PHYSICAL ACTIVITY IN THE UNITED STATES MEASURED BY ACCELEROMETER: COMMENT
It was exciting to read the landmark report by Troiano et al. (6) that, for the first time, quantified the overall level of physical activity of children and adults in the United States (US). Results from the National Health and Nutrition Examination Survey provide a much needed description of the level of active and sedentary behaviors (5) done at home, work/school, and during leisure time in the US. Such data have been sorely lacking as we try to understand the contribution of reduced activity levels to the obesity epidemic and endeavor to increase participation in a wide range of moderate-vigorous activity (3).
Although this report provides many novel insights, the exceedingly low prevalence of meeting the recommended levels of moderate-vigorous activity reported, particularly in adults, merits additional consideration. It is clear that the inability of the accelerometer to appropriately classify swimming and cycling was unlikely to account for the low prevalence estimates. However, more prevalent moderate-intensity household (e.g., vacuuming, sweeping) and lawn and garden activities (e.g., digging, raking) are likely to be important contributors to the weekly accumulation of moderate-intensity activity, and these active behaviors were unlikely to have been captured by the methods used in this report.
The activity count thresholds used to identify moderate-intensity activity were determined from purely ambulatory behaviors, but not the broader spectrum of moderate-intensity, but less-ambulatory, lifestyle activities that are recommended by public health agencies and that are assessed by current self-reported surveillance instruments. After walking, lawn and garden work is the second most commonly reported leisure-time activity (7), and high levels of moderate-intensity household activities are particularly prevalent among women (2).
Participation in these behaviors is typically associated with activity count values between 500 and 1000 counts per minute (cpm) (1). We recently used a threshold of 760 cpm to specifically assess these kinds of activities (4) and observed a prevalence of 30% for meeting the recommendations in a small sample of middle-aged adults with an overall activity level of 308 cpm·d−1 (SD 130). This cutoff has been shown to have utility in a field-based study (8). These data suggest that use of methods that more effectively capture moderate-intensity lifestyle activities-that are recommended by public health agencies and are assessed on the current surveillance instruments-would provide substantially higher estimates of the prevalence of meeting the current physical activity recommendations than those reported by Troiano et al.
That said, it should be emphasized that this difference of methodological opinion should in no way diminish the importance of the results provided by Troiano et al. (6) with respect to the description of the overall levels of physical activity in the US population. These data establish an important baseline from which to describe trends over time in activity-related behaviors. Furthermore, the prevalence estimates presented in the report are indeed sobering for their clear indication of the very low levels of frequent and sustained bouts of walking and/or running in the US early in this century.