Severe Acute Respiratory Syndrome Coronavirus 2 Incidence and Risk Factors in a National, Community-Based Prospective Cohort of US Adults.

Abstract

BackgroundProspective cohort studies of SARS-CoV-2 incidence complement case-based surveillance and cross-sectional seroprevalence surveys.MethodsWe estimated the incidence of SARS-CoV-2 infection in a national cohort of 6,738 U.S. adults, enrolled March-August 2020. Using Poisson models, we examined the association of social distancing and a composite epidemiologic risk score with seroconversion. The risk score was created using LASSO regression to identify factors predictive of seroconversion. The selected factors were household crowding, confirmed case in household, indoor dining, gathering with groups ≥ 10, and no masking in gyms/salons.ResultsAmong 4,510 individuals with ≥1 serologic test, 323 (7.3%, 95% confidence interval [CI] 6.5%-8.1%) seroconverted by January 2021. Among 3,422 participants seronegative in May-September 2020 and retested during November 2020-January 2021, 161 seroconverted over 1,646 person-years of follow-up (9.8 per 100 person-years [95%CI 8.3-11.5]). Seroincidence rate was lower among females compared to males (IRR: 0.69, 95% CI 0.50-0.94) and higher among Hispanic (IRR: 2.09, 95% CI 1.41-3.05) participants compared to White non-Hispanic. In adjusted models, participants who reported social distancing with people they did not know (IRRalways vs. never: 0.42, 95% CI 0.20-1.0) and with people they knew (IRRalways vs. never 0.64, 95%CI 0.39-1.06; IRRsometimes vs. never 0.60, 95% CI 0.38-0.96) had lower seroconversion risk. Seroconversion risk increased with epidemiologic risk score (IRRmedium vs. low 1.68, 95% CI 1.03-2.81; IRRhigh vs. low 3.49, 95% CI 2.26-5.58). Only 29% of those who seroconverted reported isolating and 19% were asked about contacts.ConclusionModifiable risk factors and poor reach of public health strategies drove SARS-CoV-2 transmission across the U.S.