Harvard study links semaglutide to rare optic nerve condition, prompting reviews

Researchers at Harvard-affiliated Massachusetts Eye and Ear reported in July 2024 that people prescribed semaglutide — sold as Ozempic, Wegovy and Rybelsus — were diagnosed with a rare optic nerve condition more often than matched patients taking other drugs for diabetes or weight loss ^[1]. The condition, non-arteritic anterior ischemic optic neuropathy, or NAION, causes sudden, usually painless vision loss, and there is no known effective treatment for it ^[2].

The study, published in JAMA Ophthalmology, reported a hazard ratio of 4.28 (95% CI, 1.62–11.29) for NAION among people with type 2 diabetes prescribed semaglutide, and 7.64 (95% CI, 2.21–26.36) among people with overweight or obesity ^[2]. A hazard ratio compares how often an outcome occurs in one group versus another, so 4.28 means the semaglutide group was diagnosed roughly four times as often over the study period ^[3].

At 36 months, the cumulative incidence of NAION was 8.9% (95% CI, 4.5–13.1) in the diabetes group taking semaglutide versus 1.8% (95% CI, 0–3.5) in matched controls on other diabetes drugs. In the overweight or obesity group, cumulative incidence was 6.7% (95% CI, 3.6–9.7) with semaglutide and 0.8% (95% CI, 0–1.8) with other weight-loss medications ^[2]. The overweight or obesity arm included 979 patients ^[2]. Patients were matched on age, sex, hypertension, type 2 diabetes, obstructive sleep apnea, obesity, hyperlipidemia and coronary artery disease ^[1].

The criticism and the conflicting evidence

The study was conducted at a single tertiary referral center that treats most of the region's patients with NAION, and the condition appeared more often than would be expected in the general population — signs of possible selection bias and overestimation of risk, according to a later review by the North American Neuro-Ophthalmology Society and the American Academy of Ophthalmology ^[2]. Hemoglobin A1c was not compared between the groups, so diabetes severity could have influenced both treatment choice and NAION risk. The control group took different medications, so it is also possible that GLP-1 drugs were simply less effective at preventing NAION than the alternatives ^[2].

Later research has not settled the question. The largest retrospective study, covering 37.1 million adults with type 2 diabetes, found an increased risk of NAION for semaglutide compared with empagliflozin (HR, 2.27; 95% CI, 1.16–4.46) using a specific diagnosis definition, but no increased risk using a sensitive definition, and about half the databases in that network showed no increased risk ^[2]. Several TriNetX database studies found elevated risk, while others found none ^[2]. A meta-analysis of 78 randomized trials of semaglutide found an odds ratio of 3.92 (95% CI, 1.02–15.02) for ischemic optic neuropathy, while a meta-analysis of 69 GLP-1 trials found no association (OR, 1.53; 95% CI, 0.53–4.44), and a review of 8 semaglutide trials found no statistically significant increase ^[2]. In June 2025, the European Medicines Agency's safety committee concluded that NAION is a very rare side effect of semaglutide, affecting up to 1 in 10,000 people, and estimated roughly a two-fold increase in risk for adults with type 2 diabetes — about one additional case per 10,000 person-years of treatment ^[4]. That estimate is smaller than the four-fold and seven-fold figures in the 2024 study.

Why it matters for patients

NAION is uncommon, but when it occurs the vision loss is often permanent, and up to one-quarter of patients may develop it in the other eye ^[2]. The 2026 consensus statement from the two ophthalmology groups says the overall magnitude of the risk remains low, that the evidence comes mostly from retrospective studies with limitations, and that patients and their care teams should decide together whether to start, continue or stop a GLP-1 receptor agonist ^[2]. The European regulator advises that patients who notice sudden loss of vision or rapidly worsening eyesight during semaglutide treatment contact their doctor without delay, and that treatment should be stopped if NAION is confirmed ^[4]. Whether risk varies by dose, by injectable versus oral semaglutide, or by how long someone has taken the drug has not been reported. Lawsuits have also followed: more than 1,800 claims against Novo Nordisk had been filed as of mid-2025, with consolidation under way in New Jersey state courts ^[3].

What happens next

Regulatory timing differs by account. The European Medicines Agency's announcement that product information should list NAION as very rare is dated June 6, 2025 ^[4], while a law firm page states the agency required updated warning labels in August 2024 and that U.S. labels had not been updated as of October 2025 ^[3]. JAMA Ophthalmology has since published several letters debating the study's methods and newer research, including a February 2026 study of semaglutide initiators among U.S. veterans with type 2 diabetes ^[1].