Objective: To explore the best method of adjusting energy expended on physical activity (AEE) for differences in body size. Many publications have expressed AEE per kg body weight (i.e. using weight 1.0 as denominator). This makes the unjustified assumption that all activities are weight-dependent.
Design: Retrospective analysis of data from ninety-two 24-h whole-body calorimetry measurements in women, and 574 doubly-labelled water measurements in men and women to calculate the optimal exponents of body weight for adjusting AEE.
Results: The analysis proved that weight 1.0 over-corrects for size differences and yields invalid conclusions about relationships between physical activity and obesity. An exponent close to 0.5 is more appropriate for sedentary lifestyles. However the correct exponent is itself dependent on the relative mix of weight-dependent and non-weight-dependent activities undertaken.
Conclusion: We conclude that it is impossible to recommend a generalizable coefficient for adjusting AEE, and that great caution must be exercised when interpreting AEE data from individuals of markedly different body sizes.