Prediction of energy expenditure from wrist accelerometry in people with and without Down syndrome

Abstract

This study examined the relationship between energy expenditure and wrist accelerometer output during walking in persons with and without Down syndrome (DS). Energy expenditure in metabolic equivalent units (METs) and activity-count rate were respectively measured with portable spirometry and a uniaxial wrist accelerometer in 17 persons with DS (age: 24.7±6.9 years; 9 women) and 21 persons without DS (age: 26.3±5.2 years; 12 women) during six over-ground walking trials. Combined groups regression showed that the relationship between METs and activity-count rate differed between groups (p<.001). Separate models for each group included activity-count rate and squared activity-count rate as significant predictors of METs (p≤.005). Prediction of METs appeared accurate based on Bland-Altman plots and the lack of between-group difference in mean absolute prediction error (DS: 17.07%; Non-DS: 18.74%). Although persons with DS show altered METs to activity-count rate relationship during walking, prediction of their energy expenditure from wrist accelerometry appears feasible.