Eur Arch Psychiatry Clin Neurosci (1993) 242 : 362-372 ,o eao P chiatry Archives of and Clinical Neuroscience (cid:14)9 Springer-Verlag 1993 Prevalence of Alcohol and Drug Abuse in Schizophrenic Inpatients M. Soyka 1, M. Albus 2, N. Kathmann x, A. Finelli 1, S. Hofstetter 1, R. Holzbach 1, B. Immler 1, and P. Sand 1 1psychiatric Hospital of the University of Munich, NussbaumstraBe 7, W-8000 Munich 2, Germany 2Mental State Hospital Haar/Munich, Germany Received December 7, 1992 Summary. All schizophrenic patients admitted consecu- tively either to the Psychiatric Hospital of the University of Munich (group 1, N = 183) or the Mental State Hospi- tal Haar/Munich (group 2, N= 447) between 1.8.1989 and 1.2.1990 were examined to assess prevalence esti- mates for substance abuse in schizophrenic inpatients. Psychiatric diagnosis were made according to ICD-9 criteria. Psychopathology and psychosocial variables were documented by means of the AMDP-protocol on admission and discharge. The diagnostic procedure in- cluded a detailed semi-structured interview concerning the individual alcohol and drug history and sociodemo- graphic data, the Munich Alcoholism Screening Test (MALT), a physical examination and the screening of various laboratory parameters such as GGT and MCV, among others. The results show that substance abuse is a very com- mon problem in schizophrenics. Lifetime prevalence rates for substance abuse were estimated at 21.8% in group 1 and 42.9% in group 2, 3-month prevalence rates for substance abuse were estimated at 21.3% resp. 29.0%. Alcohol abuse was by far the most common type of abuse with prevalence estimates being 17.4% resp. 34.6%. Prevalence rates for substance abuse were much higher in the more "chronic" sample of the Mental State Hospital and in male patients. With respect to schizo- phrenic subtype few differences could be demonstrated with drug dependence being more common in patients with paranoid schizophrenia. The MALT proved to be a valuable sceening instrument for alcohol abuse in schizo- phrenics with both a high specifity and sensitivity. "Dual diagnosis" schizophrenics had a significantly higher rate of suicide attempts and were less likely to be married. Pos- sible clinical implications of these findings are discussed. Key words: Schizophrenia - Alcoholism - Substance abuse Many clinical and epidemiological studies prove that alcoholics often suffer from an additional psychiatric dis- Correspondence to: M. Soyka order such as depression, personality and anxiety dis- order (Hesselbrock et al. 1985; Hirschfeld et al. 1989; Hasin et al. 1989; Roy et al. 1991a,b; Schuckit 1986a, b). A possible correlation between substance abuse and schizophrenia has long been suspected. In his 1975 re- view Freed reported prevalence rates for alcohol abuse in schizophrenia ranging from 3 to 63%. More recently, several authors (Alterman et al. 1980-1982; Test et al. 1985, 1989; Mueser et al. 1990) further emphasized this issue and reported a high comorbidity of alcohol or drug abuse and schizophrenia. Open questions concerning this topic have been addressed by Kesselmann et al. (1982) and Pulver et al. (1989) who pointed at the lack of valid epidemiological and clinical data on the prevalence of alcohol and drug abuse in schizophrenia, the temporal relationship between the onset of schizophrenia and an additional abuse as well as the specific psychopathology in "dual diagnosis" patients. Both clinical and epidemio- logical studies suggest that drug, especially cannabis, as well as alcohol abuse might be a risk factor for develop- ing psychosis (Knudsen and Villmar 1984, Tsuang et al. 1982, Tien and Anthony 1990, Eikmeier et al. 1991). In previous studies, the lifetime prevalence of alcohol abuse/dependence in schizophrenia was estimated at 12.3% (Alterman et al. 1981), 14.2% (McLellan and Druley 1977), 14.8% (Drake et al. 1989 for alcohol dependence), 23.0% (O'Farell et al. 1983) up to 42.8% (Barbee et al. 1989, alcohol dependence in male schizo- phrenics). Bernadt and Murray (1986) reported a preva- lence for alcoholism