|| Checking for direct PDF access through Ovid
Many wearable fitness trackers have a function measuring physical activity such as energy expenditure but the accuracy of the measures is unclear.PURPOSE: The aim of the study was to validate the energy expenditure estimation of three fitness trackers: LT, FB, and AW for treadmill walking and running.METHODS: The participants of the study were 30 college students (17 males and 13 females) from a public university in Pennsylvania. All participants completed six trials of 10-minute walking and running activities on a treadmill at speeds of 54, 80, 107, 134, 161, & 188 m•min-. Participants were wearing all the fitness trackers and connected to indirect calorimetry during the exercise protocol. All devices provided caloric expenditure while the indirect calorimetry was used as a criterion measure. Resting Metabolic Rate was collected along with a familiarization trial prior to the execution of the exercise protocol. Two-way ANOVA with repeated measures and Pearson Correlation were used to compare the caloric expenditure estimates between the criterion and three fitness trackers.RESULTS: There were no significant differences in energy expenditure estimates between the criterion and AW (mean difference 4.6 kcal, p > 0.05), nor criterion and LT (mean difference 0.8 kcal, p > 0.05). However there was a significant difference between the criterion measure and FB (mean difference 18.2 kcal, p = 0.001). Pearson correlation coefficients (r): the criterion of indirect calorimetry-derived activity energy expenditure yielded the strongest positive correlations with activity energy expenditure estimated from the AW at all speed levels (range from .55 to .85, all ps <0.01). FB was positively correlated to the criterion measure at five speed levels (range from .46 to .67, ps <0.05). No positive correlation was found between the criterion and LT measures (range from -.40 to .26).CONCLUSION: AW and FB demonstrated a moderate to high level of validity on measuring physical activity while LT had a low level of validity.