Tirzepatide: how a dual GIP/GLP-1 agonist reset expectations for obesity pharmacology

It's a 39-amino-acid synthetic peptide with a C20 fatty diacid hooked onto lysine 20 ^[2]. Eli Lilly sells it as Mounjaro for type 2 diabetes (FDA-approved May 2022) ^[3] and as Zepbound for chronic weight management (approved November 2023, with an obstructive sleep apnea indication added in December 2024) ^[4]. The fatty acid tail latches onto albumin in circulation, and that's the trick that, in practice, stretches tirzepatide's half-life to roughly five days ^[5] and makes once-weekly dosing tolerable.

Same engineering idea Lilly borrowed for semaglutide's weekly formulation, applied here to a different peptide backbone.

The molecule started as a GIP scaffold. Lilly's chemists then grafted GLP-1 receptor agonism onto it. That order matters, because it explains the pharmacology you end up with: tirzepatide binds GIPR with affinity essentially identical to native GIP, but its GLP-1R affinity sits about five times weaker than endogenous GLP-1 ^[6]. In cell-based signaling assays, the EC50 for GIPR activation falls in the low-picomolar range (comparable to native GIP), while the GLP-1R EC50 is considerably higher than that of native GLP-1.

So it's lopsided.

In the field's vocabulary, tirzepatide is an imbalanced agonist tilted toward GIPR. It also shows biased signaling at GLP-1R, preferentially recruiting cAMP over β-arrestin, and that quirk may reduce receptor internalization and keep GLP-1 signaling going even though raw affinity is weaker on that side.

At the 15 mg dose, peak plasma concentrations are high enough to exceed the GIPR EC50 by a substantial margin and the GLP-1R EC50 by a far smaller, but still meaningful, one. In practice, both receptors are activated well above EC50. The lopsided affinity isn't an accident; it's by design, and in the patient both pathways end up saturated.

Two hormones, related but not identical. GLP-1 comes from intestinal L-cells after a meal; it nudges insulin out of the β-cell in a glucose-dependent way, shuts down glucagon, slows the stomach, and reaches into the hindbrain and hypothalamus to dial down appetite. GIP, released from K-cells higher up in the duodenum and proximal jejunum, also pushes insulin. But what it does to fat tissue, and how it acts on satiety circuits centrally, doesn't really mirror GLP-1.

GIP has been the stranger sibling for decades.

The trouble is that in type 2 diabetes, β-cells go relatively deaf to GIP on its own, which is why isolating GIP as a therapy never really panned out and why nobody quite knew what to do with it until it was paired up.

The interesting finding of the past few years is that GIPR co-activation, on top of GLP-1R agonism, appears to do several things at once: it improves insulin sensitivity beyond what GLP-1R activation alone produces, it supports adipose tissue remodeling (in animal models and inferentially in humans from lipid and body composition data), and it may attenuate the gastrointestinal side effects that limit GLP-1 monotherapy. That last claim is the contested one. Across separate trials, GI adverse events with tirzepatide were not lower than those reported in semaglutide trials; if anything, the higher tirzepatide doses produce more nausea than mid-range semaglutide. What seems true is that GIPR co-activation lets the GLP-1 component do less work for the same anorectic effect, which may be why patients tolerate tirzepatide's weight loss magnitude at all.

Worth flagging: rodent data show GIP reduces food intake and body weight, but those effects have not been cleanly demonstrated in humans. The mechanism by which GIPR contributes to tirzepatide's superior weight loss in people remains, in 2026, only partially worked out. Some investigators argue the imbalanced pharmacology is the key. Others argue it's the longer half-life and higher exposure relative to comparators. The honest answer is we don't yet have a definitive mechanistic decomposition.

What is settled: patients taking tirzepatide experience reduced appetite, earlier satiety, slowed gastric emptying, and improved postprandial glucose, along with suppressed glucagon and, over months, a downward reset of body weight set point that mirrors what the bariatric surgery literature describes. The downward reset is the part that matters clinically. It's also the part that reverses when the drug is stopped, which is the central honesty problem of this whole drug class.

The cleanest comparison data come from SURMOUNT-5, the head-to-head Lilly-sponsored trial published in the New England Journal of Medicine in May 2025. Adults with obesity but without diabetes were randomized 1:1 to maximum tolerated tirzepatide (10 or 15 mg) or maximum tolerated semaglutide (1.7 or 2.4 mg) for 72 weeks. The primary endpoint, percent change in weight, came in at -20.2% for tirzepatide versus -13.7% for semaglutide (p<0.001) ^[7]. Waist circumference dropped substantially more with tirzepatide than with semaglutide, by a difference of around 5 cm. The proportion of patients hitting ≥15%, ≥20%, and ≥25% weight loss was meaningfully higher in the tirzepatide arm at every threshold.

