مدل پنج عاملی در اختلال شخصیت

Abstract Studies of the five-factor model of personality in schizotypal personality disorder (SPD) have produced inconsistent results, particularly with respect to openness. In the present study, the NEO-FFI was used to measure five-factor personality dimensions in 28 community volunteers with SPD and 24 psychiatrically healthy individuals. Standard multivariate statistical analyses were used to evaluate personality differences as a function of diagnosis and gender. Individuals with SPD had significantly higher levels of neuroticism and significantly lower levels of extraversion, agreeableness and conscientiousness than those without SPD. Female, but not male, SPD subjects had significantly higher openness levels than their healthy counterparts, and this gender-specific group difference persisted when SPD symptom severity was statistically controlled. These findings suggest that gender-associated differences in openness may account for prior inconsistent findings regarding this dimension, and they further u

Introduction 1.1. The five-factor model Personality description and classification systems can be grouped into those that delineate distinct categorical entities based upon specific features, and those that attribute individual differences to variation along one or more continuous dimensions. Axis II of the DSM-IV exemplifies the former approach, and the five-factor model is a widely known example of the latter. The five-factor model is well replicated across instruments and over time (John and Srivastava, 1999), and posits that variation along broad personality dimensions (the so-called ‘Big Five’) accounts for most inter-individual personality differences (Digman, 1990). Until recently, investigations of categorical and dimensional personality models have proceeded along mutually independent lines of research, reflecting their origins in clinical psychiatry and academic psychology, respectively (Lenzenweger and Clarkin, 1996). A widely used measure of the Big Five dimensions is the NEO Personality Inventory (Costa and McCrae, 1992), which defines each dimension in terms of related attributes, or facets. Thus, individuals who score high on neuroticism are tense, irritable, dissatisfied, shy, moody and lacking self-confidence. High extraversion indicates an individual who is sociable, forceful (assertive), energetic, adventurous, enthusiastic, and outgoing. High openness individuals are curious, imaginative, artistic, excitable, unconventional, and have wide interests. High agreeableness individuals are forgiving, not demanding, warm, compliant (not stubborn), modest (not show-offs), and sympathetic. High conscientiousness indicates someone who is efficient, organized, dutiful (not careless), thorough, self-disciplined (not lazy), and deliberate (not impulsive). 1.2. The five-factor model and personality disorders The conceptual gap between dimensional constructs of “normal” personality and clinically based categorical classifications of abnormal personality was bridged empirically by Wiggins and Pincus (1989), who demonstrated that the five-factor model accounts for much of the variance in personality disorder (PD) diagnoses. Similar results were obtained subsequently by others. For example, using several different PD measures in a community sample, Costa and McCrae (1990) concluded that the five-factor model accounts for the “major dimensions underlying personality disorder”. In another study (Blais, 1997), clinicians used five-factor model trait descriptions to rate their own patients who met DSM-IV diagnostic criteria for one or more PDs and obtained results similar to those reported by Wiggins and Pincus (1989). Based upon a review of these and other studies, Dyce (1997) concluded that high neuroticism is typical of PDs generally, whereas openness may be elevated in some (e.g., narcissistic and histrionic) but low in others (e.g., schizoid). In undergraduates schizotypal personality disorder (SPD) scores were positively related to neuroticism and openness, and negatively related to extraversion and agreeableness (Dyce and O'Connor, 1998). Morey et al. (2002) examined the five-factor model in patients with borderline, avoidant, obsessive–compulsive, and schizotypal PDs and found that they shared a common pattern of above-average neuroticism and below-average extraversion, agreeableness and conscientiousness; openness was elevated in all subgroups except avoidant PD. A limitation of that study is that subjects with major mood disorders were not excluded. 