Is religious service attendance beneficial for mental health? Causal inference using structural nested mean models

OBJECTIVES: Previous studies on the longitudinal association between religious service attendance and mental health have not adequately accounted for time-varying confounding. The current study aimed to address this limitation. STUDY DESIGN: Longitudinal analysis using nationally representative data. METHODS: Data from Waves 3, 4, 6, 8, and 9 of Understanding Society: The UK Household Longitudinal Study were used (n = 86,093). Mental health was assessed using the General Health Questionnaire-12 (GHQ-12) and the Mental Component Summary (MCS) of the 12-item Short-Form Health Survey. RESULTS: After adjustment for gender, age, and ethnicity as baseline covariates, as well as self-rated health and life satisfaction as time-varying covariates, frequent religious service attendance did not predict subsequent GHQ-12 or MCS scores. CONCLUSIONS: After accounting for time-varying confounding, frequent religious service attendance was not a statistically significant predictor of subsequent mental health among UK adults. These findings do not support the assumption that religious service attendance is universally beneficial for mental health, although the confidence intervals were wide enough to be consistent with small effects in either direction.