Syntes: Albrecht et al. (2026) — when two Swiss schools let pupils choose a later start, 95% did, slept longer, and reported better health and maths and English scores

Källa: Albrecht, J., Risch, A., Huber, R. & Jenni, O. (2026). The Power of Flexible School Start Times: Longitudinal Associations with Sleep, Health, and Academic Performance. Journal of Adolescent Health 78(6), 865–872. https://doi.org/10.1016/j.jadohealth.2026.01.011

Verified content

• [abstract] Two Swiss secondary schools; 351 students answered the baseline survey (spring 2022) and 403 the follow-up (spring 2023), 212 both; median age 14.
• [abstract] 95% of students used the later-start option; median start time moved 38 minutes later (n = 711, β = .57 [.53, .62], p < .001), with later wake times and more sleep on school days; in the paired subsample the delay was associated with more school-day sleep (n = 205, β = .51 [.05, .94], p = .03).
• [abstract] No worsening was observed; fewer problems falling asleep, fewer students with clinically low health-related quality of life, and higher scores in mathematics and English.

Limits

• [assessment] Abstract only: the abstract does not say how mathematics and English scores were measured (school grades, self-report or tests) or how large the gain was. Before/after design without control schools; survey-based sleep.

Why it is in the library

A new European study of a different policy (choice of a later start rather than a mandated one) that reports both more sleep and better results; worth reading in full before any routing, since C14 turns on whether results improved, not just sleep.

Full text read (2026-10-25, nightly attack run)

• [full text] Read on the journal's article page (open access, CC BY). Academic results came from the school administration: grades (maths, German, English) did not change (math β = .01, p = .83; English β = .06, p = .19); standardized test scores rose in maths (β = 26.61 [10.02, 43.42] points on a 200–800 scale, p = .002) and English (β = 43.81 [24.34, 63.26], p < .001), not in German (p = .09) or French (p = .98); only some second and third graders sat the tests.
• [full text] Self-reported school-day sleep 8.25 h before vs 9.00 h after (medians). Authors' limits: no control sample; the new school model also changed learning formats (more self-directed learning, learning teams), so the later start cannot be separated from the rest; all students one year older at follow-up.
• [assessment] Not a well-controlled kill-condition test for C14 (no comparison schools, bundled reform); grades unchanged restates C14's existing consistency bound. Not routed.

Added 2026-10-25 by the project (weekly source sweep, kind: new; read: abstract). [abstract] = from the published abstract; [full text] = read in the full text; [assessment] = the project's own assessment.