1.3. The five-factor model in schizotypal personality disorder Among clinical PDs, SPD is least accounted for by the five-factor model (Stone, 1993 and Lynam and Widiger, 2001), possibly because it lacks a dimension related to aberrant cognition (Costa and McCrae, 1990). Studies of the relationship between five-factor model traits and SPD have also yielded inconsistent results, which may reflect differences in personality and PD measures, analytic methods, and types of populations sampled (Dyce, 1997). Perhaps the most controversial issue in research on SPD and the five-factor model is the role of openness (Ross et al., 2002). Morey et al. (2002) found that openness was elevated in patients with SPD, but other studies found no relationship (Trull, 1992, Yeung et al., 1993 and Blais, 1997). Schizotypy is positively correlated with openness in college students (Wiggins and Pincus, 1989 and Coolidge et al., 1994), but Tien et al. (1992) reported that it is negatively related to openness in a community sample, and others (Costa and McCrae, 1990) found that openness is unrelated or negatively related to SPD depending on the PD measure used. Generally, studies reporting a positive relation between openness and SPD symptoms have used college student samples, whereas those failing to find this relation have utilized psychiatric samples (Ross et al., 2002), suggesting that symptom severity may influence this relationship. In particular, Ross et al. (2002) have proposed that discrepant results are due to failures to control for negative SPD symptom severity within samples. Intelligence (Ashton et al., 2000) and cognitive performance (Demetriou et al., 2003) are other factors that might contribute to differences in openness scores. It is also possible that different ascertainment methods could influence reported openness levels. Some subject populations (e.g., college students or paid volunteers) might be more inclined than others (e.g., patients referred by a treating clinician) to report more openness characteristics. However, this theoretical bias does not appear to account for the pattern of findings that have been reported. 1.4. SPD, personality dimensions and gender Although many potentially confounding variables have been implicated in these discrepant findings, gender has not been considered. Morey et al. (2002) compared personality dimension scores in their SPD subjects to mixed gender norms, possibly obscuring any contribution by gender. Gender effects may also have been obscured by the use of self-standardized scores in the study by Wiggins and Pincus (1989), because systematic gender effects on other personality traits could differentially affect self-standardized scores for openness in female and male subjects. However, Ross et al. (2002) reported that positive SPD symptoms were positively, and negative SPD symptoms negatively, related to openness after statistically removing gender effects, and previous work by Niznikiewicz et al. (2004) indicates that gender may be important in understanding SPD. 1.5. Study aims In the present study, men and women who met diagnostic criteria for SPD were recruited from the community and compared to a psychiatrically healthy community sample with respect to five-factor personality dimensions. Because SPD is thought to be genetically related to schizophrenia (Kendler et al., 1993 and Battaglia and Torgersen, 1996), we hypothesized that personality profiles in SPD subjects would resemble those in patients with schizophrenia (Gurrera et al., 2000). Specifically, we predicted that SPD subjects would show elevated neuroticism, reduced extraversion and conscientiousness, and reduced or normal levels of openness and agreeableness. We were also interested in examining the SPD subjects for gender-related personality differences, which have not been described previously.

