Concurrent Associations Of Physical Activity And Screen-based Sedentary Behaviors On Sleep Duration Among Us Adolescents

Abstract

Adolescence is a vulnerable period for experiencing insufficient sleep due to puberty-related changes in circadian rhythm and increasing sleep deprivation during school days. Physical activity (PA) and sedentary behaviors (SB) are well-documented, modifiable sleep hygiene behaviors independently associated with sleep duration; however, scarce data are available regarding their concurrent associations in adolescents. PURPOSE: This study examined the concurrent associations of PA and screen-based SB on sleep duration in US adolescents using a latent class analysis (LCA) approach. METHODS: A total of 11,204 adolescents who participated in the 2013 Youth Risk Behavior Survey were analyzed. The outcome variables of interest included self-reported PA regarding 1) moderate and vigorous-intensity PA ≥ 60 minutes/day; 2) sport team participation ≥ 1 per year; and 3) muscle-strengthening exercise ≥ 3 days/week; and self-reported screen-based SB regarding 4) watching TV ≥ 3 hours/day; and 5) using a computer/playing video games ≥ 3 hours/day. Self-reported sleep durations on average school nights was obtained to determine the sufficient sleep (≥8 hours). A LCA model was developed 1) to identify the latent subgroups with varying response probabilities of each PA and screen-based SB items; and 2) to examine the associations of latent subgroups with likelihood of having sufficient sleep. RESULTS: Four latent subgroups with varying levels of PA and screen-based SB were identified. Using the estimated response probability ≥ .50 as a threshold to determine ‘High’ level of respective PA and screen-based SB items, four latent subgroups were characterized as the High PA/Low SB (26.06%), High PA/High SB (23.23%), Low PA/Low SB (29.41%), and Low PA/High SB (21.29%). The likelihoods of having sufficient sleep was greater for the High PA/Low SB when comparing to the High PA/High SB (OR = 1.51) and Low PA/Low SB (OR = 1.49), whereas no difference was found when comparing to the Low PA/High SB. CONCLUSIONS: The results demonstrated the complexity of concurrent associations of PA and screen-based SB with sleep duration in adolescents. However, the findings generally implied that increasing PA and reducing screen-based SB would yield greater likelihood of having sufficient sleep in this population.