اختلالات شخصیت در جمعیت عمومی: شیوع تعریف شده DSM-IV و ICD-10 به عنوان مربوط به مشخصات اجتماعی جمعیتی

Abstract Prevalence and sociodemographic characteristics of DSM-IV and ICD-10 defined personality disorders were examined in a Swedish community sample. Data were obtained by means of the DSM-IV and ICD-10 personality questionnaire (DIP-Q) postal survey administered to 1000 randomly selected adults from the Isle of Gotland. A total of 557 individuals responded to the questionnaire. The prevalence of any ICD-10 defined personality disorder was 11.0% and 11.1% when using DSM-IV criteria. Comorbidity between personality disorders were common. Prevalence similarities between DSM and ICD definitions were obtained for paranoid, schizotypal, borderline/emotionally unstable, histrionic, avoidant, dependent, and obsessive-compulsive/anancastic but not schizoid and antisocial/dissocial personality disorders. Personality disorders were significantly more often diagnosed in the younger subjects, students and unemployed/homemakers had the highest rates. Individuals with personality disorders more often received psychiatric treatment and lacked social support. In addition, they reported significantly more psychosocial and environmental problems in the past year. Thus, personality disorders are relatively common in the community and affected individuals are more impaired than subjects without personality disorders.

Introduction Personality disorders may be significant causes of psychiatric morbidity and can be major sources of long-term disability. They relate to a poor outcome in patients with other major psychiatric disorders and general medical conditions (Reich and Vasile, 1993, Sato et al., 1994 and Ekselius et al., 1996). However, it is necessary to investigate personality disorders in the population to estimate their prevalence in the community, understand their consequences in individuals who do not seek treatment, in order to plan and evaluate the delivery of psychiatric services (Samuels, Nestadt, Romanoski, Folstein & McHugh, 1994). Two areas of great concern with high societal costs and great individual suffering for example are substance abuse and suicide. Epidemiological studies in North America and Europe conducted between 1951 and 1963 suggest a personality disorder prevalence rate of 6–10% of the total population (Bremer, 1951, Essen-Möller, 1956, Leighton, 1959 and Langner and Michael, 1963). The overall lifetime community rates or in nonreferred samples according to DSM-III/DSM-III-R criteria are 6–14% (Casey and Tyrer, 1986, Reich et al., 1989a, Zimmerman and Coryell, 1990, Maier et al., 1992 and Samuels et al., 1994). Reich et al. (1989a), ascertained DSM-III personality disorders by sending the self-completed Personality Diagnostic Questionnaire (Hyler et al., 1988) to a random sample of households in Iowa City. Reich et al. (1989a) estimated that 11.1% of the 235 adults who completed the questionnaire had personality disorders. Subjects with personality disorders were less educated and reported more alcohol problems. Of those married, individuals with personality disorders reported more marital difficulties than subjects without personality disorders. In addition, there was a significant association between the number of disordered personality traits and medical resource utilisation (Reich, Boerstler, Yates & Nduaguba, 1989b). Samuels et al. (1994) reported an adult personality disorder prevalence rate of 5.9% (9.3% when provisional cases were included) using a semistructured method to interview 810 adults in the Eastern Baltimore Mental Health Survey. Men had higher rates than women and separated or divorced subjects had the highest rates. Individuals with personality disorders were more likely to have a history of sexual dysfunctions, alcohol and drug use disorders as well as suicidal thoughts and attempts. In addition, they reported significantly more life events in the past year. Only one fifth of the individuals who qualified for diagnoses of personality disorder in the community sample received treatment. In 1992, the tenth revision of the international statistical classification of diseases and related health problems (ICD-10) (WHO, 1992) was introduced, and separate diagnostic criteria for personality disorders research purposes are now available (WHO, 1993). In the fourth version of the diagnostic manual of mental disorders (DSM-IV) (American Psychiatric Association, 1994) stringent efforts were made to increase concordance with the ICD-10. The two upgraded versions of the DSM and ICD have great similarities. DSM-IV covers ten personality disorders whereas the ICD-10 diagnostic criteria for research cover eight personality disorders. Narcissistic personality disorder is not specified in the ICD-10. Furthermore, the DSM-IV schizotypal personality disorder is a “schizophrenia spectrum disorder” classified under “Schizophrenia, schizotypal and delusional disorders” in the ICD-10. During the past years the DSM-IV and ICD-10 Personality–Questionnaire (DIP-Q) has been developed and validated (Ottosson et al., 1995 and Ottosson et al., 1998). The aim of the present study was to examine the prevalence of DSM-IV and ICD-10 personality disorders and social characteristics in a randomly selected community sample of the Isle of Gotland using the DIP-Q.

