A Pilot Trial Tracked Weight Regain After Stopping GLP-1

October 8, 2026 · Science · Nutrition

What the Trial Reported

A pilot randomized controlled trial indicates that structured post-treatment support may reduce weight regain after stopping GLP-1 receptor agonists, but the trial did not set out to prove weight loss maintenance. Across 39 adults observed for four months from the day they discontinued medication, participants assigned to usual care regained an estimated 10.75% of their body weight. Participants receiving a paid Noom subscription regained an estimated 3.60%, while participants receiving ten weekly home-delivered medically tailored meals regained an estimated 3.58%.

Weight change was a secondary outcome in this study rather than the primary target. The two primary outcomes, intervention adherence and intervention satisfaction, did not differ between the three study groups. Reading the weight numbers as proof that one support works passes over the null result on the questions the investigators set out to test.

The findings appeared in Obesity Science & Practice under the title 'Preventing Weight Regain After GLP-1 Therapy Discontinuation: A Pilot Randomized Controlled Trial.' Kelseanna Hollis-Hansen, PhD, MPH, of the Tufts University Food is Medicine Institute served as corresponding author. The University of Texas Southwestern Medical Center served as the lead study sponsor, as detailed in the ClinicalTrials.gov record NCT06273163. A summary report appeared on MedicalXpress.

The Primary Endpoints Came Back Flat

The study was designed to evaluate whether patients would accept and stick with these support programs. Investigators tracked intervention adherence at four months using the 6-item Perceived Behavior subscale of the validated Treatment Adherence Perception Questionnaire (TAPQ). They evaluated intervention satisfaction at four months using the Abbreviated Treatment Satisfaction Questionnaire for Medication (TSQM-9), adapted for behavioral programs.

Neither primary measure showed a statistical difference across the three study arms. Participants reported similar satisfaction and adherence levels whether they received meal boxes, logged meals in an application, or continued with usual care. Secondary measures for diet quality, assessed by the Healthy Eating Index-2015 on a scale from 0 to 100, and perceived health also showed no difference between arms. Some trend estimates favored the digital application, but the study was not designed to confirm those differences.

When a clinical trial misses its primary endpoints, secondary findings require cautious interpretation. A secondary outcome like weight change cannot substitute for the original study goals. Because the trial did not establish differences on its primary feasibility milestones, the numerical weight gap remains a preliminary observation.

What a Dozen People Per Arm Can Show

The trial enrolled 39 participants across three study groups, which leaves roughly a dozen individuals in each arm. The study team and authors explicitly describe the project as a pilot study with preliminary results. They evaluated whether the protocol could run smoothly before committing resources to a larger investigation.

Pilot studies are designed to test operational procedures rather than establish therapeutic efficacy. The registration record NCT06273163 does not document statistical power calculations for detecting weight differences between groups. In an arm with twelve participants, an unusual weight change in one or two individuals substantially moves the group average.

The study ran at a single medical center in Dallas. All participants spoke English. Those conditions bound who the result describes, and they do not extend to other care settings.

What the Two Active Arms Were

The two active interventions tested distinct lifestyle support models against usual care. The meal delivery group received 10 medically tailored meals per week, totaling 40 meals per month for four months, formulated for weight wellness. The digital wellness group received a four-month paid subscription to Noom, an application that pairs food and exercise logging with psychological principles to guide habit formation.

The two active groups produced nearly identical four-month weight trajectories. Participants using Noom regained an estimated 3.60% of their body weight, while those receiving tailored meals regained 3.58%. That 0.02 percentage-point difference is indistinguishable in cohorts of this size. In our comparison of Omada vs Noom, we look at how behavioral applications structure daily tracking, which requires sustained personal engagement compared to pre-portioned meal deliveries.

The comparator arm received usual care, which meant medical follow-up at the discretion of their individual healthcare provider without structured lifestyle interventions. Triple masking was planned for care providers, investigators, and outcome assessors. Complete masking was impossible in practice because study participants knew whether they were eating meals, tracking in an app, or receiving nothing.

Who This Pilot Describes

Strict enrollment rules define the exact group reflected in these findings. Eligible participants were adults aged 18 and older who had achieved greater than 10% weight loss while taking a GLP-1 receptor agonist. Every participant enrolled within 30 days of stopping their GLP-1 medication.

The trial excluded anyone with major psychiatric illness, substance misuse, or medical conditions that prevented eating standard study meals. Individuals with active eating disorders or those requiring liquid diets were also barred from entry. Anyone who had used Noom or a medically tailored meal program in the prior 12 months was excluded.

This pilot describes adults who lost more than 10% of their body weight on a GLP-1 and stopped within 30 days. It does not describe individuals actively taking GLP-1 medications, those with less than 10% weight loss, or patients managing an eating disorder. Patients who discontinued treatment months ago or who have non-English language requirements also fall outside these results.

What a Four-Month Regain Number Leaves Out

Four months of observation cannot reveal long-term weight maintenance. A four-month regain figure does not show what happens over one or two years. A short pilot cannot establish whether meal delivery or digital logging prevents weight regain over extended periods.

The trial tracked weight using Bluetooth scales and baseline electronic health record (EHR) data, but collected no data on body composition changes. This study did not evaluate whether regained weight consisted of skeletal muscle or adipose tissue.

Metabolic markers were not reported in this pilot. Our nutrition hub collects the ground this pilot leaves open, including a weight loss plateau and the hardest age to lose weight. Our research methodology sets out how we weigh a pilot result.

What Would Change This Read

A trial powered for weight as the primary endpoint, run across several centers with longer follow-up, would change this read. A definitive study would need to pre-specify weight as the primary outcome rather than a secondary measure. It would require hundreds of participants to ensure adequate statistical power across diverse backgrounds.

A head-to-head of the meals against the app, over longer follow-up, would show whether the two actually diverge. Testing participants across varied clinics outside Dallas would confirm whether these behavioral supports produce dependable results.

Anyone planning to stop should review clinical options with their prescribing clinician, because stopping a GLP-1 is a prescribing decision.

Sources

• "Preventing Weight Regain After GLP-1 Therapy Discontinuation: A Pilot Randomized Controlled Trial," Obesity Science & Practice, 2026, DOI 10.1002/osp4.70194; corresponding author Kelseanna Hollis-Hansen, PhD, MPH, Tufts University Food is Medicine Institute. Reported by MedicalXpress, October 2026 (accessed October 8, 2026). Source of the pilot randomized design and its three arms, the 39 participants, the four-month duration from GLP-1 cessation, the four-month regain figures of 10.75%, 3.60% and 3.58%, the finding that diet quality, perceived health, adherence and satisfaction did not differ significantly by group with some estimates trending toward the application, and the authors' description of the results as preliminary.
• Trial registration NCT06273163, "Identifying Strategies to Curtail Weight Regain After GLP-1 Receptor Agonist Treatment Cessation," on ClinicalTrials.gov (accessed October 8, 2026). Source of the lead sponsor, the University of Texas Southwestern Medical Center, the actual enrolment of 39, the randomized parallel design and its triple-masking record with the note that complete masking was not feasible, the March 2024 to August 2025 dates, the single Dallas location, the two primary outcomes (adherence by TAPQ and satisfaction by TSQM-9) and the two secondary outcomes (percent weight loss maintained and diet quality by the Healthy Eating Index-2015), the arm descriptions of 10 medically tailored meals per week and a paid Noom subscription against usual care, and the inclusion and exclusion criteria.