Date of Conferral

4-21-2026

Degree

Ph.D.

School

Health Sciences

Advisor

Srikanta Banerjee

Abstract

Diabetes remains disproportionately prevalent in Appalachia, and macroenvironmental contributors remain poorly understood. Guided by spatial epidemiology, this quantitative, cross-sectional ecological study examined whether county-level environmental factors including distance to mine, distance to landfill, distance to water, mean annual precipitation, mean annual temperature, forest classification, distance to impervious surface, and median income were independently and collectively associated with diabetes cases per capita in West Virginia, used as a surrogate for Appalachia. Research questions addressed environmental predictors of diabetes cases per capita with and without controlling for median income. Publicly available county-level data from 53 West Virginia counties representing adults aged 20 years and older with diagnosed diabetes were analyzed after ArcGIS-based variable preparation. Multiple linear regression was conducted in SPSS. The unadjusted model was not statistically significant, F(8, 44) = 0.90, p = .528, R² = .14, and the adjusted model controlling for median income was also not statistically significant, F(9, 43) = 1.05, p = .415, R² = .18. No individual predictor reached statistical significance. These findings suggest that county-level centroid analyses may underestimate localized environmental influences on diabetes prevalence. Implications for positive social change include informing geographically targeted surveillance, guiding future higher-resolution spatial research, and supporting more precise resource allocation in Appalachian communities.

Download

Included in

Epidemiology Commons

Share

 
COinS