Psychopathology and attrition in the epidemiologic catchment area surveys

This analysis shows the effect of sociodemographta variables and psychopathology on attrition in a 1-year follow-up of 10,167 household-residing respondents, aged 18-64 years, from four sites of the National Institute of Mental Health Epidemiologic Catchment Area Program. Attrition is defined in two ways: failure to locate (n = 1,035) and refusal to participate (n = 1,010). As compared with 8,122 respondents successfully interviewed after 1 year, those who are not located are more likely to be male, Hispanic, young, and unmarried. Refusers are older and more likely to be married and to have lower educational attainment than those successfully reinterviewed. After adjusting for sociodemographic variables, psychopathology had a small to moderate effect on attrition. Panic and depression were associated with failure to locate, but no diagnoses, or selected symptoms, were strongly or significantly associated with refusal to partici-pate in the second wave. Aw J Epidemiol 1992;135:1051-9. follow-up studies; psychopathology The problem of attrition in field surveys will attract increasing attention as well