Comparison of free-living physical activity data obtained from a Fitbit Zip, the Apple iPhone Health app and a modified Bouchard Activity Record

Issue addressed Physical activity tracking devices have potential to improve public health, but their data needs to be reliable. No study has compared movement data between the Fitbit Zip, Apple iPhone Health app and physical activity records in a community setting over 10 days. Methods University students aged 18+ years wore both a Fitbit Zip and an iPhone at/near their right waist and completed a modified Bouchard Activity Record (BAR) for 10 days in a free-living setting. Comparisons were made between the Fitbit Zip and iPhone for the number of steps and the distance travelled and between the Fitbit Zip and BAR for the minutes of activity in three different intensities. Results Eighteen students provided sufficient data for inclusion. There were strong correlations between steps per day (r = .87) and distance travelled (r = .88) between the Fitbit Zips and iPhones. However, the Fitbit Zip measured significantly more steps per day (mean 8437 vs 7303; P <= .001) and greater distances (mean 5.9 vs 4.9; P <= .001) than the iPhone. Correlations between the Fitbit Zips and the BARs were moderate for minutes of total (r = .51) and light (r = .40) activity and weak for moderate/fairly active (r = .20) and vigorous/very active (r = .25). Conclusions There were strong correlations between the physical activity data measured by Fitbit Zips and iPhones, but the iPhone Health app significantly underestimated the number of steps per day taken and the distance travelled when compared to the Fitbit Zip. So what? Understanding the comparability of accelerometer devices provides useful information for future pragmatic physical activity measurement.