Table 2.

Description of regression models included in analysis

Model	Counties included	Independent variables included
Model 1 (individual SDH)	2026 counties with ≥ 5 deaths	Population density, days since first death, % over age 65, % obesity, % diabetes, % COPD, % smokers, % hypertension, one SDH, or %Black residents
Model 2 (full SDH)	2026 counties with ≥ 5 deaths	Population density, days since first death, % over age 65, % obesity, % diabetes, % COPD, % smokers, % hypertension, % Black residents, ICE income quintile, % uninsured, % low birthweight, % adults without HS diploma, incarceration rate, % households without internet
Model 3a (subgroup low adverse SDH)	1013 counties with ≥ 5 deaths and lowest half of adverse SDH	Population density, days since first death, % over age 65, % obesity, % diabetes, % COPD, % smokers, % hypertension, % Black residents
Model 3b (subgroup high adverse SDH)	1013 counties with ≥ 5 deaths and highest half of adverse SDH	Population density, days since first death, % over age 65, % obesity, % diabetes, % COPD, % smokers, % hypertension, % Black residents
Model 3c (interaction)	2026 counties with ≥ 5 deaths	Population density, days since first death, % over age 65, % obesity, % diabetes, % COPD, % smokers, % hypertension, % Black residents, one high SDH binomial, % Black residents * high SDH binomial