The SURMOUNT-MAINTAIN results give a clear answer: reducing tirzepatide to 5 mg kept more weight off than stopping it. But some weight came back. People who continued their higher dose held onto more of their earlier loss than either group.
Lower dosing is a maintenance option for a patient and prescriber to consider. The group average cannot predict one person’s result.
What question did this tirzepatide step-down trial ask?
The earlier SURMOUNT-4 trial compared continuing tirzepatide with stopping. SURMOUNT-MAINTAIN added a middle choice: reducing the dose after successful weight loss.
As of October 1, 2026, the day this article was written, Horn and colleagues’ Lancet report showed how those three choices compared. It was published online May 12, 2026. Tirzepatide is the medicine in Zepbound and Mounjaro.
Who joined SURMOUNT-MAINTAIN, and for how long?
The trial enrolled 441 adults across 20 US sites. Their average age was about 47. Participants had a BMI, a weight measure adjusted for height, of at least 30. A BMI of at least 27 also qualified with a weight-related condition, such as high blood pressure. People with diabetes were excluded.
Everyone first received tirzepatide for 60 weeks. Doses rose toward their maximum tolerated dose: the highest dose they could handle. That was either 10 or 15 mg weekly. The comparison required tolerating at least 10 mg and losing at least 5% of starting weight.
Then 378 people were randomized, meaning assigned by chance, for another 52 weeks. Of those, 140 continued 10 or 15 mg and 144 switched to 5 mg. The remaining 94 received placebo, an inactive look-alike shot. Neither participants nor study staff knew the assignment. Diet and activity counseling continued. The registry record, NCT06047548, counts everyone enrolled; the paper’s 378 counts those assigned at the switch.
What did the tirzepatide maintenance dose study measure?
The published paper’s primary endpoint, its main outcome, was percentage weight change from the original starting weight to week 112. That includes both the initial weight loss and the maintenance year.
Its primary analysis used a “modified treatment-regimen estimand.” An estimand defines the precise question an analysis answers, including how it handles treatment changes. This analysis included people who stopped their assigned medicine.
Rescue treatment, extra medication after substantial regain, became available from week 84 if at least half the lost weight returned. Measurements after rescue were excluded. The analysis carried forward the worst weight result recorded before rescue, preventing renewed weight loss on rescue treatment from improving the assigned group’s result.
SURMOUNT-MAINTAIN results: what stayed off?
At week 112, the 5 mg group remained 16.6% below its original starting weight, compared with 9.9% below on placebo. Continuing 10 or 15 mg preserved more: 21.9% below the starting weight.
These averages use the primary analysis:
| Weight change | Continued 10 or 15 mg | Reduced to 5 mg | Switched to placebo |
|---|---|---|---|
| From the original start to week 112 | −21.9% | −16.6% | −9.9% |
| From the week-60 switch to week 112 | −0.2% | +7.0% | +15.2% |
Rescue treatment was started by 8% of the higher-dose group, 25% of the 5 mg group and 67% of the placebo group.
The second row explains the tradeoff. The lower-dose group regained weight equal to 7.0% of its weight at the switch. That is different from regaining 7% of the pounds already lost.
Among participants whose weight had leveled off before the switch, the share of earlier loss kept off averaged 96.5%, 67.9% and 42.8%, respectively. These percentages describe the amount of loss retained, not the percentage of people who maintained it. The tirzepatide plateau guide explains why weight can level off during treatment.
In a separate analysis estimating outcomes with continued assigned treatment, waist size changed by −0.6 cm, +3.8 cm and +9.3 cm after the switch. Blood pressure, blood sugar and cholesterol were also tracked. Those measurements cannot establish whether stepping down prevents heart attacks.
What side effects did the paper report?
An adverse event is a health problem recorded during a trial, whether or not the medicine caused it. The maintenance-year safety groups included 139, 142 and 91 people who received at least one assigned dose.
| During the maintenance year | Continued 10 or 15 mg | Reduced to 5 mg | Placebo |
|---|---|---|---|
| Any adverse event | 56% | 50% | 45% |
| Nausea | 6% | 4% | 2% |
| Vomiting | 6% | 1% | 0% |
| Diarrhea | 7% | 5% | 1% |
| Constipation | 1% | 4% | 4% |
| Serious adverse events | 0% | 1% | 1% |
| Stopped treatment because of an adverse event or death | 0% | 1% | 0% |
The paper’s safety table excludes events after rescue treatment. Symptoms earlier, while doses increased, were more common. These later counts describe people who had already tolerated higher doses; they cannot settle rare or long-term risks.
What did commentators and newer analyses add?
Scheen’s accompanying Lancet comment frames weight regain as a challenge for long-term treatment. As of October 1, 2026, the day this article was written, an August 20 European Heart Journal commentary described maintenance as more than a continue-or-stop choice. Lower dosing preserved some benefit; continued higher dosing preserved more.
As of October 1, 2026, the day this article was written, a Lilly-sponsored analysis published August 30 added detail about the 5 mg group. In a subset of 132 participants, 69, or 52%, kept at least three-quarters of their earlier weight loss. This was a post hoc analysis, a later look at existing trial data. It did not show in advance who would succeed after reducing the dose.
As of October 1, 2026, the day this article was written, an August 31 pooled analysis combined older semaglutide and tirzepatide withdrawal studies. It studied stopping, rather than reducing the dose. Its forecasts beyond one year were projections, not measured follow-up.
Lilly funded SURMOUNT-MAINTAIN and participated in its design and analysis. The trial did not formally compare the two active doses statistically. Its selected participants and one-year maintenance period limit what it can say about everyone starting treatment.
How does semaglutide’s STEP 4 compare?
The STEP 4 trial in JAMA 2021 also asked what happens after initial weight loss. After 20 weeks on semaglutide, 803 adults were assigned to continue 2.4 mg or switch to placebo for 48 weeks. Weight then fell another 7.9% with continued treatment and rose 6.9% with placebo.
STEP 4 had no lower-dose maintenance group. It supports the broader concern about regain after stopping, but cannot identify an equivalent semaglutide step-down dose. These separate trials also cannot rank the two medicines against each other. The STEP 4 results page explains that study’s design.
What does this mean for a prescription?
A lower maintenance dose involves a tradeoff between keeping weight off, tolerability and a person’s treatment goals. The prescriber sets the actual dose. The guide to stopping tirzepatide covers that broader decision.
SURMOUNT-MAINTAIN did not test doses below 5 mg or longer gaps between doses, the approaches discussed in microdosing tirzepatide. It cannot describe outcomes beyond week 112 or establish results for any compounded preparation.
Compounding means combining or altering ingredients to prepare a medication for a patient. Tirzepatide prescribed through Promise is dispensed by a licensed U.S. compounding pharmacy as a compounded medication, which is different from an FDA-approved product: the formulation offered here is not FDA-approved. The FDA’s compounding guidance, updated September 16, 2025, explains that these preparations are not reviewed for safety, effectiveness or quality before marketing.
A licensed provider may still prescribe — that decision is between you and your doctor. At Promise, a licensed provider reviews every request, and not everyone qualifies.
