Exploring geographic variation in US mortality rates using a spatial Durbin approach.

Area of Science:

• Public Health
• Spatial Epidemiology
• Sociology

Background:

• Traditional studies on US county mortality determinants overlook spatial dependencies.
• Ignoring neighboring county characteristics provides an incomplete understanding of mortality variations.

Purpose of the Study:

• To investigate the spatial associations between county-level features and mortality rates.
• To apply spatial Durbin modeling to account for spatial dependence in mortality research.
• To test the spillover and social relativity theories in explaining mortality variations.

Main Methods:

• Utilized spatial Durbin modeling to analyze mortality determinants in US counties.
• Incorporated spatial dependence, considering both own-county and neighboring county features.
• Compared findings against traditional methods like ordinary least squares, spatial error, and spatial lag regression.

Main Results:

• Mortality rates are significantly associated with neighboring county characteristics, supporting spatial dependence.
• Spillover effects observed: Hispanic population percentage, concentrated disadvantage, and social capital negatively associated with mortality in own and neighboring counties.
• Social relativity effects observed: Health insurance coverage, non-Hispanic other races percentage, and income inequality showed opposite associations with mortality in own versus neighboring counties.

Conclusions:

• Spatial Durbin modeling offers a more accurate approach to ecological mortality research by addressing spatial autocorrelation.
• Both spillover and social relativity perspectives are valuable for understanding complex mortality patterns in US counties.
• Findings provide unbiased estimates, offering new insights into the spatial determinants of county mortality.