Teenage night owls or early birds? Chronotype and the mental health of adolescents

Abstract

Evidence suggests that the timing of sleep (chronotype) impacts mental health in young people, but previous studies have not accounted for sleep duration or school start time in this association, or examined a broad range of mental outcomes. In this study, we investigated the association between chronotype and mental health in a representative sample of adolescents from the 2014 Canadian Health Behaviour in School-Aged Children survey (29,635 students, 362 schools). We examined positive and negative aspects of mental health, using scores for emotional problems (range 0-33), emotional well-being (0-22), behavioural problems (0-28) and prosocial behaviours (0-25). We estimated chronotype using the time of mid-sleep on weekends and examined the associations using multilevel regressions, adjusted for sleep duration, school start time, individual, family and geographic characteristics. The average time of mid-sleep (chronotype) was 04:11 hr. An hour delay in mid-sleep time was associated with more emotional problems (0.34 [95% confidence interval 0.23, 0.45] point higher score), more behavioural problems (2.0% [95% confidence interval 1.4%, 2.6%] higher score), less emotional well-being (0.19 [95% confidence interval 0.09, 0.20] point lower score), and fewer prosocial behaviours (0.18 [95% confidence interval 0.08, 0.29] point lower score). A later chronotype was associated with poorer mental health, independent of sleep duration and school start time, and across internalizing and externalizing mental health domains. Further research is needed to clarify the mechanisms underlying this association. The timing of sleep, and not just its duration, may be an additional consideration for youth mental health.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 38 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.