Social Distancing Associations with COVID-19 Infection and Mortality Are Modified by Crowding and Socioeconomic Status

Abstract

Background: The Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) virus has emerged as a worldwide public health emergency. Social distancing currently represents the primary public health approach to slowing disease transmission. However, to date, little evidence on the efficacy of these policies have been available, and nothing is known on the types of areas where these policies have been more or less effective in reducing incidence and mortality. Methods: We obtained county-level data on COVID-19 incidence and mortality, cell-phone based average social distancing (0-5, where higher numbers indicate more social distancing), and Census data on demographics and socioeconomic status. Using generalized linear mixed models with a Poisson distribution accounting for counties nested within states, we modeled associations between county-level social distancing and COVID-19 incidence and mortality. We used multiplicative interaction terms to determine if associations varied by demographics or socioeconomic status. Findings: In multivariable adjusted models, each unit increase in social distancing was associated with a 26% decrease (pInterpretation: County-level measures of social distancing lead to reductions in COVID-19 incidence and mortality, but was most effective in counties with lower percentages of black residents, higher median household incomes, and in counties with lower levels of household crowding. Funding: JEH and GA were supported by NIH/NIEHS P30 ES000002. TV was supported by NIH/NIDDK K01 DK125612. MDW was supported by CDC/NIOSH R01 OH011773 and the Brigham Research Institute Fund to Sustain Research Excellence.Declaration of Interests: Missing.