in schizophrenics of 8.8% in the year preceeding survey and 6.3% for schizoaffective psychosis. Drake et al. (1990) found a lifetime diagnosis of an "alcohol use disorder" of 50.3% in schizophrenic outpatients. In a recent study, Mueser et al. (1990) reported a prevalence rate for alcohol abuse of 33.0% and a lifetime prevalence of 47.0%. A possible relationship between substance abuse and psychosis can be considered in different ways: 1. Alcohol or drug abuse could cause a schizophrenia- like psychosis. For example McLellan et al. (1979) re- ported a higher prevalence rate for psychosis among amphetamine users when compared with barbiturate users. Andreasen et al. (1987) reported a six-fold rate of
`
`CFAD VI 1023-0001
`
`

`

`363 a schizophrenic psychosis among Swedish cannabis users as opposed to non-users. 2. A preexisting psychosis may predispose to an addi- tional alcohol or drug abuse by which the psychiatric dis- order can be masked (Bagley and Binitie 1970) leading to serious problems in differential diagnosis. Some au- thors suggested the "self-medication hypothesis" as an explanation for substance abuse in affective disorder and schizophrenia. 3. Substance abuse in schizophrenic patients can cause additional psychiatric symptoms such as anxiety, delu- sional and hallucinatory symptoms and influence course of illness. 4. Treatment-relevant variables such as social integra- tion and compliance could be significantly impaired by an additional alcohol or drug abuse in schizophrenic pa- tients (Miller and Tanenbaum 1989; Kofoed et al. 1986; Drake and Wallach 1988). This is supported by the find- ing that among city shelters, patients with substance abuse and psychotic patients predominate (Priest et al. 1985; Susser et al. 1989). These findings together with the obvious lack of valid data on the co-morbidity of alcohol/substance abuse and schizophrenia were reasons for us to perform a prospec- tive study on the prevalence of alcohol and drug abuse in schizophrenic inpatients as well as on the psycho- pathology, psychosocial integration and other socio- demographic variables in "dual diagnosis" patients. This article will focus on the prevalence of alcohol/drug abuse in schizophrenia and possible clinical implications. Methods All psychiatric inpatients admitted consecutively either to the Psy- chiatric Hospital of the University of Munich or the Mental State Hospital Haar between 1. August 1989 and 1. February 1990 were included in the study. During that period 183 patients with a schizophrenic psychosis were admitted to the former (= group 1) and 447 schizophrenic patients to the latter (= group 2). The psychiatric diagnoses were made according to ICD-9 criteria; psychopathology and sociodemographic data were assessed by means of the AMDP protocol (Guy and Ban 1982) on admission and discharge. In addition, the alcohol and drug abuse was as- sessed by means of a detailed semi-structured interview concerning all relevant variables such as the individual drinking history, past psychiatric admissions and somatic as well as psychosocial vari- ables possibly affected by either alcohol or drug abuse. Details of the diagnostic instruments and statistical analyses have been de- scribed elsewhere (Soyka et al. 1992). Furthermore, the "Munich Alcoholism Test", a diagnostic instrument with both high reliabil- ity and validity for alcoholism (Auerbach and Mechertsen 1981, Gorenc et al. 1984, Feuerlein et al. 1977) was performed for each patient. The diagnostic process included a somatic and neurologi- cal examination and the screening of various laboratory paramet- ers such as GGT, GOT, GPT and MCV on admission. Table 1. Patient characteristics Group i Group 2 N N Number of patients 183 447 Male/female ratio 70/113 237/210 Paranoid subtype 56 (30.6%) 196 (43.8%) Male/female ratio 19/ 37 117/ 79 Schizoaffective 55 (30.0%) 112 (25.1%) Male/female ratio 19/ 36 39/ 73 Residualtype 27 (14.7%) 75 (16.8%) Male/female ratio 9/ 18 34/ 41 Mean age (years) 36.1 (18-73) 38.9 (17-78) Males 32.2 (19-73) 37.2 (17-78) Females 38.5 (18-71) 40.9 (19-73) Duration of psychosis (years) 7.3 (1-36) 11.2 (1-50) No. psychiatric admissions 4.2 (1-16) 5.9 (1-40) Sociodemographic characteristics Married 48 (25.3%) 76 (17.0%) Employed 128 (69.9%) 212 (47.4%) Housewife 24 (13.1%) 46 (10.2%) Unemployed/retired/other 31 (16.9%) 189 (42.2%) resented a more "chronic" patient group with a higher rate of past psychiatric admissions, among others, and was therefore analysed separately from group 1, the sample derived from the psychiatric hospital at the Uni- versity of Munich. Prevalence Rates The prevalence rates for any kind of addiction and abuse of or dependence on legal or illegal drugs were estimated as lifetime and 3-month prevalence. As demonstrated in Table 2 and 3, prevalence rates in group 2 were much higher than in group i with alcohol abuse or dependence being by far the most common kind of addiction. Alco- hol and drug abuse were significantly more common in male than in female patients in both groups. The lifetime prevalence rate for any kind of addiction (abuse or de- pendence) was 21.8% in group 2 and 42.9% in group 2, the prevalence rate for females was 15.0% resp. 31.4%. The lifetime and three-month-prevalence rates for any kind of dependence were much lower. When comparing the estimated prevalence rates in both groups the higher percentage of male patients in group 2 must be taken into consideration. Looking only at the male subgroup in both groups, the 3-month preva- lence rates for alcohol abuse were both 30%. The "Munich Alcoholism Test" identified 81.2% resp. 90.3% of the schizophrenics with alcohol abuse/de- pendence correctly. False positive results were very rare (Table 4). Abuse Pattern Results Patient characteristics are given in Table 1. Group 2, the sample derived from the Mental State Hospital, rep- The prevalence rates in group 1 being very low, the abuse pattern of schizophrenic patients only of group 2 is given in detail. Apart from ethanol, cannabis, cocaine and amphetamines were the most commonly abused il-
`
`CFAD VI 1023-0002
`
`

`

`364 Table 2. Prevalence of alcohol and drug abuse/dependence in schizophrenic inpatients - Results of Group I Diagnoses Total Male Female Signifi- cance N (%) N (%) N (%) p Total 183 100 70 100 113 100 Any abuse/dependence Lifetime 40 21.8 23 32.8 17 15,0 <0.01 3-Month 39 21.3 22 31.4 17 15.0 <0.01 Dependence only Lifetime 13 7.1 8 11.4 5 4.4 ns 3-Month 12 6.5 7 10.0 5 4.4 ns Alcohol abuse/dependence Lifetime 32 17.4 21 30.0 11 9.7 <0.001 3-Month 32 17.4 21 30.0 11 9.7 <0.001 -dependence only Lifetime 9 4.9 5 7.1 4 3.5 ns Now 9 4.9 5 7.1 4 3.5 ns Abuse/dependence of legal drugs Lifetime 11 6.0 5 7.1 6 5.3 ns 3-Month 10 5.5 4 5.7 6 5.3 ns -dependence only Lifetime 4 2.1 3 4.2 1 0.9 ns 3-Month 3 1.6 2 2.8 1 0.9 ns Abuse/dependence of illegal drugs Lifetime 10 5,5 6 7.1 5 4.4 ns 3-Month 10 5.5 5 7.1 5 4.4 ns -dependence only Lifetime 2 1.1 2 2.8 0 - ns 3-Month 2 1.1 2 2.8 0 - ns Diagnoses Total Male Female Signifi- cance P Table 3. Prevalence of alcohol and drug abuse/dependence in schizophrenic inpatients - Results of Group II N (%) N (%) N (%) Total 447 100 237 100 210 100 Any abuse/dependence Lifetime 192 42.9 126 53.1 3-Month 130 29.0 96 40.5 Dependence only Lifetime 96 21.5 64 27.0 3-Month 59 13.2 46 19.4 Alcohol abuse/dependence Lifetime 155 34.6 109 46.0 3-Month 93 20.8 71 30.0 -dependence only Lifetime 60 13.4 42 17.7 3-Month 34 7.6 27 11.4 Abuse/dependence of legal drugs Lifetime 66 14.7 33 13.9 3-Month 39 8.7 24 10.1 -dependence only Lifetime 29 6.4 14 5.9 3-Month 13 2.9 7 3.0 Abuse/dependence of illegal drugs Lifetime 64 14.3 51 21.5 3-Month 51 11.4 44 18.6 -dependence only Lifetime 30 6.7 25 10.5 3-Month 23 5.1 21 8.9 66 31.4 <0.001 34 16.2 <0.001 32 15.2 <0.01 13 6.1 <0.001 46 21.9 <0.001 22 10.5 <0.001 18 8.6 <0.01 7 3.3 <0.01 33 15.7 ns 15 7.1 ns 15 7.1 ns 6 2.9 ns 13 6.2 <0.001 7 3.3 <0.001 5 2.4 <0.001 2 1.0 <0.001
`
`CFAD VI 1023-0003
`
`

`

