Does Semaglutide Increase Risk of a Rare Eye Condition? Meta-Analysis of Nearly 5 Million Patients

This systematic review and meta-analysis followed PRISMA guidelines and was prospectively registered on PROSPERO. Researchers searched seven databases plus Google Scholar through August 2025 for studies comparing semaglutide users to non-GLP-1 RA users with NAION as an outcome. Six observational studies were included. Random-effects models pooled odds ratios and adjusted hazard ratios, with subgroup analyses by treatment indication. Evidence certainty was graded using the GRADE framework.

Semaglutide is one of the most widely prescribed drugs globally, and even rare side effects affect large numbers of people. NAION can cause permanent vision loss. Clarifying whether there is a genuine risk is critical for the millions of patients using semaglutide for diabetes or weight management, and for the clinicians counseling them about potential side effects.

This meta-analysis addresses one of the most publicized safety concerns about GLP-1 receptor agonists. The finding that any signal may be driven by underlying diabetes rather than the drug itself is important for the broader GLP-1 safety narrative. As these drugs are prescribed to an ever-widening population, distinguishing drug effects from disease effects becomes essential for accurate risk communication.

All included studies were observational, preventing causal conclusions. There was substantial heterogeneity between studies (I² up to 99%). Evidence certainty ranged from very low to moderate by GRADE assessment. Residual confounding is likely since NAION risk factors overlap with conditions treated by semaglutide. The difference between diabetes and obesity indications may reflect underlying disease risk rather than a true drug-indication interaction. The results were fragile — sensitive to removal of a single study.