Semaglutide, Tirzepatide and Retatrutide: Similarities, Differences and Evidence

Semaglutide, tirzepatide and retatrutide are often discussed as if they were three strengths of one treatment. They are not. All three are incretin-based molecules given once a week as an injection, but they differ in the receptors they engage, in the approved products they appear in, in which indications each regulator has accepted — and in what the evidence can and cannot prove. Retatrutide, unlike the other two, is not an approved medicine anywhere: it is investigational.

This article sets out the shared mechanism, the product and indication differences, the approval picture in the United States, the European Union and South Africa, and the trial evidence, with the caveats that make cross-trial numbers readable. It does not crown a winner: no ranking survives the population and endpoint differences between these programmes, and choosing or switching these medicines is a prescribing decision, not a statistics exercise.

Receptor mechanisms: one hormone family, three signatures

GLP-1 (glucagon-like peptide-1) is a gut hormone released after eating: it prompts insulin secretion when glucose is high, reins in inappropriate glucagon release, slows stomach emptying and acts on appetite centres in the brain. Making these effects long-acting — the shared feature of all three molecules — reduces appetite and food intake. That common ground is also why the three share most of their side-effect profile.

Semaglutide (marketed as Wegovy and Ozempic) is a selective GLP-1 receptor agonist: it activates that one receptor and nothing else. This single-receptor design is the reference point the newer molecules are measured against.

Tirzepatide (Mounjaro, Zepbound) adds the receptor for glucose-dependent insulinotropic polypeptide (GIP), the other major incretin, which influences lipid handling and insulin sensitivity. In controlled trials the dual agonist outperformed the selective agonist on glucose and weight measures — the reason the two-receptor design matters — even though the mechanism behind the larger effect is still being worked out.

Retatrutide, Eli Lilly’s investigational triple agonist, adds the glucagon receptor. Glucagon mobilises stored energy and raises hepatic glucose output, which sounds paradoxical in a metabolic medicine; the design hypothesis is that pairing glucagon-receptor activity with the incretin arms shifts energy balance further while the GLP-1/GIP components hold glucose effects in check. Trials so far have produced the largest average weight effects in the field — but this is a hypothesis being tested, not a labelled, regulator-reviewed mechanism. Retatrutide Explained covers the design in depth. All three are long-acting, dose-escalated injections; a mechanism sets out what a molecule can plausibly do, and outcomes must come from trials.

One molecule, several products

One molecule, several brands: Ozempic’s label centres on type 2 diabetes, Wegovy’s on weight management. The US Wegovy label has since expanded well beyond its 2021 obesity approval: long-term weight management in adults and in young people aged 12 and older with obesity, cardiovascular risk reduction in adults with established cardiovascular disease and either obesity or overweight (the SELECT trial), and — under accelerated approval — metabolic liver disease (MASH) with moderate-to-advanced fibrosis. A 7.2 mg weekly maximum entered the label for adults who tolerate 2.4 mg, alongside a 25 mg once-daily tablet (recommended for EU approval in May 2026).

Tirzepatide splits the same way. Mounjaro is the type 2 diabetes brand; Zepbound is the obesity brand, and in the US it also carries the first drug approval for moderate-to-severe obstructive sleep apnea in adults with obesity (December 2024). In August 2026 the FDA added a further Mounjaro indication: lowering the risk of major adverse cardiovascular events in adults with type 2 diabetes at high cardiovascular risk, on the strength of SURPASS-CVOT. Approved dose ceilings can differ between products of the same molecule — Ozempic tops out at 2 mg weekly while Wegovy reaches 7.2 mg — so the same ingredient never means the same prescription — and retatrutide has no brand, product or label in any country at all.

Approval by jurisdiction

Regulators decide indication by indication, so the same brand can mean different labelled uses in different countries.

Molecule and receptors	United States (FDA)	European Union (EMA)	South Africa (SAHPRA)
Semaglutide GLP-1 receptor agonist	Ozempic: type 2 diabetes; CV-risk reduction in T2D; kidney protection (T2D, CKD). Wegovy: weight management; CV risk reduction; MASH (accelerated); also tablets	Wegovy: weight management (adults; adolescents ≥12); SELECT data in product information, no separate CV indication. Ozempic: type 2 diabetes	Ozempic: type 2 diabetes — not registered for weight loss. Wegovy: weight management in adults, to 2.4 mg (launched August 2025)
Tirzepatide GIP + GLP-1 receptor agonist	Mounjaro: type 2 diabetes; MACE risk reduction added Aug 2026. Zepbound: weight management; moderate-to-severe OSA with obesity	Mounjaro: type 2 diabetes and weight management under one brand (Nov 2023)	Mounjaro: registered; PI covers type 2 diabetes and weight management (Oct 2025)
Retatrutide GIP + GLP-1 + glucagon receptor agonist	Not approved — investigational	Not approved — investigational	Not registered — investigational

Two European wrinkles. The EU covers both tirzepatide indications under one brand (no “Zepbound” split), with sleep-apnoea trial data noted in the product information without being made separate indications. And the EMA folded SELECT’s cardiovascular findings into Wegovy’s existing weight-management indication rather than granting a separate claim, unlike the US label.

