Measurement of physical activity

This paper describes the problem of measuring physical activity in free-living populations, a premier challenge facing investigators trying to understand exercise adherence and its impact on public health. Valid methods of assessing physical activity are required that are unobtrusive, practical to administer, and specific about the type, frequency, duration, and intensity of physical activity. Existing methods can be categorized according to whether they measure a direct or indirect (e.g., self-report) observation of physical activity, energy expenditure, a physiological response during physical activity, a physiological adaptation to physical activity, or motion. On balance, self-report of physical activity offers the most practical and cost-effective method for use in population based studies. Several objective methods such as direct observation, doubly labeled water, heart rate monitors, accelerometers, and change in fitness provide standards of comparison for the validation of self-reports. No single "gold standard" exists for judging the validity of physical activity measures. Hence, the methods chosen to assess physical activity may vary according to the nature of the research question being addressed.