0299 An Examination of Discrepancies in Sleep Device Data Captured by Sleep Ring, Fitbit, AndActigraphy

Abstract Introduction Wearable devices have improved access to sleep data by empowering patients with greater control over their information. However, the choice of the best wearable for sleep measurement remains uncertain. This study investigated variations in sleep data collected by Sleep Ring, Fitbit, and Actigraphy in a diverse US group of Black adults. Methods Wearables descriptive statistics were calculated for sleep efficiency (% by time in bed/ time sleep), REM duration (in minutes), awakening (count), and total sleep duration (in minutes) where data was available. One-sample t-tests were conducted for sleep efficiency (Sleep Ring and Actigraphy), REM duration (Sleep Ring and Fitbit), and awakening (Fitbit and Actigraphy). ANOVA was conducted for total sleep duration (Sleep Ring, Fitbit, Actigraphy). Significance was specified as α=0.05. Results Participants (Sleep Ring N=199, Fitbit N=147, Actigraphy N=202) were predominantly female (71%, Mage=48.9) compared to male (28%, Mage=50.2). Using the Insomnia Severity Index, 35% were insignificant, 37% met subthreshold, 23% exhibited moderate severity, and 5% reported severe insomnia. Sleep efficiency for Sleep Ring (M=77%, SD=10.4; t(198)=104.6, p.001) was significantly different than Actigraphy (M=93%, SD=3.3; t(201)= 397.7, p=.001). REM duration for Sleep Ring (M=62 minutes, SD=29; t(198)=30.0, p.001) was also significantly different Fitbit (M=70 minutes, SD=26; t(142)= 31.6, p.001). Additionally, there was a significant difference in awakening count in Fitbit (M=15 times, SD 7; t(146)=26.6, p.001) compared to Actigraphy (M=9 times, SD=4; t(201)= 28.7, p.001). ANOVA indicated an effect of sex on sleep duration for all devices: Sleep Ring M=368 minutes, SD=106; F(1,197)= 5.4,p=.021, Fitbit M=305, SD=89; F(1,145)=13.4,p.001 and Actigraphy [M=374 minutes, SD=163; F(1,198)=4.9,p=.027). Conclusion Differences among wearable devices exist. Actigraphy may overestimate sleep efficiency, while Sleep Ring may provide poor estimates. REM duration is similar in Sleep Ring and Fitbit but more variable in Sleep Ring. Sleep duration ranked highest in Actigraphy, but also exhibited the highest variance/deviation, followed by Sleep Ring and Fitbit. Fitbit is more sensitive in detecting awakenings compared to Actigraphy. Future research should exploresleep health parameters across wearable devices to ensure accuracy in Blacks, considering discrepancies in metrics potentially due to differences in skin tone and the impact of refractory lights Support (if any) NIH R25HL10-5444; R01HL142066 and R01AG067523.