GLP-3: A Hormone That Does Not Exist, and the Molecule Sold Under Its Name

No. Searching the endocrinology literature for GLP-3 returns no human hormone, no receptor, and no gene. The name looks plausible because two numbered siblings genuinely exist, so a third seems like a reasonable extrapolation. Physiology does not work that way: the numbering reflects the order in which two specific peptides were identified inside one precursor protein, not an open-ended series waiting to be extended.

Both real peptides come from a single gene, proglucagon (GCG). The precursor protein is cut by prohormone convertases, and which fragments appear depends on the tissue doing the cutting. In pancreatic alpha cells, processing releases glucagon, while the region containing the two glucagon-like peptides stays bound together as an inactive major proglucagon fragment. In the intestinal L cells and in parts of the brain, the same precursor is processed differently: GLP-1 and GLP-2 are released as separate active peptides, while the glucagon sequence remains embedded in larger products such as glicentin and oxyntomodulin.

That inventory is complete and well mapped. A 2025 review in Physiological Reviews devoted to proglucagon-derived peptides catalogs glucagon, GLP-1, GLP-2, oxyntomodulin, glicentin, the glicentin-related pancreatic polypeptide and the intervening peptides. There is no further glucagon-like peptide to number. A hypothetical GLP-3 would require a third glucagon-like sequence in the precursor, and no such sequence exists.

This matters beyond pedantry. When a product name implies a natural hormone, it implies a known physiology, a known receptor and a known safety profile inherited from decades of endocrine research. The molecule actually sold under the GLP-3 label has none of that inheritance, because it is a designed synthetic peptide whose evidence base has to be read on its own terms. For the background biology of the hormone that does exist, see our GLP-1 overview.

GLP-2 is the quieter sibling, and knowing what it does helps explain why the numbering stops there. It is co-secreted with GLP-1 by intestinal L cells after a meal, and it acts on a distinct receptor expressed mainly in the gut. Its documented effects are intestinal rather than metabolic: it promotes mucosal growth, increases nutrient absorptive capacity, supports epithelial barrier function and slows gastric emptying.

Because of that biology, GLP-2 pharmacology developed along a completely different path from GLP-1 pharmacology. GLP-2 analogs were studied for short bowel syndrome and intestinal failure, not for body weight or glycemic control. GLP-1 pharmacology, by contrast, produced the incretin-based agents now familiar in diabetes and obesity medicine.

So the family already contains two members with two different therapeutic stories, which is precisely why a third number sounds credible to a reader skimming a product page. The logic fails on inspection. GLP-1 and GLP-2 are named for being glucagon-like sequences within one precursor. The molecule marketed as GLP-3 is not a sequence found in that precursor at all. It is a laboratory-designed peptide that happens to activate the GLP-1 receptor among others.

One further nuance is worth keeping in mind: the number in a hormone name describes molecular ancestry, while the number in a vendor name often describes nothing but marketing sequence. The two systems coincide often enough to be confusing, and this is a case where they do not coincide at all.

The label originates in research-chemical catalogs, not in the literature. Scan the listings of several suppliers and you will find the same compound under several closely related headings: GLP-3, GLP-3 Reta, GLP-3 R, GLP-3 RT, GLP-RT or simply RETA. The variation is itself informative. A real scientific designation does not drift between four spellings across vendors; a marketing shorthand does.

Two pressures appear to drive the naming. The first is search behavior. Buyers already know GLP-1 as a category, so a name that reads as the next generation in that category is commercially attractive and captures curiosity-driven searches. The second is friction avoidance. Brand names and some generic names of investigational drugs attract more scrutiny from payment processors, advertising platforms and marketplaces, so sellers of unapproved compounds frequently adopt oblique catalog labels. The same pattern shows up elsewhere in this market under names such as GLP-TZ for tirzepatide.

Note what the label is not. It is not a code name assigned by the developer. Eli Lilly's internal designation for the compound is LY3437943, and the international nonproprietary name is retatrutide. Neither the published trials, the company's filings nor the regulatory record use GLP-3 in any form.

There is also a quieter consequence: because GLP-3 is not a defined chemical identity, nothing about the name constrains what a vial contains. Two products carrying the same three-character label do not thereby share a specification, a purity standard or even a molecule. The name travels; the identity does not travel with it.

In the overwhelming majority of listings, the intended compound is retatrutide. It is a synthetic single-chain peptide of 39 amino acids, described in detail by Coskun and colleagues in Cell Metabolism in 2022, where it first appeared publicly as LY3437943. Its design began from a GIP receptor agonist scaffold that was then modified to add activity at the GLP-1 and glucagon receptors, which is why it belongs to no single hormone family.

