Comparing measures of free-living sleep in school-aged children.

Abstract

OBJECTIVE/BACKGROUND\nRecent technological advances and emerging commercially-available consumer-friendly sleep assessment products affords researchers with a host of tools to consider for capturing free-living sleep in children. The purpose of this study was to compare free-living sleep characteristics (duration and bed/wake times) across different measures in children.\n\n\nMETHODS\nElementary school-aged children (N\xa0=\xa030, mean age 7.2 years, 63% boys, 87% non-Hispanic white) wore an ActiGraph GT9X Link© and Fitbit Charge HR© on the non-dominant wrist, with a Beddit 3 Sleep Monitor© affixed to their mattress for two consecutive weekend nights of free-living sleep. Parents completed a sleep log of bed and wake times. Absolute differences in bed and wake times were examined and Bland Altman plots assessed the level of agreement across sleep measures.\n\n\nRESULTS\nAcross the four sleep measures, total sleep time (TST) ranged from 458\xa0min/night (ActiGraph GT9X Link©) to 613\xa0min/night (Parent report). Mean bed and wake times ranged from 8:46PM to 9:03PM, and 6:52AM to 7:16AM, respectively. Pearson correlation coefficients were moderate between all four sleep measures (range r\xa0=\xa00.30-0.71). Bland-Altman plots indicated the highest level of agreement for TST was between Beddit 3 Sleep Monitor© and Fitbit Charge HR© (mean difference\xa0-11.7, limits of agreement: 119.0,\xa0-142.4\xa0min).\n\n\nCONCLUSIONS\nThe findings from this study show a high level of agreement of when a child goes to sleep and wakes up across a variety of sleep measures; however, more work is needed to classify TST once the sleep period has commenced.