What Happens When People Stop Weight-Loss Medications

Stopping Weight-Loss Medications: How Fast Weight Returns, and What Happens to Blood Sugar and Blood Pressure

Will the weight come back?

If I stop taking it, will the weight come back — and what happens to the improved blood sugar, blood pressure, and cholesterol that came with treatment? For many people that is not hypothetical: a 2026 narrative review reports that up to 65% of real-world GLP-1 users discontinue within a year.

The pooled evidence answers the direction plainly: weight comes back. It also puts numbers on the pace, and it is specific about where measurement ends and projection begins.

The measured pace of regain

The central synthesis is a systematic review and meta-analysis in the BMJ (January 2026, Oxford researchers; PROSPERO CRD42024532069). It pooled 37 studies — 63 intervention arms, 9,341 participants — spanning older and newer agents: orlistat, naltrexone-bupropion, liraglutide, semaglutide, tirzepatide. Mean treatment lasted 39 weeks; mean follow-up after stopping, 32 weeks.

Over that window, people who stopped any weight-loss medication regained about 0.4 kg per month (95% CI 0.3–0.5). In randomised trials, regain ran about 0.3 kg (about 0.7 lb) per month faster than in control-assigned participants (95% CI 0.3–0.4, P<0.001). The review graded its certainty in that rate estimate as moderate.

For semaglutide and tirzepatide — newer incretin drugs the review describes as more effective — the measured rate was twice as fast: 0.8 kg per month (95% CI 0.7–0.9) across 10 arms and 1,776 participants, who had lost a mean 14.7 kg (about 32 lb) (11.1–18.4) by the time they stopped.

Everything beyond about a year is a projection, and the label matters: only one study in the pooled review followed participants more than a year after stopping, and the semaglutide/tirzepatide subset had at most 12 months off-treatment. The review’s model projects first-year regain of about 4.8 kg (about 10.6 lb) for any medication (3.6–6.0), 6.0 kg (about 13.2 lb) for incretins overall (4.8–8.4), and 9.9 kg (about 21.8 lb) for semaglutide or tirzepatide (8.4–10.8) — against that 14.7 kg mean loss, a projection that most of the lost weight returns within the first year. The review’s models also project return toward pre-treatment weight about 1.7 years after stopping any medication (95% CI 1.3–2.1), about 1.5 years after semaglutide or tirzepatide (1.0–1.9). The extrapolations assume regain stays linear — no departure from linearity appeared within the observed window — but nothing beyond 12 months has been measured for the newer drugs.

The cardiometabolic improvements recede in step: pooled marker trajectories project fasting glucose, systolic blood pressure, cholesterol, and triglycerides back to baseline within about a year, and HbA1c and diastolic blood pressure within about 1.4 years (95% CI 0.4–2.4) — projections too, and a few markers had not fully reversed even at one year in the semaglutide withdrawal study.

For pace context — an indirect comparison from separate review populations — after behavioural weight-management programmes regain ran about 0.1 kg (about 0.2 lb) per month after a mean 5.1 kg (about 11.2 lb) loss (return to baseline projected near four years, 3.9; 95% CI 2.8–4.9), while regain after stopping medication ran about 0.3 kg per month faster (0.22–0.34) independent of initial loss.

What the withdrawal trials show

Two randomised withdrawal trials give the clearest drug-by-drug pictures.

The STEP 1 extension (Diabetes, Obesity and Metabolism, 2022) followed 327 adults for 52 weeks after 68 weeks of semaglutide or placebo. Semaglutide participants averaged 105.6 kg (about 233 lb) at the trial’s start, 87.5 kg (about 193 lb) at treatment’s end, 99.0 kg (about 218 lb) a year later. In percentage points of body weight, the semaglutide arm regained 11.6 (SD 7.7) versus 1.9 (SD 4.8) on placebo — two-thirds of what the drug had taken off. But the arm still finished 5.6% below its starting weight (SD 8.9), and 48.2% of participants (95 of 197) remained at least 5% below baseline, versus 22.6% (21 of 93) on placebo. The authors call the extension exploratory: no treatment arm remained after week 68, the analysis is descriptive with observed data, the subset came from the highest-recruiting sites (possible selection bias), and 9.2%/6.1% used other obesity pharmacotherapy during follow-up.

