NAION and GLP‑1s: Do Ozempic and Mounjaro Cause Blindness?

In the third week of September 2026, a wave of stories reported that more than 200 people are suing Novo Nordisk and Eli Lilly, alleging that their GLP‑1GLP‑1 (glucagon-like peptide-1) A gut hormone released after eating that boosts insulin, slows the stomach, and signals fullness. Drugs like semaglutide mimic it. drugs caused sudden, permanent vision loss. Most of the coverage leaned on one number from a 2024 study: people taking semaglutide had four to seven times the risk of a rare optic nerve condition called NAIONNAION (non-arteritic anterior ischemic optic neuropathy) A sudden loss of blood flow to the front of the optic nerve, sometimes called an eye stroke. It usually causes painless, permanent vision loss in one eye..

That number is real, but it’s also the least reliable estimate available. More than a dozen larger studies published since have landed well below it. Here is what the full body of evidence shows, and what it means if you take one of these drugs.

What NAION Is, and Its Symptoms

NAION stands for non-arteritic anterior ischemic optic neuropathy. In plain terms, it’s a stroke of the optic nerve: blood flow to the front of the nerve drops suddenly and the nerve fibers are damaged. The classic symptom is sudden, painless vision loss in one eye, often noticed on waking. Most of the lost vision never comes back, and no treatment has been shown to restore it.

NAION existed long before GLP‑1 drugs. It’s the most common sudden optic nerve disorder in adults over 50, striking roughly 10 of every 100,000 people in that age group each year. Its risk factors look a lot like the profile of a typical GLP‑1 patient: diabetes, high blood pressure, sleep apneasleep apnea A condition in which breathing repeatedly stops and restarts during sleep, dropping oxygen levels and fragmenting rest. It is common, often undiagnosed, and a cause of daytime exhaustion., and older age. The biggest single risk factor is anatomical. Some people are born with a small, crowded optic discoptic disc The spot at the back of the eye where the optic nerve leaves for the brain. Some people are born with a small, crowded disc, which leaves the nerve fibers less room and makes them more vulnerable to a drop in blood flow., which only an eye exam can reveal.

That overlap is the central problem in this debate. The people most likely to be prescribed semaglutide were already the people most likely to develop NAION.

Where the Scare Started

In July 2024, researchers at Massachusetts Eye and Ear reported in JAMA Ophthalmology that patients with type 2 diabetes prescribed semaglutide had a hazard ratiohazard ratio A number comparing how fast an outcome happens in two groups. Above 1 means more risk in the treated group, below 1 means less, and exactly 1 means no difference. of 4.28 for NAION. For patients with overweight or obesity, it was 7.64.

It was a legitimate and important signal, but a poor guide to your personal risk. Every patient in the study had been referred to a neuro-ophthalmologyneuro-ophthalmology The medical specialty that treats vision problems coming from the optic nerve and brain rather than from the eye itself. Its clinics see the most complex, highest-risk cases. clinic, so the rates were extraordinary: 8.9% of the semaglutide-treated diabetes patients developed NAION within three years, versus a small fraction of 1% in the general population. The numbers were also tiny: 17 cases versus 6 in the diabetes group, and 20 versus 3 in the obesity group. The authors themselves noted that the study could not show the drug caused the condition.

What the Larger Studies Found

Since then, researchers have tested the question in national registries, veterans’ records, and insurance claims covering millions of patients. The strongest studies compare people starting semaglutide with people starting a different second-linefirst-line / second-line First-line is the treatment doctors try first for a condition. Second-line drugs are added or switched to when the first choice isn’t enough or isn’t tolerated. diabetes drug, usually an SGLT2 inhibitorSGLT2 inhibitor A class of diabetes drugs, including Jardiance and Farxiga, that lowers blood sugar by making the kidneys pass extra sugar into the urine. Often used as the comparison group in GLP‑1 safety studies. such as Jardiance or Farxiga, which makes the two groups more alike from the start.

How to read the table: the middle column compares people starting a GLP‑1 with similar people starting a different diabetes drug. The last column turns that into real people. For context, NAION strikes about 1 in every 10,000 adults over 50 each year, with or without any drug.

All studies above are in people with type 2 diabetes, except that the eye clinic study also included people with obesity alone.

There is probably a real signal. Different countries and research teams keep finding roughly double the risk in people with type 2 diabetes, and two 2026 meta-analysesmeta-analysis A study that pools the results of several earlier studies to get a more precise overall estimate. It is only as good as the studies that go into it. landed in the same range.

The absolute riskabsolute risk The actual chance of something happening, such as 8 cases per 10,000 women per year. It shows how big a risk really is, which a percentage change alone can hide. is small. Doubling a rare event still leaves a rare event. Across the better studies, the added risk works out to about 1 to 6 extra cases for every 10,000 people treated for a year. The European Medicines Agency reached essentially the same conclusion, classifying NAION as a “very rare” side effect of semaglutide, meaning up to 1 in 10,000 users.

Is It the Drug, or the Patients?

The most careful study so far, published this summer in the Annals of Internal Medicine using Swedish national data, found the familiar doubling in its main analysis. The researchers then limited the comparison to patients already taking metforminmetformin The standard first medication for type 2 diabetes. It lowers blood sugar mainly by reducing how much sugar the liver releases., which makes the two groups more alike in how advanced their diabetes was. The excess risk largely disappeared. It’s one analysis with few events, but it suggests “doubles the risk” is an upper estimate rather than a settled fact.

