Peptide News Digest
Evidence Brief (Updated ) 11 min read

GLP-1 Drugs and NAION: Strongest Evidence for Semaglutide, but the FDA Now Warns on the Whole Class

On November 9, 2026, the FDA required every GLP-1 drug sold in the U.S., from Ozempic and Wegovy to Mounjaro, Zepbound, and Foundayo, to warn about non-arteritic anterior ischemic optic neuropathy (NAION), a rare eye condition that can cause permanent vision loss. Australia's regulator had already updated GLP-1 labels in July 2026, while its advisory committee concluded that the evidence supported the signal for semaglutide but not yet for dulaglutide or tirzepatide. This piece explains what NAION is, what each regulator found, how large the risk is, and what people taking these drugs should do.

Key points

  • NAION is a rare eye condition that causes sudden, painless vision loss in one eye; the FDA puts the general rate at 2.3 to 10.2 cases per 100,000 adults over 50 each year, and no treatment reliably restores lost vision
  • On November 9, 2026, the FDA required every GLP-1 drug in the U.S., including Ozempic, Wegovy, Mounjaro, Zepbound, and Foundayo, to add NAION to the Warnings and Precautions section of its label, calling the risk potential
  • In the FDA's pooled analysis of 126 trials with 134,253 patients, NAION occurred at 6.4 vs. 3.8 cases per 100,000 person-years on GLP-1 drugs vs. placebo or other glucose-lowering drugs (10 events vs. 5, all 5 on placebo)
  • The FDA's Sentinel study in type 2 diabetes found a higher risk with GLP-1 drugs than DPP-4 inhibitors (hazard ratio 1.43) but no significant difference from SGLT2 inhibitors (1.19)
  • Australia's TGA updated all GLP-1 labels in July 2026; its advisory committee said the evidence supported the signal for semaglutide but not yet for dulaglutide or tirzepatide, which had been studied far less
  • The UK's MHRA issued a Drug Safety Update on semaglutide and NAION on February 5, 2026
  • The 2024 Massachusetts Eye and Ear study found hazard ratios of 4.28 in type 2 diabetes and 7.64 in overweight or obesity; larger registry and claims studies since then found smaller increases, such as 2.81 in Denmark and Norway and 1.85 in a 2026 claims study of all GLP-1 drugs
  • Absolute risk is small: about 1.4 to 2.9 extra cases per 10,000 person-years in the large observational studies, or about one extra case per 2,700 people treated in a 2026 meta-analysis
  • No study has shown that switching from semaglutide to tirzepatide lowers NAION risk
  • Get same-day eye care for sudden vision loss in one eye on a GLP-1 drug; do not stop the drug without talking to your prescriber, and if NAION occurs, discuss whether to continue

The short version

Non-arteritic anterior ischemic optic neuropathy (NAION) is a rare eye condition in which blood flow to the front of the optic nerve drops suddenly, causing painless vision loss in one eye. No treatment reliably restores the lost vision. According to the FDA, NAION occurs in about 2.3 to 10.2 of every 100,000 adults over 50 each year.

Concern about GLP-1 drugs began with a July 2024 study from Massachusetts Eye and Ear that linked semaglutide prescriptions to higher NAION rates. Larger studies since then have mostly found a smaller increase, and most of them looked at semaglutide because it had the most users. Regulators then acted one by one. The UK's MHRA issued a Drug Safety Update on semaglutide in February 2026. Australia's Therapeutic Goods Administration (TGA) updated the labels of every GLP-1 drug sold there in July 2026, while its advisory committee said the evidence supported the signal for semaglutide but not yet for dulaglutide or tirzepatide.

On November 9, 2026, the FDA went further than either. It required every GLP-1 drug in the U.S. to add NAION to the Warnings and Precautions section of its label. The FDA called the risk potential, not proven. It said NAION cases have been reported for every marketed GLP-1 drug, and it did not single out semaglutide.

The absolute risk is small. In the FDA's pooled analysis of 126 clinical trials, NAION occurred at 6.4 cases per 100,000 person-years on GLP-1 drugs and 3.8 on comparison treatments. The practical advice is the same across regulators: get same-day eye care for sudden vision loss in one eye, and do not stop the drug without talking to your prescriber.

