تفاوت های جنسیتی در شیوع و عوامل مرتبط با اختلال شخصیت ضد اجتماعی در میان کاربران وابسته به هروئین در درمان ایزولاسیون اجباری در چین

Abstract Background Little is known about gender difference in correlates of antisocial personality disorder (ASPD) among drug users. Objective To detect gender difference in correlates of ASPD in a Chinese heroin dependent sample. Methods Structured interviews were conducted among 882 heroin dependent users in two compulsory isolation settings in Changsha, China. Descriptive statistics were employed to report sample characteristics by gender. Bivariate relationships were examined between co-occurring ASPD and variables measuring demographic, drug use, and psychiatric co-morbidities. Multivariate logistic regressions with stepwise forward method were conducted to determine independent predictors for co-occurring ASPD. All analyses examining correlates of co-occurring ASPD were conducted for the total, the male and the female participants respectively to detect both the common and the unique correlates of ASPD by gender. Results Of the total participants, 41.4% (54.2% of males and 15.4% of females) met the DSM-IV criteria of ASPD. For male participants, lower educational level, unemployment, unmarried, younger age at first heroin use, previous history of compulsory treatment, larger amounts of heroin used per day and poly-drug abuse during past month before admission, as well as psychiatric co-morbidities of lifetime major depressive disorder and borderline personality disorder were independent predictors for co-occurring ASPD; while for female participants, only three variables: younger age at first heroin use, paranoid personality disorder and borderline personality disorder were independent predictors for co-occurring ASPD. Conclusions Gender differences in prevalence and correlates of ASPD among heroin dependent users were detected. The findings highlight a need for gender-specific interventions.

1. Introduction Antisocial personality disorder (ASPD) is typically the most common co-occurring diagnosis among drug abusers, with the prevalence ranging between 23 and over 80% reported among opioid abusers (Brooner et al., 1997, Grella et al., 2009, Jun–Jul, Ladd and Petry, 2003, Mackesy-Amiti et al., 2012, Neufeld et al., 2008 and Ross et al., 2005). The cognitive and emotive deficiencies that underlie ASPD, such as poor self-control, ineffective behavioral regulation, and poor decision-making, may facilitate and maintain patterns of drug use (Hasin et al., 2011). The co-morbidity of ASPD in drug users has been linked with earlier onset age of drug use, greater severity of addiction, more psychiatric co-morbidities and social disadvantages (Goldstein et al., 2007, Huang et al., 2009, Moran et al., 2006, Mueser et al., 2012, Najt et al., 2011, Rosenthal et al., 2012, Mar–Apr and Shand et al., 2011, Jan–Feb). In spite of adverse correlates of co-morbidity of ASPD in populations with drug use disorders, no studies have reported gender difference in the correlates of ASPD among these populations, which, however, has an important implication for whether a gender-specific strategy is needed when targeting to treat the co-morbidity of ASPD and drug use disorder. Compulsory isolation treatment is a major treatment modality for drug abusers adopted nowadays in China. By the end of 2008 (the year of the present research being carried out), about 1,126,700 drug abusers were officially registered in China and 103,359 of them were receiving treatment in compulsory isolation settings (National Narcotic Control Commission of China, 2009). It was stipulated in Chinese drug law that drug addicts who refused to receive community-based rehabilitation or failed to keep abstinence in community, or those found by police having a severe drug addiction would be sent for 1–3 years to compulsory isolation treatment centers managed by justice departments (Office of China National Narcotics Control Commission, 2008, Standing Committee of the National People's Congress, 2007 and State Council of the People's Republic of China, 2011). As drug users are not allowed to leave these centers, they are considered incarceration sites. Treatments provided in these centers are mainly detoxification, basic medical care, physical training, drug and HIV/AIDS education, behavior correction (however consisting mainly of required manual work) and group therapy, with psychological counseling, job skill training, and anti-relapse training and medicine for treatment of psychiatric disorders is also provided, but still far from enough. As under these conditions of admission and treatment, drug addicts in compulsory isolation treatment might be more severe in drug use and psychopathology, as well as in more social disadvantaged status, than drug users in communities. As thus, the current study went further than previous research to seek to detect gender difference in the correlates of ASPD among a sample of heroin dependent users in compulsory isolation treatment. Correlates of ASPD examined in the current study fell into three areas: demographics, drug use and psychiatric co-morbidities. To be specific, the aims of this study were: (1) to examine the demographic, drug use and psychopathological correlates of co-occurring ASPD by gender among a sample of heroin dependent users admitted in two selected compulsory isolation treatment settings in Changsha, China. (2) to detect the gender difference in correlates of ASPD among this sample of heroin dependent users.