SURPASS-2, the earlier diabetes head-to-head, pointed the same way on glycemic control: tirzepatide delivered greater HbA1c reductions and meaningfully more weight loss than semaglutide across the doses tested. One caveat. That trial predated the semaglutide 2 mg dose and used an older comparator dose, which makes the comparison less clean than the headline numbers suggest.

A 2024 propensity-matched cohort of 18,386 adults, drawing on real-world prescribing data, found hazard ratios of 1.76 for ≥5% weight loss with tirzepatide versus semaglutide. At ≥10%, the ratio climbed to 2.54. At ≥15%, 3.24 ^[8]. The gap widens the further out on the weight-loss curve you look.

Real-world numbers run softer than trial numbers. They always do. But the direction tracks.

Is this just a dose-equivalence problem? Partly. Comparing 15 mg of tirzepatide to 2.4 mg of semaglutide is comparing the top of one drug's range to the top of another's, and the two molecules are not on the same potency scale. But the proportional weight loss is larger than dose adjustment alone explains, and the glycemic effect, driven primarily by β-cell function and insulin sensitivity rather than dose-saturable receptor pharmacology, is robustly larger across the SURPASS program. Something about the second receptor, the longer half-life, the imbalanced pharmacology, or some combination, is doing real work. We don't yet know the precise breakdown. We do know the outcome.

Mounjaro received FDA approval on May 13, 2022, for adjunctive treatment of type 2 diabetes alongside diet and exercise ^[9]. Zepbound followed on November 8, 2023, for chronic weight management in adults with a BMI of 30 or higher, or 27 or higher with at least one weight-related comorbidity ^[10]. In December 2024, the FDA expanded Zepbound's label to include moderate-to-severe obstructive sleep apnea in adults with obesity ^[11], the first pharmacotherapy approved for that indication and a meaningful expansion of the therapeutic envelope. The Mounjaro label carries a boxed warning for thyroid C-cell tumors based on rodent data, with contraindications for personal or family history of medullary thyroid carcinoma and multiple endocrine neoplasia type 2.

The two brands are the same molecule in the same strengths and devices; what differs is the approved use, and with it the trials, the coverage and the price. Zepbound vs Mounjaro sets the two labels side by side.

The SURPASS program ran tirzepatide through the standard type 2 diabetes gauntlet: monotherapy in drug-naïve patients (SURPASS-1), a head-to-head against semaglutide 1 mg (SURPASS-2), add-on to insulin glargine (SURPASS-3), against titrated insulin glargine (SURPASS-4), and as add-on to insulin glargine plus metformin (SURPASS-5). Extension and switching studies have kept reporting through 2024-2025. Five core trials. Across them, tirzepatide at weekly doses cut HbA1c by meaningful margins, with the higher doses pulling the bigger drops almost every time.

Weight came down by several kilograms. For a 100 kg patient at the top end, that's roughly a quarter of body weight gone on what is, on paper, a diabetes drug. The figure endocrinologists keep circling, though, is a different one: depending on dose and trial, a substantial share of treated patients, in some studies a majority, dropped below an HbA1c of 5.7%, the upper edge of the normoglycemic range. Not "better controlled diabetes." Numbers that, in a screening clinic, would read as no diabetes at all.

And the weight numbers tracked the glycemic ones. Across the trial program, substantial proportions of patients hit weight loss of 20% or more, a magnitude the field hadn't seen from a single agent in people with diabetes, who as a rule lose weight more grudgingly than people without it.

Cardiovascular safety, the regulatory hurdle that has ended several diabetes drug candidates, was met. Meta-analytic MACE hazard ratios across the program sat below 1.3 ^[12], the conventional threshold. The dedicated outcomes trial, SURPASS-CVOT, compared tirzepatide with dulaglutide, itself a drug with proven cardiovascular benefit, in 13,299 patients with type 2 diabetes and established cardiovascular disease ^[13]. Over a median of about four years, tirzepatide was non-inferior on heart attack, stroke and cardiovascular death combined (hazard ratio 0.92, 95.3% CI 0.83 to 1.01); superiority over dulaglutide was not established ^[25]. On that basis the FDA added a cardiovascular indication to the Mounjaro label in August 2026 ^[25].

