Dynamic body acceleration as a proxy for human energy expenditure.

RATIONALE: The use of dynamic body acceleration (DBA) has previously been used as a proxy for energy expenditure (EE) in humans with promising results. Two forms of dynamic body acceleration have been used; overall dynamic body acceleration (ODBA) which comprises of the sum of acceleration data from three orthogonal axes and vectorial dynamic body acceleration (VeDBA) which constitutes of the vector of the acceleration data from three orthogonal axes. VeDBA is the mathematically correct calculation of body acceleration however there is strong biological rationale for the use of OBDA. This study sought to ascertain which DBA metric is the most accurate predictor of EE and in addition, how accelerometer orientation and placement, body anthropometries, body composition and aerobic capacity might influence these relationships. METHODS: Twenty-one voluntary participants [seventeen males, four females; age = 22.44 +/- 3.28 years, height = 1.75 +/- 0.07 m; weight = 70.66 +/- 9.78 kg] performed an incremental maximal exercise test on a motor driven treadmill [0% grade]. Volume of oxygen utilised per minute (VO2) was measured using an online gas analyser and body acceleration (g) measured simultaneously, via three tri-axial accelerometers; two attached to the upper back (one in a straight orientation and the other skewed 30°in each axis) and one attached to the right hip (in a straight orientation). Body composition data was collected using the skinfold method. RESULTS: Both ODBA and VeDBA were good proxies for VO2 with values exceeding 0.78, although ODBA accounted for slightly but significantly more of the variation in VO2 than did VeDBA (p = 0.002). There were no significant differences between ODBA and VeDBA in terms of the change in VO2 estimated by the acceleration data in a simulated situation of the accelerometer being mounted straight but becoming skewed. In terms of placement. ODBA and VeDBA values were significantly greater at the waist than the upper back (straight orientated device only) (p = 0.000) however when plotted against VO2 the differences between the hip and upper back became insignificant for both metrics. Fat-free mass, fat mass and age added significantly to the VO2 versus ODBA and VO2 versus VeDBA relationship in terms of r2. CONCLUSIONS: ODBA was found to be a marginally better proxy for VO2 than VeDBA although should only be used where researchers can guarantee a reasonably consistent device orientation. The upper back and hip are equally appropriate placements and should be chosen depending on the practicality. The ability of DBA to predict VO2 can be improved by adding additional variables to the regression equation, hi this case fat-free mass was the most significant covariate in terms of the improvement in r2