Global suicide trends across major religions from 2000 to 2021

BMC Public Health. 2026 Jul 1. doi: 10.1186/s12889-026-28354-5. Online ahead of print.

ABSTRACT

BACKGROUND: Suicide claims over 700,000 lives annually. Religious beliefs may influence suicide risk, yet large-scale, cross-national comparisons remain limited.

METHODS: This retrospective, ecological study analyzed age-standardized suicide rates of 185 countries over a 22-year period (2000-2021) using World Health Organization (WHO) mortality data. Countries were classified by dominant religion, including Buddhism, Christianity, Hinduism, Islam, and Others, per Pew Research Center data, and by income level (World Bank) and region (WHO). Suicide numbers were estimated by multiplying suicide rates and UN population data. Poisson regression models, controlling for income, region, and year, examined the association between religion and suicide rates.

RESULTS: Islam showed the lowest suicide rates across total, male, and female populations. Poisson analysis confirmed significantly lower risk for Muslims (total population RR = 0·37, 95% CI: 0·33-0·42) than Buddhists as a base. Hindu populations, particularly females, exhibited the highest suicide rates (total population RR = 2·02, 95%CI:1·82-2·24), (female population RR = 4·74, 95%CI: 4·26-5·28). Christianity showed high male suicide rates descriptively, but adjusted risk decreased after accounting for region. (total population RR = 0·65, 95%CI: 0·57-0·75).

CONCLUSIONS: Our findings indicate a statistical association between a country’s dominant religion and its suicide rates. The lower rates observed in Islamic-majority countries might be related to factors such as religious prohibitions and community cohesion, while the high rates among females in Hindu-majority nations could be linked to specific socio-cultural pressures. Given the study’s cross-sectional ecological design, these results merely highlight associations and cannot establish causality. Further research is necessary to elucidate the complex mechanisms underlying these observations.

PMID:42380917 | DOI:10.1186/s12889-026-28354-5