Shorter sleep, higher odds: a nationally representative analysis of the dose-dependent link to depression among 48.8 million U.S. youth

Objective While short sleep is linked to depression in youth, the precise dose-response relationship and age-specific risk thresholds remain poorly defined. We aimed to characterize this association and identify optimal sleep durations across development in a nationally representative sample. Methods Cross-sectional data from the 2020–2023 National Survey of Children's Health included 126,407 youth aged 6–17 years. Caregiver-reported sleep duration was modeled as a continuous variable using restricted cubic splines and as a categorical variable. Current depression was defined by caregiver-reported clinician diagnosis. Survey-weighted logistic regression adjusted for sociodemographic and health covariates. Results A strong, non-linear inverse association was observed between sleep duration and depression risk ( P -non-linearity <0.001). The dose-response curve plateaued at ∼10 h for ages 6–9, approximately 8.5 h for male aged 10–13, and ∼8.3 h for ages 14–17. Each additional hour of sleep up to these thresholds was associated with significantly reduced odds of depression. The association was strongest in children without ADHD, those in better health, and females ( P -interaction <0.05). Mediation via the tested pathways (household income, caregiver mental health, child general health, ADHD diagnosis, and caregiver education) accounted for <4% of the total effect. Conclusion This study identifies developmentally specific sleep thresholds beyond which the protective effect on depression risk plateaus. These findings provide empirical benchmarks for age-graded sleep recommendations and mental health screening strategies in pediatric populations.