Retatrutide Side Effects: Comprehensive Phase 3 Evidence Review

The honest short answer

• Retatrutide’s GI adverse-event profile is class- typical for a GLP-1 receptor agonist. At 12 mg in TRIUMPH-1: nausea 42.4%, diarrhea 32.0%, constipation 26.1%, vomiting 25.3%^[1]. These numbers fit inside the envelope already mapped by Wegovy STEP-1 (Wilding 2021 NEJM)^[5] and Zepbound SURMOUNT-1 (Jastreboff 2022 NEJM)^[6].
• Dysesthesia is the genuinely novel retatrutide signal. 12.5% on retatrutide 12 mg vs 0.9% on placebo — a ~14-fold imbalance. Dysesthesia did not surface as a notable adverse-event category in STEP-1^[5], SURMOUNT-1^[6], or ATTAIN-1^[7]. The leading mechanistic hypothesis is that the glucagon-receptor arm of retatrutide’s triple agonism contributes, but this is not yet confirmed.
• Discontinuation due to adverse events at 12 mg was 11.3% vs 4.9% placebo. Higher than STEP-1 Wegovy (7.0%) and the SURMOUNT-1 Zepbound range (4.3-7.1%). The trade-off: largest weight loss in the class, at a meaningfully higher tolerability cost.
• The AE profile is dose-dependent. Retatrutide 4 mg (TBWL −19.0%) is substantially better tolerated than 12 mg (TBWL −28.3%); 9 mg (TBWL −25.9%) is intermediate^[1]. Prescribers and patients may negotiate the dose-vs-tolerability trade-off similarly to how tirzepatide 5/10/15 mg decisions are made today.
• Retatrutide is NOT yet FDA-approved as of May 2026. The TRIUMPH-1 pivotal results are public via Eli Lilly’s investor release and the 86th ADA Scientific Sessions^[1][2]; the peer-reviewed manuscript has not yet been PubMed-indexed. TRIUMPH-2 (T2D) and TRIUMPH-3 (cardiovascular outcomes) read out later in 2026. NDA filing anticipated late 2026 to early 2027.
• Rare adverse events typically emerge in post-marketing surveillance, not in 2,339-patient pivotal trials. Until the FDA-approved label is issued and FAERS data accumulates, the trial-level safety data are the best available evidence — but they are not the last word.

What TRIUMPH-1 actually measured

TRIUMPH-1 (ClinicalTrials.gov NCT05929066)^[2] is the pivotal phase 3 randomized, double-blind, placebo-controlled trial of retatrutide in adults with obesity (BMI ≥30) or overweight (BMI ≥27) with at least one weight-related comorbidity. 2,339 participants were randomized to weekly subcutaneous retatrutide at one of three maintenance doses (4 mg, 9 mg, or 12 mg) or placebo, escalated from a 2 mg starting dose stepping up every 4 weeks (2 → 4 → 6 → 9 → 12 mg) until the assigned maintenance dose was reached. The primary endpoint was percent change in body weight from baseline to week 80; a 104-week extension reported continued weight loss in the BMI ≥35 subgroup.

The headline efficacy at 80 weeks: retatrutide 12 mg produced −28.3% mean total body-weight loss (TBWL) — approximately 70.3 pounds at a typical starting body weight — vs −2.2% on placebo^[1]. The 9 mg arm produced −25.9% TBWL; the 4 mg arm produced −19.0%. At 104 weeks in the BMI ≥35 subgroup, the 12 mg arm reached −30.3% TBWL, the largest mean weight loss any pharmacological obesity therapy has produced in a randomized controlled trial at the time of the readout.