. Results 3.1. Participant characteristics Participant characteristics of the study sample are provided in Table 1. Compared to male participants, females were significantly younger, less educated, more likely to be divorced or widowed, more likely to be unemployed, and in a significantly lower income level. Female participants also reported a younger age at first heroin use and less amounts of heroin used per day over the month before admission. Male participants had significant higher proportions of ASPD, passive–aggressive personality disorder and avoidant personality disorder than female participants. Of the total participants, 41.4% (54.2% of males and 15.4% of females) met the DSM-IV criteria of ASPD. Table 1. Participant characteristics by gender. Variable Total n = 882 % Male n = 590 % Female n = 292 % χ2/t P Demographics Age, years 12.64 0.002 ≤ 30 37.6 33.7 45.5 31–40 47.7 51.5 40.1 ≥ 41 14.6 14.7 14.4 Mean ± SD 33.2 ± 6.9 33.6 ± 6.8 32.2 ± 7.1 2.78 0.006 Education 0.89 0.478 –Junior school 76.1 75.9 76.4 High school– 23.9 24.1 23.6 Mean ± SD, years 8.8 ± 2.6 8.9 ± 2.5 8.5 ± 2.8 2.24 0.030 Marital state 8.09 0.018 Married or cohabiting 26.3 28.0 22.9 Divorced or widowed 22.9 20.2 28.4 Never married 50.8 51.9 48.6 Employment 5.90 0.015 Employed 43.1 45.9 37.3 Unemployed 56.9 54.1 62.7 Personal income, ￥/year 32.79 0.000 < 30,000 63.9 57.8 76.4 30,000–100,000 24.3 29.7 13.4 > 100,000 11.8 12.5 10.3 Drug use Age at first heroin use, years 6.07 0.014 –20 54.0 51.6 71.1 21– 46.0 48.4 28.9 Mean ± SD, years 22.9 ± 5.7 21.6 ± 4.8 19.3 ± 3.6 3.05 0.002 Heroin use career, yearsa 1.88 0.391 < 5 16.9 16.1 18.5 5–10 30.2 31.5 27.4 > 10 52.9 52.4 54.1 Mean ± SD, years 10.4 ± 4.7 10.3 ± 4.5 10.4 ± 5.0 0.20 0.839 Amounts of heroin used per dayb, g/d 2.21 0.137 < 0.5 38.0 36.3 41.4 ≥ 0.5 62.0 63.7 58.6 Mean ± SD, g/d 0.6 ± 0.5 0.7 ± 0.5 0.6 ± 0.5 2.37 0.018 Heroin injection 2.08 0.150 No 12.1 11.0 14.4 Yes 87.9 89.0 85.6 Poly-drug abusec 0.65 0.421 No 59.4 58.5 61.3 Yes 40.6 41.5 38.7 Previous history of compulsive treatment 0.87 0.351 No 13.2 13.9 11.6 Yes 86.8 68.1 88.4 Psychiatric co-morbidity Lifetime major depressive disorder 14.7 14.6 15.1 0.04 0.846 Current major depressive disorder 9.2 9.2 9.2 0.00 0.964 Dysthymia 5.0 5.4 4.1 0.72 0.399 Avoidant personality disorder 11.8 13.9 7.5 7.61 0.006 Passive–aggressive personality disorder 10.2 11.9 6.8 5.36 0.021 Paranoid personality disorder 7.5 8.5 5.5 2.53 0.112 Depressive personality disorder 7.1 8.0 5.5 1.82 0.117 Borderline personality disorder 22.1 22.7 20.9 0.376 0.540 ASPD 41.4 54.2 15.4 121.38 0.000 P < 0.05 was denoted in boldface in the table. a Heroin use career in this study refers to the period of years from first heroin use to this admission minus the period of years of continued abstinence outside drug treatment settings. b Amounts of heroin used per day over the 30 days before admission. c Poly-drug abuse was defined in this study as abusing or being dependent on at least one other