Does Ozempic Help Sleep Apnea?

Evidence for GLP-1 receptor agonists reducing the severity of obstructive sleep apnea in adults with obesity.

The strongest evidence here tested tirzepatide, which is sold as Mounjaro and Zepbound, not semaglutide, which is sold as Ozempic and Wegovy. In SURMOUNT-OSA, two 52-week phase 3 randomised placebo-controlled trials in adults with moderate-to-severe obstructive sleep apnea and obesity, tirzepatide cut apnea-hypopnea events by 23.8 an hour more than placebo. Tirzepatide is the drug that holds an FDA approval for moderate-to-severe obstructive sleep apnea in adults with obesity. Pooled across the GLP-1 class the figure is 13.89 fewer events an hour, from four randomised trials meta-analysed in Sleep and Breathing in 2026, which is smaller than the tirzepatide trial alone. Two qualifications travel with all of it: the drug took about 20 weeks to separate from placebo, and the size of the improvement tracked how much weight came off, so whether the molecule or the weight loss is doing the work is not settled.

Which drug the evidence is actually about

Ozempic and Wegovy are semaglutide. Mounjaro and Zepbound are tirzepatide, which acts on two receptors rather than one, the GLP-1 receptor and the glucose-dependent insulinotropic polypeptide receptor. They are commonly discussed as one category and they are not one molecule, and on this particular question the distinction decides the answer.

SURMOUNT-OSA, published in the New England Journal of Medicine in 2024, is the trial that moved this field. It enrolled adults with moderate-to-severe obstructive sleep apnea and obesity into two parallel 52-week double-blind randomised placebo-controlled studies, split by whether the participant was already using positive airway pressure at baseline. Participants not on PAP went into trial 1, those on PAP into trial 2. Tirzepatide reduced the apnea-hypopnea index by 23.8 events an hour against placebo.

That is the evidence behind the FDA approval for moderate-to-severe obstructive sleep apnea in adults with obesity, and the approval is tirzepatide's. A reader asking whether Ozempic treats their sleep apnea is asking about a drug that does not carry that approval and was not the drug in that trial.

What the class-wide evidence shows

Across the GLP-1 class the pooled figure is 13.89 fewer apnea-hypopnea events an hour, from four randomised controlled trials in obstructive sleep apnea, meta-analysed in Sleep and Breathing in 2026. That is a real reduction and it is smaller than the 23.8 from the tirzepatide trial on its own, which is what pooling a class around its strongest member tends to produce.

For scale on what those numbers mean: the apnea-hypopnea index counts breathing interruptions per hour of sleep, and moderate-to-severe disease starts at 15. A reduction of 23.8 an hour is large enough to move many people across diagnostic categories, which is why the trial result was treated as a significant finding rather than an incremental one.

A separate analysis of the SURMOUNT-OSA data examined the time course and the weight relationship directly. Two findings from it bear on how a reader should read the headline number. The drug took roughly 20 weeks to separate from placebo, so this is not a treatment that shows an effect in the first month. And the magnitude of the apnea improvement tracked the magnitude of the weight loss.

The question the trials do not settle

Obesity is an established causal risk factor for obstructive sleep apnea, and tirzepatide produces substantial weight loss. So an apnea improvement that scales with weight loss has two readings, and the published trials do not separate them. The drug may be acting on the airway through some mechanism of its own, or it may be producing a familiar result through a familiar route, which is that losing weight reduces apnea severity.

That distinction matters to a reader deciding what to expect, because weight loss achieved any other way carries the same expectation. It also matters because it sets what happens on stopping. If the mechanism is weight, the apnea benefit depends on the weight staying off.

Neither trial was designed to answer it, and a design that could would need to hold weight loss constant across arms. Nothing published has done that.

Common questions