Semaglutide and Your Eyes: What the NAION Research Actually Shows

NAION stands for non-arteritic anterior ischemic optic neuropathy — a mouthful that describes a simple, frightening event: the small blood vessels feeding the front of your optic nerve stop delivering enough oxygen, and part of your vision goes dark. It usually happens in one eye, it usually happens overnight (people often notice it on waking), and it is painless. That painlessness is exactly what makes it dangerous, because there is nothing to warn you it is coming.

NAION is often called "a stroke of the optic nerve," and the comparison is fair. The nerve tissue that loses its blood supply does not grow back. Most people are left with a permanent blind spot, often in the lower half of their vision, though the degree of loss varies widely. It is the most common cause of sudden optic nerve vision loss in adults over 50.

NAION landed in headlines in July 2024, when researchers at Massachusetts Eye and Ear published a retrospective study in JAMA Ophthalmology reporting that patients prescribed semaglutide appeared to develop NAION at several times the rate of patients on other medications. Semaglutide is the active ingredient in Ozempic, Wegovy, and Rybelsus. Overnight, millions of people taking a weekly injection for weight or blood sugar had a new question for their doctor.

Here is the crucial framing, and it is worth reading twice: NAION is rare to begin with. In the general population over 50, roughly 2 to 10 people per 100,000 per year develop it. Multiplying a very small number by four still leaves a very small number. The research matters, and the symptoms are worth knowing cold — but the risk of NAION should be weighed alongside everything else the medication is doing for your metabolic health, not in isolation.

The 2024 study (Hathaway et al., *JAMA Ophthalmology*, 2024) reviewed the records of 16,827 patients seen at Massachusetts Eye and Ear over six years and used a matched design to compare people prescribed semaglutide with people prescribed other medications for the same conditions.

The researchers ran two separate analyses. In the type 2 diabetes group, semaglutide was associated with a hazard ratio of 4.28 — meaning NAION occurred roughly four times more often in the semaglutide group over the study period. In the overweight and obesity group, the hazard ratio was 7.64, closer to a sevenfold difference.

Those numbers sound alarming, so it is worth understanding what a study like this can and cannot prove. This was a retrospective cohort study at a single specialty eye hospital — a place where people with serious eye problems are specifically referred. That creates what researchers call referral bias: the population is not representative of everyone taking semaglutide. The absolute number of NAION cases was also small, which widens the uncertainty around any ratio calculated from them.

The study's own authors were careful about this. They described the finding as an association that should prompt further investigation, not proof that semaglutide causes NAION. That is the honest scientific position, and it is the one to hold onto.

There is also a plausible biological question underneath all of this. GLP-1 receptors exist on the optic nerve, and rapid changes in blood sugar are already known to affect the eye — the SUSTAIN-6 trial found a small increase in diabetic retinopathy complications with semaglutide, likely tied to how fast glucose dropped rather than the drug itself. Whether something similar applies to optic nerve blood flow is still an open question.

Mostly, no — and that is the part of the story that got far less coverage than the original headline.

When a small study raises an alarm, the next step is to look at much bigger populations where the signal, if real, should still be visible. Several teams did exactly that through 2025, using national prescription registries and multi-million-patient health record networks rather than a single hospital.

A Scandinavian cohort study drawing on Danish and Norwegian national registries — countries where essentially every prescription and every hospital diagnosis is recorded — found a smaller increase in NAION risk than the Harvard study, roughly a doubling rather than a four- to sevenfold jump. Because these registries capture whole populations rather than eye-hospital referrals, this estimate is generally considered more reliable.

Other large database analyses using US health record networks found no statistically significant increase at all once patients were carefully matched for the factors that independently drive NAION risk: diabetes, high blood pressure, sleep apnea, high cholesterol, and a particular optic nerve anatomy called a "crowded disc" (a small optic nerve head with little spare room, present in most people who develop NAION).

That last point deserves emphasis. Nearly every risk factor for NAION is also a reason someone would be prescribed semaglutide in the first place. Untangling "the drug did this" from "the people who take this drug were already at higher risk" is genuinely hard, and it is why the scientific picture remains unsettled rather than resolved.

If you are weighing this against the documented benefits, our summary of [what the SELECT trial found on heart health](/blog/glp1-heart-health-what-the-select-trial-found) puts the cardiovascular side of the ledger in numbers.

Regulators reached a measured conclusion: the risk is real enough to put on the label, and small enough that it does not change who should take the medication.

