GLP-1 Medications for Sleep Apnea: A Drug That Works on Your Airway, Not Your Sleep

The first drug ever approved to treat obstructive sleep apnea does nothing to your sleep. It doesn't act on the brainstem circuits that set your breathing rhythm. It doesn't firm up the muscles that hold your throat open, and it doesn't change how easily a falling oxygen level wakes you. What tirzepatide does, along with the broader family of GLP-1 medications it belongs to, is make you smaller. That includes places most people never think about, like the base of the tongue.

That sounds like a technicality. It isn't. It explains why the drug works as well as it does, why it fails for the people it fails, and why the most useful question to ask your doctor isn't "Should I take this instead of CPAP?"

What the FDA actually approved

In December 2024 the FDA approved tirzepatide, sold as Zepbound for weight management, for moderate to severe obstructive sleep apnea in adults with obesity. The approval rested on two parallel trials published together as SURMOUNT-OSA (Malhotra et al., 2024, New England Journal of Medicine). The trials enrolled 469 adults with a body-mass index of 30 or more and an apnea-hypopnea index (AHI, the count of breathing pauses and shallow breaths per hour of sleep) of at least 15. People with type 2 diabetes were excluded. One trial enrolled people who weren't using positive airway pressure. The other enrolled people who were using it and agreed to keep using it.

After 52 weeks of weekly injections at the highest dose each person tolerated (10 or 15 mg), AHI fell by about 25 events per hour in the no-CPAP trial and about 29 in the CPAP trial. The placebo groups fell by about 5. Body weight dropped by roughly 18 to 20 percent, against 2 percent or less on placebo.

Those are large effects. For context, a person at an AHI of 45 who drops 25 events per hour moves from "severe" to "moderate." That's a real change in the burden on the heart and blood vessels, but it isn't a cure. Keep that distinction in mind. It comes up again below.

Why a weight drug works on a breathing disorder

The link between body weight and apnea severity is one of the better-established relationships in sleep medicine. The common citation is Peppard et al. (2000, JAMA), which followed 690 adults in the Wisconsin Sleep Cohort for four years. A 10 percent gain in weight predicted about a 32 percent rise in AHI. A 10 percent loss predicted about a 26 percent fall. The relationship isn't perfectly linear, and it varies a lot from person to person, but its direction has held up for a quarter century.

What changed more recently is our picture of where the weight matters. Wang et al. (2020, American Journal of Respiratory and Critical Care Medicine), from Richard Schwab's group at the University of Pennsylvania, scanned the upper airways of 67 people with obesity and OSA before and after they lost weight through diet or bariatric surgery. Shrinking fat in the tongue was the strongest anatomical predictor of improvement in AHI. It mattered more than changes in the pharyngeal walls or the soft palate.

What happens in the airway, in order

It helps to walk through the sequence as it unfolds in the body over months of treatment.

First, the drug acts on GLP-1 receptors (and, for tirzepatide, GIP receptors too) in the gut and brain. Stomach emptying slows and appetite falls. People eat less, often a good deal less, without having to fight hunger the whole time.

Second, sustained low intake draws down fat stores throughout the body. Fat isn't only under the skin. Some of it sits in the tongue, in the walls of the pharynx, and in the tissue packed around the airway inside the bony box of the jaw and spine.

Third, as that fat shrinks, the airway has more room. The rigid frame of the jaw stays the same size, but there is less soft tissue crowding it. The pressure at which the relaxed throat collapses during sleep falls.

Fourth, and only then, breathing during sleep improves. There are fewer collapses and fewer arousals, and oxygen dips are shallower. The drug never touched sleep directly. It changed the plumbing, and better sleep followed from that.

That is also why the effect builds slowly. SURMOUNT-OSA measured at a year, and weight was still falling for many participants near the end. Anyone expecting to feel different in the first month is expecting the wrong thing.

Can a GLP-1 drug replace CPAP?

For most people, no, at least not at first, and not with certainty later. CPAP works from the first night by splinting the airway open with air pressure. A GLP-1 drug works over months by reducing the tissue that collapses. In SURMOUNT-OSA, roughly half the people who took the drug still met the criteria for sleep apnea after a year. The realistic framing is that the drug may lower the severity of your apnea, sometimes by enough that a repeat sleep study shows you no longer need CPAP. You find that out by testing, not by how you feel.

The trial's own design points the same way. Half the participants stayed on CPAP the whole time, and the drug still added benefit on top of it. The FDA approved it as treatment, not as a CPAP substitute.

The number the headline left out

The trial reported "disease resolution," defined as an AHI below 5, or an AHI between 5 and 14 with little daytime sleepiness. About 42 percent of treated participants in the no-CPAP trial reached it, and about 50 percent in the CPAP trial. On placebo, the figures were 14 to 16 percent.

Read the other way, that means half or more of people on a drug that took off a fifth of their body weight still had sleep apnea.

Part of the explanation is that apnea isn't purely an anatomical disease. Eckert et al. (2013, American Journal of Respiratory and Critical Care Medicine) measured four traits in 75 people with OSA. One was airway collapsibility. The other three were nonanatomical: an unstable breathing control system ("high loop gain"), a tendency to wake too easily from small breathing disturbances, and weak muscle response when the airway starts to close. A little over half had at least one nonanatomical trait contributing to their disease. Weight loss mostly treats the anatomical one.¹

So if your apnea comes mainly from crowding, a GLP-1 drug goes after the cause. If it comes partly from an unstable breathing controller or a small, recessed jaw, the drug works on only part of the problem. Today there's no routine clinical test to tell you which group you're in. That's plausible but thin as a basis for choosing treatment, and it's an active research area rather than settled practice.

What happens when you stop

The drug's effect on the airway runs through weight, so it lasts only as long as the weight loss does. In SURMOUNT-4 (Aronne et al., 2024, JAMA), 670 participants who had lost weight on tirzepatide for 36 weeks were randomized to continue the drug or switch to placebo. The placebo group regained about 14 percent of their body weight over the next year. No one has published a trial following AHI after stopping. Still, given everything above, expecting the apnea to come back with the weight is the reasonable assumption, not the pessimistic one.

Who the trial didn't include

People with type 2 diabetes, people with a BMI under 30, and people with central rather than obstructive apnea weren't studied. Other GLP-1 drugs, including semaglutide, have no OSA approval, though they work through the same weight pathway. Gastrointestinal side effects, mainly nausea, diarrhea, vomiting, and constipation, were the most common adverse events and were usually mild to moderate. One practical detail matters to some readers: in the United States, an OSA diagnosis has opened a coverage route for some patients whose insurance, including Medicare Part D, excludes drugs prescribed purely for weight loss. Coverage still varies by plan.

A practical sorting

Honest rule of thumb

Treat the drug as a way to change your anatomy, and treat the repeat sleep study as the only verdict. Keep whatever is already working for you. Give the drug most of a year before judging it. Then get retested before you put the mask in a drawer. I read sleep-apnea drug papers differently now. I skip the weight-loss figure in the abstract and go straight to the table showing how many people still had the disease at the end. In SURMOUNT-OSA it is the least quoted number in the paper, and the most useful one.

¹ A side note for this magazine's regular readers: caffeine does nothing for the airway. It can cover up the daytime sleepiness that comes from fragmented sleep, so heavy afternoon coffee is often part of the story people tell before they're diagnosed. It isn't evidence that anything has been treated.