How to Taper Off GLP-1 Drugs: What Published Protocols and Obesity-Medicine Practice Actually Do

The published randomised data on stopping a GLP-1 is, almost entirely, cold-stop data. In the STEP 1 extension, patients who stopped semaglutide regained two-thirds of their lost weight within 52 weeks (Wilding et al., 2022). In SURMOUNT-4, tirzepatide patients regained about 14 percentage points in the year after withdrawal (JAMA Internal Medicine 2025). Both trials stopped the drug abruptly. The structured taper schedules used in most obesity-medicine clinics today were not what was studied.

That gap, between trial design and clinical practice, is the reason this page exists. We describe what published protocols and current obesity-medicine practice actually do, where the evidence is solid, and where it is thin. This page is awaiting medical-reviewer signoff. It describes practice; it does not prescribe one. A different scenario — pausing the drug briefly and then restarting — is covered separately in GLP-1 drug holidays.

The basic question: cold-stop versus structured taper

Discontinuing a GLP-1 means one of two things in the published literature. Either the patient stops at their last therapeutic dose (trial protocols call this "withdrawal"; patients usually call it cold turkey), or the dose is stepped down through the manufacturer's titration ladder over several weeks before stopping. The randomised data is almost entirely the first version.

STEP 1's 1-year off-treatment extension randomised 1,961 adults to semaglutide 2.4 mg or placebo for 68 weeks, then stopped both. The semaglutide arm had lost a mean 17.3 percent of body weight by week 68 and regained 11.6 percentage points by week 120, leaving a net loss of 5.6 percent. SURMOUNT-4 followed the same shape: by week 88, the placebo (withdrawal) arm sat at 9.9 percent below baseline versus 25.3 percent for continued treatment.

The structured-taper data is sparse. Real-world cohort studies and clinic series describe step-down schedules, but a randomised head-to-head of taper versus cold-stop with matched maintenance support has not yet been published. A 2026 systematic review in eClinicalMedicine pooled discontinuation studies and estimated that weight regain plateaus at roughly 75 percent of the on-treatment loss (Trajectory of weight regain). None of the contributing studies was a randomised taper-versus-cold-stop comparison.

This matters for how patients should read what follows. Step-down schedules used in obesity-medicine clinics are a clinically reasonable extension of pharmacokinetic logic, but they are not, in 2026, a protocol with randomised efficacy evidence behind them.

Why tapering might help

The mechanistic argument is straightforward. GLP-1 receptor agonism produces sustained satiety signalling, slowed gastric emptying, and central appetite suppression (see How GLP-1 Drugs Work for the underlying mechanism). When the drug is stopped, those signals fall as the molecule clears. Semaglutide's half-life is approximately 7 days, reaching biologically inactive levels around 35 days after the last dose; tirzepatide is closer to 25 days; liraglutide is under 3.

Stopping at the highest therapeutic dose produces the steepest drop in receptor agonism. Stepping down spreads that drop across weeks, in theory giving appetite, gastric, and food-reward signals more time to recalibrate at each rung. The behavioural argument runs parallel: a slower handover gives nutrition counselling, resistance training, and weighing habits more time to consolidate before the pharmacological scaffolding is gone.

Both arguments are plausible. Neither is settled. The literature documents weight regain after cold-stop; what it does not yet document is how much of that regain a structured taper actually prevents.

Standard step-down schedules in current obesity-medicine practice

Major obesity-medicine clinics and published clinic protocols converge on schedules that mirror the manufacturer's titration ladder in reverse. The schedules below are what practising clinicians describe in published commentary and clinic protocols; they are not from a randomised trial and they are not in the FDA label.

For semaglutide (Wegovy), the dose ladder runs 0.25 → 0.5 → 1.0 → 1.7 → 2.4 mg on the way up. The taper inverts this:

• 2.4 mg weekly (current dose)
• 1.7 mg weekly for 4 weeks
• 1.0 mg weekly for 4 weeks
• 0.5 mg weekly for 4 weeks
• 0.25 mg weekly for 4 weeks
• discontinue

The total runs roughly 12 to 20 weeks depending on cadence; some clinics step down every 4 to 6 weeks rather than every 4 and hold longer at the lower doses where appetite return is most likely.

