0666 ABILITY OF THE JAWBONE UP3 TO QUANTIFY SLEEP IN PATIENTS WITH HYPERSOMNOLENCE: A COMPARISON AGAINST POLYSOMNOGRAPHY

Abstract Introduction: Commercially available fitness activity trackers are emerging as an alternative, cost-effective measure of sleep. Unlike other such devices, the Jawbone UP3 (JB3) purports the ability to quantify light, deep, and rapid eye movement (REM) sleep. REM sleep detection is a particularly important component in the assessment of patients with suspected central disorders of hypersomnolence. Thus, this investigation evaluated the abilities of the JB3 to quantify sleep relative to polysomnography (PSG) in clinical patients with reported hypersomnolence. Methods: Seventeen patients undergoing ad libitum in-laboratory overnight PSG as part of a PSG and multiple sleep latency test (MSLT) protocol wore the JB3 on their non-dominant wrist during overnight evaluation. Bland-Altman analysis was utilized to compare total sleep time (TST), sleep efficiency (SE), sleep onset latency (SOL), wake after sleep onset (WASO), REM latency (REML), and total REM sleep time (TRT). Additionally, JB3 light sleep (LS) and deep sleep (DS) were compared to PSG N1 plus N2 and N3 sleep, respectively. Results: Compared to PSG, the JB3 significantly overestimated TST (mean difference = 20.5 minutes; p-value = .03) and SE (mean difference = 3.7%; p-value = .01). Although no other comparison reached statistical significance (p