Abstract 3991: Cardiovascular Mortality and Frequency of Attendance at Religious Services in a Large National Cohort, 1988–2000
Introduction. Social support and successful coping with stressors is thought to play a role in preventing mortality from cardiovascular disease (CVD). Organized religion may provide both. We reported elsewhere a strong, inverse association of frequency of attendance at religious services (FARS) and mortality from all causes combined in a national cohort.
Hypothesis and Methods. We assessed the hypothesis that the association of FARS with total mortality is due to an association with CVD death in the Third National Health and Nutrition Examination Survey (NHANES III). We examined data from 20,024 American men and women 17 years-of-age and older at baseline (1988–1994) and followed for mortality through 2000. After all exclusions, the analysis included 14,223 persons. There were 2,033 deaths, including 907 with CVD (ICD-10 codes I00-I99) as the underlying cause category.
Results. At baseline, 26.1% reported attending religious services weekly and 8.6% more than weekly. Compared to never attenders, the socio-demographic adjusted hazard ratio (HR) for CVD deaths (95% confidence limits) for weekly attenders was 0.77 (0.63– 0.94, p<0.01), and for more than weekly attenders was 0.76 (0.52–1.02, p<0.10). The HRs for ischemic heart disease death were 0.71 (0.53– 0.95, p<0.05) and 0.76 (0.52–1.12, NS); and for stroke 0.82 (0.49–1.35, NS) and 0.62 (0.27–1.41, NS), respectively. After also controlling for baseline health status variables, the HRs for weekly and more-than-weekly attenders were increased slightly and no longer significant. In contrast, we saw a strong, independent inverse association of FARS with cancer death.
Conclusions. Persons attending religious services weekly or more than weekly did not have significantly lower adjusted risk of mortality from CVD causes of death compared to never attenders. Therefore the association of FARS with all-cause mortality is mostly due to an inverse association with non-CVD causes in this cohort.