The essential features of personality disorder in DSM-5: the relationship between criteria A and B.

ORIGINAL ARTICLE

The Essential Features of Personality Disorder in DSM-5 The Relationship Between Criteria A and B Annett G. Hentschel, PhD* and Ralf Pukrop, PhD, MDScÞ Abstract: The essential features of the general criteria for personality disorder in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), are based on impairments in self and interpersonal functioning (criterion A) and pathological personality traits (criterion B). The current study investigated the relationship between criteria A and B in a German psychiatric sample (N = 149). Criterion A was measured by the General Assessment of Personality Disorder (GAPD); criterion B, by the Dimensional Assessment of Personality Pathology (DAPP) and the Revised NEO Personality Inventory (NEO-PI-R). There was a significant relationship between the GAPD, the DAPP, and the NEO-PI-R. The DAPP and NEO-PI-R domains increased the predictive validity of the GAPD (by 7.5% and 14.6%, respectively). The GAPD increased the variance explained by the DAPP by 1.5% and by the NEO-PI-R by 6.5%. The results suggest a substantial relationship between criteria A and B. Criterion B shows incremental validity over criterion A but criterion A only in part over criterion B. Future research should investigate whether it is possible to assess functional impairment apart from personality traits. Key Words: DSM-5 personality disorders, essential features of personality disorder, trait models, GAPD, DAPP, NEO-PI-R (J Nerv Ment Dis 2014;202: 412Y418)

T

he Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), Personality and Personality Disorder Work Group developed a proposal for diagnosing personality disorders (PDs), which is included in section III of the DSM-5, as an alternative DSM-5 model for PDs (American Psychiatric Association [APA], 2013, p. 761). This proposal includes new general criteria for PD, of which criteria A and B are defined as ‘‘essential features of personality disorder’’ (APA, 2013, p. 761). The general criterion A refers to ‘‘moderate or greater impairment in personality (self/interpersonal) functioning’’ (APA, 2013, p. 761). Self-functioning is further subdivided into identity and self-directedness; interpersonal functioning is subdivided into empathy and intimacy. The general criterion B is again defined as an essential feature of PD and requires ‘‘one or more pathological personality traits’’ (APA, 2013, p. 761). Five broad trait domains have been included in the DSM-5 as follows: a) negative affectivity, b) detachment, c) antagonism, d) disinhibition versus compulsivity, and e) psychoticism (APA, 2013). These five domains can be further subdivided into 25 traits on a lower facet level. The DSM-5, however, gives no explanation on how these two ‘‘essential features’’ of PDs, operationalized by criteria A and B, are conceptually or empirically related to each other. The central underlying motive of our investigation was to understand and explore this relationship more precisely. On the one hand, criteria A and B are the essential criteria for PD. Neither functional impairment without pathological personality trait expression nor pathological personality

*Saxonian Hospital of Psychiatry, Grosschweidnitz, Germany; and †Department of Psychiatry, University of Cologne, Cologne, Germany. Send reprint requests to Annett G. Hentschel, PhD, SKH Grosschweidnitz, Tagesklinik WSW, Gauss-Str. 3, 02943 Weisswasser, Germany. E-mail: [email protected]. Copyright * 2014 by Lippincott Williams & Wilkins ISSN: 0022-3018/14/20205Y0412 DOI: 10.1097/NMD.0000000000000129

412

www.jonmd.com

trait expression without functional impairment would justify PD diagnosis. If both criteria have to be fulfilled for a PD diagnosis, then criteria A and B should co-occur. It would therefore be expected that a significant correlation between both criteria exists. The new core elements of personality functioning need to demonstrate associations with the full range of personality pathologies; otherwise, they would not be the core elements of PD. On the other hand, both criteria also assess separate aspects of PD. A factor analytic study by Berghuis et al. (2012a) supported a distinction between general personality functioning and personality style. The factors observed in the general PD model remained largely intact in a combined factor analysis with personality traits, resulting in a seven-factor model with three general PD factors and four personality trait factors. However, there was an overlap between the general PD facets and personality trait facets for extraversion and agreeableness. Extreme expressions of personality style or traits are not necessarily related to functional impairment and do not justify a diagnosis of PD (Livesley and Jang, 2005; Trull and Durrett, 2005; Wakefield, 2008). Thus, criterion A (impairments in personality functioning) is introduced in the alternative DSM-5 model for PDs to determine a pathological threshold for PD (Skodol, 2012). Therefore, criteria A and B should also cover different aspects of PD, rather than being totally redundant. The degree of overlap between personality impairment and personality traits was investigated in two studies. Parker et al. (2004) reported correlations between self-rated personality style scores and self-reported disordered functioning scores ranging between 0.21 and 0.80. The higher-order disordered functioning scores showed large correlations with 13 of 15 personality style scores. Berghuis et al. (2012b) reported correlations of the two scales of the General Assessment of Personality Disorder (GAPD; self-pathology and interpersonal pathology) with the Dimensional Assessment of Personality Pathology (DAPP), at the facet level, for a Dutch and a Canadian sample. There were at least moderate correlations of the GAPD with 16 of the 18 facet scales of the DAPP and large correlations with 11 of the 18. Fourteen of the 18 facet scales of the DAPP showed higher correlations for the self-pathology than for the interpersonal pathology scale. The highest correlations were found between the self-pathology scale and identity problems (0.88) and cognitive distortion (0.82). Interestingly, compulsivity showed only small, negative correlations with the scales of the GAPD. The following considerations for operationalizing the two essential features of PDs had to be taken into account.

