Validity of accelerometry for the assessment of moderate intensity physical activity in the field

Purpose: This study was undertaken to examine the validity of accelerometry in assessing moderate intensity physical activity in the field and to evaluate the metabolic cost of various recreational and household activities. Methods: Twenty-five subjects completed four bouts of overground walking at a range of self-selected speeds, played two holes of golf, and performed indoor (window washing, dusting, vacuuming) and outdoor (lawn mowing, planting shrubs) household tasks. Energy expenditure was measured using a portable metabolic system, and motion was recorded using a Yamax Digiwalker pedometer (walking only), a Computer Science and Application, Inc. (CSA) accelerometer, and a Tritrac accelerometer. Correlations between accelerometer counts and energy cost were examined. In addition, individual equations to predict METs from counts were developed from the walking data and applied to the other activities to compare the relationships between counts and energy cost. Results: Observed MET levels differed from values reported in the Compendium of Physical Activities, although all activities fell in the moderate intensity range. Relationships between counts and METs were stronger for walking (CSA, r = 0.77; Tritrac, r = 0.89) than for all activities combined (CSA, r = 0.59; Tritrac, r = 0.62). Metabolic costs of golf and the household activities were underestimated by 30-60% based on the equations derived from level walking. Conclusion: The count versus METs relationship for accelerometry was found to be dependent on the type of activity performed, which may be due to the inability of accelerometers to detect increased energy cost from upper body movement, load carriage, or changes in surface or terrain. This may introduce error in attempts to use accelerometry to assess point estimates of physical activity energy expenditure in free-living situations.