substance (alcohol or illicit drugs) than heroin/opiates during 30 days before admission (with at least one current diagnosis of DSM-IV substance use disorder other than heroin/opioid use disorder). Nicotine use was not included in the analyses of this study. Table options Characteristics of participants with ASPD were provided in Table 2. Of the participants with ASPD, females were younger, less educated, with a younger age of first heroin use and used less heroin per day during past month before admission. Female participants with ASPD were also more likely to have borderline personality disorder than their male counterparts. Table 2. Characteristics of participants with ASPD by gender. Variable Total n = 365 % Male n = 320 % Female n = 45 % χ2/t P Demographics Age, years 5.85 0.054 –30 43.8 41.6 60.0 31–40 47.1 49.4 31.1 41– 9.0 9.1 8.9 Mean ± SD 33.2 ± 6.9 32.0 ± 6.6 29.8 ± 6.6 2.17 0.031 Education 0.32 0.572 –Junior school 81.4 80.9 84.4 High school– 18.6 19.1 15.6 Mean ± SD, years 8.8 ± 2.6 8.5 ± 2.6 8.0 ± 2.6 1.26 0.210 Marital state 2.31 0.315 Married or cohabiting 22.5 22.5 22.2 Divorced or widowed 16.7 15.6 24.4 Never married 60.8 61.9 53.3 Employment 0.37 0.408 Employed 34.8 35.6 28.9 Unemployed 65.2 64.4 71.1 Personal income, ￥/year 6.19 0.045 < 30,000 62.5 60.9 73.3 30,000–100,000 23.6 25.6 8.9 > 100,000 14.0 13.4 17.8 Drug use Age at first heroin use, years 10.30 0.001 –20 42.7 39.0 50.3 21– 57.3 61.0 49.7 Mean ± SD, years 22.9 ± 5.7 23.3 ± 5.7 22.1 ± 5.6 2.98 0.003 Heroin use career, yearsa 0.14 0.935 < 5 14.0 13.8 15.6 5–10 32.9 32.8 33.3 > 10 53.2 53.4 51.1 Mean ± SD, years 10.4 ± 4.7 10.5 ± 4.4 10.8 ± 5.4 0.33 0.742 Amounts of heroin used per dayb, g/d 0.44 0.509 < 0.5 33.4 32.8 37.8 ≥ 0.5 66.6 67.2 62.2 Mean ± SD, g/d 0.6 ± 0.5 0.7 ± 0.5 0.6 ± 0.5 2.37 0.018 Heroin injection 0.30 0.584 No 6.3 6.6 4.4 Yes 93.7 93.4 95.6 Poly-drug abusec 0.87 0.351 No 51.0 51.9 44.4 Yes 49.0 48.1 55.6 Previous history of compulsive treatment 0.29 0.586 No 11.0 10.6 13.3 Yes 89.0 89.4 86.7 Psychiatric co-morbidity Lifetime major depressive disorder 18.1 18.1 17.8 0.00 0.955 Current major depressive disorder 11.0 10.9 11.1 0.00 0.972 Dysthymia 5.9 5.3 2.2 0.80 0.370 Avoidant personality disorder 12.9 13.8 6.7 1.76 0.784 Passive–aggressive personality disorder 12.5 8.9 12.1 0.49 0.486 Paranoid personality disorder 12.1 10.9 20.0 3.06 0.080 Depressive personality disorder 9.9 9.7 11.1 0.09 0.764 Borderline personality disorder 31.5 29.7 44.4 0.98 0.046 P < 0.05 was denoted in boldface in the table. a Heroin use career in this study refers to the period of years from first heroin use to this admission minus the period of years of continued abstinence outside drug treatment settings. b Amounts of heroin used per day over the 30 days before admission. c Poly-drug abuse was defined in this study as abusing or being dependent on at least one other substance (alcohol or illicit drugs) than heroin/opiates during 30 days before admission (with at least one current