Effects of school start time and its interaction with the solar clock on adolescents' chronotype and sleep: A systematic review and meta-analysis

37 studies comparing different school start times on adolescent sleep and/or chronotype

Adolescents in secondary schooling across multiple countries, including double-shift systems

Random-effects meta-analysis with meta-regressions on the size of the start-time difference and on the interplay between social (bell) time and solar time. Pools designs of very mixed quality — most included comparisons are cross-sectional or pre/post rather than controlled natural experiments — so the pooled estimate inherits that confounding even though the direction is consistent. Its distinctive contribution is showing the effect scales with the size of the delay and with solar timing.

Later start times are associated with later weekday sleep timing, longer weekday sleep duration, lower social jet lag, and later chronotype. Meta-regression: the magnitude of the sleep-duration effect scales with the distance between the compared start times and is modulated by the interaction between bell time and the solar clock — i.e. the same clock time is a different biological insult depending on where in a time zone the school sits. This is the dose-response and mechanism evidence that the individual district studies cannot supply.

Medium. Chronotype is itself heritable, and cross-sectional comparisons between schools with different bell times can reflect family sorting into schools and neighbourhoods. The solar-clock meta-regression partly defuses this, because longitude within a time zone is not chosen for genetic reasons, but the pooled main effect still rests substantially on confounded comparisons.