A new obesity drug works worse in the patients who need it most
Retatrutide produced smaller weight loss in people with type 2 diabetes than in those without it.
By The Weekend · · 5 min read
In the TRIUMPH-2 trial, people with obesity and type 2 diabetes lost less weight on retatrutide than people with obesity alone lose on the same drug. The gap isn't small, and it isn't new to this drug — it shows up with every GLP-1 and multi-agonist tested so far. The question The Lancet's commentary raises is blunt: does having diabetes blunt the one treatment that was supposed to work regardless?
A triple-action drug still ran into the same wall
Retatrutide is built to hit three receptors at once — GLP-1, GIP, and glucagon — the theory being that stacking mechanisms should produce weight loss beyond what single- or dual-agonist drugs manage. In trial participants without diabetes, that theory has mostly held: retatrutide has produced some of the largest average weight reductions reported for any injectable obesity drug to date. TRIUMPH-2 tested the same drug in people who also have type 2 diabetes, and the weight loss came in lower. Not negligible — but measurably short of what the non-diabetic data promised.
That pattern isn't retatrutide's alone. Semaglutide and tirzepatide trials have shown the same split for years: people with type 2 diabetes consistently lose a smaller percentage of body weight than people with obesity and no diabetes, even at identical doses and similar starting BMI. TRIUMPH-2 matters because it shows a drug engineered specifically to be more powerful hitting the identical ceiling. If a triple agonist can't close the gap, the gap probably isn't about potency.
Who pays for the shortfall, and who was promised otherwise
The people who pay are patients with both conditions — a large and growing population, given that most people with type 2 diabetes also carry obesity. They are often the ones steered toward these drugs first, on the logic that one injection should manage two problems. TRIUMPH-2's data complicates that pitch: the same prescription, same dose, delivers a smaller result precisely in the group it's most often marketed to.
Manufacturers gain regardless of the gap, because regulatory approval and prescribing volume don't depend on closing it — retatrutide's non-diabetic results are strong enough to carry headlines, and the diabetic subgroup data lives in a journal commentary, not a press release. Insurers and health systems, meanwhile, are left pricing a drug whose cost-effectiveness calculations were often built on the larger weight-loss figures from non-diabetic trials. If the diabetic population loses less weight, the metabolic payoff — fewer cardiovascular events, less joint strain, lower long-term care costs — arrives smaller and later than projected for a meaningful share of the people actually taking the drug.
Insulin resistance, not willpower, explains the gap
The mechanism is where this stops being a marketing problem and becomes a biology problem worth understanding. GLP-1 and related agonists work partly by suppressing appetite through the brain and partly by slowing gastric emptying — but a meaningful share of their fat-loss effect in trials also tracks with improved insulin sensitivity, since better-regulated insulin reduces fat storage signaling. People with established type 2 diabetes already have chronic insulin resistance baked into their physiology, often after years of elevated insulin levels doing exactly the opposite of what these drugs are trying to achieve. Layering a new agonist on top of an insulin-resistant system doesn't fully reverse years of metabolic adaptation; it attenuates it.
There's a second factor the commentary flags: people with type 2 diabetes are frequently on other medications — insulin itself, sulfonylureas — that promote weight gain or complicate calorie balance, and those prescriptions don't automatically stop when a new weight-loss drug starts. A patient can be losing fat from retatrutide while an older diabetes drug is working, mildly, in the opposite direction. None of this is a failure of the new drug. It's a reminder that triple agonists are being asked to out-muscle a metabolic environment that single mechanisms never fully cleared in the first place.
What translating the percentage points actually means for someone taking this
Put in concrete terms: if a person without diabetes on retatrutide loses roughly a fifth of their body weight over the trial period, the equivalent person with type 2 diabetes loses a noticeably smaller fraction — enough to matter for clothing size, joint pain, and blood pressure, but not enough to match the number used to sell the drug class. For someone weighing 220 pounds, the difference between the two outcomes can be the gap between losing 45 pounds and losing 25 — both real, neither trivial, but not interchangeable when a patient is choosing a treatment based on headline results from a different population.
This isn't an argument against prescribing these drugs to people with diabetes — the smaller effect is still a meaningful effect, and improved glucose control alongside any weight loss carries its own benefit. It's an argument for telling patients which number applies to them before they start.
The next trial has to test the drugs already on the shelf, not just the new one
The open question TRIUMPH-2 leaves is whether combining retatrutide with changes to a patient's existing diabetes regimen — tapering weight-promoting drugs like certain insulins or sulfonylureas where clinically appropriate — narrows the gap seen in this trial. That's a testable claim: a follow-up trial that adjusts background diabetes medication alongside the new agonist, rather than layering it on top of unchanged prescriptions, should show whether the shortfall is pharmacological or simply additive drug interference. If weight loss in that scenario closes toward the non-diabetic numbers, the lesson is about prescribing practice. If it doesn't close, the lesson is that insulin resistance itself sets a biological ceiling no single new drug, however many receptors it hits, is likely to clear alone.