In June 2025, the European Medicines Agency's safety committee (PRAC) completed its review and concluded that NAION should be listed as a "very rare" side effect of semaglutide-containing medicines. In European regulatory language, "very rare" has a precise meaning: it affects fewer than 1 in 10,000 people treated. The committee recommended updating the product information for Ozempic, Wegovy, and Rybelsus accordingly, and advised that treatment be stopped if NAION occurs.

Notably, the EMA did not restrict who may be prescribed semaglutide, did not require new eye screening before starting, and did not extend the finding to other GLP-1 medications such as tirzepatide (Mounjaro, Zepbound), for which no comparable signal has been identified.

That combination — acknowledge it, label it, change nothing else — is how drug regulators typically handle a rare event with an uncertain but plausible link. It is a signal worth documenting, not a reason to withdraw a medication that reduces major cardiovascular events.

If you take semaglutide and want to be proactive, the reasonable step is a baseline eye exam, particularly if you have not had one recently and you have diabetes, high blood pressure, or sleep apnea. That exam can tell you whether you have a crowded optic disc, which is the single strongest anatomical predictor of NAION.

The signs of NAION are specific, and knowing them is more useful than any risk statistic — because NAION is a same-day emergency regardless of what medication you take.

Watch for all of these:

• •Sudden vision loss in one eye, typically noticed on waking. It can be a complete loss, a dark curtain over part of the field, or a dimming as though someone lowered the lights on one side.
• •Loss of the lower or upper half of your visual field in one eye. This pattern — a horizontal cut, called an altitudinal defect — is characteristic of NAION.
• •Painless onset. There is no ache, no grittiness, no headache. Eye pain points toward a different diagnosis.
• •Colors looking washed out or dull in the affected eye. Red in particular can look faded or brownish.
• •A dark or blurred central spot that does not clear when you blink or rest.

An easy self-check: cover one eye at a time and look at something with straight lines and clear color, like a doorframe or a book cover. Comparing eyes separately catches losses that both-eyes-open vision quietly compensates for.

What to do: treat sudden painless vision loss in one eye as an emergency and get to an ophthalmologist or emergency department the same day. There is no proven treatment that reverses NAION, but several conditions that look identical in the first hours — including retinal artery occlusion and giant cell arteritis — are treatable and time-critical. Do not wait to see if it clears overnight, and do not stop your medication and hope. Get examined first.

If what you are experiencing is dizziness or a brief head-rush rather than vision loss, that is a different and far more common issue on GLP-1s — our guide to [dizziness and lightheadedness on GLP-1s](/blog/glp1-dizziness-lightheadedness-causes-and-fixes) covers the usual causes.

For the overwhelming majority of people, no — and that decision belongs to you and your prescriber together, not to a headline.

Here is the arithmetic that regulators used. If NAION affects fewer than 1 in 10,000 people treated, then out of 10,000 women on semaglutide, fewer than one will experience it. Over the same period, the SELECT trial (*NEJM*, 2023) found that semaglutide reduced major cardiovascular events — heart attack, stroke, cardiovascular death — by 20% in people with overweight or obesity and existing heart disease. For most people, those two numbers are not close.

That said, a few situations warrant a real conversation with your doctor:

• •You have already had NAION in one eye. This is the clearest case for caution. Once NAION has occurred in one eye, the risk to the other eye over the following years is meaningfully elevated, and most specialists would avoid adding any uncertain risk.
• •You have a known crowded optic disc, sleep apnea that is untreated, or unusually low overnight blood pressure. Each of these independently raises NAION risk, and all three are worth addressing on their own merits.
• •You are also managing menopause. Estrogen decline affects the eye's surface and tear film, and dry-eye symptoms can muddy the picture of what is happening with your vision — our guide to [menopause and dry eyes](/blog/menopause-dry-eyes-why-it-happens-and-what-helps) explains how to tell those symptoms apart.

What is *not* a good reason to stop: general anxiety after reading a news story. Stopping a GLP-1 abruptly has its own consequences, including rapid weight regain and the return of the metabolic risks the medication was managing — we cover that in detail in [stopping GLP-1s and avoiding weight regain](/blog/stopping-glp1-in-menopause-how-to-avoid-weight-regain).

The balanced approach: know the symptoms, get a baseline eye exam if you are due for one, tell your ophthalmologist you take a GLP-1, and act immediately if your vision changes. Understanding [how GLP-1 medications work](/blog/how-glp1-medications-work-mechanism-explained) makes these trade-offs easier to weigh with your doctor.