SURMOUNT-4 (JAMA, 2024) randomised at a different point: after a 36-week open-label tirzepatide lead-in with a mean 20.9% loss, 670 of 783 enrolled were assigned to continue the drug or switch to placebo for 52 weeks. Those who continued lost a further 5.5%; those who switched regained 14.0% — a 19.4-percentage-point difference (95% CI −21.2 to −17.7, P<0.001). At week 88, 89.5% of those who continued had kept at least 80% of their lead-in loss; 16.6% of switchers had. The switchers' figure needs care: +14.0% is measured from a treated baseline, after a 20.9% loss — regain of lost weight, not weight above where they began; from the original baseline, switchers still ended 9.9% below it. STEP 4 (JAMA, 2021), a semaglutide continuation trial, shows the same shape: −7.9% with continued treatment versus +6.9% after switching to placebo over 48 weeks (−14.8 percentage points, 95% CI −16.0 to −13.5), after a run-in that had produced −10.6%.

What predicts how much comes back

The clearest predictor is how much was lost: regain runs roughly proportional to it in the STEP 1 extension’s subgroup data, in the BMJ review’s sensitivity analyses at fixed 5, 10, and 15 kg (about 33 lb) losses, and in an independent meta-analysis of eight trials (Berg et al., 2,372 participants). Larger losses mean larger regain but also a larger net loss retained — “more regained” is not “worse off than before treatment.”

Timing has been quantified too. A 2025 network meta-analysis (Wu et al., BMC Medicine) found weight already significantly higher than untreated controls by 8 weeks after discontinuation (+1.50 kg (about 3.3 lb), 95% CI 1.32–1.68), widening to +2.50 kg (about 5.5 lb) by 20 weeks (2.27–2.73). A post hoc analysis of SURMOUNT-4 (Horn et al., JAMA Internal Medicine, 2026) found most participants had regained at least a quarter of the lost weight within a year, with cardiometabolic measures climbing in step with the regain fraction.

What we don’t know

The 12-month ceiling. Every figure past one year off-treatment — the 1.5-to-2-year returns to baseline, the first-year totals — is a model projection; the newer drugs have no measured data beyond 12 months.
Who was studied. Adults with overweight or obesity without type 2 diabetes, roughly two-thirds women, baseline weights near 105–107 kg (about 236 lb); liraglutide-era agents dominate the “any medication” estimate. How the averages apply to current clinical populations, or to oral and lower-dose use, is uncertain.
Prevention. The BMJ review found no evidence that post-cessation behavioural support slows regain, and none that support intensity during treatment matters. Those are absence-of-evidence findings from few trials — not precise nulls, not proof that support cannot help. No strategy has been shown to prevent regain.
Mechanism hints. In a small Oxford experimental-medicine trial (Moolla et al., n=29), a liraglutide arm and a lifestyle arm lost matched weight; during withdrawal, only the drug arm showed elevated circulating markers (MMP-10, IL-10RB, FGF-23, Flt3L) and dysregulated adipose gene expression — a hypothesis about why drug-associated regain runs faster, not a demonstrated mechanism. (Cited at abstract level; full text not reviewed.)
A correction, reviewed. The Wu network meta-analysis has a published erratum (BMC Medicine, 2026;24:169); read in full, it adds multilevel-model and sensitivity analyses and corrects a study-count label, but does not alter the timing figures cited above.
Sponsorship. Both withdrawal RCTs were designed and run by the drugs’ manufacturers — Novo Nordisk (STEP 1 extension) and Eli Lilly (SURMOUNT-4) — and frame obesity as a chronic disease requiring ongoing treatment, aligned with the sponsors’ commercial interests. The measurements, however, converge with independent analyses. The BMJ review is publicly funded (NIHR, Wellcome) and declares a Novo Nordisk Foundation grant held by its senior author.

What this means

The averages describe populations, not individuals. On the pooled evidence, regain begins within weeks to months of the last dose and runs at roughly 0.4 kg (about 0.9 lb) per month — about twice that for semaglutide or tirzepatide in the measured window — with most of the lost weight projected to return within about a year and the blood-sugar, blood-pressure, and lipid improvements receding alongside. Two qualifications: regain is proportional to loss, so people who lost more regain more but also keep more; and partial retention is the norm — a net 5.6% below baseline in the semaglutide withdrawal study, a year after stopping. Nothing here shows stopping is futile, and nothing identifies a proven way to prevent regain. Nor is it a recommendation about anyone’s treatment: the projections, populations, and sponsor-shaped trial conclusions argue for reading these numbers as context, not personal prediction.

