Validity and reliability of measuring steps and energy expenditure with a triaxial accelerometer in healthy adults under controlled conditions
Faculty of Medicine and Health Sciences
, p. 5-8
University of Antwerp
Purpose: To determine the validity and test-retest reliability of a triaxial accelerometer in measuring steps and estimating energy expenditure in healthy adults. Method: 20 volunteers walked 3 times for 10 minutes on a treadmill following a standardized protocol. Data of the accelerometer were compared to hand tally counted steps and to breath by breath analysis of respiratory variables. Bland-Altman plots and an intra-class correlation coefficients were used to determine validity and reliability. Results: Accelerometers generated valid results at 5km/h but not at 3km/h. At 5 km/h the lanyard and belt positions were more valid than the pocket positions. The test-retest reliability was good to excellent for steps (ICC range .78 - .92) and moderate to good for energy expenditure (ICC range .64 - .87). Conclusions: The accelerometer generated valid results for the number of steps at 5 km/h with a coefficient of variation of 1-1.6% but overestimated energy expenditure.