Stopping a GLP-1: Can Meals or an App Prevent Regain?
A 39-person randomized pilot tested medically tailored meals and a digital coaching app in the four months after people stopped GLP-1 drugs. Both groups regained less than usual care. Here is what that does and does not prove.
Build your own printable peptide tracker
A personalized injection log, weekly dosing schedule and dated calendar — your compounds filled in, ready to print. Free to build and preview; $5/month to unlock printing.
How It Works
Participants picked 10 dietitian-designed "weight wellness" meals a week (breakfasts 130–310 kcal, lunches and dinners 160–460 kcal), delivered home for 4 months. They reported eating about 8 a week. Estimated regain: 3.58% of body weight.
A Noom subscription, prescribed at 15 minutes a day of lessons, food logging and weigh-ins. Reported use averaged 2.8 days a week. Estimated regain: 3.60% of body weight.
Whatever the referring clinic normally offered; not standardized. This group regained an estimated 2.94% of body weight per month, 10.75% over the 4 months.
In a Copenhagen RCT, people who stopped a year of supervised exercise regained 6.0 kg less over the following year than people who stopped liraglutide alone. The exercise-only group never took a GLP-1, so this compares stopping exercise with stopping liraglutide; it does not test support given around stopping a GLP-1.
What the Data Shows
Key Takeaways
- In a randomized pilot of 39 adults who had lost more than 10% on a GLP-1, usual care regained an estimated 10.75% of body weight over 4 months, vs 3.60% with a coaching app and 3.58% with medically tailored meals (PMID 42846130).
- In a sensitivity analysis adjusting for how much weight people had lost on the drug and their starting weight, both groups still regained significantly less than control (about 5.8 and 6.3 points less).
- Weight was an exploratory outcome. The primary outcomes, adherence and satisfaction, did not differ significantly between groups.
- Without support, regain is fast: one year after stopping semaglutide 2.4 mg, STEP 1 participants had regained two-thirds of their weight loss (PMID 35441470).
- A 2025 meta-analysis of 18 RCTs found people with obesity regained about 5.6 kg after stopping GLP-1 drugs, with HbA1c, waist circumference and blood pressure also worsening (PMID 41399474).
- Withdrawal trials show the same pattern: switching to placebo meant +6.9% over 48 weeks after semaglutide (PMID 33755728) and +14.0% over 52 weeks after tirzepatide (PMID 38078870).
- A 2026 BMJ meta-analysis of 37 studies found regain after weight-management drugs averaged about 0.4 kg a month, faster than after behavioural programmes (PMID 41500720).
- Whether the benefit lasts past 4 months. The trial stopped there; the authors call for at least 1 year of follow-up.
- Whether it replicates. Enrollment reached 39 of a planned 60; most were women and college graduates, and everyone who reported income had a household income of at least $30,000.
- How much of the gap reflects the groups being unequal at the start: the control arm was all women, lighter, and had lost more on the drug (21.2% vs 15.6–17.7%).
- Whether meals or the app is better. Their results were nearly identical, and the pilot was not powered to compare them.
- Whether tapering the dose instead of stopping outright changes regain. The authors note tapering trials are still under way.
- Whether regained weight is mostly fat or lean mass. Body composition was not measured.
Frequently Asked Questions
How much weight do people regain after stopping a GLP-1?
In the STEP 1 extension, people who stopped semaglutide 2.4 mg regained two-thirds of their lost weight within a year (11.6 of 17.3 percentage points). In SURMOUNT-4, switching from tirzepatide to placebo meant a 14.0% regain over 52 weeks. A 2026 BMJ meta-analysis put the average at about 0.4 kg a month.
Can diet or an app prevent weight regain after stopping semaglutide or tirzepatide?
One small randomized pilot suggests they may slow it. Over four months, usual care regained an estimated 10.75% of body weight, versus 3.60% with a Noom subscription and 3.58% with home-delivered medically tailored meals. With 39 participants and four months of follow-up, it is a signal worth testing, not proof.
What are medically tailored meals?
They are dietitian-designed, portion-controlled prepared meals delivered to the home. In this pilot, participants chose 10 "weight wellness" meals per week from more than 70 options; breakfasts were 130–310 kcal and lunches or dinners 160–460 kcal. Participants reported eating about eight meals a week on average.
Is the GLP-1 regain pilot trial strong evidence?
No. It was a pilot designed to test adherence and satisfaction, so weight was an exploratory outcome. It enrolled 39 of a planned 60 people, usual care was not standardized, the control arm differed at baseline, and follow-up lasted four months. The authors call for larger trials with at least a year of follow-up.
Does exercise help keep weight off after stopping a GLP-1?
In a Copenhagen trial, people who stopped a year of supervised exercise regained 6.0 kg less over the following year than people who stopped liraglutide alone. Those who had combined exercise with liraglutide were 5.1 kg lighter a year after stopping than the liraglutide-only group. That trial began after a diet-induced weight loss, so the setting differs.
🔬 Research-Grade Source
Swiss Chems publishes third-party HPLC COAs for selected products — vendor-published, a claim to verify. HighPeptides' primary vendor reference for peptides.
Browse Swiss Chems →Affiliate links — support HighPeptides at no extra cost.
🛒 Recommended Products
Support supplements and gear relevant to this protocol.
Affiliate links — support HighPeptides at no extra cost.
📚 Related HighPeptides Research
Peer-Reviewed References
Educational purposes only. Not medical advice.
Stopping, tapering or restarting a prescription GLP-1 medication is a decision to make with the clinician who prescribes it.
Results come from a 39-person pilot with 4 months of follow-up; treat them as preliminary.