Evaluation of two portable sensors for energy expenditure assessment during high-intensity running

The correct assessment of individual energy requirements of athletes remains a challenge. Available literature indicates that accelerometry-based sensors underestimated energy expenditure in athletes, especially during high-intensity exercise. The goal of our study was to evaluate two commercially available devices (Actiheart; SenseWear Pro3) during high-intensity running and to identify potential sources of error. Twenty-nine male endurance- and strength-trained athletes participated in an incremental running exercise, which started at 2.8 m · s−1. Running speed was increased by 0.4 m · s−1 every 5 minutes until 4.8 m · s−1 or individual exhaustion were reached. Energy expenditure was measured with indirect calorimetry (IC) and both sensors were worn according to the manufacturers' instructions. Twelve participants also conducted a step test in order to individually recalibrate the Actiheart data. The Actiheart monitor underestimated energy expenditure (EE) for all speeds by 1.1 to 8.3 kcal · min−1 and showed a weak correlation with IC (r=0.61). Following individual recalibration there were no significant differences to IC but validity correlation did not improve (r=0.56). Heart rate was assessed correctly with the Actiheart when compared to a Polar 610 recorder. Activity counts increased only between 2.8 and 3.6 m · s−1 but plateaued thereafter. SenseWear significantly underestimated EE for all speeds by 1.0 to 9.5 kcal min−1 and was moderately correlated with IC (r=0.66). Longitudinal and transversal acceleration increased significantly between 2.8 and 4.4 m · s−1 and were correlated with running speed (0.63 <r