Income inequality and suicide in the United States: A spatial analysis of 1684 U.S. counties using geographically weighted regression

Background: Suicide rates vary considerably across U.S. counties. Spatial non-stationarity may explain mixed findings on the relationship between suicide and income inequality. Methods: This ecological study analyzed county-level income inequality and suicide rates for the timespan 2012-2016. Ordinary least squares regression, multilevel regression, and geographically weighted regression models were constructed while adjusting for age, race/ethnicity, gender, education, median income, unemployment, and urbanicity. Results: Ordinary least squares regression and multilevel models found no significant association between income inequality and county suicide rates after adjusting for confounding variables. However, the geographically weighted regression model identified two main areas in which income inequality was negatively associated with suicide rates, as well as several counties across central U.S. in which income inequality was positively associated with suicide rates. Conclusion: Income inequality's effect on county suicide rates may vary across space. Future research should consider spatial non-stationarity when studying suicide and macro-level socioeconomic conditions.