روابط بین اختلال شخصیتی و حل مسئله اجتماعی در بزرگسالان

Abstract Personality disorders, as defined in DSM-IV, require theoretical models to guide our understanding and treatment of them, and social problem solving is one cognitive model that might contribute. In this study, the relationships between social problem solving and personality disorders were investigated in a sample of 173 men and women in treatment for personality problems. Cluster A diagnoses were infrequent and not amenable to analyses. Of the Cluster B diagnoses, Borderline predominated and was associated with an impulsive/careless problem solving style, as were Histrionic and Narcissistic. Of Cluster C diagnoses, Avoidant was associated with negativity and low impulsive/careless problem solving style, and Dependent with negativity. Thus, the social problem solving profiles of specific personality disorders in Clusters B and C mostly showed the expected associations with personality characteristics. Theoretical and practice implications are discussed.

Introduction Personality disorders (PDs), as defined in DSM-IV (American Psychiatric Association, 1994), have been criticised as having been formulated without any theoretical underpinning, with the consequence that efforts to understand and treat PD lacks direction (Arntz, 1999). Nonetheless, Arntz (1999) provided evidence that DSM PDs are reasonably coherent constructs and suggested that it is possible to propose theoretical models of PD that can be empirically tested. We have proposed a model in which the concept of social problem solving is central to adaptive functioning (McMurran, Egan, & Duggan, 2005). We postulate that innate traits are the developmental start-point for behavioural patterns. Certain personality traits limit and bias information processing, interfering with the acquisition of good social problem solving skills and consequently leading to dysfunctional ways of operating in everyday life. Interpersonal dysfunction causes stress, experienced affectively in a number of ways including anxiety, depression, and anger. Stress further impairs problem solving abilities and may also lead to problematic stress-relieving behaviours, such as substance use, which still further impair social problem solving abilities and also potentially create additional interpersonal problems. Persistent dysfunction leads to a negative approach to life’s problems and the development of maladaptive self-schemas that have a further deleterious effect on information processing and social problem solving. This model is based upon existing, albeit limited, research into problem-solving abilities in people with PDs. We have examined social problem-solving abilities using the Social Problem-Solving Inventory-Revised (SPSI-R; D’Zurilla, Nezu, & Maydeu-Olivares, 2002), a self-report questionnaire measuring problem orientation and styles of social problem solving. Comparisons of mean scores for male personality disordered offenders with sentenced male prisoners and male mature students showed personality disordered offenders to be most negative, least rational, most impulsive, and most avoidant (McMurran, Blair, & Egan, 2002; McMurran et al., 2005). In further investigations with mentally disordered offenders, we used the NEO-Five Factor Inventory (NEO-FFI; Costa & McCrae, 1992), finding that high neuroticism (N) predicted poor social problem solving, and correlations showed N to relate positively to both the impulsive/careless and avoidant subscales of the SPSI-R (McMurran, Egan, Blair, & Richardson, 2001; McMurran et al., 2005). These dimensions of impulsiveness and avoidance are positively correlated in this sample, as in a student sample (McMurran et al., 2002), indicating that people with high N use both maladaptive styles. Furthermore, high openness (O) predicted good social problem solving, possibly because O correlates highly with intelligence, which is associated with better social problem solving (Harris, 2004). Looking at the correlations between O and SPSI-R scales, those who scored high on O were more rational and less impulsive/careless in their approach to problem solving, which accords with other research where O has been shown to correlate with planful problem solving (Bouchard, 2003). One further aspect that we have investigated is impulsiveness, as measured by