Aaron Spaulding

Purpose To examine changes in obesity prevalence among US adults after the COVID-19 pandemic by level of stay-at-home order and sociodemographic characteristics. Design Quasi-experimental study using repeated cross-sectional data. Setting Behavioral Risk Factor Surveillance System (BRFSS). Sample Pooled data for US adults ages ≥26 years (n = 1,107,673) from BRFSS (2018-2021). Measures States/territories were classified into three levels of stay-at-home order: none, advisory/only for persons at risk, or mandatory for all. Individual-level sociodemographic characteristics were self-reported. Analysis The difference-in-differences method was conducted with weighted multiple logistic regression analysis to examine obesity (body mass index ≥30 kg/m 2 ) prevalence by stay-at-home order level and sociodemographic characteristics before/after the COVID-19 pandemic (January 2018-February 2020 vs March 2020-February 2022). Results After adjusting for a secular trend and multiple covariates, adults in states/territories with mandatory stay-at-home orders experienced a larger increase in obesity prevalence (adjusted odds ratio: 1.05; 95% confidence interval: 1.01, 1.11) than adults in states/territories with no stay-at-home order. Younger adults (vs ≥65 years) and individuals with <high school education (vs ≥4-year college) experienced larger increases in obesity prevalence. Conclusion Increases in obesity prevalence were perpetuated 2 years after implementation of stay-at-home orders, indicating that longer-term health implications are co-occurring during the pandemic recovery period. Future research should focus on successful methods for staying active and healthy during social distancing and beyond.