Semaglutide linked to NAION signal
- EMA’s safety committee concluded in June 2025 that semaglutide can cause NAION, a rare optic-nerve stroke, and told doctors to stop treatment if confirmed. - The key number is small but real — about one extra NAION case per 10,000 person-years, with epidemiology studies pointing to roughly doubled risk. - That matters because earlier studies conflicted sharply, so the story has shifted from “possible signal” to “recognized very rare side effect.”
Semaglutide and NAION is an eye-safety story — and the stakes are obvious. NAION is a sudden optic-nerve injury that can leave permanent vision loss. For most of 2024, the question was whether the signal was real or just noise from messy observational data. By June 2025, that changed: EMA’s safety committee finished its review and said NAION should be listed as a very rare side effect for semaglutide medicines including Ozempic, Wegovy, and Rybelsus. ### What is NAION? NAION stands for non-arteritic anterior ischemic optic neuropathy. Basically, it is a kind of optic-nerve stroke — blood flow to the front of the optic nerve drops, vision can go suddenly blurry or dim, and some loss can be permanent. It is already known to happen more often in people with diabetes, hypertension, sleep apnea, high cholesterol, and other vascular risk factors — which is exactly why this question is so hard to study in people taking semaglutide. (ema.europa.eu) ### What first raised the alarm? The first big jolt came from a July 2024 JAMA Ophthalmology paper out of Mass Eye and Ear. In a matched cohort from one academic neuro-ophthalmology registry, patients prescribed semaglutide had higher observed NAION risk than patients on non-GLP-1 comparator drugs for diabetes or weight loss. The sample was only 16,827 patients, though, and it came from one specialty center — useful as a warning light, but not enough to settle causality. (jamanetwork.com) ### Did bigger datasets back that up? Sort of — but not cleanly. A March 2025 TriNetX study in more than 3.3 million patients with diabetes found no clear excess risk early on, then a higher risk appearing at 2, 3, and 4 years, with hazard ratios around 2.05 to 2.44. Another 2025 commentary in JAMA Ophthalmology described the OHDSI network analysis as showing only a small relative increase. So the signal persisted in some large datasets, but the size of the effect shrank as the data got broader. (jamanetwork.com) ### Why were people arguing about confounding? Because semaglutide users are not random. They often have more severe diabetes, more obesity, and more cardiovascular risk — and those are exactly the traits that also raise NAION risk. That means an observational study can make the drug look guilty when the real driver is the patient mix. A 2026 narrative systematic review pulled together nine large retrospective studies and found the literature split almost down the middle: four showed increased risk, four found no significant association, and one even suggested lower risk. (jamanetwork.com) ### So what did regulators decide? EMA landed in the middle. It did not say semaglutide commonly causes blindness. It said NAION is a very rare side effect — up to 1 in 10,000 patients — and that the total evidence supports roughly a twofold increase in risk in adults with type 2 diabetes, equal to about one extra case per 10,000 person-years of treatment. If sudden vision loss or rapidly worsening eyesight happens, patients should get urgent medical evaluation, and if NAION is confirmed, semaglutide should be stopped. (pmc.ncbi.nlm.nih.gov) ### What about the “brand differences” story? That part is trickier than some headlines made it sound. A 2026 British Journal of Ophthalmology analysis using FDA adverse-event reports found stronger reporting signals for Wegovy than Ozempic, and higher odds in men than women. But FAERS is a spontaneous reporting system — good for signal detection, bad for proving incidence or comparing brands cleanly because reporting can be shaped by publicity, prescribing patterns, and dose differences. (ema.europa.eu) ### What should clinicians actually do? The practical move is simple. Ask about semaglutide when a patient shows up with sudden optic neuropathy. Warn patients to report abrupt vision changes fast. But do not treat this like proof that semaglutide is broadly unsafe. The absolute risk looks very low, and semaglutide still has major benefits in diabetes, obesity, cardiovascular disease, and kidney disease. (bjo.bmj.com) ### Bottom line The story is no longer “maybe there’s a signal.” The story is that regulators now treat NAION as a real but very rare semaglutide side effect — while the exact size, mechanism, and who is most vulnerable still are not fully nailed down. (ema.europa.eu)
