ABSTRACT
The U.S. Environmental Protection Agency uses energy expenditure metrics in some of its health risk assessments and is interested in evaluating physical activity level (PAL) data contained in its human activity database, shown previously to be an important surrogate for lifestyle characteristics. This article describes a meta-analysis of PAL measurement studies undertaken in children and adolescents ≤18 y in the United States in the 1984–2003 time period. The meta-analysis is supplemented by pooled re-analyses of individual records contained in some of the published studies. PAL and related data were obtained from 32 published studies that met objective criteria. These data were supplemented by validated Monte Carlo modeling techniques to simulate PAL from correlated total daily and resting energy expenditure measures to increase power of the meta-analysis. Statistical analyses were used to determine if age, gender, ethnic class, and temporal trend (year of the study) affected the PAL metric. Only age is consistently significantly related to PAL measures in children and adolescents. Gender, ethnic class, and year of study do not, in general, affect the PAL metric in the studies evaluated. PAL metrics seem to be an age-based physiological parameter that others have hypothesized to be largely related to body mass of the young. However, normalizing PAL by total or fat-free body mass did not provide consistent relationships with age, gender, or ethnicity.