Anna Shcherbina · Journal of Personalized Medicine 2017 · cross-sectional validation study · n=60

Accuracy in Wrist-Worn, Sensor-Based Measurements of Heart Rate and Energy Expenditure in a Diverse Cohort

Cited 651 times in the scientific literature.

Level 3 - non-randomized controlled study

Non-randomized observational validation study comparing multiple consumer devices against reference standards

OpenAlex W2618581983 · doi:10.3390/jpm7020003 · record verified 2026-08-29

What was done

Sixty volunteers (29 male, 31 female; age 38 ± 11 years) with diverse age, height, weight, skin tone, and fitness levels were evaluated using seven commercial wrist-worn devices: Apple Watch, Basis Peak, Fitbit Surge, Microsoft Band, Mio Alpha 2, PulseOn, and Samsung Gear S2. Participants completed sitting, walking, running, and cycling protocols while device measurements of heart rate (HR) and energy expenditure (EE) were compared concurrently against continuous telemetry and indirect calorimetry.

What was found

Device error varied by activity and user characteristics: error was lowest for cycling and highest for walking, and increased among males, individuals with higher body mass index, and those with darker skin tone. Six of the seven devices achieved a median HR error under 5% during cycling. Conversely, no device achieved an error below 20% for energy expenditure. Across both metrics, the Apple Watch had the lowest overall error and the Samsung Gear S2 had the highest.

Why it matters

While consumer wearables reliably track heart rate during common laboratory activities, their energy expenditure estimates are consistently inaccurate and should not be relied upon for clinical or health improvement interventions.

Limits

The sample size was modest (n = 60) and testing was restricted to controlled, laboratory-based activities rather than free-living conditions. Specific numerical error percentages were omitted in the abstract for most individual devices, walking, and running.

Cited by