without depressive DIDP (< 30) with PD (SCID-II).
Avoidant PD Dependent PD Obsessive-Compulsive PD Passive-aggressive PD Depressive PD Paranoid PD Schizotypal PD Schizoid PD Histrionic PD Narcissistic PD Borderline PD Antisocial PD Non-specified PD
GAFa Hamilton depressiona
DIDP < 30, n (%)
DIDP 30, n (%)
x2
29 19 74 25 39 32 17 11 6 33 64 7 13
92 37 97 58 114 54 35 23 1 39 98 7 5
67.6d 10.4c 15.2d 24.5d 91.1d 12.5d 10.8c 6.9d 3.0 2.2 25.5d 0.07 2.67
(16.7) (10.9) (42.5) (14.4) (22.4) (18.4) (9.8) (6.3) (3.4) (19) (36.8) (4) (7.5)
(60.9) (24.5) (64.2) (38.4) (75.5) (35.8) (23.2) (15.3) (0.7) (25.8) (64.9) (4.6) (3.3)
DIDP < 30
DIDP 30
60.1 (17.4) 12.1 (7.3)
50.9 (16.1) 15.2 (7.8)
Student’s t 4.91d 3.75d Mann-Whitney z
b
PD number
1 (2)
4 (3)
8.69d
PD: personality disorder; DIDP: Diagnostic Interview for Depressive Personality; GAF = Global Assessment of Functioning. The PDs were diagnosed using the SCID-II for DSM-IV. Hamilton Depression = 17item scale. a Mean (standard deviation). b Median (interquartile range). c P is significant at 0.01 (two-tailed). d P is significant at 0.001 (two-tailed).
L.J. Irastorza et al. / European Psychiatry 27 (2012) 582–590
587
Table 6 Principal-component analysis of DIDP traits (oblimin transformation including factors with an eigenvalue 1). Component Dimension
Factor 1 negativistic
Pessimistic Melancholic Negative reactivity Tendency to worry Low self-esteem Overly dependent Bitter Oral Hypersensitive to rejection Quiet Serious Introverted Tense Limited pleasure capacity Restricted Self-critical Burdened Remorseful Counter-dependent Self-sacrificing Passive Critical of others Difficulty in becoming angry Unassertive Underachiever Moralistic
0.741 0.663 0.677 0.567 0.493 0.564 0.519 0.446 0.392
Factor 2 introverted
Factor 3 self-sacrificing
Factor 4 passive
0.316 0.876 0.830 0.733 0.679 0.447 0.426 0.586 0.606 0.528 0.530 0.421
0.453 0.338
0.631 0.545 0.448 0.412 0.351 0.390
DIDP: Diagnostic Interview for Depressive Personality. Factor loadings with absolute values of <0.3 are omitted.
analysis (based on the principal component extraction method, with oblimin rotation and eigenvalues > 1) revealed the existence of four factors accounting for 43% of the variance: negativistic (22.7%), introverted (8.92%), passive/unassertive (6.35%), and selfsacrificing (5.24%). The factor loadings are reported in Table 6. The difficulty in becoming angry dimension is observed in the factor analysis in two factors: passive and self-sacrificing. However, we prefer to include it in the passive factor in this study, since Cronbach’s a is greater. It was also included in the original authors’ version [17] and in the Spanish version of the DIDP.
