Why it stalls, and what a plateau really is.

First: is it actually a plateau?

Most reported plateaus are two or three weeks long, and two or three weeks is not enough data to see through the noise. Bhutani and colleagues, weighing adults at baseline and again two weeks later in Physiological Reports (2017), found about 84% of a two-week change in body weight was fat-free mass rather than fat, and note that total body water alone can fluctuate by up to about 5% daily.

So the first question is not why you stalled. It is whether a rolling average (not a reading) has been flat for several weeks. Below that, the honest answer is that you cannot yet tell, and why the scale jumps is the longer version of that argument.

The models predict this

Weight loss decelerating and stopping is not a malfunction. It is what the arithmetic says will happen at a fixed intake. Hall and colleagues, in The Lancet (2011), simulated a 100 kg man cutting about 480 calories a day: the model settles at a new weight over several years, reaching half the total change in roughly the first year.

The reason is simple enough. A smaller body costs less to run, so the same intake that was a deficit at the start is a smaller deficit later, and eventually is not one at all. A plateau is the point where the two lines meet. Anything that promised you a straight line was running the 3,500-calorie rule, which cannot represent a plateau at all.

Worth knowing before reading on: the plateau people actually hit in clinical practice arrives earlier than the models say it should. Thomas and colleagues, in the American Journal of Clinical Nutrition (2014), report clinical weight loss typically stabilizing at around six months where validated energy-balance models predict one to two years, and attribute the gap to intermittent lapses in adherence rather than to metabolic adaptation. Adding metabolic adaptation to their model raised the final weight but did not move the timing.

The four mechanisms

1. Expenditure falls, and by more than the smaller body accounts for. Leibel, Rosenbaum and Hirsch measured this directly in the New England Journal of Medicine (1995): maintaining a weight at least 10% below the starting point, total energy expenditure fell by 6 to 8 calories per kilogram of fat-free mass per day, with resting and non-resting expenditure contributing three to four each. It is symmetric, gaining 10% raised expenditure by a similar amount.

2. The movement you were not counting falls too. Redman and colleagues followed 48 adults for six months in PLOS ONE (2009), measuring free-living expenditure with doubly labeled water. In the calorie-restricted groups, activity-related expenditure fell by about 420 calories a day at three months and was still about 228 down at six. No decision to move less was involved, and the group that exercised as well as restricted did not show it.

3. Appetite rises in proportion to what you have lost. Polidori and colleagues analyzed the weight trajectories of 153 patients in a year-long trial of a drug that creates a covert deficit, and reported in Obesity (2016) that appetite rises by an estimated 100 calories a day for every kilogram lost, more than three times the corresponding fall in expenditure. Hall and Kahan restate the pair in the Medical Clinics of North America (2018): per kilogram lost, expenditure falls roughly 20 to 30 calories a day while appetite rises roughly 100. It is a model-derived estimate rather than a measurement of hunger, and it is the clearest account of why a deficit that felt manageable in month one is genuinely harder in month five without anything about you having weakened.

4. Intake drifts upward without being noticed. This one is uncomfortable and it is very well evidenced. Freedman and colleagues pooled five large US validation studies against recovery biomarkers in the American Journal of Epidemiology (2014) and found people under-report energy intake by about 28% on a food frequency questionnaire and about 15% on a single 24-hour recall, with body mass index, age and education all predicting how much. Hall and colleagues concluded from the shape of the curve in The Lancet that the plateau at six to eight months implies the original intake reduction was “rapidly and progressively relaxed”. The point is not dishonesty. Estimating is simply hard, and it drifts in one direction.

Telling them apart

They leave different fingerprints, which is the practical use of separating them.

If a rolling average has been flat for several weeks while logged intake looks unchanged, the two candidates are a fallen expenditure and a drifted intake. The way to distinguish them is unglamorous: log the way you logged in week one, weighing rather than estimating, for two weeks, and see whether the numbers agree with what you thought you were eating. Given the figures above, often they will not, and that is a property of estimation rather than of you.

If the scale is flat but measurements or training loads are still moving, composition is changing underneath a stable weight, and the scale is answering a different question from the one you are asking. That the two can move independently is directly measured: Hall and colleagues ran 19 adults through a metabolic ward in Cell Metabolism (2015) and the arm that lost 42% more body weight lost no more fat, by chamber measurement it lost less. Cava, Yeat and Mittendorfer, reviewing the evidence in Advances in Nutrition (2017), put fat-free mass at 20 to 30% of weight lost in people who are overweight or obese and often above 35% in people of normal weight, and name adequate protein and resistance training as what limits it.

And if weight changes sharply with no change in habits, or there are symptoms alongside it, that is a reason to speak to a clinician rather than to adjust a target.

What actually resolves one

This is where a website has to stop, because the answer belongs to a person rather than to the mechanism. What can be said generally is that the useful move is diagnostic rather than dramatic. Find out which of the four is happening before changing anything: the response to a drifted intake and the response to a fallen expenditure are not the same response, and cutting harder in answer to the wrong one tends to make appetite and lean-tissue loss worse at once. What a sensible rate or intake is for you specifically is a question for a doctor or a registered dietitian who can see your whole situation.

How Loopa handles this

Loopa’s weight trend draws a seven-day average over your weigh-ins, which is what makes a real plateau distinguishable from two weeks of water; that, and food, weight and water logging, are free. Premium opens the full history beyond thirty days, the forecast that decelerates as you lean out and shows a plateau when your data says plateau, and reports that put intake, weight and activity on one page so a drift in one is visible against the others.