GLP-1 Drugs and Sudden Vision Loss (NAION): What the Studies Show

Several large studies have linked semaglutide and other GLP-1 drugs to a rare cause of sudden vision loss called NAION. Others found no clear link, and pooled data from placebo-controlled trials did not show one. Where an increase was seen, the extra risk was small in absolute terms: from about 1 to 16 additional cases per 10,000 people, depending on the study and how long people were followed. Europe's medicines regulator lists NAION as a very rare side effect of semaglutide; the US labels we checked on 22 October 2026 do not mention it.

This article explains what NAION is, what each type of study found, why the results differ, and what regulators and eye specialists have concluded. It covers evidence published up to 22 October 2026 and gives no treatment or dosing advice. For the wider safety picture of these medicines, see GLP-1 drugs beyond weight loss.

The short version

• The first signal came from one eye hospital. A 2024 study of patients seen by neuro-ophthalmologists in Boston found NAION was several times more common in people prescribed semaglutide. The numbers were small and the patients were already being seen for eye problems.¹
• Bigger studies mostly found a smaller increase, from about 1.3 to 2.8 times the comparison group, in people with type 2 diabetes. Some found no increase at all.^2345678
• The absolute risk stays low. In a 2026 US study, 8.5 in 10,000 GLP-1 users had an ischaemic optic neuropathy within 18 months, against 5.5 in 10,000 people starting a different diabetes drug.⁸
• Randomised trials have not settled it. Pooled placebo-controlled trials of semaglutide and liraglutide found three confirmed cases among nearly 65,000 people on the drugs and five cases among about 32,000 on placebo.⁹
• Regulators disagree on labelling. The European Medicines Agency (EMA) added NAION as a "very rare" side effect of semaglutide in 2025. The US prescribing information for Wegovy, Ozempic and Zepbound we checked does not list it.¹⁴¹⁵
• What to do: sudden loss of vision in one eye needs urgent medical care. If you take a GLP-1 drug and are worried, talk to your prescriber rather than stopping on your own.¹²¹⁴

What NAION is

NAION stands for non-arteritic anterior ischaemic optic neuropathy. In plain words, the front of the optic nerve, which carries signals from the eye to the brain, is damaged because its blood supply drops. "Non-arteritic" separates it from a type caused by inflamed arteries. It is a rare but serious cause of sudden, often permanent, loss of vision.¹⁰¹¹ NAION makes up about three-quarters of ischaemic optic neuropathy in adults.⁸

Its main risk factors overlap with the people most likely to be prescribed GLP-1 drugs: older age, diabetes, obesity, high blood pressure, high cholesterol and smoking. Every confirmed case in the pooled trial analysis had at least one of these.⁹¹¹ That overlap is the central difficulty in this research: when NAION is more common in GLP-1 users, it is hard to tell how much is the drug and how much is the illness it was prescribed for.

What the studies found

All but one of these are observational studies, so they show associations, not cause and effect. A hazard ratio (HR) or risk ratio above 1 means NAION was more common in the GLP-1 group; if its 95% confidence interval includes 1, the difference could be chance. Our guide to correlation versus causation explains why this matters.

Tirzepatide has much less data. A 2025 US study that grouped semaglutide and tirzepatide users with type 2 diabetes found NAION in 0.04% of them over two years, against 0.02% in a matched comparison group (HR 1.76).⁴

Why the results differ

Who is being compared. Studies that compared GLP-1 users with people starting another modern diabetes drug, usually an SGLT2 inhibitor, tried to match people with similar illness. Even so, GLP-1 drugs are often started later in diabetes, when people have more complications. The Swedish study found that the extra risk shrank sharply when it compared only people already taking metformin, a step meant to match disease severity more closely. The authors suggested the remaining difference "may reflect residual confounding".⁷

How NAION is counted. Most studies rely on diagnosis codes in medical records, which can miss cases or include the wrong ones. One study checked every coded case against clinicians' notes and found about 85% were correct; its result held up under stricter definitions, and GLP-1 use was not linked to other eye conditions such as diabetic retinopathy, which suggests the finding is specific to NAION.⁶

Trials versus everyday practice. Randomised trials remove most of the confounding problem but are too small for a condition this rare: the pooled trial analysis had eight confirmed cases in total.⁹ The EMA, which reviewed trial data as well, read them as pointing to a slightly higher risk.¹⁴ A 2026 meta-analysis that separated the evidence by study type found an increased risk only in observational studies of people with diabetes, not in trials or in people taking semaglutide for weight loss.¹⁰ Another 2026 meta-analysis rated the overall evidence as low or very low certainty.¹¹ Our guide to evaluating a meta-analysis explains these ratings.