4. Discussion Based on an outpatient sample, and with a semi-structured interview (the SCID-II for DSM-IV) as the comparator, we validated the Spanish-language version of the DIDP developed by Gunderson et al. [17]. Inter-rater reliability (ICC, 0.821) was similar to that reported for the original scale. The result of our reliability test– retest was a k of 0.81 after three weeks, and this result is comparable to that of other studies (k = 0.60–0.74) [23,38]. Gunderson et al. [17] obtained a k of 0.41 and an ICC of 0.62 after 1 year. Other studies have evaluated the long-term stability of the DIDP and have reported a k value of 0.55 and an ICC value of 0.62 after 1 year [48] and an ICC of 0.47–0.53 after 2 years [24]. The internal consistency of the original authors’ DIDP was very high (Cronbach’s a, 0.93) [17] and similar to that recorded in our study (0.854). By applying the SCID-II, we observed a strong association between the DIDP and DPD in many dimensions, thus indicating good convergent validity. The quiet, counter-dependent, difficulty in becoming angry, and moralistic dimensions showed the lowest association between the DIDP and DPD. This is consistent in part with other studies, where the low self-esteem, counter-dependency, and self-criticism dimensions are seen to be predictive of DPD, but feeling burdened is not [19]. If only the SCID-II is used to evaluate
DPD, measures of anger and depression correlate with DPD, and state and trait anxiety with APD [25]. The self-criticism dimension was mostly associated with DPD and was independent of a depressed mood [19]. Similarly, Klein [34] found that patients with DPD were more likely to be selfcritical and harbour negative attributes than psychiatric controls. We found that this dimension fits better in the selfsacrificing fourth factor than in the negativistic first factor. We observed higher factor loading (0.586/0.202) than in the negativistic factor, and a higher Cronbach’s a (0.398/0.204); these results are similar to those of previous studies that include this dimension in the self-sacrificing subscale [26]. Hartlage et al. [19] found that psychiatric patients classified as having DPD were more self-critical, introverted, stress-reactive, negativistic, counter-dependent, and burdened than their psychiatric counterparts. Our study shows that the self-critical dimension is found in DPD, APD, and BPD. Using the NEO-PIR, Huprich [23] observed high levels of anxiety, depression, and self-consciousness, and low levels of tendermindedness in patients with DPD. These dimensions, compared with the DSM-IV and DSM-IV-TR as diagnostic classifications, were as follows: being dejected, gloomy, cheerless (i.e. depression facet), prone to worrying and brooding (i.e. anxiety facet), believing that they are inadequate and worthless, having low self-esteem, feeling much guilt towards themselves (i.e. self-consciousness facet), and being negative, critical, and judgmental towards others (i.e. low levels of the tender-mindedness facet) [26]. The DIDP does not include the oral, psychomotor inertia, or underachiever dimensions. However, oral is ultimately included in factor three (passive/unassertive) in the initial Spanish version of the DIDP and in our version, and underachiever is included in factor four (masochistic or self-sacrificing). We did not include psychomotor inertia, as did Gunderson et al. [17]; this is consistent with the suggestions of subsequent reports and with the final version harmonised with the main author of the DIDP [17]. The first Spanish-language version of the DIDP had already obviated
588
L.J. Irastorza et al. / European Psychiatry 27 (2012) 582–590
psychomotor inertia and low sex drive from the third factor, due to very low specificity with DPD. We observed a good degree of diagnostic convergence in the DIDP and SCID-II (47.1% of the sample were DPD-positive by SCIDII and 46.5% were DPD-positive by DIDP). When both instruments were considered together, DPD affected 35.1% of the sample, a higher percentage than found in other studies. Discriminant validity was also established with APD and BPD. It could be postulated that APD is more introverted, with limited capacity for pleasure, restricted and pessimistic, and less burdened, critical of others, moralistic, and overly dependent. Consistent with our results, a high degree of comorbidity of DPD with APD has been observed elsewhere [37,42,47], with an overlap in criteria sets [34,29,30]. In their description of DPD, Gunderson et al. [17] included hypersensitivity to rejection, which is also a feature of APD. Huprich [23] found that outpatients diagnosed with DPD felt more alienated and insecure in the context of interpersonal relationships and reported less support from family members. Such findings suggest that patients with DPD may have a heightened sensitivity to loss and feel detached. Although the fourfacet trait set fails to differentiate DPD from APD, the highest weight for DPD comes from the depression trait facet, whereas the highest weight for APD comes from the self-consciousness facet [4]. As for the dimensions of the DIPD in BPD, we observed more negativistic dimensions, hypersensitivity to rejection, overly dependent, less introverted, and difficulty in