Retatrutide’s status is identical in every jurisdiction — no product, no label. Eli Lilly plans a Q1 2027 biologics licence application to the FDA; until a regulator completes its review, retatrutide has no labelled indications, doses or safety profile. Registration lists change, so current status belongs on the regulators’ own registers — SAHPRA’s repository for South Africa. The registered-versus-research-compound line, and what it means for what is sold, is the subject of Approved medicines versus research peptides.

What trials can compare

Only a study that randomises people to two medicines under identical rules can compare the two. For these three molecules, exactly two such head-to-head comparisons of approved options exist — SURMOUNT-5 in obesity and SURPASS-2 in type 2 diabetes — and neither involves retatrutide.

SURPASS-2, the type 2 diabetes head-to-head, compared tirzepatide (5, 10 and 15 mg) against semaglutide 1 mg over 40 weeks on metformin; every tirzepatide dose won on glucose control and weight. The 1 mg semaglutide dose sits in the Ozempic range, which matters when reading it.

Everything else is separate trials — important, but never a head-to-head. The landmarks: STEP 1, −14.9% mean weight on semaglutide 2.4 mg at 68 weeks; SELECT, a 20% relative reduction in major cardiovascular events versus placebo; SURMOUNT-1, about 21% weight loss with tirzepatide 15 mg at 72 weeks; SURPASS-CVOT, Mounjaro non-inferior (HR 0.92) to dulaglutide on cardiovascular events; and retatrutide’s phase 2 trial, with −24.2% at 48 weeks.

The rest of the TRIUMPH programme: TRIUMPH-2 (obesity with type 2 diabetes) reported −20.8% mean weight loss at 80 weeks; TRIUMPH-3 (obesity with established cardiovascular disease) reported −22.6% with a cardiovascular composite that was directionally favourable but not statistically significant; TRIUMPH-4 reported knee-osteoarthritis pain results in December 2025. Retatrutide Trial Results: What the Headlines Leave Out reads these in detail.

What separate trials cannot prove

League-table arithmetic between trials fails quietly. Populations differ — STEP 1 excluded diabetes, TRIUMPH-2 required it, TRIUMPH-1 averaged near BMI 40. Durations differ (68, 72, 80 weeks). Dosing philosophies differ (maximum tolerated dose versus fixed doses). Even the estimands — how trials handle people who stop treatment — differ. A 28.3% result at 80 weeks cannot be read as sitting in one column with 20.2% at 72 weeks or 14.9% at 68 weeks.

Labels move while completed trials stand still: the 7.2 mg semaglutide dose now permitted in the US label was not the dose SURMOUNT-5 tested, so even that head-to-head describes the label as it stood in 2025.

Most importantly, body weight is a surrogate. The record holds two proven clinical-outcomes claims here and none for retatrutide: SELECT proved cardiovascular benefit for semaglutide 2.4 mg against placebo; SURPASS-CVOT proved non-inferiority, not superiority, for tirzepatide; TRIUMPH-3’s cardiovascular composite fell short of significance. Outcomes-level claims about retatrutide — hearts, joints, kidneys — remain unestablished.

Safety context

Shared class profile first: gastrointestinal effects — nausea, vomiting, diarrhoea, constipation — are the most common adverse events across all three, clustering during dose escalation. Labels carry warnings for pancreatitis, gallbladder disease and gallstones (risk rises with rapid weight loss), acute kidney injury from dehydration and increases in heart rate. Low blood sugar is a risk mainly when insulin or sulfonylureas are taken alongside.

Serious warnings are consistent too. All three molecules are presumed to work through pathways that caused, or in trials caused, thyroid C-cell tumours in rodents: the US labels carry a boxed warning and the South African product information opens with the same warning, and a personal or family history of medullary thyroid carcinoma or MEN 2 syndrome rules the medicines out. Tirzepatide is not recommended in severe gastroparesis; semaglutide’s label directs monitoring of diabetic retinopathy in type 2 diabetes and carries warnings about severe gastrointestinal reactions. A prescriber works from the full label, not from this summary.

Retatrutide has no label at all. What is known comes from trials: class-consistent gastrointestinal effects and heart-rate increases, a growing but unreviewed safety database. Anything sold as “retatrutide” outside a clinical trial is by definition unlicensed, with unverifiable identity, purity and dose — SAHPRA has repeatedly warned about falsified and unregistered GLP-1 products sold online. Most of the weight removed is fat, but some is lean mass — Weight Loss and Muscle: Why Body Composition Matters covers it.

The role of qualified care

These medicines are prescription-only — Schedule 4 in South Africa. A prescriber matches each molecule to the person: whether a labelled indication exists for them (diabetes status changes this), what is ruled out (thyroid cancer history, gastroparesis, pregnancy plans), what else is taken (insulin and sulfonylureas change the risk picture), and how escalation and monitoring run. Indications, doses and ceilings differ across jurisdictions — one more reason the decision belongs with someone who can read the local label.

Which brings the comparison back to scale. The evidence supports clear differences in mechanism, labelled use and weight effect within each trial. It does not license a default switch, a dose-stacking experiment or a mail-order purchase of an unapproved molecule — and nothing reviewed here crowns one of the three. The honest answer to “which is strongest” is “strongest for whom, for which labelled purpose, measured how, over how long.”

Apply the same reading habits to any trial on this site, including those cited here: read the trial interpretation guide.

Suggest a correction