Quantity versus quality of objectively measured sleep in relation to body mass index in children: cross-sectional and longitudinal analyses

Publisher Website
Google Scholar
Abstract
Background/objectives Although sleep duration is well established as a risk factor for child obesity, how measures of sleep quality relate to body size is less certain. The aim of this study was to determine how objectively measured sleep duration, sleep timing, and sleep quality were related to body mass index (BMI) cross-sectionally and longitudinally in school-aged children. Subjects/methods All measures were obtained at baseline, 12 and 24 months in 823 children (51% female, 53% European, 18% Māori, 12% Pacific, 9% Asian) aged 6–10 years at baseline. Sleep duration, timing, and quality were measured using actigraphy over 7 days, height and weight were measured using standard techniques, and parents completed questionnaires on demographics (baseline only), dietary intake, and television usage. Data were analysed using imputation; mixed models, with random effects for person and age, estimated both a cross-sectional effect and a longitudinal effect on BMI z-score, adjusted for multiple confounders. Results The estimate of the effect on BMI z-score for each additional hour of sleep was −0.22 (95% CI: −0.33, −0.11) in cross-sectional analyses and −0.05 (−0.10, −0.004) in longitudinal analyses. A greater effect was observed for weekday sleep duration than weekend sleep duration but variability in duration was not related to BMI z-score. While sleep timing (onset or midpoint of sleep) was not related to BMI, children who were awake in the night more frequently (0.19; 0.06, 0.32) or for longer periods (0.18; 0.06, 0.36) had significantly higher BMI z-scores cross-sectionally, but only the estimates for total time awake (minutes) were significant longitudinally (increase in BMI z-score of 0.04 for each additional hour awake). Conclusion The beneficial effect of a longer sleep duration on BMI was consistent in children, whereas evidence for markers of sleep quality and timing were more variable.