Several structural features are deliberate. Non-proteinogenic residues (alpha-aminoisobutyric acid at positions 2 and 20, alpha-methyl-L-leucine at position 13) protect the peptide from enzymatic degradation, the C-terminus is amidated, and a C20 fatty diacid is attached at the lysine in position 17 through a gamma-glutamyl and AEEA linker. That lipid chain binds circulating albumin, which is what makes once-weekly administration feasible in clinical trials. The resulting molecule has a molecular formula of C₂₂₁H₃₄₂N₄₆O₆₈, a molecular weight near 4731 g/mol and CAS number 2381089-83-2.

A short caution follows from this. Because retatrutide is still investigational, there is no pharmacopeial monograph establishing an official reference standard for it, and nothing in a catalog name verifies that the material matches the published structure. The pharmacology summarized here describes retatrutide as characterized in peer-reviewed work, which is not the same as a claim about any particular vial. Our full retatrutide monograph covers the compound's pharmacology in depth; this page is deliberately limited to the terminology question.

A classic GLP-1 receptor agonist engages one receptor. Retatrutide engages three, and in the characterization published by Coskun and colleagues it bound the human GIP, GLP-1 and glucagon receptors with half-maximal effective concentrations of 0.0643, 0.775 and 5.79 nM respectively. The potencies are intentionally unbalanced, with the strongest activity at the GIP receptor and the weakest at the glucagon receptor.

Each arm contributes something different in preclinical and clinical models. GLP-1 receptor activation reduces food intake and slows gastric emptying while enhancing glucose-dependent insulin secretion. GIP receptor activation adds incretin effect and appears to modify nausea signaling and adipose tissue handling. The glucagon receptor arm is the conceptual novelty: glucagon raises energy expenditure and promotes hepatic lipid mobilization, which is why triple agonists were pursued as a route to greater weight reduction than incretin action alone. The same arm is also why hepatic glucose output and heart rate receive close attention in the trial safety data.

Calling such a molecule GLP-3 therefore misdescribes it twice. It implies membership in a hormone family it does not belong to, and it hides the two receptors that make the compound pharmacologically distinctive. A reader who believes they are looking at a stronger GLP-1 has misunderstood the mechanism they are reading about.

The closest real comparison is not GLP-1 at all but tirzepatide, a dual GIP and GLP-1 receptor agonist that shares the same design lineage. Retatrutide adds the glucagon component on top of that dual activity. We compare the two mechanisms and their trial data side by side in retatrutide vs tirzepatide.

The clinical record belongs to retatrutide, not to anything named GLP-3, and it is now extensive. The phase 2 obesity trial published by Jastreboff and colleagues in the New England Journal of Medicine in 2023 randomized 338 adults to placebo or one of several retatrutide amounts. At 48 weeks, least-squares mean body weight change was -8.7% in the 1 mg group, -17.1% in the combined 4 mg groups, -22.8% in the combined 8 mg groups and -24.2% in the 12 mg group, compared with -2.1% with placebo. Serious adverse events occurred in 4% of the placebo group and ranged from 0% to 6% across the retatrutide groups.

A parallel phase 2 trial in 281 adults with type 2 diabetes, published by Rosenstock and colleagues in The Lancet in 2023, reported dose-dependent reductions in HbA1c that exceeded placebo in all but the lowest group and exceeded the active comparator dulaglutide at the higher amounts, with no severe hypoglycemia and no deaths reported. A phase 2a trial in 98 participants with metabolic dysfunction-associated steatotic liver disease, published by Sanyal and colleagues in Nature Medicine in 2024, found mean relative liver fat changes at 24 weeks of -42.9%, -57.0%, -81.4% and -82.4% across ascending groups versus +0.3% with placebo, with 86% of the highest group reaching normal liver fat below 5%.

Phase 3 results arrived in 2026. TRIUMPH-1, published in the New England Journal of Medicine on September 29, 2026, enrolled 2,339 adults with obesity and followed them for 80 weeks. Under the treatment-regimen estimand reported in the paper, mean body weight reduction in the highest group was 25.0% versus 3.9% with placebo, alongside reductions in knee osteoarthritis pain and in apnea-hypopnea events in prespecified subgroups. Eli Lilly's earlier topline release described weight reduction of up to 28.3% at 80 weeks; the two figures differ because they answer different analytical questions, the company figure estimating effects under full adherence and the published figure estimating effects as the regimen was actually taken.

