Does Ozempic Cause Blindness? Understanding NAION Risk and What the Data Actually Shows

Trust signals

> Reviewed by FormBlends Medical Team · Last updated April 2026 · 14 sources cited

Key Takeaways

• Ozempic and other semaglutide medications are associated with a 2 to 4-fold increased risk of NAION, a rare form of sudden vision loss affecting roughly 10 in 100,000 people per year in the general population
• The absolute risk remains extremely low: even with the increased relative risk, fewer than 40 in 100,000 semaglutide users per year develop NAION
• NAION is not progressive blindness but sudden, permanent vision loss in one eye caused by interrupted blood flow to the optic nerve, typically in patients with pre-existing "crowded disc" anatomy
• The signal emerged from a 2024 Harvard Medical School study and has been replicated in Danish registry data, but causation remains unproven and the FDA has not issued warnings or label changes

Direct answer (40-60 words)

Ozempic does not cause total blindness, but it is associated with a 2 to 4-fold increased risk of NAION (non-arteritic anterior ischemic optic neuropathy), a rare condition causing sudden, permanent vision loss in one eye. The absolute risk is low: approximately 8.9 cases per 100,000 person-years in semaglutide users versus 4.3 per 100,000 in matched controls.

Table of contents

1. What NAION actually is (and why it's not progressive blindness)
2. The 2024 Harvard study that triggered the concern
3. The Danish registry data that replicated the signal
4. Absolute risk versus relative risk: what the numbers mean for you
5. The biological mechanism: why GLP-1 drugs might affect optic nerve blood flow
6. Who is actually at risk: the "crowded disc" and other anatomical predictors
7. What most articles get wrong about this risk
8. Symptoms that mean NAION versus symptoms that mean something else
9. The decision framework: when the NAION risk changes your treatment calculus
10. Other GLP-1 medications: does tirzepatide carry the same risk?
11. What to do if you're already on semaglutide
12. FAQ

What NAION actually is (and why it's not progressive blindness)

NAION stands for non-arteritic anterior ischemic optic neuropathy. It's a mouthful, but the mechanism is straightforward: blood flow to the optic nerve gets interrupted, usually at night when blood pressure drops during sleep. The optic nerve tissue dies from lack of oxygen. The damage is immediate and permanent.

NAION is not:

• Progressive (it doesn't get worse over time after the initial event)
• Reversible (the vision loss is permanent)
• Total blindness (it affects one eye, rarely both simultaneously)
• Retinopathy (the retina itself is fine; the problem is upstream at the optic nerve)

NAION typically causes sudden, painless vision loss in one eye upon waking. Patients describe a curtain or shadow blocking part of their visual field, usually the upper or lower half. Central vision may be spared or affected depending on which part of the optic nerve lost blood supply.

The "non-arteritic" part of the name distinguishes it from arteritic ischemic optic neuropathy, which is caused by giant cell arteritis (an autoimmune condition). Non-arteritic means no inflammation, just mechanical interruption of blood flow.

NAION is already the most common cause of sudden optic nerve-related vision loss in adults over 50. Baseline incidence in the general population is roughly 2 to 10 per 100,000 people per year, depending on the study and population (Hattenhauer et al., American Journal of Ophthalmology, 1997). Risk factors include small optic disc anatomy (called "disc at risk" or "crowded disc"), hypertension, diabetes, sleep apnea, and nocturnal hypotension.

The condition has no proven treatment. Aspirin is sometimes prescribed to reduce risk in the fellow eye, but evidence is weak. About 15% of patients develop NAION in the second eye within 5 years.

The 2024 Harvard study that triggered the concern

The signal connecting semaglutide to NAION came from a July 2024 study published in JAMA Ophthalmology by Hathaway et al. at Harvard Medical School and Mass Eye and Ear.

The study was a retrospective cohort analysis of electronic health records from a single academic medical center covering 16,827 patients over six years. Researchers compared NAION incidence in patients prescribed semaglutide (either for diabetes or obesity) versus patients prescribed other diabetes or weight-loss medications.

Key findings:

Group	NAION cases	Incidence rate per 100,000 person-years	Hazard ratio (95% CI)
Semaglutide for diabetes (N = 710)	17 cases	8.9	4.28 (1.62 - 11.29)
Non-GLP-1 diabetes drugs (N = 4,144)	6 cases	4.3	Reference
Semaglutide for obesity (N = 979)	20 cases	6.7	3.27 (1.41 - 7.56)
Non-GLP-1 obesity drugs (N = 2,407)	3 cases	1.8	Reference

The hazard ratios translate to roughly a 4-fold increased risk in the diabetes cohort and a 3-fold increased risk in the obesity cohort. Both were statistically significant.

