Energy compensation and why exercise calories may not add up

Exercise calories may not translate directly into extra daily expenditure. Conflicting studies show why a fixed compensation rate cannot predict your results.

日本語で読む（Japanese version）

Can you add the calories displayed after a workout straight to your daily energy expenditure? In a short experiment involving 12 adults, the increase in activity expenditure was close to the increase in total expenditure.[7] Yet a paper comparing human aerobic exercise interventions reported that daily total expenditure increased by only about 30% of the increase predicted by simple addition.[1] The answer changes with what researchers measure and how long they follow it.

That gap concerns more than what people eat after exercise. It also raises a question about the energy the body uses outside the workout. When total daily expenditure rises by less than expected from the exercise performed, researchers call the shortfall exercise-related energy compensation. Some interventions have documented it. They have not all found that a fall in resting metabolism explains it.[2]

Other studies find a more straightforward relationship: greater activity accompanies greater total expenditure, with no evidence of the proposed compensation.[3][7] The companion article on exercise and appetite considers eating after exercise. Here, the focus is the expenditure side of the budget. Before deciding that your effort was wasted, it helps to separate the energy used during movement, the change across the whole day, and the eventual change in weight.

A workout and a whole day measure different things

Total energy expenditure is usually considered the sum of energy used at rest, energy used for activity, and diet-induced thermogenesis. That last term means the energy expenditure associated with eating. Exercise belongs within the activity component; it is not the entire daily budget. The scientific question is whether the other components remain unchanged when activity expenditure increases.[7][8]

The traditional additive model assumes that activity expenditure and resting expenditure are independent. Extra expenditure from activity therefore adds to the daily total. The constrained model makes a different prediction: as activity increases, expenditure elsewhere decreases, keeping the total within a relatively narrow range. These are competing descriptions of how the body allocates energy. They can produce different predictions from the same record of exercise.[1][8]

Imagine a morning workout followed by work, travel, household tasks and rest. The workout is part of that day, but the day does not consist entirely of the workout. To test the models, researchers need to examine the whole budget. If you put “energy used in the workout” and “increase in daily expenditure” in the same column, you have already assumed the additive answer before checking whether the rest of the budget changed.[1][7]

This distinction also clarifies what a device display cannot settle. A number describing the exercise period does not, by itself, describe how expenditure changed during the remaining day. The missing information is the relationship between that activity and the rest of the total. The studies discussed here examine that relationship; a workout record alone does not provide the same comparison.[1][7]

The constrained model does not mean that total expenditure can never rise. An early study of adults found a positive association between activity and total expenditure at lower activity levels, with a plateau at higher levels. Its picture was of a changing response across the activity range, rather than no response at any level. “Less than simple addition predicts” and “no increase at all” are different findings.[9]

Weight loss adds another layer. Asking how much total expenditure rose is different from asking how much weight someone lost. The latter also involves food intake and changes in body composition. A systematic review defined compensation using changes in energy stored in body fat and fat-free mass relative to exercise expenditure. That is not the same measure as the shortfall in total daily expenditure. The word “compensation” can refer to different calculations.[19]

For a reader trying to make sense of a stalled weight-loss plan, the first useful step is therefore to identify the outcome. Was the study measuring total expenditure, resting expenditure, or changes in body energy stores? Was it comparing active and less active people, or following the same people after an exercise intervention? These questions determine what the result can tell you. They also prevent a number about one part of the budget from becoming an answer about every part.[7][19]

You can keep these distinctions in an ordinary exercise diary without trying to reproduce a laboratory. Write what activity you actually did, and keep weight change as a separate observation. A completed session remains a completed session even when the scale does not match the expected calculation. The scale cannot divide the day into resting and activity expenditure, and it cannot identify compensation on its own.[2][7]

What the studies supporting constraint found

An observational study of 332 adults from different living populations provided evidence for the constrained model. Researchers measured total expenditure using doubly labeled water and activity using accelerometers. Doubly labeled water is a research method for measuring total energy expenditure. After accounting for body size and composition, the relationship between activity and total expenditure was stronger at lower activity levels, while total expenditure plateaued at higher levels.[9]

The study’s strength was its view of ordinary life. It compared activity with the daily total rather than looking only at exercise sessions. Its observational design also sets a limit: a comparison between different people does not establish that every individual will reach the same plateau after increasing exercise. Assigning an intervention and following its effects is a different kind of evidence.[9]

A larger analysis examined total and basal energy expenditure in 1,754 adults living normal lives. It estimated average compensation of 28%, attributed to lower basal expenditure. The corresponding interpretation was that 72% of the extra energy used for additional activity translated into additional expenditure in the daily total. This was an average relationship in a large dataset, rather than a personal rule for correcting every workout.[5]

That analysis also found an association between body fatness and the strength of compensation. The direction of cause remained unresolved. People who compensate more might be more likely to accumulate fat, or accumulating fat might make compensation stronger. The paper identified establishing causality as a task for further work. Its finding does not justify telling a person with more body fat that their body is destined to cancel out exercise.[5]

The uncertainty matters when discussing results with a friend or trainer. An association describes a pattern in the data; it does not show which factor started the pattern. Turning it into a statement about someone’s fixed physiology goes beyond the finding. Nor does a group average show how much of a particular person’s activity expenditure contributed to their daily total on a particular day.[5]

