Increased risk of sexual offending in men with psychotic disorders.

ED FROM Alden A, Brennan P, Hodgins S, et al. Psychotic disorders and sex offending in a Danish birth cohort. Arch Gen Psychiatry 2007;64:1251–8. Correspondence to: Amanda Alden, MA, Department of Psychology, Emory University, 532 N Kilgo Cir, Atlanta, GA 30322, USA; [email protected] Source of funding: Research supported by grants from the National Institute of Mental Health, the National Institute of Justice, Fonds Concerte d’Aide a la Recherché (FCAR), Social Science and Humanities Research Council. c Additional notes, a table and references are published online only at http:// ebmh.bmj.com/content/vol11/issue3 C O M M EN TA R Y T he association of sexual offending and severe mental illness has not been widely studied, and we welcome the study by Alden and colleagues. In a nested case-control study of people born in Denmark 1944–7 and followed until 1991, they found a fourfold increased risk of hospitalisation for schizophrenia in sex offenders compared with the general population (adjusted odds ratio (OR) 4.2, 95% CI 2.9 to 6.0). This replicated findings of previous population-based case-control studies in Australia (crude OR 2.7), and in Sweden with follow-up until 2000 (adjusted OR 4.8, 3.4 to 6.7). Although a population study might have design advantages, low base rates for sexual offending and psychosis yielded only 152 sex offenders with psychosis (the Swedish report had 413). The authors found an increased violence risk for psychosis with versus without comorbid personality disorder and a non-significant trend for comorbid substance abuse increasing risk. They acknowledged the difficulties with secondary diagnoses in psychosis, and robust validation studies of comorbidity are lacking. The reference cited for diagnostic accuracy is a non-blinded study of mothers with schizophrenia, finding concordance rates for schizophrenia between ICD-8 and DSM-III diagnoses of 84–86% (not over 90% as stated). Further, caution is warranted in interpreting comorbidity findings as individuals are more likely to be diagnosed with personality disorder if arrested previously for a sex offence. Alden and colleagues were not precise with the terminology of offending. Specifically, they conflated paraphilias with offences. In law, a sexual offence where the victim is a child is termed ‘‘indecent assault’’, ‘‘rape’’ etc, but not ‘‘paedophilia’’. Therefore, it is incorrect to write ‘‘being convicted of paedophilia’’. Importantly, paedophilia implies a deviant sexual interest, which may or may not be associated with a child sexual offence. The same holds for ‘‘exhibitionism’’ and ‘‘voyeurism’’. This and related studies 2 suggest that psychoses are associated with sexual offending, and that 4–5% of sex offenders will benefit from antipsychotic treatment. Generalisability of the data is unclear. A 1944–7 birth cohort cannot elucidate the impact of potentially important secular trends. For example, reporting of sexual offences in Sweden sank from 49 per 100 000 inhabitants during 1960–9 to 35 in 1970–9 and more than doubled again to 84 during 1990–9. As the Alden study ended in 1991, more recent work will establish if the association has changed in the intervening 17 years. Dr Seena Fazel, MRPsych, MD Clinical Senior Lecturer in Forensic Psychiatry, Department of Psychiatry, University of Oxford, UK Dr Niklas Långström, MD, PhD Associate Professor, Centre for Violence Prevention, Karolinska Institutet, Stockholm, Sweden Competing interests: None. Prevalence EBMH August 2008 Vol 11 No 3 95 group.bmj.com on August 6, 2017 Published by http://ebmh.bmj.com/ Downloaded from