Use of glucagon-like peptide-1 receptor agonists may be associated with a higher risk of anterior ischemic optic neuropathy compared with sodium-glucose cotransporter-2 inhibitors among patients with type 2 diabetes, although the absolute risks were low and attenuated in analyses designed to better account for diabetes severity.
In the nationwide, register-based cohort study, investigators emulated a target trial using linked Swedish health and administrative registers from 2013 to 2024. They compared 107,518 patients initiating glucagon-like peptide (GLP)-1 receptor agonists with 185,898 patients initiating sodium-glucose cotransporter (SGLT)-2 inhibitors. The investigators assessed anterior ischemic optic neuropathy identified through the national patient register. Adjusted risks were estimated at 1, 3, and 5 years using propensity score weighting to balance demographic characteristics, diabetes severity, comorbidities, drugs, health care use, and ophthalmic risk factors between treatment groups. The median follow-up was approximately 1.5 years.
The investigators found that anterior ischemic optic neuropathy occurred in 62 patients receiving GLP-1 receptor agonists and 64 patients receiving SGLT-2 inhibitors. At 1 year, the absolute risks were 0.04% and 0.02%, respectively. By 5 years, the corresponding risks were 0.12% and 0.07%. The primary analysis indicated a 78% higher likelihood of anterior ischemic optic neuropathy among those who received GLP-1 receptor agonists compared with those who received SGLT-2 inhibitors.
After performing analyses restricted to patients receiving metformin at baseline, the differences were substantially reduced. At 5 years, the absolute risks were 0.10% among those receiving GLP-1 receptor agonists and 0.09% among those receiving SGLT-2 inhibitors. Similar findings were observed among patients receiving metformin without insulin, in whom the 5-year risks were 0.08% and 0.07%, respectively.
Post hoc subgroup analyses by glycated hemoglobin level, smoking status, and individual GLP-1 receptor agonists produced wide confidence intervals because of the small number of outcome events. Sensitivity analyses using alternative exposure definitions, outcome definitions, and analytic approaches were generally consistent with the primary findings.
The investigators noted that despite adjustment for numerous clinical and demographic factors, residual confounding could not be excluded. Additional limitations included the small number of anterior ischemic optic neuropathy events, possible outcome misclassification, limited ability to evaluate individual GLP-1 receptor agonists, and uncertain generalizability to patients using the agents for obesity without diabetes.
The findings suggested that GLP-1 receptor agonist use may be associated with a higher relative risk of anterior ischemic optic neuropathy among patients with type 2 diabetes. However, the very low absolute event rates and the attenuation of the association in analyses accounting for diabetes severity support cautious interpretation.
“[A]bsolute risks for [anterior ischemic optic neuropathy] were small and the risk increase among users of GLP-1 [receptor agonists] was substantially attenuated in analyses restricted to patients with metformin use, indicating that it might be due to confounding,” wrote lead study author Peter Ueda, MD, PhD, of the Centre for Pharmacoepidemiology in the Division of Clinical Epidemiology in the Department of Medicine at the Karolinska Institutet, and colleagues.
The study was supported by the Karolinska Institutet, Swedish Society of Medicine, Swedish Research Council, and Region Stockholm. Full disclosures of the study authors can be found in the study.
Source: Annals of Internal Medicine