Criterion A Before the DSM-5 was published, a self-report questionnaire, designed to measure self and interpersonal pathology similar to DSM-5 concepts, had been developed: the GAPD (Livesley, in press). The GAPD was translated from English into German and Dutch, and it has been shown that the GAPD differentiates well between patients with and without PDs in Canadian, Dutch, and German samples (Berghuis et al., 2012b; Hentschel and Livesley, 2013a; Morey et al., 2011; Verheul et al., 2008). Thus, a basic feature of a general PD criterion was met: to differentiate between PDs and other psychopathological states. Furthermore, Hentschel and Livesley (2013b) reported moderate to

The Journal of Nervous and Mental Disease

& Volume 202, Number 5, May 2014

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.



Essential Features of PD

TABLE 1. The DSM-5 Trait Model Domains With Corresponding Higher-Order Domains of the NEO-PI-R and the DAPP-BQ and One Lower-Order Facet of the DAPP-BQ DSM-5 Trait Model Domains

Negative affectivity Detachment Antagonism Disinhibition versus compulsivity Psychoticism

NEO-PI-R Higher-Order Domains

Neuroticism Low FFM extraversion Low FFM agreeableness Low FFM conscientiousness versus excessive FFM conscientiousness Openness to experience

high correlations between the GAPD scales and 9 of the 12 DSM-IV PDs, showing an association of self and interpersonal pathology with a wide range of personality pathologies, except for antisocial, obsessivecompulsive, and histrionic PDs.

Criterion B The DAPPYBasic Questionnaire (DAPP-BQ; Livesley and Jackson, 2009) was developed to measure the most fundamental traits of abnormal personality. It has been demonstrated that the DAPP-BQ is a valid and reliable measure for clinically relevant personality variants in various cultures (Gutie´rrez-Zotes et al., 2008; Maruta et al., 2006; Pukrop et al., 2009; Van Kampen, 2006). Factor analytic studies of the DAPP-BQ revealed a four-factor structure similar to personality models for normal-range personality, such as the Five Factor Model (FFM), with emotional dysregulation corresponding well to neuroticism, dissocial behavior to antagonism, inhibitedness to introversion, and compulsivity to conscientiousness (Widiger et al., 2009a). The relationships between the DAPP, the FFM, and DSM-5 domains are represented in Table 1 (Krueger et al., 2011; Thomas et al., in press). The fifth domain of the DSM-5 (psychoticism), or FFM (openness), corresponds to a facet scale of the DAPP-BQ: cognitive dysregulation (Widiger and Simonsen, 2007). Cognitive dysregulation includes depersonalization, derealization, schizotypal cognition psychosis, and brief stress psychosis as subtraits and general behaviors (Livesley and Jackson, 2009). Furthermore, the DAPP-BQ domain scales showed positive correlations for emotional dysregulation and dissocial behavior with a broad range of the DSM-IV PDs, as well as positive correlations for inhibitedness with schizoid and avoidant PDs and negative correlations for inhibitedness with histrionic and narcissistic PDs, and a positive correlation for compulsivity with obsessive-compulsive PD (Bagby et al., 2005; Bagge and Trull, 2003; Kushner et al., 2011; Pukrop et al., 2001; Simonsen and Simonsen, 2009). We also included a measure of normal personality in our study because abnormal personality could be viewed as representing extremes on normal personality dimensions (Clark, 2007; Widiger and Samuel, 2005). Two meta-analyses revealed that normal-range personality models, such as the FFM, show meaningful and unique relationships to DSM-IV PDs (Samuel and Widiger, 2008; Saulsman and Page, 2005). Personality disorders were usually associated with high neuroticism, high introversion, low agreeableness, and low conscientiousness. There were two exceptions to these patterns: histrionic PD was related to extraversion and obsessive-compulsive PD was related to high conscientiousness. The relationship between normal-range openness to experience and PD, especially schizotypal, is controversially debated (Watson et al., 2008; Widiger and Simonsen, 2005). Watson et al. (2013) reported small correlations between openness and psychoticism from the DSM-5 trait model. However, recent studies have shown modest to large factor loadings of openness and psychoticism (facet and domain scales) on the same factor (De Fruyt et al., 2013; Gore and Widiger, 2013; Thomas et al., 2013). Nevertheless, the FFM does not fully encompass the variation in pathological personalities (Krueger et al., 2011). The FFM fails to capture, for * 2014 Lippincott Williams & Wilkins

DAPP-BQ Higher-Order Domains

Emotional dysregulation Inhibitedness Dissocial behavior Compulsivity DAPP-BQ lower-order facet: cognitive dysregulation

instance, psychotic-like experiences, insecure attachment, and cognitive dysregulation (Krueger et al., 2011; Livesley and Jang, 2005). Markon et al. (2005) were able to demonstrate that normal and abnormal personality traits, according to several different instruments, can be integrated in a joint hierarchical structure. Samuel et al. (2010) reported a substantial overlap in coverage for the DAPP-BQ, the Schedule for Nonadaptive and Adaptive Personality (SNAP), and the Revised NEO Personality Inventory (NEO-PI-R). Those instruments measure similar underlying traits. However, their analysis showed that the NEO-PI-R provided more information at the lower (normal) personality trait range and the DAPP-BQ and the SNAP provided more information at the higher (abnormal) personality trait range. To our knowledge, Berghuis et al. (2012b) is the only study examining the relationship between the GAPD, measuring the essential features of PD, and personality traits. In contrast to Berghuis et al. (2012b), we were interested in the trait domains instead of the facet-level traits. Our study is based on a more recent version of the GAPD, with fewer items (85 instead of 142/144 items). Furthermore, we included the NEO-PI-R next to the DAPP in our correlational analyses, and additionally, we computed hierarchical regression analyses with the GAPD, the DAPP, and the NEO-PI-R to explore the incremental validity of each of the questionnaires. In the current study, criterion A (self and interpersonal functioning impairments) was operationalized by the GAPD. Criterion B (pathological personality trait domains) was measured by the DAPP and additionally by the NEO-PI-R. First, we expected all of the DSM-5 trait model domains represented by the DAPP or NEO-PI-R domain scales and the DAPP facet scale cognitive dysregulation to show a significant relationship to the GAPD. Second, we were interested in seeing whether the DAPP and the NEO-PI-R and the GAPD would capture different aspects of PD and show incremental validity.

