A rare optic nerve condition has followed semaglutide through three years of studies and one label change. Here is what each dataset found — and how large the risk actually is in absolute terms.
Non-arteritic anterior ischemic optic neuropathy — NAION — is a sudden loss of blood flow to the front of the optic nerve. It usually presents as painless vision loss in one eye, often noticed on waking, sometimes as a missing chunk of the visual field rather than total blackness. There is no proven treatment, and the damage is generally permanent.
It is uncommon but not exotic. NAION is the most frequent acute optic neuropathy in adults over 50, and the background risk is already elevated in people with diabetes, high blood pressure, sleep apnea, and a crowded optic disc anatomy — a population that overlaps heavily with everyone taking a GLP-1.
That overlap is exactly what makes the question hard to answer.
In July 2024, neuro-ophthalmologists at Massachusetts Eye and Ear published a retrospective cohort in JAMA Ophthalmology. Among patients seen at their clinic, semaglutide prescription was associated with a markedly higher rate of subsequent NAION — a hazard ratio of roughly 4.3 in the type 2 diabetes group and 7.6 in the overweight and obesity group. Most events occurred in the first year of exposure.
The headline numbers came with a large caveat. The cohort was drawn from a single specialist referral clinic, so the patients in it were already there because something was wrong with their eyes. That inflates absolute incidence dramatically. The authors framed it as a hypothesis, not a settled risk estimate.
Two much larger studies followed, and both landed on a smaller effect.
The Scandinavian study followed roughly 61,000 semaglutide users against SGLT-2 inhibitor users and counted 32 NAION events in total. Risk was elevated, but the absolute difference worked out to about one and a half extra cases per 10,000 person-years.
The OHDSI analysis, drawing on 14 databases and tens of millions of adults with type 2 diabetes, found no excess in the first year and a roughly doubled risk only after two or more years of use. Its authors described the association as real but smaller than previously reported, with signals concentrated in women and in Black patients that need separate confirmation.
In June 2025 the EMA's safety committee finished its review and required NAION to be added to the product information for Ozempic, Wegovy and Rybelsus — classified as very rare, meaning up to 1 in 10,000 people. The WHO issued a parallel alert the same month. The instruction attached to it is narrow: sudden or rapidly worsening vision means call your prescriber immediately, and if NAION is confirmed, semaglutide is stopped.
Sources: EMA PRAC meeting highlights, 2–5 June 2025 · WHO medical product alert, June 2025 · Hathaway et al., JAMA Ophthalmology 2024 · Simonsen et al., Diabetes, Obesity and Metabolism 2025 · Cai et al., JAMA Ophthalmology 2025 (OHDSI).
Very rare is doing real work in that sentence. A doubled risk of an uncommon event is still an uncommon event. The registry data put the excess at roughly one extra case per several thousand people treated for a year — against a drug class with documented cardiovascular and kidney benefit in the same populations.
The symptom to know is specific. Not blurriness, not dry eye, not the transient focus changes that come with shifting blood sugar. NAION is abrupt, painless, one-sided, and does not improve over hours. That warrants a same-day call, not a wait-and-see.
Nobody should stop a GLP-1 over this on their own. Stopping abruptly has its own consequences, and the decision belongs with the prescriber who knows your risk factors — particularly if you already have diabetic eye disease, sleep apnea, or a prior NAION event in the other eye.
GLP1ShotDay logs side effects against the exact dose and week they appeared, so if something does need reporting, the timeline is already written down rather than reconstructed from memory at an appointment.
This article summarizes published research and regulatory decisions for general education. It is not medical advice and does not replace guidance from your prescriber. Report any sudden change in vision to a clinician immediately, and never start, stop, or adjust a prescription based on an article.