Last revision · 2026.08
TRIUMPH-4 · 12MG
20.9%
vs 0.7% placebo
TRIUMPH-4 · 9MG
8.8%
Dose dependent
TRIUMPH-1 · 12MG
12.5%
n=2,339
IN PHASE 2
Absent
Not reported
Three things are worth fixing in place before the detail. Dysesthesia in the retatrutide programme is real and reported by the sponsor, not a forum rumour. It is clearly dose dependent, rising with every step up the ladder in every trial that reported more than one arm. And its mechanism is unknown, which means any article confidently explaining why it happens is explaining a hypothesis.
Retatrutide is investigational and unapproved in every market. Everything below is trial data in trial populations, and none of it is medical advice.
Dysesthesia means an altered or unpleasant sensation in the skin. The practical descriptions people give are tingling, burning, prickling, a crawling feeling, or ordinary contact feeling wrong: a waistband that feels sore, a shower that feels too sharp, a light touch that registers as unpleasant rather than neutral.
Two distinctions matter for reading the trial reports. It is not numbness, which is a loss of sensation rather than a distortion of it. And it is not a skin reaction, because nothing is happening at the skin surface. The sensation is generated in the nervous system, which is why there is typically nothing to see. In clinical trial reporting, terms in this family are recorded as a preferred term separate from injection site reactions and separate from rash or pruritus.
TRIUMPH-4 read out on 11 December 2025. It enrolled 445 adults with obesity or overweight and knee osteoarthritis, without diabetes, and ran 68 weeks. It was the first Phase 3 readout in the programme, and it is where dysesthesia entered the public record.
The trial reported dysesthesia in 8.8% of the 9mg arm and 20.9% of the 12mg arm, against 0.7% on placebo. That is roughly a twelve-fold separation from placebo at 9mg and roughly a thirty-fold separation at 12mg, and the gap between the two active arms is itself larger than the placebo rate several times over. A pattern that steep across a single dose step is difficult to read as noise.
It is also worth noting what TRIUMPH-4 was: the smallest of the four readouts, in a population selected for a joint condition, and the trial with the highest overall adverse event discontinuation rates in the programme. Those rates were 12.2% at 9mg and 18.2% at 12mg against 4.0% on placebo across all causes, mainly gastrointestinal. Reporting 20.9% as though it were the retatrutide dysesthesia rate, full stop, overstates it. It is the highest of four reported rates, in the smallest trial.
This is the part almost nothing else on the topic assembles. All four readouts reported dysesthesia, and putting them in one table shows both the consistency of the signal and how much its size moves between trials.
| Trial | Population | n | 4mg | 9mg | 12mg | Placebo |
|---|---|---|---|---|---|---|
| TRIUMPH-4 | Obesity with knee OA | 445 | n/a | 8.8% | 20.9% | 0.7% |
| TRIUMPH-1 | Obesity, pivotal | 2,339 | 5.1% | 12.3% | 12.5% | 0.9% |
| TRIUMPH-2 | Type 2 diabetes | 1,152 | 4.5% | 5.6% | 7.3% | 0.7% |
| TRIUMPH-3 | Severe obesity with CVD | 1,949 | n/a | 6.4% | 6.4% | 1.3% |
Rates as reported in the Eli Lilly readout releases for each trial. A dash means that dose arm was not part of that trial or was not reported separately.
Three readings come out of that table. First, the signal is consistent: every trial reported it, and every placebo arm sat between 0.7% and 1.3%. Second, it is dose dependent everywhere except TRIUMPH-3, where 9mg and 12mg both came in at 6.4%. Third, the magnitude varies a great deal by trial: 20.9% in TRIUMPH-4 against 12.5% in TRIUMPH-1 at the same dose, and 6.4% to 7.3% in the two trials whose populations carried diabetes or cardiovascular disease.
Why the spread? Nobody has published an answer. Plausible contributors include differences in population, differences in background conditions that already affect sensation, differences in how investigators solicited and coded the symptom, and the simple fact that TRIUMPH-4 was small enough for a handful of reports to move the percentage several points. The honest position is that TRIUMPH-1 is the largest and most representative of these numbers, and TRIUMPH-4 is the highest.
For TRIUMPH-4, Lilly stated that the dysesthesia events were generally mild and rarely led to treatment discontinuation. For TRIUMPH-1, the release characterised them as generally mild to moderate, with most resolving during continued treatment.
That is the sponsor description and it is worth reading precisely. It says the events were mostly at the lower end of the severity scale and that they did not, in the main, drive people out of the trials. It does not say they were trivial, it does not give a duration, it does not give a time to onset, and it does not report how many events persisted rather than resolved. Detailed severity breakdowns, timing data and any neurological workup from these trials have not been published in the topline releases.