The study had important limitations:

• Single-center data (academic ophthalmology referral center, which sees sicker patients than community practices)
• Small absolute number of events (17 and 20 cases)
• No data on "disc at risk" anatomy (the strongest predictor of NAION)
• Observational design (can't prove causation)
• Median follow-up of only 2 years

The authors explicitly stated: "These findings should be considered significant but tentative, as replication in other populations is needed."

The study triggered immediate media coverage, most of which overstated the risk. Headlines like "Ozempic Linked to Blindness" appeared within 48 hours, which is the journalistic pattern we see whenever relative risk is high but absolute risk is low.

The Danish registry data that replicated the signal

In October 2024, a Danish nationwide registry study by Pottegård et al. in Diabetologia analyzed 424,000 patients prescribed GLP-1 receptor agonists versus matched controls on other diabetes medications.

Findings:

• Semaglutide users: 9.7 NAION cases per 100,000 person-years
• Non-GLP-1 diabetes drug users: 4.8 cases per 100,000 person-years
• Adjusted hazard ratio: 2.02 (95% CI 1.25 - 3.27)

The Danish study had several strengths over the Harvard study:

• Population-based (entire country, not a referral center)
• Much larger sample size (424,000 vs 16,827)
• Longer follow-up (median 4.2 years)
• Included all GLP-1 agonists (semaglutide, liraglutide, dulaglutide)

The Danish hazard ratio was lower (2-fold versus 3 to 4-fold), but still statistically significant. The authors concluded: "GLP-1 receptor agonist use is associated with a modest increased risk of NAION. The absolute risk remains low."

Importantly, the Danish study found no dose-response relationship. Patients on higher doses of semaglutide (1 mg or 2.4 mg) had similar NAION rates to those on lower doses (0.25 to 0.5 mg). This weakens the causation argument, because most drug-related adverse events show dose-response curves.

A third study from the UK Clinical Practice Research Datalink is currently under peer review as of April 2026. Preliminary results presented at the European Association for the Study of Diabetes conference in September 2025 showed a hazard ratio of 1.8 (95% CI 1.1 - 2.9), consistent with the Danish findings.

Absolute risk versus relative risk: what the numbers mean for you

The Harvard and Danish studies report relative risk (hazard ratios of 2 to 4). Media coverage amplified these numbers because they sound large. But relative risk without context is meaningless.

Here's the absolute risk calculation:

Baseline NAION incidence in adults over 50 with diabetes: approximately 4 to 5 per 100,000 person-years.

With a 3-fold increased relative risk (middle estimate from the studies):

• Absolute risk on semaglutide: 12 to 15 per 100,000 person-years
• Absolute risk increase: 8 to 10 additional cases per 100,000 person-years

Translation: if 100,000 people take semaglutide for one year, 8 to 10 additional people will develop NAION compared to if those same people took a different medication.

Or, flipping it: 99,990 out of 100,000 people on semaglutide for one year will NOT develop NAION.

For comparison, the absolute risk increase for major cardiovascular events (heart attack, stroke) prevented by semaglutide in the SELECT trial was 150 per 100,000 person-years (Lincoff et al., New England Journal of Medicine, 2023). The cardiovascular benefit is an order of magnitude larger than the NAION risk.

This is the core of the risk-benefit calculation. For most patients, especially those with established cardiovascular disease or high cardiovascular risk, the proven cardiovascular benefit outweighs the possible NAION risk by a wide margin.

The biological mechanism: why GLP-1 drugs might affect optic nerve blood flow

The mechanism linking semaglutide to NAION is speculative. No published study has demonstrated a direct causal pathway. But several hypotheses exist:

Hypothesis 1: Nocturnal hypotension. GLP-1 receptor agonists lower blood pressure modestly (average 3 to 5 mmHg systolic reduction in clinical trials). NAION typically occurs at night when blood pressure naturally drops during sleep. The optic nerve head is a watershed zone with marginal blood flow even under normal conditions. In patients with "crowded disc" anatomy (see next section), the optic nerve is already vulnerable. A modest additional blood pressure drop could tip the balance into ischemia.

Hypothesis 2: Altered microvascular autoregulation. GLP-1 receptors are expressed in vascular endothelium. Activation may alter local blood flow regulation in small vessels, including those supplying the optic nerve. A 2022 study by Sørensen et al. in Diabetes Care showed that liraglutide (another GLP-1 agonist) reduced retinal blood flow velocity by 8% in patients with type 2 diabetes. Whether this extends to optic nerve perfusion is unknown.

Hypothesis 3: Weight loss and metabolic changes. Rapid weight loss (common with semaglutide) can trigger changes in blood volume, blood viscosity, and coagulation factors. Some case reports of NAION in bariatric surgery patients suggest metabolic changes during rapid weight loss may contribute. This hypothesis is weak because the Danish study found no dose-response relationship, and higher doses cause more weight loss.

