Metabolism and function · 20 October 2026

Fuelling changes expenditure: why energy needs are not fixed

Overfeeding experiments show an adaptable expenditure response—and the limits of that adaptation.

Billy Craig, MSc · Independent Researcher

This article discusses published research and an unvalidated EPK hypothesis. It reports no new experimental results.

An estimate is a starting point

A calorie estimate can look more precise than the physiology it represents. A number calculated from age, height and mass is useful for getting started, but it is not a permanent personal allowance. Food intake, body composition and the conditions under which expenditure is measured all matter. The question for EPK is how to recognise a meaningful change in needs without treating every fluctuation as a new requirement.

EPK means energy per kilogram: an approach to considering food energy in relation to body mass and function. Its practical ambition is adequate, consistent fuelling. Testing that approach requires separating a useful organising idea from claims that a particular ratio or surplus has already been validated.

What the carbohydrate experiment measured

Welle and Campbell studied seven men through six baseline days and twenty days of carbohydrate overfeeding. Mean intake rose from 9.12 to 16.69 MJ/day, with absolute fat and protein held constant. Basal expenditure was measured by indirect calorimetry. BMR increased by 11.5%, while body weight rose by 4.8%. Acute propranolol did not abolish the basal thermogenic response. [1]

The experiment changed carbohydrate and total energy together. Its fixed order, small male sample and short duration prevent a clean claim that carbohydrate alone caused the change. Increased body mass also complicates the comparison. It establishes neither a universal response nor an optimal surplus. [1]

Adaptation does not make excess energy disappear

A later comparison by Horton and colleagues gave sixteen men equal excess energy as fat or carbohydrate for fourteen days each. Whole-room calorimetry measured expenditure and oxidation. Estimated retention was 75–85% of the carbohydrate excess and 90–95% of the fat excess. More expenditure with carbohydrate did not mean no energy storage. [2]

This distinction matters when an experiment becomes a slogan. “Expenditure responds to intake” and “surplus energy cannot accumulate” are different propositions. The first has experimental support here; the second does not. These were large, short-term surpluses, not a test of maintaining performance on an individually adjusted diet.

Longer overfeeding in twelve pairs of identical twins also produced substantial but variable weight gain. Similarities within pairs pointed to differences in susceptibility between people. A response observed in a group is therefore a poor basis for promising an identical outcome to every user. [3]

Resting expenditure is only one part of the day

Basal expenditure, the cost of processing food and expenditure during activity are different quantities. An increase in one component cannot simply be applied to the whole day. Nor does an increase in body weight tell us exactly how much tissue energy has accumulated: a body-composition measure answers a different question from a scale reading.

For interpretation, write down what was measured. Was it a resting calorimetry session, a day in a chamber, or a wearable estimate? Were intake and movement controlled? Were body composition and measurement conditions comparable? The Welle and Horton experiments are useful precisely because their measurement settings are described, rather than hidden inside a single “calories burned” number. [1,2]

How this informs EPK

The metabolism discussion in Consistent Eating places emphasis on the work the body continues to do, rather than treating food as something earned only by visible exercise. These experiments offer a research connection to that theme: expenditure deserves to be reassessed, not assumed fixed.

For the app, a reasonable research proposal is to treat an initial estimate and the first days of wearable data as provisional. Subsequent evaluation should consider intake, changes in mass, activity and functional outcomes together. This is a proposed validation strategy; the studies above did not test the app, a five-day learning period or its performance pathway.

A trial could compare a fixed starting estimate with an updating approach, while recording strength or performance, hunger, adherence and body composition. The analysis should ask whether updating improves outcomes, not merely whether its calorie figures change. A performance surplus around 15% remains a choice to test, not a threshold established by these overfeeding studies.

The useful conclusion

Energy needs can change, and the response to additional food is not identical across diets or individuals. That supports a flexible view of fuelling. It also places a boundary around the claim: an adaptable body still stores energy, and adequate fuelling should be judged by what it enables over time, rather than by a promise that eating more automatically raises expenditure enough to compensate.

Research and guidance

Explore more studies in the research library.

Related working papers

Read the relevant PDFs: EPK–001 to 007 are the revised version 1.2 papers; EPK–008 to 010 are version 1.0 supplements. Earlier PDFs and DOI records remain on the paper archive.