For comparison: Wegovy STEP-1 (Wilding 2021 NEJM)^[5] reported −14.9% TBWL at 68 weeks on semaglutide 2.4 mg weekly. Zepbound SURMOUNT-1 (Jastreboff 2022 NEJM)^[6] reported −20.9% TBWL at 72 weeks on tirzepatide 15 mg weekly. Foundayo ATTAIN-1 (Wharton 2025 NEJM)^[7] reported −11.2% TBWL at 72 weeks on orforglipron 36 mg oral daily. Retatrutide at 12 mg produced roughly 1.4x the magnitude of tirzepatide 15 mg, 1.9x of semaglutide 2.4 mg, and 2.5x of orforglipron 36 mg in cross-trial indirect comparison.

Important caveat on the TRIUMPH-1 source data. As of May 2026, the TRIUMPH-1 pivotal manuscript is not yet PubMed-indexed. The data summarized here are from Eli Lilly’s investor release on PR Newswire^[1] and the 86th ADA Scientific Sessions abstract and oral presentation. The peer-reviewed manuscript is anticipated late 2026, almost certainly in the New England Journal of Medicine based on the publication pattern of every prior pivotal GLP-1 obesity trial. The TRIUMPH program rationale and design (Giblin, Kaplan, Somers 2026 Diabetes Obes Metab)^[3] is already published. The retatrutide phase 2 lipid and metabolite profile (Pearson 2026 JCEM)^[4] is also already published. This article will be updated once the TRIUMPH-1 pivotal manuscript reaches PubMed indexing.

GI adverse-event rates — nausea, diarrhea, constipation, vomiting

The gastrointestinal adverse-event pattern in TRIUMPH-1 at the 12 mg dose^[1]:

• Nausea: 42.4% on retatrutide 12 mg vs 14.8% on placebo. Class-typical magnitude. Mostly mild to moderate; concentrated during dose escalation; usually resolves or attenuates after the patient has spent 2 to 4 weeks at the dose step.
• Diarrhea: 32.0% on retatrutide 12 mg vs 13.5% on placebo. Also class-typical. The combined GLP-1 + GIP + glucagon agonism slows gastric emptying and alters small-bowel motility; diarrhea is the predictable counterpart to the slowed-upper- GI nausea signal.
• Constipation: 26.1% on retatrutide 12 mg vs 10.9% on placebo. The slowed gastric emptying and reduced food intake combine to produce constipation in roughly one-quarter of patients at the 12 mg maintenance dose — consistent with the Zepbound and Wegovy patterns.
• Vomiting: 25.3% on retatrutide 12 mg vs 4.8% on placebo. Higher than the Zepbound SURMOUNT-1 12.2% and roughly equal to the Wegovy STEP-1 24.8%. The placebo-arm vomiting rate is notably low (4.8%), giving a ~20 percentage-point absolute imbalance.
• Treatment-emergent AE discontinuation at 12 mg: 11.3% vs 4.9% placebo. This is the cleanest integrated tolerability metric because it captures patients who stopped therapy for any AE-related reason. The 6.4 percentage-point absolute increase is the headline tolerability cost.

The clinical interpretation is the standard GLP-1-class one: GI symptoms are the dominant tolerability challenge, are largely time-limited per dose step, and are usually manageable with the same strategies that work for injectable monoagonists and dual agonists — smaller meals, avoiding very rich or greasy foods, eating slowly, stopping at the first sign of fullness, staying hydrated, and slowing the titration if any particular dose step is intolerable. For patient-facing GI management advice see the GLP-1 side effects Q&A and the interactive GLP-1 side-effect timeline tool.

Dysesthesia — the unique retatrutide signal

Dysesthesia is the genuinely novel adverse-event signal in the retatrutide pivotal data. The reported rate at TRIUMPH-1 12 mg is 12.5% vs 0.9% on placebo^[1] — an absolute increase of 11.6 percentage points and roughly a 14-fold relative imbalance. By contrast, dysesthesia did not surface as a prominent adverse-event category in any of the established class’s pivotal trials: STEP-1 semaglutide^[5], SURMOUNT-1 tirzepatide^[6], or ATTAIN-1 orforglipron^[7].