diagnosis of DSM-IV substance use disorder other than heroin/opioid use disorder). Nicotine use was not included in the analyses of this study. Table options 3.2. Bivariate correlates of co-occurring ASPD In bivariate analyses (Table 3), variables correlated with co-occurring ASPD among the total participants (males and females combined) were younger age, lower educational level, never married, unemployment, younger age at first heroin use, larger amounts of heroin used per day, poly-drug abuse, heroin injection, previous history of compulsory treatment, lifetime major depressive disorder, paranoid personality disorder, depressive personality disorder and borderline personality disorder (P < 0.05). For the male participants, variables correlated to co-occurring ASPD were quite the same as for the total participants, except for amounts of heroin used per day and depressive personality disorder, which were near the significant level (P < 0.1). While for the female participants, only four variables: younger age at first heroin use, poly-drug abuse, paranoid personality disorder and borderline personality disorder were found to correlate with co-occurring ASPD. Table 3. Bivariate odds ratio of participant characteristics to co-occurring ASPD by gender. Variable Subjects with ASPD among total participants Subjects with ASPD among males Subjects with ASPD among females OR 95% CI P OR 95% CI P OR 95% CI P Demographics Age, years 0.000 0.000 0.106 –30 1 1 1 31–40 0.37 0.24–0.58 0.000 0.54 0.37–0.78 0.001 0.53 0.27–1.08 0.079 41– 0.74 0.56–0.99 0.044 0.25 0.15–0.42 0.000 0.41 0.13–1.26 0.120 Education –Junior school 1 1 High school– 0.60 0.43–0.83 0.002 0.55 0.38–0.81 0.002 0.55 0.23–1.29 0.171 Marital state 0.000 0.000 0.760 Married or cohabiting 1 1 1 Divorced or widowed 0.79 0.53–1.19 0.256 0.94 0.58–1.51 0.786 0.87 0.35–2.19 0.769 Never married 1.80 1.30–2.49 0.000 2.37 1.61–3.49 0.000 1.16 0.52–2.59 0.718 Employment Employed 1 1 1 Unemployed 1.80 1.36–2.37 0.000 2.51 1.80–3.50 0.000 1.57 0.78–3.13 0.206 Personal income, ￥/year 0.243 0.067 0.164 < 30,000 1 1 1 30,000–100,000 0.99 0.72–1.37 0.952 0.66 0.46–0.95 0.026 0.658 0.22–1.97 0.455 > 100,000 1.42 0.93–2.16 0.103 1.04 0.62–1.73 0.884 2.094 0.86–5.10 0.104 Drug use Age at first heroin use, years –20 1 1 1 21– 0.46 0.35–0.60 0.000 0.30 0.21–0.43 0.000 0.35 0.18–0.71 0.003 Heroin use career, yearsa 0.098 0.235 0.602 < 5 1 1 1 5–10 1.58 1.04–2.39 0.031 1.50 0.91–2.47 0.108 1.55 0.59–4.10 0.377 > 10 1.37 0.93–2.08 0.113 1.44 0.91–2.28 0.124 1.14 0.46–2.84 0.772 Amounts of heroin useb, g/d < 0.5 1 1 1 ≥ 0.5 1.40 1.06–1.85 0.019 1.39 0.99–1.94 0.057 1.20 0.62–2.30 0.588 Heroin injection No 1 1 1 Yes 2.89 1.78–4.68 0.000 2.77 1.60–4.79 0.000 4.16 0.97–17.85 0.055 Previous history of compulsive treatment No 1 1 1 Yes 1.40 0.93–2.11 0.107 1.82 1.13–2.92 0.013 0.83 0.32–2.14 0.701 Poly-drug abusec No 1 1 1 Yes 1.82 1.38–2.39 0.000 1.83 1.31–2.55 0.000 2.26 1.19–4.30 0.013 Psychiatric co-morbidity Lifetime