Does Calories In, Calories Out Hold Up?

The short answer

Energy balance is not a diet theory, it is bookkeeping: what the body stores changes by what you eat minus what you spend, and no food, timing trick or hormone escapes the ledger. That part is not in dispute and cannot be falsified. What fails is one specific use of it, the claim that cutting 500 kcal a day removes about 0.45 kg of fat a week for as long as you keep cutting.

That prediction assumes expenditure holds still, and it does not. Eat less and the body spends less: there is less tissue to move, less food to digest, and a measurable fall in spontaneous movement. Hunger and under-reporting push from the other side. The arithmetic survives; the assumption of a constant does not.

What the equation actually says

The identity is simple. Body stores change when intake and expenditure stop matching, and fat holds roughly 7,700 kcal per kilogram. A 500 kcal daily deficit is therefore about 0.45 kg of fat a week while nothing else moves. That is exactly the sum the deficit calculator runs when it turns a target rate of loss into a daily calorie figure, and it is the same arithmetic used to size a target for a person.

Because it is an accounting identity rather than a mechanism, it says nothing about which diet is best, how quickly you can change safely, or how you will feel while doing it. Those questions are answered by appetite, adherence and training. The model is a map of the terrain rather than a description of the walk, and most arguments about it are really arguments about the walk.

Why the arithmetic overshoots after a few kilos

Start at 75 kg with maintenance of about 2,662 kcal a day and a 500 kcal deficit, which puts the target at 2,162 kcal. The first weeks behave exactly as promised. Then the weight comes off, and maintenance falls with it, because a lighter body costs less to run. The same calculator at the same age and height returns a BMR of 1,518 kcal at 55 kg, 1,718 kcal at 75 kg and 1,918 kcal at 95 kg, so twenty kilos of body weight is worth about two hundred calories a day before activity is considered.

Keep the original target and the real deficit quietly shrinks. Nothing about the plan changed, the result did, which is why a plateau at week eight is usually arithmetic rather than personal failure. The fix is to re-run the numbers at the new weight rather than to cut again on top of a deficit that is no longer the size you think it is.

Metabolic adaptation, the part beyond body size

Some of the fall in expenditure is just mass, because less tissue burns less. The rest is the body adjusting to a shortfall. Spontaneous movement drops, digesting food costs a little less, and the tissue itself becomes marginally more efficient. Careful measurements put the extra fall in the region of fifty to a hundred and fifty calories a day below what weight and body composition alone would predict.

That effect is real and it is modest. It does not amount to a broken metabolism and it cannot make a deficit disappear on its own. It does mean the deficit you designed is smaller than the one you are actually running, which is another reason to re-measure the target every two or three kilograms instead of defending the number you started with.

Adherence is the biggest error term

In a metabolic ward intake is weighed and expenditure is measured, and the arithmetic lands where it should. In ordinary life neither side is measured. Under-reporting of food runs at something like twenty to thirty percent in adults who are overweight, while exercise is usually over-estimated at the same time, so the gap between the plan and reality opens from both directions at once.

The error is invisible from the inside, because nobody believes they are misremembering their own week. The practical test is to weigh and log food for two weeks and compare the daily average with the target the calculator produced. If the two disagree by a wide margin, the log is the first suspect, well ahead of metabolism.

Compensation: the body pushes back on activity

Exercise burns calories and then influences appetite and movement for the rest of the day. The compensation is partial and varies enormously between people, from almost nothing to more than half. A thirty minute run costs about 386 kcal at 75 kg by the calculator's own figures; if half of that comes back through hunger and a quieter afternoon, the net is nearer 190 kcal.

This is why adding activity rarely delivers the full sum on the treadmill display, and why the scale sometimes ignores a new training block entirely for a few weeks. It is not an argument against exercise. Regular activity is among the strongest predictors of keeping weight off, but it earns that role through muscle, fitness and habit rather than through the calorie count of any single session.

What the population models got wrong

Model-based projections of how much weight a whole population would lose from a modest reduction in intake have had to be revised downwards once measured intake and measured expenditure replaced estimates. Two errors compound: food surveys under-count what people eat, and activity questionnaires over-count what they do. A model fed both will confidently project a change that never arrives.

The lesson is about inputs rather than physics. When one person tracks honestly, the balance model describes their trend well. When analysts model a nation on survey data, the errors swamp the signal. The same caution applies to any online calculator, including the ones here: the output is a starting estimate that expects to be corrected by the scale.

What the model still gets right

As a planning tool it remains the best one available. It gives the direction, since a deficit loses weight and a surplus gains it; the rough size, at about 0.45 kg a week per 500 kcal; and the levers, because intake and expenditure can each be moved. Every credible programme, clinical or athletic, is built on this frame.

It also tells you what to do when progress stops. If weight has been flat for three weeks on a recorded intake, then either true intake is higher than recorded or expenditure has fallen below the assumption, and one of the two has to change. That is a diagnosis that calorie counters and their critics both end up using.

What the model cannot do is promise a rate. It describes the direction of travel and the size of the slope for a few weeks at a time. Anyone who quotes it as a guarantee for a year is quoting an assumption rather than the equation.

How to use the model without being misled

Treat the 7,700 kcal per kilogram as a planning constant rather than a promise. It sizes a deficit well over four to six weeks and poorly over a year, because the expenditure side moves underneath it. Set three hundred to five hundred calories below measured maintenance, weigh yourself daily, and judge the weekly average rather than any single morning.

• Re-measure maintenance every two or three kilograms of loss, because the target drifts down with the body.
• Expect the rate to slow near the halfway point and read that as arithmetic, not as a personal failing.
• If the weekly average has been flat for three weeks, correct one side of the ledger before cutting harder.
• Keep protein high and keep training, so the weight that comes off is the weight you wanted to lose.

How to use this

Take maintenance from the TDEE calculator at an honest activity level, cut three to five hundred calories with the deficit calculator, and cover protein with the protein calculator. The macro calculator turns whatever target you settle on into grams. If your weight has been flat for three weeks, re-run the TDEE figure at the current weight before you cut anything further, because the old number was calculated for a heavier person.

Then run the loop the model actually needs: four weeks at the target, a comparison of weekly averages, and a fresh calculation at the new weight. Used that way the arithmetic holds up well enough to plan with, which is all it ever promised.

Frequently asked questions