The Real Maths of Weight Loss (It Is Not 3,500 Calories a Pound)

Energy balance is real: to lose fat you must use more energy than you take in. What is not real is the tidy arithmetic built on top of it, and understanding why explains most stalled diets.

The 3,500-calorie rule is wrong

The familiar rule — a pound of fat is 3,500 calories, so a 500-calorie daily deficit loses a pound a week — comes from a calculation made in 1958. Followed to its conclusion it predicts that a modest daily deficit maintained for five years would take a person down to nothing.

It fails because it treats both sides of the equation as fixed. Neither is.

What changes as you lose weight

You get smaller. A lighter body costs less to run and less to move. Someone who has lost 15 kg burns meaningfully fewer calories doing exactly the same day.

Adaptive thermogenesis. Beyond what body size predicts, energy expenditure falls further during sustained restriction — thyroid output shifts, and the body becomes more efficient. The magnitude is debated and it is not zero.

NEAT falls, invisibly. Non-exercise activity thermogenesis — fidgeting, pacing, standing, gesturing, taking the stairs — can vary by hundreds of calories a day between people, and it drops substantially in a deficit without any conscious decision. You move less, and you do not notice.

Together, these mean a deficit shrinks over time on its own. The plateau is not a failure of willpower. It is arithmetic.

The other side: nobody tracks accurately

Studies using doubly labelled water — the reference method for measuring energy expenditure — consistently find that people under-report intake by 20 to 50%. This holds for dietitians. It holds for people who are certain they are tracking carefully.

It is not dishonesty. It is oil in the pan, the handful while cooking, the underestimated portion, the weekend not logged, the drink not counted. Small and systematic in one direction.

So when someone says they are eating 1,400 calories and not losing weight, the likeliest explanation is not a broken metabolism. It is that they are eating closer to 1,900.

What this means in practice

Expect the rate to slow. Build it into your expectations rather than treating it as failure.
Aim for 0.5 to 1% of body weight per week. Faster costs more muscle and is harder to sustain.
Weigh yourself often, judge by weekly averages. Daily weight swings by 1 to 2 kg from water, glycogen, sodium and digestion. A single reading tells you almost nothing; a seven-day average tells you a lot.
Recalculate periodically. The intake that produced a deficit at the start will be maintenance later.
Take planned maintenance breaks. A week or two at maintenance every couple of months makes long stretches more sustainable and appears to blunt some of the adaptation.

The two things that protect the outcome

Protein. Requirements go up in a deficit, not down — roughly 1.6 to 2.4 g per kg of body weight per day. This is what determines how much of the loss comes from fat rather than muscle.

Resistance training. The signal that tells your body to keep the muscle it has while the fat comes off. Without it, a substantial share of what you lose is lean tissue — which lowers expenditure further and makes the weight easier to regain.

The variable that beats all the others

When diets are compared head to head with calories and protein matched, differences between them largely disappear. Low-carb, low-fat, Mediterranean, intermittent fasting — in controlled comparisons they perform similarly.

What differs is adherence. The best diet is the one you will still be following in eight months, which is a question about your life rather than about metabolism. Choose on that basis, not on mechanism.