Findings from a recent prospective, registry-based cohort study revealed the use of once-weekly semaglutide, a glucagon-like peptide-1 receptor agonist, more than doubles the risk of nonarteritic anterior ischemic optic neuropathy in patients with type 2 diabetes. 

Semaglutide has been shown to improve glycemic control and reduce cardiovascular outcomes in patients with type 2 diabetes (T2D) by mechanisms such as enhanced β–cell response, postponed gastric emptying, inhibition of glucagon secretion, and weight loss. Although the exact pathway linking semaglutide to nonarteritic anterior ischemic optic neuropathy (NAION) remains unclear, associations between the drug and ocular disease have been shown in evidence of diabetic retinopathy worsening in previous research and reports.

“After the introduction of once-weekly semaglutide in Denmark in November 2018,” Jakob Grauslund and Andreas Abou Taha, of the Department of Ophthalmology at Odense University Hospital, and the Department of Clinical Research at the University of Southern Denmark, along with colleagues observed, “the annual number of first-time NAION episodes reached an all-time high for the years 2019–2023. Likewise, at the same time, the rate of T2D in newly-diagnosed NAION increased from 1 in 20 to 1 in 4.”

The study, recently published in the International Journal of Retina and Vitreous, included 424,152 individuals with T2D in Denmark from 2018 to 2024, who were split into those exposed to once-weekly semaglutide (106,454 patients) and those who were not (317,698 patients). A total of 218 cases of NAION were identified during 1,915,120 person-years of observation. The incidence rate was significantly higher in semaglutide users (0.228 per 1,000 person-years) compared to non-users (0.093 per 1,000 person-years). Users exhibited a 2.19-fold increased risk of NAION after adjusting for confounding factors, including age, sex, duration of diabetes, hemoglobin A1c, cardiovascular disease, and medication use. The median time from first semaglutide prescription to NAION diagnosis was 22.2 months (interquartile range: 10.2–37.8 months). Results remained consistent after excluding patients with diabetic retinopathy and those using other GLP-1 RAs.

The researchers did not find a high-risk window between GLP-1 RA exposure and NAION onset, and further explained that, while “expression of GLP-1 RA receptors have been identified in the optic nerve, it is unknown whether continuous stimulation of these with specific GLP-1 RAs may alter vascular perfusion of the optic nerve head.”

Previous research that linked GLP-1 RAs to worsening diabetic retinopathy may be attributed to early worsening after rapid glycemic control improvement in the semaglutide arm of previous research. But here, the investigators described, the increased risk of NAION may indicate a different mode of action, given the statistically significant risk after adjustment for glycemic control. 

“The observed incidence rate of NAION of 0·228 per 1000 person-years for persons with T2D exposed with once-weekly semaglutide may not discourage semaglutide treatment but needs to be acknowledged as a potential risk,” they wrote.

The researchers suggested future studies should focus on identifying high-risk subgroups and exploring whether the risk of NAION is specific to semaglutide or whether it extends to the entire GLP-1 RA drug class.

A full list of author disclosures can be found in the published research.

This article provides further data on 1 of the studies referenced here.