SURPASS and SURMOUNT: what the tirzepatide trials measured

What are the SURPASS and SURMOUNT trials?

SURPASS and SURMOUNT are Eli Lilly clinical programmes that tested tirzepatide, identified in research as LY3298176. SURPASS enrolled adults with type 2 diabetes and made change in HbA1c its main endpoint. SURMOUNT enrolled adults defined by BMI and made percent change in body weight a main endpoint. The programmes share a molecule, not a research question. SURPASS-1 is registry NCT03954834, SURPASS-2 is NCT03987919, and SURMOUNT-1 is NCT04184622. These trials studied tirzepatide, not retatrutide. Readers seeking the molecule's earlier identity can see where tirzepatide came from without turning this programme guide into another origin story.

The wider SURPASS record also names SURPASS-3 as NCT03882970, SURPASS-4 as NCT03730662, SURPASS-5 as NCT04039503, and SURPASS-CVOT as NCT04255433. Those registry numbers identify studies. They do not carry outcome claims here.

The shared sponsor does not make the programmes one trial. Each paper defines its own enrolled group, comparator, endpoint, and duration. The programme label is useful only when those details stay attached to it.

What did the SURPASS trials measure?

SURPASS trials measured change in HbA1c from baseline in adults with type 2 diabetes. HbA1c is a blood measure that reflects average blood sugar over roughly the past three months. SURPASS-1 used tirzepatide alone against placebo and reported mean HbA1c changes of −1.87%, −1.89%, and −2.07% across the three trial arms, compared with +0.04% for placebo at 40 weeks (Rosenstock et al., The Lancet, 2021). SURPASS-2 added treatment to metformin and reported changes of −2.01%, −2.24%, and −2.30%, compared with −1.86% for semaglutide 1-mg at 40 weeks (Frías et al., NEJM, 2021).

SURPASS asked a blood-sugar question first. Body weight was also recorded, but it remained a secondary measurement within those glycaemia trials. SURPASS-1 reported body-weight loss ranging from 7.0 to 9.5 kg across the tirzepatide arms (Rosenstock et al., The Lancet, 2021). SURPASS-2 reported body-weight differences of −1.9, −3.6, and −5.5 kg against semaglutide (Frías et al., NEJM, 2021). Those figures do not turn SURPASS into SURMOUNT.

The distinction matters because the same percent sign can hide a different unit of thought. HbA1c is itself written as a percentage, but a change in HbA1c is reported in percentage points. A percent change in body weight describes a proportion of starting weight. The symbols may look alike while the measurements remain different.

What did the SURMOUNT trials measure?

SURMOUNT trials measured percent change in body weight and the proportion of participants reaching at least 5% weight reduction. These were coprimary endpoints, meaning two questions the trial committed to answering before it started, with both counted as main results. SURMOUNT-1 ran for 72 weeks, including a 20-week dose-escalation period (Jastreboff et al., NEJM, 2022). Mean weight change at week 72 was −15.0%, −19.5%, and −20.9% across the three tirzepatide arms, compared with −3.1% for placebo (Jastreboff et al., NEJM, 2022). The endpoint was body weight, not HbA1c.

SURMOUNT-1 also reported that 85%, 89%, and 91% of participants in the tirzepatide arms reached at least 5% weight loss, compared with 35% on placebo (Jastreboff et al., NEJM, 2022). That second main result answers a threshold question: how many people crossed a line set before the trial began? The average-change endpoint answers a group question. The threshold endpoint counts individual crossings.

Both results belong to the same body-weight programme, but they are not duplicates. One describes the size of the average change. The other describes the share of people who reached a stated threshold.

What is the difference between SURPASS and SURMOUNT?

SURPASS and SURMOUNT differ in the question, the enrolled population, and the primary endpoint, not in the molecule studied. SURPASS centred on HbA1c change for adults living with type 2 diabetes. SURMOUNT centred on body-weight change for adults selected by BMI, with diabetes included or excluded according to the individual trial. A SURPASS percentage is an HbA1c change in percentage points. A SURMOUNT percentage is a percent change in body weight. The two values are not interchangeable and cannot be averaged into one result. Readers separating brand and molecule names can also review how the three molecules differ.

The comparison table keeps the question beside the population and endpoint that produced it. That is the safest way to read a programme headline. Start with what the trial promised to measure. Then check who entered, what the comparison group received, and how long the trial ran. Only after those pieces match does a result belong in the same conversation.

Who took part in the SURMOUNT trials?

SURMOUNT trial populations differ from one study to another. SURMOUNT-1 and SURMOUNT-4 enrolled adults defined by BMI and excluded diabetes. SURMOUNT-2 enrolled adults with BMI of at least 27 and required HbA1c from 7% to 10%, so type 2 diabetes was part of its entry criteria (Garvey et al., The Lancet, 2023). The results should not be pooled as if the participants came from one population. Diabetes status, starting measurements, and trial design shape what each result describes. A programme name is not enough to erase those differences.

SURMOUNT-1 enrolled 2,539 adults; their baseline mean weight was 104.8 kg (Jastreboff et al., NEJM, 2022). In SURMOUNT-2, 938 adults were randomised with a baseline mean weight of 100.7 kg (Garvey et al., The Lancet, 2023). The SURMOUNT-4 lead-in enrolled 783 adults; mean weight at randomisation was 107.3 kg (Aronne et al., JAMA, 2024).

The inclusion rules explain why a result cannot travel freely from one trial to another. SURMOUNT-1 and SURMOUNT-4 answer questions in populations without diabetes. SURMOUNT-2 answers a related body-weight question in adults who had type 2 diabetes. Similar headings do not make those groups interchangeable.

What did SURMOUNT-4 test?

SURMOUNT-4 tested a maintenance question after an open-label lead-in, not a first-treatment question. All participants received tirzepatide during the first 36 weeks, then 670 people were randomised to continue or switch to placebo through week 88 (Aronne et al., JAMA, 2024). From week 36 to week 88, mean weight changed by −5.5% among those who continued and +14.0% among those switched to placebo (Aronne et al., JAMA, 2024). At week 88, 89.5% of the continuation group maintained at least 80% of the lead-in loss, compared with 16.6% on placebo (Aronne et al., JAMA, 2024).

The design begins with everyone walking the same path. The split comes later. That makes SURMOUNT-4 a study of what happened after an initial period, rather than a clean comparison from the starting line. Bodies keep responding to changing conditions, so the point of randomisation matters when the two groups are read.

The placebo group in the withdrawal phase was not a group that had received placebo from the first day. Its members had already completed the open-label lead-in. Reading the week 36 starting point preserves the question the trial was built to answer.

Evidence limits

What do the SURPASS and SURMOUNT trials not show?

No source here shows that research-use peptide material is a medicine, a substitute, or a route to an approved product. Research-use peptide material is not a medicine, not a substitute for one, and not for human or animal use.

Every paper cited here was funded by Eli Lilly. That is a conflict-of-interest disclosure, not a finding about the results.

The catalog page for tirzepatide reference material lists batch certificate-of-analysis and release-status paperwork. That listing does not change the limits of the clinical record.