Retatrutide: what the trials show so far, and what is still a press release

Retatrutide looks like the strongest weight-loss drug yet tested, but most of the biggest numbers you will see quoted have not been peer reviewed. The published phase 2 trial found 24.2% weight loss at 48 weeks on the top dose [2], and the first published phase 3 trial, in type 2 diabetes, found up to 15.3% at 40 weeks [5]. The 28% figures come from Eli Lilly press releases [7,8]. It is not approved anywhere we checked, and outside trials and a narrow US access programme it isn't legally available.

What it is

Retatrutide (development code LY3437943) is a 39-amino-acid peptide from Eli Lilly that activates three hormone receptors at once: GIP, GLP-1 and glucagon [1]. The first two are the same targets as tirzepatide. Glucagon is the odd one out, because it normally raises blood sugar. The reasoning is that glucagon also increases energy expenditure and pushes the liver to burn fat, while the GLP-1 and GIP activity holds blood sugar down and cuts appetite [1].

That is mechanism theory, drawn mostly from animal work and early human pharmacology in the discovery paper [1]. Nobody has yet shown in a controlled human trial how much of retatrutide's extra effect comes from the glucagon arm. The molecule carries a fatty side chain for albumin binding, which gives it a half-life long enough for weekly injection [1].

The human evidence

Two tiers of evidence exist, and they should be kept apart: peer-reviewed papers you can read and check, and company announcements of trials that have not yet been published.

Peer-reviewed: the phase 2 obesity trial

Side effects were mostly gastrointestinal and dose-related [2], and starting at 2 mg rather than 4 mg reduced them. Heart rate rose with dose, peaked at 24 weeks and then eased [2].

Peer-reviewed: type 2 diabetes and liver fat

A phase 2 trial in 281 people with type 2 diabetes compared retatrutide with placebo and with dulaglutide, an established GLP-1 drug [3]. At 24 weeks, HbA1c fell by up to 2.02 percentage points on retatrutide against 1.41 on dulaglutide. By 36 weeks, weight had fallen 16.94% on the 12 mg dose against 2.02% on dulaglutide [3].

A substudy of the obesity trial looked at 98 participants who had fatty liver (at least 10% liver fat) [4]. At 24 weeks, liver fat fell by 81.4% on 8 mg and 82.4% on 12 mg, against a 0.3% rise on placebo. On 12 mg, 86% got their liver fat back into the normal range [4]. That is imaging data, not biopsy evidence of reduced liver inflammation or scarring, which is what regulators want for a liver-disease licence.

A post-hoc analysis of both phase 2 trials, written by Lilly staff, reported falls in apolipoprotein B of about 21–24% and, in the non-diabetic trial, a 54.8% fall in C-reactive protein, a marker of inflammation [6]. These are risk markers, not heart attacks prevented.

Peer-reviewed: the first phase 3 paper

TRANSCEND-T2D-1 is the only phase 3 trial published in a journal as of this writing [5].

Weight loss is smaller here than in the obesity trials, partly because it was shorter and partly because people with diabetes consistently lose less on these drugs [5]. Two deaths occurred in the 4 mg group; investigators judged both unrelated to the drug [5].

Company-reported: the TRIUMPH phase 3 trials

The following results come only from Lilly press releases. They have not been through peer review, the full data are not public, and companies choose which numbers go in a headline. We report them because they are the largest trials, but treat them as provisional.

Those headline figures use what Lilly calls the efficacy estimand, an estimate of what would have happened if everyone had stayed on the drug. Placebo groups lost 2.1% in TRIUMPH-4 and 2.2% in TRIUMPH-1 [7,8]. For TRIUMPH-1, Lilly also gave the treatment-regimen figure, which counts people who stopped: 25.0% on 12 mg against 3.9% on placebo [8]. That's still a very large effect, but it is the number closer to real-world use.

