The 30-second summary
- A pilot trial published on 7 October 2026 in Obesity Science and Practice followed 39 adults for four months after they stopped a GLP-1.
- With usual care, people gained an estimated 10.75% of their body weight. With a Noom subscription it was 3.60%, and with 10 dietitian-designed meals a week it was 3.58%.
- Both kinds of support beat usual care by about 7 percentage points. The app and the meals did about equally well.
- It is a small, early study. Weight was not its main outcome, and it stopped before reaching its planned size.
What did the trial test?
Most people who start a GLP-1 stop at some point, for cost, side effects, insurance or choice. The trial asked a practical question: if you give people structured help on the day they stop, does less weight come back?
Researchers at UT Southwestern Medical Center in Dallas ran it, led by Kelseanna Hollis-Hansen, who is now at the Tufts University Food is Medicine Institute. Adults who had lost more than 10% of their body weight on a GLP-1 and had stopped within the past 30 days were randomly assigned to one of three groups (trial record NCT06273163):
| Group | What they got for four months | People |
|---|---|---|
| Usual care | Care from their own provider | 14 |
| Wellness app | A paid Noom subscription | 12 |
| Medically tailored meals | 10 delivered meals a week | 13 |
The app arm used Noom, which tracks food, exercise and weight and teaches behaviour change. Participants were asked to use it for at least 15 minutes a day, according to coverage of the study in Newsweek. The meals arm picked 10 meals a week from a list of 70. The dishes were designed by dietitians and had fewer calories, less saturated fat and the same or less sugar than conventional prepared meals.
Weight was measured every month on Bluetooth scales at home. The average starting BMI was 31.4.
What did it find?
Over four months, the usual care group gained an estimated 10.75% of body weight. The app group gained 3.60% and the meals group 3.58%.
Compared with usual care, the app group gained 7.16 percentage points less (95% confidence interval 2.04 to 12.27) and the meals group 7.17 points less (2.14 to 12.20). The intervals are wide, which is what you expect with 12 to 14 people per group, but neither crosses zero.
The trial's main outcomes were adherence and satisfaction, and those did not differ significantly between groups. Diet quality and how healthy people felt did not differ significantly either, though the authors note some estimates leaned towards the app.
Why did the meals and the app work about the same?
The study cannot say. One reading is that both did the same job: they gave structure at the moment the appetite came back. A delivered meal decides portion and content for you. An app makes you look at what you eat and weigh yourself. Either way, the first weeks off the drug are no longer unplanned.
What is notable is that diet quality scores did not separate the groups. So the benefit may have come from structure and routine rather than from eating measurably better food.
What does this study not tell us?
Several things, and they matter.
- It is small. The trial planned 60 people and enrolled 39. The registry lists it as terminated because funding ended and recruitment fell short.
- Weight was exploratory. The trial was built to test whether people would use and like the support, not to prove a weight effect.
- Four months is short. Coverage in FoodNavigator noted there is no data on whether the lower regain would last a full year.
- Nobody was blinded. People knew what they were getting, and that alone can change behaviour.
- It does not say which drug or dose people were on, at least in the published abstract.
For scale: in the STEP 1 trial extension, people who had lost 17.3% of their weight on semaglutide regained 11.6 percentage points of it within a year of stopping, about two-thirds (Wilding et al., 2022). The new pilot shows early regain can be slowed. It does not show it can be stopped.
What does it mean if you take a GLP-1?
If you are still on treatment, nothing changes today. If stopping is on the horizon, this is the first randomised evidence, as of October 2026, that support started on the day you stop may matter. Our guide to what happens when you stop a GLP-1 covers the larger maintenance trials.
The practical lesson is about timing. The support in this trial did not start months later, when the scale had already moved. It started at cessation, when appetite returns but habits are still in place. Three things are worth having ready before that day:
- Meals you do not have to think about. Whether delivered, batch-cooked or a short list of repeat dinners, the meals arm suggests decided food helps when hunger comes back.
- A way to see what you eat and weigh. The app arm logged food and weight. A weekly weigh-in and a simple food log do part of the same job.
- A protein and fibre floor. These are what keep meals filling once the drug is no longer doing that work. Protein on a GLP-1 explains how much and why.
Talk to your prescriber before stopping, and ask what support is available. The trial suggests that what happens in the first four months is worth planning, not leaving to chance. As the trials behind these drugs showed from the start, the medicine was never tested on its own.
Sources
- Hollis-Hansen K, Vonderschmidt A, Haskins C, Kane RM, Almandoz JP, Lingvay I. Preventing Weight Regain After Glucagon-Like Peptide-1 Therapy Discontinuation: A Pilot Randomized Controlled Trial. Obesity Science and Practice 2026. DOI
- Identifying Strategies to Curtail Weight Regain After GLP-1 Receptor Agonist Treatment Cessation. ClinicalTrials.gov NCT06273163. ClinicalTrials.gov
- Wilding JPH, Batterham RL, Davies M, et al. Weight regain and cardiometabolic effects after withdrawal of semaglutide: The STEP 1 trial extension. Diabetes Obes Metab 2022. PubMed