Sleep Duration and Adiposity in Children and Adults: Observational and Mendelian Randomization Studies.

<bold>Objective: </bold>This study used two complementary designs, an observational and a Mendelian randomization (MR) study, to assess whether sleep duration causes adiposity in children and adults.<bold>Methods: </bold>In Hong Kong's "Children of 1997" birth cohort, the adjusted cross-sectional associations of sleep duration with BMI z score and obesity and overweight were assessed at ~11 years of age. Generalized estimating equations were also used to examine longitudinal associations of sleep duration at ~11 years with annual BMI z score and obesity and overweight at about 11 to 16 years of age. Using MR, this study assessed the association of genetically predicted sleep duration, based on 54 single-nucleotide polymorphisms, applied to genetic studies of adiposity in children (n = 35,668), men (n = 152,893), and women (n = 171,977).<bold>Results: </bold>Longer sleep was cross-sectionally associated with lower BMI z score at ~11 years of age (-0.13 per category, 95% CI: -0.22 to -0.04) and at about 11 to 16 years of age longitudinally in girls (-0.39, 95% CI: -0.66 to -0.13). Using MR, sleep duration was inversely associated with BMI in children (-0.29 SD per hour, 95% CI: -0.54 to -0.04), but was not clearly associated with BMI in adults, particularly for women.<bold>Conclusions: </bold>A small beneficial effect of sleep on BMI in children cannot be ruled out. [ABSTRACT FROM AUTHOR]