Ozempic, Wegovy, Mounjaro: what are the risks for the eyes?
Prescribed for type 2 diabetes or weight loss, GLP-1 agonists — semaglutide (Ozempic, Wegovy, Rybelsus) and tirzepatide (Mounjaro) — are among the most widely used treatments of recent years. Since 2024, several studies have raised a legitimate question: do these medicines increase the risk of a rare but serious optic nerve event? Here is what the data say today, without dramatising or minimising.
Non-arteritic anterior ischaemic optic neuropathy (NAION) is an infarction of the optic nerve: the small vessels that supply it become blocked, and vision in one eye drops suddenly and painlessly. It is the second most common cause of optic neuropathy after glaucoma, and there is currently no curative treatment — the loss of vision is most often permanent. It mainly affects people whose optic nerve is anatomically narrower, a congenital feature that can be spotted on a fundus examination.
What studies show on semaglutide
A first American study (Hathaway et al., JAMA Ophthalmology, July 2024), conducted on about 700 patients with diabetes and 980 overweight patients followed at a single centre, reported a higher risk of NAION in patients treated with semaglutide. A Danish cohort of more than 420,000 patients with diabetes then found a roughly doubled risk over five years. On this basis, the European Medicines Agency (EMA) concluded in June 2025 that NAION is a very rare adverse effect of semaglutide: at most 1 treated patient in 10,000 would be affected. The information has been added to the patient leaflets of Ozempic, Rybelsus and Wegovy.
In May 2026, a joint consensus of the North American Neuro-Ophthalmology Society (NANOS) and the American Academy of Ophthalmology confirmed this signal while recalling its limits: most of the available studies are retrospective, a causal link has not been formally established, and the risk remains low in absolute terms. The consensus is clear on one point: there is no reason to stop an ongoing treatment as a mere precaution.
What about tirzepatide (Mounjaro)?
The data are more recent and less abundant than for semaglutide, but they point in the same direction: since 2025, several publications and case reports have suggested a similar signal with tirzepatide. There is still too little hindsight to conclude with certainty, and health authorities continue to monitor these data. In the meantime, the same vigilance applies as with semaglutide.
Diabetes: the retina can also react at the start of treatment
Another phenomenon, distinct from NAION, has been known since the SUSTAIN-6 study: in patients with diabetes, a rapid and substantial drop in glycated haemoglobin in the first weeks of treatment may be accompanied by a transient worsening of diabetic retinopathy — especially when the retina was already affected before treatment. This phenomenon is not specific to any one medicine: it is linked to the speed of improvement in blood sugar. It generally stabilises after 12 to 18 months and does not call into question the value of the treatment, but it justifies closer monitoring at the start in the patients concerned.
In practice, what this changes
Before starting GLP-1 treatment, a fundus examination can assess the condition of the retina and the appearance of the optic nerve. In patients with diabetes who already have retinal involvement, closer monitoring is recommended during the first year. At any time during treatment, a sudden, painless drop in vision in one eye should prompt you to see an ophthalmologist urgently, on the same day if possible.