Eye doctors' group says tirzepatide's NAION risk needs more study
The eye doctors' groups say a semaglutide-NAION link is possible but unproven and rare in absolute terms, and that tirzepatide's risk has not been studied enough.
The American Academy of Ophthalmology and the North American Neuro-Ophthalmology Society issued a joint consensus statement in May 2026 concluding that the available evidence suggests a possible link between semaglutide and a rare form of vision loss called non-arteritic anterior ischemic optic neuropathy (NAION), but that the association is not established and the overall risk remains low [1]. The statement also says the link has been studied less often for other GLP-1 drugs and calls for more research, particularly on tirzepatide [1].
What the evidence shows
NAION is rare, with an annual incidence of 2.3 to 10.2 per 100,000 people, and there is no known effective treatment [1]. Most patients have sudden, painless vision loss; about one-third may improve somewhat on their own, but the loss is often permanent, and up to one-quarter of patients develop it in the other eye [1]. Semaglutide is sold as Ozempic, Wegovy and Rybelsus; tirzepatide is sold as Mounjaro and Zepbound.
The statement reviewed retrospective studies and meta-analyzes and found results pointing in different directions. A 2024 study of nearly 17,000 patients at a single referral clinic reported much higher NAION risk among patients taking semaglutide than among matched controls: a hazard ratio of 4.28 (95% CI, 1.62-11.29) at 36 months in patients with type 2 diabetes and 7.64 (95% CI, 2.21-26.36) in patients with overweight or obesity [1]. Cumulative incidence was 8.9% versus 1.8% in the diabetes group and 6.7% versus 0.8% in the weight group [1]. The statement notes this study came from one center that treats most of its region's NAION patients, may have overestimated risk, and did not compare diabetes severity between the two groups [1].
Larger database studies reported smaller increases. Among 37.1 million adults with type 2 diabetes in the OHDSI network, semaglutide was linked to higher risk than empagliflozin (HR 2.27; 95% CI, 1.16-4.46) but not sitagliptin or glipizide, and about half the databases in that network showed no increase [1]. TriNetX studies found hazard ratios of 2.39 at two years, 2.44 at three years and 2.05 at four years in more than 170,000 patients with diabetes taking semaglutide [1]. One TriNetX study that grouped semaglutide and tirzepatide together found a NAION rate of 0.04% versus 0.02% in controls over two years (HR 1.76; 95% CI, 1.01-3.07) [1]. A separate analysis of more than 185,000 people taking GLP-1 drugs including tirzepatide found a small increase that was not statistically significant (HR 1.26; 95% CI, 0.94-1.70) [1].
Other TriNetX work found no increased risk at any time point, with cumulative incidence of 0.039% at one year and 0.065% at five years in patients with type 2 diabetes taking semaglutide, and 0.036% and 0.038% at one and two years in patients with overweight or obesity [1]. Meta-analyzes of randomized trials also disagreed: one of 78 semaglutide trials found an odds ratio of 3.92 (95% CI, 1.02-15.02), one of 69 GLP-1 trials found 1.53 (95% CI, 0.53-4.44), and one of 8 semaglutide trials found no statistically significant increase for diabetes (relative risk 1.76; 95% CI, 0.43-7.25) or weight loss (relative risk 2.18; 95% CI, 0.33-14.34) [1].
Why it matters for patients
Numbers like these are hard to turn into personal risk. Because NAION is rare to begin with, even a two- or three-fold relative risk means a small absolute risk, and the statement says the overall magnitude of risk remains low [1]. Many people who take these drugs already have conditions that raise their NAION risk on their own, including diabetes, cardiovascular disease, high blood pressure, metabolic syndrome and obstructive sleep apnea [1].
The statement does not tell patients to stop or avoid GLP-1 drugs. It calls for shared decision-making between each patient and their care team about whether to start, continue or discontinue them [1]. It also notes how widely these drugs are now used: more than one-quarter of patients with diabetes received a GLP-1 drug in 2024 [1].
What happens next
The statement says new evidence is emerging rapidly and that the existing studies are mainly retrospective, which limits what can be concluded [1]. It says the association with other GLP-1 drugs has been less frequently studied and warrants further investigation, particularly for tirzepatide [1]. No specific timeline for new research or regulatory action has been reported.
Sources
Semaglutides.org provides information only, not medical advice. Talk to a licensed provider about your own care before making changes. Some telehealth links are affiliate links and are labeled where they appear.