METHODS Participants Psychiatric patients (N = 149) from two German rural psychiatric clinics were interviewed after obtaining informed consent. We included patients who had considerably remitted acute states and who were free from acute psychosis, acute substance intoxication or withdrawal, or organicity. The sample consisted of 100 women and 49 men ranging from 18 to 78 years old (mean, 41.0; SD, 13.1). Seventy-three were inpatients, 39 were outpatients, 30 patients were from day care facilities, and 7 were forensic patients. All of the patients had at least one axis I disorder, as assessed with the German version of the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I; First et al., 1996; Wittchen et al., 1997). Half of the sample (n = 75) was determined to have at least one PD diagnosis as measured by the Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SCID-II; First et al., 1997; Wittchen et al., 1997). The DSM-IV-TR PDs diagnosed in the sample were avoidant PD (49.3%, n = 37), depressive PD (32.0%, n = 24), borderline PD (24.0%, n = 18), paranoid PD (20.0%, n = 15), www.jonmd.com


413

Hentschel and Pukrop


obsessive-compulsive PD (18.7%, n = 14), PD not otherwise specified (16.0%, n = 12), dependent PD (13.3%, n = 10), schizoid PD (10.7%, n = 8), antisocial PD (9.3%, n = 7), narcissistic PD (2.7%, n = 2), passive-aggressive PD (2.7%, n = 2), histrionic PD (1.3%, n = 1), and schizotypal PD (1.3%, n = 1).

Instruments The GAPD The GAPD (Livesley, in press) is a self-report questionnaire that is used to determine the general presence or absence of a PD diagnosis. It is based on the core features of PD as suggested by Livesley (1998): self-pathology and interpersonal pathology. These concepts are very similar to criterion A of the general criteria for PD of DSM-5 (section III). Both GAPD and criterion A of the general criteria for PD of DSM5 are based on the PD definition of Livesley (Livesley, 1998; Skodol, 2012). The self-pathology scale is conceptualized by problems involving a poorly defined self-construct, characterized by difficulty differentiating self-experiences from those of others, problems forming an integrated self-structure resulting in an unstable sense of self, and the consequences of this deficit expressed by a lack of authenticity, as well as problems with self-directedness involving a lack of autonomy and agency or difficulty setting and attaining rewarding goals. The interpersonal pathology scale is defined by problems with attachment, intimacy, and affiliation and problems with prosocial behavior and cooperativeness. For more information about the development and content of the GAPD scales, refer to Hentschel and Livesley (2013b). The GAPD consists of 85 items; the self-pathology scale includes 60 items and the interpersonal pathology scale includes 25 items. Each item is measured on a 1- to 5-point scale, from 1 (very unlike me) to 5 (very like me). In our study, the internal consistencies assessed by Cronbach’s > coefficients were 0.98 for the self-pathology scale (60 items), 0.85 for the interpersonal pathology scale (25 items), and 0.98 for the GAPD score (sum score of all 85 items). The data of the current study showed a one-factor structure for the GAPD (Hentschel and Livesley, 2013b). The GAPD score can be understood as a measure of severity of personality pathology in general.

The DAPP-BQ The DAPP (Livesley and Jackson, 2009) is a self-report questionnaire that measures traits delineating PD. The 290 items are rated on a 5-point Likert scale, ranging from 1 (very unlike me) to 5 (very like me). The questionnaire yields 18 factorially derived scales: affective lability, anxiousness, callousness, cognitive dysregulation, compulsivity, conduct problems, identity problems, insecure attachment, intimacy problems, narcissism, passive-oppositionality, rejection, restricted expression, self-harm, social avoidance, stimulus seeking, submissiveness, and suspiciousness. Each scale contains 16 items, except self-harm (12 items) and suspiciousness (14 items). The higher-order factor structure of the DAPP-BQ revealed four higher-order dimensions (emotional dysregulation, dissocial behavior, inhibitedness, and compulsivity; Bagge and Trull, 2003). The Cronbach’s > coefficients for the DAPP-BQ dimensions based on the present study range, for higher-order dimensions, were from 0.87 to 0.98. Test-retest reliability of the 18 dimensions during a 3-week period ranges from 0.84 to 0.93 (Livesley and Jackson, 2009). Research has confirmed the stability of the postulated four-factor structure across different cultures, such as Chinese, Dutch, German, Japanese, and Spanish, and in diverse clinical and healthy populations (Gutie´rrez-Zotes et al., 2008; Maruta et al., 2006; Pukrop et al., 2001; Van Kampen, 2006; Zheng et al., 2002).

The NEO-PI-R The NEO-PI-R (Costa and McCrae, 1992) is a self-report questionnaire with 240 items, which are answered on a 5-point Likert 414

www.jonmd.com


scale. It is designed to measure the FFM of personality, which is based on the lexicographic approach of personality. The five personality domain scales (neuroticism, extraversion, openness, agreeableness, and conscientiousness) are subdivided into six facet scales with eight items each. Internal consistency coefficients (Cronbach’s >) for domain scales of the German version have ranged from 0.87 to 0.94 (Ostendorf and Angleitner, 2004). Cronbach’s > coefficients for NEO-PI-R domain scales based on the present study range from 0.86 to 0.94. Retestreliability coefficients (1Y2 months) for domain scales of the German version of the NEO-PI-R range from 0.82 to 0.91 (Ostendorf and Angleitner, 2004).

