A validation study of the WHOOP strap against polysomnography to assess sleep. Issue 22 (16th November 2020)

• Record Type: Journal Article
• Authors: Miller, Dean J.; Lastella, Michele; Scanlan, Aaron T.; Bellenger, Clint; Halson, Shona L.; Roach, Gregory D.; Sargent, Charli
• Is Part Of: Journal of sports sciences. Volume 38:Issue 22(2020)
• Page Start: 2631
• Page End: 2636
• Publication Date: 2020-11-16
• Subjects: Wearable sleep monitor -- consumer sleep technology -- sleep quality -- sleep quantity -- sleep staging; Sports -- Periodicals; Sports -- Physiological aspects -- Periodicals; Sports -- Psychological aspects -- Periodicals; 612.044
• DOI: 10.1080/02640414.2020.1797448
• Languages: English
• ISSNs: 0264-0414

Abstract

ABSTRACT: The aim of the study was to compare the WHOOP strap – a wearable device that estimates sleep based on measures of movement and heart rate derived from actigraphy and photoplethysmography, respectively. Twelve healthy adults (6 females, 6 males, aged 22.9 ± 3.4 years) participated in a 10-day, laboratory-based protocol. A total of 86 sleeps were independently assessed in 30-s epochs using polysomnography and WHOOP. For WHOOP, bed times were entered by researchers and sleeps were scored by the company based on proprietary algorithms. WHOOP overestimated total sleep time by 8.2 ± 32.9 minutes compared to polysomnography, but this difference was non-significant. WHOOP was compared to polysomnography for 2-stage (i.e., wake, sleep) and 4-stage categorisation (i.e., wake, light sleep [N1 or N2], slow-wave sleep [N3], REM) of sleep periods. For 2-stage categorisation, the agreement, sensitivity to sleep, specificity for wake, and Cohen's kappa were 89%, 95%, 51%, and 0.49, respectively. For 4-stage categorisation, the agreement, sensitivity to light sleep, SWS, REM, and wake, and Cohen's kappa were 64%, 62%, 68%, 70%, 51%, and 0.47, respectively. In situations where polysomnography is impractical (e.g., field settings), WHOOP is a reasonable method for estimating sleep, particularly for 2-stage categorisation, if accurate bedtimes are manually entered.