What Happens When People Stop GLP-1s — GLP-1 Data Series
Page 4 of 5

What Happens When People Stop GLP-1s

Weight regain data from every trial withdrawal study and real-world dataset, plus metabolic parameter rebound. The ~67% regain from RCTs is a floor, not a ceiling — real-world evidence points to 74–100% regain at 12–18 months for patients who truly discontinue.

Last updated: April 2026

The question the persistence data forces

The persistence data shows that the vast majority of patients stop GLP-1 therapy within three years. The question this page addresses is what happens next.

The answer from clinical trial withdrawal studies is that approximately two-thirds of weight lost is regained within one year of stopping — a figure remarkably consistent across semaglutide, tirzepatide, and liraglutide. But this ~67% figure, derived from ideal conditions with selected responders, structured withdrawal, and ongoing monitoring, is almost certainly a floor, not a ceiling.

Real-world data paint a more complex picture. The largest observational dataset (Optum, n=18,228) shows 74% of lost weight regained at one year. The most granular study (Twin Cities, n=130) found that 49% of patients who had lost weight ended up heavier than their pre-treatment baseline. A 2026 BMJ meta-analysis projects complete return to baseline weight within 18 months. One notable exception — Cleveland Clinic's 0.5% regain — appears driven by the fact that over half the cohort transitioned to alternative obesity treatment after discontinuation, demonstrating that treatment continuity can meaningfully attenuate regain.

The combination of discontinuation rates and regain rates means that the durable weight-loss benefit at the population level may be substantially smaller than headline trial numbers suggest.

What RCT withdrawal studies showed

Every published withdrawal study shows substantial regain. The ~67% figure replicates across different drugs and treatment durations — but these trials measured regain under ideal conditions that systematically favor lower estimates.

Study Drug N Tx duration Off-tx follow-up Wt loss on tx % of loss regained Net from baseline
STEP 1 Extension
Wilding et al., Diabetes Obes Metab, 2022
Semaglutide 2.4 mg
Injection
22868 wk52 wk−17.3%~67%−5.6%
STEP 4
Rubino et al., JAMA, 2021
Semaglutide 2.4 mg
Injection
26820 wk run-in48 wk−10.6%~65%~−3.7%
STEP 10
McGowan et al., Lancet D&E, 2024
Semaglutide 2.4 mg
Injection
~13852 wk28 wk−13.9%~43%−7.9%
SURMOUNT-4
Aronne et al., JAMA, 2024
Tirzepatide 10/15 mg
Injection
33536 wk lead-in52 wk−20.9%~67%−9.9%
SURMOUNT-CN
Chen et al., Life Metabolism, 2025
Tirzepatide 10/15 mg
Injection
15252 wk26 wk−17.8 to −22.9%~51–54%−8.7 to −10.6%
Critical caveat

Weight regain was still continuing at the final study visit in the STEP 1 Extension — the trajectory had not plateaued at 52 weeks off-treatment. A SURMOUNT-4 post-hoc showed 82.5% of withdrawn participants regained ≥25% of their loss, and 24% regained ≥75%. The 67% figure at 12 months is a snapshot of an ongoing process, not an endpoint.

What real-world data actually show

Multiple independent real-world datasets converge: regain in clinical practice exceeds the RCT benchmark of ~67%. The most likely figure is 74–100% regain at 12–18 months for patients who truly discontinue without alternative treatment.

