Does Religiosity Reduce Mortality Risk Over the Second Half of Life Among Southern Californians? A Multidimensional Model Within a Hazard Modeling Framework

This study explored the relationship between a multidimensional typology of religiosity and risk of mortality. Using data from middle-aged Southern California respondents in the Longitudinal Study of Generations, we performed latent class analysis to construct a typology of religiosity, which was then used to predict mortality from 1971 to 2020. We identified four religiosity classes: strongly, weakly, privately, and liberally religious. Cox proportional hazard models revealed that privately religious males had greater mortality risk than strongly religious males, even after controlling for the mediators self-rated health and psychological wellbeing. In women, risk of mortality was explained by these mediators rather than by religiosity.