. Results 3.1. Descriptive statistics Demographic data, by subgroup, are presented in Table 1. There were no significant subgroup differences for age, SES, PSES or IQ. The male comparison subgroup had significantly more education than other subgroups, which did not differ from one another (Table 1). Female SPD subjects had significantly higher mean total SPD scores than males (Table 2), but SPD subgroups did not differ on other symptom measures. Table 1. Mean (S.D.) demographic data by subgroup Variable SPD-m SPD-f C-m C-f F[3,48] p N 13 15 11 13 Age (years) 37.6(11.8) 33.1(10.4) 31.2(10.4) 28.0(9.9) 1.844 .152 Education (years) 15.3(2.6) 16.5(1.8) 19.0(3.4) 16.8(1.3) 5.042 .004a SES 2.8(1.3) 2.5(1.6) 2.8(1.6) 2.3(1.8) .342 .795 PSES 4.1(1.1) 4.0(1.1) 4.2(1.0) 4.0(1.0) .078 .971 Full scale IQ 114.8(10.1) 119.1(13.8) 121.1(15.9) 119.7(8.9) .601 .618 aMale comparison subgroup (C-m) had significantly more years of education than did female comparison (C-f) (p = .030), male SPD (SPD-m) (p < .001) and female SPD (SPD-f) (p = .011) subgroups (post hoc contrasts by Least Significant Difference method). Other subgroups did not differ from one another (p ≥ .101). Table options Table 2. Mean (S.D.) clinical symptom measures for SPD subgroups Symptom measure SPD-m SPD-f tb p SANS total score 4.38 (3.40) 4.00 (1.48) .35 .732 SAPS total score 5.00 (1.08) 4.73 (1.95) .43 .670 SPD negative symptoms 2.00 (0.82) 2.47 (0.92) − 1.41 .169 SPD positive symptoms 3.46 (0.66) 3.67 (0.82) − .72 .476 SPD total score 5.46 (0.78) 6.13 (0.92) − 2.08 .048 bDegrees of freedom = 22 for SANS and SAPS comparisons due to missing data for 4 female SPD subjects; df = 26 for SPD symptom comparisons. Table options 3.2. Diagnosis, personality and gender T scores computed with gender-specific normative data ( Costa and McCrae, 1992) differed significantly between SPD and comparison subjects on all personality dimensions (multivariate F[5,46] = 9.763, p < .001). SPD subjects scored significantly higher on neuroticism and openness, and lower on extraversion, agreeableness and conscientiousness ( Table 3). Gender subgroup T scores were consistent with group differences, except that female SPD subjects scored much higher on openness than their male counterparts, whose scores were similar to the male comparison subjects ( Table 4). Also, female SPD subjects had somewhat lower conscientiousness scores than the male SPD subgroup. These subgroup differences were unexpected since T scores were computed using gender-specific normative data. In contrast, male and female comparison subjects had similar scores on all dimensions ( Table 4). Table 3. Personality T scoresc, by diagnostic group Personality dimension SPD group Comparison group F[1,50] p Neuroticism 57.7 (12.8) 39.8 (9.2) 32.554 < .001 Extraversion 43.4 (13.0) 57.2 (8.5) 19.873 < .001 Openness 67.0 (10.9) 57.0 (7.9) 13.772 .001 Agreeableness 41.4 (11.4) 56.5 (9.0) 27.288 < .001 Conscientiousness 41.3 (13.0) 51.3 (12.0) 8.172 .006 cT scores computed from gender-specific normative data reported by Costa and McCrae (1992). Table options Table 4. Personality dimension mean (S.D.) T scores by subgroup Personality dimension SPD-m SPD-f C-m C-f Neuroticism 58.5 (15.1) 57.1 (10.9) 41.2 (8.4) 38.7 (10.1) Extraversion 43.6 (15.0) 43.3 (11.5) 55.4 (9.1) 58.8 (8.0) Openness 59.1 (10.3) 73.8 (5.6) 59.1 (7.7) 55.3 (7.9) Agreeableness 41.1 (13.2) 41.6 (10.1) 56.1 (10.7) 56.8 (7.8) Conscientiousness 46.3 (14.8) 37.0 (9.6) 52.3 (10.8) 50.5 (13.2) Table options To further evaluate the effects of gender and diagnosis on personality, two-way MANOVA with diagnosis and gender as between-subjects factors was performed using raw scores. This analysis confirmed a statistically significant main effect for diagnosis (F[5,44] = 9.528, p < .001) and a significant gender × diagnosis interaction (F[5,44] = 3.988, p = .005), but only a weak trend for a main effect of gender (F[5,44] = 2.138, p = .079). Post hoc univariate ANOVAs demonstrated that the effect of diagnosis was statistically significant for all personality dimensions (F[1,48] ≥ 7.937, p ≤ .007), but only openness was significantly associated with gender (F[1,48] = 5.792, p = .020) and a gender × diagnosis interaction (F[1,48] = 17.342, p < .001). To identify possible sources of personality variance other than gender and diagnosis, product moment correlations were computed between personality dimensions and demographic variables, and between personality dimensions and symptom scales. A Bonferroni probability of .001, reflecting an alpha level of .05 applied to 50 comparisons, was used to evaluate the outcomes. In