becoming angry, which is greater in APD and DPD. We decided to include it in the passive factor, where there is greater differentiation between BPD and DPD, although it could also be included in the self-sacrificing factor. We also observed comorbid conditions with personality disorders and DPD (according to the DIDP), especially in depressive, avoidant, borderline, and obsessive–compulsive personality disorders. We agree with other authors on the higher comorbidity of DPD with APD and BPD [34,42,24,41,45]. When we compared these three personality disorders dimensionally, we observed that they were all negativistic and hypersensitive to rejection; BPD is not introverted and is less self-sacrificing than the other two personality disorders. APD is less dependent, burdened, and critical of others than DPD. In BPD, mood is unstable, and the individual is affected by anger explosions. In DPD, anger is not usually expressed. Although patients with APD are timid and introverted, as are patients with DPD [24,41,45], patients with DPD less commonly reject contact with people, and thus they have more frequent and better relationships with others [3]. Due to their high sense of personal responsibility, patients with DPD try to relate to others, apparently with good results; however, they suffer tension and doubts about whether they deserve the love and friendship of others [3]. In addition, patients with APD are not chronically sad or unhappy. The validity of the DIDP construct can also be assessed by comparing patients presenting a positive threshold ( 30) with dysthymic patients. This is not the subject of the present study, although more data can be found in the literature [31,32,26]. In Irastorza, a longitudinal study differentiating dysthymia from DPD using the Spanish-language DIDP, comorbidity was found between the two disorders, more so with group C personality disorder in DPD. We also found predictors that were specific to DPD, namely, dependent, avoidant, self-defeating, and schizoid personality disorders. In contrast, dysthymia was more associated with the personality disorders of groups A and B of DSM-IV, and with predictors such as antisocial, passive-aggressive, and narcissistic personality disorders. The present study does not address the recent discussion as to whether DPD belongs to the dysthymia group [52,49–51] or not [34,37,20,22,48,36,42,25,41,3]. Other authors propose the need for
clearer differentiation between the two disorders or the development of an alternative classification model [56]. Factor analysis revealed the four factors of the original scale [17] which, when compared with our study, corresponded to the following factors and variances: depressive/negativistic (15.4% versus 22.7%), introversion/tense (11.8% versus 8.9%), passive/ unassertive (9.8% versus 6.35%), and masochistic (9.8%) or selfsacrificing (5.24%). Using a cutoff point of 30 or higher, we found 76.6% of the patients to be correctly classified, with a sensitivity of 74.5% and a specificity of 78.5%. This is lower than in the initial study [17], although in this case the sample was much smaller (cutoff point 42, sensitivity 87%, and specificity 83%). As in other studies [48], no significant differences were observed in the DIDP scores between patients with or without major depression episodes according to the HAMD-17. We recorded differences only in the non-depressive range (15.23 versus 12.08, respectively), i.e. for a HAMD of <17. We also found functional impairment to be significant in patients with DPD, and significant differences were observed between those with a DIDP score above and below the cutoff point (50.80 versus 60.12). This effect is also seen in other studies [20,48,36]. The association we observed between the DIDP and the DPD subscale of the SCID-II for DSM-IV was quite good (76.6%) – even though other authors report little association with DPD questionnaires – and is consistent with other studies using DPD questionnaires [24,44]. The present study has a number of limitations, including a possible influence of the depressive state on the evaluation. Some previous studies [5,21] have indicated that self-rating personality assessments can be affected by depressive state, even if the level of depression is minimal. Such a state of depression has been shown to significantly affect the harm avoidance (HA), self-direction, and cooperation scales of the TCI [12,9,8,33,57,6,59], thus raising an interesting question as to whether the state of depression can damage the proposed factor structure in the DIDP. Clinical state (i.e. depression) has been shown to affect the presentation of DPD [41,39]. However, most of the clinical studies on DPD have been conducted in patients with unipolar depression, while others have included patients with a variety of Axis I or II disorders [22,52,23,42,30,41,39]. We attempted to obviate this by asking patients about their character since adolescence, and not only within the context of a possible depressive state. A further limitation may be the high comorbidity of personality disorders. This could be due to the selection process, as we chose patients with specific diagnostic criteria and degrees of severity at each mental health centre. Despite its limitations, our