Researchers from the Harvard School of Public Health analyzed data from a prospective study of 89,708 women aged 30-55 years (98% white) participating in the Nurses’ Health Study, examining predictors of completed suicide during the 14-year follow-up.
Religious service attendance was assessed in 1992 and 1996 with the question “How often do you go to religious meetings or services?” (never, almost never, less than once per month, 1-3 times per month, once a week, or more than once a week).
Death by suicide were documented using state mortality files, the US National Death Index, and reports from next of kin, using the standard definition contained in the 8th version of the International Classifiaction of Disease (ICD-8).
Multiple covariates possibly related to suicide were assessed and controlled for including age, employment status, family history of alcoholism, BMI, physical activity level, caffeine intake, alcohol intake, smoking status, depressive symptoms, history of chronic medical illness, living situation, social integration (including marital status, other group participation, number of close friends, number of close relatives, number of close friends seen at least once per month, number of close relatives seen at least once per month), and religious service attendance in 1992.
The primary predictor was religious service attendance in 1996. To ensure that depressive symptoms or chronic health problems at baseline were not responsible for low religious attendance (reverse causation), women who were depressed or used antidepressant medications or had a history of cancer or other cardiovascular condition in 1996 were excluded from the analysis. Multivariable Cox proportional hazards regression models were used to analyze the data, and mediation analysis was used to examine factors that might help to explain the relationship (depressive symptoms, alcohol intake, social integration). In addition sensitivity analysis was used to assess the robustness of the results.
Results: Women who attended religious services at least once per week were 84% less likely to commit suicide than women who never attended services (hazard ratio=0.16, 95% CI=0.06-0.46), with more than a five-fold reduction in incidence rate from 7 per 100,000 person-years to only 1 per 100,000 person-years. Results were similar when excluding women who were depressed or had chronic illness at baseline in 1996. Effects were particularly strong among Catholic women (HR = 0.05, 95% CI = 0.006-0.48).
Adjustments for age, lifestyle, demographic, and religious service attendance in 1992, as well as further adjustment for social integration in 1992, could not explain the effect. Sensitivity analysis indicated that unmeasured confounders would have to both increase the likelihood of religious service attendance and decreased the likelihood of suicide by 12- fold above and beyond the measured confounders to explain the effect (highly unlikely).
VanderWeele TJ, Li S, Tsai AC, Kawachi I (2016). Association between religious service attendance and lower suicide rates among US women. JAMA psychiatry (formerly Archives of General Psychiatry), June 29.
Comment: The quality of this study (large sample size and long follow-up) and of the statistical methods used to analyze the data (careful control for confounding, elimination of possible reverse causation, sensitivity analysis) make it a seminal one in the study of religion and suicide. Given the CDC’s recent announcement of an alarming 60% increase in suicide rate between 1999 and 2014 among white women in the US, these results are relevant to public health.
Koenig HG. Association of religious involvement and suicide. JAMA Psychiatry, June 29, 2016