TRIUMPH-1 answered the question "what does retatrutide do in people with obesity?" TRIUMPH-2 asks the harder one: what does it do when the same obesity comes with type 2 diabetes? The trial was published in The Lancet in October 2026, following topline results in July12.
Why diabetes is a different trial
Every incretin-based agent studied to date produces less weight loss in people with type 2 diabetes than in people without it. Semaglutide did. Tirzepatide did. The phase 2 retatrutide programme already showed it: the diabetes trial reported smaller weight reductions than the obesity trial at matched doses4.
The reasons are not fully settled, but the leading candidates are consistent across the literature: insulin resistance blunts the metabolic response, background diabetes medications such as insulin and sulfonylureas promote weight gain, and the diabetic population tends to be older with longer-standing obesity. Whatever the mechanism, the pattern is robust enough that a diabetes trial producing the same weight loss as a non-diabetes trial would be the surprising result.
The weight endpoint
Mean percentage change from baseline, adults with obesity and type 2 diabetes
Lilly topline release, July 2026 [2]. Chart by TrueLab Care.
At 80 weeks, topline mean weight change was −12.7% at 4 mg, −19.1% at 9 mg and −20.8% at 12 mg, against −4.0% on placebo2.
Topline mean weight change, TRIUMPH-1 (no diabetes) vs TRIUMPH-2 (type 2 diabetes)
Lilly topline releases [2][3]. Chart by TrueLab Care.
Set against TRIUMPH-1's topline −28.3% at the same 12 mg dose3, the diabetes population lost roughly seven to eight percentage points less. That ratio, approximately three quarters of the non-diabetic effect, is in line with what the class has shown before. Retatrutide does not escape the pattern; it reproduces it at a higher absolute level.
The glycaemic endpoint
In a diabetes trial, body weight is only half the primary question. Glycated haemoglobin, HbA1c, is the standard measure of glucose control over the preceding two to three months. TRIUMPH-2 reported reductions of up to 1.6 percentage points at the higher doses2.
For context, a reduction of one percentage point is considered clinically meaningful, and most oral diabetes agents produce between 0.5 and 1.0. The incretin class routinely exceeds that, and retatrutide's figure sits at the upper end of what has been reported for the class.
The glucagon receptor component of the molecule makes this result less obvious than it looks. Glucagon raises blood glucose. A triple agonist that includes glucagon agonism could, in principle, worsen glycaemic control. The phase 2 diabetes trial had already indicated that the GLP-1 and GIP components dominate at the doses used4; TRIUMPH-2 confirms it at scale.
Tolerability
Percentage of participants per arm
Lilly topline release [2]. Chart by TrueLab Care.
Discontinuation due to adverse events was 3.8% at 4 mg, 11.6% at 9 mg and 7.7% at 12 mg, against 4.9% on placebo2.
Two features of those numbers are worth noting. First, the 4 mg arm discontinued less often than placebo, which suggests that at the lowest dose the compound's side-effect burden was not distinguishable from background. Second, the 9 mg arm discontinued more often than the 12 mg arm, which is not what a simple dose, response relationship would predict. Differences of this size between adjacent arms are common in trials of this scale and usually reflect chance variation in who happened to be randomised where, rather than a real inversion. The full paper's adverse-event tables are the place to check, not the topline summary.
What the trial does not show
The usual limits apply. Eighty weeks is not long-term. The population was selected and protocol-managed. Cardiovascular outcomes were not an endpoint here; that is the job of a separate trial. And the comparison with TRIUMPH-1 is between two trials, not within one, so the seven-point gap is an observation rather than a tested hypothesis.
Status
As with every retatrutide trial to date, this is data in support of a filing that has not yet been made. The compound is not approved for diabetes, obesity or anything else. The Lancet publication is what a regulator will read; it is not a licence.