ABSTRACT
OBJECTIVE: Coronavirus disease 2019 (COVID-19) has disproportionally affected communities of color. We aimed to determine what factors are associated with COVID-19 testing and test positivity in an underrepresented, understudied, and underreported (U3) population of mothers. METHODS: This study included 2996 middle-aged mothers of the Boston Birth Cohort (a sample of predominantly urban, low-income, Black and Hispanic mothers) who were enrolled shortly after they gave birth and followed onward at the Boston Medical Center. COVID-19 testing and test positivity were defined by the SARS-CoV-2 nucleic acid test. Two-probit Heckman selection models were performed to identify factors associated with test positivity while accounting for potential selection associated with COVID testing. RESULTS: The mean (SD) age of study mothers was 41.9 (±7.7) years. In the sample, 1741 (58.1%) and 667 (22.3%) mothers were self-identified as Black and Hispanic, respectively. A total of 396 mothers had COVID-19 testing and of those, 95 mothers tested positive from March 2020 to February 2021. Among a multitude of factors examined, factors associated with the probability of being tested were obesity (RR = 1.27; 95% confidence interval (CI): 1.08-1.49); and presence of preexisting chronic medical conditions including hypertension, asthma, stroke, and other comorbidities (coronary heart disease, chronic kidney disease, and sickle cell disease) with a corresponding RR = 1.40 (95% CI: 1.23-1.60); 1.29 (95% CI: 1.11-1.50); 1.44 (95% CI: 1.23-1.68); and 1.37 (95% CI: 1.12-1.67), respectively. Factors associated with higher incident risk of a positive COVID-19 test were body mass index, birthplace outside of the USA, and being without a college-level education. CONCLUSIONS: This study demonstrated the intersectionality of obesity and social factors in modulating incident risk of COVID-19 in this sample of US Black and Hispanic middle-aged mothers. Methodologically, our findings underscore the importance of accounting for potential selection bias in COVID-19 testing in order to obtain unbiased estimates of COVID-19 infection.