What is dysesthesia clinically? Dysesthesia is an abnormal sensory experience — typically tingling, burning, pins-and-needles, numbness, itching, or a pricking sensation — that does not correspond to a clear external stimulus. It is distinct from paresthesia (which is the same sensory category but commonly transient and benign, like a foot “falling asleep”) and from neuropathy (which implies established structural nerve damage). In the TRIUMPH-1 pivotal release the dysesthesia rate is reported as an adverse-event category without granular description of distribution (peripheral, oral, scalp, etc.), severity, or reversibility upon discontinuation. Those details will surface in the peer-reviewed manuscript and the FDA-approved label once issued.

Why dysesthesia in retatrutide and not in the comparator molecules? The mechanistic hypothesis most consistent with the cross-trial pattern is that the glucagon-receptor arm of retatrutide’s triple agonism contributes. Retatrutide is a triple agonist of GLP-1, GIP, and glucagon receptors; semaglutide is a pure GLP-1 monoagonist; tirzepatide is a dual GIP+GLP-1 agonist (no glucagon arm); orforglipron is a pure GLP-1 monoagonist (small molecule). The dysesthesia signal cleanly differentiates the molecule that has glucagon agonism from the three that do not. Glucagon receptors are expressed on hepatic, adipose, cardiac, and central tissues, with some evidence of expression in peripheral sensory pathways. The causal connection from glucagon-receptor activation to a dysesthesia phenotype is not established; this remains a hypothesis pending dedicated mechanistic study and post- marketing surveillance.

Is it dose-related? Based on the TRIUMPH-1 dose-response pattern, dysesthesia is reported at the 12 mg maintenance dose at 12.5%. Granular dysesthesia rates at the 4 mg and 9 mg arms will be reported in the peer-reviewed manuscript; based on the overall AE dose-response gradient, a lower rate at lower doses is the expected pattern but is not yet confirmed.

Practical implication for prescribers. Dysesthesia is a novel adverse-event category that prescribers will need to elicit actively at follow-up visits once retatrutide is approved — patients may not volunteer sensory symptoms unless asked, particularly because tingling or burning sensations are often dismissed as nonspecific. Pre-treatment counseling should include the symptom category and the instruction to report new sensory symptoms. Discontinuation criteria for dysesthesia will be specified in the FDA-approved label.

Practical implication for patients. If you experience new, persistent tingling, burning, pins-and-needles, numbness, or unexplained itching while on retatrutide (post-approval), report it to your prescribing clinician. The symptom is on the radar of the trial investigators; it is labeled and tracked; it is not a mystery symptom.

Discontinuation rate — the integrated tolerability metric

Treatment-emergent adverse-event discontinuation is the cleanest cross-trial tolerability metric because it integrates symptom severity, persistence, and patient-level decision-making across the full trial duration. In TRIUMPH-1 at the 12 mg dose, AE discontinuation was 11.3% vs 4.9% placebo^[1] — an absolute increase of 6.4 percentage points.

Cross-trial context:

• STEP-1 (Wegovy semaglutide 2.4 mg, 68 wk): AE discontinuation 7.0% vs 3.1% placebo (absolute increase 3.9 pp)^[5].
• SURMOUNT-1 (Zepbound tirzepatide, 72 wk): AE discontinuation 4.3 to 7.1% across 5/10/15 mg arms vs 2.6% placebo (absolute increase 1.7 to 4.5 pp)^[6].
• ATTAIN-1 (Foundayo orforglipron, 72 wk): AE discontinuation 5.3 to 10.3% across 6/12/36 mg arms vs 2.7% placebo (absolute increase 2.6 to 7.6 pp)^[7].
• TRIUMPH-1 (retatrutide 12 mg, 80 wk): AE discontinuation 11.3% vs 4.9% placebo (absolute increase 6.4 pp)^[1].

