GLP-3 side effects: what the retatrutide trials actually report

Photograph: editorial illustration for this article. It does not show retatrutide.

A note on the name first: there is no hormone called GLP-3. The term is a nickname for retatrutide, a triple GIP, GLP-1 and glucagon receptor agonist developed by Eli Lilly, as explained in our GLP-3 peptide article. Everything below is the side-effect record of retatrutide from its published and announced trials.

What are the most common side effects?

Gastrointestinal effects dominate, as with every drug in this class. The largest data set is TRIUMPH-1, the 80-week phase 3 trial in 2,339 adults without diabetes. Lilly's topline announcement lists the most common adverse events on 4, 9 and 12 mg versus placebo as nausea (28.6%, 38.4%, 42.4% versus 14.8%), diarrhoea (25.2%, 34.1%, 32.0% versus 13.5%) and constipation (23.8%, 25.9%, 26.1% versus 10.9%)^[1].

The published phase 2 obesity trial describes the same pattern in more detail: gastrointestinal adverse events were dose-related, mostly mild to moderate, and occurred mainly during dose escalation, with a slower escalation schedule reducing their frequency^[2]. In the phase 2 diabetes trial, gastrointestinal events were again the most frequent and again dose-related, and no severe hypoglycaemia was reported on retatrutide^[3]. In TRIUMPH-2, diarrhoea was the most common event (27.4% to 33.6% across doses versus 13.2% on placebo), ahead of nausea (13.7% to 28.0% versus 8.0%)^[4].

Does retatrutide raise heart rate?

Yes. This is the effect that most distinguishes the triple agonist from a pure GLP-1 drug. The phase 2 obesity trial reported a dose-dependent increase in heart rate that peaked at around 24 weeks and declined thereafter^[2]. Both the GLP-1 and the glucagon receptor components are plausible contributors. Whether a transient rise in resting heart rate translates into cardiovascular harm or is outweighed by the benefits of large weight loss is exactly the question TRIUMPH-Outcomes was designed to answer, and that trial has an estimated primary completion of February 2029^[6].

How many people stopped because of side effects?

Figures from the two Lilly announcements^[1][4]. For comparison, in tirzepatide's SURMOUNT-1 trial, 6.2% of the 15 mg group discontinued for adverse events versus 2.6% on placebo^[5]. So at the top dose, retatrutide's dropout for side effects is roughly twice tirzepatide's, in different trials. The lower doses look closer to what is already on the market.

Why does dose escalation matter so much?

Every retatrutide trial started participants on a low dose and stepped up over several months. The phase 2 obesity trial included arms with different starting doses to test this directly and and reported that a lower starting dose was associated with fewer gastrointestinal events^[2]. That is why the side-effect frequencies above cannot be transferred to any use outside a trial: without the protocol, there is no basis for expecting the same tolerability.

Which class effects apply?

• Gallbladder events are reported across GLP-1-based drugs after rapid weight loss; they were monitored in the phase 2 trials and are part of the phase 3 safety package that has not yet been published in full.
• Pancreatitis is a labelled precaution for the whole class and will be assessed by the FDA from the adjudicated phase 3 data.
• Thyroid C-cell tumours in rodents are the reason approved GLP-1 receptor agonists carry a boxed warning; the same warning is likely for any approved triple agonist.
• Hypoglycaemia was not an issue on retatrutide alone in the diabetes trial, but the combination with insulin or sulfonylureas is a known class risk^[3].
• Lean mass loss accompanies all large weight loss; phase 3 includes body-composition substudies whose results are pending.

What is still unknown?

Three things. First, the full adverse-event tables from phase 3 are not public; only the top-line frequencies are. Second, follow-up ends at 104 weeks in the longest announced extension, so nothing is known beyond two years. Third, the heart-rate signal has no outcome data behind it until around 2029. Anyone weighing these numbers should also read what happened to weight in the same trials, on our clinical trials page, and remember that they come from supervised, escalated dosing of a pharmaceutical-grade product.