The Weight-Loss Penalty

A pattern nobody advertises

There is a fact about this entire drug class that rarely makes it into the marketing, and it is not a small one.

People with type 2 diabetes lose less weight on these drugs than people without diabetes, taking the identical molecule at the identical dose.

Not a little less. A meaningfully smaller share of them hit the thresholds that matter.

In the STEP-UP programme, semaglutide 7.2 mg produced 10 percent or greater weight loss in 86 percent of participants without diabetes. In participants with type 2 diabetes, on the same drug and dose, that fell to 66 percent.

In the SURMOUNT programme, tirzepatide 15 mg produced 10 percent or greater weight loss in 90 percent of participants without diabetes. In participants with type 2 diabetes, 70 percent.

Twenty points, both times. That is not noise. That is a reproducible penalty, and it has a name for a reason: insulin resistance and elevated insulin levels themselves promote fat storage, and diabetes duration correlates with a blunted response to these drugs' appetite and energy pathways. The disease you are trying to treat is quietly working against the treatment.

Which means every weight loss headline for a GLP-1 drug in a general population tells you less than you think about what will happen to the person that drug is arguably needed by most.

What this trial was built to test

Part 1 covered the population: newly diagnosed, drug-naive, a mean disease duration of 2.5 years. If the weight loss penalty is driven by years of accumulated insulin resistance, catching people this early should blunt the penalty. If it is driven by something more fundamental to having diabetes at all, catching them early should not matter much.

Here is what happened, at 40 weeks, treatment regimen estimand.

Under the efficacy estimand, which assumes full adherence to the assigned dose for the full 40 weeks, the 12 mg group reached 17 percent.

Set that beside the numbers you were just given. Sixty six to 70 percent of people with type 2 diabetes on the best prior drugs in this class reached 10 percent weight loss. Here, the average outcome at every dose of retatrutide exceeded 10 percent, and the top dose averaged 15 to 17 percent, in a population that was, by definition, entirely diabetic.

This trial did not find a smaller version of the usual weight loss penalty. It found numbers that read like a population without diabetes at all.

Why, probably

The paper is careful here, and so should we be. This is a single-arm comparison against historical trial data from different studies, different populations, different eras. It is not a randomised head-to-head, and the authors say so directly.

But two explanations sit on the table, and they are not mutually exclusive.

The first is timing. This population had a mean disease duration of 2.5 years and was overwhelmingly drug-naive. Insulin resistance and the metabolic remodelling that comes with years of hyperglycaemia had less time to entrench. If the penalty is partly a function of how long the system has been running in a diabetic state, catching it early would be expected to blunt it, and that is exactly the shape of what happened.

The second is the drug itself. Every trial cited above used a GLP-1 or GIP-plus-GLP-1 mechanism. This trial used a third receptor: glucagon. If you read the first series in this publication, you know what that arm does. It drives lipolysis, increases energy expenditure, and does it through a pathway that operates somewhat independently of insulin signalling. A mechanism that does not lean entirely on insulin-sensitive pathways would be expected to hold up better in a population defined by insulin resistance.

The honest answer is that this trial cannot separate the two. Both are plausible, both point the same direction, and untangling them needs the head-to-head trials the researchers themselves flag as still running.

One more number worth sitting with

The trial also tracked waist circumference, and it fell alongside body weight across every dose group. The paper notes this as a marker of reduced visceral adipose fat specifically, the fat wrapped around internal organs that drives the worst of diabetes and cardiovascular risk, as opposed to fat lost more generally across the body.

That distinction matters clinically. Losing fat is not uniformly protective; where it comes from matters. A treatment that preferentially reduces visceral fat in a population defined by insulin resistance is treating closer to the actual mechanism of the disease, not just the number on the scale.

What this doesn't tell you yet

Weight loss and diabetes control are not automatically the same story, and the field has learned that lesson the hard way with other drugs before. A number on a scale is not a glycaemic outcome.

So the obvious next question: with a drug-naive population losing this much weight this fast, what happened to their blood sugar? Because there is a second well known ceiling in this drug class, entirely separate from the weight loss penalty, and this trial had something to say about that one too.

That is Part 3.

This series covers a peer-reviewed phase 3 clinical trial published in The Lancet. Retatrutide has not been approved by the FDA for type 2 diabetes or any other indication. Nothing here is medical advice.