A meta-analysis of accelerometer sleep outcomes in healthy children based on the Sadeh algorithm: the influence of child and device characteristics

Abstract Study Objectives Children often experience sleep problems, with a negative impact on mood, behavior, cognitive function, and other aspects of mental and physical health. Accelerometers are widely used to assess sleep, but general reference values for healthy children do not yet exist. The aim of this meta-analysis was to determine mean values for wake after sleep onset (WASO), sleep efficiency (SE), total sleep time (TST) and sleep onset latency (SOL), and to determine the effect of child and accelerometer-characteristics. Methods A search included studies with healthy children, 0–18 years, reporting WASO, SE, TST, and/or SOL, calculated with the Sadeh algorithm. Meta-analyses with random effects produced pooled estimate means per outcome. Meta-regression analyses determined the effect of age, sex, placement site and accelerometer type. Results Eighty-three studies (9,068 participants) were included. Pooled means were 63 min (95% CI 57 to 69) for WASO, 88% (95% CI 87 to 89) for SE, 477 min (95% CI 464 to 491) for TST and 19 min (95% CI 17 to 22) for SOL. Heterogeneity was high (95%–99%). TST decreased with age and there was an age-effect on SOL. SE differed between wrist and ankle (used in age 0–24 months) placement, and between piezoelectric and MEMS-type accelerometers. No differences were found between boys and girls, although this number of studies was small. Conclusions We found differences in almost all investigated outcomes and heterogeneity was high. Therefore, we advise to use a study-specific control sample until more robust reference values are available. Future research should narrow the methodological heterogeneity and produce larger datasets, needed to establish these reference values.