Energy balance

Total daily energy expenditure breaks down into four parts, and they differ enormously both in size and in how much they vary from person to person.

In a study that overfed volunteers by 1,000 kcal per day for eight weeks, the amount of fat they gained varied roughly ten-fold from one person to the next. That spread was not explained by their resting metabolic rate or by exercise. It tracked changes in non-exercise activity: some people unconsciously moved a great deal more when overfed, and some barely changed.

The same mechanism runs in reverse during a deficit, and it is largely invisible to the person it is happening to. It is a substantial part of why two people following an identical plan get different results.

BMR cannot be measured without lab equipment, so instead it is estimated using formulas built from measurements of large groups of people. The Mifflin-St Jeor equation is the most widely used, and it is what the on-page calculator uses.

An estimate is only an estimate. These formulas can be meaningfully off for any one person, so the honest way to use one is as a starting point that you adjust against what actually happens over several weeks — which is how the calculator here presents its result.

When weight is lost, energy expenditure falls — partly because a smaller body costs less to run, and partly by an extra amount beyond what the size change alone would predict. That extra drop is called adaptive thermogenesis, and it can be measured in controlled studies.

Appetite shifts in the same direction, toward eating more. Together, these adaptations are the physiological reason regaining weight is common — a far better explanation than a failure of willpower.

It is worth being precise about what “a calorie is a calorie” gets right and what it misses. Under tightly controlled feeding, with total energy held equal, the differences in body fat lost between very different diets — low-fat versus low-carbohydrate — turn out to be small. One metabolic-ward study even found a slight edge for cutting fat rather than carbohydrate, the opposite of the popular expectation, and a separate result reporting higher energy expenditure on low-carbohydrate diets has been openly disputed. At the level of pure energy accounting, the physics largely holds.

That is a statement about a metabolic ward, not a kitchen. In everyday life calories are not created equal, because what a food is made of changes how much of it you eat. In a controlled study where participants lived on-site and the diets were matched for calories, sugar, fat, fibre and macronutrients, people ate around 500 kcal per day more on the ultra-processed diet — without being asked to, and without reporting feeling any hungrier.

The same energy also arrives with very different company. Five hundred calories can come carrying protein, fibre and a range of micronutrients, or almost none of them — and protein is the macronutrient most consistently linked in trials to feeling full and eating less afterwards. It is why the food database on this site reports what each food supplies, not only what it costs in calories.

So the evidence points to something unglamorous: a deficit large enough to work but small enough to sustain, enough protein and resistance training to protect lean mass, and a timescale measured in months rather than weeks. How lean mass is protected in a deficit is in Fat loss.