`Table 4. Results of the "Munich Alcoholism Test" 365 Group I Group II N (%) N (%) Total N 183 100 447 100 Patients with score 11 or more points 23 12.5 108 24.2 "False positive" results (score -> 11 points) 2 1.1 5 1.1 Patients with alcohol abuse/dependence (lifetime diagnosis) 32 100 155 100 Score >- 11 points 21 65.6 103 66.5 Patients with alcohol abuse/dependence (3-month-prevalence) 32 17.4 93 20.8 Score -> 6 und < 11 points 5 15.6 15 16.1 Score -> 11 points 21 65.6 69 74.2 Score _> 6 points 26 81.2 84 90.3 Average score in patients with alcohol abuse/dependence 19.30 ((cid:127) 15.40 ((cid:127) 8.03) Munich Alcoholism Test: Score -> 11 points confirming alcoholism, score _> 6 < 11 points suggesting alcoholism opiates opiate-analoga benzodiazepines clomethiazol other hypnotics analgesics amphetamines amphetamines 2 cannabis halluzinogens petrol sniffing etc. cocaine polytoxicomania alcohol barbiturates m 13-month prevalence ~"lifetime" diagnosis I I 0 20 40 60 % Fig. 1. Abuse pattern in "dual diagnosis" schizophrenics (results of group 2). 1 stimulant, 2 weight reducer legal drugs (see Fig. 1). The number of patients taking hallucinogenes was surprisingly high. Among legal sub- stances, benzodiazepines were by far the most com- monly abused drug. While 89 of the 155 patients with a lifetime history of alcoholism had no history of an addi- tional drug abuse, an isolated drug abuse was rarely found (Fig. 2). None of the patients abused antiparkin- sonian medication. Prevalence Rates in Schizophrenic Subgroups In both groups of schizophrenic patients the paranoid subtype was the most common schizophrenic psychosis, followed by schizoaffective and residual type (Table 1). While for the schizophrenic patients in group i no statis- tically significant differences could be demonstrated, schizophrenic patients with paranoid subtype from the larger group 2 did have a significantly higher lifetime (9.6% vs. 4.3%, P < 0.05, Table 5) and 3-month (8.2% vs 2.7%, P < 0.01) prevalence rate for drug dependence and a higher 3-month prevalence for any kind of depen- dence (16.6% vs 10.8%, P = 0.08) than the other schizo- phrenics in this group, whereas no significant differences could be demonstrated for the variables such as alcohol abuse as listed in detail in Tables 2-4. Patients with schizoaffective psychosis did not differ significantly from other schizophrenics. Patients with chronic (residual) type schizophrenia had the highest lifetime prevalence rates for any kind of abuse/dependence (54.7% vs. 40.6%, P < 0.05) in comparison with non-residual schizophrenics. Order of Onset of Psychosis and Addiction Firstly, a drug abuse was found nearly exclusively in schizophrenics younger than 30 years. All 11 patients with drug abuse in group 1 and 43 of the 64 patients in group 2 were younger than 30 years. Secondly, a detailed analysis of the data of the larger group 2 showed that the onset of addiction in general was later than the onset of a psychosis, especially in female patients (Table 6). An additional substance abuse more often followed than preceeded onset of psychosis. The onset of psychosis had to be defined retrospectively and was estimated at the first time positive or negative
`
`CFAD VI 1023-0004
`
`

`

`366 fl is 3 Fig. 2. Comorbidity of alcohol and substance abuse lifetime diag- noses in 192 schizophrenic inpatients (group 2); 1 abuse of legal drugs (n = 66); 2 alcohol abuse (n = 155); 3 abuse of illegal drugs (n = 64) Table 5. Differences in prevalence rates for alcohol/substance abuse among schizophrenic subtypes (results of Group 2) Paranoid Other P subtype schizophrenics N (%) N (%) 196 100 251 100 Any kind of addiction Lifetime diagnosis 32 16.3 27 10.8 0.08 Drug addition Lifetime diagnosis 19 9.6 11 4.3 <0.05 3-Monthprevalence 16 8.2 7 2.7 <0.01 Residual Other type schizophrenics 75 100 372 100 Any kind of abuse Lifetime diagnosis 41 54.7 151 40.6 <0.05 symptoms were reported. Since many patients had been admitted repeatedly to psychiatric hospitals the past psy- chiatric history could be assessed quite accurately. With respect to alcohol abuse (155 patients) in 85 (54.8%) of the cases psychosis preceeded, in 62 (40.0%) followed, the onset of alcohol/drug abuse, in 8 (5.2%) patients both disorders had a simultaneous onset. In patients with onset of psychosis prior to begin of alcohol abuse (mean age 38.5 + 11.5 years) onset of psychosis had been 14.8 (+ 10.2) years, onset of abuse 4.9 (+ 4.5) years before examination. In patients with alcohol abuse prior to onset of psychosis (mean age 38.2 + 10.9 years), onset of abuse had been 11.9 (_+ 8.2) years, onset of psychosis 5.3 (+ 5.6) years before examination. Interestingly, 36% of the 74 patients with first mani- festation of schizophrenic psychosis fullfilled the criteria for lifetime diagnosis of an additional alcohol/drug abuse, the 3-month prevalence rate was 31%. Alcohol abuse/dependence (23%) was more common than abuse of illegal (13%) or legal drugs (13%). Sociodemographic Parameters When compared with patients without addiction (Tables 7 and 8), both the numbers of married individuals and descended children were found to be markedly smaller among schizophrenics with an additional alcohol/drug abuse. Also, patients with a history of alcohol/drug abuse had made significantly more suicide attempts than those without (Table 9) and had been homeless signifi- cantly more often in the past. Discussion The results of the study suggest a high prevalence rate for alcohol and drug abuse among schizophrenic patients. Table 6. Age at onset of psychosis, duration of illness and number of psychiatric admissions in schizophrenics with and without the diagnosis of an alcohol/drug abuse (results of group 2) Schizophrenics P With abuse Without abuse N N Age at onset of psychosis (years) in patients with - Any kind of abuse (lifetime) - With alcohol abuse (N = 155) - Any kind of abuse (3-Month-prevalence, N = 130) Alcohol abuse (N = 93) Illegal drugs (N = 51) Duration of illness (years) Age at examination (years) Age at begin of abuse (years) - Alcohol abuse - Illegal drugs Duration of abuse at examination (years) - All patients - Alcohol abuse - Drug abuse No. of psychiatric admissions 27.1 (_+10.4) 28.4 (_+9.9) 27.4 (_+ 10.3) 27.3 (+ 10.9) 28.0 (+- 10.8) 20.8 (_+4.8) 10.8 (_+9.5) 37.9 (+ 11.5) 29.7 (+ 10.7) 30.3 (+ lO.3) 22.0 (+_4.6) 8.3 (+ 6.9) 8.9 (+7.6) 5.9 (_+3.4) 5.79 (+ 6.1) 11.4 (+9.9) 39.8 (+ 13.2) 4.60 (+- 5.7) <0.05
`
`CFAD VI 1023-0005
`
`

`

`Table 7. Sociodemographic parameters in schizophrenic inpatients of Group II with and without addiction (3-month-prevalence) Schizophrenics With Without Signifi- addiction addiction cance P N (%) N (%) 130 100 317 a 100 Martial status Single 86 66.2 176 57.0 ns Married yet 10 7.7 66 21.4 <0.001 Divorced 25 19.2 50 16.2 ns Separated 4 3.1 4 1.3 ns Widowed 5 3.8 13 4.2 ns Number of children None 95 73.1 199 63.0 <0.05 One 16 12.3 65 20.6 <0.05 Two 11 8.5 29 9.2 ns More than two 8 6.1 24 7.2 ns Homelessness Prior 54 41.5 68 21.4 <0.05 On admission 17 13.1 28 8.8 ns a In this subgroup no valid data available in 8 patients. Statistics and level of significance only given for patients with complete data Table 8. Sociodemographic parameters in schizophrenic patients of Group II with and without "lifetime" diagnosis of addiction Schizophrenics With Without Signifi- addiction addiction cance P N (%) N (%) 192 100 255 a 100 Martial status Single 126 65.6 136 55.1 <0.05 Married 18 9.4 58 23.5 <0.001 Divorced 36 18.8 39 15.8 ns Separated 5 2.6 3 1.2 ns Widowed 7 3.6 11 4.5 ns Numb er of children None 134 69.8 160 62.7 ns One 32 16.7 49 19.2 ns Two 16 8.3 24 9.4 ns More than two 10 5.2 22 8.7 ns Homelessness Prior 72 37.5 50 19.5 <0.01 At admission 22 11.5 23 9.0 ns a In 8 cases incomplete data available 367 The 3-month and lifetime prevalence rates for any kind of abuse/dependence were significantly higher in the group of schizophrenic patients examined at a mental state hospital than in the group of patients examined at an university hospital and significantly higher among male patients. Unlike some studies in the USA, alcohol was by far the most commonly abused drug. Cannabis, amphet- amines and cocaine were the most frequently abused il- legal drugs, followed by opioids and hallucinogens. Con- cerning legal drugs the