Drug facts

Semaglutide and vision loss

By David Chen·Editor·Reviewed September 2026

What the US labels say¹

The eye warning in both US labels is about the retina, not the optic nerve, and it applies to people with type 2 diabetes. Wegovy’s section 5.8 and Ozempic’s 5.3 report the same two-year trial of semaglutide 0.5 mg and 1 mg in type 2 diabetes with high cardiovascular risk:

The Wegovy label adds two more readings. In its own weight-reduction trial in people with type 2 diabetes, diabetic retinopathy was reported by 4% on the injection against 2.7% on placebo; and in section 6.1, retinal disorders of any kind were reported by 6.9% on semaglutide 2.4 mg, 6.2% on 1 mg and 4.2% on placebo.

Then the sentence readers ask about most, quoted from both labels: “Rapid improvement in glucose control has been associated with a temporary worsening of diabetic retinopathy.” The labels add that “the effect of long-term glycemic control with semaglutide on diabetic retinopathy complications has not been studied,” and that people with a history of it should be monitored for progression.

What the labels ask patients to do is narrow. Section 17 instructs prescribers to tell patients with type 2 diabetes “to contact their physician if changes in vision are experienced during treatment,” and each Medication Guide repeats it in plain words.²

The census¹

Because the question is really “does my label warn me about this,” here is the count. Both PDFs were taken from the manufacturer’s own site and read end to end — prescribing information, Medication Guide and Instructions for Use together:

So the answer is plain: the current US labels do not name NAION, and neither label’s postmarketing section lists an eye event of any kind. Of the seven uses of “vision” in the Wegovy document, three are the pen and syringe instructions telling people who are blind or have vision problems not to inject without help; three are the instruction to report changes in vision — once in section 17 and twice in the Medication Guide’s bullet, which uses the word in both of its sentences; and one is “blurred vision” in the list of low-blood-sugar symptoms. None of them describes a sudden loss of sight.

What Europe’s regulator concluded³

The European Medicines Agency reviewed the question and announced its conclusion in June 2025. In the EMA’s words, its safety committee “has concluded that NAION is a very rare side effect of semaglutide (meaning it may affect up to 1 in 10,000 people taking semaglutide).” The review covered Ozempic, Rybelsus and Wegovy, and drew on non-clinical studies, clinical trials, post-marketing surveillance and the literature.

The size of it, again in the EMA’s words:

Results from several large epidemiological studies suggest that exposure to semaglutide in adults with type 2 diabetes is associated with an approximately two-fold increase in the risk of developing NAION compared with people not taking the medicine. This corresponds to approximately one additional case of NAION per 10,000 person-years of treatment.

The EMA also says there is no identified time interval for when NAION may develop after treatment starts, that patients who experience “a sudden loss of vision or rapidly worsening eyesight” should contact their doctor without delay, and that if NAION is confirmed, treatment with semaglutide should be stopped. Those are the EMA’s recommendations, reported here, not this site’s advice.

That conclusion is now in the European label. Section 4.4 of the EU product information carries a NAION heading, stating that epidemiological data indicate an increased risk during treatment with semaglutide and that “a sudden loss of vision should lead to ophthalmological examination and treatment with semaglutide should be discontinued if NAION is confirmed.” In the section 4.8 table of side effects it sits under eye disorders, in the very rare column, beside diabetic retinopathy.⁴

Two regulators, the same molecule, different labels. The EU document names the condition and tells prescribers what to do about it; as of the revisions above, the US documents do not mention it at all.

What NAION is⁶

The EMA describes NAION as “an eye condition that may cause loss of vision,” and the European package leaflet tells patients it “may cause loss of vision to one of your eyes without any pain,” with instructions to contact a doctor immediately on noticing sudden or gradually worsening eyesight.⁴ The Danish investigators who published one of the cohorts put the stakes plainly: NAION, they write, is “an untreatable condition often causing severe and irreversible visual loss in the affected eye.”

It is also uncommon. In the largest study to report a rate on semaglutide, 14.5 cases turned up per 100,000 person-years of treatment⁸, and the Danish cohort found 218 cases among 424,152 people with diabetes over five years⁶. A doubling of a number that small is still a small number, which is the EMA’s own framing.

The studies behind it⁷

The evidence is observational and it does not agree with itself. The second column is the one to read: several of these studies compare semaglutide with a different drug, one pools the whole GLP-1 class, and one cannot separate semaglutide from tirzepatide at all.

Seven of these eleven report an increased risk, usually around twofold — one of the seven for the GLP-1 class rather than semaglutide on its own. Two report no association, one reports a lower risk in diabetes, and one cannot separate semaglutide from tirzepatide. The neuro-ophthalmology societies looked at the same set and wrote a joint statement: the evidence is mostly retrospective records research with real limitations, “many report a small possible increased risk of NAION in patients taking GLP-1 RAs such as semaglutide, some studies report no correlation, and the overall magnitude of the risk of NAION remains low,” with the decision to start, continue or stop left to each patient and their care team.¹⁸

What the randomized trials found¹⁶

The randomized record points the other way, and it is thin for the reason you would expect: NAION is too rare for trials to count well. A pooled safety analysis of the placebo-controlled trials of the manufacturer’s two molecules — liraglutide and semaglutide, 96,829 participants, 205,777 participant-years — found three confirmed cases among people on a GLP-1 drug and five among people on placebo: about three cases per 100,000 participant-years on the drugs against about six on placebo. Its conclusion is that these data “do not show an increased incidence of NAION.” It pools two molecules, and the trials in it were the manufacturer’s own.

A meta-analysis of 78 semaglutide trials with 73,640 participants found no increase in eye disorders overall (odds ratio 1.01) and none in diabetic retinopathy (1.04), but did find an association with NAION — odds ratio 3.92, confidence interval 1.02 to 15.02, resting on a handful of events. Its own sequential analysis concluded there is not yet enough data for a definitive answer on NAION, while there is for retinopathy.¹⁷

Blurred vision, and whether it goes away¹

“Blurred vision” is the most searched eye complaint on this drug, and here is the honest state of the documents. Neither US label lists blurred vision as a side effect of semaglutide. The phrase appears once in each Medication Guide, in the list of symptoms of low blood sugar — alongside dizziness, sweating, shakiness and confusion — which matters because hypoglycemia is a labeled risk when semaglutide is combined with insulin or a sulfonylurea (see interactions).

Neither label says whether vision changes resolve. The nearest thing to an answer is the retinopathy sentence: rapid improvement in glucose control “has been associated with a temporary worsening of diabetic retinopathy” — temporary is the label’s word, about retinopathy, in people with diabetes. NAION is a different thing, and the cohort authors above describe its visual loss as often irreversible.⁶

So the labels give one instruction and no reassurance: report changes in vision to the prescriber. Which of the possibilities applies to any one person — blood sugar, the retina, the optic nerve, or none of them — is not something a page can work out.

What none of this settles

Whether semaglutide causes NAION is unresolved. The studies that find an association are records research, which can be confounded by who gets prescribed what; the studies that find nothing are also records research, with the same limitation; the randomized trials are too small for the question. One regulator has judged the evidence sufficient to label a very rare risk, and the other has not changed its label. Both positions are reported above rather than averaged into one.

This page reports the labels, the EMA’s conclusion and the published studies, quoted and attributed, and makes no causal claim of its own. The census counts were run over the full text of each PDF, extracted twice to guard against the two-column layout hiding a word. Nothing here is medical advice, and no one should start or stop a prescription on the strength of it — see the medical disclaimer. Related: the full side-effect table, long-term safety, and the boxed warning.