Abstract
BACKGROUND: Against a background of interest in rates of diabetes in schizophrenia and related psychoses and claims that data from historical periods demonstrate a link that antedates modern antipsychotics, we sought to establish the rate of diabetes in first onset psychosis and subsequent prevalence in historical and contemporary cohorts. METHODS: Analysis of two epidemiologically complete databases of individuals admitted for mental illness. 3170 individuals admitted to the North Wales Asylum between 1875-1924 and tracked over 18,486 patient years and 394 North West Wales first admissions for schizophrenia and related psychoses between 1994 and 2006 and tracked after treatment. RESULTS: The prevalence of Type 2 diabetes among patients with psychoses at time of first admission in both historical and contemporary samples was 0%. The incidence of diabetes remained 0% in the historical sample throughout 15 years of follow-up but rose in the contemporary sample after 3, 5 and 6 years of treatment with an incidence rate double the expected population rate so that the 15 year prevalence is likely to be over 8%. CONCLUSION: No association was found between diabetes and serious mental illness, but there may be an association between diabetes and treatment.