Measuring the validity of self-monitoring heart rate and activity tracking wearables

PURPOSE: To examine the validity of wearable physical activity tracking devices. METHODS: Participants were 62 students. Participants wore a Polar HR chest strap, Actigraph GT3X+ Acceleremetor, Apple Watch, Fitbit Charge HR, Garmin Forerunner 225 and were connected to a metabolic cart. Participants completed a seated 10-min baseline period, 4-min stages of light, moderate and vigorous intensities, and a 10-min seated recovery. Heart rate (HR), energy expenditure (EE) and step count were examined for each bout of exercise. ANALYSIS: Two-way RM-ANOVA were performed to compare the ability of the wearable devices to accurately measure each outcome relative to the criterion. Paired-samples t-tests compared the number of steps in observed videos and those reported for Fitbit. RESULTS: For HR, Apple Watch was accurate at all stages except in light and moderate intensities, in which the device measured lower HR. The Fitbit Charge HR produced accurate results in moderate PA, but measured significantly higher HR readings at baseline and light activity and lower HR readings at vigorous intensity. The Garmin Forerunner 225 was only accurate at vigorous intensity PA and measured significantly higher HR readings at all other intensities. For EE, the Fitbit measured significantly higher EE for all stages except vigorous intensity and recovery stages. The Apple Watch and Garmin measured significantly higher EE at all PA intensities. The Fitbit measured significantly lower step count than the criterion at all PA intensities. DISCUSSION: This study provides novel findings for Apple Watch and Garmin devices and provides new information regarding Fitbit accuracy. No studies have reported accuracy of these devices to measure HR. Future studies should investigate why differences between the devices exist.