study had a large sample of psychiatric outpatients who underwent semi-structured diagnostic interviews administered by trained interviewers. In addition, all the diagnostic criteria of DSM-IV Axis II disorders were assessed. A new approach might involve the development of a set of consensus prototypical ratings [40] to describe the main text personality disorders using facet traits from the five-factor model. The results of a study in which Q factor analysis and the Shedler-Westen Assessment Procedure 200 were applied [55,60] suggest that a depressive or dysphoric personality could represent an internalising spectrum of personality disorder [7]. Similar to the construct underlying DPD, the central features of this dysphoric factor [60] reflect a global internalising style, including depression, shame, guilt, self-blame, fear of rejection, anxiety, low assertiveness, and emptiness. On the basis of our results, we suggest that the fifth edition of the DSM, within the types of personality disorders, should include a supra-category comprising several personality disorders, as
L.J. Irastorza et al. / European Psychiatry 27 (2012) 582–590
suggested by Shedler and Westen [55], when speaking of depressive or dysphoric personality disorder. This supra-category could include dependent, avoidant, borderline (especially its affective subtype, if we can label it as such), and obsessive personality disorders. We observed greater factor loading and Cronbach’s a in negativistic and introverted in patients with DPD. Other authors report high levels of neuroticism [30,61], high negative affect [41,11], or high HA [1]. 5. Conclusion The DIDP proved to be a reliable and valid instrument for assessing DPD, at least among our outpatients. The psychometric properties of the DIDP support its clinical usefulness in assessing depressive personality. Conflict of interest statement There is no conflict of interest. Acknowledgements This study was supported by a grant from Glaxo-GSK and from the Instituto de Salud Carlos III, for Health Technologies and Evaluation and CIBERSAM. We are grateful to the Fundacio´n para la ˜ o´n (Madrid) Investigacio´n Biome´dica of Hospital Gregorio Maran for editorial assistance. References [1] Abrams KY, Yune SK, Kim SJ, Jeon HJ, Han SJ, Hwang J, et al. Trait and state aspects of harm avoidance and its implication for treatment in major depressive disorder, dysthymic disorder, and depressive personality disorder. Psychiatry Clin Neurosci 2004;58(3):240–8. [2] Akiskal HS. Dysthymic disorder: psychopathology of proposed chronic depressive subtypes. Am J Psychiatry 1983;140(1):11–20. [3] Akiskal HS. Validating affective personality types. In: Robins I, Barrett J, editors. The validity of psychiatric diagnosis. New York: Raven Press; 1989. [4] Bagby RM, Schuller DR, Marshall MB, Ryder AG. Depressive personality disorder: rates of comorbidity with personality disorders and relations to the fivefactor model of personality. J Personal Disord 2004;18(6):542–54. [5] Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961;4:561–71. [6] Black KJ, Sheline YI. Personality disorder scores improve with effective pharmacotherapy of depression. J Affect Disord 1997;43(1):11–8. [7] Bradley R, Shedler J, Westen D. Is the appendix a useful appendage? An empirical examination of depressive, passive-aggressive (negativistic), sadistic, and self-defeating personality disorders. J Pers Disord 2006;20(5):524–40. [8] Brown SL, Svrakic DM, Przybeck TR, Cloninger CR. The relationship of personality to mood and anxiety states: a dimensional approach. J Psychiatry Res 1992;26(3):197–211. [9] Chien AJ, Dunner DL. The Tridimensional Personality Questionnaire in depression: state versus trait issues. J Psychiatry Res 1996;30(1):21–7. [10] Cicchetti DV, Sparrow SA. Developing criteria for establishing interrater reliability of specific items: applications to assessment of adaptive behavior. Am J Ment Defic 1981;86(2):127–37. [11] Clark LA, Watson D. Personality, disorder, and personality disorder: towards a more rational conceptualization. J Personal Disord 1999;13(2):142–51. [12] Cloninger CR, Przybecl TR, Svrakic DMT, Wetzel RD. The Temperament and Character Inventory (TCI): a guide to its development and use. St Louis (Miss): Center for Psychology of Personality, Washington University; 1994. [13] Costa PT, McCrae RR. The NEO PI-R professional manual. Odessa, FL: Psychological Assessment Resources; 1992. [14] Dreessen L, Hildebrand M, Arntz A. Patient-informant concordance on the Structured Clinical Interview for DSM-III-R personality disorders (SCID-II). J Personal Disord 1998;12(2):149–61. [15] First MB, Spitzer RL, Gibbon M, Williams JB, Benjamin LS. The structured clinical interview for DSM-IV Axis II disorders. New York: Biometrics Department, New York State Psychiatric Institute; 1996. [16] First MB, Gibbon M, Spitzer RL, Williams JB, Benjamin LS. User’s guide for the structured clinical interview for DSM–IV axis II personality disorders. Washington, DC: American Psychiatric Press; 1997. [17] Gunderson JG, Phillips KA, Triebwasser J, Hirschfeld RM. The Diagnostic Interview for Depressive Personality. Am J Psychiatry 1994;151(9):1300–4. [18] Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry 1960;23:56–62.
