Skip to content
GLP-1 stopping trials

Stopping a GLP-1: What the Trials Show

Three big trials that tested stopping semaglutide or tirzepatide, side by side, with what came back and how fast.

  1. 1On average, half to two-thirds of the lost weight came back within a year of stopping.1,2,3
  2. 2Blood pressure, blood sugar and waist size rose again as the weight came back.1,2,3,4
  3. 3A year after stopping, average weight was still below where it began; a review projects full regain at 1.5 years.1,3,5
Read the short answer

The short answer

Semaglutide is the drug in Wegovy and Ozempic; tirzepatide is the drug in Zepbound and Mounjaro.6,7,8,9 Three big trials tested stopping them.1,2,3 On average, half to two-thirds of the lost weight came back within a year.1,2,3 Blood pressure, blood sugar and waist size went back up too.1,2,3,4 Still, a year after stopping, the groups weighed less than before treatment.1,3 For these two drugs, an Oxford review projects weight back where it began about 1.5 years after stopping.5

Three stopping trials at a glance1,2,3
STEP 1 extension (semaglutide)
On the drug
68 weeks, lost 17.3%
Then off it
52 weeks, no drug
Share of loss back
About two-thirds
STEP 4 (semaglutide)
On the drug
20 weeks, lost 10.6%
Then off it
48 weeks on placebo
Share of loss back
More than half (our math)
SURMOUNT-4 (tirzepatide)
On the drug
36 weeks, lost 20.9%
Then off it
52 weeks on placebo
Share of loss back
About half (our math)

Placebo means a dummy shot with no drug in it. Weight lost is the group average. Two-thirds is the STEP 1 paper's own figure. For STEP 4 and SURMOUNT-4, the share is our math from each paper's averages.

1 in 6

people switched to placebo in SURMOUNT-4 kept at least 80% of their weight loss a year later3

9 in 10 who stayed on tirzepatide

The full PDF, free.

  • Three big stopping trials in one table, on the same measures, from weight lost to blood pressure.
  • A timeline of what happened after stopping, from the first weeks to the 1.5-year mark.
  • Who regained most, and what came back with the weight.
  • How many people stop: 157,822 first-time users in Denmark, tracked for a year.

A source on every line. No paid placements. No doses.

Instant download. You also get the Wednesday email, which you can leave on the first click. Privacy

The fair line
All three trials stopped the drug at once, and STEP 1 ended the diet and exercise program too. None tested a slow taper, so these numbers can't say what tapering does.1,2,3

This guide was last checked on 28 October 2026. Research and educational information only. Nothing here is medical advice.

  1. [1] Wilding et al., Diabetes Obes Metab, 2022 (STEP 1 extension) (opens in new tab). Checked 28 Oct 2026Back to text (first mention of source 1)
  2. [2] Rubino et al., JAMA, April 2021 (STEP 4 trial) (opens in new tab). Checked 28 Oct 2026Back to text (first mention of source 2)
  3. [3] Aronne et al., JAMA, January 2024 (SURMOUNT-4 trial) (opens in new tab). Checked 28 Oct 2026Back to text (first mention of source 3)
  4. [4] Horn et al., JAMA Intern Med, 24 Nov 2025 (study of SURMOUNT-4's placebo group) (opens in new tab). Checked 28 Oct 2026Back to text (first mention of source 4)
  5. [5] West et al., The BMJ, 7 Jan 2026 (Oxford review of 37 studies) (opens in new tab). Checked 28 Oct 2026Back to text (first mention of source 5)
  6. [6] Zepbound label (Eli Lilly, revised Aug 2026), section 14 (opens in new tab). Checked 28 Oct 2026Back to text (first mention of source 6)
  7. [7] Wegovy label (Novo Nordisk, revised Jun 2026), sections 12.3 and 14 (opens in new tab). Checked 28 Oct 2026Back to text (first mention of source 7)
  8. [8] Ozempic label (Novo Nordisk, revised May 2026) (opens in new tab). Checked 28 Oct 2026Back to text (first mention of source 8)
  9. [9] Mounjaro label (Eli Lilly, revised Aug 2026) (opens in new tab). Checked 28 Oct 2026Back to text (first mention of source 9)
  10. [10] Mailhac et al., JAMA Network Open, 28 Aug 2026 (Denmark) (opens in new tab). Checked 28 Oct 2026