Abstract
Amount of daily physical activity is strongly associated with the prevalence of lifestyle-associated diseases, and thus maintaining a certain level of physical activity is recommended. Amount of physical activity, namely energy expenditure, is commonly estimated from linear regression analysis of oxygen uptake during steady-state exercise of different intensities. Short-duration exercise, mainly walking lasting less than 3 minutes, has been reported to account for over 90% of the daily energy expenditure of healthy adults. In a rodent study that measured the oxygen consumption of rodents running for different durations, the energy cost of exercise was shown to decrease as exercise duration increased, so that the oxygen consumption level asymptotically converged at a steady-state level. Wearable accelerometer-based activity counters have widely been used for physical activity assessment. The estimation of energy expenditure by these devices, however, is based on steady-state exercise of more than 5 minutes, and energy consumption of shorter duration exercise is not commonly considered. Oxygen consumption remains elevated for some period of time after exercise. Then, the estimation of energy expenditure for short duration exercise must include excess post-exercise oxygen consumption (EPOC). The amount of EPOC change in duration exercise. Therefore, we attempted to elucidate the energy expenditure for VO_2 kinetics. Healthy men (age 20-40) walked for 1, 3 and 5 minutes on a treadmill at constant speed. Oxygen consumption during and post-exercise was measured using a portable gas analyzer in breath-by-breath mode. A best-fit exponential equation to estimate oxygen consumption from exercise duration and walking speed was generated. The measured data of 3 subjects not used in generating the formula were assessed for systematic error. Cross-validation of estimated oxygen consumption revealed 3.86(ml/kg/min) RMS, and the results of the Bland-Altman analysis revealed neither fixed nor proportional bias. An equation for the estimation of energy expenditure at shorter durations of exercise was successfully generated. Accumulation of additional data may further improve the equation.
