Impact of EPOC adjustment on estimation of energy expenditure using activity monitors

Abstract

Purpose: To examine the accuracy of activity monitors in estimating energy expenditure (EE) during activities of varying mode and intensity and to evaluate the impact of including energy expended during recovery from activity (EPOC) on the EE estimate. EE estimates obtained from the Fitbit Surge (FBS), Garmin Vívofit (GV) and SenseWear Armband Mini (SWA) were compared to criterion EE with and without EPOC adjustments during moderate- and vigorous-intensity treadmill and cycling activities.Methods: Participants (N = 34; 23 males) completed counterbalanced treadmill or cycling conditions, comprised of a resting metabolic rate measurement, 10-min bouts of moderate- and vigorous-intensity activity and an EPOC measurement. Participants simultaneously wore the three activity monitors and a portable metabolic analyser.Results: The FBS provided lowest percent error (PE) during treadmill walking (4.4%) and the GV during moderate (6.4%) and vigorous (−0.1%) cycling bouts. EPOC-adjusted PE was higher than non-EPOC PE across all monitors and activities. Mean absolute error rate (MAPE), indicating overall measurement error, was the smallest for the FBS (14.1%) during moderate treadmill walking and the largest for the SWA (53.5%) for vigorous intensity cycling. Only the FBS had comparable non-EPOC (14.6%) and EPOC-adjusted (17.6%) MAPE during treadmill walking.Conclusion: The activity monitors tended to underestimate EE during moderate and vigorous treadmill and cycling activities. The EE estimates from the activity monitors did not account for the energy cost met by anaerobic means during activity, as suggested by the higher EPOC-adjusted EE error rates.