Evening Phone Use and Sleep Outcomes Among Youth

Importance Smartphone use has been associated with poor sleep, but existing evidence primarily relies on self-reported, cross-sectional data. Objective To examine within-person and between-person associations between passively sensed evening smartphone use and objectively measured sleep outcomes in adolescents. Design, Setting, and Participants This intensive longitudinal cohort study assessed passively sensed data from US youths in the Adolescent Brain Cognitive Development Study. Participants were Android users with at least 14 days of concurrent Effortless Assessment Research System application data and Fitbit-measured sleep. Data were collected between November 30, 2020, and February 5, 2024, and analyzed from February 2 to June 1, 2026. Analysis used within-between decomposition in multilevel models to separate within-person from between-person associations. Exposures Smartphone use after each hourly cutoff from 8 pm to midnight (cumulative) and within 5 to 30 minutes before sleep onset (proximal) were quantified. Main Outcomes and Measures Fitbit-derived sleep measures were assessed, including total sleep time (TST), onset, offset, and sustained and total wake after sleep onset (WASO). Results A total of 425 participants (mean [SD] age, 14.8 [1.1] years; 216 [50.8%] female) contributed 7591 person-nights of data. On nights when adolescents used their phones more than their own evening average (after 10 pm ), they slept a mean (SD) of 20.05 (49.42) minutes less and fell asleep a mean (SD) of approximately 64.73 (50.95) minutes later. Adjusting for covariates, within-person analyses revealed that each additional minute of smartphone use after 10 pm above a participant’s own average was associated with sleep onset that was 0.87 (95% CI, 0.84 to 0.90) minutes later, sleep offset that was 0.37 (95% CI, 0.33 to 0.41) minutes later, and a TST change of −0.43 (95% CI, −0.47 to −0.39) minutes (all P < .001). Estimates increased monotonically across cutoffs from 8 pm to midnight. Proximal use was associated with greater sustained WASO, largest in the final 5 minutes of use before sleep onset (0.74 [95% CI, 0.50 to 0.98] additional minutes of wakefulness; P < .001) and negligible when looking at the final 20 minutes of use before sleep onset. This association was concentrated in the first 30 minutes after sleep onset. Conclusions and Relevance This cohort study found that greater cumulative late-evening phone use was associated with later sleep onset, delayed sleep offset, and shorter TST. Smartphone use immediately before sleep onset was associated with greater WASO, primarily during the first 30 minutes after sleep onset. These findings suggest that reducing evening smartphone use, particularly after 10 pm and immediately before bedtime, may help improve sleep health in adolescents.