Childhood emotional abuse, dissociation, and suicidality among patients with drug dependency in Turkey
Aim: The aim of the present study was to determine the prevalence and correlates of dissociative disorders among patients with drug dependency.
Methods: The Dissociative Experiences Scale (DES) was used to screen 104 consecutive patients at an addiction treatment center. Thirty-seven patients who had scores ≥30 were compared with 21 patients who scored <10 on the DES. Both groups were then evaluated using the Dissociative Disorders Interview Schedule and the Structured Clinical Interview for Dissociative Disorders (SCID-D). The interviewers were blind to the DES scores.
Results: Twenty-seven patients (26.0%) had a dissociative disorder according to the SCID-D. Dissociative patients were younger than the non-dissociative group. History of suicide attempt and/or childhood emotional abuse was significant predictors of a dissociative disorder. The majority (59.3%) of dissociative drug users reported that dissociative experiences had existed prior to substance use. More patients in the dissociative disorder than in the non-dissociative group stopped their treatment prematurely.
Conclusion: A considerable proportion of drug users have a dissociative disorder, which may also interfere with treatment process. The relatively young age of this subgroup of patients and frequent reports of childhood emotional abuse underline potential preventive benefits of early intervention among adolescents with developmental trauma history and dissociative psychopathology.
PATIENTS WITH DISSOCIATIVE disorder report substance abuse frequently,1 while studies conducted on populations with substance dependency also demonstrated concurrent dissociative psychopathology. One study demonstrated that 39.0% of 100 chemically dependent patients had a dissociative disorder.2 A further study yielded a prevalence of 15.0% for 100 patients with substance use disorder at the end of their inpatient treatment.3 Alcohol and drug addiction occur in a large proportion of patients with dissociative identity disorder; drug abuse is severe and begins at an early age in many of them.1
The prevalence of substance use in Turkey is considered lower than that in Western Europe and North America. Most clinicians agree, however, about a recent increase of alcohol and drug use among adolescents and young adults in particular. A 1996 study among high-school students in Istanbul yielded a prevalence of 34.2% for alcohol use for the last month. The lifetime use for cannabis was 4.2%, for solvents 4.0%, for heroin 0.8% and for cocaine 0.8%.4 In a 1998 high school survey, the prevalence of alcohol use was 17.3% for the last month. Lifetime use for solvents was 8.2%, for cannabis 3.6%, for sedative hypnotics 3.3%, for heroin 1.6%, and for cocaine 1.4%.5 A replication study in 2001 yielded not only an increase in lifetime alcohol use (45.0%), but increases for heroin (2.5%) and for ecstasy (2.5%) as well.6 These data document clearly that the rise in both alcohol and use of hard drugs such as heroin among youngsters in Turkey is alarming in scope.
A relationship is consistently reported between childhood trauma and dissociation both in the general population and among psychiatric patients.7–9 Reports of childhood abuse and neglect are also common among substance dependents.10 Dependent inpatients with a history of distressing traumatic event report not only higher levels of dissociation, but more self-mutilative acts, and a greater degree of impulsivity than did patients without such histories.11 In fact, general psychiatric outpatients with higher scores on the DES are likely to have attempted suicide also.12 Among substance users, patients with higher childhood trauma scores attempt suicide at an earlier age and tend to have a greater number of suicide attempts.13
The aim of the present study was therefore to determine the prevalence of dissociative disorders among consecutive inpatients with drug dependency and its correlates. In order to minimize possible effects of drug use, all interviews were carried out after completion of the detoxification period. Conducted in a blind fashion, both self-rating and clinician-administered assessment tools were used in the present study. In order to determine the predictors of the disorder, we compared patients with and without dissociative disorders on clinical and sociodemographic variables including psychiatric comorbidity and childhood trauma history.
The study was conducted at the Alcohol and Substance Research, Treatment and Training Center (AMATEM) of the Bakirkoy Research and Training Hospital for Psychiatric and Neurological Diseases in Istanbul. As the largest center specialized for treatment of drug and alcohol use disorders in Turkey, the AMATEM accepts referrals from all over the country. All admitted patients in a 1-year period (1 March 2003–31 March 2004) who were able to complete the detoxification period and entered the treatment program were considered for participation in the study.
All study interviews were conducted after completion of detoxification period, that is, 2–4 weeks after the last day of substance use. A clinician decided if withdrawal symptoms had disappeared. Exclusion criteria were current epilepsy, mental retardation, cognitive deficit, illiteracy and current acute psychotic disorders. The comparison subjects were drawn from the same patient population. After complete description of the study to the patients, written informed consent was obtained.