The SCID-II The SCID-II (First et al., 1997; Wittchen et al., 1997) is a wellestablished semistructured interview for diagnosing DSM-IV PDs. The interrater reliability varied from adequate to excellent depending on the study method (Lobbesteal, 2011). Both the SCID-II screening questionnaire and the interview were conducted with all patients. The SCID-II interview yielded the number and the kind of PD diagnosis and a dimensional score for each PD. Each PD criterion was rated as absent = 1, subthreshold = 2, or true = 3. SCID-II dimensional scores are computed by summing the results of each criterion for each PD. The SCID-II interviews were conducted by the first author, who is clinically experienced and well trained in administering the SCID-II.

Statistical Analyses First, Pearson’s product-moment correlation coefficients were computed for the GAPD, DAPP higher-order domains, and NEO-PI-R higher-order domains. To map the five trait domains of the DSM-5 trait model, we included the DAPP facet scale cognitive dysfunction as corresponding to the psychoticism domain. The scale of the GAPD included in the calculation was a GAPD score (sum score of all items) because of the one-factor structure obtained from the sample presented here (Hentschel and Livesley, 2013b). Berghuis et al. (2012b) had obtained a two-factor structure of a former version of the GAPD (142 items) and reported their results separated into self-pathology and interpersonal pathology. To maximize comparability, the self-pathology and interpersonal pathology subscales were additionally included into the correlation analysis. We followed the guidelines for interpretation of correlation coefficients provided by Cohen (1988). Second, we conducted a hierarchical regression analysis to calculate the amount of variance explained by the GAPD, DAPP, and NEO-PI-R domain scales, using the SCID-II dimensional total scores as the dependent measure. The GAPD score was entered as the predictor, in a block, followed by the DAPP domain scales (model 1), and separately by the NEO-PI-R domain scales (model 2). Next, the DAPP domain scales were entered as a block followed by the GAPD score (model 3). The NEO-PI-R domain scales were also entered separately as a block followed by the GAPD score (model 4). Furthermore, we conducted a hierarchical regression analysis to determine the amount of variance explained by the DAPP and NEO-PI-R domain scales, whereas the GAPD score served as the dependent measure. The DAPP domain scales were entered as the predictor, in a block, followed by the NEO-PI-R domain scales (model 5), and vice versa (model 6).

RESULTS Convergent and Discriminant Validity Table 2 shows the Pearson’s correlation coefficients between the NEO-PI-R and DAPP domains and the GAPD scales (GAPD score, self-pathology, and interpersonal pathology scale). All domain traits of the DAPP and the NEO-PI-R were substantially correlated with the GAPD score, except the DAPP domain * 2014 Lippincott Williams & Wilkins





TABLE 2. Pearson’s Correlation Coefficients Between DAPP-BQ Domains, NEO-PI-R Domains, and GAPD Scales DAPP-BQ ED

DAPP-BQ facet DAPP-BQ domains

NEO-PI-R domains

GAPD-scales

DB

IH

Cognitive 0.90** 0.44** 0.68** dysregulation ED 0.42** 0.72** DB 0.16 IH CO N E O A C S P

NEO-PI-R CO

0.10

N

E

O

0.74** j0.46** j0.03

0.17* 0.86** j0.54** j0.09 0.01 0.21** 0.18* 0.21** 0.21** 0.66** j0.76** j0.33** 0.14 j0.13 0.05 j0.62** j0.06 0.46**

GAPD A

C

j0.23** j0.62** j0.16* j0.68** j0.08 0.17* j0.13 j0.10 j0.06

S

0.87**

P

0.53**

GAPD score

0.85**

j0.65** 0.86** 0.54** 0.85** j0.37** 0.36** 0.42** 0.40** j0.45** 0.71** 0.64** 0.75** 0.47** 0.08 0.08 0.09 j0.58** 0.74** 0.44** 0.73** 0.37** j0.52** j0.53** j0.56** 0.17* j0.15 j0.25** j0.18* 0.34** j0.18* j0.46** j0.25** j0.63** j0.44** j0.63** 0.60** 0.98** 0.73**

N = 149. Correlation coefficients are based on GAPD, DAPP-BQ, and NEO-PI-R raw score. *Correlation is statistically significant at the 0.05 level. **Correlation is statistically significant at the 0.01 level. A indicates agreeableness; C, conscientiousness; CO, compulsivity; DB, dissocial behavior; E, extraversion; ED, emotional dysregulation; IH, inhibitedness; N, neuroticism; O, openness to experience; P, interpersonal pathology scale; S, self-pathology scale

compulsivity and the NEO-PI-R domain openness to experience. The highest correlations with the GAPD score were obtained for the DAPP domains emotional dysregulation and inhibitedness, the DAPP facet scale cognitive dysregulation, and the NEO-PI-R domain neuroticism. Most domain scores were more highly correlated with the selfpathology than with the interpersonal pathology subscale. The only clear exception was obtained by the NEO-PI-R domain agreeableness, which was significantly more highly correlated with the interpersonal pathology scale of the GAPD.