the Barratt Impulsiveness Scale (BIS; Patton, Stanford, & Barratt, 1995). Using our mature students’ data, all scales of the SPSI-R correlated significantly and in the expected direction with the BIS, indicating that impulsivity may be one trait that adversely affects problem solving (McMurran et al., 2002). Our research has focused on examining traits rather than PDs in relation to social problem solving. However, if PDs are coherent constructs (Arntz, 1999), then it is worth examining them in relation to social problem solving abilities. Herrick and Elliott (2001) studied 117 people in a US dual-diagnosis treatment programme using the Problem Solving Inventory (Heppner, 1988) and the Millon Clinical Multiaxial Inventory – II (Millon, 1987). They found ineffective problem solving to be associated with Cluster A and C PDs, but not Cluster B, although the interpretation was that Cluster B respondents may have overestimated their problem-solving abilities. These studies suggest that the relationships between PD and social problem solving warrant further examination. This may have clinical utility in that there is growing evidence of the effectiveness of social problem solving therapy for people with PDs (Huband, Duggan, McMurran, & Evans, in press; McMurran, Egan, Richardson, & Ahmadi, 1999; McMurran, Fyffe, McCarthy, Duggan, & Latham, 2001). Here, we expand knowledge of PD diagnoses and social problem solving by exploring their relationship in a large sample of personality disordered adults from the community. Our aim was to investigate the problem solving profiles associated with different types of PD

. Results 3.1. Personality disorder diagnoses We examined clear-cut diagnostic cases where there was a definite presence or a definite absence of a diagnosis; people with probable diagnoses were excluded. Frequencies are shown in Table 1, with the distributions of men and women in each diagnostic category being comparable, with the exception of Borderline PD, where women predominate View the MathML source(χ(1)2=10.54,p<0.001). PDs are not equally distributed; Cluster A essentially equates to Paranoid PD, Cluster B to Borderline PD, and Cluster C to Avoidant PD. Analysis by Cluster is, therefore, contraindicated. Table 1. Frequencies of presence and absence of diagnoses for each personality disorder (N = 173; 73 men, 100 women) Cluster/diagnosis Present Absent Men Women Men Women Cluster A Paranoid 8 11 51 71 Schizoid 1 2 70 89 Schizotypal 1 0 70 97 Cluster B Antisocial 10 9 59 83 Borderline 23 59 33 27 Histrionic 0 3 73 91 Narcissistic 2 1 69 99 Cluster C Avoidant 32 42 33 38 Dependent 2 6 65 84 Obsessive–compulsive 9 10 46 71 Not otherwise specified 18 17 55 83 Table options 3.2. Personality disorders and social problem solving Comparisons of SPSI-R scale scores for people with and without personality diagnoses were examined using ANOVAs with age included as a covariate. Because there are multiple comparisons in each Cluster, the risk of familywise error was corrected by Bonferroni correction (α/6), resulting in the criterion for significance being p = .01. Table 2 shows the mean SPSI-R scores for people with PD diagnoses in Cluster A. One-way ANOVAs comparing SPSI-R scores between people with presence or absence of a Paranoid PD diagnosis showed no significant differences for any scale. Insufficient numbers with Schizoid and Schizotypal diagnoses precluded separate analyses in these cases. Table 2. Mean SPSI-R scores for Cluster A SPSI-R Paranoid Schizoid Schizotypal Present (N = 19) Absent (N = 122) Present (N = 3) Absent (N = 159) Present (N = 1) Absent (N = 167) PPO 5.09 5.01 1.28 5.15 7.91 5.01 (4.01)a (3.90) (2.31) (4.07) (–) (3.96) NPO 29.48 28.57 31.98 28.43 37.95 28.36 (8.47) (7.42) (8.00) (7.66) (–) (7.55) RPS 18.87 21.61 0.76 21.40 9.39 21.09 (13.89) (15.83) (0.58) (15.48) (–) (15.41) ICS 21.07 20.45 27.08 21.03 31.67 20.73 (9.77) (9.14) (4.04) (9.26) (–) (9.14) AS 14.82 15.67 17.54 15.27 20.25 15.05 (7.23) (6.41) (6.81) (6.82) (–) (6.80) SPS 6.80 6.96 3.91 6.99 4.21 7.01 (2.95) (3.20) (2.26) (3.25) (–) (3.18) a Standard deviations in parentheses. Table options Table 3 shows the mean SPSI-R scores for people with PD diagnoses in Cluster B. Insufficient numbers with Histrionic and Narcissistic diagnoses precluded separate analyses in these cases. One-way ANOVAs comparing