SourceNSettingRegain at ~1 yearKey finding
West et al.
BMJ, 2026
9,34137-study meta-analysis100% projected at 18 moSema/tirz discontinuers regain 0.8 kg/month (9.9 kg in first year). Projected full return to baseline at 1.5 years.
Weintraub et al.
Obesity Week, 2025
18,228Optum EHR-claims~74% of loss regainedProgressive regain (+4.5% at 3 mo, +7.5% at 12 mo). No plateau evident. 58% regained at least some weight.
Abdel-Bary et al.
Obesity Medicine, 2025
130Allina Health (Twin Cities)65.4% gained weight49% of patients who had lost weight exceeded their pre-treatment baseline. Mean loss at discontinuation was only −2.26%.
Shah et al.
Diabetes Obes Metab, 2025
289,000+Narrative review synthesis60–90% regainGlycemic, BP, and lipid parameters return to baseline within 12–18 months.
Gasoyan et al.
Diabetes Obes Metab, 2026
7,938Cleveland Clinic EHR0.5%~55% received alternative treatment after stopping. See Cleveland Clinic analysis below.
The population-level math

If 50–68% of patients discontinue within a year (see persistence data), and most regain 74–100% of lost weight within 18 months, the durable weight-loss benefit at the population level may be substantially smaller than headline trial numbers suggest.

Rate and trajectory of regain

The RCT trajectory shows the V-shape under ideal conditions. The real-world trajectory — shown below it — is steeper and does not plateau within the measurement window.

~23 weeks
Half-life of regain (RCT model) — 50% of max regain by ~5.5 months
0.8 kg/mo
Real-world regain rate for sema/tirz discontinuers (BMJ, 2026)
18 months
Projected full return to baseline (BMJ meta-analysis)
49%
Of patients exceeded pre-treatment weight at 1 year (Twin Cities)
RCT Weight Trajectory: On Treatment → Off Treatment
Based on STEP 1 Extension (68 wk on → 52 wk off). Selected responders, structured withdrawal, ongoing monitoring.
Real-World Trajectory: What Actually Happens
Based on Optum (n=18,228) and BMJ projection. Less weight lost, faster regain, return to baseline by ~18 months. No plateau.

The two charts illustrate a crucial divergence. The RCT trajectory shows patients losing 17% and retaining −5.6% at one year off treatment. The real-world trajectory shows patients losing ~10%, regaining faster without monitoring, and projecting back to baseline within 18 months. The exponential model from RCTs may describe the shape of regain, but the magnitude is worse in practice.

Why RCTs systematically underestimate real-world regain

The gap is not random noise. Seven structural features of withdrawal studies all push in the same direction — toward making RCTs look more favorable than reality.

1. Run-in enrichment eliminates the worst cases before measurement
SURMOUNT-4 used a 36-week open-label lead-in; only those who completed it (85.6%) entered the withdrawal phase. Non-responders (10–17%), those who stopped for side effects (28%), and those who stopped for cost (13%) are invisible in RCT regain data.
2. Follow-up captures only partial regain
Weight was still rising at the final visit in STEP 1 Extension. The BMJ meta-analysis projects 100% return at 18 months — the 67% at 12 months is a midpoint, not an endpoint.
3. Completion bias compounds enrichment bias
The STEP 1 Extension analyzed only 327 of 1,961 originally randomized participants — those from the highest-recruiting sites who completed 68 weeks and agreed to extended follow-up. Weight losses in this subset were "slightly greater" than the full population.
4. Monitoring attenuates regain even in the "off-treatment" group
Participants still attended study visits and were weighed regularly. A meta-analysis of the Hawthorne effect found an OR of 1.41 for behavior change from being observed. Real-world patients who stop typically have no ongoing monitoring.
5. Lifestyle support continues in trials but not in practice
STEP 4 maintained structured diet and exercise counseling throughout the withdrawal phase, yet still saw ~65% regain. Only 34% of employers covering GLP-1s require any lifestyle program participation (KFF 2025).
6. Abrupt, chaotic real-world discontinuation differs from structured withdrawal
BCBS data show >40% of real-world users stop after just 4 weeks. Insurance denials, shortages, cost burden, and stop-start patterns from high deductibles create chaotic trajectories unlike clean trial withdrawal.
7. Real-world patients lost less weight to begin with
Cleveland Clinic data show real-world weight loss of 6.8–11.9% vs. 15–21% in RCTs. With less lost, even modest regain pushes patients past baseline. The Twin Cities mean loss was only 2.26% at discontinuation.

