JMIR Publications

ABSTRACT

This case study examined the feasibility of using consumer-grade wearable devices for longitudinal sleep tracking and explored how changes in sleep patterns relate to balance performance. Two college students participated over four months: Participant 1 (P1) used an Apple Watch Series 5 and an OURA ring; Participant 2 (P2) used a Fitbit Charge 5 and an OURA ring. Participants were assigned different wearable devices to assess device-specific feasibility and variability in sleep-tracking accuracy. Sleep data were collected continuously, including time in bed (TIB), total sleep time, sleep efficiency, wake after sleep onset (WASO), sleep stages, and sleep onset timing. Both participants also completed a daily sleep diary and underwent monthly balance assessments using the Bertec® Balance Advantage Sensory Organization Test (SOT). Wearables showed varying accuracy in estimating TIB: the OURA ring overestimated TIB by 15–22 minutes, the Fitbit by 27 minutes, while the Apple Watch slightly underestimated it by 9 minutes. Excellent agreement was observed in sleep duration estimates between the OURA ring and Apple Watch (ICC=0.97) and between the OURA and Fitbit (ICC=0.99), but agreement was lower for WASO, deep sleep, and sleep efficiency. Sleep variability appeared to influence balance outcomes. Fluctuations in sleep timing and duration corresponded to changes in SOT visual subscale scores, suggesting increased postural sway with irregular sleep patterns. Missing data rates were acceptable, ranging from 0–25% across devices. For P1, missingness was highest for the OURA (25%) and Fitbit (20.3%), but zero for the sleep diary. For P2, the Apple Watch had a 14.1% missing rate, the OURA 9.4%, and the sleep diary 6.25%. In conclusion, all tested wearables demonstrated feasibility for long-term sleep monitoring, though measurement discrepancies highlight the need to align device choice with research goals. Variations in sleep consistency may affect postural stability, reinforcing the importance of accurate, continuous sleep tracking in balance research. Due to the small sample size, findings are illustrative and not generalizable.