The U.S. Food and Drug Administration directed application holders of all marketed GLP-1 RAs to add information about a potential increased risk of NAION to the Warnings and Precautions section of product labeling. The agency characterized NAION as a rare cause of acute, painless decreased vision in one eye that can include permanent vision loss. The FDA based its action on a bundled review of randomized clinical trials, observational research, postmarketing reports, and animal data, and the agency explicitly noted limitations in individual datasets.
The regulatory action requires companies to update labeling language; the source did not report specific wording that must be used, nor did it provide a deadline for when application holders must submit revised labels. The FDA concluded that, given available evidence across multiple data sources, informing clinicians and patients via labeling changes was warranted while additional data accumulate.
FDA investigators performed a meta-analysis across GLP‑1 RA development programs because individual trials contained few or no NAION cases. The agency included 126 randomized, controlled trials, encompassing 134,253 patients and 288,879 person‑years (PY) of exposure. Within that pooled dataset the reviewers identified 15 probable NAION events: 10 among patients treated with a GLP‑1 RA and 5 among comparator-treated patients. These counts translated to an incidence of 6.4 events per 100,000 PY for GLP‑1 RA users versus 3.8 events per 100,000 PY for comparators.
FDA also ran a retrospective cohort analysis using the Sentinel Real‑World Evidence Data Enterprise Commercial Network. That analysis examined adults with type 2 diabetes who initiated a GLP‑1 RA, an SGLT‑2 inhibitor, or a DPP‑4 inhibitor between 2019 and 2024. Incidence estimates in the Sentinel study ranged from 26 to 32 events per 100,000 PY (GLP‑1 RA: 30–32 events; SGLT‑2 inhibitor: 31 events; DPP‑4 inhibitor: 26 events). The agency reported no statistically significant increase in NAION risk for GLP‑1 RAs versus SGLT‑2 inhibitors, while a higher risk was observed when GLP‑1 RAs were compared to DPP‑4 inhibitors.
Postmarketing reports submitted to the FDA Adverse Event Monitoring System were reviewed as well. The agency described these spontaneous reports as challenging to interpret because recognized NAION risk factors—older age, smoking, hypertension, obstructive sleep apnea, anemia, kidney disease, and diabetes itself—are commonly present in patients prescribed GLP‑1 RAs. FDA investigators therefore weighed these reports alongside controlled trial data and observational studies rather than relying on them in isolation.
Published observational studies entered the agency’s assessment. The FDA noted that methodologically rigorous observational analyses generally reported an increased NAION risk in patients initiating GLP‑1 RAs compared with those starting DPP‑4 or SGLT‑2 inhibitors, results that were consistent with signals seen in the clinical trial meta-analysis and the Sentinel cohort in certain comparisons.
Animal data were examined to explore biologic plausibility. FDA review of available preclinical information did not identify a mechanism by which GLP-1 RAs would increase the risk of NAION, and the source reported that no clear mechanistic explanation emerged from those data.
Collectively, the agency concluded that NAION cases have been reported across all currently marketed GLP‑1 RA products and that a potential increased risk exists, even as uncertainties and confounding factors remain.
The FDA advised clinicians to counsel patients about the signs of NAION—sudden, painless decreased vision in one eye, new visual field defects, or other acute visual disturbances—and to instruct immediate reporting of those symptoms. If NAION is suspected, the agency recommended prompt ophthalmology referral for evaluation. The guidance further states clinicians should consider discontinuing the GLP‑1 RA in patients who develop NAION unless the benefits of continued treatment clearly outweigh the risk.
The agency flagged several limitations that clinicians should keep in mind. Risk factors for NAION overlap substantially with the comorbidities seen in adults treated with GLP‑1 RAs, which complicates attempts to infer causality from observational data or spontaneous reports. The FDA also noted limitations in the Hathaway et al. semaglutide study referenced in its review and emphasized the need to interpret that study in the broader evidence context. The source did not provide granular trial-level methodology, dosing regimens, or individual patient demographics beyond aggregate counts used in the meta-analysis.
Unanswered clinical questions remain. FDA review did not identify an animal‑model mechanism linking GLP-1 RA exposure to NAION, and the agency has not specified whether further prospective studies or active surveillance programs will be required beyond the labeling change. The agency similarly did not provide a timeline for additional regulatory steps or requirements for postmarketing studies in the public notice summarized here.
Clinicians balancing cardiometabolic and glycemic benefits of GLP-1 RAs against these potential ocular risks should document shared decision‑making, screen for ocular risk factors where appropriate, and refer rapidly for ophthalmologic assessment if acute visual symptoms appear. The FDA action formalizes a safety signal across multiple evidence streams, mandates labeling updates, and leaves open specific questions about mechanism, causality, and prospective risk quantification that will need further data to resolve.