major depressive disorder 1.56 1.08–2.27 0.019 1.91 1.18–3.10 0.009 1.27 0.55–2.94 0.581 Current major depressive disorder 1.43 0.90–2.26 0.127 1.62 0.91–2.91 0.104 1.28 0.46–3.57 0.639 Dysthymia 0.98 0.53–1.82 0.948 0.95 0.47–1.95 0.897 0.49 0.06–3.87 0.497 Avoidant personality disorder 1.19 0.79–1.80 0.401 0.97 0.61–1.55 0.910 0.86 0.24–3.03 0.811 Passive–aggressive personality disorder 1.40 0.91–2.17 0.128 1.14 0.69–1.89 0.603 1.41 0.45–4.43 0.558 Paranoid personality disorder 3.08 1.81–5.24 0.000 2.09 1.11–3.91 0.022 8.57 3.01–24.45 0.000 Depressive personality disorder 1.99 1.18–3.33 0.009 1.70 0.91–3.19 0.096 2.68 0.89–8.13 0.081 Borderline personality disorder 2.51 1.82–3.48 0.000 2.50 1.65–3.79 0.000 4.02 2.04–7.91 0.000 P < 0.05 was denoted in boldface in the table. a Heroin use career in this study refers to the period of years from first heroin use to this admission minus the period of years of continued abstinence outside drug treatment settings. b Amounts of heroin used per day over the 30 days before admission. c Poly-drug abuse was defined in this study as abusing or being dependent on at least one other substance (alcohol or illicit drugs) than heroin/opiates during 30 days before admission (with at least one current diagnosis of DSM-IV substance use disorder other than heroin/opioid use disorder). Nicotine use was not included in the analyses of this study. Table options 3.3. Independent predictors for co-occurring ASPD In order to determine which factors were independently associated with co-occurring ASPD (with ASPD = 1, without ASPD = 0), multivariate logistic regressions with stepwise forward method were performed. Variables analyzed in the bivariate analyses that were significant or not significant were all listed as potential independent variables. Specifically, variables entered into the final model for the total participants were: gender (male = 1, female = 0), education (“high school or higher” = 1, “junior school or lower” = 0), employment (unemployed = 1, employed = 0), age at first heroin use (years, “≤ 20” = 1 “≥ 21” = 0), heroin injection (yes = 1, no = 0), poly-drug abuse (yes = 1, no = 0), paranoid personality disorder (yes = 1, no = 0), and borderline personality disorder (yes = 1, no = 0; Table 4), suggesting that social–economic disadvantages, a younger onset age of drug use, a more severe pattern of drug abuse, and psychiatric co-morbidities are involved. Table 4. Logistic regression model predicting co-occurring ASPD for the total participants. Variable Value β Wald df Odds ratio 95% confidence interval P Gender Male = 1, female = 0 2.29 121.55 1 9.86 6.56–14.81 0.000 Education “High school–” = 1, “–junior school” = 0 − 0.49 6.46 1 0.61 0.42–0.89 0.011 Employment Unemployed = 1, employed = 0 0.53 9.89 1 1.70 1.22–2.36 0.002 Age at first heroin use, years “21–” = 1, “–20” = 0 − 0.90 27.39 1 0.41 0.29–0.57 0.000 Heroin injection Yes = 1, no = 0 0.64 5.62 1 1.91 1.12–3.27 0.018 Poly-drug abuse Yes = 1, no = 0 