The COVID-19 pandemic has disrupted daily life in many ways, including changes that could either improve or impede sleep duration among adolescents. Periods of remote learning may have provided opportunities for adolescents to sleep late; findings from some small studies suggest adolescents may have shifted to later bedtimes and wake times and slept longer (1–3). Short sleep duration among adolescents is linked to higher risk of injury, worse metabolic and mental health, and difficulty focusing (4). The prevalence of short sleep duration and its association with difficulty doing schoolwork and poor mental health during the COVID-19 pandemic have yet to be explored in a nationally representative sample. Examining this association and estimating the co-occurrence of short sleep duration with poor mental health might provide schools with an additional rationale to adopt policies that lead to improvements in sleep duration within a comprehensive approach to support student academic achievement and mental health.

We used data from the Adolescent Behaviors and Experiences Survey (ABES) — a one-time, nationally representative, cross-sectional survey of high school students (grades 9–12) — fielded from January through June 2021 (N = 7,705). The overall ABES response rate during the study period was 18% (school response rate [38%] × student response rate [48%]) (5). Details on ABES are available elsewhere (www.cdc.gov/healthyyouth/data/abes.htm). ABES was reviewed and approved by institutional review boards at the Centers for Disease Control and Prevention and ICF International (5).

Data on sleep duration were derived from the question “On an average school night, how many hours of sleep do you get?” (Table 1). To align with American Academy of Sleep Medicine recommendations (8–10 h for ages 13–18 y) (4), we restricted analyses to participants who reported their age as 13 years or older. We combined the response options for 8, 9, and 10 or more hours into a single category (≥8 h), which resulted in 5 categories (≤4, 5, 6, 7, or ≥8 h). We defined short sleep duration as sleeping ≤4, 5, 6, or 7 hours and ≥8 hours as meeting sleep recommendations. Data on the primary outcome was derived from the question “Do you agree or disagree that doing your schoolwork was more difficult during the COVID-19 pandemic than before the pandemic started?” with “strongly agree” and “agree” coded as experiencing “more difficulty doing schoolwork” vs “not sure,” “disagree,” or “strongly disagree.”

Bivariate analyses (χ² tests, Pearson correlation) and univariate logistic regression models (Wald F test) assessed the associations between difficulty doing schoolwork, short sleep duration, and selected covariates (poor mental health, high level of screen time [≥5 h/d], and self-reported sex, race and ethnicity, grade, and hunger). An adjusted logistic regression model predicted prevalence ratios (PRs) for experiencing more difficulty doing schoolwork, including the covariates. We tested whether poor mental health modified the association between sleep duration and more difficulty doing schoolwork (interaction term: sleep duration × poor mental health). Analyses were conducted in SAS-callable SUDAAN version 11.0.3 (RTI International) and used sample weights to account for complex sampling and nonresponse. Statistical significance was set at P < .05.

The sample was evenly distributed across sex and grade and racially and ethnically diverse; no racial or ethnic group comprised a majority (Table 2). Most high school students (76.5%) experienced short sleep duration, and 66.6% reported more difficulty doing schoolwork during the COVID-19 pandemic than before the pandemic.

Overall, 37.1% reported poor mental health during the pandemic, which correlated with short sleep duration (Pearson correlation r = 0.22, P < .001; χ² ₄ = 347.48, P < .001). Among students who met sleep recommendations, 25.2% reported poor mental health. About half of students who slept 5 (49.1%) or 4 hours or less (55.9%) reported poor mental health (P < .001).

The unadjusted models confirmed the hypothesized association between short sleep duration and greater difficulty doing schoolwork, which remained robust after adjusting for covariates (Table 3). Students who slept less than 7 hours during an average school night had a significantly greater prevalence of experiencing more difficulty doing schoolwork during the COVID-19 pandemic compared with students who met sleep duration recommendations (6 h sleep: PR = 1.17 [95% CI, 1.08–1.27]; 5 h sleep: PR = 1.18 [95% CI, 1.09–1.28]; ≤4 h sleep: PR = 1.20 [95% CI, 1.08–1.33] vs ≥8 h sleep).

Students who experienced poor mental health had 17% higher prevalence of more difficulty doing schoolwork compared with students who did not report poor mental health (PR = 1.17; 95% CI, 1.10–1.25). Poor mental health did not moderate the association between short sleep duration and more difficulty doing schoolwork. Students with 5 or more hours of screen time were slightly less likely to report more difficulty doing schoolwork than students who spent less than 5 hours per day using screens (PR = 0.95; 95% CI, 0.91–0.997). Self-reported sex, grade, and hunger were not associated with more difficulty doing schoolwork in the adjusted model. Hispanic or Latino students were more likely to report more difficulty doing schoolwork than White students.

Before the COVID-19 pandemic, short sleep duration was becoming more prevalent among US high school students (74.6% [73.1%–76.0%] in 2017 and 77.9% [76.3%–79.4%] in 2019, P < .001) (6). We found that short sleep duration remained widespread during the COVID-19 pandemic, affecting roughly three-quarters of students. Students who slept less than 7 hours during an average school night were more likely to report greater difficulty doing schoolwork during the pandemic compared with before the pandemic, as were those who experienced poor mental health. This study contributes to the literature by highlighting the co-occurrence of short sleep duration and poor mental health during the COVID-19 pandemic. Teachers have identified behavioral and mental health challenges among the leading barriers to addressing learning gaps during the 2022–2023 school year (7). Policies known to improve sleep duration among students, including later school start times and family practices, such as good sleep habits and parent-set bedtimes, might help support both learning and mental health (8,9).

Previously noted limitations to ABES include the low response rate and the inability to draw causal inferences about the impact of the COVID-19 pandemic (5). We note additional limitations. The recall period differed across some of the items: the sleep and screen time questions reference the average school day or night, whereas items about mental health and greater difficulty doing schoolwork reference “during the COVID-19 pandemic.” We do not know how students interpretated “average school day.” Rapid survey development during the COVID-19 pandemic precluded item validation and testing. Previous univariate models confirmed small significant associations in the expected direction between short sleep duration, self-reported hunger (10), and poor mental health (4), which suggests concurrent validity. We cannot account for some potential confounders, such as socioeconomic status and school instruction modality (eg, remote, hybrid, in-person). The latter may have influenced wake times and sleep duration, even if school start times were unchanged. Instruction modality may also have influenced self-reported screen time; the question excludes time doing schoolwork but does not specify whether to count time using screens to attend school. This may help explain the counterintuitive finding that a high level of screen time was inversely associated with greater difficulty doing schoolwork. ABES addressed a single dimension of sleep; we could not assess sleep quality, sleep schedules, or sleep disorders.

Nevertheless, our findings show that most students were not sleeping enough and that many students concurrently experienced poor mental health and insufficient sleep. Schools can consider addressing sleep duration within a broader strategy to bolster adolescent mental health and learning, including addressing protective factors (11,12).