REST trial protocol published, formalizing the taper question

Researchers have published the full protocol for the REST trial, a randomized study designed to answer a question that millions of people taking GLP-1 drugs face but that has never been tested head to head: does slowly lowering the semaglutide dose before stopping work better than simply stopping? The protocol appeared in PLOS One on July 27, 2026 ^[2].

The paper lays out the hypothesis plainly. The authors write that they expect "gradual reduction of semaglutide will be associated with less weight regain and cardiometabolic deterioration compared with immediate cessation" ^[1]. They also note that there "is little clinical data on the best strategy to achieve weight maintenance and cardiometabolic benefits following semaglutide discontinuation" ^[2]. Publishing the design ahead of results puts the taper question on the record as an open, testable one, with the outcomes and methods locked in before any data are collected.

How the trial is built

REST is an open-label, parallel-arm randomized controlled trial running 32 weeks in total: a 16-week discontinuation phase followed by a post-discontinuation follow-up period ^[2]. Participants will be randomized 1:1 into two groups. One group continues its current semaglutide dose for 16 weeks and then stops completely. The other reduces the semaglutide dose by 25% every four weeks until complete discontinuation at week 16 ^[2]. Everyone receives standard lifestyle advice, including physical activity, portion control and self-monitoring of weight ^[2].

The entry criteria are narrow. Eligible adults are ages 18 to 75 with obesity, or overweight with adiposity-related complications, who are already on weekly injectable semaglutide at a minimum dose of 1 mg, have lost at least 10% of their pre-treatment body weight, and have held that weight steady for the previous 12 weeks ^[2]. People with existing cardiovascular disease, type 2 diabetes, pregnancy, prior bariatric surgery, eating disorders, or significant kidney, liver or malignant disease are excluded ^[2].

The primary outcome is the difference between groups in percent body weight change ^[1]. Secondary outcomes include 24-hour ambulatory blood pressure and fasting ghrelin, the hormone tied to hunger signaling ^[1]. Study visits occur at baseline, 16 weeks and 32 weeks, with phone check-ins in between. Assessments include body measurements, fasting glucose, hemoglobin A1c, lipids, C-reactive protein, blood counts, electrolytes, creatinine and liver enzymes, plus eating-behavior questionnaires and an online 24-hour dietary recall tool ^[2].

The trial is registered as NCT07294950 and funded by the Heart and Stroke Foundation of Canada, which the authors state had no role in the study design or the decision to publish ^[2]. Authors include Yevusiak, Weisman, Retnakaran, Wharton, Drucker and Kramer ^[2].

Why it matters for patients

The background section of the protocol restates what the existing trial literature already shows: semaglutide produced mean weight loss of roughly 16.9% in clinical trials, along with improvements in blood pressure, glucose metabolism, inflammatory markers and lipids — but after abrupt withdrawal, about two-thirds of the weight lost is regained within one year, with the cardiometabolic gains reversing as well ^[2]. The authors attribute that rebound to compensatory changes in energy balance after weight loss, including increased appetite and reduced energy expenditure ^[2].

In practice, people stop these drugs for many reasons — cost, insurance changes, supply, side effects, or reaching a goal. Some clinicians already taper doses on the way out, but that is currently a judgment call, not an evidence-based standard. REST is designed to show whether the taper actually changes the outcome or only feels gentler. The inclusion of ambulatory blood pressure and ghrelin means the trial can look at whether a slower exit blunts the hormonal hunger rebound and the blood pressure creep, not just the number on the scale.

It is also worth noting what REST will not answer. The design tests one specific taper schedule — 25% every four weeks — in people without diabetes or established heart disease who have already lost at least 10% and stabilized ^[2]. Results will not automatically apply to tirzepatide (Mounjaro, Zepbound), to oral semaglutide (Rybelsus), or to people who stop earlier in treatment.

What happens next

The protocol was received June 22, 2026, accepted June 29, 2026 and published July 27, 2026 ^[2]. The sources do not state the target sample size, the enrollment start date, the study sites, or the expected date of results. The authors say no datasets have yet been generated or analyzed, and that all relevant data will be made available upon study completion ^[2].