0.52 10.23 1 1.68 1.22–2.32 0.001 Paranoid personality disorder Yes = 1, no = 0 0.68 4.80 1 1.98 1.08–3.65 0.028 Borderline personality disorder Yes = 1, no = 0 0.87 19.06 1 2.39 1.62–3.53 0.000 Constant – − 2.74 59.85 1 0.07 – 0.000 Model: χ2 = 260.89, df = 8, P = 0.000. Hosmer–Lemeshow χ2 = 3.67, df = 7, P = 0.817. Note: Referent group = participants (males and females combined) without ASPD. Table options Variables entered into the final model for the male participants were education (“high school–” = 1, “–junior school” = 0), never married (dummy variables: never married = 1, divorced or widowed = 1, married or cohabiting = 0), employment (unemployed = 1, employed = 0), age at first heroin use (years, “–20” = 1 “21–” = 0), poly-drug abuse (yes = 1, no = 0), lifetime major depressive disorder (yes = 1, no = 0) and borderline personality disorder (yes = 1, no = 0; Table 5), suggesting that related factors of co-occurring ASPD in the male participants are quite similar to those in males and females combined. Table 5. Logistic regression model predicting co-occurring ASPD for the male participants. Variable Value β Wald df Odds ratio 95% confidence interval P Education “High school–” = 1, “–junior school” = 0 − 0.50 5.39 1 0.61 0.40–0.93 0.020 Marital state Never married = 1, divorced or widowed = 1, married or cohabiting = 0 7.74 2 0.021 Divorced or widowed 0.02 0.01 1 1.02 0.61–1.71 0.931 Never married 0.54 6.20 1 1.72 1.12–2.64 0.013 Employment Unemployed = 1, employed = 0 0.58 9.37 1 1.78 1.23–2.57 0.002 Age at first heroin use, years “21–” = 1, “–20” = 0 0.86 18.64 1 2.37 1.60–3.52 0.000 Poly-drug abuse Yes = 1, no = 0 0.55 8.59 1 1.73 1.20–2.49 0.003 Lifetime major depressive disorder Yes = 1, no = 0 0.60 4.95 1 1.83 1.07–3.11 0.026 Borderline personality disorder Yes = 1, no = 0 0.76 10.94 1 2.15 1.37–3.37 0.001 Constant − 1.08 23.20 1 0.34 0.000 Model: χ2 = 105.12, df = 8, P = 0.000. Hosmer–Lemeshow χ2 = 8.58, df = 8, P = 0.379. Note: Referent group = male participants without ASPD. Table options Three variables entered into the final model for the female participants are: age at first heroin use (years, “–20” = 1, “21–” = 0), paranoid personality disorder (yes = 1, no = 0) and borderline personality disorder (yes = 1, no = 0; Table 6), without suggesting any relationship between socioeconomic disadvantages or severe pattern of drug abuse and ASPD in female participants. Table 6. Logistic regression model predicting co-occurring ASPD for the female participants. Variable Value β Wald df Odds ratio 95% confidence interval P Age at first heroin use, years “21–” = 1, “–20” = 0 − 0.90 5.94 1 0.41 0.20–0.84 0.015 Paranoid personality disorder Yes = 1, no = 0 1.60 7.88 1 4.93 1.62–15.03 0.005 Borderline personality disorder Yes = 1, no = 0 1.12 9.19 1 3.06 1.49–6.32 0.002 Constant − 1.80 50.41 1 0.17 0.000 Model: χ2 = 30.91, df = 3, P = 0.000. Hosmer–Lemeshow χ2 = 0.37, df = 3, P = 0.947. Note: Referent group = female participants without ASPD.