Abstract
Extreme storms across the United States pose substantial and unequally distributed health burdens. We linked 111,252 county-event observations from the US National Oceanic and Atmospheric Administration Storm Events Database (2005-2022) to the Index of Concentration at the Extremes (ICE), a measure of racialized economic concentration reflecting the relative concentration of high-income non-Hispanic White households compared with low-income Black households within U.S. counties, to examine whether ICE predicts storm-related injuries and deaths. Using multilevel zero-inflated negative binomial models with county random intercepts and population offsets, each one-quintile increase in ICE (greater concentration of high-income, non-Hispanic White households relative to low-income Black households) was associated with 52% lower rates of direct injuries (IRR = 0.48, 95% CI: [0.44, 0.53]) and 36% lower rates of direct deaths (IRR = 0.64, 95% CI: [0.59, 0.69]). Associations for indirect outcomes were less precise but directionally consistent (indirect injuries IRR = 0.70, 95% CI: [0.54, 0.91]; indirect deaths IRR = 0.66, 95% CI: [0.48, 0.89]). Spline models suggested elevated risk concentrated in the most disadvantaged counties, with risk flattening in more advantaged areas. County-level maps show that structural disadvantage and storm exposure frequency do not systematically co-locate, suggesting that realized harm is driven by social and infrastructural context rather than exposure alone. These findings identify racialized economic concentration as a structural determinant of disaster harm and support targeting resilience investments toward the most structurally disadvantaged and economically marginalized counties.