How Cleveland Clinic achieved only 0.5% regain

One study reported substantially less regain than other datasets. The difference appears to reflect what happened after patients stopped their initial GLP-1 — and offers a model for how health systems can mitigate regain through structured transitional care.

Gasoyan et al. (Cleveland Clinic, 2026): 0.5% regain at 1 year

Among 7,938 patients who discontinued injectable semaglutide or tirzepatide, the obesity cohort regained only 0.5% at one year. The key to understanding this result is that a majority of patients transitioned to alternative treatments rather than simply stopping all therapy.

19.6%
Restarted their original GLP-1
27.4%
Switched to a different medication
13.7%
Continued structured lifestyle visits
~55%
Total receiving continued treatment

Gasoyan acknowledged that many patients "restart the medication or transition to another obesity treatment, which may explain why they regain less weight than patients in randomized trials." The Cleveland Clinic data demonstrates that treatment continuity — medication switching, lifestyle visits, GLP-1 restart — can meaningfully attenuate regain. The study did not report weight outcomes separately for the ~45% who received no further treatment, so the regain rate for true discontinuers remains unknown. Most employer benefit designs do not currently replicate this level of transitional care, but the data suggests they could benefit from doing so.

Metabolic parameter rebound

Metabolic improvements do not all reverse at the same rate. Blood pressure reverts fastest, glycemic markers track weight, and lipid/inflammatory improvements show partial durability.

Rebounds fastest (weeks)

Blood pressure: +4.15 mmHg SBP (meta-analysis); +6.8 to +10.4 mmHg (SURMOUNT-4)
~70–80% of the initial SBP reduction regained within 12 weeks. Heart rate decreased −3.22 bpm post-cessation (normalization of GLP-1-induced tachycardia, not deterioration).
Tzang et al., eClinicalMedicine, 2025; Horn et al., JAMA Intern Med, 2025.

Tracks weight regain (months)

HbA1c: +0.25% (obesity); +0.65% (T2D)
Dose-response with regain: those regaining <25% saw only +0.14% HbA1c vs. +0.35% for ≥75% regain. Waist circumference: +0.8 cm (<25% regain) to +14.7 cm (≥75% regain).

Most durable improvements

HDL cholesterol persisted across all regain categories
The most durable metabolic benefit observed. CRP also remained improved at 1 year off semaglutide. LDL, VLDL, triglycerides remained below baseline — likely reflecting residual body composition changes.
Evidence gap

Liver fat/MASH: No major RCT has reported liver fat data after GLP-1 cessation. ESSENCE demonstrated MASH resolution on treatment but had no off-treatment extension.

Cardiovascular protection and discontinuation

SELECT established a 20% MACE reduction with continuous semaglutide. Observational data now provide initial evidence on what happens to cardiovascular protection after stopping.

Al-Aly et al., BMJ Medicine, March 2026

Target trial emulation: 333,687 U.S. veterans with T2D (132,551 GLP-1 users vs. 201,136 sulfonylurea users).

−18%
MACE reduction with continuous 3-yr use
+4–8%
MACE risk increase after stopping 6 mo
+14%
After stopping 1 year
+22%
After stopping 2 years

GLP-1 use for less than 18 months followed by discontinuation showed no significant residual CV protection. Interrupted-and-resumed use yielded only a 12% risk reduction vs. 18% with continuous therapy. The investigators described this pattern as "metabolic whiplash." These findings suggest that cardiovascular benefits may require sustained treatment to maintain, though RCT-level evidence on post-discontinuation CV outcomes remains absent.

Regain is similar across drugs when scaled to initial loss

More potent agents cause more absolute regain, but proportional regain converges around ~65–67% in RCTs. Real-world data push this higher.

% of Weight Loss Regained After Stopping, by Study
Orange = RCT data. Red = real-world data. Dashed line = 100% (full return to baseline).

