Mental Disorders and Violence in a Total,::Birth Cohorts Results from the Dunedin Study
Background: Because most individuals with mental illness are not hospitalized and most violent individuals are not convicted of crimes, hospital-and prison-based research underestimates the rates of mental illness and violence found in the general population. This study examined the overlap of mental disorders and violence in a birth cohort. Method: A total of 961 individuals born in Dunedin, New Zealand, from April 1, 1972, through March 31, 1973, (i.e., 94% of the total city birth cohort) were studied. DSM-III-R interviews were used to identify pastyear prevalence of mental disorders, and self-report of criminal offense and search of official conviction records were employed to measure past-year violence. The variables of substance use before the violent offense, excessive threat perception, and adolescent conduct disorder were studied as possible explanations for the link between mental disorders and violence. Results: Individuals with DSM-III-R alcohol dependence were 1.9 times (95% confidence interval [CI] = 1.0 to 3.5), those with marijuana dependence were 3.8 times (95% CI = 2.2 to 6.8), and those with schizophrenia-spectrum disorder were 2.5 times (95% CI = 1.1 to 5.7) more likely to be violent than individuals without a psychiatric disorder. Although individuals with at least 1 of these disorders committed half of the violent crimes reported in this study (one tenth of the violence accounted uniquely for by patients with schizophrenia-spectrum disorder), they constituted only one fifth of the study cohort. Substance use before the violent event accounted for the violence in individuals with alcohol dependence. Adolescent history of conduct disorder best explained violence in individuals with marijuana dependence. Both excessive threat perception and adolescent history of conduct disorder accounted for violence in individuals with schizophrenia-spectrum disorder. Conclusions: Individuals with mental illness were responsible for a substantial percentage of the violent acts committed by persons within their age group. Because the explanations for violence varied between groups of individuals with different mental disorders, multiple treatment and intervention strategies may be necessary to prevent the occurrence of violent acts.
Arseneault, L; Moffitt, TE; Caspi, A
Volume / Issue
Start / End Page
International Standard Serial Number (ISSN)