A total of 111 consecutive patients with drug dependency were able to complete the detoxification period during the study period. Seven of these patients were excluded from the study due to current epilepsy (n = 2), current acute psychosis (n = 2), or illiteracy (n = 3).
The Dissociative Experience Scale (DES) is a 28-item self-rating scale of good reliability and validity.14,15 It is not a diagnostic tool but serves as a screening device for dissociative disorders. Possible scores range from 0 to 100. The Turkish version of the DES16 has good reliability and validity.17,18 A cut-off score of 30 has been shown to be useful in screening dissociative disorders in Turkey.19
The Dissociative Disorders Interview Schedule (DDIS) is a structured clinical interview consisting of 131 items.20 It was designed to diagnose somatization, major depression, borderline personality disorder and five classes of dissociative disorders according to the DSM-IV. The schedule also inquires about childhood abuse and neglect and a variety of features associated with dissociative identity disorder, including 11 Schneiderian symptoms, 16 secondary features of dissociative identity disorder, and 16 extrasensory experiences. The validity and reliability of the Turkish version (unpublished 1993/1994 translation by V. Sar et al.) has been reported elsewhere.19
The Structured Clinical Interview for DSM-IV Dissociative Disorders (SCID-D) is a semi-structured diagnostic interview, and investigates five dissociative disorders according to DSM-IV criteria and also rates five symptom areas (depersonalization, derealization, amnesia, identity confusion, and identity alteration) of dissociation and systematically rates the severity of individual symptoms.21 Information about reliability and validity of the Turkish version of the instrument (unpublished 1996 translation by V. Sar et al.) has been reported elsewhere.22
A structured history form designed for this study was given to each patient, and it included sociodemographic data, history of alcohol and drug use and clinical features including the sequence of onset of dissociative experiences and substance use. We also determined the dropout rates during the index hospitalization, that is, the number of patients who did not complete the treatment program.
The study consisted of two phases. In the first phase, all patients were asked to complete the DES and the history form. One psychiatrist (C.E.) and one psychiatry resident (M.K.) collected these data. All subjects who scored >30 or <10 on the DES were referred to the second phase of the study. A previous study in Turkey demonstrated that these cut-off scores classify dissociative and non-dissociative patients at the highest sensitivity and specificity.19
In the second phase of the study, the DDIS and the SCID-D were administered to patients in both groups by two psychiatrists. The interviewers were blind to the patients' diagnoses and DES scores. Two of the authors (F.K. and D.T.G.) conducted all structured interviews, both of whom had experience using these instruments before the initiation of the study.
A total of 104 patients participated in the study. Preference drugs reported were cannabis (n = 33), solvents (n = 27), heroin (n = 21), ecstasy (n = 18), and anti-cholinergic drugs or cocaine (n = 5). Only 11 patients had monodrug dependency, including cannabis (n = 1), heroin (n = 4), ecstasy (n = 1), and solvents (n = 5). The remaining 93 patients had alcohol dependency as well. The mean age of the participants was 25.6 ± 6.2 years (range, 17– 46 years). Four patients (3.8%) were under 18 years of age. Only 13 patients were female (12.5%). Women were younger (mean, 21.2 ± 3.9 years) than men (mean, 26.2 ± 6.3 years; z = 2.90, P = 0.004).
The mean DES score of the patients was 29.0 ± 18.2. There was no significant difference between women (mean, 35.7 ± 15.4) and men (mean, 28.1 ± 18.5) on the scale (z = 1.57, P = 0.117). Age correlated negatively with scale score (r = −0.20, n = 104, P = 0.038). Forty-eight patients (46.2%) had a score >30. Thirty-nine of them (81.3%) were men and nine (18.7%) were women, these rates were 52 (92.9%) and four (7.1%) for the remaining, respectively; that is, there was no significant differences in gender distribution for a high scorer or not (χ2 = 3.18, d.f. = 1, P = 0.074).
Thirty-seven (77.1%) of the 48 patients with scores >30 and 21 (87.5%) of the 24 patients with scores <10 were considered for an evaluation using structured interviews. Eleven (22.9%) high scorers and three (12.5%) low scorers were excluded from the second phase because they were hospitalized for too short a period (χ2 = 1.11, d.f. = 1, P = 0.292). Thus, the second phase of the study consisted of 58 patients including 10 women (17.2%).