Predictors of regain

Greater initial loss → more absolute regain, but more residual benefit
STEP 1 Extension: those who lost ≥20% retained −12.1% at 1 year off treatment. Those who lost <5% showed +4.2% above baseline. But in real-world populations where many lose only 2–10%, even modest regain pushes past baseline.
Exercise: 7.2× odds of maintaining ≥10% weight loss after stopping
S-LiTE (n=195, 4-arm RCT): combined exercise + liraglutide led to 6.0 kg less regain than liraglutide alone. Exercise alone maintained weight one year after stopping. The strongest evidence for any behavioral intervention.
Lundgren et al., NEJM, 2021; Jensen et al., eClinicalMedicine, 2024.
T2D patients show less regain; non-T2D patients gain weight at higher rates
70.8% of non-T2D patients gained weight vs. 41.7% of T2D patients (P=.007). T2D patients also show higher reinitiation rates, likely driven by glycemic deterioration.
Treatment duration does not reduce proportional regain
"Rapid regain occurs regardless of the duration of treatment." No validated biomarkers predict individual susceptibility.
Quarenghi et al., J Clin Med, 2025.

The chronic disease argument

The consensus is remarkably uniform. The AMA, Obesity Society, World Obesity Federation, Endocrine Society, AACE/ACE, AAP, EASO, AGA, and WHO all classify obesity as a chronic, relapsing disease requiring chronic management. The World Obesity Federation states that "short-term treatments do not change the underlying biology."

Dissenting positions exist. The AMA's own Council on Science recommended against the disease classification in 2013 — the House of Delegates overrode this. Academic critics argue the model promotes pharmaceutical over structural solutions. A 2024 JAMA Viewpoint proposed "short-term, intermittent GLP-1 therapy paired with sustained lifestyle interventions."

The gap between model and practice

If the medical consensus is indefinite pharmacotherapy, and most patients stop within 3 years, there is a significant gap between the clinical model and real-world usage patterns. The regain data helps explain why this gap matters: without ongoing treatment or structured transitional care, most of the clinical benefit appears to be temporary.

Over one-third restart within a year

47.3%
Of T2D patients reinitiated within 1 year
36.3%
Of non-T2D patients reinitiated within 1 year
2.5%
Higher restart likelihood per 1% weight gain

Outcomes on reinitiation remain a critical evidence gap. No study reports whether patients achieve equivalent weight loss on a second course. The Al-Aly data suggest interrupted use yields inferior CV protection (12% vs. 18%), arguing against cycling. The pattern — treatment, discontinuation, regain, reinitiation — creates a recurring cost cycle that complicates actuarial modeling.

Source: Rodriguez et al., JAMA Network Open, 2025 (125,474 adults, Truveta). DOI: 10.1001/jamanetworkopen.2024.57349

What happens to your population?

Start with a cohort on GLP-1 therapy. Apply persistence and regain rates. See how many maintain ≥5% weight loss over time. Default uses real-world regain assumptions (75%).

Interactive model

GLP-1 population outcome model

Connects the persistence data with the regain data from this page.

Population & assumptions
Employees starting GLP-1 therapy100
Average weight loss among those on treatment12%
RCT ITT: 15–21%. Real-world average: 6–12%.
% of weight regained after stopping (at 1 year)75%
RCT benchmark: ~67%. Real-world (Optum): ~74%. BMJ 18-mo projection: ~100%.
Still on treatment at 12 mo
At ≥5% weight loss at 12 mo
At ≥5% weight loss at 36 mo
Simplified model. Does not account for reinitiation (36–47% restart within a year) or switching to alternative treatments, which would improve outcomes. Real populations show heterogeneous responses.

