Psychosocial predictors of mortality among the elderly poor. The role of religion, well-being, and social contacts.

Mortality data during a two-year follow-up were obtained on some 400 elderly poor residents of New Haven, Hartford, and West Haven, Connecticut, in 1972-1974. These subjects were cases and controls in a study of the health effects of residential relocation. Initial data collection included a detailed health history, sociodemographic and background variables, and a variety of behavioral and psychological data. The variables selected for analysis in this report were: religious beliefs, social contacts, feelings of well-being, and affective states. Stepwise logistic regressions were used to determine the role of these psychosocial variables in predicting mortality, while controlling for case/control status, demographic variables, and health status (measured by an index maximally predictive of mortality in this sample). Three psychosocial variables were significant predictors: religiousness, happiness (as rated by the interviewers), and presence of living offspring. The first two reduced the risk of mortality primarily among the elderly who were in poor health, while the third one did not interact with health status.