For tirzepatide (Zepbound), the dose ladder is longer (2.5 → 5 → 7.5 → 10 → 12.5 → 15 mg), so the taper is usually longer too:

• 15 mg weekly (current dose)
• 12.5 mg weekly for 4 weeks
• 10 mg weekly for 4 weeks
• 7.5 mg weekly for 4 weeks
• 5 mg weekly for 4 weeks
• 2.5 mg weekly for 4 weeks
• discontinue

That runs about 16 to 24 weeks. A live question that obesity-medicine clinicians are increasingly raising: whether to stop at 5 mg or 2.5 mg and continue indefinitely as a maintenance dose rather than coming off completely. SURMOUNT-MAINTAIN randomised participants who had reached a weight plateau on tirzepatide to continue at maximum tolerated dose, step down to 5 mg, or switch to placebo. At week 112, mean body weight change from baseline was -21.9 percent (max-dose), -16.6 percent (5 mg), and -9.9 percent (placebo). Translated: tirzepatide 5 mg held roughly 70 percent of the weight loss versus placebo. An equivalent randomised maintenance trial for semaglutide has not yet read out.

These schedules are descriptions of practice patterns. They are not a treatment plan. Cadence, the threshold for pausing or reversing a step, and the decision to stop entirely versus transition to maintenance are clinical decisions depending on weight trajectory, glycemic status, comorbidities, and intent.

What goes with the taper

The literature on weight-loss maintenance after pharmacotherapy is clearer than the literature on the taper itself. The through-line across professional-society guidance is that behavioural and nutritional infrastructure matters at least as much as the drug schedule.

The 2025 AACE algorithm for the medical care of adults with obesity frames obesity as a chronic relapsing disease that requires long-term treatment and explicitly cautions against discontinuation as a default endpoint. The Obesity Medicine Association uses the same framing. Both societies recommend that any planned discontinuation occur alongside a structured maintenance plan including nutrition follow-up with a registered dietitian, behavioural support, monitored exercise, and ongoing weight and glycemic tracking.

The components practising obesity-medicine clinicians most consistently emphasise during and after a taper:

• Protein intake at the upper end of dietary guidance, typically 1.2 to 1.6 g/kg/day at 20 to 40 g per meal. Lean-mass loss during GLP-1 weight reduction is documented at roughly 30 to 40 percent of total loss in older cohort studies; recent data suggests structured protein intake and resistance training reduce that fraction substantially (Medscape, June 2025).
• Resistance training two to three times weekly with compound movements. Preserving lean mass through weight loss and maintenance protects resting energy expenditure.
• Routine weight tracking, daily or near-daily. National Weight Control Registry data on patients maintaining substantial weight loss for 10+ years consistently identifies regular self-weighing as one of the strongest behavioural correlates of maintenance (Thomas et al., 2014).
• Glycemic and blood-pressure tracking for patients with prediabetes, type 2 diabetes, or treated hypertension. Cardiometabolic benefits of GLP-1 therapy partially reverse with weight regain (SURMOUNT-4 post hoc).
• Behavioural support, typically CBT for obesity or structured mindful-eating programmes. The return of food noise and pre-treatment eating patterns is one of the most commonly reported subjective experiences after a taper.

What is novel here is that these patterns are being layered on top of a pharmacological scaffolding being deliberately withdrawn. The question is how much maintenance the behavioural infrastructure can carry on its own.

What to expect physically during the taper

The patterns below are what clinicians and patients commonly describe. None is universal; each is a reported pattern, not a guarantee.

Appetite return, often gradual at first and more pronounced at the lower dose steps. Most patients describe the return of food noise and pre-meal hunger cues. Timing varies; some notice it within the first week of a dose step, others not until the second or third.

Possible weight movement, in either direction. Some patients lose weight during early taper steps if appetite returns slowly enough to maintain deficit; others regain from the first reduction. The cold-stop data is the floor of plausibility, not the average expectation.

Transient fatigue and mood shifts at dose steps, as food-reward signalling returns gradually. Return of gastric-emptying speed also means meals that felt large on the drug feel smaller off it; portion-size habits learned on therapeutic dosing become more important to maintain through the taper.

If any of these intensify sharply or trigger weight regain above a clinician-defined threshold (often around 5 percent of the prior loss), clinicians commonly pause or reverse the taper step. That decision belongs to the prescriber.

When tapering is not the right approach

Several patient situations call for a different conversation than the generic step-down.

Type 2 diabetes glycemic control. Patients prescribed semaglutide or tirzepatide for diabetes are managing a chronic metabolic disease, not a time-bounded weight-loss intervention. Discontinuation here is rarely about coming off the drug entirely and usually about transitioning between regimens with the endocrinologist or primary-care prescriber managing glycemic targets.

Active eating-disorder concerns. Both the AACE algorithm and obesity-medicine practice patterns flag eating disorders as a contraindication or relative caution for GLP-1 initiation; the taper conversation in this population is highly individualised and should occur alongside the eating-disorder treatment team.