Sources

Wilding JPH, et al. "Weight regain and cardiometabolic effects after withdrawal of semaglutide (STEP 1 Extension)." Diabetes Obes Metab. 2022;24(8):1553–1564. DOI: 10.1111/dom.14725
Rubino D, et al. "Effect of Continued Weekly Subcutaneous Semaglutide vs Placebo on Weight Loss Maintenance (STEP 4)." JAMA. 2021;325(14):1414–1425. DOI: 10.1001/jama.2021.3224
McGowan BM, et al. "Semaglutide treatment effect in people with obesity (STEP 10)." Lancet Diabetes Endocrinol. 2024;12(9):631–642. DOI: 10.1016/S2213-8587(24)00182-7
Aronne LJ, et al. "Continued Treatment With Tirzepatide for Maintenance of Weight Reduction (SURMOUNT-4)." JAMA. 2024;331(1):38–48. DOI: 10.1001/jama.2023.24945
Horn DB, et al. "SURMOUNT-4 post hoc: Weight regain and cardiometabolic changes after tirzepatide withdrawal." JAMA Intern Med. 2025. DOI: 10.1001/jamainternmed.2025.6112
Chen Y, et al. "Follow-up after tirzepatide discontinuation (SURMOUNT-CN)." Life Metabolism. 2025;4(5):loaf024. DOI: 10.1093/lifemeta/loaf024
West HW, et al. "Weight regain after anti-obesity medication discontinuation: systematic review and meta-analysis." BMJ. 2026. DOI: 10.1136/bmj-2025-085304
Weintraub M, et al. (NYU Langone). "Weight regain after GLP-1 RA discontinuation in a real-world cohort." Presented at Obesity Week 2025. (Optum Market Clarity data, n=18,228) obesityweek.org
Abdel-Bary R, et al. "Weight outcomes after GLP-1 RA discontinuation in clinical practice." Obesity Medicine. 2025. (Allina Health/Twin Cities, n=130) PII: S2451847625000788
Shah A, et al. "Weight regain and metabolic changes after GLP-1 receptor agonist discontinuation: a narrative review." Diabetes, Obesity and Metabolism. 2025. (Synthesis of 289,000+ patients) dom-pubs.onlinelibrary.wiley.com
Gasoyan H, et al. "Real-world weight change after injectable GLP-1 RA discontinuation." Diabetes Obes Metab. 2026. DOI: 10.1111/dom.70660
Xie Y, Choi T, Al-Aly Z. "Cardiovascular outcomes after GLP-1 receptor agonist discontinuation." BMJ Medicine. Mar 2026. DOI: 10.1136/bmjmed-2025-002150
Rodriguez PJ, et al. "Reinitiation of GLP-1 RA therapy after discontinuation." JAMA Network Open. 2025;8(1):e2457349. DOI: 10.1001/jamanetworkopen.2024.57349
Budini M, et al. "Nonlinear meta-regression of weight regain after GLP-1 RA discontinuation." eClinicalMedicine. Mar 2026. DOI: 10.1101/2025.06.09.25328726
Tzang BS, et al. "Meta-analysis of weight and metabolic changes after GLP-1 RA discontinuation." eClinicalMedicine. 2025. DOI: 10.1016/S2589-5370(25)00614-5
Lundgren JR, et al. "Healthy Weight Loss Maintenance with Exercise, Liraglutide, or Both Combined (S-LiTE)." NEJM. 2021;384:1719–1730. DOI: 10.1056/NEJMoa2028198
Jensen SBK, et al. "Long-term follow-up of the S-LiTE trial." eClinicalMedicine. 2024;69:102475. DOI: 10.1016/j.eclinm.2024.102475
Pi-Sunyer X, et al. "A Randomized, Controlled Trial of 3.0 mg of Liraglutide in Weight Management (SCALE)." NEJM. 2015;373:11–22. DOI: 10.1056/NEJMoa1411892
Lincoff AM, et al. "Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes (SELECT)." NEJM. 2023;389:2221–2232. DOI: 10.1056/NEJMoa2307563
Quarenghi M, et al. "Weight regain after anti-obesity medication discontinuation: a narrative review." J Clin Med. 2025;14(11):3791. DOI: 10.3390/jcm14113791

Questions about this data? Corrections or updates?

Get in touch

Compiled and maintained by Ray Wu, MD — physician-founder working on metabolic health technology.

Last updated: April 2026

← Back to GLP-1 Data Series