Article Title



Patricia Pagan Lassalle, Lauren C. Bates, Peter Halpin, Zachary Y. Kerr, Erik D. Hanson, FACSM, Michelle L. Meyer, Lee Stoner, FACSM. The University of North Carolina at Chapel Hill, Chapel Hill, NC.

BACKGROUND: Beneficial 24-hour activity behaviors (i.e., limiting sedentary behavior, physical activity, sleep) are critical for chronic disease prevention, but were altered by the COVID-19 pandemic. Less time in beneficial 24-hour activity behaviors may place individuals at greater risk for contracting COVID-19. Further, some populations may be more impacted than others, with demographic and environmental (rural/urban/suburban) factors influencing access to healthcare and COVID-19 information. METHODS: In December 2020, we recruited an online convenience sample of 746 adults aged 18 years or older residing in the US. Of the sample, 400 (39±14 years old, 52% female, 15% Hispanic, 78% White, and 28% had or survived cancer) had complete data for our variables of interest. Participants self-reported demographic information, COVID-19 diagnosis (yes/no), total weekday and weekend sedentary behavior (hours/day), moderate to vigorous physical activity (mins per week), sleep (hours/day), and environment (rural/urban/suburban). a) A k-fold cross-validation, machine learning variable selection (glmnet) approach was used to identify which variables were most strongly associated with COVID-19 prevalence, as determined by specifying the tuning parameter, λ, at its minimum value. This approach minimizes the elastic net penalty and optimizes the model fit. b) Logistic regression was used to calculate odds ratios (OR) and 95% confidence intervals (95%CI) for the strongly associated variables. RESULTS: a) The variables most associated with COVID-19 prevalence were total sedentary behavior during weekdays, sleep, biological sex, race, and environment (λ=0.012). b) Holding all other parameters constant, the odds of having COVID-19 were lower for every additional hour of sleep (OR=0.87, 95%CI=0.79, 0.96); identifying as non-white versus white (OR=0.46, 95%CI=0.23, 0.87); living in a rural versus urban area (OR=0.42, 95%CI=0.18, 0.92); and higher for being male versus female (OR=2.79, 95%CI=1.69, 4.67). Non-significant decreases were found for weekday sedentary behavior (OR=0.97; 95%CI=0.92, 1.02) and living in a suburban versus urban areas (OR=0.63, 95%CI=0.38, 1.05). CONCLUSION: Less sleep, identifying as non-white, female, and residing in rural areas were associated with a lower COVID-19 prevalence.

This document is currently not available here.