Pregnancy planning. The Wegovy and Zepbound labels both recommend discontinuation at least 2 months before planned conception (Wegovy label; Zepbound label). That timeline is driven by pharmacokinetics and fetal safety, not weight-maintenance logic.

Elective surgery. Perioperative GLP-1 discontinuation has its own decision framework, driven by aspiration risk from delayed gastric emptying (multisociety perioperative guidance, 2024). The timeline is short (days, not months) and is set by the surgical and anaesthesia team.

In each of these cases, the framing is not "should I taper or stop cold" but "what is the goal of stopping, and on what timeline." Those questions belong with the prescriber. If a taper is being forced by losing access rather than chosen — for instance because your compounded supply ended — a deliberate taper is only one of several options, and our guide to your choices when access ends sets it beside the alternatives. If the trigger is specifically cost, our guide to being forced off a GLP-1 by price covers handling an involuntary stop.

What a structured maintenance plan looks like after the taper

If the taper is the bridge, the maintenance plan is the far side. The behavioural patterns the long-term weight-loss literature documents as most strongly associated with sustained maintenance are not specific to the post-GLP-1 patient, but the post-GLP-1 patient may need them more deliberately because the appetite scaffolding is gone.

National Weight Control Registry participants who have maintained ≥30 lb of loss for 6+ years (Thomas et al., 2014) consistently report regular weighing (daily or near-daily, for early detection of drift); sustained physical activity at meaningful volume (registry participants average around 60 minutes of moderate activity daily); high-protein structured eating; resistance training preserved from the taper period; and external accountability via a clinician follow-up cadence, peer or group programme, or structured tracking app.

This is what obesity medicine has known for decades about long-term maintenance, and what GLP-1 pharmacotherapy temporarily made less necessary. After the taper, it becomes necessary again.

The honest caveat

Even with the best structured taper and the most disciplined maintenance plan, published average regain after GLP-1 discontinuation is substantial. The STEP 1 extension's 11.6 percentage point regain in 52 weeks and SURMOUNT-4's ~14 percentage point regain are population averages from randomised trials, and they are the numbers patients should anchor on rather than the best-case testimonial content.

We do not yet know what fraction of patients maintain most of their loss long-term off a GLP-1, or with confidence what distinguishes them. The honest framing: the average outcome is substantial regain, some patients do better, and the maintenance infrastructure above appears to be the best lever we currently have on which side of that average a given patient lands.

A framing some obesity-medicine clinicians now use: a low maintenance dose indefinitely may be a better long-term plan than a clean off-ramp for many patients, particularly given the SURMOUNT-MAINTAIN result on tirzepatide 5 mg. Whether this applies to your situation is a conversation to have with your prescriber.

For drug-specific detail on what cold-stop looks like, see our companion pages on stopping Wegovy and stopping Zepbound once those publish.

What to discuss with your prescriber

Six questions worth bringing to the appointment, regardless of which drug you are on:

1. What is the goal of stopping, in your view: completion of a defined weight-loss phase, transition to maintenance dosing, or a clinical reason such as side effects, cost, or pregnancy planning?
2. If we taper rather than stop, what schedule do you use, and on what evidence do you base that schedule?
3. What weight or symptom thresholds would prompt you to pause the taper, reverse a step, or transition me to indefinite maintenance dosing?
4. What does follow-up look like during the taper: visit cadence, what we track, when we adjust?
5. How will we handle the nutrition, resistance-training, and behavioural-support side, and who is responsible for what?
6. If I am on this drug for type 2 diabetes or cardiovascular risk reduction rather than primarily for weight management, does that change the conversation about stopping at all?

These questions are also a screening exercise. A prescriber who can answer them concretely is the prescriber you want for the off-ramp; a prescriber who cannot is signalling that the off-ramp may not be where their model is set up to serve you.

How we keep this article current

We refresh this page on any randomised trial readout, professional-society guidance update, or label change that touches GLP-1 discontinuation. Taper-protocol research is one of the most active areas of obesity medicine in 2026, and we expect new evidence within 12 to 24 months on three specific questions:

• Taper versus cold-stop. A randomised head-to-head trial of taper versus cold-stop with matched behavioural support.
• A semaglutide maintenance-dose trial. A semaglutide equivalent of SURMOUNT-MAINTAIN.
• Behavioural-infrastructure data. Randomised efficacy data on behavioural support specifically in the post-pharmacotherapy population.

If you spot an outdated claim or a missing source, email [email protected].

Editorial status: This article is awaiting medical-reviewer signoff. It describes published protocols and current obesity-medicine practice patterns; it does not constitute medical advice and is not a treatment regimen. The dose schedules described are practice patterns from clinical commentary and clinic protocols, not FDA-label recommendations. See our methodology and disclaimer.