Retatrutide cut sleep-apnoea severity in two phase 3 trials

Retatrutide, an investigational once-weekly injection that activates three metabolic hormone receptors at once — those for GIP, GLP-1 and glucagon — reduced the severity of obstructive sleep apnoea in two phase 3 trials published in the New England Journal of Medicine [s1]. The result extends the triple agonist's weight-loss story into a hard clinical endpoint, but the two trials did not agree on how large the benefit is, and that gap is the part worth reading closely.

Obstructive sleep apnoea (OSA) is the repeated collapse of the upper airway during sleep, measured by the apnoea–hypopnoea index (AHI) — the average number of breathing interruptions per hour. Obesity is the dominant modifiable driver, and losing weight reliably lowers the AHI, which is why a drug that produces large weight loss was expected to help. The question these trials set out to answer was whether retatrutide does so, and by how much, against a placebo [s1].

What the trials did

The two trials were nested inside larger "basket" studies of retatrutide. TRIUMPH-1-OSA enrolled 243 adults without type 2 diabetes; TRIUMPH-2-OSA enrolled 448 adults who also had type 2 diabetes [s1]. Both were double-blind and placebo-controlled, registered as NCT05929066 and NCT05929079 [s2][s3]. All participants had moderate-to-severe OSA and a body-mass index of at least 27, and were randomly assigned to subcutaneous retatrutide at 4, 9 or 12 mg, or placebo, for 80 weeks [s1]. The primary endpoint was the change in AHI at 80 weeks for the 9 mg and 12 mg doses [s1]. At baseline the mean AHI was 58.6 events per hour in TRIUMPH-1 and 55.9 in TRIUMPH-2 — both squarely in the severe range — with mean body weights near 120 kg and 116 kg respectively [s1].

What it found

In TRIUMPH-1-OSA, the AHI fell by 34.3 events per hour on 9 mg and 31.7 on 12 mg, versus 9.9 on placebo — differences from placebo of 24.4 (95% CI 14.7–34.1) and 21.9 (11.5–32.2) events per hour [s1]. In TRIUMPH-2-OSA, the group with type 2 diabetes, the reductions were more modest: 24.7 on 9 mg and 25.4 on 12 mg against 10.3 on placebo, for placebo-adjusted differences of 14.4 (7.2–21.6) and 15.1 (7.7–22.5) [s1]. Intention-to-treat estimates were close to these figures, so the result is not an artefact of how dropouts were handled [s1].

Weight loss tracked the apnoea benefit: mean body-weight reductions ran from roughly 12.5% to 22.2% in TRIUMPH-1 and 12.1% to 18.5% in TRIUMPH-2, versus about 5% on placebo [s1]. That parallel is the mechanistic heart of the result — the drug is widely assumed to relieve apnoea chiefly by shrinking the fat around the airway and neck, so the smaller weight loss in the diabetes trial plausibly explains its smaller apnoea effect. The most common adverse events were gastrointestinal — nausea, vomiting and diarrhoea, the familiar signature of this drug class — and the trials do not suggest a new safety signal specific to OSA [s1].

How to read it

The honest headline is that retatrutide lowered a clinically meaningful measure of OSA severity, not just a scale number, and did so at both of the doses the trials were powered to test. But the split between the two trials matters. People with type 2 diabetes got roughly a third to a half less AHI reduction than people without it — a pattern seen before with incretin drugs, whose weight loss, and often their metabolic effects, tend to be smaller in diabetes [s1]. A reader told only the TRIUMPH-1 numbers would come away with an inflated sense of the average benefit.

Two further cautions belong in any fair summary. First, these were 80-week efficacy trials with AHI as the endpoint; they were not designed to show that lowering the AHI translates into fewer daytime symptoms, cardiovascular events or deaths, which is the outcome patients ultimately care about. Second, both trials were funded by Eli Lilly, the drug's maker [s1], and retatrutide remains investigational rather than approved for OSA. The confidence intervals in the diabetes trial, while clearly excluding no effect, reach down toward single-digit AHI reductions at their lower bounds [s1] — a benefit that is real but not uniformly large.

What it means

For now, retatrutide joins a short list of weight-loss drugs with trial evidence against sleep apnoea rather than against the scale alone. Whether it becomes a treatment will depend on regulators, on how the apnoea benefit holds up outside a trial, and on cost and access — none of which these two papers settle. What they do establish is narrower and worth stating plainly: in people with obesity and moderate-to-severe OSA, the drug reduced breathing interruptions during sleep more than placebo over 80 weeks, by more when diabetes was absent and by less when it was present [s1].