New data clarify rare optic nerve risk with GLP-1 receptor agonists

Clinical takeaway: Continue prescribing GLP-1 receptor agonists when clinically indicated, as their proven cardiometabolic benefits are likely to outweigh this potential rare risk. Advise patients to seek urgent ophthalmologic evaluation for sudden, painless vision loss, but avoid discontinuing therapy solely because of these findings.

As GLP-1 receptor agonists continue to see widespread use for type 2 diabetes and obesity, clinicians need clearer information about a rare but potentially vision-threatening adverse event that has generated growing attention.

Two large observational studies published in Annals of Internal Medicine found a small increase in the relative risk of ischemic optic neuropathy among patients with type 2 diabetes initiating GLP-1 receptor agonists compared with other glucose-lowering medications, but both studies emphasized that the absolute risk was extremely low and that residual confounding may explain some or all of the observed association.

In the U.S. target trial emulation, investigators analyzed commercial insurance claims from more than 369,000 adults aged 18 to 65 years who initiated a GLP-1 receptor agonist, an SGLT2 inhibitor, or a DPP-4 inhibitor. Over 18 months, ischemic optic neuropathy occurred in fewer than 1 in 1,000 patients in every treatment group. The estimated risk was about 8 to 9 cases per 10,000 GLP-1 users versus 4 to 6 cases per 10,000 comparator patients—an excess of roughly 3 to 4 additional cases per 10,000 treated patients. Most events occurred in men and adults older than 50 years, with somewhat higher risks among those with cardiovascular disease or preexisting eye disease.

The nationwide Swedish target trial emulation reached similar conclusions. Among more than 293,000 patients, anterior ischemic optic neuropathy developed in 62 GLP-1 users and 64 SGLT2 inhibitor users. The 1-year absolute risk was 0.04% with GLP-1 therapy versus 0.02% with SGLT2 inhibitors, increasing to 0.12% versus 0.07% after 5 years. Notably, when investigators restricted analyses to patients receiving metformin at baseline (a strategy intended to better balance diabetes severity) the already small risk differences became even smaller and were no longer statistically distinguishable, suggesting underlying differences in patient health may account for much of the observed association.

Both research teams cautioned against inferring causality from these observational data. As the Swedish investigators concluded, "absolute risks were small," while the U.S. authors emphasized that "observed differences may reflect residual confounding." Together, the findings strengthen evidence for a possible safety signal but also reinforce that any excess risk, if present, appears to be rare and should be weighed against the well-established cardiovascular, renal, metabolic, and weight-loss benefits of GLP-1 receptor agonists.

Sources: Reynolds KR, et al. (2026 July 14) Ann Intern Med. Glucagon-Like Peptide-1 Receptor Agonists and Risk for Ischemic Optic Neuropathy: A Target Trial Emulation; Ueda P, et al. (2026 July 14) Ann Intern Med. Glucagon-like Peptide-1 Receptor Agonists and Risk for Anterior Ischemic Optic Neuropathy: A Nationwide Cohort Study