Consumer-grade sleep trackers are still not up to par compared to polysomnography.

Abstract

Although polysomnography (PSG) is the gold standard for monitoring sleep, it has many limitations. We aimed to prospectively determine the validity of wearable sleep-tracking devices and smartphone applications by comparing the data to that of PSGs. Patients who underwent one night of attended PSG at a single institution from January, 2015 to July 2019 were recruited. Either a sleep application or wearable device was used simultaneously while undergoing PSG. Nine smartphone applications and three wearable devices were assessed. We analyzed the results of 495 cases of smartphone applications and 170 cases of wearables by comparing each against PSG. None of the tested applications were able to show a statistically significant correlation between sleep efficiency, durations of wake time, light sleep or deep sleep with PSG. Snore time correlated well in both of the two applications which provided such information. Deep sleep duration and WASO measured by two of the three wearable devices correlated significantly with PSG. Even after controlling for transition count and moving count, the correlation indices of the wearables did not increase, suggesting that the algorithms used by the wearables were not largely affected by tossing and turning. Most of the applications tested in this study showed poor validity, while wearable devices mildly correlated with PSG. An effective use for these devices may be as a tool to identify the change seen in an individual's sleep patterns on a day-to-day basis, instead of as a method of detecting absolute measurements.