Agreement of different methods for assessing sleep characteristics: a comparison of two actigraphs, wrist and hip placement, and self-report with polysomnography

Abstract

Objective To assess the agreement of sleep parameters measured by two actigraphs (SOMNOwatch plus, ActiGraph GT3X+) at two different placements (wrist, hip) and of self-reported sleep with polysomnography (PSG).Methods We estimated agreement with PSG for total sleep time (TST), sleep onset latency (SOL), wake after sleep onset (WASO), number of awakenings after sleep onset (NASO), and sleep efficiency (SE%) for 100 participants of the general population, aged 18-75 years by judging mean differences to PSG and intervals of agreement using Bland-Altman plots.Results Mean difference to PSG for TST was 8.3 min (95% confidence intervals [CI] -7.4; 24.1) for SOMNOwatch plus (wrist), 39.8 min (95% CI 24.3; 55.3) for self-report, -79.0 min (95% CI -89.0; -68.9) for SOMNOwatch plus (hip), and -81.1 min (95% CI -91.9; -70.4) for GT3X+ (hip), respectively. The width of intervals of agreement differed with the placement of the devices. Mean differences to PSG were higher for hip-based measurements compared with wrist placement for most parameters.Conclusions Agreement of sleep parameters assessed by actigraphy with PSG differs with the placement of the device and is limited for hip-based measurements. Agreement of self-report with PSG is comparable to that of actigraphy for some parameters.