Validity and reliability of Fitbit activity monitors compared to ActiGraph GT3X+ with female adults in a free-living environment

Abstract

Objectives: Inexpensive activity monitors have recently gained popularity with the general public. Researchers have evaluated these consumer-based monitors in laboratory-conditions. Given the current wide-spread consumer use of these devices, it is important to ensure users are attaining accurate information compared to previously validated measures. This study investigates the accuracy of Fitbit One and Flex activity monitors in measuring steps, sedentary time, and time spent in light, moderate, and vigorous intensity activities with ActiGraph GT3X+ with female adults in free-living conditions.

Design: Cross-sectional study.

Methods: Twenty-two women, 21.23±1.63 years, BMI: 22.35±2.34kg/m² wore two Fitbit Ones (bra and waist), one Fitbit Flex on the wrist, and one ActiGraph GT3X+ on the waist for seven-consecutive days. Repeated measures ANOVA was used to explore differences in steps, sedentary time, and time spent in light, moderate and vigorous intensity activities among the four devices.

Results: No differences were found in number of steps recorded across the four devices. Fitbit One, waist and bra, overestimated time spent in light intensity activities. Fitbit One (waist) and Fitbit Flex overestimated time spent in moderate intensity activities. Fitbit One, waist and bra, and Fitbit Flex overestimated time spent in vigorous intensity activities. All Fitbit activity monitors overestimated MVPA and underestimated sedentary time compared to the ActiGraph.

Conclusions: Regardless of wear-location all Fitbit devices provide similar activity monitoring and users can wear the devices wherever best accommodates their lifestyle or needs. Users should not rely solely on these monitors when tracking vigorous and MVPA activities.

Keywords: Fitbit; Physical activity; Reliability; Sedentary behavior; Validation.