Yes, and by a lot. Two phase 3 trials followed 469 people for 52 weeks. In those not using positive airway pressure, breathing interruptions fell 20.0 events an hour more than placebo. In people already using it, the gap was 23.8 events an hour.
Treating apnea and preventing it are different questions. A new-user cohort study in type 2 diabetes asked whether these drugs stop apnea appearing at all [2], and the two are different purchases.
One practical gap sits behind the result. A pan-European survey of sleep clinicians found few services rechecking a patient’s machine after weight loss [3]. So a pressure setting can outlive the body it was titrated for.
Sleep apnea is also the one obesity-related condition where the trial evidence and the price question collide most directly. The result is large and unusually clean. The dose it was produced at is near the top of the ladder, and the ladder is what this market publishes worst. Tirzepatide is already the dearer molecule before any of that, as the tirzepatide premium sets out.
What the trials found
SURMOUNT-OSA ran as two parallel phase 3 trials in adults with moderate-to-severe obstructive sleep apnea and obesity [1]. Trial 1 enrolled people not receiving positive airway pressure therapy; trial 2 enrolled people who were. Both randomized 1:1 to the maximum tolerated dose of tirzepatide — 10 mg or 15 mg — or placebo, for 52 weeks.
Participants arrived severely affected. Mean BMI was 39.1 in the first trial and 38.7 in the second, and the index itself averaged just over 50 events an hour in both.
By week 52 the drug arms had improved by 25.3 and 29.3 events an hour. The placebo arms are what make those numbers interpretable. They moved only 5.3 and 5.5. Everyone in both trials also received lifestyle intervention, so the placebo change is what that bought on its own. Subtracting leaves the drug’s own contribution. That is 20.0 events an hour in the first trial (95% CI 25.8 to 14.2). The second gave 23.8 (95% CI 29.6 to 17.9). Each at p<0.001. Every prespecified key secondary endpoint moved with it.
An index above 30 is severe by convention. These participants started at around 50 and finished, on average, in the twenties. That is a category change rather than an improvement at the margin, which is rare enough in this literature to be worth saying plainly.
The dose is the price
Every number above was produced at 10 mg or 15 mg, held for a year. That is not where a buyer starts. It is where a buyer arrives several months in, if escalation goes as planned. What the bill does on the way there is the question this roster mostly does not answer.
Across 430 sellers, 258 — 60% of them — publish nothing about what happens to the price as the dose rises. The rest split between stating a flat rate and stating that the figure climbs. Only the first is a promise a reader can hold anyone to. We count the whole distribution in nobody says what happens at a higher dose.
Before buying for this reason
A sleep study is the thing that establishes there is a problem, how bad it is, and whether anything changed. It is also the one part of this that a telehealth intake form does not do. If the reason for interest is apnea rather than weight, diagnosis comes first. The seller comparison comes second, which is what what a GLP-1 actually costs and where choice runs out are for.
The dose these trials used also sits at the far end of an escalation most people never complete. The discontinuation figures in two-thirds stop within a year describe the same journey from the other side.