Retatrutide side effects: the Phase 3 numbers compared

Key takeaways

• Mostly gastrointestinal. Nausea, diarrhea, vomiting and constipation were the most common adverse events in every retatrutide trial that has reported, dose-related and mostly mild to moderate .
• The two Phase 3 obesity trials disagree at the same dose. On 12 mg, nausea was 42.4 percent in TRIUMPH-1 and 43.2 percent in TRIUMPH-4, but dysesthesia (an abnormal skin sensation) was 12.5 percent in the first and 20.9 percent in the second .
• Dropout at 12 mg spans a wide range: 5.1 percent in TRANSCEND-T2D-1, 11.3 percent in TRIUMPH-1, 18.2 percent in TRIUMPH-4, against placebo rates of 0.0 to 4.9 percent .
• Dysesthesia turns up across the GLP-1 class, and the retatrutide dose gradient is uneven. A 2026 VigiBase analysis associates semaglutide with dysesthesia and burning sensation and tirzepatide with hyperaesthesia, reports these sensory reactions as dose-dependent, and describes recovery after stopping . Across 4, 9 and 12 mg the trial rates ran 5.1, 12.3 and 12.5 percent in TRIUMPH-1 and 4.5, 2.3 and 4.4 percent in TRANSCEND-T2D-1, where about 134 people per arm puts those last three at roughly six, three and six events .
• Retatrutide is still in clinical development with no approved label. Only the type 2 diabetes trial has been peer-reviewed; the obesity tolerability numbers come from company releases .

Skip to:

• Three Phase 3 trials, three different tolerability numbers
• The gut effects arrive with the dose climb and then settle
• Between 5 and 18 percent stopped over side effects
• Dysesthesia shows up across the class, and the dose gradient is uneven
• Gut rates like semaglutide's, dropout roughly double
• What the safety record still does not answer
• What is measured about retatrutide bought outside a trial
• The bottom line on retatrutide tolerability

Three Phase 3 trials, three different tolerability numbers

Three Phase 3 retatrutide trials have reported adverse-event data. TRIUMPH-1 ran 80 weeks in adults with obesity or overweight and randomized 2,339 participants across 4, 9 and 12 mg and placebo . TRIUMPH-4 ran 68 weeks in adults with obesity or overweight plus knee osteoarthritis . TRANSCEND-T2D-1 ran 40 weeks in 537 adults with type 2 diabetes and is the only one of the three published in a journal . The two obesity trials exist so far as Lilly press releases, which is a weaker source than a paper with a full adverse-event table, and worth holding in mind while reading the grid below .

Lilly's own framing of the June 2026 data is that the types of adverse events in TRIUMPH-1 and TRANSCEND-T2D-1 were generally consistent with trials of other incretin-based therapies . The categories are familiar across the class. The rates at a fixed dose are where the three trials part company, which is the part a single headline percentage hides.

The gut effects arrive with the dose climb and then settle

Nausea, diarrhea, vomiting and constipation were the most common adverse events in the Phase 2 trial, where they were dose-related and mostly mild to moderate . A 2025 meta-analysis of the randomized trials available at the time put nausea at roughly four times the placebo rate on the higher doses (RR 4.27 at 8 mg, RR 4.00 at 12 mg) and vomiting at eight to nine times placebo (RR 8.13 at 8 mg, RR 8.98 at 12 mg) .

Timing is the consistent part. Lilly's Phase 2 summary put the gastrointestinal events in the dose-escalation period , and the peer-reviewed Phase 3 in type 2 diabetes describes them as mild to moderate and says they subsided over time .

• The Phase 2 trial randomized the starting dose as well as the target, and gastrointestinal events were partially mitigated with a 2 mg start rather than a 4 mg start .
• That start-dose comparison only ran at the 4 mg and 8 mg targets. The 12 mg group had one tested route, starting at 2 mg, so nobody has measured what a fast start to 12 mg does .
• Phase 3 settled on a 4, 9 and 12 mg dose set . TRIUMPH-4 tested only the 9 and 12 mg arms .

Between 5 and 18 percent stopped over side effects

Adverse events pushed 6 to 16 percent of Phase 2 participants off the drug, against none on placebo . In Phase 3 at 12 mg the answer depends on which trial you read: 5.1 percent in TRANSCEND-T2D-1 against 0.0 percent on placebo, 11.3 percent in TRIUMPH-1 against 4.9 percent, and 18.2 percent in TRIUMPH-4 against 4.0 percent .