number of patients abusing ben- zodiazepines was surprisingly high. Several methodological difficulties are encountered in assessing alcohol/drug abuse in schizophrenic pa- tients. Thus, a matter of concern has been psychiatric diagnosis which in the present study was made according to ICD-9 criteria. Since most patients had been treated in psychiatric hospitals a number of times, reliable infor- mations on the past psychiatric history was available. Psychopathology was documented by means of the AMDP system. An obvious problem was to find a con- sensus on the patients' alcohol and drug history. Apart from records of prior psychiatric admissions and the patients' self-reports a detailed semistructured interview was performed on each patient to collect any possible in- formation on the patient's alcohol/drug history. We feel that this was an adequate procedure to ensure psychiat- ric diagnosis. From a methodological point of view it seems noteworthy that the "MALT" correctly identified 81.2% of the 90.3% patients clinically diagnosed as alco- holics, with false positive results being very rare, and therefore proved to be a valuable diagnostic instrument for assessing alcohol abuse in schizophrenic patients. Only two pilot studies have addressed the problem of validity of alcoholism screening instruments in schizo- phrenics so far (Smith and Pistrach 1990; Toland and Moss 1989). In comparison with the results reported in these studies the "MALT" proved to be a valuable diag- nostic instrument of good sensitivity and remarkably high specifity in schizophrenic patients. Another important issue is the order of onset of psychosis and addiction, which has been a source of speculation. In our own sample the onset of psychosis more often preceeded than followed onset of an alcohol/ drug abuse which again was the case in about a third of the "dual diagnosis" patients. Since both alcoholism and drug abuse can result in chronic schizophrenia-like psychosis in some cases (Benedetti 1952; Glass 1989a, b; Bron 1982) it may be argued whether some of the patients Table 9. Suicide attempts in schizophrenic patients Schizophrenics Without addiction With addiction N (%) N (%) Group I Patients with suicide attempts Group II Patients with suicide attempts Number of suicide attempts/patients P 143 40 19 (13.2) 12 (30) <0.05 255 192 95 (37.3) 96 (50) <0.01 0.76 ((cid:127) 1.34 ((cid:127) <0.01
`
`CFAD VI 1023-0006
`
`

`

`368 diagnosed as being schizophrenic indeed suffer from a schizophrenia-like psychosis caused or induced by alco- hol/drug abuse. Since valid biological or genetic markers for the diagnosis of schizophrenia have not yet been es- tablished and the psychopathology in chronic alcohol hallucinosis closely resembles chronic schizophrenia (Benedetti 1952) this possibility cannot be ruled out completely. Yet clinical and psychopathological studies have shown that some psychopathological symptoms such as psychotic ego disturbances and schizophrenic thought disorder, both being rarely found in alcohol psychosis, as well as age at onset and different family history, among others, can facilitate differential diag- nosis between both conditions (Soyka 1989, 1990). Some other findings appear to count against this hypothesis, too. With respect to alcohol hallucinosis, catamnestic studies have shown that only in a minority of cases does a chronic schizophrenia-like psychosis finally emerge (review by Glass 1989b). Thus chronic alcohol hallu- cinosis which might be mistaken for schizophrenia can be considered a very rare phenomenon. Since alcohol hallucinosis resembles paranoid schizophrenia, most of all but patients with paranoid schizophrenia did not show higher prevalence rates for alcohol abuse/depen- dence than other schizophrenics in our study, there is little reason to assume that patients diagnosed as schizo- phrenics really had alcohol hallucinosis in disguise. Other findings further corroborate this view: alcoholics usually develop alcohol hallucinosis in their late thirties following a long-time alcoholism and therefore later in life