This section provides supplemental information about some key measurement challenges and issues that come about when using METs to standardize activity intensities in youth physical activity assessments. METs are often used to impute energy expenditure values in self-reports or/and to classify activity intensities using calibration equations developed for activity monitors. However, the use of METs assumes standard resting energy expenditure rates and therefore requires a basic understanding of changes that occur in boys and girls as they transition from childhood into adolescence and adulthood.

Children undergo systematic changes in body composition as a result of growth and maturation and these changes are particularly relevant when considering the use of MET values in youth. In general, both boys and girls (though more so in girls) experience gains in fat mass during pre-adolescent ages, but patterns change after puberty. Boys experience increases in muscle mass across adolescence and reductions in fat mass, while girls’ fat mass continues to increase throughout adolescence. The marked events in body composition are strongly related with the onset of maturity, which also varies between boys (usually at an average age of 14.0 years) and girls (usually at an average age of 12.0 years), and within sex, with some children maturing earlier or later than others. These patterns are relevant considering that muscle mass is a strong predictor of resting energy expenditure and that resting energy expenditure is often expressed as ml or kcal per kg of body weight, which reflects a composite of the partitions of body fat, muscle mass, and other tissues. The MET as it stands assumes a linear and positive relation between body weight and energy demand, which would be defensible if most changes in body weight were to occur as a result of muscle mass alone (i.e., metabolically active tissue). However, as changes in body weight occur and the fractions of muscle mass and body fat are altered as a result of growth, the MET assumption of linearity between body weight and energy expenditure is violated. The main cause of this violation is changes in body fat that can account for a significant portion of the changes in body weight (such as in girls during adolescence). These changes explain the differences in resting energy expenditure when comparing children with adults, children of different body weight, children of different ages, and children of different sexes.

The variability in body composition during development has important implications for activity intensity classifications based on METs and requires that cut-points be adjusted for differences in resting energy expenditure that occur during growth. The MET was not intended to capture these differences, but interestingly, it has become a popular energy expenditure metric when establishing cut-points or converting activity behaviors into estimates of energy expenditure or classifications of intensity. The simplicity of METs and relative ease for comparison across different subgroups of the population might well justify its use. However, researchers and practitioners are now more aware of the limitations of this metric and how it might affect activity classification and measurement of physical activity in general. By definition, METs assume resting energy expenditure as being 3.5 ml/kg/min (based on adult values and therefore often referred to as adult-METs), and multiples of this value (e.g., 3 METs) are used to distinguish light-intensity from moderate-intensity activity. The systematic change in resting energy expenditure during growth in youth implies that moderate activity might be better characterized by 4.0 METs and not 3.0 if resting energy expenditure is higher than 3.5 ml/kg/min (e.g., 4.2 to 6.0 ml/kg/min as seen in previous research). Adjusted MET values for sedentary, light, and moderate-to-vigorous intensity are directly affected by changes in resting energy expenditure and result in MET values that are higher in younger ages but decrease as youth get older, reaching adult values at about the age of 18 years. Figure 6 is an illustration of the discrepancy between adjusted and unadjusted MET values. The dashed lines indicate the traditional thresholds of 1.0, 1.5, and 4.0 adult METs (using 3.5 ml/kg/min), respectively, used to classify youth activity intensity, while the solid lines of same color indicate these same thresholds when accounting for maturation and differences in resting energy expenditure as growth occurs.

The proposed values were extrapolated based on average resting energy expenditure values published by Harrell and colleagues.³⁰ The resulting, adjusted resting energy expenditure values in METs were then multiplied by 1.5, 3.0, and 6.0 to obtain energy expenditure thresholds for light, moderate, and vigorous activity, respectively. Based on the published values from Harrell and colleagues, the resting energy expenditure in children (ages 8 to 12 years) can be on average 5.92 ml/kg/min, which is equivalent to 1.7 METs (substantially higher than the commonly used 1 MET value). When multiplied by 3.0, this value results in a threshold of 5.1 METs that should be used to classify activities of moderate intensity. Again, the 5.1 METs is substantially higher than the traditional threshold of 3.0 METs, which implies that if the latter is used, time spent in moderate physical activity is likely to be overestimated in children. Table 3 provides adjusted values for boys and girls ages 8 to 19+ years that can be used to overcome this misclassification and inappropriate assumption of the 1 MET value for resting energy expenditure when assessing activity levels in youth.

These new MET thresholds can be easily matched with available, published energy costs of physical activities in youth, and researchers and practitioners can alter their interpretation of the intensity of the activity performed or reported using the thresholds described above. The Youth Compendium of Physical Activity provides the energy cost using different metrics of the MET values (with both unadjusted and adjusted values for youth resting energy expenditure). Therefore, the above thresholds are appropriate only for energy cost when expressed in adult-METs (using 3.5 ml/kg/min). For example, according to the corrected thresholds, a reported activity for a boy age 10 years that is estimated at a value of 4.4 adult-METs would be interpreted as a light intensity activity (because it is <5.1 METs as shown in Table 3) and not as being moderate as it would be if based on the traditional and unadjusted cut-points (i.e., 3.0 METs). Alternatively, crude data obtained from activity monitors also calibrated against adult-MET values and the output activity intensity would need similar adjustments.