Millions of people now take semaglutide for weight loss or diabetes management, but a small number of them face a risk that no dose adjustment or lifestyle change can reverse: sudden, painless vision loss in one eye. Europe’s drug safety committee concluded in June 2025 that non-arteritic anterior ischemic optic neuropathy, often called an “eye stroke,” is a very rare side effect of semaglutide medicines including Wegovy, Ozempic, and Rybelsus. The classification, affecting up to 1 in 10,000 patients, followed a review of all available data and two large Scandinavian registry studies that found the drug doubled the five-year risk of the condition in people with type 2 diabetes.
Why the NAION finding changes the calculus for semaglutide users
NAION occurs when blood flow to the optic nerve is suddenly cut off. Unlike a retinal detachment or glaucoma, it typically strikes without warning symptoms and leaves permanent damage before a patient can seek help. There is no established treatment. The condition has long been associated with cardiovascular risk factors such as hypertension, diabetes, and sleep apnea, but the new regulatory finding isolates semaglutide itself as an independent contributor.
The European Medicines Agency’s Pharmacovigilance Risk Assessment Committee (PRAC) met from June 2 through 5, 2025, and recommended that product information for all semaglutide medicines be updated to list NAION at a frequency of “very rare,” defined as up to 1 in 10,000 treated patients. That frequency label is the lowest tier on the European regulatory scale, yet it applies to a drug class prescribed to tens of millions of people worldwide, which means even a very rare event can translate into a meaningful number of affected individuals.
A central question is whether the risk stems from the drug’s mechanism or from the underlying metabolic disease it treats. A Danish cohort study covering 424,152 persons with type 2 diabetes between December 2018 and June 2024 found that once-weekly semaglutide doubled the five-year NAION risk compared with other glucose-lowering treatments. If the elevated risk were driven purely by diabetes severity, patients on comparator drugs with similar glycemic profiles should show similar rates. They did not. That gap points toward a drug-specific effect, though the registry data cannot yet confirm whether cumulative exposure over months or years increases the hazard further. Comparing time-to-event curves after matching on diabetes duration and HbA1c across the Danish and Norwegian registries would be one way to test that hypothesis directly.
Registry data from Denmark and Norway anchor the signal
Two population-level studies form the backbone of the regulatory conclusion. The first, drawn from Danish national registries, tracked all adults with type 2 diabetes over roughly five and a half years and used standardized diagnostic codes to identify NAION events. The doubling of five-year risk it reported was large enough to reach statistical significance in a condition that, by definition, is rare.
A second, cross-national study used health registries in Denmark and Norway to compare semaglutide initiators against people who started SGLT-2 inhibitors, a different class of diabetes drug. The researchers also ran a self-controlled analysis among all semaglutide users, comparing each patient’s risk during exposed and unexposed periods. Replicating the signal in a second country and against a different comparator drug strengthened the case that semaglutide, rather than diabetes alone, plays a role.
Randomized controlled trials, which are the gold standard for detecting drug side effects, have not confirmed the link. A pooled safety evaluation of completed placebo-controlled trials of Novo Nordisk-manufactured liraglutide and semaglutide found no clear NAION signal. But that absence is expected: clinical trials are typically too small and too short to detect events that occur in fewer than 1 in 10,000 patients. The registry studies, with hundreds of thousands of participants observed over years, are better suited to capture signals at that frequency.
Unanswered questions as a higher-dose Wegovy reaches the market
Several gaps in the evidence remain. No published study has stratified NAION incidence by semaglutide dose or duration of use, so clinicians cannot yet tell patients whether a lower dose carries less risk or whether the hazard grows over time. Individual-level clinical details, such as baseline visual acuity, optic-disc anatomy, and the exact timing of vision loss relative to the first injection, are absent from the registry analyses. Long-term visual outcomes and recurrence rates among affected patients have not been reported either.
These gaps take on added weight because the FDA recently approved Wegovy HD, a 7.2 mg formulation that delivers a higher weekly dose than any previously available semaglutide product. The agency noted that the pivotal trials for this higher-dose formulation did not show a clear safety signal for NAION, but those trials were not powered to detect such a rare outcome. As higher doses roll out to a broader population, the real-world data that first revealed the association at standard doses will become even more important for ongoing surveillance.
For now, regulators have not placed formal restrictions on semaglutide prescribing. Instead, they emphasize informed decision-making. Clinicians are urged to discuss the very small but serious risk of sudden vision loss with patients, particularly those who already carry multiple vascular risk factors or who have experienced NAION in one eye, a group considered at higher baseline risk for involvement of the fellow eye. Patients who notice abrupt blurring or a dark area in their vision in one eye are advised to seek emergency ophthalmic care and to report their medication history in detail.
How regulators and agencies coordinate on rare safety signals
The NAION decision illustrates how European regulators coordinate across borders when a potential safety issue emerges. The PRAC operates within a wider network of agencies that share pharmacovigilance data and expertise. At the EU level, national regulators participate in structured cooperation frameworks such as the EU Agencies Network, which helps align responses when new risks are identified with widely used medicines. This structure is particularly important for rare events, where pooling data across countries can make the difference between a missed signal and a confirmed association.
Communication is another challenge. Safety updates must reach clinicians and patients in all official EU languages, often on short timelines. Rather than drafting each language version from scratch, institutions rely increasingly on tools like the Commission’s machine translation services to produce rapid multilingual notices that can then be checked and finalized by human experts. For a widely prescribed drug such as semaglutide, that combination of automation and oversight allows consistent warnings about NAION to appear quickly on national websites, patient leaflets, and professional guidance documents.
Outside Europe, regulators are watching the evolving evidence closely. Some have opened their own safety reviews, while others are waiting for additional observational data before changing product labels. Divergent timelines are common with emerging signals, but the core scientific questions are shared: Is the risk consistent across populations? Does it vary with dose, duration, or indication? And can specific subgroups be identified in whom the benefits clearly outweigh the small but irreversible risk to vision?
What this means for patients and prescribers
For most people already taking semaglutide, the new classification does not mean they should stop treatment. The absolute risk of NAION remains very low, and for many patients with severe obesity or poorly controlled diabetes, the benefits of weight loss and glycemic improvement are substantial. Instead, the finding adds a new element to the risk–benefit discussion: a rare, unpredictable, and currently untreatable threat to sight.
In practical terms, prescribers may become more cautious about starting semaglutide in patients with a prior history of NAION, crowded optic discs, or multiple uncontrolled vascular risk factors, especially when alternative therapies are available. They may also counsel patients more explicitly about early visual symptoms and document shared decision-making in the medical record. Ophthalmologists, for their part, are likely to begin asking about GLP-1 agonist use when evaluating new cases of optic neuropathy and may report suspected associations to pharmacovigilance systems more systematically.
The story of semaglutide and NAION is still unfolding. As more data accumulate from higher-dose formulations and longer follow-up, the estimated risk may shift up or down, and researchers may uncover biological mechanisms that explain why some patients are vulnerable while most are not. For now, the PRAC’s conclusion that NAION is a very rare side effect provides a clearer framework for decisions: semaglutide remains an effective therapy for many, but it is no longer assumed to be neutral with respect to the optic nerve. Balancing its powerful metabolic benefits against a small but serious risk to vision is the new calculus that patients and clinicians must navigate together.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
