VOL 19, NO. 4, 1993 The Epidemiology of Late-Onset Schizophrenia

We report an analysis of a large catchment area sample of pa-tients with nonaffective func-tional psychoses presenting across all ages at onset. The male:female ratio was 1.56:1 in the 16-25-year age group; it reached unity around 30 years of age and declined to 0.38:1 in the 66-75-year group. Contrary to ex-pectation, a higher proportion of patients with onset of illness af-ter 45 years than of younger onset patients fulfilled DSM-III-R criteria for schizophrenia (52% vs. 38%). The distribution by age at onset was much the same irre-spective of stringency of diag-nosis. The highest rates were in the 16-25-year age group, with a slight second peak in the 46-55-year group, and a third (more emphatic) peak in the over-65 group. A closer analysis of dem-ographic and phenomenologic variables revealed distinct dif-ferences between patients with early and late (after 44 years) onset of illness. Despite the plethora of epidemio-logic investigations of schizophre-nia, late-onset illness has all too often been ignored. This lack of interest (until recently) in late-onset schizophrenia is due, at least in part, to the tacit acceptance (particularly in the United States) that schizophrenia cannot manifest for the first time after 44 years of age. This view is embodied in the DSM-III criteria for schizophreni