What NAION actually is

The optic nerve carries visual signals from the retina to the brain. The front portion of the optic nerve (the optic nerve head, visible during a routine dilated eye exam as the optic disc) receives its blood supply from a small network of vessels branching from the ophthalmic artery.

Non-arteritic anterior ischaemic optic neuropathy (NAION) occurs when the blood supply to the optic nerve head is acutely reduced or interrupted, damaging the nerve fibers that carry visual information out of the eye. The word 'non-arteritic' distinguishes NAION from a similar condition (arteritic AION) that involves inflamed arteries and behaves differently clinically.

The presentation is characteristic and abrupt. A person wakes up in the morning, or notices during the day, that vision in one eye has gone partially dark, blurry, or lost in a particular field. There is no pain. The affected eye typically shows a specific pattern on visual field testing, most often loss of the lower half of the field of vision.

Risk factors for NAION include age over 50, a small optic disc (the anatomical setup called 'disc at risk'), high blood pressure, diabetes, sleep apnea, smoking, anemia, and kidney disease. The absolute risk is low. The FDA puts the rate at 2.3 to 10.2 cases per 100,000 adults over 50 each year, and people with the anatomical risk factor have higher rates.

Treatment options are limited. Once NAION has occurred, no intervention has been shown to reliably improve visual acuity outcomes. Some patients recover partial vision on their own over months. Most do not recover the lost vision. About 15-20% of people who have NAION in one eye will develop NAION in the other eye within five years.

That prognostic reality is why regulatory agencies pay close attention to any drug that might elevate NAION risk, even at low absolute rates. The condition is permanent, functionally significant, and untreatable after the fact.

How the signal emerged

The first widely publicized concern came from a July 2024 study in JAMA Ophthalmology by researchers at Massachusetts Eye and Ear. They searched records of patients seen by the hospital's neuro-ophthalmologists and compared people prescribed semaglutide with people taking other diabetes or weight-loss drugs. Among 710 patients with type 2 diabetes, the hazard ratio for NAION was 4.28; among 979 patients with overweight or obesity, it was 7.64. A hazard ratio of 4 means the event happened about four times as often. Because the patients came from a specialist eye clinic, the study likely overstated how common NAION is in everyday practice, and the authors said it could not show cause.

Larger studies followed, and most found a smaller increase:

  • A Danish and Norwegian registry study of about 61,000 semaglutide users with type 2 diabetes, published in 2025, found a hazard ratio of 2.81 compared with SGLT2 inhibitors, or about 1.4 extra cases per 10,000 person-years.
  • A 2025 study across 14 claims and health-record databases in the OHDSI network, covering 810,390 new semaglutide users with type 2 diabetes, found a modest increase, smaller than the 2024 study. Compared with three other diabetes drugs, the risk was significantly higher only against empagliflozin and only under a strict definition of NAION, while a within-person analysis found a 32% higher rate during semaglutide use.
  • An insurance-claims study of 482,912 matched pairs with type 2 diabetes, published in 2026, found NAION 85% more often in people starting any GLP-1 drug than in people starting an SGLT2 inhibitor (hazard ratio 1.85), or about 0.29 extra cases per 1,000 person-years.
  • A 2026 meta-analysis in Diabetes Care of 15 studies with more than 1.5 million patients found odds of NAION about 1.7 times higher with GLP-1 drugs, with an absolute risk of 0.09%. That works out to about one extra case for every 2,700 people treated.

Almost all of these studies were of people with type 2 diabetes, and most measured semaglutide or GLP-1 drugs as a group. Few looked at tirzepatide on its own. Case reports of NAION after tirzepatide have been published, but there is little comparative data either way.

The FDA's November 2026 warning

On November 9, 2026, the FDA issued a Drug Safety Communication requiring the makers of all GLP-1 drugs to add information about NAION to the Warnings and Precautions section of their labels. The list covers generic exenatide, Victoza, Trulicity, Saxenda, Xultophy, Soliqua, Ozempic, Rybelsus, Wegovy, Mounjaro, Zepbound, and Foundayo.