589
[19] Hartlage S, Arduino K, Alloy LB. Depressive personality characteristics: state dependent concomitants of depressive disorder and traits independent of current depression. J Abnorm Psychol 1998;107(2):349–54. [20] Hirschfeld RM, Holzer 3rd CE. Depressive personality disorder: clinical implications. J Clin Psychiatry 1994;55(Suppl.):10–7. [21] Hisli N. The validity and reliability of the Beck Depression Inventory. Turk Psikol Derg 1989;7:3–13. [22] Huprich SK. Depressive personality and its relationship to depressed mood, interpersonal loss, negative parental perceptions, and perfectionism. J Nerv Ment Dis 2003;191(2):73–9. [23] Huprich SK. Evaluating facet-level predictions and construct validity of depressive personality disorder. J Personal Disord 2003;17(3):219–32. [24] Huprich SK. Convergent and discriminant validity of three measures of depressive personality disorder. J Personal Assess 2004;82(3):321–8. [25] Huprich SK. Differentiating avoidant and depressive personality disorders. J Personal Disord 2005;19(6):659–73. [26] Huprich SK. What should become of depressive personality disorder in DSMV? Harvard Rev Psychiatry 2009;17(1):41–59. [27] Huprich SK, Margrett J, Barthelemy KJ, Fine MA. The Depressive Personality Disorder Inventory: an initial examination of its psychometric properties. J Clin Psychol 1996;52(2):153–9. [28] Huprich SK, Sanford K, Smith M. Psychometric evaluation of the depressive personality disorder inventory. J Personal Disord 2002;16(3):255–69. [29] Huprich SK, Porcerelli J, Binienda J, Karana D. Functional health status and its relationship to depressive personality disorder, dysthymia, and major depression: preliminary findings. Depress Anxiety 2005;22(4):168–76. [30] Huprich SK, Zimmerman M, Chelminski I. Disentangling depressive personality disorder from avoidant, borderline, and obsessive-compulsive personality disorders. Compr Psychiatry 2006;47(4):298–306. [31] Irastorza LJ. Distimia y personalidad depresiva: diferenciacio´n clı´nica. Actas Esp Psiquiatr 2001;29(5):318–26. [32] Irastorza LJ. Trastorno de personalidad depresiva y distimia:diferenciacio´n clı´nica. Persona 2002;1:12–9 [Available from Instituto Argentino para el estudio de la personalidad, http://www.iaepd.com.ar.]. [33] Kleifield EI, Sunday S, Hurt S, Halmi KA. The effects of depression and treatment on the Tridimensional Personality Questionnaire. Biol Psychiatry 1994;36(1):68–70. [34] Klein DN. Depressive personality: reliability, validity, and relation to dysthymia. J Abnorm Psychol 1990;99(4):412–21. [35] Klein DN. Commentary on Ryder and Bagby’s Diagnostic viability of depressive personality disorder: theoretical and conceptual issues. J Personal Disord 1999;13:118–27. [36] Klein DN, Miller GA. Depressive personality in nonclinical subjects. Am J Psychiatry 1993;150(11):1718–24. [37] Klein DN, Shih JH. Depressive personality: associations with DSM-III-R mood and personality disorders and negative and positive affectivity, 30-month stability, and prediction of course of Axis I depressive disorders. J Abnorm Psychol 1998;107(2):319–27. [38] Kwon JS, Kim YM, Chang CG, Park BJ, Kim L, Yoon DJ, et al. Three-year follow-up of women with the sole diagnosis of depressive personality disorder: subsequent development of dysthymia and major depression. Am J Psychiatry 2000;157(12):1966–72. [39] Laptook RS, Klein DN, Dougherty LR. Ten-year stability of depressive personality disorder in depressed outpatients. Am J Psychiatry 2006;163(5):865–71. [40] Lynam DR, Widiger TA. Using the five-factor model to represent the DSM-IV personality disorders: an expert consensus approach. J Abnorm Psychol 2001;110(3):401–12. [41] Markowitz