A total of 27 patients had a dissociative disorder according to the SCID-D, yielding an overall prevalence of 26.0% among participants who completed the interview. Six patients (5.8%) had dissociative identity disorder. One patient (1.0%) had dissociative fugue, and one (1.0%) had depersonalization disorder. Nineteen patients (18.3%) had dissociative disorder not otherwise specified. Eleven of them had distinct personality states without fitting the criteria of dissociative identity disorder fully. Four patients had a combination of amnesia and depersonalization, two had amnesia and derealization, one had derealization without depersonalization, and one had dissociative fugue, depersonalization and derealization.
Mean age of onset of dissociative experiences was 15.1 ± 6.0 for dissociative patients. According to the responses to the SCID-D items on the chronology of symptom formation, 16 patients (59.3%) with dissociative disorder had dissociative experiences prior to onset of substance use. Thus, for 16 patients dissociative experiences had started 3.9 ± 3.2 years before onset of substance use, on average.
In accordance with the diagnoses obtained on SCID-D, the patients with a dissociative disorder had significantly higher scores in three main symptom clusters of the DDIS (Table 1). These three symptom clusters were number of associated features of dissociative identity disorder, extrasensory perception experiences, and Schneiderian symptoms. There were no significant difference between two groups on borderline personality disorder criteria and somatoform symptoms.
Table 1. Inpatients with drug dependency according to dissociative disorder status
|Age (years)||24.6||6.8||27.4||6.4||1.98|| 0.048 |
|Age of onset of substance use||16.4||5.4||16.0||4.9||0.50||0.620|
|Longest remission (months)||2.2||2.2||7.8||15.4||1.30||0.194|
|Types of substance use||2.8||1.6||2.8||1.4||0.03||0.974|
|DES score||40.9||16.0||23.5||20.6||3.03|| 0.002 |
|DDIS Main symptom clusters (no. symptoms)|
|Features associated with dissociative identity disorder||6.7||4.0||2.0||2.8||4.42||0.001|
|Borderline personality disorder criteria||5.8||2.7||4.8||2.9||1.39||0.164|
Patients with a dissociative disorder were younger than the non-dissociative patients, but there were no differences in age of onset of substance use, the longest remission duration (months) and number of substance use types between the two groups (Table 1). There were no differences on gender, comorbid major depression, borderline personality disorder, or somatization disorder either, but dissociative group had higher prevalence of suicide attempts and a history of childhood emotional abuse. Significantly more patients in the dissociative disorder group left the treatment program prematurely (Table 2).
Table 2. Comorbid diagnoses, suicide attempt and childhood trauma history among dependent inpatients vs dissociative disorder status
|Borderline personality disorder||20||74.1||18||58.1||1.64||0.201|
|Major depression (lifetime)||18||66.7||21||67.7||0.01||0.931|
|Major depression (current)||1||5.6||3||14.3||0.80||0.609|
|One of the above||23||85.2||23||74.2||1.06||0.303|
|Premature cessation of treatment||15||55.6||9||29.0||4.2||0.041|
To assess the relative abilities of the variables to predict dissociative disorder, multivariate logistic regression was used, with presence of dissociative disorder as the dependent variable. Independent variables were age, gender, suicide attempt, somatization disorder, major depression, physical abuse, emotional abuse, sexual abuse, neglect, and borderline personality disorder. History of suicide attempt and of childhood emotional abuse were the only significant predictors of dissociative disorder diagnosis (Table 3).
Table 3. Predictors of dissociative disorder on backward stepwise multiple regression
|Suicide attempt||−1.27||0.59||4.61||1||0.032||0.28 (0.09–0.90)|
|Emotional abuse||−1.18||0.58||4.17||1||0.041||0.31 (0.10–0.95)|
|Sexual abuse||1.94||1.38||2.00||1||0.158||7.00 (0.47–103.8)|
|Major depression||1.02||0.75||1.84||1||0.175||2.77 (0.64–12.06)|
|Somatization disorder||0.35||1.05||0.11||1||0.738||1.42 (0.18–11.21)|
|Physical abuse||−0.12||0.67||0.03||1||0.859||0.89 (0.24–3.27)|
|Borderline personality disorder||−0.02||0.93||0.00||1||0.960||0.98 (0.16–6.09)|
Twenty-four (64.9%) of the 37 high scorers on the DES were diagnosed as having a dissociative disorder according to the SCID-D. This rate was three (14.3%) of the 21 patients with scores <10.0 (χ2 = 13.77, d.f. = 1, P < 0.001). When the SCID-D is taken as the gold standard for diagnosis, sensitivity and specificity of the DES were 88.9% and 58.1%, respectively, positive predictive value was 64.9% and negative predictive value was 85.7%. Although it is a highly sensitive screening tool, specificity of the DES for dissociative disorder diagnosis is relatively low among patients with drug abuse.
