Correlations among study variables.

Objective
Religious and spiritual (R/S) engagement is generally associated with better health. However, it is not known which comes first between R/S engagement and health due to a lack of longitudinal studies. We examined this issue in a sample assessed six years apart.
Methods
Using a large nationwide sample (N = 3,010 at Wave 1; n = 607 at Wave 2) and structural equation modeling (SEM), we developed composite latent measures of R/S engagement and self-rated physical health (SRH). R/S engagement included identity, prayer, commitment, attendance, and coping. SRH included two subjective ratings of health alongside number of chronic illnesses. We examined the measurement invariance and reciprocal relationships of R/S engagement and SRH at two timepoints (six years apart), and controlled for multiple demographic variables (age, sex, education, income, race/ethnicity) as well as R/S engagement or SRH at Wave 1. We then assessed whether the strength of their relationships with each other differed.
Results
Higher R/S engagement at Wave 1 significantly predicted better SRH at Wave 2 (β =.07, b = 0.09, SE = 0.04, p =.026), whereas higher SRH at Wave 1 did not significantly predict higher R/S engagement at Wave 2 (β =.02, b = 0.03, SE = 0.03, p =.224); however, such associations were not significantly different. Post-hoc weighted analyses indicated the findings were driven by older participants.
Conclusion
Our findings demonstrate that R/S engagement predicts better SRH six years later, whereas better SRH does not significantly predict future R/S engagement. Future areas of growth in the R/S and health research field include addressing heterogeneity in the measurement of both constructs, increasing sample diversity/representation, and considering contextual nuances. Experimental methods or multiple-timepoint data, with a focus on mediators (e.g., inflammation), may help further disentangle the longitudinal relationships between R/S engagement and self-rated health.