The FDA drew on four kinds of evidence:

  • Clinical trials. It pooled 126 trials with 134,253 patients and 288,879 person-years of follow-up. It found 15 probable NAION events: 10 in people on GLP-1 drugs, 5 on placebo, and none on other diabetes drugs. The rates were 6.4 and 3.8 per 100,000 person-years.
  • Its own Sentinel study. In adults with type 2 diabetes who started a drug between 2019 and 2024, NAION occurred at 26 to 32 cases per 100,000 person-years across the groups. GLP-1 drugs showed a higher risk than DPP-4 inhibitors (hazard ratio 1.43, 95% CI 1.02 to 2.01) but not a significant difference from SGLT2 inhibitors (1.19, 0.91 to 1.54).
  • Published studies. The FDA said the most rigorous observational studies consistently found a higher NAION risk in people with type 2 diabetes who started GLP-1 drugs than in those who started DPP-4 or SGLT2 inhibitors.
  • Adverse event reports and animal data. Reports were hard to interpret because NAION's risk factors are common in people who take these drugs, and animal studies did not show how the drugs could cause NAION.

The agency concluded there is "a potential increased risk of rare NAION events." It noted that NAION cases have been reported for every GLP-1 drug on the market and applied the warning to the whole class.

Australia's July 2026 action and the semaglutide-tirzepatide split

The TGA action on July 25, 2026 has two parts.

Part 1: Class-wide label update. All five GLP-1 receptor agonists currently marketed in Australia get updated Product Information (equivalent to the US FDA-approved label) that adds NAION under Adverse Reactions and includes the standard patient guidance about seeking urgent medical attention for sudden vision loss. The five products are:

  • Trulicity (dulaglutide, Eli Lilly)
  • Ozempic (semaglutide, Novo Nordisk)
  • Wegovy (semaglutide, Novo Nordisk)
  • Mounjaro (tirzepatide, Eli Lilly)
  • Saxenda (liraglutide, Novo Nordisk)

Part 2: The Advisory Committee on Medicines analysis. The committee reviewed the underlying evidence base (published case-control studies, cohort studies, and pharmacovigilance data) and reached the more specific conclusion that the current evidence supports the NAION signal for semaglutide, but does not currently support it for dulaglutide or tirzepatide.

The underlying case data from the Australian Database of Adverse Event Notifications (DAEN) as of the review:

  • Total cases of optic ischaemic neuropathy reported with GLP-1 receptor agonists: 36
  • Semaglutide (Ozempic + Wegovy): 23 cases
  • Tirzepatide (Mounjaro): 10 cases
  • Liraglutide (Saxenda): 3 cases

The raw case counts alone do not resolve the drug-specific question because they reflect exposure volume as much as any real risk difference. Semaglutide has been on the Australian market longer and been prescribed to substantially more patients than tirzepatide (which launched later). The published observational studies available at the time mostly examined semaglutide and found elevated risk, while far less data existed for tirzepatide.

The FDA's review four months later did not draw that line. It applied the warning to every GLP-1 drug, noting that NAION cases have been reported for all of them. The two positions are less contradictory than they look. Evidence that does not yet support a signal for tirzepatide may largely reflect how little tirzepatide had been studied, and it is not evidence that tirzepatide is safer. No study has shown that switching from semaglutide to tirzepatide lowers NAION risk.

Why the risk might differ between drugs

GLP-1 receptor agonists share a mechanism: they bind and activate the glucagon-like peptide-1 (GLP-1) receptor. But the individual drugs are structurally distinct molecules with different pharmacokinetic profiles, different receptor-binding kinetics, different secondary receptor activities, and different exposure duration.

The differences that could matter for NAION risk:

Half-life and steady-state exposure. Semaglutide has a half-life of about a week. Tirzepatide has a half-life of about five days. Dulaglutide has a half-life of about five days. Liraglutide has a half-life of about 13 hours (requires daily injection). Longer steady-state exposure to a drug that affects blood-vessel dynamics or glycemic swings might correlate with different vascular risk profiles.

Receptor specificity. Tirzepatide is a dual GIP/GLP-1 receptor agonist (activates both receptors). Semaglutide is a selective GLP-1 receptor agonist. GIP receptor activation has downstream effects on lipid handling and blood-vessel function that partially differentiate the classes.

