Effects of school start time on students' sleep duration, daytime sleepiness, and attendance: a meta-analysis
Bowers JM, Moyer A · 2017
grade Cmeta-analysisindependentreplicated
Sample
5 longitudinal studies and 15 cross-sectional comparison-group studies for sleep duration; 7 studies for daytime sleepiness; 3 for tardiness
Population
Children and adolescents in schools with differing start times
Design
Random-effects meta-analysis pooling five longitudinal and fifteen cross-sectional comparison studies. Three-quarters of the pooled evidence is cross-sectional, so the pooled estimate inherits substantial selection confounding; the authors themselves urge cautious interpretation and call for more longitudinal primary research.
Key findings
Later school start times are associated with longer sleep durations, less daytime sleepiness (7 studies), and less tardiness (3 studies). Direction is consistent with every other synthesis of this literature.
Genetic confound
Medium-high for the cross-sectional majority of the pooled sample, where families sort into schools; low for the longitudinal minority where the same students are observed across a bell-time change. The meta does not separate the two.
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| Sleep duration (health/behavioural) | pooled random-effects association | later start times associated with longer sleep duration | mixed | varies | none | unclear | health |
| Daytime sleepiness | pooled across 7 studies | less sleepiness with later start times | mixed | varies | none | unclear | health |
| Tardiness (attendance) | pooled across 3 studies | less tardiness with later start times | standardized | varies | none | unclear | behaviour |