Abstract
The influence of level of activity examined with data from an eight-year (1976 to 1984) longitudinal study of 508 older Mexican Americans and Anglos. Over the study interval, 119 subjects were confirmed to have died. Activity was a significant predictor of mortality at the univariate level. However, when age, gender, education, marital status, ethnicity, and self-rated health were controlled for in the analysis, activity was not a significant predictor of mortality. The popular notion that an active life among elderly persons might lead to extended longevity was not supported by these data.
