The evidence for constrained total energy expenditure in humans and other animals

Energy is a limited resource for living organisms, with trade-offs in expenditure evident among physiological tasks. The impact of physical activity on these trade-offs is currently debated. Additive models, which predominate in human nutrition and animal ecology, suggest physical activity does not affect other expenditure. Recently proposed constrained models propose that increases in physical activity lead to decreases in other expenditure, maintaining total energy expenditure within a narrow range. Here, we develop quantitative frameworks for both models and compare their predictions against data from experimental studies that manipulate physical activity and ecological studies that measure physical activity and expenditure in free-living populations. In human aerobic exercise interventions, total daily energy expenditure increased by only ∼30% of the change expected from additive models. Compensation appeared to be reduced with resistance training and amplified when aerobic exercise is paired with diet restriction. In animal experiments, which often involve some form of dietary restriction, compensation is generally ∼100%. Results from experimental studies are consistent with those of ecological studies, which indicate the degree of compensation to physical activity may be greater in the presence of limited food availability. Reductions in basal metabolic rate and sleeping metabolic rate contribute to energy compensation, particularly in animal studies and longer-duration human studies, but do not fully account for the observed compensation in total daily energy expenditure.