Thousands of adults signed up for SURMOUNT-1, published in NEJM in June 2022, all with obesity or overweight plus at least one weight-related condition, none diabetic. At 72 weeks, mean weight loss landed at 15.0% on 5 mg, 19.5% on 10 mg, and 20.9% on the 15 mg dose in the primary analysis ^[14]. Per-protocol, the top dose pushed past that, hitting 22.5% ^[15].

About 9 in 10 participants on tirzepatide lost weight.

The 15 mg arm shed roughly 52 lbs on average. For someone starting at 250 lbs, that's about a fifth of starting body weight, magnitudes previously associated with bariatric surgery.

SURMOUNT-2 took the harder population, adults with obesity and type 2 diabetes, where weight loss is structurally more difficult, and at 10 and 15 mg produced reductions around 12.8% ^[16] and 14.7% respectively, against 3.2% on placebo. This is the trial that led W. Timothy Garvey to propose the term "second-generation obesity medications" for drugs that produce 15%+ weight loss durably in diabetics ^[17], a population in which first-generation agents typically plateaued earlier and lower.

SURMOUNT-3 layered tirzepatide onto a successful 12-week intensive lifestyle intervention. After participants had already lost an average of 6.9% on lifestyle alone, those randomized to tirzepatide lost an additional 18.4% from randomization to week 72; the placebo arm regained 2.5% ^[18]. SURMOUNT-4 examined weight regain after withdrawal, the data point that anchors the long-term conversation about whether GLP-1-class drugs are chronic therapy or temporary intervention. The answer in the trial was unambiguous: stopping the drug means losing most of the benefit. SURMOUNT-5, the semaglutide head-to-head already covered, closed the comparative loop.

Tirzepatide comes in single-use pens at 2.5, 5, 7.5, 10, 12.5, and 15 mg ^[19]. The label says start at 2.5 mg weekly for four weeks, a non-therapeutic introductory dose by design, then move to 5 mg for at least another four weeks, then step up in 2.5 mg increments no sooner than every four weeks until you hit maintenance ^[20]. Maintenance is 5, 10, or 15 mg ^[21].

Titration is the whole tolerability strategy here.

Most of the GI trouble (nausea, vomiting, the constipation-or-diarrhea coin flip) shows up during escalation, not at steady state. Climbing too fast is the single most common reason patients quit. Climbing too slowly delays benefit but rarely hurts anyone. Reading the trial discontinuation curves alongside what our editorial team has watched play out across several DTC providers, the patients who do best are the ones whose prescriber parks them at a tolerable dose instead of reflexively chasing 15 mg. In practice, plenty of people get most of their weight loss at 10 mg and never need the ceiling.

Some people go the other way and stay below the starting dose, or stretch the interval between doses, to spend less or to avoid nausea. No trial has tested that, and GLP-1 microdosing covers what is known and where it goes wrong.

A practical note from FDA postmarketing: adverse events linked to compounded tirzepatide products often involved off-label dosing schedules, including larger single doses, more frequent administration, or accelerated titration. The label exists for a reason, and the reason is that the GI system tolerates this drug class only when it has weeks to adapt.

For Zepbound (weight management): adults with obesity, or adults with overweight with at least one weight-related condition. That includes type 2 diabetes, hypertension, dyslipidemia, obstructive sleep apnea, or cardiovascular disease, the usual suspects that push someone from "overweight" into a category insurers will actually cover.

Mounjaro is the other label. In practice it's the same molecule, prescribed for type 2 diabetes as an adjunct to diet and exercise, which is the standard regulatory phrasing for "the drug works better if you also change what you eat." Since December 2025 that covers children 10 and older as well as adults, and since August 2026 Mounjaro is also indicated to reduce the risk of heart attack, stroke and cardiovascular death in adults with type 2 diabetes at high risk of them ^[25].

Not indicated for type 1.

And it hasn't been studied in patients with a history of pancreatitis, so prescribers tend to steer around that group rather than guess.

Contraindications are absolute: personal or family history of medullary thyroid carcinoma, and MEN type 2. The basis goes like this: rodent C-cell tumors showed up in the toxicology work, and although nothing similar has surfaced in humans at scale, the contraindication stands.

Read the FDA label in order and the warnings stack up fast. Severe GI disease, gastroparesis included. Acute kidney injury, almost always downstream of dehydration from vomiting or diarrhea. Acute gallbladder disease, because dropping weight quickly is itself a stone-forming setup. Acute pancreatitis. Hypersensitivity, up to anaphylaxis. Hypoglycemia, but really only when you stack tirzepatide on insulin or a sulfonylurea. Diabetic retinopathy worsening in people who already have retinopathy on the books. And a suicidal ideation precaution that the FDA applies across the whole anti-obesity class, not tirzepatide specifically.

