Why weight returns after stopping: the withdrawal trials and adaptive thermogenesis

Two trials were built to answer one question: what happens when you take the drug away? They agree with each other, and with thirty years of physiology that came before them.

The withdrawal trials

SURMOUNT-4. 783 adults with obesity took tirzepatide for a 36-week open-label lead-in and lost a mean 20.9%. They were then randomised to continue tirzepatide or switch to placebo for 52 weeks. The tirzepatide group lost a further 5.5%. The placebo group regained 14.0%. At week 88 the continuing group was 25.3% below baseline and the withdrawn group 9.9% below (PubMed 38078870). Roughly two thirds of the lead-in loss came back within a year, and the curve had not flattened at the end.

STEP 4. 902 adults took semaglutide 2.4 mg for a 20-week run-in and lost 10.6%. Randomised to continue or switch to placebo for 48 more weeks, the continuing group lost another 7.9% and the placebo group regained 6.9% (PubMed 33755728). Cardiometabolic markers that had improved during the run-in moved back toward baseline in the placebo arm.

STEP 1 extension. Participants from STEP 1 were followed for a year after both semaglutide and the lifestyle programme stopped. They regained two thirds of the weight they had lost, and the improvements in blood pressure, lipids and HbA1c largely reverted (PubMed 35441470).

The physiology underneath

None of this was surprising to people who study weight regulation. The body defends its weight through several mechanisms that switch on with weight loss from any cause and do not switch off on a schedule.

Adaptive thermogenesis. In 1995 Leibel, Rosenbaum and Hirsch showed that after a 10% weight loss, total energy expenditure fell by more than the change in body size predicted; the body was running cheaper than its new weight should require (PubMed 7632212). Rosenbaum and Leibel's later review put the persistent deficit at roughly 10 to 15% below predicted (PubMed 20935667). Fothergill and colleagues found the effect still present six years after the large losses of The Biggest Loser contestants, with resting metabolic rate about 500 kcal per day below what their body composition predicted (PubMed 27136388).

Hormonal adaptation. Sumithran and colleagues measured appetite hormones one year after a 10-week very-low-energy diet. Leptin was still lower and ghrelin still higher than baseline, and participants still reported more hunger, a full year on (PubMed 22029981). The signals that drive eating do not return to baseline when the weight stabilises; they stay tilted toward regain.

A GLP-1 medicine works by holding intake down against these signals. Remove it and the signals are still there, now unopposed. That is the whole explanation, and it is why the regain curves look like regain after any other method of losing weight, only steeper because the losses were larger.

What is not known

Whether GLP-1 medicines themselves worsen adaptive thermogenesis, leave it unchanged or blunt it is not established. The trials measured weight, not energy expenditure. Whether tapering rather than stopping abruptly slows regain has not been tested in a randomised trial as of the date on this page. What replaces the drug matters: structured support in STEP 1 ended at the same time as the drug, which the extension authors note as a limitation.

Where to go next

• Maintenance planning after stopping is covered practically at Weight Loss on GLP-1s.
• Both trials are digested with confidence intervals at FormBlends Research and STEP 4.
• The lean mass page covers what kind of weight comes back.