Retatrutide’s 11.3% discontinuation rate is the highest among the four molecules. The interpretation depends on the denominator the prescriber uses. Per-patient-on-therapy, approximately 9 of 10 retatrutide 12 mg patients tolerated the drug well enough to remain on it through 80 weeks. Per- unit-of-weight-loss, retatrutide produced approximately −28.3% TBWL at the 12 mg dose — meaning the discontinuation cost was 11.3% to access roughly 1.4x the weight loss of tirzepatide 15 mg or 1.9x of semaglutide 2.4 mg. Whether that trade-off is acceptable will be a patient- and-prescriber-specific decision once the drug is approved.

Class comparison — retatrutide vs tirzepatide vs semaglutide vs orforglipron

Side-by-side adverse-event rates at the labeled high-dose arms of the four molecules (cross-trial indirect comparison):

• Nausea: retatrutide 12 mg 42.4%^[1], Wegovy semaglutide 2.4 mg 44.2%^[5], Zepbound tirzepatide 15 mg 29.6%^[6], Foundayo orforglipron 36 mg in the qualitative GI-dominant range from ATTAIN-1^[7].
• Diarrhea: retatrutide 12 mg 32.0%, Wegovy 31.5%, Zepbound 23.0%. Comparable across the class.
• Constipation: retatrutide 12 mg 26.1%, Wegovy 23.4%, Zepbound 16.8%. Retatrutide marginally higher than Wegovy.
• Vomiting: retatrutide 12 mg 25.3%, Wegovy 24.8%, Zepbound 12.2%. Retatrutide and Wegovy approximately equal; both ~2x Zepbound.
• AE discontinuation: retatrutide 12 mg 11.3% vs 4.9% placebo, Wegovy 7.0% vs 3.1% placebo, Zepbound 4.3-7.1% vs 2.6% placebo, Foundayo 5.3-10.3% vs 2.7% placebo. Retatrutide is the highest.
• Dysesthesia: retatrutide 12 mg 12.5% vs 0.9% placebo — unique to retatrutide. Not a prominent adverse-event category in STEP-1, SURMOUNT-1, or ATTAIN-1.
• Magnitude of weight loss: retatrutide 12 mg −28.3% TBWL at 80 wk^[1], Zepbound 15 mg −20.9% at 72 wk^[6], Wegovy 2.4 mg −14.9% at 68 wk^[5], Foundayo 36 mg −11.2% at 72 wk^[7].

The taxonomy is clean. Retatrutide produces the largest mean weight loss in the class, with a GI profile that fits inside the existing class envelope, plus a novel dysesthesia signal and a meaningfully higher AE discontinuation rate. For the full pipeline landscape including survodutide, MariTide (maridebart cafraglutide), and ecnoglutide, see our GLP-1 pipeline review.

Dose-response — AE rates at 4 mg vs 9 mg vs 12 mg

TRIUMPH-1 tested three maintenance doses (4 mg, 9 mg, 12 mg) and reported both efficacy and tolerability across the dose range^[1]. The weight-loss gradient:

• Retatrutide 4 mg: −19.0% TBWL at 80 wk. Approximately equal to Zepbound 15 mg (−20.9% in SURMOUNT-1)^[6].
• Retatrutide 9 mg: −25.9% TBWL at 80 wk. Exceeds Zepbound 15 mg by ~5 percentage points.
• Retatrutide 12 mg: −28.3% TBWL at 80 wk; −30.3% at 104 wk in the BMI ≥35 subgroup — the largest mean weight loss in any GLP-1-class pivotal trial reported to date.

The AE rates rise with dose. Granular per-dose AE rates for each event category will be reported in the peer-reviewed manuscript; the overall pattern from the investor release and ADA presentation is that the 12 mg arm produced the highest rates of every GI category and the highest dysesthesia rate. The 4 mg arm produced AE rates closer to placebo and is the most tolerable maintenance dose — at the cost of giving up roughly one-third of the weight-loss magnitude available at 12 mg.