SPSI-R scores between people with presence or absence of Antisocial PD and Borderline PD showed significant differences only for Borderline PD in ICS (F(1146) = 18.38, p < 0.001). In summary, compared with those with no Borderline diagnosis, people with a Borderline diagnosis are significantly more impulsive/careless in their approach to social problem solving. Table 3. Mean SPSI-R scores for Cluster B SPSI-R Antisocial Borderline Histrionic Narcissistic Present (N = 19) Absent (N = 142) Present (N = 82) Absent (N = 60) Present (N = 3) Absent (N = 164) Present (N = 3) Absent (N = 168) PPO 5.41 4.96 5.02 4.83 1.35 5.11 3.06 4.96 (3.32)a (4.13) (4.14) (3.93) (1.53) (4.01) (2.65) (3.95) NPO 28.68 28.20 29.51 27.93 23.40 28.35 33.09 28.43 (7.53) (7.83) (7.92) (7.44) (9.29) (7.70) (4.36) (7.70) RPS 20.03 21.14 19.80 21.89 17.52 20.94 8.55 20.71 (14.27) (15.93) (16.15) (14.90) (9.00) (15.62) (10.39) (15.47) ICS 21.29 20.45 23.53⁎⁎⁎ 17.46 13.67 20.76 21.00 20.72 (7.72) (9.27) (8.91) (8.38) (5.13) (9.14) (3.22) (9.12) AS 13.55 15.14 14.74 15.10 11.90 15.19 11.46 15.21 (6.86) (6.70) (7.01) (6.66) (9.54) (6.62) (7.02) (6.77) SPS 7.18 7.03 6.61 7.37 7.75 7.00 6.01 6.95 (2.55) (3.28) (3.21) (3.16) (2.18) (3.22) (1.76) (3.22) a Standard deviations in parentheses. ⁎⁎⁎ p < .001. Table options Table 4 shows the mean SPSI-R scores for people with PD diagnoses in Cluster C. An insufficient number with a Dependent diagnosis precluded analyses. One-way ANOVAs comparing SPSI-R scores between people with presence or absence of Avoidant and Obsessive–Compulsive diagnoses showed no significant differences at p < .01. Table 4. Mean SPSI-R scores for Cluster C SPSI-R Avoidant Dependent Obsessive–compulsive Present (N = 74) Absent (N = 71) Present (N = 8) Absent (N = 149) Present (N = 19) Absent (N = 117) PPO 4.41 5.58 2.76 5.24 4.84 4.69 (3.58)a (4.27) (2.71) (4.09) (3.67) (3.94) NPO 29.83⁎ 27.34 34.66 27.51 30.13 28.65 (7.07) (7.88) (2.83) (7.76) (7.71) (7.74) RPS 19.86 21.24 13.33 21.21 24.38 19.61 (14.61) (16.51) (5.66) (15.54) (17.25) (15.39) ICS 21.04 21.86 25.87 20.11 21.55 20.72 (9.47) (8.85) (8.70) (9.15) (11.36) (9.08) AS 15.69 14.74 20.45 14.39 15.93 15.80 (6.58) (7.02) (3.30) (6.61) (6.44) (6.62) SPS 6.56 7.16 4.28 7.30 6.74 6.74 (3.18) (3.25) (1.23) (3.22) (3.51) (3.21) a Standard deviations in parentheses. ⁎ p < .05. Table options 3.3. Dimensional personality scores and social problem solving Age, as a known predictor of SPSI-R scale scores, was entered into a hierarchical regression, then each scale of the SPSI-R was entered in a stepwise manner to identify which scales best predict dimensional IPDE scores for each PD (see Table 5). There were no significant predictors for Paranoid, Schizoid, Schizotypal, or Antisocial PDs. Borderline PD was predicted by high ICS and low AS, which describes a profile of impulsive/careless problem solving along with active confronting of problems, with age also a significant predictor. Histrionic PD was predicted solely by high ICS, indicating that those in our sample with Histrionic PD were impulsive and careless in their approach to solving problems. Narcissistic PD was predicted by high PPO and high ICS, giving a profile of optimism and impulsive/careless problem solving. Avoidant PD was predicted by high NPO and low ICS, a profile of pessimism and caution. Dependent PD was predicted by high NPO, that is negativity towards problems. Obsessive–compulsive PD was predicted only by age; the older people were the greater the tendency to obsessive–compulsive traits. Neither RPS nor the total SPS scores predicted any PD type. Table 5. IPDE dimensional scores and SPSI-R scale scores Dimension Acceptability of model Adjusted R2 Predictors Beta Confidence interval Lower Upper Paranoid No predictors Schizoid No predictors Schizotypal No predictors Antisocial No predictors Borderline F(3,169) = 10.03 ⁎⁎⁎ 0.14 ICS +0.18⁎⁎⁎ +0.11 +0.25 AS −0.13⁎⁎ −0.22 −0.04 Age 0.06⁎ −0.12 −0.001 Histrionic F(1,170) = 7.32 ⁎⁎⁎ 0.07 ICS +0.09⁎⁎⁎ +0.04 +0.13 Narcissistic F(3,169) = 3.62 ⁎ 0.04 ICS +0.07⁎⁎ +0.02 +0.12 PPO +0.14⁎ +0.03 +0.25 Avoidant F(3,168) = 6.23 ⁎⁎⁎ 0.08 NPO +0.17⁎⁎⁎ +0.09 +0.24 ICS −0.08⁎ −0.14 −0.01 Dependent F(2,170) = 8.65 ⁎⁎⁎ 0.08 NPO +0.13⁎⁎⁎ +0.07 +0.20 Obsessive–compulsive F(1,171) = 7.57 ⁎⁎ 0.04 Age +0.07⁎⁎ +0.02 +0.13 ⁎ p < .05. ⁎⁎ p < .01. ⁎⁎⁎ p < .001.