GLP-1 medicines and rapid discontinuation — what the withdrawal-arm data show

The withdrawal-arm evidence in plain English

Randomised withdrawal trials are a specific trial design in which participants first receive active treatment for a run-in period, and are then randomised to either continue active treatment or switch to placebo, with outcomes compared between the two groups for the remainder of the study. This design isolates the effect of continued treatment from the effect of the initial treatment period, and is a more rigorous way of answering "what happens if you stop" than simply observing people who choose to stop on their own, which is subject to confounding (people who stop for medical reasons, cost reasons, or because treatment already achieved their goal are systematically different from those who continue).

STEP-4 (semaglutide) and SURMOUNT-4 (tirzepatide) findings

STEP-4 applied this randomised withdrawal design to semaglutide: after an initial run-in period of active treatment, participants were randomised to continue semaglutide or switch to placebo. Those switched to placebo experienced substantial weight regain relative to those who continued treatment, who generally maintained or extended their weight loss. SURMOUNT-4 applied the same design to tirzepatide, with a closely comparable pattern of outcomes: substantial regain in the placebo-switch group against continued maintenance or further loss in the continued-treatment group. The consistency of this result across two different molecules and trial programmes is one of the stronger pieces of evidence in the GLP-1 literature, precisely because of the randomised design. For context on the original treatment-phase results these withdrawal arms build on, see STEP-1 and SURMOUNT-1. The broader implications of this pattern are discussed further at weight regain after GLP-1 discontinuation.

Why tapering has no established evidence base for these agents

Dose titration during initiation of GLP-1 therapy is a tolerability strategy for managing gastrointestinal side effects as treatment begins — it is not derived from, or evidence for, an equivalent strategy in reverse when stopping. STEP-4 and SURMOUNT-4 both studied an abrupt switch to placebo rather than a gradual dose reduction, so neither trial provides evidence about whether tapering the dose down before full discontinuation would change the pattern or pace of weight regain, reduce any withdrawal-related symptoms, or offer any other advantage over stopping at the treatment dose. No pivotal trial to date has been specifically designed to test a tapering protocol against abrupt discontinuation for these agents. This is a genuine, acknowledged gap in the published evidence base rather than a settled question with an answer this page is simply not providing.

What patients and prescribers should discuss before stopping

Given the consistency of the withdrawal-arm weight-regain pattern, a conversation with the prescriber managing treatment before stopping — planned or otherwise — allows expectations to be set based on the available trial evidence, and allows any individual monitoring or support plan to be discussed in advance rather than reactively. Relevant topics generally include the reason for stopping, what change in weight and other metabolic parameters might reasonably be expected based on published withdrawal-arm data, whether tapering is appropriate for the individual case despite the limited evidence base, and what non-pharmacological strategies — dietary, activity-based, behavioural — might help support outcomes after stopping. See long-term use and discontinuation for the related maintenance-strategy discussion.

What the evidence does not establish

STEP-4 and SURMOUNT-4 are relatively short-duration trials by the standard of a lifelong or multi-year treatment decision, and neither trial nor any other published pivotal study has directly compared tapering against abrupt discontinuation. Withdrawal-arm data describe average outcomes across trial populations; they do not predict the specific trajectory for any individual, whose response to stopping may be influenced by factors such as duration of prior treatment, degree of weight lost, and concurrent lifestyle changes that are not fully captured in the trial data. This page does not provide a tapering schedule, a stopping timeline, or individual guidance — these require direct discussion with the prescriber managing treatment.

Frequently asked questions

What happens if I stop taking a GLP-1 medicine?

Randomised withdrawal-arm trial data consistently show that most participants regain a substantial proportion of the weight they lost after switching from active GLP-1 or dual-agonist therapy to placebo, typically over the following months to about a year of follow-up. This reflects the fact that these medicines treat an underlying, ongoing physiological drive toward weight regain rather than producing a permanent metabolic reset — appetite and weight-regulation pathways generally return toward their pre-treatment state once the medicine is no longer present.

What did STEP-4 show about stopping semaglutide?

