Measuring Sleep-wake State Discrepancy in the Home across Multiple Nights

Introduction
Sleep-wake state discrepancy, the mismatch between subjective and objective measurements of sleep/wake status, is estimated to affect 9-50% of individuals with insomnia, often undetected due to lack of objective measures. Consumer sleep devices allow efficient measurement of sleep-wake state across multiple nights in the habitual sleep environment. This study assessed sleep-wake state discrepancies obtained by two devices across multiple nights in individuals with insomnia.

Methods
Individuals with insomnia completed daily sleep diaries, wore a Fitbit Charge 4 device, and used a Withings Sleep Analyser under-mattress sensor for 14-days. Sleep metrics were obtained from all three data sources to calculate sleep-wake state discrepancy (device metric − diary metric).

Results
15 people participated in the study. Fitbit data were available for 12 and mattress data for 11 participants. There were no significant differences between the sleep-wake state discrepancy values derived from Fitbit and Withings devices for total sleep time (−78.7±111.8 and −95.7±124.4 minutes respectively; p .123) or wake after sleep onset (−15.4±52.9 and −17.6±110.6 minutes respectively; p .826). Significant differences were found in the sleep-wake state discrepancy values derived from Fitbit (−17.5±23.3, p <.001) and Withings devices for number of awakenings (3.2±23.1, p .175).

Discussion
Both devices comparably detected sleep-wake state discrepancy in people with insomnia, noting a device-difference for number of awakenings. These preliminary findings suggest that consumer devices may be useful to detect sleep-wake state discrepancy. Larger studies may reveal devices do so with different levels of sensitivity. Future studies should include polysomnography to assess relative magnitudes across devices.