Not everybody regains at the same rate, and nobody knows why

What is thinly evidenced is the composition of the regained weight. The withdrawal trials measured mass, not composition, with the single exception of a maintenance study that scanned participants throughout. The general concern — that weight lost as a favourable fat-to-lean ratio returns as a less favourable one — is physiologically plausible, is supported by fragments of the dietary weight-loss literature, and has not been measured in anybody who stopped a current agent. The Journal reports the concern as a hypothesis rather than as a finding.

The STEP 1 extension: two-thirds back within a year

The pivotal semaglutide obesity trial ran for sixty-eight weeks with a mean weight reduction of approximately 14.9 per cent on 2.4 mg weekly against 2.4 per cent on placebo.¹ An extension followed a subset of participants for a further fifty-two weeks after both the drug and the lifestyle intervention were withdrawn, which makes it an off-treatment observation rather than a randomised withdrawal.

By week 120 — a year after stopping — participants who had received semaglutide had regained approximately two-thirds of the weight they had lost, finishing on average around 5.6 per cent below their original baseline against approximately 0.1 per cent for the former placebo group.² Improvements in glycaemic parameters, blood pressure and lipids reverted broadly in step with the weight.

Two details are consistently dropped from summaries. The residual benefit was real: a mean 5.6 per cent reduction sustained a year after stopping is not nothing, and it is more than most non-pharmacological interventions achieve while they are still being delivered. And the lifestyle support was withdrawn at the same time as the drug, so the extension describes the removal of an entire intervention package rather than of a molecule.

The physiology was described before the drugs existed

The homeostatic response to weight loss was characterised in a study that predates this drug class and remains the clearest account of it. After a substantial diet-induced reduction, circulating leptin fell and remained suppressed, ghrelin rose and remained elevated, satiety hormones including peptide YY and cholecystokinin fell, and subjective hunger was significantly greater than baseline — and all of these persisted at twelve months, long after active dieting had ended.³

Alongside the hormonal changes sits a reduction in resting energy expenditure larger than the loss of metabolically active tissue predicts, an effect usually termed adaptive thermogenesis. Its magnitude and persistence are contested, but in the most extreme documented cohort — participants in a televised competitive weight-loss programme followed for six years — resting metabolic rate remained substantially suppressed relative to prediction long after most of the weight had returned.⁴

Both findings explain regain after drug withdrawal without invoking anything specific to incretins. A drug that suppresses appetite is holding a system away from a defended state. Remove the drug and the system, which has been signalling for restoration throughout, gets what it has been asking for. That is not a drug effect; it is the condition the drug was treating becoming visible again.

What comes back, and in what proportion

The composition of regained weight is the thinnest part of this literature. None of the three withdrawal designs measured body composition after cessation. The concern most often voiced — that weight lost in a favourable fat-to-lean ratio returns in a less favourable one, so that repeated cycles progressively worsen composition — is physiologically plausible and, in this drug class, entirely unmeasured.

What exists comes from the dietary weight-cycling literature, where the picture is mixed rather than alarming: several studies find that regained weight is disproportionately fat, several find no such asymmetry, and the meta-analytic position is that weight cycling has not been shown to produce a cumulative composition penalty in humans. The Journal regards that as genuinely unresolved rather than as reassurance.

One dataset does bear on it indirectly. In the maintenance trial that randomised exercise, a GLP-1 receptor agonist, both or neither after a diet-induced loss, composition was tracked throughout, and the arms that trained retained a more favourable composition through the maintenance year.⁵ That is a statement about maintenance rather than about regain after withdrawal, and it is the closest thing to relevant evidence anybody has.

The variance in regain, which is as large as the variance in loss

Mean regain trajectories conceal a distribution as wide as the one that characterises the initial response. In the reported withdrawal arms, some participants returned to within a percentage point of their original baseline within a year while others retained most of their loss with no pharmacological support whatever. The interquartile ranges in the published figures are broad, and the trials were not designed to explain them.

Nothing measured at randomisation predicts an individual regain trajectory usefully. Baseline body mass index, the magnitude of the initial loss, age, sex and diabetes status all shift the mean modestly and leave the spread largely intact. The behavioural variables that plausibly matter — what a person was eating and doing during the loss, and whether any of it persisted — were not measured with the granularity required to test them.

This is the same structural gap that runs through the whole field: the mean effects are well characterised and the variance is not. It has a specific practical consequence here. A person deciding whether to stop cannot be given a personal probability of holding their weight, because no such probability has been established, and any source offering one is offering a group mean dressed as a prediction.

The first six weeks after the last injection

Exposure falls on a predictable schedule. A week after a final weekly injection roughly half the steady-state concentration remains; at two weeks a quarter; at four weeks a few per cent. Appetite returns along a curve that lags the concentration curve somewhat, and delayed gastric emptying resolves over a similar interval, so the sensation of early fullness that has been governing meal size for months disappears over about a month.

What people report, and the Journal reports it as report rather than as measurement, is that the appetite return is experienced as abrupt rather than gradual — as a threshold being crossed somewhere in the third or fourth week rather than as a smooth ramp. That is consistent with a non-linear relationship between receptor occupancy and the perceived effect, which is what the dose-response data would predict, and it is not consistent with the exposure curve alone.

Glycaemic parameters follow their own timescale. Fasting glucose responds within days to weeks; HbA1c, which reflects the preceding three months with the recent weeks weighted most heavily, will not show the full consequence of stopping for a full quarter. A panel drawn four weeks after cessation will understate the change, and this is a specific and common misreading. In the trial population with type 2 diabetes, where glycaemic control was a co-primary concern, the treatment effect on HbA1c was of the order of one and a half percentage points, which is the scale of what a cessation eventually undoes.⁶

The lifestyle intervention is a confound in both directions

Every trial in this class delivers a behavioural intervention alongside the drug: energy-restriction targets, activity targets, and regular contact with a study team. That contact is itself an intervention of measurable effect, which is why placebo arms in these programmes lose two to three per cent of body weight rather than nothing. Where the behavioural component was deliberately intensified, the placebo arm lost around 5.7 per cent over sixty-eight weeks, which is a useful upper bound on what contact and counselling alone achieved in these populations.⁷

It matters for the withdrawal question in a way that is usually elided. The semaglutide off-treatment extension withdrew the drug and the lifestyle support together, so its regain figure describes the removal of a package.² The STEP 4 and SURMOUNT-4 withdrawal arms kept the lifestyle component running, so their regain figures describe the removal of a molecule with support maintained.⁸⁹ Those are different experiments and the second is the more conservative.

Anybody comparing regain figures across the three should therefore expect the extension to look worse, and it does. The Journal states which withdrawal design a figure comes from every time it quotes one, because the alternative is pooling two different experiments into a single number that describes neither. The same caution applies to the frequent comparison with dietary weight-loss regain, where the behavioural intervention is the whole of the treatment.

Readers deciding whether to stop should be told the two things the trials do establish and the one thing they do not. Regain begins promptly and decelerates; the residual benefit at a year is real but modest; and no dataset anywhere permits a personal prediction. Any source offering an individual probability of holding weight after cessation is presenting a group mean as a forecast.