43.8% Fewer Breathing Events, and No Device Maker Is Hurting: Eli Lilly's Sleep-Apnea Pill Is a Funnel, Not a Kill Shot

Eli Lilly reported ATTAIN-OSA on Nov. 9, 2026: a 52-week, randomized, double-blind, placebo-controlled Phase 3 trial of once-daily orforglipron in 712 adults with moderate-to-severe obstructive sleep apnea and obesity or overweight. It ran two studies side by side — 363 patients not using positive airway pressure and 349 patients who were — and in both, the pill beat placebo by a wide margin. The number that matters for investors is the second study: orforglipron cut the apnea-hypopnea index a further 27.5% in patients already wearing the mask every night.

The convenience case is strong. The potency case is not. Cross-trial, the pill sits below both of Eli Lilly's injectable sleep-apnea assets: tirzepatide cut moderate-to-severe severity by up to 62.8% in SURMOUNT-OSA, and retatrutide reduced apnea-hypopnea severity by 60.6% at its top dose. Orforglipron delivered 43.8%. These are different trials and different estimands, but the direction is consistent, and the conclusion matters: the pill's value proposition is access, adherence and price, not a better number than the needle.

The strongest evidence against the CPAP-disruption trade comes from the device makers themselves. ResMed's fiscal 2026 Form 10-K names the risk explicitly, noting that injectable weight-loss drugs are already approved for sleep apnea in patients with obesity while oral versions are still being developed. Yet on Oct. 29, 2026 the company told investors that patients holding scripts for both CPAP and a GLP-1 are more than 11% likelier to start therapy, and it raised fiscal 2027 guidance. A pill that gets an obese patient diagnosed and treated hands the device chain more paying patients rather than taking its machines away.

Here is what the device-versus-drug framing misses. Medicare already funds the standard of care for sleep apnea as durable medical equipment: a machine, then recurring masks. A pill enters the same disease through a different benefit, at a price that is multiples of the annual device cost it might one day displace, and it arrives on top of a benefit that is already paid. At $649 a month, the pill adds a recurring pharmacy cost to a benefit built around a one-time purchase, and the cheap bridge price is off the table for exactly these patients.

The chain explains why the device makers are not the first-order losers. A pill label moves sleep-apnea treatment volume out of the durable-medical-equipment channel and into the pharmacy benefit, where distributors and telehealth platforms earn a recurring script instead of a one-time machine sale. Every additional prescription that a payer authorizes for a PAP-naive patient creates a diagnostic, titration and mask opportunity before it creates any device-displacement risk, and only about a third of those patients would reach remission or mild disease on the pill alone.