Patients with diabetes who received semaglutide had a higher long-term risk of nonarteritic anterior ischemic optic neuropathy compared with matched patients who received non–glucagon-like peptide-1 receptor agonist antidiabetic medications, according to a retrospective cohort study published in JAMA Ophthalmology.
Researchers analyzed TriNetX US research network data from 2019 to 2023, including approximately 3.35 million patients with diabetes. After propensity score matching, the analysis included 174,584 patients who received semaglutide and 174,584 patients who received non–glucagon-like peptide-1 receptor agonist (GLP-1 RA) therapies, including insulin, metformin, sulfonylureas, sodium-glucose cotransporter 2 inhibitors, and other antidiabetic medications.
The study was designed in part to address limitations of a smaller 2024 JAMA Ophthalmology study by Hathaway and colleagues that reported an association between semaglutide use and nonarteritic anterior ischemic optic neuropathy (NAION) in 710 patients from a single institution.
The primary outcome was first diagnosis of NAION following treatment initiation. Semaglutide use was not associated with a statistically significant increase in NAION risk through 1 year of follow-up. However, elevated relative risk emerged at 2 years and persisted through 4 years.
Across the full study period, NAION occurred in 45 semaglutide-treated patients compared with 29 patients receiving non–GLP-1 RA therapies, corresponding to absolute event rates of approximately 0.026% and 0.017%, respectively. Semaglutide exposure was associated with a hazard ratio of 2.22 — approximately twice the hazard — compared with non–GLP-1 RA use across the full study period.
Subgroup analyses found elevated relative risk among patients aged 40 to 64 years, female patients, White patients, and patients with hypertension. The researchers did not adjust for multiple subgroup comparisons, however, and event counts were small in several analyses, making these findings hypothesis-generating rather than practice-changing.
Analyses by semaglutide formulation showed increased NAION risk among patients with Ozempic prescription history, including standalone Ozempic use. Similar statistically significant associations were not observed for oral semaglutide or semaglutide prescribed for weight management, although these subgroup analyses were limited by sparse events and wide confidence intervals.
The researchers noted several important limitations, including the retrospective observational design, inability to verify medication adherence, and potential residual confounding despite propensity matching. They also noted that the ICD-10 diagnostic code used for NAION is not fully specific and may capture other ischemic optic neuropathies, potentially introducing outcome misclassification. In addition, patients prescribed semaglutide may have had greater engagement with the health care system and more complete longitudinal follow-up, which could increase the likelihood of NAION detection compared with control patients. The study design did not allow investigators to establish causality.
"In this cohort study, among patients with diabetes, an elevated risk of NAION was associated with semaglutide use compared with non–GLP-1 RA use," wrote lead study researcher Alan Y. Hsu, MD, of China Medical University Hospital, and colleagues. The researchers added that "future, carefully designed observational studies" are needed to further evaluate the association.
The researchers reported no conflicts of interest.
Source: JAMA Ophthalmology