BackgroundThere is evidence that religious attendance is associated with positive outcomes for mental health; however, there are few longitudinal studies, and even fewer, which take into account the possibility of bi-directional associations. This study aimed to investigate bi-directional associations between religious attendance and mental health.MethodsParticipants were 2125 study members who provided data at age 68–69 from the Medical Research Council National Survey of Health and Development (1946 British birth cohort study). Mental health was assessed using the 28-item General Health Questionnaire at ages 53, 60–64 and 68–69. Religious attendance was measured using a 4-point scale (weekly=3, monthly=2, less than monthly=1 or never=0) at ages 43, 60–64 and 68–69. Cross-lagged path analysis was used to assess reciprocal associations between mental health and religious attendance, adjusting for gender and education.ResultsPrevious religious attendance was strongly related to later attendance (r=0.62–0.74). Similarly, mental health at baseline was strongly associated with subsequent mental health scores (r=0.46–0.54). Poor mental health at age 53 and 60–64 was associated with more frequent religious attendance at age 60–64 (b=0.04; 95% CI: 0.02 to 0.06; p<0.05), and 68–69 (b=0.03; 95% CI: 0.02 to 0.06; p<0.05), respectively. There was no evidence that religious attendance at age 43, 60–64 or 68–69 was associated with later or concurrent mental health.ConclusionUsing birth cohort data from the UK, it was found that poor mental health was associated with later religious attendance but not vice versa. Future research should confirm these novel findings and explore the underlying mechanisms between religious attendance and mental health.