Feasibility and impact on daytime sleepiness of an experimental protocol inducing variable sleep duration in adolescents

Although most research on sleep and adolescent health has focused on how long each youth sleeps on average, variability in sleep duration may be just as problematic. Existing findings have been inconsistent and unable to address cause-effect relationships. This study piloted an experimental protocol to induce sleep variability and explore its impact on daytime sleepiness in adolescents. Healthy adolescents aged 14-17 participated in a 3-week, at-home protocol. Sleep was monitored by sleep diaries and actigraphy. Following a run-in period to stabilize wake times (set at 6:30am throughout the protocol), participants were randomly counterbalanced across two 5-night experimental conditions. Bedtimes were consistent at 11:00pm during the stable sleep condition (7.5-hour sleep period each night) but changed on alternating nights during the variable sleep condition (ranging from 9:30pm to 12:30am) so that sleep duration averaged 7.5 hours across the condition with a standard deviation of 1.37 hours. Difficulty waking was assessed each morning and daytime sleepiness was assessed by end-of-condition parent- and adolescent-reports. Of the 20 participants who completed the study, 16 met the predetermined adherence definition. For those who were adherent, there were no differences in overall sleep duration between the stable and variable sleep conditions (p>.05) but adolescents had 58.6 minutes greater night-to-night variation in sleep duration in the variable condition (p < .001). Across all nights, youth reported greater difficulty waking following nights of shorter assigned sleep (p = .004) and greater overall sleepiness during the variable condition (p = .03). It is feasible to experimentally vary how long adolescents sleep on a nightly basis while holding average sleep duration constant. Such a protocol will promote tests of the acute effects of day-to-day changes in sleep duration on health.