Retatrutide: What the Evidence Actually Shows

Retatrutide is a triple agonist still in development at Eli Lilly. It switches on three receptors at once (glucagon alongside GIP and GLP-1), where tirzepatide reaches two of them and semaglutide only one. Four phase 3 papers are now published: TRIUMPH-1 and TRIUMPH-2 in obesity, TRANSCEND-T2D-1 in type 2 diabetes, and the sleep apnea results from inside the TRIUMPH trials.^{[5][6][11][12]} Lilly’s stated timetable is an FDA filing in early 2027 (Q1).^[8] It is not approved anywhere, cannot be prescribed, and is sold anyway across a gray market with no pharmacy behind it.

The numbers people repeat are 28.7% and 28.3%. Both are real, but neither is the figure the peer-reviewed paper leads with. The reason tells you how to read every number below.

Peer-reviewed figures and press-release figures

TRIUMPH-1 is the lesson in a single trial. Lilly’s announcement led with 28.3% on the top 12 mg dose.^[7] The NEJM paper leads with 25.0%, against 3.9% on placebo.^[5] Nobody misreported anything. The press figure answers a hypothetical: what would have happened had every participant stayed on the drug. The journal figure answers what happened to everyone assigned to it, dropouts included. A seller quoting “28%” is quoting the hypothetical.

Two trials are still known only from Lilly’s announcements. A PubMed search for TRIUMPH-4 results still returns nothing; what is indexed is the protocol paper describing how the program was designed.^[2] TRIUMPH-3, in people with severe obesity and established heart disease, has also not been published.^[10] Lilly’s release is the only source so far for its heart event counts: 27 cases of cardiovascular death, heart attack or stroke on retatrutide against 23 on placebo (hazard ratio 1.12, 95% CI 0.64 to 1.96), and 44 against 52 on a broader five-part measure (hazard ratio 0.82, 95% CI 0.55 to 1.22).^[10] Both intervals span no difference, and Lilly says events were less frequent than anticipated in both arms, so these counts cannot show whether the drug helps or harms the heart.

The number sellers do not quote

The December 2025 release reported a side effect that rarely appears in the enthusiastic coverage: dysesthesia (abnormal, often unpleasant skin sensation such as burning, prickling or numbness) in 20.9% of the 12 mg arm and 8.8% of the 9 mg arm, against 0.7% on placebo.^[4] Lilly described the events as generally mild.

The larger trials put it lower but did not make it go away. In TRIUMPH-1, Lilly reported dysesthesia in 12.5% on 12 mg and 12.3% on 9 mg against 0.9% on placebo.^[7] The published TRIUMPH-2 paper reports 7% on 12 mg against 1% on placebo, and low blood pressure in 6% against under 1%.^[6] Every trial's figures, dose by dose, are on retatrutide side effects.

Dropouts are the other number. In TRIUMPH-1, 11.3% of the 12 mg arm stopped because of adverse events, against 4.9% on placebo.^[7] In TRIUMPH-2, 12% of the 9 mg arm and 8% of the 12 mg arm stopped for adverse events or death, against 5%.^[6] Phase 3 exists to surface findings like these, and you lose them when you buy a molecule outside the trial system: nobody records what happens to you, and nobody compares it against a placebo arm.

Why you can buy it anyway

You can buy it today, at scale. The vials arrive marked not for human consumption and described as laboratory material. That phrase is a legal device: it is meant to place the sale outside drug regulation while the surrounding marketing addresses people who plainly intend to inject it.

“No medicine containing retatrutide has yet been approved for human use by any regulatory agency in the world.”^[9]

— Eli Lilly and Company, August 12, 2026

• No prescriber. Nobody has assessed whether this is appropriate for you, and nobody is monitoring what happens.
• No pharmacy. Not a 503A, not a 503B, no state licensure, no USP <797> sterile compounding standard, no inspection history to look up.
• No way to know what is in the vial. Identity, purity, concentration and sterility are all unverified. A certificate of analysis supplied by the seller is a document the seller produced.
• No accountability if it goes wrong. Avoiding that accountability is what the label’s disclaimer is for.
• No labeled dose. Lilly has not yet filed its application. Until the FDA reviews the data there is no approved dose, schedule or monitoring plan; what exists are the doses the trials tested.

If you want what retatrutide promises

The approved drugs already deliver a large part of it. Tirzepatide reported 20.9% at 72 weeks in a published phase 3 trial of 2,539 people, and you can be prescribed it tomorrow with a clinician and a pharmacy attached.^[3] Retatrutide’s comparable published figure is 25.0% at 80 weeks.^[5] Those are separate trials of different lengths, so the gap is indicative, not measured. CagriSema and retatrutide have never been tested against each other. In separate phase 3 trials, CagriSema reached −20.4% at 68 weeks in REDEFINE-1 and retatrutide 12 mg −25.0% at 80 weeks in TRIUMPH-1.^[5]

If retatrutide is approved, it will be prescribable, monitored and covered by the same accountability every other drug has, and that is worth waiting for. Our comparison of retatrutide against Zepbound sets the two side by side, and the research-chemical market covers what buying outside the system involves.

Common questions