Questions a reader could bring to a clinician — as questions, not a checklist of actions: what trajectory would you expect for someone with my history, and how would we monitor it? What would you check — weight, labs, blood pressure — and how often around any treatment change? If weight returns, what would we discuss then? Where does the evidence on support after stopping run out?

Any decision to start, stop, or change one of these medications belongs to a prescriber–patient conversation; the numbers here describe what happened, on average, in studies — not what will happen to any one person.

Sources

West S, Scragg J, Aveyard P, et al. “Weight regain after cessation of medication for weight management: systematic review and meta-analysis.” BMJ 2026;392:e085304. doi.org/10.1136/bmj-2025-085304 (DOI 10.1136/bmj-2025-085304) · free full text, PMC12776922 · PMID 41500720 · PROSPERO CRD42024532069.
Wilding JPH, Batterham RL, Calanna S, et al. “Weight regain and cardiometabolic effects after withdrawal of semaglutide: The STEP 1 trial extension.” Diabetes, Obesity and Metabolism 2022;24(8):1553–1564. doi.org/10.1111/dom.14725 (DOI 10.1111/dom.14725) · free full text, PMC9542252 · PMID 35441470 · parent-trial registration NCT03548935 (the extension has no separate registry record).
Aronne LJ, Sattar N, Horn DB, Bays HE, Wharton S, et al. “Continued Treatment With Tirzepatide for Maintenance of Weight Reduction in Adults With Obesity: The SURMOUNT-4 Randomized Clinical Trial.” JAMA 2024;331(1):38–48. doi.org/10.1001/jama.2023.24945 (DOI 10.1001/jama.2023.24945) · PMID 38078870 · NCT04660643.
Moolla A, Poolman T, Othonos N, et al. “Randomised trial comparing weight loss through lifestyle and GLP-1 receptor agonist therapy in people with MASLD.” JHEP Reports 2025;7(5):101363. doi.org/10.1016/j.jhepr.2025.101363 (DOI 10.1016/j.jhepr.2025.101363) · PMC12060445 · PMID 40342635.
Wu et al. BMC Medicine 2025;23:398. doi.org/10.1186/s12916-025-04200-0 (DOI 10.1186/s12916-025-04200-0) · PMID 40691794. Erratum: BMC Medicine 2026;24:169 (DOI 10.1186/s12916-026-04817-9) — scope not reviewed for this draft.
Berg et al. Obesity Reviews 2025;26(8):e13929. doi.org/10.1111/obr.13929 (DOI 10.1111/obr.13929) · PMID 40186344.
Rubino et al. (STEP 4). JAMA 2021;325(14):1414–1425. doi.org/10.1001/jama.2021.3224 (DOI 10.1001/jama.2021.3224) · PMID 33755728.
Horn et al. JAMA Internal Medicine 2026;186(2):157–167. doi.org/10.1001/jamainternmed.2025.6112 (DOI 10.1001/jamainternmed.2025.6112) · PMID 41284285.
Moiz et al. eClinicalMedicine 2026;96:103992. doi.org/10.1016/j.eclinm.2026.103992 (DOI 10.1016/j.eclinm.2026.103992) · PMID 42256676.

Context sources identified but not cited (no claim rests on them): Cochrane living reviews of tirzepatide (CD016018) and semaglutide (CD015092.pub2), flagged in the packet as challenge/benchmark checks and not read.

SEO suggestions (unreviewed)

Title: Stopping weight-loss medications: how fast weight returns, and what happens to blood sugar and blood pressure
Description: Pooled evidence from 37 studies and two withdrawal trials: about 0.4 kg regained per month after stopping (0.8 kg after semaglutide or tirzepatide), most lost weight typically returning within about a year, with cardiometabolic gains fading alongside — plus what remains projected, not measured. About 0.4 kg (about 0.9 lb) regained per month after stopping weight-loss medications — 0.8 kg (about 1.8 lb) after semaglutide or tirzepatide — and cardiometabolic gains fade too.