Geographically Weighted Modeling to Explore Social and Environmental Factors Affecting County-Level Cardiovascular Mortality in People With Diabetes in the United States: A Cross-Sectional Analysis

Abstract

Disparities exist in the cardiovascular mortality rates among people with type 2 diabetes (T2D). Research has established that these disparities are often related to the environmental and social determinations of health. This study explores the spatial variation between air pollution, social determinants of health and T2D related age-adjusted cardiovascular mortality (aa-CVM) in the United States. We obtained county-level T2D related to aa-CVM (per 100,000 residents) from Centers for Disease Control and Prevention WONDER (Wide-ranging Online Data for Epidemiologic Research) (2010 to 2019). We fit a geographically weighted linear regression with aa-CVM as the outcome and the following covariates (ambient air pollution [particulate matter of 2.5 µm size], median annual household income, racial/ethnic minorities, higher education, rurality, food insecurity, and primary health care access) were included. Overall, the median aa-CVM rate was 92.9 and highest in the South (102.2). In the West, aa-CVM was significantly associated with particulate matter of 2.5 µm size, annual median household income, racial minority status and primary health care access. Food insecurity was the most significant exposure in the Midwest and Northeast, while in the South, annual median household income and food insecurity were significant. In conclusion, this study demonstrated a substantial regional variation of exposure to determinants of T2D related aa-CVM in the United States. These findings should be considered in policy frameworks and interventions as part of community-level approaches to addressing T2D related aa-CVM, and within broader state and national discussions of the importance of population health.

Keywords: air pollution; cardiovascular mortality; diabetes; food insecurity; rural health; social determinants of health.

Figure 1.

County level rates (per 100,000 residents) of diabetes related cardiovascular mortality in the US (2010 –2019).

Figure 2.

County-level map of p-values from geographically weighted linear regression model reflecting association between age-adjusted diabetes related CVM in the US (2010–2019) for racial/ethnic minorities, median household income, PM2.5, and primary healthcare access Footnote: We fit a multivariable geographically weighted linear regression model to evaluate the association between chosen exposures and age-adjusted diabetes related cardiovascular mortality at the county-level in the US (2010–2019). We obtained and mapped the adjusted p-values from this model grouping them as p ≥ 0.05, p < 0.05 and p < 0.01. This map presents results for annual average PM2.5, racial/ethnic minorities, median annual household income, and primary health-care access.

Figure 3.

County-level map of p-values from geographically weighted linear regression model reflecting association between age-adjusted diabetes related CVM in the US (2010–2019) for food insecurity, rurality, and higher education Footnote: We fit a multivariable geographically weighted linear regression model to evaluate the association between chosen exposures and age-adjusted diabetes related cardiovascular mortality at the county-level in the US (2010–2019). We obtained and mapped the adjusted p-values from this model grouping them as p ≥ 0.05, p < 0.05 and p < 0.01. This map presents results for food insecurity, rurality, and higher education.