Do Ozempic and similar GLP-1 drugs increase the risk of a rare form of vision loss?
Current evidence suggests an association between GLP-1 drugs and NAION, but it has not established that these drugs cause this rare form of vision loss.
Where the claims stand
Reports linking Ozempic (semaglutide) and other GLP-1 drugs (formally, GLP-1 receptor agonists) to blindness have drawn widespread attention. The current evidence does not establish that these drugs cause blindness, but several observational studies have reported an association with non-arteritic anterior ischemic optic neuropathy (NAION), a rare form of sudden optic nerve injury. Separately, semaglutide has a longer-standing warning about diabetic retinopathy complications — a different eye problem that public discussion sometimes conflates with NAION. We are tracking whether accumulating evidence supports a causal relationship between GLP-1 drugs and NAION, or whether the observed association is explained by underlying risk factors such as diabetes and vascular disease.
We'll only notify you when something material changes.
Additional information
Status
as of July 1, 2026Multiple independent observational studies have reported an association between GLP-1 drugs and NAION, but none can establish causation. Researchers continue to investigate whether the relationship reflects a drug effect, characteristics of patients prescribed these medications, or other confounding factors. Regulators and ophthalmologists are monitoring the evidence, but no consensus has concluded that semaglutide causes NAION. Other ocular concerns — especially diabetic retinopathy complications noted in clinical trials and labeling — are tracked here as a separate question from the NAION signal.
Confidence — current state
We are cautiously confident that there is a real signal worth investigating because multiple independent observational studies have reported elevated rates of NAION among GLP-1 drug users. However, the existing evidence remains observational and cannot determine causation.
This is our best read given the published evidence we have reviewed — not a claim of absolute truth.
Open questions
Does semaglutide itself increase the risk of NAION?
Current studies are observational and cannot fully separate drug effects from patient risk factors.
Are some GLP-1 drugs associated with higher risk than others?
Different drugs may have different pharmacology or prescribing populations.
Which patients are most susceptible?
Understanding individual risk could improve prescribing decisions without unnecessarily discouraging treatment.
What biological mechanism could explain the association?
A plausible mechanism would strengthen evidence for causality.
Are other reported ocular risks of GLP-1 drugs — especially diabetic retinopathy complications — related to the NAION signal, or separate questions?
Public discussion of “Ozempic blindness” often conflates sudden optic-nerve injury with retinopathy and other vision changes that have different evidence and mechanisms.
What would change our mind
- Large prospective studies finding no increased NAION risk after careful adjustment for diabetes severity and vascular risk factors.
- Multiple independent studies consistently reproducing the association across different populations.
- Mechanistic evidence demonstrating how GLP-1 drugs could directly contribute to optic nerve injury.
- Updated regulatory conclusions based on substantially larger bodies of evidence.
Claims & evidence
Each claim is tracked separately — not a single verdict.Several observational studies have reported an association between GLP-1 drugs and non-arteritic anterior ischemic optic neuropathy (NAION).
CorroboratedEvidence basisPeer-reviewed · independently corroborated- Peer-reviewedJuly 3, 2024Risk of Nonarteritic Anterior Ischemic Optic Neuropathy in Patients Prescribed Semaglutide
Retrospective study reporting increased incidence of NAION among semaglutide users compared with matched controls.
- Official statementMay 1, 2026Glucagon-Like Peptide-1 Receptor Agonists and the Risk of Non-Arteritic Anterior Ischemic Optic Neuropathy: A Consensus Statement by NANOS and the American Academy of Ophthalmology
Professional society summarizes the observational findings — that there is a possible link between semaglutide and NAION — while emphasizing they do not establish causation.
Current evidence does not establish that Ozempic or semaglutide causes NAION.
SupportedEvidence basisPeer-reviewed · independently corroborated- Peer-reviewedJuly 3, 2024Semaglutide and Nonarteritic Anterior Ischemic Optic Neuropathy
Observational studies identify an association but cannot determine causality.
- Official statementMay 1, 2026Glucagon-Like Peptide-1 Receptor Agonists and the Risk of Non-Arteritic Anterior Ischemic Optic Neuropathy: A Consensus Statement by NANOS and the American Academy of Ophthalmology
Experts caution patients not to discontinue medication based solely on the observational findings.
NAION is a rare condition, even among patients taking GLP-1 drugs.
CorroboratedEvidence basisPeer-reviewed · independently corroborated- Scholarly bookStatPearls / NCBI Bookshelf (Kaur & Margolin)September 14, 2025Nonarteritic Anterior Ischemic Optic Neuropathy
The annual incidence of NAION in the United States is estimated to be 2.3 to 10.3 cases per 100,000 individuals aged 50 or older.
- Peer-reviewedJuly 3, 2024Risk of Nonarteritic Anterior Ischemic Optic Neuropathy in Patients Prescribed Semaglutide
Absolute numbers of NAION events remained small.
People with diabetes already have a higher baseline risk of NAION than the general population.
CorroboratedEvidence basisOfficial statement · independently corroborated- Official statementJanuary 1, 2023Non-Arteritic Anterior Ischemic Optic Neuropathy
Diabetes mellitus is an established vascular risk factor for NAION.
- Official statementJanuary 1, 2023NAION Clinical Guidance
Diabetes and vascular disease are recognized risk factors.
The FDA-approved prescribing information for Ozempic does not currently identify NAION as an established adverse reaction.
