Heterogeneous Associations Between Objectively Measured Screen Time and Sleep in Early Adolescence

Early adolescence is a time of risk for disrupted or insufficient sleep, and widespread use of electronic media may exacerbate this risk. However, findings linking screen time to sleep in adolescent samples are inconsistent, rely primarily on self-report, and seldom consider sleep variability. We examined associations between 24-hour total screen time scraped from mobile devices and actigraphy-assessed indicators of sleep timing and quality across one week (means, variability, and day-to-day fluctuations) in a community sample of 677 adolescent twins (Mean _age = 13.53, SD = 1.13; 50.96% female; 52.88% non-Hispanic White, 21.27% Hispanic/Latino), of whom 389 had actigraphy data and 488 had screen time data. For a subsample of Android users (n=171), hours of social media and mobile gaming were also considered. We conducted person-level associations using path analysis, with standard errors corrected for clustering of twins within families, and day-level analyses using three-level models. After multiple testing correction, greater person-level screen time was associated with later bedtime, and more variable wake time and duration, but not efficiency, latency, catch-up sleep, or social jetlag. More social media use was associated with later bedtime and greater duration variability. Mobile gaming was unrelated to sleep. Day-level screen time was weakly associated with later bedtime that night, with no other significant associations. Associations were mixed and modest, but not reliant on adolescent self-report. Findings do not provide strong evidence that moderate screen use is detrimental to sleep, but future studies should consider individual and contextual moderators and more nuanced objective measures.