Skip to main content
Call Us Provider Portal

Does Ozempic Cause Vision Loss? Why Association Is Not the Same as Causation

Wagner Kapoor Institute Blog Does Ozempic Cause Vision Loss? Why Association Is Not the Same as Causation

The Most Honest Answer Is More Nuanced Than “Yes” or “No”

Does Ozempic cause vision loss?

Current research does not support a simple answer. Several observational studies have found that patients starting semaglutide experienced NAION more often than comparison patients. International regulators now list NAION as a very rare adverse effect of semaglutide. However, observational research cannot prove that the medication directly caused every event, and randomized-trial analyses have not consistently shown a statistically significant increase.

As we explain in our full guide to GLP-1 medications and NAION, the responsible conclusion is that a credible safety signal exists—especially for semaglutide—while the size, mechanism and class-wide extent of the risk remain under study.

That distinction matters for patients in Virginia Beach and across Hampton Roads who need useful medical context, not a frightening headline.

What Does “Association” Mean?

An association means two things occurred together more often than expected in a study. It does not automatically show that one caused the other.

GLP-1 medications are commonly prescribed to people with type 2 diabetes or obesity. Those patients may also have high blood pressure, high cholesterol, sleep apnea, cardiovascular disease or other factors associated with NAION.

Researchers try to balance comparison groups for those differences. But they cannot measure or remove every possible influence, including subtle optic nerve anatomy, medication adherence, care received outside a health system or differences in how diagnoses are coded.

What Do the Newer Observational Studies Show?

A 2026 nationwide study of more than 102,000 U.S. veterans with type 2 diabetes compared people who began semaglutide with those who began an SGLT2 inhibitor. After statistical weighting, cumulative NAION risk was 0.29% in the semaglutide group and 0.13% in the SGLT2 inhibitor group over a median follow-up of 2.1 years.

That finding supports an association and provides a different estimate from the first widely discussed 2024 study, which came from a specialty neuro-ophthalmology population and reported a much larger relative difference.

Differences between results do not necessarily mean one study was “right” and another was “wrong.” Study populations, comparison drugs, outcome definitions, follow-up and methods all affect estimates.

Why Have Randomized Trials Not Settled the Question?

Randomized trials are designed to reduce many forms of bias, but most GLP-1 trials were not designed specifically to study NAION. The condition is rare, diagnostic coding may not be uniform and the number of recorded events is very small.

A 2026 systematic review noted that meta-analyses of randomized GLP-1 trials had not shown a statistically significant increase in NAION, while emphasizing that rare event counts were extremely low and high-risk patients may be underrepresented.

This leaves physicians weighing two imperfect bodies of evidence:

That is what uncertainty in medicine looks like. It is not an excuse to ignore the signal, and it is not proof that every case was caused by medication.

What Do Regulators Say?

The European Medicines Agency and the UK Medicines and Healthcare products Regulatory Agency describe NAION as a very rare semaglutide adverse effect, meaning it may affect up to 1 in 10,000 people taking semaglutide. Australia updated warnings across GLP-1 products in July 2026 after reviewing conflicting evidence and postmarketing reports.

The U.S. FDA has publicly identified NAION as a potential serious-risk signal for the GLP-1 class and states that it is evaluating the need for regulatory action. A safety signal means the issue warrants evaluation; it is not, by itself, a final conclusion about causation or equal risk across every product.

What Should Patients Do With This Information?

Patients do not need to choose between two extremes:

Neither statement reflects the evidence.

A more accurate takeaway is:

NAION remains uncommon, but the association with semaglutide is credible enough to discuss, especially for patients with previous NAION or other important risk factors.

Do not stop prescribed treatment on your own. Discuss the benefit-risk balance with your prescriber and share any history of NAION or optic nerve disease. If vision suddenly becomes dim, blurred, clouded or partially missing, seek urgent medical evaluation.

See NAION Symptoms: When Sudden Vision Loss Requires Urgent Care for a clear action guide.

At Wagner Kapoor Institute, we provide specialty ophthalmic care throughout Hampton Roads. For appointments or guidance, call 757-481-4400.

Related reading

Sources
Author
Kapil G. Kapoor, MD, FACS, FICS, FASRS Kapil G. Kapoor, MD, FACS, FICS, FASRS Dr. Kapil G. Kapoor, is a board certified ophthalmologist specializing in vitreoretinal surgery. He is Co-Founder and President of the Wagner Kapoor Research Institute which has been actively engaged in over 75 clinical trials, often one of the only sites in the region offering sight-saving treatment. He has been passionately involved in ophthalmology research as part of his mission to deliver the best care possible to his patients. With over a decade’s experience in the field, Dr. Kapoor is a progressive leader in the healthcare space.

You Might Also Enjoy...

5 Easy Habits for Lifelong Healthy Vision
May is Healthy Vision Month, and it’s a reminder that taking care of your eyes doesn’t have to be complicated. In fact, some of the most effective ways to protect your vision come down to small, everyday choices.

Seeing Beyond the Diagnosis: Living Better with Low Vision
Low vision is when vision loss interferes with daily activities like reading, driving, recognizing faces, or using a phone. For some, it’s a blind spot in the center of their vision. For others, it’s missing pieces in their peripheral field. No two people