No metformin. No background drug to lean on. 478 people with early type 2 diabetes, randomized to tirzepatide or placebo — nothing else.
SURPASS-1 was the first of five global registration trials that supported tirzepatide's approval for type 2 diabetes, and it was also the simplest by design. It enrolled 478 adults whose diabetes was inadequately controlled by diet and exercise alone — just over half (54.2%) had never taken a diabetes medication at all. Average diabetes duration was a short 4.7 years, baseline HbA1c was 7.9%, and average body weight at enrollment was 85.9kg.
Participants were randomized to once-weekly tirzepatide (Mounjaro) at 5mg, 10mg, or 15mg, or to a matching placebo injection, for 40 weeks. There was no metformin, no sulfonylurea, no other background therapy running underneath — tirzepatide, or nothing, is what separates SURPASS-1 from most of the other SURPASS trials, which tested tirzepatide as an add-on.
That makes it the cleanest read on the drug by itself. Head-to-head or add-on trials always carry some interaction with the background medication. A monotherapy-versus-placebo design strips that out, at the cost of only being ethically feasible in people whose diabetes is still mild enough that withholding other treatment for 40 weeks is reasonable.
The trial ran under ClinicalTrials.gov identifier NCT03954834, sponsored by Eli Lilly, and results were published in The Lancet in July 2021 and presented at the American Diabetes Association's 81st Scientific Sessions.
Every dose beat placebo by a wide margin, and the gap didn't need a second drug to show up. The highest dose dropped HbA1c by just over two full points while the placebo group actually drifted slightly upward — expected, since diabetes tends to progress without treatment. By the trial's efficacy-estimand analysis, up to 92% of people on tirzepatide reached the standard 7% HbA1c target, versus 1 in 5 on placebo.
More than a third of participants left the trial without diabetes, by lab numbers. Between 31% and 52% of people on tirzepatide reached an HbA1c under 5.7% — the threshold used to define normal glucose metabolism — compared with essentially none (1%) on placebo. That's a striking result for a monotherapy trial in people who, on average, had been diagnosed less than five years earlier.
Weight loss tracked with dose, up to about 11% of body weight. The 15mg arm lost an average of 9.5kg (about 21 lb), versus a negligible 0.7kg on placebo. Lilly also reported favorable secondary changes in lipids at the top dose — total cholesterol down 8.4%, triglycerides down 21.0%, LDL down 12.4%, HDL up 7.5% — though these were exploratory endpoints, not the trial's primary focus.
No severe hypoglycemia was reported in any tirzepatide arm. That matters for a monotherapy trial specifically, since drugs that lower blood sugar independent of meals or insulin dosing can, in some classes, push glucose too low. Tirzepatide's mechanism is glucose-dependent, and the trial's safety data are consistent with that. Gastrointestinal side effects — nausea, diarrhea, vomiting, constipation — were the most common adverse events and mostly mild to moderate, concentrated in the dose-escalation weeks. Discontinuation specifically due to adverse events stayed under 7% in every tirzepatide arm.
As a standalone therapy with no other diabetes drug in the background, tirzepatide cut HbA1c by up to 2.07 points and body weight by up to 9.5kg over 40 weeks — versus almost no change on placebo — with no severe hypoglycemia reported in any dose group.
Sources: Rosenstock J, Wysham C, Frías JP, et al. Efficacy and safety of a novel dual GIP and GLP-1 receptor agonist tirzepatide in patients with type 2 diabetes (SURPASS-1): a double-blind, randomised, phase 3 trial. The Lancet, 2021;398(10295):143-155 (doi:10.1016/S0140-6736(21)01324-6) · Eli Lilly, SURPASS-1 topline results release, June 26 2021. The Lancet full text · ClinicalTrials.gov NCT03954834.
This trial describes the drug, not a combination. Most people who end up on tirzepatide are already on metformin or another agent, so SURPASS-1's numbers won't map exactly onto every real-world case. But it's the clearest evidence for what tirzepatide itself is doing, separate from any interaction with a background drug.
Higher discontinuation at 15mg wasn't mostly about side effects. Overall dropout was highest in the 15mg arm (21.5%) and lowest at 5mg (9.1%), but Lilly's release notes most of the 15mg and placebo discontinuations were for reasons unrelated to adverse events — pandemic-era disruptions, work, and family circumstances were cited. AE-specific discontinuation stayed under 7% at every dose.
A short diabetes duration may be doing some of the work. Participants averaged 4.7 years since diagnosis, and over half had never taken a diabetes medication. Beta-cell function tends to hold up better earlier in the disease, which may partly explain the high rates of near-normal glucose control. Someone 15 years into type 2 diabetes shouldn't expect identical numbers.
Mounjaro is the brand tested here. Tirzepatide is approved for type 2 diabetes under the Mounjaro name and for weight management under the Zepbound name; SURPASS-1 was conducted in the diabetes population and used to support the diabetes indication specifically.
This article summarizes a published, peer-reviewed clinical trial for general education. It is not medical advice. Individual results vary, and any decision about starting, switching, adjusting, or stopping any GLP-1 or GIP/GLP-1 medication should be made with a prescriber.