than (paranoid) schizophrenia on average (Soyka and Zugs 1989) which was not the case in the schizo- phrenic patients with alcohol abuse in our sample. As for drug abuse the setting is equally complex. A major psychopathological feature, the paranoid-hallu- cinatory symptomatology in drug (such as cocaine-in- duced) psychosis differs from equivalents in schizophre- nia (Unnithan and Cutting 1992) but there is little doubt that some drugs, especially amphetamines and hallucino- genes, but also cannabis, can cause acute schizophrenia- like psychosis. Still, there is very limited evidence that drug abuse can result in a chronic psychosis. Drug abus- ers with schizophreniform psychosis were found to de- velop psychosis at an earlier age than schizophrenics in general (Alterman et al. 1984; Breakey et al. 1974; Erard et al. 1980; Richardson et al. 1985) which was also the case in our material. Tsuang et al. (1982) found a sig- nificantly higher rate of schizophrenia in the families of patients with prolonged drug-induced psychosis, i.e. with a minimum duration of 6 months. Bron (1982) and McGuire (1992) reported similar results. Other authors have reported a high rate of affective disorders in families of patients with drug-induced psychosis (Bowers 1977; Horowitz 1975; Pope 1979). A catamnestic study of the patients examined by Tsuang et al. (1982) showed that patients with an initially chronic course and severe psychotic symptoms had a poorer outcome than those with an initially faster remission (Perkins et al. 1986) and Vardy and Kay (1983) reported high rates for alcohol and especially drug abuse in families of dual diagnosis schizophrenics but the methodology of both studies has attracted considerable concern (Kane and Selzer 1991). Thus there is some evidence that chronic cannabis and drug abuse can result in a chronic schizophrenia-type psychosis but the question remains unanswered whether these patients might have developed schizophrenia in any possible case, regardless of drug abuse. Tsuang esti- mated that some 60000 drug users would have to be can- vassed in a follow-up study to answer that question. Up to date, few studies have focused on the order of onset of alcohol/drug abuse and schizophrenia (Pulver et al. 1989). Hays and Aidroos (1986) reported cases in which alcoholism clearly antedated the onset of schizo- phrenic psychosis. Werner et al. (1981) mentioned a group of 45 alcoholics with "schizophrenic syndromes". Of these, in 33 cases alcoholism preceeded the onset of psychosis. Bland et al. (1987) reported cases in which alcohol/drug abuse both preceeded and followed the onset of schizophrenia. The most important findings in our study have been the high prevalence rates for alcohol/drug abuse and de- pendence in both groups of schizophrenic patients. Most previous studies on this topic suffered from serious methodological problems concerning sample size, rep- resentivity of the sample and diagnostic instruments (for review see Mueser et al. 1990). The prevalence rates for alcohol/drug abuse found in our study were considerably higher than those reported for the general population. Epidemiological field studies in the Upper Bavarian re- gion suggested a prevalence rate for alcoholism of 7.9% (Fichter 1990), the US "Epidemiological catchment area" study (Regier et al. 1988) a 6-month prevalence rate for alcohol abuse/dependence of 4.7% in general and 9% for males. The "lifetime" prevalence was re- ported to be 13.3%. The latter study also showed a high co-morbidity of alcoholism and schizophrenia: among alcoholics, schizophrenia was four times more common than among non-alcoholics (Helzer and Pryzbeck 1988), the co-morbidity of alcoholism and affective disorder being comparatively rare. Prevalence estimates for sub- stance abuse among patients with schizophrenia were es- timated at 47.0% and 33.7% for alcohol abuse resp. alcohol-related disorders (Regier et al. 1990), thus show- ing similar results to our study. In another epidemiologi- cal study Bland et al. (1987) reported a "lifetime" preva- lence of alcohol