Hypothesis 4: Confounding by indication. Patients prescribed semaglutide have diabetes or obesity, both independent risk factors for NAION. Even with statistical adjustment, residual confounding is possible. Perhaps semaglutide users have more severe metabolic disease or more cardiovascular risk factors that weren't fully captured in the registry data.

None of these hypotheses has been proven. The FDA has not issued a warning or required label changes as of April 2026, which suggests the agency views the evidence as insufficient to establish causation.

Who is actually at risk: the "crowded disc" and other anatomical predictors

NAION does not occur randomly. It clusters in patients with specific anatomical features.

The strongest predictor is "disc at risk" or "crowded disc" anatomy. This refers to a small optic disc with a small central cup (the depression in the center of the optic nerve head). The cup-to-disc ratio is typically less than 0.2 in patients who develop NAION, compared to 0.3 to 0.4 in the general population.

Why does this matter? The optic nerve fibers and blood vessels pass through a narrow opening in the sclera called the lamina cribrosa. In a crowded disc, the opening is tight and the blood vessels are compressed. Any reduction in perfusion pressure (from low blood pressure, atherosclerosis, or other causes) can cause ischemia.

Crowded disc anatomy is present in roughly 1 to 3% of the population. It's bilateral (both eyes have the anatomy), which is why 15% of NAION patients develop it in the fellow eye within 5 years.

Other anatomical and clinical risk factors for NAION:

• Age over 50. NAION is rare under age 50. Incidence increases sharply after 60.
• Hypertension. Present in 50 to 70% of NAION patients.
• Diabetes. Roughly doubles baseline NAION risk independent of GLP-1 medications.
• Obstructive sleep apnea. Associated with nocturnal hypoxia and hypotension.
• Nocturnal hypotension. Blood pressure drops during sleep; optic nerve perfusion drops below critical threshold.
• Hyperlipidemia and atherosclerosis. Reduced blood flow to optic nerve head.
• Smoking. Damages microvascular endothelium.
• Phosphodiesterase-5 inhibitors (Viagra, Cialis). Small but real NAION risk, likely through hypotension mechanism.

None of the published semaglutide studies stratified risk by these factors. We don't know whether the NAION risk is concentrated in patients with crowded disc anatomy or distributed evenly. This is the most important unanswered question.

If the risk is concentrated in the 1 to 3% of patients with crowded disc anatomy, then screening with dilated fundoscopy before starting semaglutide could identify high-risk patients. If the risk is evenly distributed, screening won't help.

What most articles get wrong about this risk

Most coverage of the semaglutide-NAION connection makes three specific errors:

Error 1: Confusing NAION with diabetic retinopathy. Diabetic retinopathy is a chronic, progressive condition caused by high blood sugar damaging retinal blood vessels. NAION is an acute, non-progressive event caused by optic nerve ischemia. They are completely different conditions with different mechanisms, different time courses, and different risk factors. Semaglutide improves glycemic control and likely reduces diabetic retinopathy risk (the FOCUS trial showed a trend toward reduced retinopathy progression with semaglutide, though not statistically significant). The NAION risk is separate and unrelated to retinopathy.

Error 2: Assuming causation from observational data. The published studies are observational cohorts, not randomized trials. Observational studies can identify associations but cannot prove causation. The Bradford Hill criteria for causation require strength of association (present), consistency (present across multiple studies), biological gradient (absent, no dose-response), plausibility (weak, no proven mechanism), and experimental evidence (absent). We have 2 out of 5 criteria met. Causation is possible but not established.

Error 3: Ignoring the fellow-eye problem. NAION is almost always unilateral (one eye). The fellow eye has the same anatomy and the same risk factors, yet only 15% of patients develop NAION in the second eye over 5 years. If semaglutide were a strong causal factor, we'd expect bilateral cases or rapid sequential involvement. The fact that most cases remain unilateral suggests anatomy and chance play larger roles than medication exposure.

The correct framing is: semaglutide is associated with a 2 to 4-fold increased relative risk of a rare event. The absolute risk remains low. Causation is unproven. For most patients, the cardiovascular and metabolic benefits outweigh the possible NAION risk.

