One-year semaglutide follow-up shows why maintenance matters

The best-established evidence is that stopping an effective obesity medication can be followed by meaningful weight recurrence. In the STEP 1 extension, adults who had received semaglutide 2.4 mg for 68 weeks had lost an average of 17.3% of body weight at week 68. During the following year off treatment, they regained an average of 11.6 percentage points of body weight, about two-thirds of their prior loss. The average net loss from the original baseline at week 120 was 5.6%.

This was an exploratory off-treatment extension involving a selected subset, including 232 participants exposed to semaglutide in the extension analysis. It describes an average group pattern, not a prediction that every person will regain the same amount, or regain at the same pace.

Alongside weight regain, most measured cardiometabolic improvements moved back toward baseline after semaglutide withdrawal. The extension was not designed to determine whether these changes lead to differences in clinical events.

A four-month pilot tested support after GLP-1 discontinuation

A 2026 exploratory pilot randomized controlled trial examined whether two structured approaches might help in the early period after GLP-1-based therapy ends. The fully remote study enrolled 39 adults who had lost more than 10% of body weight while using a GLP-1-based medication and had stopped it within the prior 30 days.

Participants had previously used semaglutide, tirzepatide, or cagrilintide plus semaglutide. They were randomized to:

• Usual care, n=14: Care was left to the referring provider, so its content and intensity could vary.
• A digital wellness application, n=12: The program included self-monitoring of diet, exercise, and weight, plus educational content and behavioral strategies.
• Medically tailored meals, n=13: Participants selected 10 dietitian-developed, portion-controlled weight-wellness meals per week, delivered to their homes.

Follow-up lasted four months. The study's main emphasis was feasibility, adherence, and satisfaction. Weight was an exploratory outcome, which is important when interpreting the results.

Estimated regain was lower in both structured-support groups

Over four months, the study's mixed-effects model estimated 10.75% weight gain with usual care, compared with 3.60% with the digital wellness application and 3.58% with medically tailored meals. Compared with usual care, the estimated differences were -7.16 percentage points for the digital intervention and -7.17 percentage points for medically tailored meals.

The model also estimated slower monthly weight-gain slopes: 2.94% in usual care, 0.42% with the digital application, and 0.61% with medically tailored meals.

These figures are model-estimated changes in body weight, not pounds, and they should not be read as proof that either intervention prevents GLP-1 weight regain. The study was small, only 34 of 39 participants completed a final study weigh-in, confidence intervals were wide, and follow-up ended after four months. It also cannot show whether differences would persist after the intervention support ended.

The two intervention groups had similar estimated results, but this pilot was not designed to establish whether one was better than the other. It tested specific structured programs, not the broad idea that any wellness app, meal delivery service, or general advice to eat differently will produce the same result.

Why regain can happen after medication ends

Weight regain after weight loss occurs in the context of chronic, relapsing obesity and is not evidence of weak willpower. Prior research on the weight-reduced state describes increased appetite, changes in appetite-related signals, reduced energy requirements, and changes in thermogenesis or energy expenditure that can oppose maintenance of a lower body weight.

These are established areas of weight-loss physiology, but they were not measured in the 2026 discontinuation pilot. The trial did not measure appetite hormones, energy expenditure, resting metabolic rate, body composition, fat mass, or lean mass. Therefore, it cannot establish why the digital and meal groups had lower estimated regain, or whether a specific biological mechanism explained the difference.

What may help, and what remains uncertain

The pilot adds early evidence that post-discontinuation support is a researchable target. It does not yet identify one reliable strategy for everyone. Approaches being studied or considered include structured nutrition, exercise, behavioral support, digital tools, continued or changed pharmacotherapy, and possibly tapering. Better, larger, and longer trials are needed before any of these can be presented as a universally effective GLP-1 weight-maintenance plan.

Exercise is one relevant, but separate, area of emerging evidence. In a different post-treatment analysis, participants who had previously received exercise plus liraglutide had more favorable weight and body-fat outcomes one year after treatment termination than those who had received liraglutide alone. That study had a different medication, population, prior weight-loss protocol, and design, so it does not prove that exercise alone prevents GLP-1-associated regain or validate the interventions in the 2026 pilot.

Likewise, the pilot does not test a tapering protocol, a maintenance dose, restarting medication, or switching to another medication. There is not enough evidence here to recommend a particular way to stop treatment.

Plan discontinuation with the prescribing clinician

Current guidance generally views obesity pharmacotherapy as potentially long-term treatment. The American Diabetes Association notes that medications should generally be continued after treatment goals when appropriate, because discontinuation often leads to weight recurrence and worsening or reemergence of cardiometabolic risk factors.

That does not mean every person should remain on medication indefinitely. Adverse effects, pregnancy, medical conditions, cost, access, personal preferences, and inadequate response can all affect treatment decisions. Medication discontinuation, dose reduction, switching, or tapering should be discussed with the prescribing clinician, particularly when the medication is also used to treat diabetes or another condition.

For now, the practical conclusion is measured: GLP-1 weight regain is common on average after treatment stops, while the best prevention strategy remains unresolved. The 2026 pilot justifies larger and longer research into structured support, rather than proving that an app or meal program can prevent rebound.