Tirzepatide is a single molecule sold under two brand names. As Mounjaro it carries an FDA approval for type 2 diabetes, granted May 13, 2022. As Zepbound it carries a separate approval for chronic weight management, granted November 8, 2023, and a third approval for moderate-to-severe obstructive sleep apnoea in adults with obesity, granted December 20, 2024. The drug is a dual agonist: it activates both the GIP receptor and the GLP-1 receptor. Semaglutide, the comparator most people reach for, hits GLP-1 alone.
Two phase 3 programs sit behind those approvals. SURPASS studied glycaemic control in type 2 diabetes. SURMOUNT studied weight and, later, sleep apnoea. What follows is a report of what those trials measured and found, with the numbers attributed. None of it is medical or dosing advice.
SURPASS: the diabetes program
The headline comparison came from SURPASS-2, an open-label trial that ran tirzepatide head-to-head against injectable semaglutide 1 mg, both added to metformin. Published in the New England Journal of Medicine in 2021 (PMID 34170647), it randomised 1,879 adults with a baseline HbA1c of 8.28% across four arms over 40 weeks.
The primary endpoint was change in HbA1c at week 40. Tirzepatide beat semaglutide at all three doses:
| Treatment | HbA1c change | Weight difference vs semaglutide |
|---|---|---|
| Tirzepatide 5 mg | -2.01 pp | -1.9 kg |
| Tirzepatide 10 mg | -2.24 pp | -3.6 kg |
| Tirzepatide 15 mg | -2.30 pp | -5.5 kg |
| Semaglutide 1 mg | -1.86 pp | reference |
The differences in HbA1c were statistically significant (P<0.001 for the 10 mg and 15 mg doses; P=0.02 for 5 mg). One caveat the trial design forces on any reading of these results: SURPASS-2 was open-label, so participants and investigators knew which drug was being given.
The monotherapy arm of the program, SURPASS-1, was published in The Lancet the same year. Against placebo in drug-naive patients, tirzepatide cut HbA1c by up to 2.07 percentage points from a baseline near 7.9%.
SURMOUNT: weight, and then sleep apnoea
SURMOUNT-1 tested tirzepatide in people with obesity but without diabetes. It was a double-blind, placebo-controlled trial of 2,539 adults run over 72 weeks, published in NEJM in 2022 (PMID 35658024). The co-primary endpoints were percentage change in body weight and the share of participants reaching at least 5% weight loss.
Mean weight change at week 72:
- Tirzepatide 5 mg: -15.0%
- Tirzepatide 10 mg: -19.5%
- Tirzepatide 15 mg: -20.9%
- Placebo: -3.1%
All comparisons against placebo were significant at P<0.001. The press figure of "up to 22.5%" that circulated at publication came from a secondary analysis (treatment-regimen estimand) rather than the efficacy estimate quoted above; both numbers appear in the paper, and they answer slightly different questions about adherence.
SURMOUNT-OSA, published in NEJM in 2024 (DOI 10.1056/NEJMoa2404881), extended the program to obstructive sleep apnoea. It was actually two double-blind trials: one in people not using PAP therapy, one in people already on it. The primary endpoint was change in the apnoea-hypopnoea index (AHI), the count of breathing interruptions per hour of sleep.
- Trial 1 (no PAP): AHI fell 25.3/hour on tirzepatide vs 5.3/hour on placebo at week 52.
- Trial 2 (on PAP): AHI fell 29.3/hour vs 5.5/hour.
Both differences were significant (P<0.001). This is the evidence the FDA cited for the December 2024 sleep-apnoea indication, which made Zepbound the first drug approved for OSA.
Safety and limits
The adverse-event profile across the program was dominated by gastrointestinal effects, mostly mild to moderate and concentrated during dose escalation. In SURPASS-2, nausea ran roughly 17-22% on tirzepatide versus 18% on semaglutide, with diarrhoea and vomiting in similar ranges. In SURMOUNT-1, treatment discontinuation for adverse events climbed with dose: 4.3% at 5 mg, 7.1% at 10 mg, 6.2% at 15 mg, against 2.6% on placebo.
A few things the trials do not establish. SURMOUNT-1 and SURMOUNT-OSA ran to 72 and 52 weeks; they are not lifetime data, and weight regain after stopping has been documented in follow-on work. SURPASS-2's open-label design and its single semaglutide dose (1 mg, not the 2 mg later approved) limit how far the head-to-head should be pushed. And the cardiovascular-outcomes trial, SURPASS-CVOT, reads out separately from the glycaemic and weight endpoints above.
A note on what is being sold
Everything here describes branded, FDA-approved tirzepatide tested in regulated trials. That is not what circulates in the grey market. Compounded and research-labelled tirzepatide is a different supply chain, and independent testing keeps finding the gap. One public dataset of vendor samples (Finnrick) reports identity and purity that often clear 98% but quantity that can diverge from the label by large margins, with occasional failed identity flags. Material sold for research use is not approved for human consumption, and none of the trial figures above transfer to an unverified vial. The evidence summarised here is the published clinical record, not a recommendation.