An analysis of US insurance claims data published in the Annals of Internal Medicine suggests a potential association between the use of GLP-1 receptor agonists (RAs) and a higher incidence of nonarteritic anterior ischemic optic neuropathy (NAION) in adults with type 2 diabetes. While the study observed a higher frequency of optic nerve damage compared to patients taking other diabetes medications, researchers emphasize the absolute risk remains very low and the study’s observational design warrants caution.
Study Findings on GLP-1 RAs and Optic Neuropathy
Researchers led by Chintan V. Dave, PharmD, PhD, of Rutgers University, analyzed health insurance claims for thousands of adults with type 2 diabetes between 2017 and 2022. The study compared patients who initiated GLP-1 RA treatment against those prescribed SGLT2 inhibitors or DPP-4 inhibitors.
The data, published online July 13, revealed that GLP-1 RA users had an 18-month incidence rate of ischemic optic neuropathy (ION) of 8.5 cases per 10,000 patients. In comparison, the rate was 5.5 cases per 10,000 for SGLT2 inhibitor users and 7.8 cases per 10,000 for DPP-4 inhibitor users. This resulted in a calculated risk difference of three cases per 10,000 patients when compared to SGLT2 inhibitors and 3.6 cases per 10,000 when compared to DPP-4 inhibitors.
Demographic Trends in Reported ION Events
The study highlighted that the majority of ION events occurred in specific patient subgroups. Among the 81 total ION cases identified in the GLP-1 RA group:
- Age: 85.2% of events (69 cases) occurred in patients older than 50 years.
- Gender: 70.3% of events (57 cases) occurred in men, despite women comprising 55% of GLP-1 RA initiators overall.
Researchers noted that these demographic findings align with the known baseline risk factors for ION, which is more common in older adults and males.
Limitations and Clinical Interpretation
The study authors cautioned that the findings are based on observational data, which limits the ability to establish a direct causal link. Several factors could have influenced the results:
- Unmeasured Confounders: The researchers lacked access to data on body mass index (BMI), smoking history, and hemoglobin A1c levels, all of which are relevant to diabetes management and vascular health.
- Diagnostic Coding: The study relied on administrative diagnostic codes for ION as a proxy for NAION, rather than direct clinical assessment.
- Patient Selection: The severity and duration of a patient’s diabetes often dictate the choice of medication, potentially introducing selection bias that the propensity score matching could not fully eliminate.
"Despite the very low absolute risk for ION, the rapidly expanding use of GLP-1 RAs in patients with and without [type 2 diabetes] increases the clinical and public health importance of any potential association," the authors reported. They advise that any potential safety signal must be balanced against the well-established cardiometabolic benefits of GLP-1 receptor agonists.
The study was primarily funded by the National Institutes of Health. Patients currently taking these medications are advised to consult their healthcare providers regarding any concerns about vision changes or existing ocular health risks.
Related reading