A sizeable proportion (24.3%) of detoxified inpatients (26.0% of the patients who received the interview) with drug dependency had a dissociative disorder in the present study. This rate is considerably higher than the dissociative disorder prevalence among general psychiatric inpatients in Turkey (10.2%) obtained in a previous study.23 The present findings are not at odds with those yielded in three screening studies conducted in North America: 41.0% of substance-dependent inpatients had high scores of dissociation,24 and the prevalence of dissociative disorders was 15.0% and 39.0% among substance users.2,3
Previous studies on case series gathered from general psychiatric settings demonstrated that patients with dissociative disorder have high comorbidity of somatization disorder, borderline personality disorder, and major depression.23,25–28 There was no difference between patients with and without dissociative disorder in the present study for comorbidity of these disorders (Table 2). Thus, dissociative disorders among dependent patients cannot be considered merely as an epiphenomenon of comorbid conditions in the present study. Moreover, a considerable proportion (59.3%) of the patients with a dissociative disorder in the present study reported that dissociative experiences had started before drug use. Thus, dissociative disorders among drug-dependent patients may not be a phenomenon limited to cross-sectional observation or to a crisis period.29
Nevertheless, patients with dissociative disorder were overrepresented among dropouts from the treatment program. Patients with comorbid dissociative disorders were overrepresented among dropouts from a treatment program in the present study as reported previously.30 Overlooking of the dissociative disorder in these patients can be a handicap for their treatment.31 It is probable that patients with dissociative disorder experience shifts in personality states during inpatient stay, which usually leads to self-destructive behavior such as leaving the treatment prematurely. The recognition of the dissociative disorder by attending clinicians would prevent dropout through facilitating communication by working through dissociative barriers and establishing contracts targeted directly at self-destructive aspects of the patient.32
Although significant relations have been reported between dissociation and childhood physical and/or sexual abuse or a lifetime history of sexual abuse among substance users,7,33 there are also studies conducted on substance users that reported no link between childhood trauma and dissociation.2,3,23,34–36 Nevertheless, lifetime psychological trauma is significantly associated with worse psychiatric status, more psychiatric hospitalizations, and more outpatient treatment despite receiving similar intensive addiction treatment.37 Although a high proportion (85.2%) of dissociative patients in the present study reported at least one type of childhood trauma, this rate was not significantly higher than that of the non-dissociative group. Among specific childhood trauma types, however, emotional abuse was more common among dissociative group.
In the present study, beside childhood emotional abuse, a history of suicide attempt was more frequent in the dissociative group. Regression analysis also demonstrated that only history of childhood emotional abuse and of suicide attempt predicted dissociative disorder diagnosis. In one recent screening study, suicidality was demonstrated to be associated with dissociative disorder independent of an associated personality disorder.38 In contrast, childhood emotional abuse has been found to be a predictor of dissociation among patients with conversion disorder25 and depersonalization disorder.39 Thus, childhood emotional abuse, dissociation, and suicidality comprise a triad probably not limited to subjects with substance dependency. Rather, this triad may provide a predisposition to drug abuse as a desperate self-soothing strategy.
Drug abuse has been in tremendous increase in Turkey among adolescents and young adults in last two decades. The prevalence of dissociative disorders is also high in Turkey and many of these subjects remain either not treated at all,9 except for crisis intervention due to visits to emergency psychiatric wards,40 or they are underdiagnosed.41 The risk of overlooking a dissociative disorder is highest among children and adolescents due to symptom overlap with other disorders,42,43 and even with normative adolescent behavior such as impulsivity.44 Thus, appropriate specific psychotherapeutic intervention is usually postponed to early adulthood for patients with a dissociative disorder, which may be one of the hidden factors behind the rise of substance use among adolescents in Turkey.
Some study limitations should be noted. First, we recruited patients from an inpatient treatment program, therefore the results cannot be generalized to outpatients and non-treatment groups in Turkey. Second, there may also be potential measurement problems due to the retrospective recall of child abuse and neglect, which may reduce the reliability of assessment. Finally, this study was cross-sectional; thus, longitudinal studies are necessary to make causal attributions about the relationship between childhood trauma, dissociative disorders, and substance dependence.
In conclusion, although rarely taken into consideration by clinicians, dissociative disorders have an important place in the drug dependency field. Dissociative disorders should not be considered merely as a consequence of drug use in this population, but they may play an important role in prevention and treatment. Premature cessation of treatment by a significant part of dissociative patients in the present study clearly demonstrates that dissociative disorders may have negative influences on outcome.29 Carefully designed specific intervention strategies may change this fate for this little understood and usually neglected patient group.