(Not so) Smart sleep tracking through the phone: Findings from a polysomnography study testing the reliability of four sleep applications. (31st October 2019)

• Record Type: Journal Article
• Authors: Fino, Edita; Plazzi, Giuseppe; Filardi, Marco; Marzocchi, Michele; Pizza, Fabio; Vandi, Stefano; Mazzetti, Michela
• Is Part Of: Journal of sleep research. Volume 29:Number 1(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2019-10-31
• Subjects: polysomnography -- sleep applications -- sleep tracking -- smartphone; Sleep -- Periodicals; Sleep disorders -- Periodicals; 612.821
• DOI: 10.1111/jsr.12935
• Languages: English
• ISSNs: 0962-1105

Abstract

Abstract: An increasing number of sleep applications are currently available and are being widely used for in‐home sleep tracking. The present study assessed four smartphone applications (Sleep Cycle‐Accelerometer, SCa ; Sleep Cycle‐Microphone, SCm ; Sense, Se ; Smart Alarm, SA ) designed for sleep−wake detection through sound and movement sensors, by comparing their performance with polysomnography. Twenty‐one healthy participants (six males, 15 females) used the four sleep applications running on iPhone (provided by the experimenter) simultaneously with portable polysomnography recording at home, while sleeping alone for two consecutive nights. Whereas all apps showed a significant correlation with polysomnography‐time in bed, only SA offered significant correlations for sleep efficacy. Furthermore, SA seemed to be quite effective in reliable detection of total sleep time and also light sleep; however, it underestimated wake and partially overestimated deep sleep. None of the apps resulted capable of detecting and scoring rapid eye movement sleep. To sum up, SC (functioning through both accelerometer and microphone) and Se did not result sufficiently reliable in sleep−wake detection compared with polysomnography. SA, the only application offering the possibility of an epoch‐by‐epoch analysis, showed higher accuracy than the other apps in comparison with polysomnography, but it still shows some limitations, particularly regarding wake and deep sleep detection. Developing scoring algorithms specific for smartphone sleep detection and adding external sensors to record other physiological parameters may overcome the present limits of sleep tracking through smart phone apps.