Incremental Validity The results of the hierarchical regression analyses are shown in Table 3. For models 1 through 4, the SCID-II dimensional total scores served as the dependent measure. The GAPD score explained 50.6% of the total variance. When the DAPP domain scales were entered in the second step, the variance increased significantly, by 7.5% (model 1). The separately entered NEO-PI-R domain scales (after entering the GAPD score) showed an increase in the variance by 14.6% (model 2). When the DAPP score was entered first, the DAPP explained 56.6% of the variance and the GAPD increased the variance, not significantly, by 1.5% (model 3). The NEO-PI-R explained 58.7% of the variance, and the GAPD increased the variance significantly, by 6.5% (model 4). In models 5 and 6, the GAPD score served as the dependent measure. In the first step, the DAPP domain scales accounted for 77.7% of the total variance and the NEO-PI-R domain scales accounted for 63.5% of the total variance, respectively. When the DAPP domain scales were entered after the NEO-PI-R domain scales, the DAPP scales increased the total variance by 15.2% (model 5), as can be seen in Table 3, whereas the NEO-PI-R scales improved the total variance, not significantly, by 1% (model 6).

DISCUSSION The results of the current study demonstrate a substantial relationship between criterion A, measured by the GAPD, and criterion B, measured by the DAPP and the NEO-PI-R, for all corresponding DSM-5 trait model domains. The domains DAPP emotional dysregulation/ * 2014 Lippincott Williams & Wilkins

NEO neuroticism, DAPP inhibitedness/NEO extraversion, NEO conscientiousness (but not DAPP compulsivity), and DAPP facet cognitive dysregulation (but not NEO openness) showed large correlations with the total score of the GAPD. The DAPP dissocial behavior and NEOPI-R agreeableness showed moderate and low correlations, respectively. The degree of overlap is comparable with the studies of Parker et al. (2004) and Berghuis et al. (2012b), which also showed large correlations between personality impairment and most of the personality style/trait scores. One exception needs to be discussed in more detail. The DAPP domain scale compulsivity showed no relationship to the GAPD scales, which is in accordance to the results of Berghuis et al. (2012b). The GAPD showed only a small, although significant, relationship to obsessive-compulsive PD (Hentschel and Livesley, 2013b). However, DAPP compulsivity itself showed significant relationships to obsessive-compulsive PD (e.g., Pukrop et al., 2009). Berghuis et al. (2012b) argue that the DAPP-BQ compulsivity domain might reflect a unidimensional construct that is specific to obsessive-compulsive PD but not shared with general personality pathology. Otherwise, the GAPD score showed a large correlation with NEO conscientiousness. Future studies will be necessary to clarify the relationship between criterion A and compulsivity, as well as obsessive-compulsive PD. Self and interpersonal pathology will not be useful as general criteria if they are not able to capture the full range of personality pathology. Furthermore, the self-pathology subscale of the GAPD had generally higher correlations with the DAPP and the NEO-PI-R than the interpersonal pathology subscale, similar to the correlational patterns of the total GAPD score. It needs to be considered, given the higher number of items on the self-pathology scale of the GAPD, that the self-pathology scale variance dominates the GAPD total score. However, the clear single-factor structure of the GAPD seems to justify the use of the total GAPD score as the most effective measure for determining severity of personality pathology in general. In addition to the substantial overlap of criteria A and B, we investigated the relative contribution of both criteria to PD. When SCID-II PD diagnoses (dimensional scores) were used as dependent measures, the GAPD explained 50.6% of PD scores; the DAPP, www.jonmd.com


415




TABLE 3. Hierarchical Regression Analysis Showing Incremental Variance of the GAPD Relative to the DAPP and the NEO-PI-R in the Prediction of DSM-IV Personality Disorder Severity as Measured by the SCID-II and Incremental Variance of the DAPP and the NEO-PI-R in the Prediction of Personality Dysfunction as Measured by the GAPD Model

1 2 3 4 5 6

Step

Predictors

R2

Adjusted R2

$ R2

F Change

Significance

1 2 1 2 1 2 1 2 1 2 1 2

GAPD ED, DB, IH, CO GAPD N, E, O, A, C ED, DB, IH, CO GAPD N, E, O, A, C GAPD ED, DB, IH, CO N, E, O, A, C N, E, O, A, C ED, DB, IH, CO

0.506* 0.581* 0.506* 0.652* 0.566* 0.581* 0.587* 0.652* 0.777* 0.786* 0.635* 0.786*

0.502 0.567 0.502 0.637 0.554 0.567 0.572 0.637 0.771 0.773 0.622 0.773

0.506 0.075 0.506 0.146 0.566 0.015 0.587 0.065 0.777 0.010 0.635 0.152

150.470 6.438 150.470 11.883 46.921 5.256 40.568 26.530 125.395 1.238 49.715 24.677

G0.001 G0.001 G0.001 G0.001 G0.001 0.023 G0.001 G0.001 G0.001 0.295 G0.001 G0.001

N = 149. *p G 0.001; model 1: df (step 1) = 1,147; df (step 2) = 4,143; model 2: df (step 1) = 1,147; df (step 2) = 5,142; model 3: df (step 1) = 4,144; df (step 2) = 5,143; model 4: df (step 1) = 5,143; df (step 2) = 6,142; model 5: df (step 1) = 4,144; df (step 2) = 5,139; model 6: df (step 1) = 5,143; df (step 2) = 4,139; dependent variable of models 1, 2, 3, and 4: SCID-II dimensional total score; dependent variable of models 5 and 6: GAPD score; predictors: GAPD score. NEO-PI-R-scales: N indicates neuroticism; E, extraversion; O, openness to experience; A, agreeableness; C, conscientiousness. DAPP-BQ-scales: ED, emotional dysregulation; DB, dissocial behavior; IH, inhibitedness; CO, compulsivity.