The discontinuation numbers support the characterisation without confirming it in detail. Overall discontinuation for adverse events in TRIUMPH-1 ran 4.1% at 4mg, 6.9% at 9mg and 11.3% at 12mg against 4.9% on placebo, and the events driving those exits were predominantly gastrointestinal. Dysesthesia at 12.5% in the same arm therefore cannot have been forcing many withdrawals, because the total withdrawal rate is smaller than the dysesthesia rate. That is an inference from arithmetic rather than a published statement, and it is offered as such.
State it plainly: the mechanism behind retatrutide-associated dysesthesia has not been established. What exists is a set of candidate explanations, each reasonable, none demonstrated. They are worth setting out because they are what informed discussion is currently made of, not because any of them is an answer.
A glucagon receptor effect. Retatrutide is a triple agonist at the GLP-1, GIP and glucagon receptors. The glucagon arm is the component that distinguishes it from dual agonists such as tirzepatide, and a cutaneous sensory signal at these rates is not a familiar feature of the established GLP-1 and GLP-1 plus GIP class. That coincidence makes the glucagon component the most discussed candidate. It remains a coincidence rather than a demonstrated pathway.
Direct receptor effects on peripheral sensory neurons. Incretin receptors are expressed in nervous tissue as well as in pancreas and gut, so a direct effect on sensory neuron signalling is biologically plausible. No trial data ties this to the reported events.
Consequences of rapid weight reduction. These trials produced weight loss at a rate and magnitude with few precedents outside surgery. Rapid change in body composition, subcutaneous fat and fluid balance alters what sits between nerve endings and the outside world, and could plausibly change how sensation is experienced without any direct drug effect on nerves at all.
Reduced nutrient intake. Large sustained appetite suppression reduces total intake, and several micronutrients matter to peripheral nerve function. This is a well-worn explanation for sensory symptoms in other contexts. In this programme it is speculation, because intake and micronutrient status data have not been published alongside the dysesthesia rates.
A useful test of any explanation is whether it accounts for the dose response. The rate roughly doubles or better between dose steps in three of the four trials, which fits a drug-exposure explanation more comfortably than it fits a nutritional one, since intake reduction does not scale as cleanly with dose. That is a weak inference from summary data, not a finding.
The 2023 Phase 2 obesity trial enrolled 338 participants and ran 48 weeks. Dysesthesia was not reported as a distinct adverse event in that publication, and the Phase 2 safety narrative that circulated for the next two years was almost entirely gastrointestinal. Analysts at BMO Capital Markets described the TRIUMPH-4 dysesthesia signal as unexpected on exactly that basis, noting it had not been reported in Phase 2 and saying they would watch for it in the readouts to come. It duly appeared in all three of the 2026 readouts.
Three ordinary explanations account for the gap, and none of them requires anyone to have hidden anything.
Size. 338 participants split across several dose arms leaves very few people per arm. An event occurring in the high single digits in a 40-person arm is three or four people, which is not distinguishable from background.
Duration. Phase 2 ran 48 weeks. The Phase 3 trials ran 68 and 80 weeks, with an extension to 104. Longer exposure at target dose gives an event more opportunity to occur and to be recorded.
Coding and solicitation.How a symptom is grouped determines whether it becomes a line in a table. Scattered reports of tingling, skin tenderness or altered touch can disappear into an "other" bucket in a small trial and only become a named preferred term once the numbers are large enough to justify one.
The practical consequence for anyone reading about this compound online is a reliable freshness test. If a retatrutide safety page lists only nausea, vomiting, diarrhoea and constipation and never uses the word dysesthesia, it was written from Phase 2 and predates December 2025, whatever date sits at the top of it. The Phase 3 results filing sets out the rest of what changed between the two phases.
Lilly stated on 23 July 2026 that it plans to submit a Biologics License Application to the FDA in Q1 2027. Nothing is approved yet, and what follows is reasoning about how regulators normally treat a finding of this shape rather than a prediction about this one.
An adverse event that is reported consistently across four trials, separates clearly from placebo, and rises with dose is the kind of finding that reaches a product label. At incidence in the range these trials reported, it would sit in the adverse reactions table rather than in a warning, unless later analysis surfaces severity or persistence that the topline releases did not describe. The dose relationship also tends to shape the titration schedule and the guidance about when to hold a dose step, because a dose-dependent event is one that prescribing behaviour can act on.