CorroboratedEvidence basisOfficial statement · single source- Official statementJanuary 1, 2025Ozempic Prescribing Information
Current labeling does not list NAION as an established adverse reaction.
Semaglutide has a separate, longer-standing concern about diabetic retinopathy complications that is distinct from the more recent NAION signal.
CorroboratedEvidence basisPeer-reviewed · independently corroborated- Peer-reviewedSeptember 15, 2016Semaglutide and Cardiovascular Outcomes in Patients with Type 2 Diabetes
Diabetic retinopathy complications occurred in 3.0% of patients in the semaglutide group and 1.8% in the placebo group (hazard ratio, 1.76); authors noted rapid glucose lowering as a possible explanation while not ruling out a direct drug effect.
- Official statementJanuary 1, 2025Ozempic Prescribing Information
Warns that diabetic retinopathy complications were reported in a clinical trial and that patients with a history of diabetic retinopathy should be monitored; rapid improvement in glucose control has been associated with temporary worsening of diabetic retinopathy.
What this doesn’t establish
Claims commonly associated with this story that the available evidence does not establish. Confirming a narrow fact here is not confirmation of the broader narrative around it. As such, these claims are not included in the claims bar above.
Ozempic (semaglutide) or other GLP-1 drugs cause NAION, the rare form of vision loss.
UnverifiedEvidence basisPeer-reviewed · independently corroborated- Peer-reviewedJuly 3, 2024Semaglutide and Nonarteritic Anterior Ischemic Optic Neuropathy
Observational studies identify an association but cannot determine causality; a causal relationship has not been established.
- Official statementMay 1, 2026Glucagon-Like Peptide-1 Receptor Agonists and the Risk of Non-Arteritic Anterior Ischemic Optic Neuropathy: A Consensus Statement by NANOS and the American Academy of Ophthalmology
The society emphasizes the findings do not establish causation and cautions patients not to discontinue medication based on them.
How we got here
6 updates · append-only- New evidence
AAO and NANOS reject a blanket recommendation to stop semaglutide after NAION
Responding to the EMA advice, the American Academy of Ophthalmology and the North American Neuro-Ophthalmology Society said they do not support a blanket recommendation that all patients stop semaglutide if NAION is diagnosed, citing potential serious risks to overall health from discontinuation. They advised shared decision-making among the patient and care team, noting that available evidence shows a possible link but has not established that semaglutide causes NAION, and that absolute risk remains low.
What changed
- Clinical guidance (US ophthalmology): Early 2024 caution; await more research Shared decision-making; no blanket stop-if-NAION rule
- Official confirmation
EMA concludes NAION is a very rare side effect of semaglutide medicines
After completing its review, PRAC concluded that NAION is a very rare side effect of semaglutide (up to about 1 in 10,000 people treated) and recommended updating EU product information accordingly. Epidemiological evidence suggested roughly a two-fold higher risk in adults with type 2 diabetes — about one additional case per 10,000 person-years of treatment. EMA advised patients with sudden or rapidly worsening vision loss to seek care promptly and recommended stopping semaglutide if NAION is confirmed. U.S. FDA labeling for Ozempic still did not list NAION as an established adverse reaction.
What changed
- Regulatory status (EU): PRAC review underway NAION listed as a very rare side effect; stop if confirmed
- US vs EU labeling: Neither US nor EU product information listed NAION EU recommends NAION labeling; US FDA label still does not list NAION
- New evidence
EMA safety committee opens a formal review of semaglutide and NAION
The European Medicines Agency’s Pharmacovigilance Risk Assessment Committee (PRAC) started a review of semaglutide medicines (Ozempic, Rybelsus, Wegovy) after recent observational studies raised concern about NAION risk. PRAC noted that some studies suggested an increased risk while others did not, and said it would assess clinical-trial, post-marketing, mechanistic, and literature data.
What changed
- Regulatory status (EU): No formal EMA safety review of NAION with semaglutide PRAC review of NAION risk underway
- New evidence
Large Danish cohort study reports roughly doubled NAION risk with once-weekly semaglutide
A nationwide Danish cohort of more than 424,000 people with type 2 diabetes (Grauslund et al.) found that once-weekly semaglutide independently more than doubled the five-year risk of NAION. Absolute event rates remained low. Other contemporaneous observational studies reported mixed results — some finding an association and others not — leaving the magnitude of any risk unsettled.
What changed
- Claim: association-naion: supported corroborated
- New evidence
Professional societies urge caution in interpreting the findings
The American Academy of Ophthalmology and the North American Neuro-Ophthalmology Society noted that the Hathaway analysis identified a potential link but was not the type of study that can show semaglutide causes NAION. They cautioned patients not to stop the medication on that basis alone and called for further research.
What changed
- Claim: no-causation-established: unverified supported
- New evidence
Large observational study reports elevated NAION incidence among semaglutide users
Researchers published a retrospective matched-cohort analysis (Hathaway et al., JAMA Ophthalmology) reporting higher observed rates of NAION among patients prescribed semaglutide. The study received widespread attention because it suggested a possible safety signal, while its authors emphasized that observational data cannot establish causality.
What changed
- Story status: breaking developing
Suggest a source
Point us to a primary source or a publisher correction. Sign in to submit — every suggestion is reviewed by a human before anything changes.
Confidence last reviewed July 1, 2026. Updates are append-only; nothing here is edited silently.