Dysesthesia shows up across the class, and the dose gradient is uneven

Dysesthesia is an abnormal skin sensation: tingling, altered touch, or burning. The three Phase 3 trials put very different numbers on it. TRIUMPH-4 reported 8.8 percent at 9 mg and 20.9 percent at 12 mg, against 0.7 percent on placebo . TRIUMPH-1 reported 5.1, 12.3 and 12.5 percent across 4, 9 and 12 mg, against 0.9 percent. TRANSCEND-T2D-1, in adults with type 2 diabetes, reported 4.5, 2.3 and 4.4 percent across the same three doses, against 0.0 percent . Read that last set with its denominators: TRANSCEND-T2D-1 randomized 134, 133 and 136 people to the three doses, so those percentages are roughly six, three and six events per arm . Neither release explains why the same 12 mg dose landed at 4.4 percent in one trial and 20.9 percent in another.

The same reaction turns up across the GLP-1 class. A 2026 disproportionality analysis of VigiBase, the WHO global pharmacovigilance database, found exenatide associated with hypoesthesia and oral paraesthesia, semaglutide and tirzepatide with hyperaesthesia, and semaglutide with dysesthesia and burning sensation. The authors conclude that the pharmacovigilance data strengthens evidence for dysesthesias already observed in clinical trials of semaglutide, tirzepatide and retatrutide .

Two findings from that analysis are worth having if you are already dosing. The authors report the reaction as dose-dependent, appearing more often at higher doses and with the more potent agonists in the class, which is a statement about the class rather than about any one retatrutide arm. And in the pharmacovigilance case narratives, discontinuation was often followed by spontaneous favourable outcomes, with rechallenge cases observed . The analysis also describes skin burning sensations as a distinctive form of dysesthesia .

The other non-gut finding is cardiac. Phase 2 saw dose-dependent increases in heart rate that peaked at 24 weeks and declined after that . The Phase 2 paper reports the rise settling rather than compounding over the 48 weeks it ran.

Gut rates like semaglutide's, dropout roughly double

TRIUMPH-1's 12 mg arm averaged 28.3 percent body-weight loss over 80 weeks . The gut column does not scale with that. Nausea at 12 mg was 42.4 percent, and the Wegovy label reports 44 percent for semaglutide 2.4 mg .

Read the gaps as rough: these are different trials, different populations and different lengths of follow-up, and the label figures are pooled across trials. The row that survives the caveat is the last one. Retatrutide's gut rates sit alongside semaglutide's, while its discontinuation rate is roughly double the 6.8 percent on the Wegovy label and the 6.7 percent on Zepbound's 15 mg arm, and TRIUMPH-4's 18.2 percent is nearly triple . Something is pushing people off retatrutide that the nausea column does not capture.

What the safety record still does not answer

• Why the same 12 mg dose produced 4.4 percent dysesthesia in the diabetes trial, 12.5 percent in one obesity trial and 20.9 percent in the other, and why the per-dose rates rose cleanly in TRIUMPH-4 but not in the other two. The releases publish the numbers and no explanation .
• Whether retatrutide dysesthesia resolves after stopping. The class pharmacovigilance data reports spontaneous favourable outcomes after discontinuation, but it pools GLP-1 receptor agonists rather than breaking the outcome out per drug .
• What the obesity adverse-event tables look like under peer review. TRANSCEND-T2D-1 is in the Lancet; TRIUMPH-1 and TRIUMPH-4 are company releases, so nobody outside Lilly has checked the denominators .
• Whether a low start helps at the 12 mg target specifically. Phase 2 only compared a 2 mg against a 4 mg start at the 4 mg and 8 mg targets .
• How tolerability behaves long term. The adverse-event windows in these three trials run 40 to 80 weeks .
• What a vial bought outside a trial contains. Every number on this page describes supervised participants dosed with Lilly's material on a fixed schedule.

What is measured about retatrutide bought outside a trial

Retatrutide has no approved product anywhere, so everything supplied outside a trial comes through an unregulated channel. Two 2026 measurements give that population a size. On 29 May 2026 the UK medicines regulator reported its largest ever seizure of unlicensed weight-loss medicines, around 12,000 doses, naming retatrutide and tirzepatide among them . In the US, trade coverage of a CBS News investigation reported human exposures logged by poison control centers up 265 percent in the first four months of 2026, alongside more than 120 websites and over 50 clinics selling or promoting the drug .

Poison-center contacts count calls, not confirmed harms, and none of them say what was in the vial. The 265 percent measures how many people are now dosing retatrutide outside a trial, which is a different quantity from a side-effect rate.

