Prediction of Metabolic Energy Cost From Upper Body Accelerations

Abstract

Upper body accelerations were measured from human subjects as they walked on a treadmill at different speeds and inclines, while carrying different loads in a backpack. Metabolic energy cost was also measured using indirect calorimetry. The objective of the study was to determine if metabolic energy cost, under the above conditions, can be reliably estimated from upper body accelerations. Several parameters were extracted from the acceleration signals, and used in a multiple regression analysis. A statistical model, derived using the acceleration parameters as independent variables and energy cost as the dependent variable, explained 60% of the variance in the measured energy cost. If load was included with the acceleration parameters in the model, 72.3% of the variance was explained. These results indicate that a reasonable estimate of physiological workload can be obtained from upper body accelerations, if the load carried is known.