Two other details from TRIUMPH-1 get quoted a lot. Lilly says 45.3% of people on 12 mg lost at least 30% of their body weight by 80 weeks, against 0.5% on placebo, and that in a subgroup of 532 people who started with a BMI of 35 or more and stayed on for 104 weeks, the 12 mg group averaged 30.3% [8]. Both are efficacy-estimand figures from a press release, and the second comes from a subgroup, so they are the most optimistic slice of the data rather than a typical result.

In TRIUMPH-4, the knee trial, pain scores on the WOMAC scale fell by about 75% on retatrutide against 40% on placebo [7]. A 40% improvement on placebo is large, which is common in pain trials and a reminder of why a control group matters.

The diabetes trials tell a more modest story. In TRIUMPH-2, weight fell 12.7%, 19.1% and 20.8% on 4, 9 and 12 mg, with HbA1c falling by up to 1.6 points [9]. That's in line with the pattern seen for semaglutide and tirzepatide, where diabetes blunts the weight response.

What the animal and lab work suggests

The discovery work in mice and early human dosing established that the molecule hit all three receptors and lowered weight and glucose [1]. Beyond that, animal studies matter less than usual here, because the human trial programme is so large. We are not aware of a published human study isolating what the glucagon component does over the long term, for instance to muscle or bone.

Safety and side effects

Gut. As with other drugs in this class, nausea, diarrhoea, constipation and vomiting dominate. In TRIUMPH-1, nausea affected 42.4% on 12 mg against 14.8% on placebo [8].

Stopping for side effects. This rises with dose and weight loss. In TRIUMPH-1, 11.3% on 12 mg stopped because of adverse events, against 4.9% on placebo; in TRIUMPH-4 it was 18.2% on 12 mg against 4.0% [7,8]. In TRIUMPH-2 and TRIUMPH-3 the figures were 7.7% and 13.5% on 12 mg, against about 5% on placebo [9]. The peer-reviewed phase 3 diabetes trial, at lower weight loss, saw 2–5% [5].

Dysaesthesia. This is the new signal. Dysaesthesia means abnormal skin sensations such as tingling, burning or sensitivity to touch. Lilly reported it in 5.1–12.5% of retatrutide groups in TRIUMPH-1 against 0.9% on placebo, and in 20.9% on 12 mg in TRIUMPH-4 [7,8]. The company describes it as mostly mild to moderate and says most cases resolved during treatment [8]. Nobody has published an explanation for it yet.

Heart rate. Dose-dependent rises in heart rate appeared in phase 2, peaking at 24 weeks [2]. The phase 3 cardiovascular data will matter here.

Legal and regulatory status

Retatrutide is not approved as a medicine in the UK or US. In August 2026 the MHRA said it has not been authorised for medicinal use in the UK and cannot be legally supplied outside clinical trials [12]. In the US, Lilly says it plans to submit an application to the FDA in the first quarter of 2027 [9].

Since mid-2026 Lilly has run a pre-approval expanded access programme in the US, registered on ClinicalTrials.gov, for a small number of patients whose doctors apply on their behalf [10,11]. That isn't a licence and doesn't make the drug available to the public.

For athletes, the 2026 WADA Prohibited List bans at all times any pharmacological substance not approved for human therapeutic use by a regulator, a category (S0) that applies to retatrutide for as long as it stays unapproved [13].

Anything sold online as retatrutide is, by definition, not the trial product.

What we still don’t know

• The full peer-reviewed TRIUMPH results, including all side-effect tables and the treatment-regimen numbers for every trial [7–9].
• What causes dysaesthesia and whether it fully resolves after stopping [8].
• Whether glucagon activity affects lean mass, bone or heart rhythm over years rather than months.
• Whether it reduces heart attacks and deaths, not just risk markers [6].
• What happens to weight when people stop; no published retatrutide trial has yet tested withdrawal.

Educational content only — not medical advice. Many peptides discussed on HPR are not approved for human use. Talk to a qualified clinician before making any decision about your health.