56.6%; and the NEO-PI-R, 58.7%. Trait models (criterion B) added significantly to functional impairments (criterion A; 7.5% by the DAPP and 14.6% by the NEO-PI-R). Functional impairments (criterion A) significantly added information to the NEO-PI-R (6.5%) but not significantly to the DAPP (1.5%). The higher amount of shared variance between the GAPD and the DAPP (77.7%), compared with the shared variance of the GAPD and the NEO-PI-R (63.5%), reflects the higher similarity between the DAPP and the GAPD in contrast to the GAPD and the NEO-PI-R. Thus, the DAPP already encompasses information that is supposed to be expressed separately in criteria A and B. For instance, one of the facet scales of the DAPP domain emotional dysregulation is identity problems, which is very similar to the self-pathology scales of the GAPD. There seems to be a higher distinction between these aspects in the GAPD and the NEO-PI-R. That distinction was intended by the DSM-5 work group of personality and PD. The GAPD and the NEO-PI-R showed a higher amount of explained variance (65.3%) than the GAPD and the DAPP (58.1%). What does a substantial relationship between criteria A and B imply for the classification of PD in the DSM-5? The trait system is supposed to describe the pathological features of a disordered personality. As stressed by some authors (Livesley and Jang, 2005; Wakefield, 2008), more extreme expression of a trait does not justify a diagnosis of PD. Therefore, criterion A was implemented to decide whether a person has, in fact, a PD (i.e., functionally impaired) while meeting criterion B (having pathological personality domain or facet trait expressions). Thus, on the one hand, a high content overlap is justified. On the other hand, criterion A should include separate informational gains, to make the decision clearer as to whether the trait profile implies functional impairment. This information does not have to be included in the trait profile, as it seems to be the case for the DAPP in our study. Furthermore, abnormal personality traits are usually favored for the description of personality pathology (Hopwood et al., 2012; Krueger and Eaton, 2010; Watson et al., 2008; Widiger and Simonsen, 2007). These abnormal trait models might interfere in the distinction between criteria A and B. Our results for the GAPD and the DAPP challenge the theory that personality impairment and abnormal personality traits are distinct. The 416

www.jonmd.com

GAPD had incremental validity over the NEO-PI-R, and the GAPD and the NEO-PI-R explained an even higher amount of variance (65.2%) in the prediction of DSM-IV PD severity than did the GAPD and the DAPP (58.1%). Future studies should focus on this question of separate informational gain between criteria A and B, especially by including the Personality Inventory for DSM-5 (Krueger et al., 2011) for measuring criterion B. There are limitations in our study. First of all, no formal assessment of interrater reliability was obtained for SCID-II interviews. Furthermore, we used a DAPP facet trait to resemble the psychoticism trait of the DSM-5 trait model. This is problematic for two reasons. First, this facet trait is part of a different domain scale in the DAPP. Second, as a facet, it does not cover the whole range of the psychoticism trait. Furthermore, the current study investigated the relationship between the essential features of PD, with traits at the domain level. A facet-level analysis would provide still more insight into the various associations. The current study used validated questionnaires to measure normal and abnormal personality traits. The DAPP and the NEO-PI-R correspond to the DSM-5 trait model but are not identical. Particularly, the psychoticism facet of the DSM-5 trait model was difficult to map, and the DAPP seems to also cover aspects of criteria A. We also acknowledge that the GAPD and criterion A are not identical; the GAPD does not capture the facet of empathy as part of interpersonal functioning. However, Morey et al. (2011) also used the GAPD, together with the 118-item Severity Indices of Personality Problems (Verheul et al., 2008), for a secondary data analysis, to validate impairments in self and interpersonal functioning as core dimensions of personality pathology for the DSM-5. To our knowledge, this was the only study of the general criteria for PD conducted by members of the DSM-5 work group of personality and PD. Finally, our study relies on self-report data for assessing self and interpersonal pathology. Self-report questionnaires show disadvantages in the diagnostics of PD, compared with semistructured interviews (Clark, 2007; Widiger and Samuel, 2009). To our knowledge, there are no interviews acquiring the DSM-5 essential features of PD. The development and validation of an interview would be a very valuable aim for future PD research. * 2014 Lippincott Williams & Wilkins




CONCLUSIONS The current study revealed a substantial relationship between criteria A and B of the general criteria for PD in the DSM-5. On the one hand, this relationship is desired because both criteria need to be fulfilled for a diagnosis of PD. On the other hand, both criteria also need to capture distinct information, to differentiate between personality functioning (criteria A) and personality style (criteria B). The distinction between these aspects could be shown only in part in our study. Future research should investigate whether functional impairment can be measured more distinctly from personality traits, as intended by the general criteria for PD of the alternative DSM-5 model for PDs. DISCLOSURE This report presents independent and nonfunded research. The views expressed are those of the authors. The authors declare no conflict of interest.