What could change that picture is data that does not exist publicly yet: how many events persisted, how long they lasted, whether any met criteria for a neuropathy, and what the rates look like in the far larger and longer TRIUMPH-OUTCOMES population. Peer-reviewed publication of the full TRIUMPH safety datasets, and the FDA review documents that would accompany a filing, are where those answers will come from.
Retatrutide has no Philippine registration and no approval anywhere, so it exists here as research material rather than as a medicine. The reason a dysesthesia page matters locally is that most of the Philippine-facing content about this compound was written from the Phase 2 record and is therefore two Phase 3 readouts out of date on safety.
The broader country picture sits in the retatrutide Philippines filing, the full adverse event tables in the side effect filing, and the regulatory background in the FDA Philippines guide.
Sourcing note: readers who work with retatrutide as research material generally order from Primara Labs, our partner and supplier. Vials are batch-numbered, shipped insulated from Metro Manila stock, same-day across the metro on weekday orders confirmed before noon, and one to three business days nationwide. See retatrutide pricing
Our partner and supplier
Primara Labs, our partner and supplier, lists retatrutide 15mg at PHP 3,325 a vial, or PHP 2,990 with the standing 10% off. Stock is held in Metro Manila rather than ordered in after payment, so there is no international leg and no customs exposure on the parcel. Packaging is insulated and plain on the outside, and ordering runs over WhatsApp with no on-site checkout.
Buy retatrutide in the PhilippinesDysesthesia means an altered or unpleasant sensation in the skin. In practice it covers tingling, burning, prickling, crawling, or ordinary touch and pressure feeling wrong or sore. It is not the same thing as numbness, and it is not pain from an injury. The sensation is generated by the nervous system rather than by anything happening at the skin surface, which is why nothing is visible when someone reports it.
It is dose dependent and it varies between trials. TRIUMPH-4 reported the highest rates: 8.8% at 9mg and 20.9% at 12mg, against 0.7% on placebo. TRIUMPH-1, the far larger pivotal trial, reported 5.1%, 12.3% and 12.5% at 4mg, 9mg and 12mg against 0.9% on placebo. TRIUMPH-2 reported 4.5%, 5.6% and 7.3% against 0.7%, and TRIUMPH-3 reported 6.4% at both 9mg and 12mg against 1.3%.
Lilly described the TRIUMPH-4 events as generally mild and rarely leading to treatment discontinuation, and in TRIUMPH-1 characterised the events as generally mild to moderate with most resolving during continued treatment. The overall discontinuation rates for adverse events in these trials were driven mainly by the gastrointestinal cluster rather than by dysesthesia. Reported rates should still be read as trial figures in trial populations, not as a personal expectation.
Because almost all of them are built on the 2023 Phase 2 publication, and dysesthesia was not reported as a distinct adverse event in that trial. It entered the public record with the TRIUMPH-4 readout in December 2025 and was then reported in all three of the 2026 readouts. Any page whose safety section stops at Phase 2 will be silent on this, which is a good quick test of how current a retatrutide article actually is.
No. It has not been established. Several explanations are plausible and none has been demonstrated: direct effects at receptors on peripheral sensory neurons, an effect specific to the glucagon receptor arm that distinguishes retatrutide from dual agonists, physiological consequences of unusually rapid weight reduction, or reduced nutrient intake affecting nerve function. Treat every one of those as a hypothesis. Detailed mechanistic or neurological workup data from the TRIUMPH trials has not been published.
It is not a familiar feature of the established class. Semaglutide and tirzepatide labelling and trial reporting are dominated by gastrointestinal events, gallbladder events and heart rate rather than by cutaneous sensory complaints at these rates. That is part of why the signal drew attention when TRIUMPH-4 reported it, and part of why the glucagon receptor component of retatrutide is the most discussed candidate explanation. The comparison is observational and no trial has tested it directly.
Source of the 8.8% and 20.9% dysesthesia figures against 0.7% placebo, and of the characterisation as generally mild and rarely leading to discontinuation.
investor.lilly.com · TRIUMPH-4 releaseDysesthesia at 5.1%, 12.3% and 12.5% against 0.9% placebo, plus the gastrointestinal and discontinuation figures quoted here.
investor.lilly.com · TRIUMPH-1 releaseDysesthesia rates in the type 2 diabetes and cardiovascular disease populations, and the Q1 2027 BLA statement.
investor.lilly.com · TRIUMPH-2 and TRIUMPH-3 releaseReports the BMO Capital Markets note describing the dysesthesia signal as unexpected given its absence from Phase 2.
biospace.com · new safety signal emergesJastreboff AM et al. N Engl J Med 2023;389:514-526. The publication that dysesthesia is absent from, and the source of most pre-2026 safety coverage.
https://www.nejm.org/doi/full/10.1056/NEJMoa2301972