A quantitative intersectional approach to measuring change in built environment amenities and the association with mortality in U.S. counties

BACKGROUND: The relationship between health and the built environment is well established, but common methods conceptualizing "place" rely on static or demographic-based interpretations that do not capture its complexity. Marrying Intersectionality Theory and political economics, we conceptualize place as a manifestation of complex, mutually constructive forces that shape the lived experience, and assert that quantitative intersectionality is essential to understanding the role of place in health outcomes. METHODS: NaNDA neighborhood data were converted to person-area density across social determinants of health (SDoH) domains at the county level for each year between 2014 and 2019. Log-transformed densities were regressed against year, summed, and exponentiated to calculate the composite average annual change in a county's SDoH amenities. This composite change trend was regressed against county-level 2020 and 2021 COVID-19 and cardiovascular disease (CVD) mortality using multilevel Bayesian models adjusted for year and older population percentage, with state and SDoH intersectional strata as random intercepts. RESULTS: A 0.1-unit increase in the composite change trend-approximately equivalent to the IQR of the composite trend-was associated with a 4.10% reduction in COVID-19 mortality rate (IRR = 0.959, 95% CrI = [0.944, 0.975]), all else held equal. A 0.1-unit increase in the composite change trend was significantly associated with a 3.15% reduction in a county's CVD mortality rate (IRR = 0.969, 95% CI = [0.959, 0.986]). CONCLUSIONS: This novel approach to conceptualizing the built environment in epidemiological research supports the growing conversations among the fields of economics, urban planning, and public health.