Semaglutide and the Risk of Nonarteritic Ischemic Optic Neuropathy: A Systematic Review and Certainty of Evidence Meta-Analysis.

Abstract

This systematic review and meta-analysis evaluates whether semaglutide use, compared with non-glucagon-like peptide-1 receptor agonist (GLP-1 RA) receptor agonist therapies, is associated with nonarteritic anterior ischemic optic neuropathy (NAION). Nonarteritic anterior ischemic optic neuropathy is a rare, vision-threatening optic neuropathy whose major risk factors overlap with populations prescribed semaglutide, making clarification of any associated risk clinically important for treatment decisions and patient counseling. This systematic review and meta-analysis followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines and was prospectively registered (PROSPERO CRD420251107346). Seven databases and Google Scholar were searched from inception to August 20, 2025. Eligible studies enrolled adults, compared semaglutide with non-GLP-1 RA users, and reported NAION as a distinct outcome with extractable odds ratios (ORs) or adjusted hazard ratios (aHRs). Two reviewers independently performed study selection, data extraction, and Newcastle-Ottawa risk of bias assessment. Random-effects models pooled ORs and aHRs. Prespecified subgroup analyses examined treatment indication. Certainty of evidence was assessed using Grading of Recommendations Assessment, Development and Evaluation. Six observational studies (n = 4 831 654) met inclusion. In pooled OR analyses, semaglutide was not associated with a statistically significant increase in NAION compared with non-GLP-1 RA comparators (OR = 2.44; 95% confidence interval [CI], 0.59-10.15; I² = 99.34%; very low certainty). Time-stratified ORs at 1 to 3 years were likewise nonsignificant (very low certainty). Compared with SGLT2 inhibitors, the pooled OR was 0.72 (95% CI, 0.38-1.35; I² = 81.59%; very low certainty). In adjusted time-to-event analyses, the pooled aHR was 1.63 (95% CI, 0.88-2.39; I² = 67.15%; low certainty). Exclusion of 1 influential study shifted the estimate (aHR = 1.92; 95% CI, 1.03-2.81; low certainty), indicating fragility. By indication, pooled aHRs were 1.70 (95% CI, 1.10-2.64; low certainty) for type 2 diabetes and 0.47 (95% CI, 0.19-1.13; moderate certainty) for obesity/overweight. Time-stratified aHRs were insufficient for meta-analysis. Across observational studies comparing semaglutide with non-GLP-1 RA therapies, pooled adjusted evidence does not demonstrate a consistent overall increase in NAION risk. However, estimates are limited by substantial heterogeneity, residual confounding, outcome misclassification, and fragility, resulting in low to very low certainty. Apparent indication-based differences should be interpreted cautiously and may reflect underlying disease-related risk rather than a drug-specific effect. Causality cannot be established, and further well-designed studies with standardized NAION ascertainment are needed. Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.