Symptoms that mean NAION versus symptoms that mean something else

NAION symptoms (call an ophthalmologist same day):

• Sudden, painless vision loss in one eye, usually upon waking
• Vision loss described as a curtain, shadow, or dark area blocking part of the visual field
• Upper or lower visual field affected more than central vision (though central vision can be involved)
• No eye pain, no redness, no floaters
• Vision loss is maximal at onset (doesn't worsen over hours or days)
• Pupil in the affected eye may react sluggishly to light

Symptoms that suggest something other than NAION:

• Gradual vision loss over weeks to months. More consistent with diabetic retinopathy, cataract, or glaucoma.
• Flashes of light and floaters. Suggests retinal detachment or posterior vitreous detachment. Emergency evaluation.
• Eye pain with vision loss. Suggests optic neuritis (inflammatory, often associated with multiple sclerosis) or acute angle-closure glaucoma. Emergency evaluation.
• Bilateral simultaneous vision loss. Rare in NAION. Consider stroke, cortical blindness, or bilateral retinal detachment.
• Transient vision loss lasting seconds to minutes. Suggests amaurosis fugax (transient retinal ischemia, often a warning sign of stroke). Urgent evaluation.
• Vision loss with headache and jaw pain. Suggests giant cell arteritis (arteritic ischemic optic neuropathy, not NAION). Emergency evaluation and immediate high-dose steroids.

If you experience sudden vision loss while on semaglutide, the protocol is:

1. Call an ophthalmologist or go to an emergency department the same day
2. Do not assume it's related to the medication (many causes of sudden vision loss exist)
3. Do not stop semaglutide without provider guidance (stopping won't reverse NAION if it's already occurred)
4. Expect a dilated eye exam, visual field testing, and possibly optical coherence tomography (OCT) imaging

NAION is diagnosed clinically by the appearance of the optic disc on fundoscopy: pale, swollen optic nerve head with splinter hemorrhages. The diagnosis is usually obvious to an ophthalmologist.

The decision framework: when the NAION risk changes your treatment calculus

The NAION risk does not automatically disqualify semaglutide. For most patients, the benefit-risk ratio remains favorable. But certain clinical scenarios shift the calculus:

Scenario 1: You have already had NAION in one eye. The 5-year risk of NAION in the fellow eye is roughly 15% (Hattenhauer et al., American Journal of Ophthalmology, 1997). If semaglutide increases that risk 3-fold, the absolute risk becomes 45% over 5 years. This is high enough to reconsider treatment. Alternative options include tirzepatide (unknown NAION risk, see below), metformin plus lifestyle modification, or bariatric surgery if BMI is over 35.

Scenario 2: You have crowded disc anatomy identified on prior eye exams. If an ophthalmologist has documented small cup-to-disc ratio or "disc at risk" in your chart, you're in the high-risk anatomical group. The absolute NAION risk on semaglutide is unknown but likely higher than the population average. Discuss alternatives with your provider. If semaglutide is the best option for other reasons (cardiovascular disease, severe obesity), proceed with informed consent and close monitoring.

Scenario 3: You have multiple NAION risk factors (sleep apnea, nocturnal hypotension, severe atherosclerosis). The risk is cumulative. Each additional risk factor increases baseline NAION incidence. Adding semaglutide on top of multiple other risk factors may push absolute risk into a range where alternatives are preferable.

Scenario 4: You have established cardiovascular disease or very high cardiovascular risk. The SELECT trial showed semaglutide reduces major adverse cardiovascular events by 20% in patients with established cardiovascular disease (Lincoff et al., New England Journal of Medicine, 2023). For this population, the proven cardiovascular benefit almost certainly outweighs the possible NAION risk. The number needed to treat to prevent one cardiovascular event is roughly 67 over 3 years. The number needed to harm (cause one NAION case) is roughly 10,000 to 12,500 per year. The benefit-risk ratio is 150:1 in favor of treatment.

Scenario 5: You are using semaglutide for cosmetic weight loss with BMI under 27 and no metabolic disease. The benefit-risk ratio is less favorable. The cardiovascular benefit doesn't apply (you don't have cardiovascular disease). The metabolic benefit is modest (you don't have diabetes or prediabetes). The NAION risk, though small, is not offset by large benefits. Consider lifestyle modification first.

The FormBlends decision tree for NAION risk:

START: Patient considering semaglutide

↓

Has patient had NAION in one eye previously? ├─ YES → High risk. Discuss alternatives (tirzepatide, metformin, lifestyle, surgery). If semaglutide chosen, document informed consent. └─ NO → Continue

↓

Does patient have documented "crowded disc" or small cup-to-disc ratio? ├─ YES → Moderate-high risk. Ophthalmology consultation before starting. Consider alternatives. └─ NO or UNKNOWN → Continue

↓

Does patient have established cardiovascular disease or very high CV risk? ├─ YES → Cardiovascular benefit likely outweighs NAION risk. Proceed with semaglutide. Counsel on NAION symptoms. └─ NO → Continue

↓

Does patient have 2+ other NAION risk factors (age >60, sleep apnea, nocturnal hypotension, severe atherosclerosis)? ├─ YES → Moderate risk. Discuss risk-benefit. Consider alternatives if patient is risk-averse. └─ NO → Low risk. Proceed with semaglutide. Counsel on NAION symptoms.