Adolescents often prefer late evenings and late mornings, a chronotype influenced by circadian and homeostatic sleep regulation changes during puberty. The circadian system naturally delays sleep onset, while homeostatic sleep pressure is also modified, allowing extended wakefulness. This conflict between delayed sleep and early social obligations leads to sleep deprivation. Most adolescents sleep less than the recommended 8-10 hours daily. Delaying school start times is proposed as an intervention to increase sleep, although objective data supporting this has been lacking. Increased daily sleep is crucial for adolescent physical and mental health, and for learning and memory. However, the link between longer sleep and better school performance in field studies hasn't been definitively established, especially in relation to delayed school start times.

Existing studies, primarily relying on self-reported data, suggested a correlation between later school start times and increased reported sleep duration in adolescents. However, this study aimed to provide objective, quantitatively measured data to confirm the hypothesis that delaying school start times directly increases sleep and subsequently impacts academic performance. Previous research on adolescent sleep regulation highlighted the interplay between circadian rhythms and homeostatic sleep drive, emphasizing the need for interventions targeting this complex dynamic. The connection between adequate sleep and improved academic outcomes has also been explored, yet field studies have yielded inconsistent results, necessitating further investigation in this specific area of intervention.

This pre-/post-study utilized data from two Seattle public high schools before and after the implementation of a 55-minute delay in secondary school start times (from 7:50 a.m. to 8:45 a.m.). Sophomores in the same science class were studied in each year. Students wore Actiwatch Spectrum Plus wrist activity monitors for two weeks during the spring of both years (2016 pre-intervention and 2017 post-intervention), providing objective sleep-wake cycle data. Sleep diaries, the Beck Depression Index II, the Epworth Sleepiness Scale, and chronotype questionnaires (Munich and Horne-Östberg) were also administered. Data analysis included two-way ANOVAs for activity and light exposure, Wilcoxon signed-rank tests for sleep parameters (onset, offset, duration), and generalized linear models to assess the relationships between sleep, academic performance (second-semester grades), sleepiness, mood, chronotype, and attendance (absences and tardies). The study was approved by the University of Washington's Human Subject Division and the Seattle Public School District Board.

The later school start time resulted in a 44-minute later median sleep offset (P < 0.0001) and a 34-minute increase in median sleep duration (P = 0.0007) on school days in 2017 compared to 2016. There were no significant differences on non-school days. Students in 2017 also showed a significant decrease in social jet lag (P = 0.0118), indicating better alignment between school-day and non-school-day sleep. Morning light exposure was significantly delayed in 2017, but evening light exposure wasn't. Generalized linear models revealed that later sleep offset (P = 2.8 × 10⁻⁵), increased academic performance (P = 0.0261), and reduced daytime sleepiness (P = 0.0370) were significant factors distinguishing the two years. Franklin High School (FHS), with a higher percentage of economically disadvantaged students and ethnic minorities, showed significant improvements in attendance and punctuality in 2017 compared to Roosevelt High School (RHS).

This study provides objective evidence that delaying high school start times leads to increased sleep duration in adolescents. The 34-minute increase in sleep is substantial, especially considering the historical trend of decreased sleep in children over the past century. The association between later start times and improved academic performance, though correlational, suggests that better rest enhances alertness and learning. The significant improvement in attendance and punctuality at FHS, more so than at RHS, highlights the potential for such interventions to address inequalities in educational outcomes. These findings align with previous research, but this study is the first to utilize actimetry to demonstrate a significant increase in objectively measured daily sleep following a district-wide change in school start times. While the study didn't achieve a full 55-minute increase in sleep, indicating potential bedtime adjustments, it still provides compelling support for adjusting school start times to better accommodate adolescent sleep needs.

This study demonstrates that a later high school start time is associated with significantly increased sleep duration, improved academic performance, reduced sleepiness, and better attendance, particularly among economically disadvantaged students. These results highlight the importance of aligning school schedules with adolescents' biological sleep rhythms. Future research could investigate the optimal start time for maximizing benefits, the long-term effects of later start times, and the interaction between start times and other factors influencing adolescent sleep and academic success, such as sleep hygiene interventions.

The pre-/post-study design lacks a control group, limiting the ability to definitively establish causality. The reliance on second-semester grades may reflect teacher bias related to the school start time change. While the Actiwatch provided objective sleep data, self-reported data from diaries could be subject to recall bias. The study's focus on two high schools in Seattle may limit the generalizability of findings to other contexts with differing student demographics or school environments.