Daily fluctuations in adolescents' sleep predict next-day attention, sleepiness, and fatigue: an ecological momentary assessment study over 28 days

Abstract

Background: Current understanding of the associations between adolescents' daily sleep and daytime alertness and fatigue under naturalistically occurring restricted (school) and unrestricted (vacation) sleep opportunities is limited. Methods: A convenience sample of adolescents (n = 205; 54.1% females, Mage ± SD = 16.9 ± 0.87 years) completed daily measures of sleep, alertness, and fatigue over 28 days (2 weeks during school, and the subsequent 2-week vacation). Actigraphy and sleep diary total sleep time (TST) and sleep efficiency (SE) were measured. Participants self-reported sleepiness and fatigue every morning and afternoon, and completed a tablet-based, 3.2-min psychomotor vigilance task (PVT) every afternoon. Cross-lagged multilevel models tested daily TST and SE as predictors of next-day subjective sleepiness/fatigue and PVT performance. Between- (i.e., differences between individuals) and within-person associations (i.e., whether nights with higher-than-individual's-average TST/SE, predict next-day outcomes) were tested simultaneously. Covariates included previous-day outcome, day of the week, study day (1–28), school/vacation, chronotype, and sociodemographic variables. Results: Within-persons, higher-than-average TST and SE (both actigraphy and diary) predicted better next-day PVT performance (all p ≤.006), and lower subjective sleepiness and fatigue the following morning and afternoon (all p ≤.032). Between-persons, adolescents with higher overall diary SE had lower morning subjective sleepiness (p