Ischemic Optic Neuropathy With Semaglutide: Global Observational Analysis of Sex- and Formulation-specific Risk
Moiz Lakhani1, Abdullah Al-Ani2, Marko Popovic3, Etienne Benard-Seguin2, Edward Margolin3
1University of Ottawa Faculty of Medicine, 2University of Calgary, Department of Surgery, Section of Ophthalmology, 3University of Toronto, Department of Ophthalmology and Visual Sciences
Objective:
We conducted the first global, population-based observational study of over 30 million reports to compare ischemic optic neuropathy (ION) risk with semaglutide formulations across diabetes and obesity.
Background:
Glucagon-like peptide-1 receptor agonists (GLP-1RAs) have transformed cardiometabolic care by improving glycemic control, weight loss, and cardiovascular risk. Recent reports link semaglutide to nonarteritic anterior ischemic optic neuropathy (NAION), with severe vision loss in up to 15% and complete blindness in up to 5%, prompting regulatory action by the European Medicines Agency. No prior study has evaluated formulation- or sex-specific patterns in semaglutide-associated NAION.
Design/Methods:
We retrieved and deduplicated FDA Adverse Event Reporting System (FAERS) reports from December 2017–December 2024. We included reports naming a GLP-1RA as the primary suspect for ION. Semaglutide was analyzed both as a pooled category and by brand. Comparator drugs were metformin, insulin, and tirzepatide. Disproportionality was assessed using reporting odds ratios (RORs, 95%CIs). Signals met adapted Evans criteria (ROR>2, χ²>4, n≥3), Bonferroni-adjusted significance (p<0.0056; p<0.0028 for sex analyses), and Bayesian confirmation (IC025>0.3). To complement disproportionality estimates, we conducted multivariable logistic regression adjusted for age and sex, generating adjusted odds ratios (AORs, 95%CIs).
Results:
Among 30,668,520 FAERS reports, 31,774 involved semaglutide (mean age=56.5±11.1 years; 54.1% female). Wegovy showed the strongest ION signal (n=28, ROR=74.89, 95%CI=51.79–108.29), exceeding Ozempic (n=47, ROR=18.81, 95%CI=14.11–25.07) and pooled semaglutide (n=85, ROR=21.37, 95%CI=17.40–26.65). Other comparators showed no signal: tirzepatide (ROR=0.56), metformin (ROR=1.35), and insulin (ROR=1.61). Sex-stratified analyses showed the highest signal for Wegovy in males (ROR=116.37) and for Ozempic in females (ROR=26.86). In multivariable analysis adjusted for age and sex, odds were higher for Wegovy vs. Ozempic (AOR=4.74, 95%CI=2.54–8.77, p<0.001) and for males vs. females (AOR=3.33, 95%CI=1.89–5.88, p<0.001).
Conclusions:
Overall, ION risk appears dose- and formulation-dependent, highest with Wegovy. Although Wegovy produced fewer ION reports than Ozempic, its lower overall reporting yielded a higher ROR.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.