Frequency of attendance at religious services and mortality in a US national cohort

OBJECTIVE: Few nationally representative cohort studies have appeared on frequency of attendance at religious services and mortality. We test the hypothesis that > weekly attendance compared with nonattendance at religious services is associated with lower probability of future mortality in such a study. METHODS: Data were analyzed from a longitudinal follow-up study of 8450 American men and women age 40 years and older who were examined from 1988 to 1994 and followed an average of 8.5 years. Measurements at baseline included self-reported frequency of attendance at religious services, sociodemographics, and health, physical and biochemical measurements. RESULTS: Death during follow-up occurred in 2058. After adjusting for confounding by baseline sociodemographics and health status, the hazards ratios (95% confidence limits) were never 1.00 (reference); < weekly 0.89 (0.75-1-04), P = 0.15; weekly 0.82 (0-71-0.94) p = 0.005; and > weekly attenders 0.70 (0.59-0.83), p < 0.001. Mediators, including health behaviors and inflammation, explained part of the association. CONCLUSIONS: In a nationwide cohort of Americans, predominantly Christians, analyses demonstrated a lower risk of death independent of confounders among those reporting religious attendance at least weekly compared to never. The association was substantially mediated by health behaviors and other risk factors.