Consumer wearable technologies are widely used for tracking physical activity, sleep, and various health metrics. However, the accuracy of these devices varies significantly depending on the measurement, device brand, and usage context. A recent umbrella review compiling 24 systematic reviews and over 430,000 participants revealed that only about 11% of wearable devices on the market have been validated for even a single biometric measure (Doherty et al., 2024). Even among validated devices, systematic errors and inconsistencies were found in key health metrics.

In laboratory settings, step counts were found to be the most reliable metric, with brands like Fitbit and Apple Watch showing good accuracy. Heart rate readings were moderately accurate, especially from Apple and Garmin, but energy expenditure estimates were consistently inaccurate across all brands (Fuller et al., 2020). For example, the average error in measuring physical activity intensity ranged from 29% to 80%, highlighting major discrepancies in detecting how vigorously a person is moving.

The Oura Ring, a newer wearable focused on sleep and daily readiness, was evaluated in both laboratory and free-living conditions. In a controlled setting, its energy expenditure estimates correlated strongly with gold-standard methods (r = 0.93), but it tended to underestimate activity at higher intensities.

In real-life conditions, step counts and energy use estimates from Oura also correlated well with accelerometer data, yet it systematically over- or underestimated various metrics depending on the intensity and body placement of reference devices (Kristiansson et al., 2023).

For sleep tracking, the Oura Ring showed high sensitivity (94.5%) and overall accuracy (~92%), closely matching clinical polysomnography results (Svensson et al., 2024).