Large Matched-Cohort Study Finds Modest Retinopathy Risk With GLP-1 Drugs, No Significant Link to Rare Optic Nerve Disorder

A study published online August 27, 2026 in the journal Diabetes, Obesity and Metabolism reports that patients who started a GLP-1 receptor agonist had a modestly higher risk of diabetic retinopathy and retinopathy-related eye procedures than matched patients who started a DPP-4 inhibitor, but found no statistically significant increase in the rare optic nerve condition non-arteritic anterior ischemic optic neuropathy, known as NAION. The study was conducted by researchers at HCA Florida Healthcare and the University of Central Florida College of Medicine in Orlando, together with colleagues at Texas Tech University Health Sciences Center and the VA North Texas Healthcare System.

The researchers built a retrospective, propensity-score-matched cohort using electronic health records of new users of either drug class, comparing GLP-1 receptor agonists against DPP-4 inhibitors as an active comparator, a design intended to reduce bias from comparing treated patients against untreated ones. Patients who started either drug class between 2006 and 2021 were eligible. Two matched cohorts were built: an overall cohort of patients with diabetes, and a separate cohort restricted to patients with diabetes and chronic kidney disease, a group the authors describe as at higher likelihood of diabetes-related microvascular complications. Primary outcomes were non-vision-threatening diabetic retinopathy, vision-threatening diabetic retinopathy or diabetic macular edema, and NAION; secondary outcomes included vision impairment or blindness and the occurrence of retinopathy-related procedures.

In the overall cohort of 98,082 matched pairs, GLP-1 receptor agonist users had a higher rate of non-vision-threatening diabetic retinopathy than DPP-4 inhibitor users (11.05% versus 9.98%; odds ratio 1.12, 95% confidence interval 1.09 to 1.15), a higher rate of vision-threatening retinopathy or diabetic macular edema (2.17% versus 1.98%; odds ratio 1.09, 95% CI 1.03 to 1.16), and a higher rate of retinopathy-related procedures (4.16% versus 3.68%; odds ratio 1.14, 95% CI 1.09 to 1.19). There was no statistically significant difference in risk of NAION (odds ratio 1.11, 95% CI 0.96 to 1.29) or of vision impairment or blindness (odds ratio 1.04, 95% CI 0.97 to 1.11). In the smaller chronic kidney disease cohort of 15,440 matched pairs, GLP-1 receptor agonist use was again associated with higher risk of non-vision-threatening retinopathy (odds ratio 1.10, 95% CI 1.03 to 1.18) and retinopathy procedures (odds ratio 1.14, 95% CI 1.03 to 1.26), but not with the other outcomes measured.

The study authors concluded that GLP-1 receptor agonist use was associated with a modest increase in risk of diabetic retinopathy and retinopathy-related procedures, independent of whether a patient already had retinopathy at baseline. They reported no association with vision impairment or blindness, while cautioning that administrative diagnosis codes have low validity for identifying blindness specifically. The paper is observational and the authors state it cannot establish causality.

Diabetic retinopathy is a labeled risk for at least one GLP-1 receptor agonist already. The FDA-approved prescribing information for Ozempic (semaglutide), citing data from the SUSTAIN-6 cardiovascular outcomes trial, states that over two years, diabetic retinopathy complications occurred in 3.0% of patients treated with semaglutide compared with 1.8% on placebo, with a larger absolute difference among patients who already had diabetic retinopathy at baseline (8.2% versus 5.2%) than among those without prior retinopathy (0.7% versus 0.4%). The label attributes this partly to a well-documented phenomenon in which rapid improvement in blood sugar control can temporarily worsen existing diabetic retinopathy, and states that the effect of long-term glycemic control with semaglutide on retinopathy complications has not been separately studied.

The new study's NAION finding contrasts with an earlier, smaller study that first raised the possibility of a link. A 2024 study in JAMA Ophthalmology, led by researchers at Massachusetts Eye and Ear and Harvard Medical School, used a propensity-matched design at a single academic neuro-ophthalmology practice to compare patients prescribed semaglutide against patients prescribed non-GLP-1 medications for diabetes or weight management. Among 710 patients with type 2 diabetes, semaglutide was associated with a higher 36-month cumulative incidence of NAION (8.9% versus 1.8%; hazard ratio 4.28, 95% CI 1.62 to 11.29). Among 979 patients who were overweight or obese, the association was stronger still (6.7% versus 0.8%; hazard ratio 7.64, 95% CI 2.21 to 26.36). That study's authors said their findings suggested an association but that, as an observational study, further research was needed to assess causality. The new, larger study did not find a statistically significant NAION signal in its GLP-1 receptor agonist class-wide comparison against DPP-4 inhibitors.

GLP-1 receptor agonists are approved by the FDA, the European Medicines Agency, and Australia's Therapeutic Goods Administration for type 2 diabetes and, for some agents including semaglutide and liraglutide, for chronic weight management. The new study does not change the approved indications, prescribing information, or regulatory status of any GLP-1 receptor agonist, and it did not report results broken down by individual drug within the class. This report describes research findings, not treatment or dosing guidance; patients with concerns about vision changes while taking a GLP-1 receptor agonist should consult a health professional.