REFERENCES American Psychiatric Association (2013) Diagnostic and statistical manual of mental disorders (5th ed). Arlington, VA: American Psychiatric Publishing. Bagby RM, Marshall MB, Georgiades S (2005) Dimensional personality traits and the prediction of DSM-IV personality disorder counts in a nonclinical sample. J Pers Disord. 19:53Y67. Bagge CL, Trull TJ (2003) DAPP-BQ: Factor structure and relations to personality disorder symptoms in a non-clinical sample. J Pers Disord. 17:19Y32. Berghuis H, Kamphuis JH, Verheul R (2012a) Core features of personality disorder: Differentiating general personality dysfunctioning from personality traits. J Pers Disord. 26:704Y716. Berghuis H, Kamphuis JH, Verheul R, Larstone R, Livesley WJ (2012b) The General Assessment of Personality Disorder (GAPD) as an instrument for assessing the core features of personality disorder. Clin Psychol Psychother. 20:544Y557. Clark LA (2007) Assessment and diagnosis of personality disorder: Perennial issues and an emerging reconceptualization. Annu Rev Clin Psychol. 58:227Y257. Cohen J (1988) Statistical power analysis for the behavioral sciences (2nd ed). Hillsdale, NJ: Lawrence Erlbaum Associates Publishers. Costa PT, McCrae RR (1992) Revised NEO Personality Inventory (NEO PI-R) and NEO Five Factor Inventory. Odessa, FL: Psychological Assessment Resources. De Fruyt F, De Clercq B, De Bolle M, Wille B, Markon KE, Krueger RF (2013) General and maladaptive traits in a Five-Factor framework for DSM-5 in a university student sample. Assessment. 20:295Y307. First MB, Spitzer R, Gibbon M, Williams JBW (1996) Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I). Washington, DC: American Psychiatric Press. First MB, Gibbon M, Spitzer RL, Williams JBW, Benjamin LS (1997) Structured Clinical Interview for DSM-IVAxis II Personality Disorders (SCID-II). Washington, DC: American Psychiatric Press. Gore WL, Widiger TA (2013) The DSM-5 dimensional trait model and Five-Factor Models of general personality. J Abnorm Psychol. 122:816Y821. Gutie´rrez-Zotes JA, Gutie´rrez F, Valero J, Gallego Baille´s E, Torres X, Labad A, Livesley WJ (2008) Structure of personality pathology in normal and clinical samples: Spanish validation of the DAPP-BQ. J Pers Disord. 22:389Y404. Hentschel AG, Livesley WJ (2013a) Differentiating normal and disordered personality using the General Assessment of Personality Disorder (GAPD). J Pers Ment Health. 7:133Y142. Hentschel AG, Livesley WJ (2013b) The General Assessment of Personality Disorder (GAPD): Factor structure, incremental validity of self pathology and relations to DSM-IV personality disorders. J Pers Assess. 95:479Y485. Hopwood CJ, Thomas KM, Markon KE, Wright AGC, Krueger RF (2012) DSM-5 personality traits and DSM-IV personality disorders. J Abnorm Psychol. 121:424Y432. Krueger RF, Eaton NR (2010) Personality traits and the classification of mental disorders: Toward a more complete integration in DSM-V and an empirical model of psychopathology. Pers Disord. 1:97Y118. Krueger RF, Eaton NR, Derringer J, Markon KE, Watson D, Skodol AE (2011) Personality in DSM-5: Helping delineate personality disorder content and framing the metastructure. J Pers Assess. 93:325Y331.

* 2014 Lippincott Williams & Wilkins


Kushner SC, Quilty LC, Tackett JL, Bagby RM (2011) The hierarchical structure of the Dimensional Assessment of Personality Pathology (DAPP-BQ). J Pers Disord. 25:504Y516. Livesley WJ (1998) Suggestions for a framework for an empirically based classification of personality disorder. Can J Psychiatry. 43:137Y147. Livesley WJ (in press) Manual for the General Assessment of Personality Disorder (GAPD). Port Huron, MI: Sigma Assessment Systems. Livesley WJ, Jackson D (2009) Dimensional Assessment of Personality PathologyY Basic Questionnaire (DAPP.BQ)VTechnical manual. Port Huron, MI: Sigma Assessment Systems. Livesley WJ, Jang KL (2005) Differentiating normal, abnormal, and disordered personality. Eur J Pers. 19:257Y268. Lobbestael J, Leugarns M, Arntz A (2011) Inter-rater reliability of the Structured Clinical Interview for DSM-IV Axis I disorders (SCID I) and Axis II disorders (SCID II). Clin Psychol Psychot. 18:75Y79. Markon KE, Krueger RF, Watson D (2005) Delineating the structure of normal and abnormal personality: An integrative hierarchical approach. J Pers Soc Psychol. 88:139Y157. Maruta T, Yamate T, Iimori M, Kato M, Livesley WJ (2006) Factor structure of the Dimensional Assessment of Personality PathologyYBasic Questionnaire and its relationship with the revised NEO personality inventory in a Japanese sample. Compr Psychiatry. 47:528Y533. Morey LC, Berghuis H, Bender DS, Verheul R, Krueger RF, Skodol AE (2011) Toward a model for assessing level of personality functioning in DSM-5, part II: Empirical articulation of a core dimension of personality pathology. J Pers Assess. 93:347Y353. Ostendorf F, Angleitner A (2004) NEO-Perso¨nlichkeitsinventar nach Costa und McCrae, Revidierte Fassung (NEO-PI-R). Go¨ttingen, Germany: Hogrefe. Parker G, Hadzi-Pavlovic D, Both L, Kumar S, Wilhelm K, Olley A (2004) Measuring disordered personality functioning: To love and to work reprised. Acta Psychiatr Scand. 110:230Y239. Pukrop R, Gentil I, Steinbring I, Steinmeyer E (2001) Factorial structure of the German version of the Dimensional Assessment of Personality PathologyYBasic Questionnaire in clinical and nonclinical samples. J Pers Disord. 15:450Y456. Pukrop R, Steinbring I, Gentil I, Schulte C, Larstone R, Livesley WJ (2009) Clinical validity of the ‘‘Dimensional Assessment of Personality Pathology (DAPP)’’ for psychiatric patients with and without a personality disorder diagnosis. J Pers Disord. 23:572Y586. Samuel DB, Simms LJ, Clark LA, Livesley WJ, Widiger TA (2010) An item response theory integration of normal and abnormal personality scales. Pers Disord. 1:5Y21. Samuel DB, Widiger TA (2008) A meta-analytic review of the relationships between the Five-Factor Model and DSM-IV-TR personality disorders: A facet level analysis. Clin Psychol Rev. 28:1326Y1342. Saulsman LM, Page AC (2005) Corrigendum to ‘‘the Five-Factor Model and personality disorder empirical literature: A meta-analytic review.’’ Clin Psychol Rev. 25:383Y394. Simonsen E, Simonsen S (2009) The Danish DAPP-BQ: Reliability, factor structure, and convergence with SCID-II and IIP-C. J Pers Disord. 23:629Y646. Skodol AE (2012) Personality disorders in DSM-5. Annu Rev Clin Psychol. 8:317Y344. Thomas KM, Yalch MM, Krueger RF, Wright AGC, Markon KE, Hopwood CJ (2013) The convergent structure of DSM-5 personality trait facets and Five Factor Model trait domains. Assessment. 20:308Y311. Thomas KM, Yalch MM, Krueger RF, Wright AGC, Markon KE, Hopwood CJ (in press) The convergent structure of DSM-5 personality trait facets and Five Factor Model trait domains. Assessment. Trull TJ, Durrett CA (2005) Categorical and dimensional models of personality disorder. Annu Rev Clin Psychol. 1:355Y380. Van Kampen D (2006) The Dutch DAPP-BQ: Improvements, lower- and higherorder dimensions, and relationship with the 5DPT. J Pers Disord. 20:81Y101. Verheul R, Andrea H, Berghout C, Dolan CC, van Busschbach JJ, Van der Kroft PJA, Bateman AW, Fonagy P (2008) Severity Indices of Personality Problems (SIPP-118): Development, factor structure, reliability, and validity. Psychol Assess. 20:23Y34. Wakefield JC (2008) The perils of dimensionalization: Challenges in distinguishing negative traits from personality disorders. Psychiatr Clin North Am. 31:379Y393. Watson D, Clark LA, Chmielewski M (2008) Structures of personality and their relevance to psychopathology: II. Further articulation of a comprehensive unified trait structure. J Pers. 76:1545Y1586.

