The Food and Drug Administration on Monday ordered a new warning on every GLP-1 drug sold in the US. It covers a rare cause of sudden blindness in one eye, a condition called NAION. The Ozempic vision loss warning also goes on Wegovy, Mounjaro, Zepbound and eight other drugs.
The FDA’s drug safety communication calls it a “potential increased risk” of rare events, not a proven effect. The agency’s own trial numbers put the extra risk at a few cases per 100,000 patients a year. That is small for any one person, but the vision loss can be permanent.
The scale of use is what makes a rare risk matter. Lilly’s Mounjaro and Zepbound brought in $36.51 billion in 2025, and Novo’s semaglutide drugs $36.19 billion, BioSpace reports.
What NAION is, and which drugs now carry the warning
NAION stands for non-arteritic anterior ischemic optic neuropathy. The FDA says it is “caused by decreased blood flow to the front part of the optic nerve.” It causes sudden, painless loss of vision in one eye, which can be permanent. NBC News reports there is no known treatment.
The new text goes in the Warnings and Precautions section of each label. That is a step below the boxed warning these drugs already carry for thyroid tumors. The FDA’s list covers 12 products approved since 2005:
| Drug ingredient | Brand names on the FDA list |
|---|---|
| Semaglutide | Ozempic, Wegovy, Rybelsus |
| Tirzepatide | Mounjaro, Zepbound |
| Liraglutide | Victoza, Saxenda, Xultophy (with insulin) |
| Dulaglutide | Trulicity |
| Exenatide | Byetta (now sold only as generic exenatide) |
| Lixisenatide | Soliqua (with insulin) |
| Orforglipron | Foundayo (a pill approved in 2026) |
That goes further than Europe did. In June 2025 the European Medicines Agency’s safety committee concluded that NAION is a “very rare” side effect of semaglutide alone. It defines very rare as up to 1 in 10,000 people.
How the FDA reached its conclusion
The FDA says its review began after a July 2024 study in JAMA Ophthalmology. Jimena Hathaway of Massachusetts Eye and Ear led it. Diabetes patients prescribed semaglutide had 4.28 times the hazard of NAION of similar patients on other drugs.
The agency then pooled 126 randomized trials with 134,253 patients, 78,217 of them on a GLP-1 drug. There were only 15 probable NAION cases: 10 among GLP-1 users and 5 on placebo.
It also ran a real-world study in its Sentinel system. It followed adults with type 2 diabetes who started one of three drug classes from 2019 to 2024. GLP-1 users had a 43% higher rate than people starting DPP-4 drugs such as Januvia. Against SGLT-2 drugs such as Jardiance, there was no statistically significant difference.
The FDA found NAION reports for every GLP-1 drug on the market, but no biological mechanism in animal data. It also says adverse-event reports are hard to read. That is because diabetes, high blood pressure and sleep apnea raise NAION risk, and they are common among people on these drugs.
The risk in numbers
A “four times the risk” headline sounds alarming until you see what it is four times of. We put every absolute figure from the sources we read into one table. The last column is our subtraction from the published rates.
| Source | Rate on GLP-1 drug | Rate on comparison | Extra cases per 100,000 patient-years |
|---|---|---|---|
| FDA, 126 clinical trials | 6.4 per 100,000 | 3.8 (placebo or other drugs) | About 2.6, or 1 per roughly 38,000 patient-years |
| FDA Sentinel, type 2 diabetes | 30 to 32 per 100,000 | 26 (DPP-4 drugs); 31 (SGLT-2 drugs) | About 4 to 6 vs DPP-4; none vs SGLT-2 |
| Danish-Norwegian registry study, 2025 | 21.9 to 29.0 per 100,000 (semaglutide) | 9.2 to 11.8 (SGLT-2 drugs) | 14.1 after adjustment |
| PLOS Medicine meta-analysis, 2026 | Not pooled | Not pooled | About 14, given by the authors as 1 per 7,000 patients a year |
| EMA, semaglutide | “Very rare”: up to 1 in 10,000 people | Not given | Not given |
For scale, the FDA cites estimates that NAION already strikes 2.3 to 10.2 people per 100,000 over age 50 each year.
Why the scariest figure is the wrong one to quote
The 2024 Massachusetts Eye and Ear study found an 8.9% three-year rate of NAION among diabetes patients on semaglutide. But its patients had already been referred to neuro-ophthalmologists at one hospital. That rate says nothing about the general population, and the authors called their result an association that needs further study.
The evidence is also split. A 2026 review of nine large cohort studies found four showing higher risk and four showing none. Its authors blamed confounding by indication: sicker patients are the ones who get the drug.
The FDA’s notice does not tell people without eye symptoms to stop their medicine. Its advice on stopping applies only after NAION occurs, and it routes that choice through a doctor. Here is what it asks of patients and caregivers:
- Know that GLP-1 drugs carry a potential increased risk of NAION.
- Watch for sudden, painless vision loss in one eye, often noticed on waking. Other signs are blind spots, tunnel vision, odd colors, or blurry central vision.
- Contact a health care professional right away if any of these appear.
- After NAION, ask your clinician whether to keep taking the drug. The other eye is then at higher risk.
Doctors are told to send suspected cases promptly to an ophthalmologist. The FDA says to consider stopping the drug after NAION unless the benefits clearly outweigh the risk. Europe is firmer: it says semaglutide should be stopped once NAION is confirmed.
Sudden vision trouble can also signal a stroke, which is why the “E” in BE FAST stands for eyes. Either way, the FDA’s advice is to get help right away.
What Novo Nordisk and Eli Lilly say
Novo Nordisk spokesperson Liz Skrbkova told USA TODAY the data do not “establish a cause-and-effect relationship.” She said the company remains “confident in the benefit-risk profile” of its GLP-1 drugs. Lilly said it is “carefully evaluating the update and will respond to the FDA promptly,” USA TODAY reported. USA TODAY also counts dozens of lawsuits filed against the two companies.
BMO analysts told BioSpace they see “limited risk” for Novo and Lilly “given rarity in incidence.” The drugs also have approved uses beyond weight loss, including cutting major heart events in some adults with heart disease. That matters because heart disease is the leading cause of death for US women, according to the CDC.
Drugmakers can propose changes to the wording before it is final, Investing.com notes. The next step is new prescribing information for each of the 12 products. Anyone who sees a suspected case can report it to the FDA’s MedWatch program at 1-800-332-1088.