STEP-4 used a randomised withdrawal design: participants first completed a run-in period on semaglutide, then were randomised either to continue semaglutide or switch to placebo for the remainder of the trial. Those switched to placebo experienced substantial weight regain relative to those who continued active treatment, while the continued-treatment group generally maintained or continued to lose weight. This trial is one of the clearest pieces of evidence that continued treatment, not just the initial treatment period, is what sustains the weight-loss effect.

What did SURMOUNT-4 show about stopping tirzepatide?

SURMOUNT-4 used a similar randomised withdrawal design for tirzepatide: an initial run-in period on active treatment followed by randomisation to continued tirzepatide or switch to placebo. As with STEP-4, participants switched to placebo regained a substantial proportion of the weight lost during the run-in period, while those continuing active treatment maintained their weight loss or continued to improve. The consistency of this pattern across both semaglutide and tirzepatide withdrawal trials strengthens the general conclusion that ongoing therapy, rather than a fixed treatment course, is what sustains outcomes for these medicines as studied.

Is there an official tapering schedule for stopping a GLP-1 medicine?

No. There is no established, evidence-based tapering protocol published in product labelling or pivotal trial design for GLP-1 or dual-agonist medicines — the withdrawal-arm trials described above studied abrupt switch to placebo, not a gradual dose reduction, so they do not provide evidence either for or against tapering as a strategy. Any decision to taper rather than stop abruptly is an individualised clinical judgement made by the prescriber, not a protocol this page can set out.

Why is there no tapering evidence if the medicines are titrated up slowly?

Dose titration on the way up is primarily a tolerability strategy, designed to reduce gastrointestinal side effects as the body adjusts to the medicine's pharmacological effects — it is not evidence about what happens, or what is optimal, when stopping. No pivotal trial to date has been specifically designed to compare a gradual dose reduction against abrupt discontinuation for outcomes such as weight regain, appetite return, or tolerability of stopping. This is a genuine evidence gap, not an oversight in how this page summarises the literature.

Does weight regain mean the medicine stopped working?

No — weight regain after stopping reflects the return of underlying appetite and metabolic regulation toward the pre-treatment state once the medicine's pharmacological effect is no longer present, not a failure or loss of effect of the medicine while it was being taken. This is consistent with how these medicines are understood to work: they modulate appetite and metabolic signalling while present in the body, rather than producing a lasting structural change that persists independently of ongoing treatment.

Is weight regain the only thing that happens after stopping?

Weight regain is the most extensively studied outcome in withdrawal-arm trials, but other metabolic parameters that improved during treatment — such as blood pressure, blood glucose and lipid measures in some studies — have also been reported to trend back toward baseline alongside weight regain in withdrawal-arm follow-up, broadly tracking the weight change rather than behaving as an independent, persistent benefit.

Should I plan to stay on a GLP-1 medicine indefinitely?

Whether a GLP-1 medicine is intended as long-term, indefinite therapy or a defined-course treatment is a decision made with the prescriber managing treatment, informed by the original indication, individual response, tolerability and broader health goals. Current trial evidence on weight maintenance after stopping supports the general principle that continued benefit generally requires continued treatment, but this page does not make an individual recommendation about whether or for how long any specific person should remain on therapy.

What should I discuss with my prescriber before stopping?

Relevant discussion points generally include the reason for stopping (cost, side effects, achieving a goal weight, medical reasons, planned pregnancy), expectations about weight and metabolic parameter trends based on the withdrawal-arm evidence described here, whether any tapering or monitoring plan is appropriate for the individual case, and what non-pharmacological support (dietary, activity-based, behavioural) might help sustain outcomes after stopping. This page provides background evidence for that conversation; it does not replace it.

Is this withdrawal pattern different from other weight-loss treatments?

Weight regain after stopping an effective weight-loss intervention is not unique to GLP-1 medicines — it has also been described after stopping other pharmacological weight-loss treatments and, to varying degrees, after non-surgical lifestyle interventions once the intervention is no longer actively maintained. What is comparatively well documented for GLP-1 medicines specifically is the randomised, controlled trial evidence (STEP-4, SURMOUNT-4) quantifying the pattern under controlled conditions, which is a stronger form of evidence than the observational data available for many other interventions.