Quantifying the Effects of Mental Health on U.S. Suicide and Mortality Rates

December 2025

Authors

Brigg Trendler
MS Student
Brigham Young University

Brianne Weaver
MS Student
Brigham Young University

Chris Groendyke, FSA, PhD
Professor and Actuarial Science Director
Robert Morris University

Brian Hartman, ASA, PhD
Professor and Actuarial Program Coordinator
Brigham Young University

Robert Richardson, ASA, PhD
Associate Professor
Brigham Young University

Davey Erekson, PhD
Assistant Director – Research and Technology, Clinical Professor, Licensed Psychologist
Brigham Young University

Executive Summary

Understanding how mental health influences mortality is increasingly important for actuaries involved in pricing, reserving, and population risk assessment. This study quantifies the relationship between mental health indicators, socio-economic conditions, and both suicide and all-cause mortality across U.S. counties from 2010 through 2023.

Using a Bayesian hierarchical spatiotemporal model with a conditional autoregressive (CAR) structure, mortality and suicide patterns are estimated by age, sex, and geography. The framework incorporates county-level data on education, income, housing, marriage, race, household size, unemployment, and multiple measures of mental health, including rates of severe depression, suicidal ideation, and self-reported poor mental health days.

Key findings include:

Strong geographic clustering: Neighboring counties show highly correlated mortality and suicide outcomes, confirming that regional social and economic context meaningfully influences risk.
Socio-economic disparities: County-level education, housing prices, and marriage rates are among the strongest predictors of suicide risk, though effects differ by age and sex. Higher education and home values are generally associated with reduced suicide risk for men but have mixed or opposite effects for women in later life.
Mental health as a leading indicator: County-level mental health distress is consistently associated with higher mortality and suicide rates. The relationship is most pronounced among youth and young adults.
Temporal persistence: Spatial and temporal correlations suggest stable, long-term regional patterns in both overall mortality and suicide.

These results demonstrate the value of integrating mental health surveillance data into traditional mortality modeling. Doing so can improve experience studies, risk segmentation, and forecasting for life, disability, and group health portfolios. The findings also highlight potential areas for designing interventions that address both mental health and underlying socio-economic factors.

Material

Quantifying the Effects of Mental Health on U.S. Suicide and Mortality Rates

Acknowledgements

The researchers thank the members of Project Oversight Group and staff from the Society of Actuaries Research Institute whose helpful suggestions, guidance, and comments have served to improve the quality of this project.

Project Oversight Group:
Tamara Bogojevic-Catanzano, FSA, MAAA
Bryan Burningham, ASA, MAAA
Jean-Marc Fix, FSA, MAAA
Rachelle Jacobs
Maggie Ma, FSA, FCIA
Murali Niverthi, FSA, MAAA
Marianne Purushotham, FSA, MAAA
Becca Reppert, FSA, CERA, MAAA
Zhanxiong Song, ASA

At the Society of Actuaries Research Institute:
Kara Clark, FSA, MAAA, Senior Practice Research Actuary
Barbara Scott, Senior Research Administrator

The researchers also thank Ben Lander, who added valuable insights into the data.

Finally, the researchers are grateful for the support of the Society of Actuaries Research Institute, Munich American Reassurance Company, the BYU Department of Statistics, and Robert Morris University.

Questions or Comments?

Give us your feedback! Take a short survey on this report. Take Survey

If you have comments or questions, please send an email to Research@soa.org