A 2026 paper compared experimental studies with studies of activity and expenditure in free-living populations. In human aerobic exercise interventions, the rise in total daily expenditure was about 30% of the increase expected under additive models. The comparison suggested that compensation may be smaller with resistance training and greater when aerobic exercise is combined with dietary restriction. Exercise type and food conditions therefore belong in the interpretation.[1]

The 28% and approximately 30% figures look deceptively similar. Their meanings point in different directions. The first describes the proportion estimated to be compensated; the second describes the proportion of the predicted increase that actually appeared in total expenditure. If you write both as “the percentage of exercise calories that counts,” you reverse the meaning of one of them. Before comparing percentages, check what the percentage is of.[1][5]

These findings support taking compensation seriously, but they do not establish a universal subtraction rule. The observational datasets describe relationships across people. The comparison of interventions describes results across exercise conditions. Neither supplies a single coefficient that can turn your displayed exercise calories into an exact meal allowance. That practical limitation follows from the differences in population, design and calculation, rather than from dismissing the evidence.[1][5][9]

Resting metabolism falls in some interventions, but not all

If expenditure outside exercise decreases, where does the saving occur? A study of 16 sedentary adults with overweight examined a 12-week supervised aerobic walking intervention. Its target exercise expenditure was 20 kcal per kilogram of body weight per week. Researchers assessed total expenditure, resting and sleeping metabolism, diet-induced thermogenesis, physical activity and food intake. This allowed them to look beyond the final weight change.[4]

Total expenditure increased, but resting and sleeping metabolic rates fell. Those reductions accounted for most of the compensatory response. Walking became more efficient, and the increase in daily moderate-to-vigorous activity was smaller than expected. Reported intake and diet-induced thermogenesis did not change. Weight loss was minimal despite improved body composition: limited movement on the scale did not mean that every measured outcome stayed the same.[4]

The methods were unusually detailed. Total expenditure was measured with doubly labeled water. Whole-room calorimetry, which measures metabolism in a room designed for that purpose, assessed resting and sleeping expenditure and diet-induced thermogenesis. Magnetic resonance imaging measured the volumes of metabolically active organs. The researchers also monitored activity and assessed walking efficiency.[4]

Liver and kidney volumes decreased by 5%, while brain volume did not change. A parallel mouse experiment found increased cellular density and mitochondrial content in the liver. These findings add detail to the proposed mechanisms, but the distinctions between them must remain visible. Findings in mouse tissue do not establish the mechanism of compensation in humans. The human organ results were observations from an intervention involving 16 adults.[4]

The study offers a concrete example of why “exercise burns energy” and “the whole-day increase equals the exercise cost” are separate statements. Total expenditure rose while other components fell. Both can be true at the same time. The walking sessions were not erased; the rest of the budget changed alongside them. The small weight change also coexisted with improved body composition.[4]

A different analysis used a randomized controlled exercise intervention lasting 24 weeks. It identified exercise-related compensation in 48% of participants, with a reported magnitude of −308 ± 158 kcal per day. However, participants with and without compensation did not differ significantly in metabolic adaptation measured over the full day, during sleep, or at rest.[2]

There were also no statistically significant differences between these groups in sex, age or body mass index. Compensation was negatively associated with baseline total expenditure, with a correlation coefficient of −0.50. Total expenditure was assessed using doubly labeled water over 14 days, while room calorimetry covered 24 hours. An association with a measured baseline characteristic is different from being able to predict compensation from someone’s age or body size.[2]

The negative calorie value needs careful reading. It describes a shortfall relative to the increase expected from the prescribed exercise. It does not mean that daily total expenditure necessarily fell below its pre-exercise value by that amount. The reference point is the expected addition. Leaving out that reference point changes the result into a different claim.[2]

Together, these interventions show why an observed outcome and its mechanism should be discussed separately. One found reductions in resting and sleeping metabolism that explained much of the compensation. The other found compensation without significant group differences in the measured metabolic adaptations. A smaller-than-expected result on the scale is not enough to diagnose a fall in basal metabolism.[2][4]

Walking efficiency adds another possible part of the picture. In the walking intervention, improved efficiency meant that the way participants used energy for the same movement had changed. That finding does not provide a percentage adjustment for your commute or shopping trip. Evidence that efficiency can change is different from a formula for calculating how much less energy your own walk now uses.[4]

This is where everyday records can tempt us into filling gaps with a preferred story. The workout display is visible; much of the body’s remaining expenditure is not. You might write “metabolism slowed” because weight loss was disappointing, or “nothing changed” because the workout felt familiar. Yet even studies measuring these components do not all give the same answer. Leaving an item marked “not measured” is more accurate than converting an interpretation into a fact.[2][4]

Numbers: energy compensation

1,754 adults: an analysis estimated average compensation of 28%; this is not a fixed individual rate.[5]

332 adults: an observational study found a plateau in total expenditure at higher activity levels.[9]

24 weeks: 48% of participants showed compensation, without significant group differences in resting or sleeping metabolic adaptation.[2]

16 adults over 12 weeks: total expenditure rose while resting and sleeping metabolic rates fell.[4]

12 adults in a short experiment: activity expenditure rose by about 250 kcal per day and total expenditure by 272 kcal per day.[7]