Pregnancy sits in its own category. The label says tirzepatide may cause fetal harm and should be stopped when pregnancy is recognized ^[26]. It also affects contraception: because the drug slows stomach emptying, it can reduce how well oral contraceptives are absorbed, so the label advises switching to a non-oral method, or adding a barrier method, for 4 weeks after starting and for 4 weeks after each dose increase ^[26]. Anyone planning a pregnancy should talk to their prescriber about when to stop.

Who should pause before starting: anyone with a history of pancreatitis, established severe gastroparesis, active gallbladder disease, eating disorders, or plans for pregnancy in the near term. Patients on insulin or sulfonylureas need those doses trimmed to avoid hypoglycemia.

The figures from both US labels, side by side with semaglutide's, are set out in GLP-1 side effects.

Gut stuff dominates. Across the SURMOUNT and SURPASS programs, the events that showed up most were nausea, diarrhea, decreased appetite, vomiting, constipation, dyspepsia, and abdominal pain. Mostly mild to moderate. Mostly clustered during the dose-escalation window, when the body hasn't yet acclimated to each new step.

And here's the bit that catches people off guard: discontinuation rates for adverse events in the major trials sit in the high single digits, roughly on par with semaglutide and well below what you'd guess from how often patients report feeling queasy. In the SURMOUNT-5 head-to-head specifically, GI-related discontinuations trended slightly lower with tirzepatide than with semaglutide, a different framing from the across-trial comparison above. In practice, most people ride it out.

Serious adverse events are rarer but worth naming. Acute pancreatitis is the one prescribers watch hardest; the absolute incidence is low, but the signal is real and the consequences serious. Gallbladder disease and cholelithiasis are more common than pancreatitis, and partly mechanical: rapid weight loss promotes gallstones whether or not a drug is involved. Acute kidney injury shows up almost entirely in patients who've been vomiting hard or running severe diarrhea long enough to get volume-depleted. Hydration is the prevention. Angioedema and other hypersensitivity reactions have been reported too, though rarely.

Then there's FAERS.

The FDA's postmarketing adverse event system flags chronic kidney disease and nausea among the most frequently reported events for tirzepatide. But FAERS is passive surveillance, and it overweights whatever patients and clinicians happen to file spontaneously, which means you can't pull an actual incidence rate out of the raw counts, no matter how tempting the numbers look. What it's useful for is signal detection, basically a sanity check on whether the trial safety profile holds up once the drug is loose in the world. So far, it does.

A post hoc analysis of nutritional status in SURMOUNT-1 through 4, published in Obesity Pillars in early 2026 (Almandoz et al., large pooled cohort), found that macronutrient malnutrition and vitamin deficiency-related adverse events were rare; albumin and total lymphocyte count remained stable; and few participants dropped to a BMI below 18.5. The drug produces weight loss but doesn't, in monitored populations, produce malnutrition. Whether that holds in unmonitored DTC populations on prolonged high doses is a different question, and one the field has not yet answered.

Across SURPASS, tirzepatide produced consistent improvements in lipid profile (triglyceride reductions, modest HDL increases, ApoB reductions) and blood pressure (systolic reductions of 5-8 mmHg at therapeutic doses), along with markers of insulin sensitivity. The drug increases adiponectin levels, a finding distinct from semaglutide's profile and possibly attributable to the GIPR arm of activity. Hepatic steatosis improves, with MRI-PDFF reductions in dedicated substudies that have driven the parallel investigation of tirzepatide for MASH (metabolic dysfunction-associated steatohepatitis).

The cardiovascular outcomes question now has an answer. SURPASS-CVOT showed tirzepatide non-inferior to dulaglutide on major cardiovascular events, without proving it better, and since August 2026 Mounjaro carries a cardiovascular risk-reduction indication for adults with type 2 diabetes at high risk ^[25]. That puts it alongside semaglutide and dulaglutide among GLP-1-class drugs with a cardiovascular claim on the label; the indication is on Mounjaro, not Zepbound. The OSA approval in December 2024 confirmed what the trial subgroup data had been showing: the drug's effects on weight and on upper airway anatomy translate into measurable apnea-hypopnea reductions, not just weight on a scale.

Through 2023 and most of 2024, compounded tirzepatide was widely available through 503A state-licensed pharmacies and 503B outsourcing facilities, legally compounded under the FDA's drug shortage framework. Telehealth-to-pharmacy DTC pipelines emerged at scale. Prices in the $150-$450/month range made tirzepatide accessible to patients for whom Lilly's $1,000+ list price was out of reach. That market is now gone.