Rate and magnitude of glycemic change. One hypothesis about the NAION signal is that it may be mediated by rapid changes in blood glucose during treatment initiation and dose escalation. Sudden reductions in chronic hyperglycemia can trigger transient changes in optic-nerve microvascular perfusion. Drugs with different titration schedules and different rates of glycemic reduction could produce different NAION risk patterns.

Weight loss velocity. Rapid weight loss has been documented as a risk factor for other ischemic events. Tirzepatide typically produces faster weight loss than semaglutide over the same time window, which if anything argues against the drug-specific split observed in the data.

These differences are hypotheses. No study has tested them directly, and the FDA's review of animal data found no mechanism by which GLP-1 drugs could cause NAION. Whether the risk truly differs between drugs, or whether semaglutide simply has the longest record and the most users, is not yet known.

Absolute risk in real-world context

The relative numbers can sound alarming, but the absolute numbers are small. Here is how the main estimates translate:

  • In the FDA's pooled trials, the difference between 6.4 and 3.8 cases per 100,000 person-years is about 2.6 extra cases for every 100,000 people treated for a year.
  • In the Danish-Norwegian semaglutide study, the difference was about 1.4 extra cases per 10,000 person-years.
  • In the 2026 claims study of all GLP-1 drugs, it was about 2.9 extra cases per 10,000 person-years.
  • The Diabetes Care meta-analysis estimated one extra case for about every 2,700 people treated.

Rates are higher in people with type 2 diabetes, who already have more NAION: in the FDA's Sentinel study, they ranged from 26 to 32 cases per 100,000 person-years whatever drug people started.

Those numbers need to be weighed against what the drugs are prescribed to prevent. In the SELECT trial of semaglutide in people with heart disease and overweight or obesity, heart attack, stroke, or cardiovascular death occurred in 6.5% of people on semaglutide and 8.0% on placebo over about three years. For someone with that kind of risk, the expected benefit is much larger than the expected NAION harm. For someone taking the drug mainly for weight loss, the balance depends more on individual risk factors, which is a conversation for a prescriber.

Signs to watch for and what to do

The FDA lists these symptoms of NAION: sudden, painless vision loss in one eye that people may notice on waking; blurred or missing parts of vision, such as blind spots or tunnel vision; colors that look distorted; and blurry or hazy central vision. Contact a health care professional right away if any of these happen. In practice that means an emergency department or an eye doctor the same day, not an appointment next week.

Do not stop GLP-1 therapy on your own. Stopping suddenly can raise blood sugar in people with diabetes and lead to weight regain. If you have had NAION, the FDA advises talking with your health care professional about whether to continue or stop the drug. It tells clinicians to refer suspected cases to an ophthalmologist and to consider stopping the GLP-1 drug in people who develop NAION unless the benefits clearly outweigh the risk.

Ask about your eyes before you start. A dilated eye exam can show whether you have a small, crowded optic disc, the anatomy most strongly linked to NAION. No guideline requires this exam before starting a GLP-1 drug, but it can be reasonable if you are over 50 or have other risk factors.

Report symptoms. In the U.S., patients and clinicians can report side effects to the FDA through MedWatch. Reports help regulators detect and size safety signals.

Questions to ask your prescriber

If you take a GLP-1 drug or are thinking about starting one, these questions are worth raising.

What is my baseline NAION risk? Age, high blood pressure, sleep apnea, diabetes, smoking, kidney disease, and a small optic disc all raise it. A prior NAION in one eye raises the risk for the other eye.

Does the drug I take matter? The strongest evidence involves semaglutide, mostly because it has been studied the most. The FDA's warning applies to every GLP-1 drug, and no study has shown that switching to a different one lowers NAION risk.

How large is my expected benefit? If you take the drug to lower heart or kidney risk, the benefit is likely to be much larger than the NAION risk. If you take it mainly for weight, ask how your other risk factors change the balance.

When should I get an eye exam? If it has been more than two years since your last dilated eye exam, this is a reasonable time to book one, especially if you are over 50 or have diabetes.

What should I do if my vision changes suddenly? Get same-day care. Agree on the plan before you need it.

What we do not yet know

Whether the signal will hold in trials built to measure it. The FDA's pooled analysis of 126 trials found 15 probable cases in total, too few to give a precise estimate, because the trials were designed for other outcomes. Most other evidence is observational. Purpose-built prospective studies would take years.

