Relative Contributions of Weekday Sleep Duration and Subjective Sleep Sufficiency to Mental Health in Adolescents

Abstract

Objectives: This study examined the associations of weekday sleep duration and subjective sleep sufficiency with mental health outcomes among Korean adolescents. It also compared the relative contribution of these two sleep indicators to model explanation.Methods: Data were obtained from the 2024 Korea Youth Risk Behavior Survey, and 54,653 adolescents were included in the analysis. Weekday sleep duration was calculated from reported bedtimes and wake-up times, and subjective sleep sufficiency was assessed using perceived recovery after sleep. Survey-weighted logistic regression models were used to estimate associations with high stress, depressive symptoms, suicidal ideation, and anxiety. Because sleep duration and sleep sufficiency differed in scale and interpretation, their odds ratios were not directly compared. Instead, Shapley value-based pseudo-R² decomposition was used to compare how much each sleep indicator contributed to model explanation.Results: Longer weekday sleep duration was associated with lower odds of high stress, depressive symptoms, suicidal ideation, and anxiety, with adjusted odds ratio (ORs) ranging from 0.90 to 0.93. Poorer subjective sleep sufficiency was associated with higher odds of these outcomes, with adjusted ORs ranging from 1.39 to 1.65. In the Shapley-based decomposition, sleep duration accounted for 74.6% to 89.6% of the contribution attributable to the two sleep indicators, whereas sleep sufficiency accounted for 10.4% to 25.4%.Conclusions: Both sleep indicators were associated with the four mental health outcomes. However, sleep duration accounted for a larger share of the Shapley-based contribution than sleep sufficiency, ranging from 74.6% to 89.6% across outcomes. In this sample, actual weekday sleep time appeared to explain more of the mental health variation than perceived recovery after sleep.