the total sample only education was significantly correlated with personality measures (for neuroticism, r = − .531, p < .001; for conscientiousness, r = .452, p = .001). Extraversion was correlated with SPD negative symptoms (r = − .664, p < .001), and there was a strong trend for a similar correlation with total SPD symptoms (r = − .557, p = .002). Thus, education differed between subgroups and was correlated with neuroticism and conscientiousness; and SPD symptoms differed between SPD subgroups and were correlated with extraversion. Openness was not correlated with any symptom measure in SPD subjects (− .087 ≤ r ≤ .146, p ≥ .460) or SPD gender subgroups (− .193 ≤ r ≤ .138, p ≥ .526). To assess the possible contribution of differences in education and SPD symptom severity to personality differences between subgroups, a second MANOVA was performed on residual personality dimension scores after removing the variance due to educational achievement and SPD total symptoms. First, raw scores for each personality dimension were regressed on SPD total score (Table 5), and residual personality scores were saved. All personality dimensions were significantly predicted by SPD total score. Next, because educational achievement remained correlated with residual neuroticism (r = − .439, p = .001) and residual conscientiousness (r = .363, p = .008), a second linear regression was carried out in which residual personality scores were regressed on educational achievement ( Table 6). Note that only residual neuroticism and conscientiousness scores were significantly predicted by educational achievement (all other F[1,50] ≤ .486, p ≥ .489), so only those residual scores from this second regression were saved separately. Table 5. Linear regression of personality dimensions on SPD symptoms Personality dimension R R2 Sum of squares df Mean square beta F p NEO-N Regression .650 .422 2520.311 1 2520.311 .650 36.495 < .001 Residual 3452.920 50 69.058 NEO-E Regression .612 .375 1121.032 1 1121.032 − .612 30.018 < .001 Residual 1867.276 50 37.346 NEO-O Regression .513 .263 532.349 1 532.349 .513 17.842 < .001 Residual 1491.881 50 29.838 NEO-A Regression .592 .350 723.639 1 723.639 − .592 26.928 < .001 Residual 1343.668 50 26.873 NEO-C Regression .391 .153 467.167 1 467.167 − .391 9.038 .004 Residual 2584.583 50 51.692 Table options Table 6. Linear regression of residual personality dimensions on educational achievement Personality dimension R R2 Sum of squares df Mean square beta F p NEO-N (resid.) Regression .439 .193 666.093 1 666.093 − .439 11.951 .001 Residual 2786.827 50 55.737 NEO-C (resid.) Regression .363 .132 340.791 1 340.791 .363 7.594 .008 Residual 2243.793 50 44.876 Table options The second MANOVA therefore included twice-regressed (on SPD total score and educational achievement) residual neuroticism and conscientiousness scores, and once-regressed (on SPD total score only) extraversion, openness, and agreeableness residual scores. With variance due to SPD symptoms and educational achievement extracted from personality scores, there was no longer a significant main effect for diagnosis (F[5,44] = .513, p = .765). However, a strong gender × diagnosis interaction persisted (F[5,44] = 3.508, p = .009). This interaction was evaluated by post hoc univariate analyses. Post hoc one-way ANOVAs revealed that the gender × diagnosis interaction was due entirely to openness (F[3,48] = 6.928, p = .001; all other F[3,48] ≤ 1.838, p ≥ .153). Subgroup contrasts by LSD showed that residual openness scores in female SPD subjects were significantly higher than those in female (p = .023) but not male (p = .247) comparison subjects, whereas male SPD subjects had lower residual openness scores than female SPD subjects (p <.001), male comparison (p = .005) and female comparison (p = .051) subjects. Thus, female and male SPD subjects differed not only from controls, but also from one another. In fact, after accounting for variance due to subgroup differences in SPD symptom severity, female and male SPDs actually had divergent openness scores. Of the four subgroups, male SPD subjects had the lowest, and female SPD subjects the highest, mean residual openness scores. Conversely, male comparison subjects scored higher than their female counterparts on openness, although this difference was not statistically significant.