3. Results 3.1. Prevalence of personality disorders Of the 557 subjects who completed and returned their questionnaires and demographic information, 62 subjects (11.1%) fulfilled criteria for any DSM-IV personality disorder. The corresponding figure for ICD-10 was 61 subjects (11.0%). Two individuals with DSM-IV personality disorders had no such disorders according to ICD-10, and one individual with an ICD-10 personality disorder was negative with respect to the corresponding DSM-IV disorder. The Kappa coefficient of agreement for any personality disorder between the two classification systems was 0.97. Table 1 shows the prevalence of personality disorders. Prevalence rates did not differ significantly between urban/rural regions (χ2=0.41, df=2, n.s.). Table 1. The prevalence of various personality disorders in a randomly selected Swedish community sample (n=557), and the diagnostic agreement between the DSM-IV and the ICD-10 classification systems DSM-IV ICD-10 Kappa coefficient n % n % Paranoid 31 5.6 33 5.9 0.80 Schizoid 5 0.9 25 4.5 0.32 Schizotypal 29 5.2 42 7.5 0.80 Antisocial/dissocial 10 1.8 21 3.8 0.50 Borderline/emotionally unstable 30 5.4 27a 4.8 0.76 Impulsive type – – 15 2.7 – Borderline type – – 25 4.5 – Histrionic 12 2.2 11 2.0 0.78 Narcissistic 16 2.9 – – – Avoidant/anxious 37 6.6 31 5.6 0.84 Dependent 10 1.8 13 2.3 0.78 Obsessive-compulsive/anancastic 43 7.7 40 7.2 0.91 Any personality disorder 62 11.1 61 11.0 0.97 a Thirteen (2.3%) of the subjects fulfilled criteria for both borderline type and impulsive type. Table options The highest prevalence rates were seen for the ICD-10 schizotypal disorder and anancastic personality disorder, and for the DSM-IV obsessive-compulsive and avoidant personality disorders. Differences between the prevalence obtained by the two classification systems was most obvious for schizoid personality disorder which was diagnosed in 25 (4.5%) of the subjects according to ICD-10 as compared to only five individuals (0.9%) using the DSM-IV criteria. ICD-10 dissocial personality disorder was diagnosed in 21 (3.8%) of the subjects as compared to 10 (1.8%) DSM-IV diagnoses. Cohen’s Kappa of agreement for the presence or absence of a personality disorder according to the two systems ranged from 0.32 for schizoid personality disorder to 0.91 for obsessive-compulsive disorder (Table 1). Multiple personality disorders diagnoses were common. The mean number of diagnoses in those with at least one ICD-10 personality disorder were 3.9 (SD 2.2) as compared to 3.5 (SD 2.0) in those with a DSM-IV personality disorder (t=2.03, p<0.05). 3.2. Sociodemographic characteristics Table 2 presents the prevalence of ICD-10 personality disorders by sociodemographic characteristics. The Bonferroni correction used in the chi-square analyses yielded an adjusted alpha level of 0.05/9=0.0056. The overall prevalence of personality disorders was nonsignificantly higher in women than in men. Personality disorders were significantly more often diagnosed in younger subjects. Of subjects aged 55–70 years, only 5.0% were given a personality disorder diagnosis as compared to 17.1% of those aged between 18 and 34 years. There was a significant relationship between personality disorders and occupational status; students and unemployed/homemakers had the highest rates. In addition, the prevalence of personality disorders was higher in singles, although this difference did not reach significance. No significant relationship was seen for educational level. Subjects with personality disorders more often received psychiatric treatment and lacked social support. Table 2. Sociodemographic and descriptive characteristics of subjects with and without ICD-10 personality disorders Characteristic Personality disorder No personality disorder, n χ2 (df) p n % Age, years 11.3 (2) 0.003 18–34 29 17.1 141 35–54 26 9.8 240 55–70 6 5.0 115 Sex 1.5 (1) n.s. Female 38 12.4 268 Male 23 9.2 228 Marital status 9.7 (2) n.s. Married 35 9.1 348 Divorced/widowed 2 4.5 42 Single 23 18.0 105 Educational level 5.9 (2) n.s. Lowa 22 11.6 168 Mediumb 33 13.4 214 Highc 6 5.0 114 Occupational status 12.8 (3) 0.005 Working 31 8.2 349 Student 12 22.2 42 Retiredd 7 11.9 52 Unemployed/homemaker 11 17.5 52 Immigration 3.1 (1) n.s. Born in Sweden 58 10.6 488 Born abroad 3 27.3 8 Psychiatric treatment 19.5 (1) <0.001 Yes 13 31.7 28 No 48 9.3 467 Psychiatric medication 6.7 (1) n.s. Yes 9 23.7 29 No 52 10.1 463 Social support 18.1 (1) <0.001 Yes 38 8.4 414 No 23 23.2 76 a Grades 1–9 or elementary school. b High school, trade school or technical education below university level. c University or university college. d Age or health reasons. Table options According to the Axis IV life-event scale, subjects with personality disorders were significantly more likely to report psychosocial and environmental problems than those without personality disorders, 3.6 (SD 2.3) vs 2.3 (SD 1.4) (t=4.40, p<0.001). They also rated their distress level with respect to these problems significantly higher, 3.5 (SD 0.9) vs 3.1 (SD 1.1) (t=2.41, p<0.02). To get a rough estimate of the prevalence of alcohol problems and suicide gestures the following statements from DIP-Q were analysed separately: “I often act impulsively without thinking which causes me to drink to much”, and “I have never threatened to commit suicide”. Subjects with personality disorders were more likely to report impulsive alcohol drinking, 34% (n=21) vs 8.6% (n=40) (χ2=38.8, df=2, p<0.001). They were also more likely to report suicidal threat, 47.5% (n=29) vs 28.6% (n=142) (χ2=9.1, df=1, p<0.003). These differences remained statistically significant also when emotionally unstable personality disorder was excluded.