Validation of Exercise and Postexercise Energy Expenditure Estimates Using Hip and Wrist Accelerometer Algorithms

Abstract

Listen

Translate article

To estimate physical activity intensity, accelerometer-based devices are often calibrated to energy expenditure (EE) measures using indirect calorimetry (IC). Most EE estimation algorithms are based on steady-state data and do not consider excess postexercise oxygen consumption (EPOC). Purpose: The purpose of this study is to determine the effect of incorporating EPOC into linear and nonlinear accelerometer methods for estimating EE after high-intensity running. Methods: Nine adult males participated in three visits using IC to measure EE while wearing wrist and hip ActiGraph accelerometers. Each visit is described as follows: (a) Visit 1: a treadmill VO2peak test determined treadmill speed for subsequent visits; (b) Visit 2: 20-min seated baseline and three vigorous treadmill running bouts (30, 60, and 120 s) with 20-min seated rest between; and (c) Visit 3: 60-min supine baseline and a 30-min treadmill run followed by 3-hr supine rest. Fifteen EE estimation algorithms were compared with IC-measured EE. The bias (ActiGraph accelerometers − IC) and 95% confidence interval of the bias were used to determine significance. Results: Linear regression EE algorithms tended to overestimate EE after each exercise bout (mean bias kilocalories [95% confidence intervals]; 30 s: 12.5 [3.3, 21.6], 60 s: 9.6 [0.4, 18.9], 120 s: 6.5 [−2.7, 15.7], and 30 min: 177.5 [84.6, 262.1]). The nonlinear algorithms underestimated EE after the short bouts and, due to the wide confidence interval, demonstrated no bias after the 30-min bout (30 s: −7.9 [−10.2, −5.6], 60 s: −10.9 [−14.0, −7.9], 120 s: −15.4 [−15.9, −14.8], 30 min: 14.8[−39.5, 69.2]). Adding EPOC did not improve EE estimates. Conclusions: Generally, the addition of measured EPOC did not improve device-based EE estimates. Nonlinear methods demonstrated less bias in estimating postexercise EE than linear methods.