GLP-1 and eye health research points to two separate issues. First, NAION, a sudden loss of blood flow to the optic nerve, is now listed in Europe as a very rare side effect of semaglutide, affecting up to 1 in 10,000 people.[1] Second, in people with diabetes and existing retinopathy, a fast drop in blood sugar can temporarily worsen eye disease: in SUSTAIN-6, retinopathy complications occurred in 3.0% on semaglutide versus 1.8% on placebo.[6]
For the drug basics, see our semaglutide profile and the GLP-1 basics hub.
Bottom line
| Issue | What’s known | Who is most affected |
|---|---|---|
| NAION (optic nerve stroke) | Rare; about 2-3 times higher risk in some cohorts, roughly 1 extra case per 10,000 person-years[1][3] | Mostly shown in people with type 2 diabetes |
| Diabetic retinopathy complications | Early worsening with rapid glucose lowering[7] | People with diabetes who already have retinopathy |
| Other eye problems | No overall increase across 78 trials[5] | — |
What NAION is
Non-arteritic anterior ischemic optic neuropathy (NAION) happens when blood flow to the front of the optic nerve drops. It usually causes sudden, painless vision loss in one eye, often noticed on waking. Known risk factors include diabetes, high blood pressure, sleep apnea and a small, crowded optic disc. Vision loss is often permanent.
Regulatory status in 2026
Europe. In June 2025, the EMA’s safety committee (PRAC) reviewed trials, post-marketing reports and published studies. It concluded that NAION is a very rare side effect of semaglutide (Ozempic, Rybelsus and Wegovy), affecting up to 1 in 10,000 people, and recommended adding it to the product information. PRAC noted that several large studies in type 2 diabetes suggested about a twofold risk, or about one extra case per 10,000 person-years.[1]
United States. The June 2026 versions of the US Ozempic and Wegovy labels on DailyMed do not list NAION. They do carry a warning on diabetic retinopathy complications.[7][11] The Zepbound (tirzepatide) label added a diabetic retinopathy warning for people with type 2 diabetes in August 2026.[8]
Key NAION studies
Mass Eye and Ear cohort (JAMA Ophthalmology, 2024). This study first raised the alarm. Among patients seen at a neuro-ophthalmology service, semaglutide was linked to a higher NAION risk in type 2 diabetes (hazard ratio 4.28) and in overweight or obesity (7.64). The numbers were small (17 and 20 NAION events on semaglutide), and patients referred to a specialist eye clinic may not represent the general population.[2]
Danish-Norwegian registry (2025). Among 61,377 semaglutide users with type 2 diabetes, NAION occurred at 2.19 to 2.90 per 10,000 person-years, versus 0.92 to 1.18 among SGLT2 inhibitor users. The adjusted hazard ratio was 2.81, meaning about 1.4 extra cases per 10,000 person-years.[3]
Ophthalmology meta-analysis (2026). Pooling 8 randomized trials and 8 observational studies (about 1.6 million people), observational data in diabetes showed a higher risk (hazard ratio 1.85). Data in people using semaglutide for weight loss did not show a clear increase, and trials had too few events to be conclusive.[4]
Neurology meta-analysis (2026). Across five observational studies with about 1.6 million people, semaglutide was linked to about 2.5 times the risk of NAION compared with non-GLP-1 drugs. The authors rated the certainty as low to moderate.[9]
Trials of all eye problems (JAMA Ophthalmology, 2025). Across 78 randomized trials and 73,640 people, semaglutide did not raise overall eye disorders or diabetic retinopathy. It was linked to NAION (odds ratio 3.92), but with few events and a wide confidence interval.[5]
Data for other drugs in the class, such as dulaglutide and exenatide, are much thinner.[9]
Diabetic retinopathy
In SUSTAIN-6, a 2-year trial of semaglutide 0.5 and 1 mg in 3,297 people with type 2 diabetes and high heart risk, retinopathy complications occurred in 3.0% on semaglutide versus 1.8% on placebo.[6] The Wegovy label explains that the increase was larger in people who already had retinopathy, and that rapid improvement in blood sugar is linked to a temporary worsening of retinopathy. It advises monitoring people with a history of retinopathy.[7] The same trial’s heart results are covered in our GLP-1 heart outcomes guide.
What’s established vs preliminary
Established:
- NAION is a recognized very rare side effect of semaglutide in the EU.[1]
- People with existing diabetic retinopathy need monitoring when starting a GLP-1 drug.[7][8]
Preliminary:
- whether the NAION risk applies equally to people without diabetes
- whether tirzepatide and other GLP-1 drugs share the NAION risk
- how GLP-1 drugs affect retinopathy over many years
Limitations
- NAION is rare, so trials can’t measure it well, and observational studies can’t rule out confounding, such as sleep apnea being more common in people prescribed these drugs.
- Diagnosis codes for NAION in health records can be inaccurate.
- Most data are for semaglutide in type 2 diabetes.[4]
What’s next
FOCUS, a Novo Nordisk trial of 1,500 people with type 2 diabetes, is testing semaglutide versus placebo specifically on diabetic eye disease. It is active but not recruiting, with primary completion expected in November 2027.[10]
Questions to discuss with a clinician
- Do I have diabetic retinopathy, and how often should my eyes be checked after starting?
- Do I have risk factors for NAION, such as sleep apnea? Our Zepbound and sleep apnea guide covers one link.
- Should my dose go up more slowly if my blood sugar is high and retinopathy is present? Our GLP-1 dose conversion chart shows typical dose steps.
- What eye symptoms should prompt an urgent call?