www.jonmd.com


417



Watson D, Stasik SM, Ro E, Clark LA (2013) Integrating normal and pathological personality: Relating the DSM-5 trait-dimensional model to general traits of personality. Assessment. 20:312Y326. Widiger TA, Clark LA, Livesley WJ (2009) An integrative dimensional classification of personality disorder. Psychol Assess. 21:243Y255. Widiger TA, Samuel DB (2005) Diagnostic categories or dimensions? A question for the Diagnostic and Statistical Manual of Mental DisordersYFifth Edition. J Abnorm Psychol. 114:494Y504. Widiger TA, Samuel DB (2009) Evidence-based assessment of personality disorders. Pers Disord. S3YS17.

418

www.jonmd.com


Widiger TA, Simonsen E (2005) Alternative dimensional models of personality disorder: Finding a common ground. J Pers Disord. 19:110Y130. Widiger TA, Simonsen E (2007) Alternative dimensional models of personality disorder. In Widiger TA, Sirovatka PJ, Simonsen E, Reiger DA (Eds), Dimensional models of personality disorders: Refining the research agenda for DSM-V (pp 1Y22). Washington, DC: American Psychiatric Publication Inc. Wittchen H-U, Wunderlich U, Gruschwitz S, Zaudig M (1997) Strukturiertes klinisches Interview fu¨r DSM-IV. Go¨ttingen, Germany: Hogrefe. Zheng W, Wang W, Huang Z, Sun C, Zhu J, Livesley WJ (2002) The structure of traits delineating personality disorder in a Chinese sample. J Pers Disord. 16:477Y486.

* 2014 Lippincott Williams & Wilkins


The relationship between borderline personality disorder and bipolar disorder.

The relationship between self-reported borderline personality features and prospective illness course in bipolar disorder.

The Relationship between Major Depressive Disorder and Personality Traits.

The Relationship between Survival Sex and Borderline Personality Disorder Symptoms in a High Risk Female Population.

Continuity between DSM-5 Categorical Criteria and Traits Criteria for Borderline Personality Disorder.

Psychopathy and the DSM-IV criteria for antisocial personality disorder.

Relationship between personality disorder functioning styles and the emotional states in bipolar I and II disorders.

Identifying criteria for narcissistic personality disorder.

The dimensionality of DSM5 alcohol use disorder in Puerto Rico.

The Relationship between Aggression and Serum Thyroid Hormone Level in Individuals Diagnosed with Antisocial Personality Disorder.

Relationship between the basic personality inventory and the NEO-Personality Inventory in a nonpatient sample.

Longitudinal Associations between Posttraumatic Stress Disorder Severity and Personality Disorder Features among Female Rape Survivors.

Alcohol consumption mediates the relationship between ADH1B and DSM-IV alcohol use disorder and criteria.

Dysfunctional responses to emotion mediate the cross-sectional relationship between rejection sensitivity and borderline personality features.

The impact of childhood traumas, depressive and anxiety symptoms on the relationship between borderline personality features and symptoms of adult attention deficit hyperactivity disorder in Turkish university students.

Cognitive features of borderline personality disorder.

Difficulties with emotion regulation mediate the relationship between borderline personality disorder symptom severity and interpersonal problems.

A negative relationship between ventral striatal loss anticipation response and impulsivity in borderline personality disorder.

Stability and fluctuation of personality disorder features in daily life.

Personality features related to generalized anxiety disorder.

Convergence between DSM-5 Section II and Section III diagnostic criteria for borderline personality disorder.

An examination of the relationship between childhood emotional abuse and borderline personality disorder features: the role of difficulties with emotion regulation.

The relationship between expressed emotion and wellbeing for families and carers of a relative with Borderline Personality Disorder.

Relationships between mixed features and borderline personality disorder in 2811 patients with major depressive episode.