Relative risk versus absolute risk

A "two-fold" risk sounds alarming, but it applies to a rare event. In the Danish and Norwegian data, NAION affected about 2 to 3 per 10,000 semaglutide users each year, compared with about 1 per 10,000 on SGLT2 inhibitors.² The US claims study estimated that about 3,333 people would need to take a GLP-1 drug instead of an SGLT2 inhibitor for 18 months for one extra case of ischaemic optic neuropathy to occur, if the link were causal.⁸

These figures come mainly from people with type 2 diabetes. Less is known about people without diabetes, and NAION, when it happens, can mean permanent loss of sight in an eye.¹⁰

What regulators and specialists say

• Europe. In June 2025 the EMA's safety committee concluded that NAION is a very rare side effect of semaglutide, affecting up to 1 in 10,000 people, about one extra case per 10,000 person-years. It said clinical trial data also pointed to a slightly higher risk with semaglutide than placebo, asked for the product information of Ozempic, Rybelsus and Wegovy to be updated, and said treatment should be stopped if NAION is confirmed.¹⁴
• United States. The current prescribing information for Wegovy (revised June 2026), Ozempic (August 2026) and Zepbound (September 2026) does not mention NAION or ischaemic optic neuropathy.¹⁵
• Eye specialists. A 2026 consensus statement from the North American Neuro-Ophthalmology Society and the American Academy of Ophthalmology said many studies report "a small possible increased risk", some report none, "the overall magnitude of the risk of NAION remains low", and the decision to start, continue or stop a GLP-1 drug should be shared between patient and care team.¹² A September 2026 editorial in the same journal described NAION with these drugs as "an important rare issue, lacking adequate data".¹³

What is still unknown

• Whether the link is causal. No study so far separates the drug from the diseases it treats with confidence.⁷⁸¹¹¹²
• Who might be most at risk. In the US claims study, the extra risk was concentrated in men, people over 50 and those with heart disease or existing eye conditions.⁸ These are hypotheses, not a screening rule.
• Risk outside type 2 diabetes, and for tirzepatide and newer drugs, which have far less data.⁴¹⁰

What this means for you

1. Treat sudden vision loss as an emergency. Sudden loss of vision or rapidly worsening eyesight in one eye needs same-day medical care, whatever medicines you take. The EMA asks people on semaglutide to contact their doctor without delay if this happens.¹⁴
2. Talk to your prescriber, don't stop on your own. The specialists' advice is a shared decision that weighs your own eye and heart risks against the benefits of treatment.¹² Stopping a GLP-1 drug has its own effects; see stopping semaglutide or tirzepatide.
3. Mention your medicines at eye appointments, including any GLP-1 drug, so an eye specialist can take it into account.
4. Report suspected side effects through FDA MedWatch or your country's reporting system. Our article on peptide clinical safety explains how these reports are used.

Frequently asked questions

Does semaglutide cause blindness?

There is no proof that it does. Some studies found NAION, which can cause permanent vision loss in one eye, was more common in semaglutide users, mainly people with type 2 diabetes; others did not, and randomised trials have not confirmed it. Where a link was found, the extra risk was small.^2391012

Is tirzepatide affected too?

There is much less data. One 2025 study that grouped semaglutide and tirzepatide users found a small increase in NAION, but its published summary does not give a separate result for tirzepatide.⁴ The US Zepbound label does not mention NAION.¹⁵

Why does Europe list NAION as a side effect when the US doesn't?

The EMA judged that the evidence was strong enough to describe NAION as a very rare side effect of semaglutide.¹⁴ The US labels we checked had not added it as of 22 October 2026.¹⁵ Different regulators can weigh the same studies differently; neither position proves that the drug causes NAION.

Should I get my eyes checked before starting a GLP-1 drug?

None of the sources reviewed here sets a screening rule. If you already have an eye condition or other NAION risk factors, raise it with your prescriber; the specialist consensus recommends deciding together.¹²