That points to confoundingconfounding When a third factor explains a link between a drug and an outcome. If sicker people are more likely to be prescribed a drug, the drug can look harmful even when the underlying illness is to blame.: the patients who get GLP‑1 drugs may simply be sicker in ways health records don’t fully capture, with more obesity, more untreated sleep apnea, and longer-standing diabetes. Awareness adds a second bias. Since 2024, eye doctors have been primed to connect optic nerve swelling with semaglutide, which makes a diagnosis, and a report, more likely in GLP‑1 users.

If the drug does play a role, the leading hypothesis is a rapid metabolic shift. In the UK study, the risk was highest in the first six months and in patients whose HbA1cHbA1c (hemoglobin A1c) A blood test that shows your average blood sugar over the past ~3 months. fell by a full percentage point or more. That mirrors a known effect: fast drops in blood sugar can temporarily worsen diabetic retinopathydiabetic retinopathy Damage to the small blood vessels of the retina, the light-sensing layer at the back of the eye, caused by long-term high blood sugar. It is a leading cause of vision loss in people with diabetes., which is already on the US label. Dehydration and nighttime dips in blood pressure could also reduce blood flow to a vulnerable nerve. If that’s right, most of the risk would fall on a small group with a crowded optic disc to begin with.

Mounjaro, Zepbound, and Weight Loss Only

People taking semaglutide for weight loss without diabetes. The dramatic 7.6× figure came from this group, in the small specialty-clinic study, and larger studies haven’t reproduced it. The 21-country study found no significant link in obesity alone, and neither did a 2026 meta-analysis. A risk hasn’t been shown here, but the data are too thin to rule one out.

Tirzepatide (Mounjaro, Zepbound). Eli Lilly is named in the lawsuits, but the evidence against tirzepatide is weak. The most direct comparison, a 2026 study of 205,000 patients with diabetes, paired up so the two groups had similar ages and health,, found that tirzepatide users had less than half the NAION rate of people on older GLP‑1 drugs. Regulators’ warnings so far cover only semaglutide. That’s reassuring, but not proof of safety. Whether any risk extends across the whole class is also open: one large study found a similar signal for all GLP‑1 drugs combined, and a 2026 study flagged liraglutide (Saxenda, Victoza).

What Regulators Have Done

The European Medicines Agency concluded in June 2025 that NAION is a very rare side effect of semaglutide and added it to the product information for Ozempic, Rybelsus, and Wegovy, advising that semaglutide be stopped if NAION is confirmed. The World Health Organization and the UK’s medicines regulator, the MHRA, issued similar alerts. In May 2026, the American Academy of Ophthalmology and the North American Neuro-Ophthalmology Society called the possible increase in risk small and recommended that patients and prescribers decide together whether to start, continue, or stop.

In the US, the FDA has not added a warning. The current Wegovy label, revised February 2026, mentions only diabetic retinopathy.

What the Headlines Got Wrong

The FDA database numbers. Several stories cited nearly 4,000 reports of vision problems tied to Ozempic and Wegovy in the FDA’s adverse eventadverse event Any harmful or unwanted medical event during a study or treatment, whether or not the drug caused it. database. Anyone can file those reports, including patients and lawyers. A report doesn’t show that the drug caused the problem, and the category lumps common blurred vision together with optic nerve damage. Reports also surge after media coverage and once litigation begins.

“Blindness.” NAION almost always affects one eye at a time. The vision loss is serious and permanent, but it isn’t usually total blindness. The real concern is the second eye: in a major follow-up study, about 15% of people with NAION in one eye developed it in the other within five years.

This pattern isn’t new. An analysis of the first round of coverage in 2024 and 2025 found that nearly half of news stories contained misleading claims.

What to Actually Do

If you take a GLP‑1 and have never had NAION: don’t stop on your own because of the news. The proven benefits, including fewer heart attacks and strokes, kidney protection, and substantial weight loss, rest on far stronger evidence than this risk does. Stopping also has consequences of its own, covered in what happens when you stop semaglutide.

Know the warning sign. Sudden, painless loss or dimming of vision in one eye, especially on waking, needs an eye exam the same day, whether or not you take a GLP‑1. Other causes, including giant cell arteritisgiant cell arteritis Inflammation of medium and large arteries, mostly in people over 50. It can cause sudden vision loss and is an emergency because prompt steroid treatment can protect the other eye., are treatable if caught quickly and can threaten both eyes if missed.

If you’ve had NAION before: this is where the calculation genuinely changes. Your other eye is already at elevated risk, and both European regulators and the latest expert review advise against semaglutide in this situation. Discuss alternatives with your prescriber and eye doctor.

If you have diabetes: keep up your annual dilated eye exam, where drops widen the pupil so the doctor can see the optic nerve. It’s also how a crowded optic disc gets found. Mention other risk factors, such as untreated sleep apnea, to your prescriber.

No one has shown that slower dose increases, staying hydrated, or switching to tirzepatide lowers NAION risk. These are reasonable questions to ask, but they aren’t proven protection.