Whether the effect is dose-related. Semaglutide is prescribed at 0.25-2.4 mg per week depending on indication. The observational studies to date have not consistently disentangled dose effects from exposure duration. A dose-response relationship would strengthen the case for a real drug effect; a lack of one would argue for exposure-duration or other confounding as the driver.

Whether the mechanism is direct drug effect or mediated by glycemic change. If NAION risk is driven by rapid changes in blood glucose during treatment initiation, then slower dose titration might mitigate the risk. If the effect is a direct receptor-mediated action on the optic-nerve microvasculature, titration wouldn't help. The FDA's review of animal data did not identify a mechanism.

Whether tirzepatide will eventually show the same signal at longer exposure duration. Tirzepatide has been on the market for less time than semaglutide and has been studied less for NAION. The FDA's class-wide warning means U.S. labels now treat it the same way; whether its true risk differs is still unknown.

The bottom line

NAION is a rare but serious possible side effect of GLP-1 drugs. As of November 9, 2026, the FDA requires every GLP-1 drug in the U.S. to warn about it, following earlier actions in the UK and Australia. The FDA calls the risk potential, and the absolute numbers are small: a few extra cases for every 10,000 to 100,000 person-years of treatment, depending on the study.

The strongest evidence involves semaglutide, and Australia's advisory committee said in July 2026 that the evidence did not yet support the signal for tirzepatide or dulaglutide. That reflected limited data, not proof of safety, and the FDA's review applied the warning to the whole class. Switching drugs is not an established way to lower the risk.

For most people with heart or kidney disease, the benefits of these drugs remain much larger than the NAION risk. Anyone taking one should know the symptoms, get same-day care for sudden vision loss in one eye, and talk with their prescriber before stopping.

Updates and corrections

Update and correction (November 9, 2026): We updated this article for the FDA's November 9, 2026 Drug Safety Communication requiring a NAION warning on all GLP-1 drugs, and changed the title from "The GLP-1 Eye-Safety Signal: NAION Cases Cluster Around Semaglutide, Not Tirzepatide" because the FDA did not single out semaglutide. We removed statements that the FDA had not acted and a prediction about when it would. We also removed advice to consider switching from semaglutide to tirzepatide to lower NAION risk, which no study supports. We corrected the description of the 2024 Massachusetts Eye and Ear study (its analysis covered 710 patients with type 2 diabetes and 979 with overweight or obesity, not 16,827, and found hazard ratios of 4.28 and 7.64). We replaced uncited descriptions of Danish, U.S. insurance-claims, and Nordic studies, including a claim that a Nordic study compared semaglutide with tirzepatide that we could not verify, with the published studies now cited. We replaced estimated absolute-risk figures and an uncited SELECT benefit estimate with figures from the FDA, the cited studies, and the SELECT trial publication.

Limitations

  • Most evidence linking GLP-1 drugs to NAION is observational; the FDA's pooled trial analysis found only 15 probable events in total, too few for a precise estimate
  • Most studies covered people with type 2 diabetes taking semaglutide or GLP-1 drugs as a group; there is little data on tirzepatide alone or on people taking the drugs only for weight loss
  • The mechanism, if there is one, is unknown; the FDA's review of animal data did not identify one, and confounding by diabetes, sleep apnea, and blood pressure is hard to rule out
  • Studies define NAION differently, often from diagnosis codes, which changes the estimates
  • The TGA details in this article come from its July 2026 safety alert and were not re-checked for this update because the TGA website could not be reached
  • This piece does not address risk in specific subpopulations (pediatric, pregnancy, prior NAION history, specific comorbid conditions) where individual risk-benefit calculations may differ substantially from the general adult population
  • Treatment decisions about starting, continuing, switching, or stopping GLP-1 receptor agonist therapy belong with a prescribing clinician who can evaluate individual medical history, current medications, indication, and clinical goals; this piece exists to inform that conversation, not to substitute for it

Citations

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
    Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes
    peer-reviewed New England Journal of Medicine 2023
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.

Peptides in this article

Full peptide profiles with evidence levels, dosing data, and safety notes live on peptidelist.org.

Related insights