Two large studies from the United States and Sweden indicated that patients with type 2 diabetes who start treatment with glucagon-like peptide-1 receptor agonists, known as GLP-1 drugs, may face a slight increase in risk of a rare optic nerve disorder that can cause sudden vision loss.

But the researchers said the absolute risk remained very low, with fewer than one in a thousand patients affected in the US study, while the one-year probability in the Swedish study was 0.04% among GLP-1 users compared to 0.02% among users of other drugs.

They added that the findings do not prove the drugs directly cause the optic nerve damage, because differences in diabetes severity and patients' underlying health may explain part of the observed increase.

Impact of obesity and diabetes drugs

The two studies were published in the Annals of Internal Medicine at a time when use of GLP-1 drugs for diabetes and obesity is growing, amid increasing interest in identifying rare side effects that may not be apparent in initial clinical trials.

This class includes drugs that help lower blood sugar by stimulating insulin secretion, slowing gastric emptying, and reducing appetite, and includes compounds such as semaglutide, dulaglutide, and liraglutide, among others.

In the first study, researchers from Rutgers University analyzed US health insurance claims data for patients aged 18 to 65 with type 2 diabetes who started using a diabetes treatment between 2017 and 2022.

Ischemic optic neuropathy

The researchers compared rates of ischemic optic neuropathy among patients who started treatment with GLP-1 receptor agonists and those who started using sodium-glucose cotransporter-2 (SGLT2) inhibitors or dipeptidyl peptidase-4 (DPP-4) inhibitors.

Ischemic optic neuropathy occurs when blood flow to the nerve that carries visual signals from the eye to the brain is reduced, potentially causing sudden, painless vision loss, often affecting one eye.

During an 18-month follow-up, the researchers recorded fewer than one case per thousand patients in all treatment groups. However, rates were modestly higher among those who started GLP-1 drugs compared to users of the other two classes.

The researchers used a statistical approach known as "target trial emulation," which aims to analyze real-world data in a way that mimics some features of randomized clinical trials, by pre-specifying treatment start date, patient selection criteria, follow-up period, and health outcome.

But that approach does not turn the study into an actual randomized trial, and cannot exclude all differences between patients for whom doctors prescribe each drug class.

For example, people receiving GLP-1 drugs may be more likely to be obese, or have more uncontrolled diabetes or cardiovascular disease, factors that may themselves be associated with reduced blood flow to the optic nerve.

The researchers noted that the observed association may reflect a possible drug effect, but may also reflect what is known as "residual confounding," meaning the continued influence of health factors that statistical analyses could not fully measure or adjust for.

4 cases per 10,000

The Swedish study reached a similar conclusion after analyzing national health data for people aged 35 to 84 with type 2 diabetes who started treatment between 2013 and 2024.

The study included 107,518 patients who started GLP-1 receptor agonists and 185,898 patients who started SGLT2 inhibitors.

After one year, the risk of non-arteritic anterior ischemic optic neuropathy was 0.04% in the GLP-1 group versus 0.02% in the SGLT2 group.

These percentages roughly mean four cases per 10,000 users of GLP-1 drugs versus two cases per 10,000 users of SGLT2 inhibitors within a year.

Thus the relative risk appeared about twice as high, but the absolute difference was only about two additional cases per 10,000 patients.

Risk factors

Researchers from the Karolinska Institute said the size of the difference decreased in some analyses that attempted to more accurately account for diabetes severity, reinforcing the possibility that patients' health status is responsible for part of the increase.

The condition studied by the Swedish group is known as "non-arteritic anterior ischemic optic neuropathy" and occurs due to a sudden lack of blood supply to the front part of the optic nerve, without arterial inflammation.

The disease is associated with risk factors such as diabetes, high blood pressure, lipid disorders, sleep apnea, and certain anatomical variations in the optic nerve head, making it difficult to separate the effect of the drug from the effect of comorbidities in observational studies.

Relative increase

The two studies highlight the difference between "relative risk" and "absolute risk." Doubling the risk may seem significant, but when the baseline incidence is extremely rare, the number of additional cases remains limited.

For the Swedish study, the risk increased from about two to four cases per 10,000 people over a year, not from common to high levels.

Also, the studies did not compare GLP-1 drugs with no treatment, but with other diabetes drugs. The comparator classes may have different effects on cardiovascular health, blood pressure, and kidney function, which could affect the results.

The available data also do not allow determining whether the potential risk applies equally to all GLP-1 drugs or is concentrated in a specific drug, or is related to dose, rate of weight loss, or rapid blood sugar reduction.

These findings follow previous observational studies that linked semaglutide use to a potential increased risk of non-arteritic anterior ischemic optic neuropathy, while other studies found no clear increase, leaving the question of causality under debate.

The American Academy of Ophthalmology said the available evidence indicates that the overall risk remains low, and some studies found an association while others did not, which requires balancing the potential risk against the benefits of treatment for each patient.