Abstract

This article presents a prospective analysis of an antecedent psychopathologic features and sociodemographic risk factors in schizophrenia with data from five community sites in the National Institute of Mental Health Epidemiologic Catchment Area Program. Three nonoverlapping psychotic cases were defined using DSM-III definitions as implemented by the Diagnostic Interview Schedule (DIS): (1) DSM-III Schizophrenia Criterion A; (2) Criterion A and Affective Episode; and (3) full Schizophrenia. In a 1-year follow-up period, the cumulative incidence rate of Criterion A was 0.79 per 100, for Criterion A with Affective Episode it was 0.17 per 100, and for Schizophrenia the rate was 0.20 per 100. In multivariable logistic regression modeling, the patterns of associations between sociodemographic factors and DIS/DSM-III Schizophrenia resembled patterns in clinically based registry data. Male subjects had an earlier peak onset than female subjects, and marital status and employment were strongly related to odds of developing DIS/DSM-III Schizophrenia. An interaction between gender and never marrying showed never-married men at 50 times higher odds of developing DIS/DSM-III Schizophrenia, never-married women at 14 times higher odds, and married women at 2.5 times higher odds, relative to married men. Adjusting for sociodemographic factors, DIS/DSM-III Obsessive Compulsive Disorder and Social Phobia were both associated with more than 3.5 times increased odds of developing DIS/DSM-III Schizophrenia. Several other psychopathology items, including panic attacks, were associated with increased odds of developing DIS/DSM-III Schizophrenia. There were both similarities and differences in risk factor structure between DIS/DSM-III Schizophrenia and the other two defined categories of case.