Incretins Explained: GLP-1, GIP and Glucagon

Start with the gut, not the drug

An incretin is a hormone your gut releases when food arrives, which then tells the pancreas to release insulin. The word predates the drugs by decades.

Two hormones carry the name: glucagon-like peptide 1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP). A third hormone, glucagon, keeps turning up in this class of drugs and is not an incretin at all — it does close to the opposite job.

That is the whole vocabulary. What follows is what each one does, and what happened when drug developers pushed on it.

GLP-1: the one with the most human data behind it

GLP-1 is released from cells in the small intestine and colon after eating. It increases insulin release when glucose is high, suppresses glucagon, slows how fast the stomach empties, and reduces appetite through receptors in the brain¹.

The critical fact about native GLP-1 is how briefly it exists. The label for one GLP-1 drug states that endogenous GLP-1 has a half-life of 1.5 to 2 minutes, because the enzyme dipeptidyl peptidase 4 and neutral endopeptidases degrade it almost immediately².

The GLP-1 drugs in this class are, at bottom, answers to that one problem. We cover the engineering in half-life and dosing frequency.

Human results: in STEP 1, 1,961 adults with overweight or obesity and no diabetes were randomised 2:1 to once-weekly semaglutide 2.4 mg or placebo for 68 weeks, and mean body weight changed by −14.9% against −2.4%, funded by Novo Nordisk³. Our semaglutide page carries the full record.

GIP: the receptor the field disagrees about

GIP comes from the upper small intestine and, like GLP-1, increases insulin release when glucose is high. Beyond that the story gets genuinely unsettled.

Tirzepatide activates both the GIP and GLP-1 receptors. In SURMOUNT-1, 2,539 adults were randomised to tirzepatide or placebo for 72 weeks, and the highest dose arm produced −20.9% against −3.1% on placebo, funded by Eli Lilly⁴.

Maridebart cafraglutide does the opposite at GIP: it is a peptide-antibody conjugate that activates the GLP-1 receptor while *blocking* the GIP receptor. In a 592-participant phase 2 trial funded by Amgen, the obesity cohort's mean weight change at week 52 ranged from −12.3% to −16.2% depending on dose and escalation schedule⁵.

So one drug turns GIP on, another turns it off, and both lose weight. Anyone telling you the GIP mechanism is settled is ahead of the evidence. We take that contradiction apart in why a triple agonist is not simply better than a dual.

Glucagon: the counter-hormone

Glucagon is the hormone that raises blood glucose — it is what your pancreas releases when you have not eaten. Adding it to a weight-loss drug sounds backwards, and the reasoning is that glucagon also increases energy expenditure and drives fat breakdown in the liver.

Survodutide activates the glucagon and GLP-1 receptors. Its phase 2 dose-finding trial randomised 387 participants across four doses and placebo for 46 weeks, and reported −14.9% at the highest dose against placebo, funded by Boehringer Ingelheim⁶.

Its phase 3, SYNCHRONIZE-1 (NCT06066515), randomised 725 adults to survodutide 3.6 mg, 6.0 mg or placebo for 76 weeks. Mean weight change under the treatment-regimen estimand was −12.2%, −13.0% and −5.4%⁷.

Note what did *not* happen: adding a glucagon arm to a GLP-1 backbone did not produce a step change past the GLP-1-only numbers. Our survodutide page holds the detail.

Three receptors at once

Retatrutide activates all three — GIP, GLP-1 and glucagon. Its phase 2 obesity trial (NCT04881760) randomised 338 adults and reported a least-squares mean weight change of −24.2% in the 12 mg arm at 48 weeks against −2.1% on placebo, funded by Eli Lilly⁸.

Its first reported phase 3, TRANSCEND-T2D-1, ran 40 weeks in 537 adults with type 2 diabetes. The primary endpoint there was HbA1c, not weight; weight was a key secondary and came in at −15.3% in the 12 mg arm against −2.6% on placebo⁹.

Those are different populations, different durations and different endpoint hierarchies — not the same experiment run twice. Retatrutide is investigational; the retatrutide page tracks it as it reports.

What the receptor count does not tell you

Line the numbers up and the pattern is not "more receptors, more weight loss."

One receptor, GLP-1 only — semaglutide 2.4 mg, −14.9% at 68 weeks in 1,961 people³.

Two receptors, GIP plus GLP-1 — tirzepatide, −20.9% at 72 weeks in 2,539 people⁴.

Two receptors, glucagon plus GLP-1 — survodutide, −13.0% at 76 weeks in 725 people⁷.

Two receptors, GIP blocked plus GLP-1 on — maridebart cafraglutide, up to −16.2% at 52 weeks in a phase 2 cohort of 465⁵.

Three receptors — retatrutide, −24.2% at 48 weeks in a phase 2 of 338⁸, and −15.3% at 40 weeks in a phase 3 of 537 people with type 2 diabetes⁹.

The two largest figures come from the two smallest and earliest trials. That is not a coincidence, and it is not a claim that the drugs are equivalent — it is a warning that these numbers were produced by trials that differ in population, duration, estimand and phase, and cannot be ranked as if they were arms of one experiment.

What is actually settled, and what is not

Settled, in the sense that large randomised trials have reported it: activating the GLP-1 receptor produces substantial, sustained weight reduction against placebo, and does so across many thousands of randomised participants.

Not settled: whether GIP should be switched on or off. As of August 2026 both directions have produced double-digit weight reduction in randomised trials. A PubMed search in August 2026 for a randomised comparison of GIP receptor agonism against GIP receptor antagonism on a shared GLP-1 backbone returned no such trial — and that is the experiment that would answer it.

Also not settled: whether glucagon agonism adds anything a higher GLP-1 dose would not. Answering that needs a head-to-head. A ClinicalTrials.gov search on "survodutide" in August 2026 returned 24 studies; the phase 2 and phase 3 records listing efficacy arms compare survodutide doses against placebo, and the two records listing a semaglutide arm — NCT06745284 and NCT05202353 — are phase 1.

For how these compounds line up against a specific goal rather than against each other, see weight loss.