A large-scale retrospective cohort study has found an association between semaglutide use and an increased long-term risk of nonarteritic anterior ischemic optic neuropathy in patients with diabetes.

According to the recent JAMA Ophthalmology article, the investigators—led by Alan Y. Hsu, MD, of the Department of Ophthalmology at the China Medical University Hospital in Taiwan—analyzed data from more than 3 million patients with diabetes using the TriNetX global health research network from October 1, 2019, through December 31, 2023. They matched 174,584 patients with diabetes who were prescribed semaglutide with an equal number of patients prescribed non–glucagon-like peptide-1 receptor agonist (GLP-1 RA) antidiabetic medications. The cohorts were balanced across demographics, comorbidities, and baseline clinical characteristics using propensity score matching. Mean age for both cohorts of patients was approximately 58.3 years.

No increased risk of nonarteritic anterior ischemic optic neuropathy (NAION) was observed in the semaglutide cohort within the first year:

• 1-month hazard ratio (HR) = 2.99 (95% confidence interval [CI] = 0.31–28.75)
• 3-month HR = 1.33 (95% CI = 0.30–5.93)
• 6-month HR = 1.79 (95% CI = 0.60–5.35)
• 1-year HR = 1.94 (95% CI = 0.93–4.02).

However, from year 2 onward, semaglutide users showed a significantly increased risk:

• 2-year HR = 2.39 (95% CI = 1.37–4.18)
• 3-year HR = 2.44 (95% CI = 1.44–4.12)
• 4-year HR = 2.05 (95% CI = 1.26–3.34)
• Overall HR for full study duration = 2.22 (95% CI = 1.37–3.60).

The investigators also found variations in risk across subgroups:

• Age: Patients aged 40 to 64 years had higher risk (HR = 3.31, 95% CI = 1.68–6.53). Patients aged 65 years or older had no significant risk.
• Sex: Female patients had an increased risk (HR = 3.24, 95% CI = 1.68–6.25) compared with male patients. This finding contrasted with previous research that found “a sizable relationship between male sex and NAION,” the study authors wrote. They suggested that this difference could reflect previous work that found conditions such as retinopathy were more commonly observed in female patients; however, they noted that the effect of sex-related factors on actual risk of NAION in semaglutide users with diabetes remained unclear.
• Race: White patients had an increased risk (HR = 2.34, 95% CI = 1.40–3.90) compared with Black or Asian patients who had no significant risk.
• Comorbidities: Patients with hypertension had an increased risk (HR = 2.42, 95% CI = 1.19–4.92), but no increased risk was associated with obesity, dyslipidemia, sleep apnea, ischemic heart disease, or hypothyroidism.

Standalone users of Ozempic had the highest observed risk (HR = 6.27, 95% CI = 2.92–13.47). The investigators did not find a significant increased risk with Rybelsus or Wegovy individually. Patients with exclusive semaglutide use (excluding prior other GLP-1 RA use) had an elevated risk (HR = 2.36, 95% CI = 1.45–3.86).

The investigators noted limitations caused by the retrospective design and reliance on diagnostic codes, which may have introduced misclassification or selection biases. Specifically, they noted that diagnostic coding for NAION is not specific and may have included patients with other types of ischemic optic neuropathy. They were also unable to establish causality. 

To address limitations from previous research, the investigators “stratified [their] baseline group based on a broader range of covariates, including cataract surgery, steroid use, HbA1c level, and [body mass index], which helps mitigate the influence of these factors on [the] results.”

Still, they concluded, the remaining limitations of their study and other existing literature “warrant future, carefully designed observational studies to explore this association further.”

No conflicts of interest were reported.