Three qualifications belong with those numbers. All of them come from supervised trials with eligibility criteria, titration schedules, dietary counseling and monitoring that have no counterpart outside a trial. Gastrointestinal adverse events were the most common reason for discontinuation across the program, and our article on retatrutide side effects covers that profile in detail. And none of this evidence transfers to a vial of unverified material, however the label is spelled. These figures are reported trial outcomes, not guidance for use.

The first problem is factual. A name that invokes a nonexistent hormone transfers borrowed credibility from fifty years of proglucagon research to a product that has not earned it. Readers who accept the name tend to assume the physiology is settled, when in fact what they are reading about is an investigational compound whose regulatory review has not begun.

The second problem is identity. Because GLP-3 has no chemical definition, the label cannot tell you what you are looking at. Pharmacopeial names map to monographs with reference standards and defined assay methods; catalog names map to nothing. Two vials bearing identical labels may differ in peptide identity, in salt and counter-ion content, in water content, in related substances and in actual peptide mass per vial.

The third problem is measurable. In a study published in the Journal of Medical Internet Research in 2024, Ashraf and colleagues test-purchased semaglutide products from online sellers operating without a prescription and analyzed them. Measured purity ranged from 7.7% to 14.37% against a 99% label claim, and endotoxin was detected in every sample analyzed, at 2.16 to 8.95 EU/mg. That study examined a different peptide and a small number of samples, so it does not characterize any specific supplier of retatrutide. What it does establish is that label claims in this channel can diverge from contents by an order of magnitude, which is the reason to ask for analytical evidence rather than to trust nomenclature.

The practical consequence is a shift in what you verify. A name is not evidence. A batch-specific certificate of analysis showing identity by mass spectrometry and purity by HPLC is evidence, and our guide on how to read a peptide certificate of analysis explains how to tell a meaningful document from a decorative one.

No. Nothing has ever been approved under the name GLP-3, because no regulator recognizes the term. The relevant question is the status of retatrutide, and as of October 2026 retatrutide is not approved by the FDA or the EMA for any indication. It remains an investigational compound.

Eli Lilly has stated that it plans to submit a Biologics License Application to the FDA in the first quarter of 2027, following completion of its phase 3 program across obesity, obstructive sleep apnea and knee osteoarthritis pain. A filing is not an approval, and a regulatory decision would follow a review period after submission. Until a decision is issued, any product sold to the public under either retatrutide or GLP-3 is an unapproved compound irrespective of how it is described.

That is why vendor pages carry research-use-only language. In the United States and the European Union, material sold on that basis is not a medicine: it is not manufactured under the quality systems required for human use, it is not reviewed for safety or efficacy, and it comes with no legal guarantee of identity or sterility. Legal treatment also varies by jurisdiction, and in several countries importing or possessing such material raises separate questions that fall outside the scope of this article.

If you are evaluating listings in this market, the relevant criteria are supplier transparency and analytical documentation rather than nomenclature. Our overview of where retatrutide is sold and how suppliers differ addresses that landscape. Nothing here should be read as an endorsement of personal use of an unapproved compound. Discuss any metabolic or weight concern with a licensed healthcare professional, who can speak to approved options.

Start with identity. A listing that only says GLP-3 has told you nothing testable. A credible one names the compound as retatrutide or LY3437943, states the CAS number 2381089-83-2 and gives a quantity of peptide per vial. If a page avoids the nonproprietary name entirely, that avoidance is a data point about how the seller handles identity in general.

Then ask for analytical evidence tied to the specific batch you would receive. Three elements carry most of the weight: identity confirmed by mass spectrometry, with an observed mass consistent with the published structure near 4731 g/mol; purity by HPLC with the actual chromatogram rather than a bare percentage; and the lot number on the document matching the lot on the vial. A certificate with no lot number, no date, no method section and no laboratory identity is not a certificate in any useful sense.

Be alert to specific failure modes that a name cannot reveal. Peptide content by weight differs from gross vial weight because of counter-ions and residual water, so a vial can contain meaningfully less peptide than its label implies even when purity looks acceptable. Sterility and endotoxin are separate questions from chemical purity and are rarely documented at all in this channel. And a certificate supplied by the seller rather than an independent laboratory tests the seller's claim, not the material.

None of this makes an unapproved compound equivalent to a reviewed medicine, and it is not meant to. It simply distinguishes a listing that can be checked from one that cannot. Our certificate of analysis guide walks through a real document line by line. This article is educational and does not constitute medical advice. Retatrutide is an investigational compound that is not approved for human use, and research peptides should be handled only in appropriate laboratory settings by qualified personnel.