• Start low. The one escalation lever with randomized data behind it is the starting dose, where gastrointestinal events were partially mitigated by beginning at 2 mg instead of 4 mg .
• Do the concentration math before the syringe is in your hand. The peptide reconstitution calculator converts a vial and a diluent volume into a draw, and the GLP-1 ramp planner holds the schedule.
• Expect the rough stretch during escalation and log it. The dose-climb clustering shows up in both the Phase 2 summary and the peer-reviewed Phase 3, which describes the events subsiding over time .
• Log new tingling or burning skin against the dose you were on, and do not assume a steady dose rules it out. The 9 to 12 mg step more than doubled the rate in TRIUMPH-4 (8.8 to 20.9 percent) but barely moved it in TRIUMPH-1 (12.3 to 12.5 percent), and TRANSCEND-T2D-1 showed no dose trend on arms of about 134 people, where 4.5, 2.3 and 4.4 percent across 4, 9 and 12 mg is roughly six, three and six events, too few to call either way . The class pharmacovigilance analysis reports the reaction as dose-dependent overall .

The bottom line on retatrutide tolerability

As the trials measured it, retatrutide's profile is gastrointestinal, front-loaded into the dose climb, and steeper at 12 mg. The part worth carrying away is the spread. The same 12 mg dose produced 11.3 percent dropout in one Phase 3 obesity trial and 18.2 percent in another, and dysesthesia at that dose ran 4.4 percent in the diabetes trial, 12.5 percent in TRIUMPH-1 and 20.9 percent in TRIUMPH-4 . Any single percentage quoted from a single trial is a thinner guide than it looks.

Retatrutide is still in clinical development with no approved label . The type 2 diabetes trial has been through peer review; the obesity tolerability numbers everyone quotes have not, and the ranges above will move when they are published .

For research and educational purposes only. Not medical advice.

pepSmart has not commissioned independent clinical review of this article.

More on how we write and source these pieces: Editorial process and contributor disclosure and Sourcing posture.

Spot an error? Email corrections via /about.

Sources: 12 entries, primary canon (two PubMed trial records, a PubMed pharmacovigilance analysis, a PMC meta-analysis, and two FDA drug labels via DailyMed) plus reputable secondary sources acknowledged inline (two Lilly trial releases, an independent CME summary, a UK regulator notice, and trade coverage of a news investigation), last reviewed 2026-07-21.

Related tools

• Tirzepatide dose calculator - Run tirzepatide-focused vial draw math.
• GLP-1 conversion calculator - Convert a GLP-1 mg dose to U-100 units and ml.
• GLP-1 ramp planner - Preview a linear educational dose-step table.
• Peptide half-life calculator - Estimate single-dose decay from cited half-life constants.
• PK simulator overview - Public overview of the Pro pharmacokinetic simulator.
• Semaglutide dose calculator - Run semaglutide-focused vial draw math.

Frequently asked questions

How long does retatrutide nausea last?

In the trials the gastrointestinal effects clustered during the dose-escalation period and settled after it. The peer-reviewed Phase 3 in type 2 diabetes describes them as mild to moderate and says they subsided over time.

Does retatrutide cause tingling or burning skin?

Yes, it is reported as dysesthesia, and the three Phase 3 trials disagree on how often. At 12 mg it was 4.4 percent in TRANSCEND-T2D-1 against 0.0 percent on placebo, 12.5 percent in TRIUMPH-1 against 0.9 percent, and 20.9 percent in TRIUMPH-4 against 0.7 percent. A 2026 VigiBase analysis finds the same reaction across GLP-1 receptor agonists, reports it as dose-dependent at class level, and describes spontaneous favourable outcomes after stopping.

How many people stop retatrutide because of side effects?

At 12 mg, 5.1 percent stopped in TRANSCEND-T2D-1, 11.3 percent in TRIUMPH-1 and 18.2 percent in TRIUMPH-4. In the Phase 2 trial, adverse events led 6 to 16 percent of retatrutide participants to discontinue, against none on placebo.

Are retatrutide side effects worse than tirzepatide's?

On the gut, retatrutide at 12 mg is close to semaglutide 2.4 mg and higher than tirzepatide 15 mg: nausea 42.4 percent in TRIUMPH-1 against 44 percent on the Wegovy label and 28 percent on the Zepbound label. On discontinuation the gap is wider, 11.3 to 18.2 percent for retatrutide against 6.8 percent for semaglutide and 6.7 percent for tirzepatide. These are separate trials, not a head-to-head.

Does a lower starting dose reduce retatrutide side effects?

In the Phase 2 trial, gastrointestinal events were partially mitigated by starting at 2 mg rather than 4 mg. That comparison was only run at the 4 mg and 8 mg targets, so it has not been tested at the 12 mg target.