Sleep disruption is known to be an obesity risk factor, however, little is known about its interplay with genetic predisposition and pathways involved in obesity pathogenesis, especially in the longitudinal setting. Leveraging a large cohort from the Beijing Child and Adolescent Metabolic Syndrome (BCAMS) study, we aimed to examine a possible sleep-gene interaction for childhood obesity risk, including longitudinal assessment in a 10-year follow-up and further test if there is any mediation through the leptin pathway. A total of 3,211 children (6-18 years) were recruited from this cohort study. Baseline leptin levels and twelve established adult BMI loci were examined for the associations with habitual sleep duration. After adjusting for pubertal stages, diet score and activity, short sleep duration at baseline was significantly associated with increased overweight/obesity risk at both baseline and follow-up. Genetic predisposition scores, particularly consisting of the central nervous system (CNS) gene variation (GPSCNS), were robustly correlated with baseline overweight/obesity in children who slept ≤8h/day (P<0.001), whereas the association was ablated in those who slept ≥10h/day (P>0.05). Comparable observations were made at follow-up. Mediation analysis revealed a modest direct effect of the GPSCNS-sleep interaction on BMI at baseline, while an indirect effect of this interaction was found to be mediated principally through elevated leptin (52.6%); moreover, the mediation effect via leptin remained stable over 10 years. This study provides longitudinal evidence that short sleep duration in children has a long-term impact on the association of polygenic risk for obesity from childhood to young adulthood and leptin pathway explains a key mechanism via a modification effect. Therefore, adequate sleep duration during childhood is important for the early prevention of obesity, especially if there is a genetic predisposition to this trait.


M. Li: None. J. Fu: None. L. Han: None. G. Li: None. S. Gao: None. S.F. Grant: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at