What is the risk of blindness associated with the use of semaglutide (glucagon-like peptide-1 receptor agonist) in patients with type 2 diabetes, particularly those with pre-existing diabetic retinopathy or other ocular conditions?

Risk of Blindness with Semaglutide

Semaglutide carries a documented risk of worsening diabetic retinopathy complications, particularly in patients with pre-existing retinopathy, but does not increase the overall risk of diabetic retinopathy development and has not been definitively linked to permanent blindness. 1, 2

Diabetic Retinopathy Complications

The primary ocular risk with semaglutide is rapid worsening of pre-existing diabetic retinopathy, not new-onset disease or blindness per se. 1, 2

• In the SUSTAIN-6 cardiovascular outcomes trial, diabetic retinopathy complications occurred in 3.0% of semaglutide-treated patients versus 1.8% with placebo 1
• The absolute risk increase was substantially higher among patients with baseline diabetic retinopathy (8.2% with semaglutide vs 5.2% placebo) compared to those without known retinopathy (0.7% vs 0.4%) 1
• This worsening is attributed to rapid A1C reduction rather than direct drug toxicity—a well-established phenomenon with any aggressive glycemic control 2
• Meta-analyses confirm no association between GLP-1 receptor agonists and retinopathy development except through the mechanism of rapid A1C reduction at 3-month and 1-year follow-up 2

Nonarteritic Anterior Ischemic Optic Neuropathy (NAION)

• A 2025 meta-analysis found semaglutide associated with increased odds of NAION (OR 3.92,95% CI 1.02-15.02), though trial sequential analysis indicated insufficient sample size for definitive conclusions 3
• Current evidence remains inadequate to establish a causal relationship between semaglutide and NAION 3

Risk Stratification and Pre-Treatment Requirements

All patients must undergo comprehensive dilated eye examination before initiating semaglutide if not performed within the last 12 months. 2, 4, 1

Highest Risk Patients:

• Pre-existing proliferative diabetic retinopathy 4, 1
• Baseline A1C >9% with concurrent insulin therapy 4
• Diabetes duration ≥10 years 5
• Age ≥60 years 5

Lower Risk Patients:

• No baseline retinopathy 1
• Well-controlled glycemia at baseline 4

Risk Mitigation Strategies

For patients with established retinopathy, slow the rate of A1C reduction through gradual dose titration and concomitant medication adjustment. 4, 2

• Reduce total daily insulin dose by 20% when initiating semaglutide to slow glycemic improvement 4
• For patients on sulfonylureas with well-controlled baseline A1C, reduce sulfonylurea dose by 50% 4
• Start at the lowest semaglutide dose and up-titrate slowly, particularly in patients with established retinopathy 4
• Implement strategies to optimize blood pressure and lipid control, which independently reduce retinopathy progression 2, 4

Monitoring Protocol

• Baseline comprehensive dilated eye examination before treatment initiation 4, 6
• Consider more frequent ophthalmologic monitoring (every 6 months rather than annually) in high-risk patients 4
• If any level of diabetic retinopathy is present at baseline, repeat dilated retinal examinations at least annually 6
• If retinopathy is progressing or sight-threatening, more frequent examinations by an ophthalmologist are required 6

Clinical Decision Algorithm for Proliferative Diabetic Retinopathy

For patients with proliferative diabetic retinopathy, consider alternative GLP-1 receptor agonists or other drug classes first, such as SGLT2 inhibitors for cardiovascular risk reduction. 4

• If cardiovascular disease is present and cardiovascular benefit is the primary goal, the 13% reduction in major adverse cardiovascular events with semaglutide may outweigh retinopathy risks in many patients 4
• Document pre-treatment ophthalmologic assessment and informed consent discussion about retinopathy worsening risks 4
• Ensure close collaboration with ophthalmology to grade baseline retinopathy severity before initiating therapy 7

Reassuring Evidence

• Large-scale real-world data from 810,390 semaglutide users showed no increased risk of proliferative diabetic retinopathy compared to dulaglutide, empagliflozin, or sitagliptin 8
• Risk of treatment-requiring diabetic retinopathy/diabetic macular edema was actually lower with semaglutide compared to dulaglutide, sitagliptin, and glipizide 8
• Overall meta-analysis of 22,096 patients found semaglutide was not associated with increased diabetic retinopathy risk when all trials were combined (RR 1.14,95% CI 0.98-1.33) 5
• GLP-1 receptor expression is low in normal human eyes and was not detected in eyes with advanced proliferative diabetic retinopathy, suggesting limited direct ocular drug effect 9

Critical Caveat

Do not initiate semaglutide in uncontrolled diabetic patients whose last eye examination was more than 12 months ago until a comprehensive dilated eye examination is performed. 4, 2