abuse of 61.2% and for drug abuse of 50.8% in schizophrenics but the number of schizophren- ic patients was small. Likewise, our data on the preva- lence of alcohol and drug abuse among schizophrenics support the hypothesis that schizophrenia is associated with a higher rate of alcohol and drug abuse. Whether the increasing prevalence estimates for substance abuse in recent surveys indicate a real increase in alcohol/drug abuse in schizophrenics as suggested by Cuffel (1992) or should rather be attributed to improved methods of as- sessing or differences in defining substance abuse or instead merely reflect differences in study samples is de- bated (Alterman 1992, Siris 1992). As the results of other studies have already indicated, there seems to be a quite distinct abuse pattern in schizo- phrenia. Apart from alcohol, the most frequently abused drugs in the study of Mueser et al. (1990) were cannabis
`
`CFAD VI 1023-0007
`
`

`

`369 (42%), psychostimulants (25%) and hallucinogens (18%), while narcotics (4%) and sedatives (7%) were abused only infrequently. In other predominantly US studies the proportion of cannabis abusers among schizophrenic patients was estimated at 35% (Barbee et al. 1989), 45% (Breakey et al. 1974) up to 65.8% (Negrete et al. 1986). Prevalence rates for psychostimulants and cocaine abuse were reported to be 4.5% (Hansell and Willis 1977), 11.3% (Barbee et al. 1989), 13.0% (Siris et al. 1988), 15.0% (Breakey et al. 1974) up to 23.1% (Richard et al. 1985), while hallucinogens were reported to be consumed by 5.7% (Barbee et al. 1989), 9.9% (McLellan and Druly 1977), 11% (Siris et al. 1988) or up to 20% (Breakey et al. 1974). These data must be seen in comparison with drug exposure and abuse pattern in the general population. While alcohol abuse is very frequent in Germany with the per-capita consumption of alcohol being approximately 11L in 1989, most experts agree that drug abuse still is a more severe problem in the US. In Germany, until recently, the abuse of narcotics was much more prominent than the abuse of psychostimulants and cocaine (Leune 1991). With respect to schizophrenic subtypes few differ- ences could be demonstrated. Patients with paranoid schizophrenia had a higher 3-month (8.2% vs. 2.7%, P < 0.01) and lifetime (9.6% vs. 4.3%, P < 0.05) preva- lence for drug abuse. These results are well in line with other findings (Bernadt and Murray 1986; Mueser et al. 1990). The highest overall prevalence rates for any kind of abuse were found among patients with residual type. This could suggest that an additional alcohol/drug abuse might either indicate a more chronic course of illness or reflect less social support and an inadequate psychoso- cial integration, as indicated by a large number of un- married patients in our study. As mentioned before, sev- eral studies have shown a frequent association of both al- coholism and schizophrenia with homelessness (Priest et al. 1985; Susser et al. 1985, 1989). Social status, race and educational level have been considered to be of some importance for the development of alcohol/drug abuse in schizophrenics (Mueser et al. 1990) but apart from sex neither educational nor social background nor marital status were found to be predictive in any way (Pulver et al. 1989; Zeiler 1990, 1991). Probably alcohol/drug abuse significantly accentuates non-compliance in schizophren- ics (Alterman et al. 1980; Drake and Wallach 1988; Drake et al. 1989; Miller and Tanenbaum 1989; Wallen and Weiner 1989) and contributes to a disappointing out- come and poor social integration. A third possible explanation for the high prevalence rates for alcohol/drug abuse in schizophrenia and the dif- ferent abuse patterns related to various schizophrenic subtypes is offered by the "self-medication"-hypothesis. It has been suggested that certain psychopathological symptoms such as post-psychotic depression, hallucina- tions or "negative" symptoms such as anhedonia or aner- gia might predispose to a particular abuse pattern (Pope 1979; Schneier and Siris 1987; Zeiler 1990, 1991). For exampl