Time-varying within-person association between daily physical activity and subsequent sleep in children and adolescents.

Abstract

While daily physical activity may benefit sleep in young people, it is unclear whether this association differs by time of day. This study examined the within-person time-varying association between physical activity and sleep in children and adolescents, and whether it differed based on age or sex. A total of 1175 children and adolescents (51% female, Mage = 9 years, SD = 4) from the National Health and Nutrition Examination Survey National Youth Fitness Survey were included. Physical activity and sleep were assessed using accelerometers. Activity volume was quantified using Monitor Independent Movement Summary (MIMS) units, aggregated in 10-min blocks from wake time to sleep onset. The 30-min window with the highest average MIMS/min was identified as the most intense period of daily activity. Sleep onset, duration, and efficiency were estimated using a machine learning algorithm. Time-varying associations were estimated using generalized additive mixture models, stratified by sex and age. Greater MIMS/min on a given day was associated with earlier sleep onset and longer sleep duration that night across nearly all time points in the 800 min before sleep onset, with associations strengthening closer to sleep onset. The relationship between MIMS/min and sleep efficiency was positive and more consistent across the day. Results were consistent across sexes while the association between MIMS/min close to bedtime was stronger in adolescents. Peak physical activity periods closer to sleep onset were associated with earlier sleep onset and longer sleep duration. These findings highlight the potential benefits of late afternoon and evening physical activity on sleep in young people.