How did trends in sleep duration in 2020 compare to previous years and how did they vary by sex, race/ethnicity, and educational attainment?

Recent evidence utilizing online samples indicates that sleep patterns were significantly altered during the initial months of the SARS-CoV-2 (COVID-19) pandemic/lockdown. However, it remains less clear how sleep duration changed in population-based samples, in the later months of 2020, and across subpopulations. Here we used a population-based sample to document sleep duration trends for the entire year of 2020, compared these trends to the previous years of 2013, 2014, 2016, and 2018, and systematically analyzed whether self-reported sleep duration patterns in 2020 varied by sex, race/ethnicity, and educational attainment. Data were from the Behavioral Risk Factor Surveillance System (n = 2,203,861) and focused on Americans aged 18 years and older. Respondents self-reported the hours of sleep they got in a 24-h period. We fit multinomial and linear regression models to predict the category of sleep duration (six or fewer hours, seven to eight h (base), and nine or more hours) and the raw reports of sleep duration, net of demographic, socioeconomic, and behavioral health covariates. Results revealed significant increases in sleep duration during the months directly after the COVID-19 lockdown (March and April in particular). However, these increases were short lived; reports of sleep duration reverted to historical levels by the Fall of 2020. We also found that the changes in sleep duration trends in 2020 were similar by sex, race/ethnicity, and educational attainment, cumulatively leading to little impact to disparities in sleep duration. In a dramatic, but brief, alteration of population-level sleep duration patterns, disparities in self-reported sleep duration remained intractable.