Validation of a multi-sensor activity monitor for assessing sleep in children and adolescents

ObjectiveTo assess the validity of a multi-sensor activity monitor in estimating sleep and wake compared to polysomnography in children and adolescents.MethodsA total of 43 children and adolescents (29 boys, 14 girls), aged 7–17 years (mean age [SD] = 11.0 [2.4] years) participated in the study. Participants wore the SenseWear Pro3 Armband™ (SWA) body monitor (BodyMedia Inc) during an overnight polysomnographic assessment in a paediatric sleep laboratory. Sleep measures included sleep onset latency (SOL), wake after sleep onset (WASO), total sleep time (TST) and sleep efficiency (SE).ResultsNo systematic bias of the SWA was noted for any of the sleep measures assessed, but limits of agreement were wide and amounted to −76 to +58 min for SOL, −75 to 102 min for WASO, −109 to +99 min for TST and −22 to +20% for SE. In addition, no effect of gender, age group (children versus adolescents) or overweight on the accuracy of the SWA was found.ConclusionsThe SenseWear Armband™ showed good agreement with polysomnography at the group level, while at the individual level rather, poor agreement between the two methods was observed. Consequently, at this point the use of the SWA in the clinical evaluation of sleep cannot be advocated.