JC, Skodol AE, Petkova E, Xie H, Cheng J, Hellerstein DJ, et al. Longitudinal comparison of depressive personality disorder and dysthymic disorder. Compr Psychiatry 2005;46(4):239–45. [42] McDermut W, Zimmerman M, Chelminski I. The construct validity of depressive personality disorder. J Abnorm Psychol 2003;112(1):49–60. [43] McLean P, Woody S. Commentary of depressive personality disorder: a false start. In: Livesley WS, editor. The DSM-IV Personality Disorders. New York: Guilford Press; 1995. p. 303–11. [44] Miller JD, Tant A, Bagby RM. Depressive personality disorder: a comparison of three self-report measures. Assessment 2009. [45] Orstavik RE, Kendler KS, Czajkowski N, Tambs K, Reichborn-Kjennerud T. Genetic and environmental contributions to depressive personality disorder in a population-based sample of Norwegian twins. J Affect Disord 2007;99(1– 3):181–9. [46] Phillips KA, Gunderson JG, Hirschfeld RM, Smith LE. A review of the depressive personality. Am J Psychiatry 1990;147(7):830–7. [47] Phillips KA, Hirschfeld RM, Shea MT, Gunderson JG. Depressive personality disorder. In: Livesley WJ, editor. The DSM-IV personality disorders, 7. New York: Guilford Press; 1995. p. 287–302. [48] Phillips KA, Gunderson JG, Triebwasser J, Kimble CR, Faedda G, Lyoo IK, et al. Reliability and validity of depressive personality disorder. Am J Psychiatry 1998;155(8):1044–8. [49] Ryder AG, Bagby RM. Diagnostic viability of depressive personality disorder: theoretical and conceptual issues. J Personal Disord 1999;13(2):99–117 [discussion 8–27, 52–6]. [50] Ryder AG, Bagby RM, Dion KL. Chronic, low-grade depression in a nonclinical sample: depressive personality or dysthymia? J Personal Disord 2001;15(1):84–93.
590
L.J. Irastorza et al. / European Psychiatry 27 (2012) 582–590
[51] Ryder AG, Bagby RM, Schuller DR. The overlap of depressive personality disorder and dysthymia: a categorical problem with a dimensional solution. Harvard Rev Psychiatry 2002;10(6):337–52. [52] Ryder AG, Schuller DR, Bagby RM. Depressive personality and dysthymia: evaluating symptom and syndrome overlap. J Affect Disord 2006;91(2–3):217–27. [53] Schneider K. Psychopathic personalities. London: Cassell; 1958. [54] SCID-II. Cuestionario de Personalidad. Masson S.A; 1999. [55] Shedler J, Westen D. Refining personality disorder diagnosis: integrating science and practice. Am J Psychiatry 2004;161(8):1350–65. [56] Sprock J, Fredendall L. Comparison of prototypic cases of depressive personality disorder and dysthymic disorder. J Clin Psychol 2008;64(12):1293–317 [discussion 318–22]. [57] Strakowski SM, Faedda GL, Tohen M, Goodwin DC, Stoll AL. Possible affectivestate dependence of the Tridimensional Personality Questionnaire in firstepisode psychosis. Psychiatry Res 1992;41(3):215–26.
[58] Svrakic DM, Whitehead C, Przybeck TR, Cloninger CR. Differential diagnosis of personality disorders by the seven-factor model of temperament and character. Arch Gen Psychiatry 1993;50(12):991–9. [59] Tanaka E, Kijima N, Kitamura T. Correlations between the temperament and character inventory and the self-rating depression scale among Japanese students. Psychol Rep 1997;80(1):251–4. [60] Westen D, Shedler J. Revising and assessing axis II. Part I: developing a clinically and empirically valid assessment method. Am J Psychiatry 1999;156(2):258–72. [61] Widiger TA, Trull TJ, Clarkin JF, Sanderson C, Costa PT. A description of the DSM-IV personality disorders with the five factor model of personality. In: Costa PT, Widiger TA, editors. Personality disorders and the five-factor model of personality. 2nd ed., Washington, DC: American Psychological Association; 2002. p. 89–102.