Stopping semaglutide: what STEP 4 measured and what the labels require

Most of the semaglutide evidence is about starting. One trial is about stopping, and both labels have rules about pausing. This page covers those and nothing speculative.

STEP 4: the withdrawal trial

902 adults with obesity (or overweight plus a weight-related condition), without diabetes, took semaglutide for 20 weeks: 16 weeks of escalation and 4 at 2.4 mg. Mean weight loss over that run-in was 10.6%. The 803 who reached 2.4 mg were then randomised 2:1 to continue semaglutide or switch to placebo for 48 more weeks, both with lifestyle intervention.

From week 20 to week 68, the continued group lost a further 7.9%. The placebo group regained 6.9%. The difference was 14.8 percentage points (95% CI 13.5 to 16.0). Waist circumference (9.7 cm), systolic blood pressure (3.9 mm Hg) and physical-functioning score also diverged in favour of continuing. Gastrointestinal events were reported by 49.1% of those continuing and 26.1% on placebo; discontinuation for adverse events was similar, 2.4% versus 2.2% (PubMed 33755728).

Three things the trial shows and one it does not. It shows that the drug's effect depends on the drug being present. It shows that regain begins within the first year off treatment. It shows that the regain over 48 weeks was, on average, smaller than the loss over 20 weeks on drug (6.9 points regained against 10.6 lost), so at week 68 the placebo group was still below its starting weight. It does not show where weight settles beyond 68 weeks, because follow-up ended there.

Label rules that pause treatment

Before pregnancy. Both injection labels say to discontinue semaglutide at least 2 months before a planned pregnancy because of the long washout period. Wegovy's weight and cardiovascular indications also say to stop when pregnancy is recognised.

Before anaesthesia or deep sedation. Both labels (section 5.10, added to Ozempic in January 2025) warn of retained gastric contents during procedures and say to tell the team about planned procedures. They do not give a fixed number of days to stop; they say to follow fasting recommendations and that temporarily discontinuing could reduce the incidence of retained gastric contents. The FormBlends Guides site covers surgery and anaesthesia guidance in depth.

Gastrointestinal illness. The labels' dehydration and acute kidney injury warnings, and the pancreatitis warning, both direct you to your prescriber rather than to a dose.

Restarting

The Wegovy label's rule for two or more consecutive missed doses is the closest thing to a restart protocol: resume dosing as scheduled or, if needed, reinitiate Wegovy and follow the dosage escalation schedule, which may reduce gastrointestinal reactions on reinitiation. The judgment about which applies belongs to the prescriber. Ozempic's label handles a single missed dose (5 days) and is silent on longer gaps; the same logic about gastrointestinal tolerability applies and the same person decides. The missed-dose helper applies the written rules and stops where they stop.

Maintenance without the drug

The trials that inform this are STEP 4 for semaglutide and SURMOUNT-4 for tirzepatide, and both point the same way. What is done with that information, protein, resistance training, a lower maintenance dose, or a planned continuation, is covered on the Weight Loss on GLP-1s site, which has a maintenance planner built on those two trials. Nothing in STEP 4 involved compounded semaglutide, which is not FDA approved; the regain data are for the brand product at 2.4 mg.