Why the FDA Put a Vision-Loss Warning on Every GLP-1 Label

The Food and Drug Administration has directed the makers of every marketed GLP-1 receptor agonist to add information about a potential increased risk of non-arteritic anterior ischemic optic neuropathy, or NAION, to the Warnings and Precautions section of their labels. The requirement, laid out in a safety communication issued November 9, 2026, covers the whole class: Novo Nordisk's Ozempic, Wegovy, Rybelsus, Victoza, Saxenda and Xultophy, Eli Lilly's Mounjaro, Zepbound, Trulicity and the new oral orforglipron product Foundayo, and Soliqua, approvals spanning 2005 through this year's orforglipron launch.

NAION is a rare condition in which blood flow to the front of the optic nerve drops, causing sudden, painless vision loss in one eye that can be permanent. Its underlying cause is not completely understood, and its known risk factors (smoking, diabetes, high blood pressure, sleep apnea, anemia, kidney disease) are precisely the conditions that cluster in the population taking GLP-1 drugs. That overlap is what made the agency's own adverse-event database nearly unreadable: reports filed through the Adverse Event Monitoring System could not be separated from background risk, so FDA set them aside and turned to harder evidence.

What the numbers actually show

The centerpiece is a meta-analysis of 126 randomized controlled trials across GLP-1 development programs, 134,253 patients and nearly 289,000 patient-years of exposure. Fifteen probable NAION events emerged: ten among GLP-1 users, five among placebo patients, none among patients on other glucose-lowering drugs. Expressed as rates, that is 6.4 events per 100,000 patient-years on a GLP-1 versus 3.8 on comparator. The relative difference is real. The absolute base is tiny, fifteen events across that exposure, and the general-population incidence in adults over 50 runs from 2.3 to 10.2 cases per 100,000 persons per year, a band the trial rates sit inside.

The second leg, a Sentinel retrospective cohort of adults with type 2 diabetes who started a GLP-1, an SGLT-2 inhibitor or a DPP-4 inhibitor between 2019 and 2024, cuts both ways. Against SGLT-2 inhibitors, GLP-1 users showed no statistically significant excess (hazard ratio 1.19, confidence interval 0.91 to 1.54). Against DPP-4 inhibitors, the hazard ratio was 1.43, with a confidence interval of 1.02 to 2.01 that only barely excludes 1.0. Together the analyses say the signal is present but fragile, sensitive to comparator and method.

A label decision without a mechanism

That fragility makes the agency's conclusion notable. FDA reviewed animal data and found no mechanism by which GLP-1 activity could raise NAION risk. What tipped the balance was the external literature: after Hathaway and colleagues reported elevated NAION risk among semaglutide patients in JAMA Ophthalmology in mid-2024, a study FDA flagged with several limitations, methodologically rigorous observational work kept arriving at the same direction of effect, consistently showing higher risk in GLP-1 initiators versus DPP-4 or SGLT-2 comparators. Cases had also been reported for every marketed product in the class. None of the individual legs was decisive. Stacked, they were enough.

The remedy FDA chose is telling. A Warnings and Precautions entry discloses the potential risk without a boxed warning, without a contraindication, and without any restriction on prescribing. Analysts at Citi, cited by Fierce Pharma, argued the treatments should preserve a favorable benefit-risk profile given NAION's rarity and the drugs' substantial weight, glycemic, cardiovascular, renal and sleep apnea benefits, and noted that PDE5 inhibitor erectile-dysfunction therapies such as Viagra have carried NAION warnings since 2005 without denting adoption. The precedent is an analyst's framing, but it describes a real regulatory pattern: vision-loss warnings on widely used drugs are survivable events.

The specialist community had largely converged before the agency moved. In August 2026, a consensus statement from the North American Neuro-Ophthalmology Society and the American Academy of Ophthalmology put the relative risk at roughly two-fold while stressing that the overall magnitude remains low, and declined to recommend screening for NAION among patients on GLP-1s. Lilly, in a statement, said the communication applies to all medicines with GLP-1 activity and references no findings specific to its products, and that it is evaluating the update and will respond promptly.

What changes for patients and prescribers

For the many people on these drugs, the practical shift is modest. FDA advises anyone on a GLP-1 who experiences sudden painless vision loss in one eye, visual field deficits or other visual disturbances to contact a health care professional immediately. Clinicians are told to refer suspected NAION cases to an ophthalmologist promptly and to consider discontinuing the medication unless its benefits clearly outweigh the risk, a hedged instruction rather than a stop order. Patients who have already had a NAION event in one eye face elevated risk in the other and are asked to discuss whether to continue.

The deeper takeaway is how FDA weighed incomplete evidence. Two years of meta-analysis, claims data and case review produced a signal without a mechanism, a relative increase without a clean denominator, and a confidence interval brushing 1.0. The agency concluded the potential risk was real enough to name on every label in the class, and small enough to name quietly. Prescribers now hold the disclosure; the evidence, on both sides, remains open.