Objective: Frequent attendance at religious services has been reported by several studies to be independently associated with lower all-cause mortality. The present study aimed to clarify relationships between religious attendance and mortality by examining how associations of religious attendance with several specific causes of death may be explained by demographics, socioeconomic status, health status, health behaviors, and social connections.
Method: Associations between frequent religious attendance and major types of cause-specific mortality between 1965 and 1996 were examined for 6545 residents of Alameda County, California. Sequential proportional hazards regressions were used to study survival time until mortality from circulatory, cancer, digestive, respiratory, or external causes.
Results: After adjusting for age and sex, infrequent (never or less than weekly) attenders had significantly higher rates of circulatory, cancer, digestive, and respiratory mortality (p < 0.05), but not mortality due to external causes. Differences in cancer mortality were explained by prior health status. Associations with other outcomes were weakened but not eliminated by including health behaviors and prior health status. In fully adjusted models, infrequent attenders had significantly or marginally significantly higher rates of death from circulatory (relative hazard [RH] = 1.21, 95 percent confidence interval [CI] = 1.02 to 1.45), digestive (RH = 1.99, p < 0.10, 95 percent CI = 0.98 to 4.03), and respiratory (RH = 1.66, p < 0.10, 95 percent CI = 0.92 to 3.02) mortality.
Conclusions: These results are consistent with the view that religious involvement, like high socioeconomic status, is a general protective factor that promotes health through a variety of causal pathways. Further study is needed to determine whether the independent effects of religion are mediated by psychological states or other unknown factors.