In December 2024, tirzepatide became the first medication the FDA approved specifically for obstructive sleep apnea (OSA), rather than for weight loss or diabetes with OSA as a side benefit. That approval covers the branded, manufactured product on the strength of the SURMOUNT-OSA trials. It says nothing about a compounded version — more on that below.
Nearly two years on, two new analyses ask a more interesting question than “does it work”: who is actually getting it. The answers don’t line up with each other the way you’d expect, and the gap between them is the story.
The national picture: still mostly untreated
A health-data company called Komodo Health analyzed claims from close to 20 million adults diagnosed with OSA, comparing treatment patterns before and after the December 2024 approval. This is proprietary claims analysis from a health-data vendor, not a peer-reviewed study, so treat the precision of the numbers loosely — but the direction is consistent with the peer-reviewed work below, and the scale is worth knowing.
Before the approval, only 14% of adults diagnosed with OSA had ever received a GLP-1 prescription. Sixty-six percent had received device-based treatment, almost entirely CPAP. In the six months after approval, new tirzepatide starts among OSA patients rose 16% — real movement, but movement from a small base. Primary care clinicians, not sleep specialists, wrote about 84% of the new prescriptions; sleep medicine accounted for roughly 0.2%.
There’s a second finding in the same analysis worth flagging carefully. Among patients newly starting tirzepatide, a smaller share went on to start CPAP than among patients who didn’t start the medication, and a meaningful minority of existing CPAP users showed extended gaps in mask usage after beginning the drug. That pattern is consistent with some patients treating a GLP-1 as a substitute for airway therapy rather than an addition to it. The data can’t say whether that substitution was clinically appropriate for those individual patients or not — claims data shows what was filled and worn, not what a clinician recommended or why a patient stopped. It’s association, not a verdict on anyone’s care.
The clinic-level picture: eligible, but not prescribed
A separate, peer-reviewed study in Otolaryngology–Head and Neck Surgery looked at the question from the opposite direction: not who’s getting a GLP-1, but who could be, and how many of those people actually are.
Researchers reviewed 384 patients seen at a single sleep surgery clinic between October 2023 and March 2025, all of whom had a sleep study on file. Average BMI was 29.4, average AHI (the count of breathing interruptions per hour that defines OSA severity) was 28.7 — moderate-to-severe by most staging systems — and just over half had already tried CPAP.
Under FDA eligibility criteria — diabetes, obesity, or a BMI of 27 or higher with a weight-related condition — 64.8% of the cohort qualified for a GLP-1. Of that eligible group, only 28.1% were actually taking one. Semaglutide and tirzepatide accounted for most of the prescriptions among those who were.
The strongest predictor of who fell into that gap wasn’t clinical. Patients enrolled in Medicaid were dramatically less likely to be on a GLP-1 than patients with other coverage — an adjusted odds ratio of 0.25, meaning roughly a quarter the likelihood, after controlling for other factors. Patients with a BMI of 30 or higher were more likely to be on one, which tracks with clinical need. Insurance coverage and side effects were the reasons patients cited most often for stopping or switching medications.
One clinic isn’t the whole country, and a sleep surgery population — people sick enough or frustrated enough with CPAP to see a surgeon — skews toward more severe, more treatment-seeking patients than OSA patients generally. But the eligibility-to-treatment gap and the insurance disparity are exactly the kind of finding a single-site study is well suited to surface, even if the precise percentages wouldn’t hold nationally.
Two studies, one shape
Neither of these is a trial of whether tirzepatide treats OSA — that question already has an answer from SURMOUNT-OSA, and we’ve covered what those results do and don’t establish elsewhere. What both of these add is a different, less flattering finding: access to an approved treatment tracks insurance status and which type of clinician a patient happens to see, more than it tracks clinical need. A patient with Medicaid and a BMI of 32 is, by the numbers here, less likely to be offered the medication than a patient with commercial insurance and a BMI of 28 — a gap running in the opposite direction of medical urgency.
That’s also where the CPAP-substitution signal from the Komodo analysis matters most. If a shrinking share of new tirzepatide patients ever start CPAP, and existing CPAP patients let it lapse after starting the medication, some of that may be sound clinical decisions made with a provider. Some of it may be patients who never had their OSA severity re-measured after starting a GLP-1 and simply assumed the pill made the mask unnecessary. AHI reduction on tirzepatide is real in the trial population, but it’s not uniform — a subgroup identified at the 2026 ATS conference improved roughly twice as much as everyone else — and nothing about starting a medication tells you which group you’re in without measuring.
Where compounded medications fit — and don’t
Everything above concerns tirzepatide as an FDA-approved, manufactured product. Compounded tirzepatide and semaglutide — the medications NoBsRx provides provider-guided access to — are prepared by state-licensed 503A pharmacies for an individual patient. They are not FDA-approved products, they were not the drug studied in SURMOUNT-OSA, and the sleep-clinic eligibility criteria described above apply to the approved label, not automatically to a compounded preparation. Whether a compounded medication is appropriate for a given patient’s sleep apnea is a decision for a licensed provider, not a default carried over from a study of a different formulation.
The part we can actually fix
We can’t change what insurance covers. What we can change is whether a patient starts any of this — a GLP-1, CPAP, or nothing — with an actual AHI number instead of a guess. A 3–5 night at-home sleep study on an FDA-cleared ring, read by a board-certified sleep clinician, gives you a baseline before treatment and something to re-check against afterward. If your data doesn’t support a diagnosis, that’s useful to know too. Our pricing is listed up front, which is more than can be said for the insurance maze in the data above.
You can’t tell whether a medication, a mask, or neither is doing the job if you never measured what you started with.
This post is general information, not medical advice. It does not describe a benefit of any NoBsRx program and it is not a claim that any medication will treat sleep apnea. Compounded medications are not FDA-approved. Treatment is never guaranteed — eligibility, diagnosis, and every treatment decision are made by independent licensed providers based on an individual clinical assessment, and availability varies by state. Do not start, stop, or change any treatment, including CPAP, based on an article. If you think you may have sleep apnea, talk to a licensed clinician.