The timeline, briefly. On October 2, 2024, the FDA initially declared the tirzepatide shortage resolved. The Outsourcing Facilities Association filed suit; on December 19, 2024, the FDA re-evaluated and again declared the shortage resolved, this time with structured enforcement timelines: February 18, 2025 for 503A pharmacies, March 19, 2025 for 503B outsourcing facilities. A district court denied the OFA's preliminary injunction motion on March 5, 2025, and upheld the FDA's shortage determination on May 7, 2025 in Outsourcing Facilities Association v. FDA. The case is on appeal to the Fifth Circuit; the enforcement posture is settled. On April 30, 2026, the FDA went further and proposed permanently excluding semaglutide, tirzepatide, and liraglutide from the 503B Bulks List, the regulatory mechanism that governs which substances large-scale compounders may use as starting materials. Public comment was extended to 30 July 2026 and has closed; no final decision had been published as of 15 October 2026.

Mass-compounded tirzepatide is, as of 2026, illegal. The FDA has issued more than 135 warning letters to GLP-1 compounders and telehealth companies since September 2025. A narrow 503A exception remains for genuinely patient-specific compounding, situations where a clinical evaluation establishes that the FDA-approved drug is not suitable for a specific patient, typically for excipient allergy or dosage form requirement. That exception is not a workaround for ordinary access. Anything labeled "compounded tirzepatide" sold through telehealth channels in 2026 is operating outside the regulatory framework.

We would be cautious here. The pharmacovigilance data on compounded tirzepatide is thin by design, since 503A pharmacies have no federal adverse event reporting requirement, but FDA postmarketing has flagged adverse events including dosing errors at concentrations that differ from the FDA-approved 2.5 mg/0.5 mL formulation. Some products on the market reportedly contain salt forms (tirzepatide acetate, tirzepatide sodium) not equivalent to the FDA-approved active ingredient and not characterized in the SURMOUNT or SURPASS trials. Patients pricing brand against gray-market product in 2026 are not comparing like with like.

Retatrutide is Eli Lilly's triple agonist — GIP and GLP-1, as here, plus glucagon — and its Phase 3 results are larger than anything in the SURMOUNT programme. It is also unapproved, with a Biologics License Application stated for the first quarter of 2027, and the trial comparing it directly with tirzepatide does not reach primary completion until November 2026. Tirzepatide vs retatrutide sets the trial numbers side by side and explains why the gap between them is not a subtraction.

Counted from PubMed and ClinicalTrials.gov on October 9, 2026. A study counts only if participants were given tirzepatide. How we count, and the same figures for every peptide we cover.

Most recent published trials

• Cost-effectiveness of tirzepatide versus semaglutide for patients with obesity or overweight in the US: evidence from the SURMOUNT-5 head-to-head phase-3 trial. J Med Econ, 2026.
• Estimating the True MACE Benefits From Tirzepatide in SURPASS-CVOT Using an Imputed Placebo Analysis of REWIND. Diabetes Care, 2026.
• Ixekizumab With Tirzepatide Achieved Greater Disease Control Than Ixekizumab Alone in Adults With Psoriatic Arthritis and Overweight or Obesity: Results From a Randomized Clinical Trial. Arthritis Rheumatol, 2026.
• Liver Fat Changes in Patients With Type 2 Diabetes by Presence of Genetic Hepatic Steatosis: A Post Hoc Analysis of SURPASS-3 MRI. Liver Int, 2026.
• Tirzepatide Efficacy and Tolerability According to Early Weight Response: A Post Hoc Analysis of the SURMOUNT-1 and SURMOUNT-2 Trials. Diabetes Obes Metab, 2026.

Most recent registrations

• Tirzepatide for the Treatment of Cannabis Use Disorder NCT07265752, not yet recruiting, planned start 2027.
• Efficacy of Tirzepatide for Weight Management in Patients With Suboptimal Weight Loss After Bariatric Surgery: A Randomized, Double-Blind, Placebo-Controlled Clinical Trial NCT07829198, not yet recruiting, planned start 2027.
• Tirzepatide in Women With Obesity and Endometrial Intra-epithelial Neoplasia or Grade 1 Endometrial Cancer NCT07078838, not yet recruiting, planned start 2027.
• Tirzepatide for the Treatment of Concurrent Type 1 Diabetes and Overweight or Obesity NCT06180616, not yet recruiting, planned start 2026.
• Trial of Tirzepatide for the Treatment of Cannabis Use Disorder NCT07468552, not yet recruiting, planned start 2026.

A registration is a plan filed before a study starts, not a result.