Estimation of the metabolic rate in the occupational field: a regression model using accelerometers

Abstract

The aim of this study is to investigate the ability of the acceleration vector magnitude (VM) to predict the metabolic rate (M) for a variety of work activities. Accurate knowledge of M is of vital importance to evaluate the energy cost of work activity. While precise estimates of M can be achieved using direct calorimetry, this is not a viable option in workplaces. Ten healthy subjects (5 males, 5 females) were enrolled in six tasks simulating specific work activities. During each test, oxygen consumption rate (V˙O2), carbon dioxide production rate (V˙CO2) and accelerations at three body positions were measured. V˙O2 and V˙CO2 were used to obtain M and accelerations to calculate VM. Three mono-linear zero-intercept regression models based on center-of-mass VM were developed. A zero-intercept functional relation between M and VM was applied to the whole data set yielding the Global model. A R2 of 0.30 was obtained which indicates limited predictive ability. Two other (Vertical and Horizontal) models were also developed by separating tasks with different kinematics. With R2 = 0.56 and 0.61 they represent a significant improvement over the Global model. The results reveal a low R2 for the Global model due to collecting in the same sample many tasks that have little in common. A large R2 improvement is achieved when the dominant direction of movement is considered. The proposed method is based on a non-invasive easy-to-measure physiological parameter and can be used to track the activity level during work activity in the industrial sector.