1005 Comparison of Commercially-Available Sleep Tracking Devices With Sleep Diary and Actigraphy. (12th April 2019)

• Record Type: Journal Article
• Authors: Rieck, Thomas M; Gaz, Daniel V; Peterson, Nolan W; Jenkins, Sarah M; Rosedahl, Jordan K; Krzoska, Ashley A; Mansukhani, Meghna P; Clark, Matthew M; Kolla, Bhanu P; Morgenthaler, Timothy I; Hagen, Philip T
• Is Part Of: Sleep. Volume 42(2019)Supplement 1
• Page Start: A404
• Page End: A405
• Publication Date: 2019-04-12
• Subjects: Sleep -- Physiological aspects -- Periodicals; Sleep disorders -- Periodicals; Sommeil -- Aspect physiologique -- Périodiques; Sommeil, Troubles du -- Périodiques; Sleep disorders; Sleep -- Physiological aspects; Sleep -- physiological aspects; Sleep Wake Disorders; Psychophysiology; Electronic journals; Periodicals; 616.8498
• DOI: 10.1093/sleep/zsz067.1002
• Languages: English
• ISSNs: 0161-8105

Abstract

Abstract: Introduction: Commercially-available devices which measure activity and sleep have increased exponentially in popularity. Users of such devices are turning to clinicians and fitness professionals for guidance on how to best utilize these devices for sleep, but little evidence exists to guide these discussions. This study compares three commercially-available sleep tracking devices, the Fitbit Charge HR ® (Fitbit ® ), Beddit 2 ® (Beddit ® ), and ResMed S+ ® (ResMed ® ), for accuracy at estimating total sleep time (TST) over seven consecutive nights against the clinically-validated Philips Actiwatch 64 ® (Actiwatch ® ). Methods: Forty-eight participants, aged 19-64 years (mean=35.1, SD=10.2, 72.9% female) used three tracking devices over the seven day trial. Each device has different placement requirements - the wrist (Fitbit ® ), under the bedsheet (Beddit ® ), and on the nightstand (ResMed ® ). All participants were determined to be good sleepers, as defined by the Pittsburgh Sleep Quality Index (PSQI). Participants with a Global PSQI score greater than four were excluded. Throughout the study, participants completed the American Academy of Sleep Medicine sleep diary. Each device's estimated TST was compared with that from the participant's recall sleep diary and the Actiwatch ® . Results: There were significant differences between the three devices and the sleep diary, with the devices underreporting sleep by 25 to 49 minutes on average (p≤0.05). The Fitbit ® and the Beddit ® estimated mean TST did not differ significantly from the Actiwatch ® (p=0.12 and 0.59 respectively). The ResMed ® estimated mean TST was slightly lower than the Actiwatch ® (by 12.8 minutes, p=0.045). Users rated the disruptiveness of the devices at an average of 1.2 on a scale of 0 to 10 with a score of "0" being indicative of not being disruptive, while "10" was extremely disruptive. Conclusion: Commercially-available sleep tracking devices can reasonably estimate TST in adults with "good" sleep, although they tend to slightly underestimate TST compared to self-report. The devices tested do not seem to disrupt sleep. Clinicians and fitness professionals could review data from these devices, in conjunction with clinical data, to help ascertain whether healthy subjects are obtaining sufficient sleep. Support (If Any): Mayo Clinic Healthy Living Program