Comprehensive comparison of Apple Watch and Fitbit monitors in a free-living setting

Abstract

Objectives The aim of this study was to evaluate the accuracy of three consumer-based activity monitors, Fitbit Charge 2, Fitbit Alta, and the Apple Watch 2, all worn on the wrist, in estimating step counts, moderate-to-vigorous minutes (MVPA), and heart rate in a free-living setting. Methods Forty-eight participants (31 females, 17 males; ages 18–59) were asked to wear the three consumer-based monitors mentioned above on the wrist, concurrently with a Yamax pedometer as the criterion for step count, an ActiGraph GT3X+ (ActiGraph) for MVPA, and a Polar H7 chest strap for heart rate. Participants wore the monitors for a 24-hour free-living condition without changing their usual active routine. MVPA was calculated in bouts of ≥10 minutes. Pearson correlation, mean absolute percent error (MAPE), and equivalence testing were used to evaluate the measurement agreement. Results The average step counts recorded for each device were as follows: 11,734 (Charge2), 11,922 (Alta), 11,550 (Apple2), and 10,906 (Yamax). The correlations in steps for the above monitors ranged from 0.84 to 0.95 and MAPE ranged from 17.1% to 35.5%. For MVPA minutes, the average were 76.3 (Charge2), 63.3 (Alta), 49.5 (Apple2), and 47.8 (ActiGraph) minutes accumulated in bouts of 10 or greater minutes. The correlation from MVPA estimation for above monitors were 0.77, 0.91, and 0.66. MAPE from MVPA estimation ranged from 44.7% to 55.4% compared to ActiGraph. For heart rate, correlation for Charge2 and Apple2 was higher for sedentary behavior and lower for MVPA. The MAPE ranged from 4% to 16%. Conclusion All three consumer monitors estimated step counts fairly accurately, and both the Charge2 and Apple2 reported reasonable heart rate estimation. However, all monitors substantially underestimated MVPA in free-living settings.

Accuracy and Precision of Energy Expenditure, Heart Rate, and Steps Measured by Combined-Sensing Fitbits Against Reference Measures: Systematic Review and Meta-analysis

“…Additional research across the age span is needed to close the gap in our understanding of how well the Fitbits measure physical activity in young individuals and older adults. Second, many different statistical strategies and related effect sizes are used to estimate the validity of these devices [ 37 ]. Researchers have used, separately or in combination, analysis of variance, correlations (eg, intraclass coefficient correlation), and measures of agreement (eg, Bradley-Blackwood test, Bland-Altman analyses, and mean absolute percentage error).…”

“…Researchers have used, separately or in combination, analysis of variance, correlations (eg, intraclass coefficient correlation), and measures of agreement (eg, Bradley-Blackwood test, Bland-Altman analyses, and mean absolute percentage error). At present, there is no specific framework for meta-analyzing statistics, such as mean absolute percentage error, although it is a preferred metric for understanding validity [ 37 ]. Thus, the meta-analysis was restricted to the mean bias and SD from the Bland-Altman analyses.…”

Effectiveness of gamified team competition in the context of mHealth intervention for medical interns: a micro-randomized trial

“…Second, heterogeneity between Apple Watch and Fitbit charge activity monitors was not accounted for during the analysis. Previous studies have shown that Fitbit Charge 2 and Apple Watch 2 had similar accuracy in terms of estimating step counts 29,30 , however, due to longer battery life, the Fitbit Charge series is more likely to be worn continuously, thus more likely to yield a higher step count and longer sleep duration in real life settings, compared with Apple Watch. Third, the results of this study may not extrapolate to a more general population because medical interns are different from the general population in terms of age, education level, and stress level.…”

Quantitative Analysis of Performance Recovery in Semi-Professional Football Players after the COVID-19 Forced Rest Period

“…Some studies [ 23 , 31 ] showed that the precision of the estimations decreases with the exercise speed, while others seemed to find good accuracies even during running [ 32 , 33 ]. As we found in the literature, the main problem is that the measurement rate could be not consistent especially at high intensities [ 23 , 31 ] but we can speculate that the missing heart rate values from the Apple Watch could be the ones evaluated physiologically implausible by the software and, so, discarded in the estimation of the mean heart rate [ 23 ]. Nevertheless, it is worth mentioning that the Apple watch resulted in being the most reliable device among several others for heart rate monitoring [ 32 , 33 ].…”