Clinical implication. Once retatrutide is FDA-approved, prescribers will likely make explicit dose-vs- tolerability trade-off decisions, similar to the existing tirzepatide 5/10/15 mg titration decisions but with a wider range. A patient who tolerates the 4 mg arm well but does not need maximum-magnitude weight loss may reasonably stay at 4 mg for maintenance. A patient who needs maximum-magnitude weight loss and accepts the higher AE burden can titrate to 12 mg. A patient who tolerates 9 mg may not need or want to push to 12 mg if the additional 2.4 percentage points of TBWL come with disproportionately higher AE rates.

What we do not yet know about retatrutide safety

The TRIUMPH-1 readout is a 2,339-patient pivotal trial with an 80-week primary endpoint and a 104-week extension. That is substantial but does not exhaust the safety question. Several categories of safety information are not yet available:

• Rare adverse events. Events that occur at rates below approximately 1 in 1,000 patient-years typically do not surface in pivotal trials of this size. Thyroid C-cell tumor signals, severe pancreatitis at population scale, and other rare-but-serious adverse events will only become visible in post-marketing surveillance via the FDA Adverse Event Reporting System (FAERS) and post-approval observational studies.
• Dysesthesia mechanism, distribution, severity, and reversibility. The peer-reviewed manuscript will report the dysesthesia phenotype in granular detail; the post-approval FDA label will specify management and discontinuation criteria. Whether dysesthesia attenuates or resolves after discontinuation, whether it can transition to a structural neuropathy with sustained exposure, and whether it is reproducible with dose re-challenge are all currently unknown.
• Long-term cardiovascular safety. TRIUMPH-3 is the dedicated cardiovascular outcomes trial; it reads out later in 2026. Until those data are public, the cardiovascular safety story rests on the cardiometabolic biomarker improvements reported in earlier phase 2 and phase 3 retatrutide data — meaningful but not the same as a hard MACE outcome.
• Adverse events in the type 2 diabetes population. TRIUMPH-2 is the dedicated T2D pivotal; it reads out later in 2026. The TRIUMPH-1 cohort was adults with obesity without diabetes (the parallel of STEP-1 and SURMOUNT-1 for Wegovy and Zepbound). T2D AE patterns, including hypoglycemia risk with background insulin or sulfonylureas, will be characterized in TRIUMPH-2.
• The FDA-approved label. The boxed warning (almost certainly the GLP-1-class thyroid C-cell tumor warning), the formal warnings and precautions section, the dosing instructions, and the contraindications will be finalized in the FDA-approved label, which has not yet been issued. Until then, the trial-level data are the best available evidence but not the regulatory authority.
• Real-world tolerability in clinical practice. Trial populations are pre-screened for adherence likelihood and exclude many comorbidities. Real-world tolerability typically tracks somewhat worse than trial tolerability, as patients on multiple concomitant medications, with broader comorbidity profiles, and without trial-grade titration support encounter the drug.

Practical implications for prescribers and patients

Retatrutide is not yet FDA-approved, so the immediate practical implication is that no prescription can be written for it through legal US channels today. Compounded retatrutide is not legal in the US — retatrutide has no USP/NF monograph, is not on the FDA 503B bulks list, and is not on the FDA shortage list. The FDA has issued multiple Warning Letters in 2025 to compounders distributing retatrutide commercially. See our retatrutide approval and access timeline and the broader retatrutide triple agonist evidence review for the regulatory boundary and the published phase 2 data.

Once retatrutide is approved (anticipated H2-2027 to H1-2028 based on typical NDA timelines), the practical clinical considerations:

• Counsel patients pre-treatment that nausea, diarrhea, constipation, and vomiting are expected, especially during the 16-week titration schedule (2 → 4 → 6 → 9 → 12 mg every 4 weeks). The same patient-management strategies that work for Wegovy and Zepbound translate.
• Counsel patients pre-treatment about the dysesthesia signal. Patients should know that tingling, burning, pins-and-needles, or numbness should be reported promptly. This is novel for the class; most patients on Wegovy or Zepbound will not have heard of it.
• Plan the dose-vs-tolerability trade-off explicitly. The 12 mg dose produces the largest weight loss but the highest AE burden. The 4 mg dose is well-tolerated but gives up roughly one-third of the magnitude available at 12 mg. The 9 mg dose offers intermediate weight loss with intermediate AE rates.
• Monitor renal function during titration. The GLP-1-class warning for acute kidney injury secondary to volume contraction from severe GI losses applies here too. Patients with chronic kidney disease, on diuretics, or with a history of dehydration require closer monitoring.
• Coordinate with diabetes medications in the T2D subset once TRIUMPH-2 reads out. Hypoglycemia risk when retatrutide is combined with insulin or a sulfonylurea is the predictable class-typical concern, managed by downward titration of the background secretagogue or insulin.
• Use the same off-ramp planning as for Wegovy and Zepbound. Weight regain after stopping is the expected pattern for every GLP-1 receptor agonist. The STEP-4 data on semaglutide discontinuation and the SURMOUNT-4 data on tirzepatide discontinuation both showed substantial weight regain after treatment cessation. Plan continuation, dose tapering, or transition to a different molecule deliberately with the patient.

For patients evaluating where retatrutide may eventually fit in their treatment plan, the comparison-tool walkthrough for the closest peer drug is the Foundayo side effects comprehensive evidence review — orforglipron is the most recently approved GLP-1 for weight management and the most relevant near-term comparator for cost, access, and route-of-administration questions. The retatrutide trial data establish that the molecule will be more efficacious than Foundayo and roughly 1.4x as efficacious as Zepbound, at the cost of a higher AE burden and the novel dysesthesia signal.

Bottom line

Retatrutide’s TRIUMPH-1 pivotal phase 3 readout established two things at once. First, the molecule produces the largest mean weight loss in the GLP-1 receptor agonist class to date: −28.3% TBWL at 80 weeks on the 12 mg dose, −30.3% at 104 weeks in the BMI ≥35 subgroup^[1]. Second, the adverse-event ledger carries a class-novel dysesthesia signal (12.5% on 12 mg vs 0.9% on placebo) on top of class-typical but slightly worse gastrointestinal rates and a higher treatment discontinuation rate (11.3%) than the established injectable comparators.

For patients evaluating retatrutide once it is FDA-approved, the question is not whether retatrutide is more efficacious than Wegovy^[5], Zepbound^[6], or Foundayo^[7] — it clearly is, in cross-trial indirect comparison. The question is whether the additional weight-loss magnitude is worth the additional GI burden, the higher discontinuation rate, and the genuinely novel dysesthesia signal. That is a patient-and-prescriber-specific call that will be informed by the FDA-approved label, the TRIUMPH-2 T2D data, the TRIUMPH-3 cardiovascular outcomes data, and the post-marketing surveillance evidence that accumulates over years of broad use.

The TRIUMPH-1 manuscript is anticipated late 2026 in the New England Journal of Medicine; the FDA NDA filing is anticipated late 2026 or early 2027; FDA action is anticipated H2-2027 to H1-2028. This article will be updated as each milestone lands. Until then, the trial-level safety and efficacy data summarized here are the best available evidence on retatrutide side effects.

Important disclaimer. This article is educational and does not constitute medical advice. Retatrutide is investigational and is not FDA-approved as of May 2026; no legal US prescription channel exists for the drug today. Compounded retatrutide is not legal in the US per FDA Warning Letters issued throughout 2025. The TRIUMPH-1 pivotal manuscript was not yet PubMed-indexed at time of writing; adverse-event rates summarized here are from Eli Lilly’s investor release and the 86th ADA Scientific Sessions presentation. Comparator AE rates for Wegovy, Zepbound, and Foundayo were independently verified against the PubMed E-utilities API on 2026-05-24. This article will be updated when the peer-reviewed TRIUMPH-1 manuscript lands and again when the FDA-approved label is issued.

Last verified: 2026-05-24. Next review: triggered by (1) TRIUMPH-1 peer-reviewed publication, (2) TRIUMPH-2 or TRIUMPH-3 readout, (3) NDA filing, (4) FDA-approved label issuance — whichever comes first.