Religious service attendance and mortality: A population-based prospective cohort study in southern Sweden.

Aims: The aim is to investigate associations between attendance in religious service during the past year and all-cause, cardiovascular (CVD), cancer and other cause mortality. Study design: Prospective cohort study. Methods: A public health survey with three reminders was sent to a stratified random sample of the adult 18-80 population in southernmost Sweden in 2008. The response rate was 54.1%, and 24,855 participants were included in this study. The cross-sectional baseline survey was connected to mortality data with 8.3-year follow-up. Analyses were conducted in Cox regression models. Results: 13.9% had attended religious service at least once during the past year, and 86.1% had not attended. The group with religious attendance contained significantly higher proportions of women, high and medium position non-manual employees, participants born abroad, never alcohol consumers, respondents with high trust in others and respondents with high social participation. It also contained significantly lower proportions with low leisure-time physical activity (LTPA) and daily smokers. Religious service attendance during the past year was significantly associated with lower hazard rate ratios (HRRs) of all-cause mortality compared to non-attendance until social participation items were introduced in the final model. HRRs of CVD mortality were significantly lower for religious attendance in the multiple models until BMI and health-related behaviors were introduced. No significant results were observed for cancer and other cause mortality. Conclusions: The results suggest that religious service attendance in a highly secularized country such as Sweden is significantly associated with lower all-cause mortality, which may be explained by a social network pathway in this highly secularized population.