The plateau is mostly arithmetic, and partly not
Weight loss slows and then stops, and there are two stories about why. One says your metabolism broke. The other says you are eating more than you think. The measurements say: mostly the arithmetic, with a real and smaller contribution from adaptation.
The arithmetic first
A mathematical model of adult metabolism, built to simulate how energy expenditure changes during weight loss, produced a result worth memorising: the body weight response to a change in energy intake is slow, with a half-time of about one year.
Read that again, because almost every expectation about dieting contradicts it. Cut your intake by a fixed amount and you do not drop to a new weight in six weeks and stop. You approach a new steady weight over years, taking about a year to cover half the distance. The model also found that people carrying more fat have a larger expected loss for the same change in intake, and take longer to reach their steady state.
The plateau, on this account, is not a failure. It is the shape of the curve. You are still moving; you are moving more slowly because you are closer to where that intake puts you.
There is a second, less comfortable half of the arithmetic. A smaller body needs fewer calories to run. The deficit you created at your starting weight is a smaller deficit — or none — at your current one. Nothing broke. The sum changed while you were doing it.
The adaptation, which is real
The other story is not fiction, it is just smaller than people claim. Fourteen participants from a televised weight-loss competition were followed for six years. Resting metabolic rate was measured by indirect calorimetry, body composition by DXA, and metabolic adaptation was defined as what remained of the change in resting rate after adjusting for changes in body composition and age.
That residual persisted six years later. The authors describe it as a proportional but incomplete response to the ongoing effort to hold a lower weight.
Two things are worth keeping straight. It is real — the adjustment was made, and something was left over. And it is an extreme case: fourteen people who lost 58 ± 25 kg under competition conditions. Reading their numbers as your own is reading the far tail of the distribution.
What to do about a plateau
Confirm it is a plateau first. Most are not. Daily weight noise can hide a real trend for weeks, so compare two 7-day averages a fortnight apart before concluding anything. A month of “no progress” often turns out to be a slow, ordinary decline underneath the noise.
Then check the deficit, not the metabolism. If the trend really is flat over a month of averages, the sum has caught up with you. The lever is intake or activity, and the size of the correction is smaller than most people apply.
Do not stop measuring. Among 2,838 weigh-ins from 40 people, weight loss occurred during periods of daily self-weighing, while gaps longer than a month were associated with weight gain. That is an association in a small group, and the likeliest direction is that people stop weighing when things go badly. Which is exactly the moment a plateau arrives.
What this does not say
None of it says a plateau means you should eat less than is sustainable, or that a slower loss is a worse one. Half-times of a year cut both ways: the same slowness that makes progress feel invisible also means a bad fortnight does not undo anything.