What are the recommendations for a patient with a history of diabetes taking GLP-1 (Glucagon-like peptide-1) receptor agonists, such as liraglutide (Victoza) or semaglutide (Ozempic), to minimize the risk of blindness?

GLP-1 Receptor Agonists and Blindness Risk

GLP-1 receptor agonists, particularly semaglutide, are associated with diabetic retinopathy complications related to rapid A1C reduction, but recent real-world evidence demonstrates they actually reduce sight-threatening complications and blindness by 23% in patients with established retinopathy. 1, 2

Pre-Treatment Ophthalmologic Assessment

All patients must undergo comprehensive dilated eye examination before initiating any GLP-1 RA if not performed within the last 12 months. 3, 1 This is non-negotiable and serves to:

• Identify pre-existing diabetic retinopathy, especially proliferative disease 1
• Document baseline retinopathy status for risk stratification 1
• Enable informed consent discussion about potential retinopathy worsening 1

Do not initiate GLP-1 therapy in patients whose last eye exam was more than 12 months ago until current examination is completed. 1

Understanding the Retinopathy Risk Mechanism

The retinopathy complications associated with semaglutide are not due to direct drug toxicity but rather to the rapidity of A1C reduction. 1, 4 This phenomenon occurs with any aggressive glycemic control strategy. 1

• Meta-analyses show no association between GLP-1 RAs and retinopathy except through rapid A1C reduction at 3-month and 1-year follow-up 1, 4
• The worsening is typically transient and occurs early in treatment 1
• Patients with poor baseline glycemic control (HbA1c >9%) and concurrent insulin therapy face highest risk 1

Risk Mitigation Strategies During Initiation

For Patients with Pre-Existing Proliferative Retinopathy

Consider alternative GLP-1 RAs or SGLT2 inhibitors first in patients with proliferative diabetic retinopathy. 1 If semaglutide is chosen despite proliferative retinopathy:

• Start at the lowest dose (0.25 mg weekly for semaglutide) 1
• Titrate more slowly than standard protocol, extending each dose level beyond the typical 4 weeks 1
• Reduce concurrent insulin dose by 20% when starting to slow glycemic improvement 1
• Reduce sulfonylurea dose by 50% in well-controlled patients 1

For All Patients Initiating GLP-1 RAs

Implement slower dose titration in patients with established retinopathy of any severity. 1, 4 The standard escalation protocol should be extended to minimize rapid A1C drops. 1

Optimize blood pressure and lipid control independently, as these reduce retinopathy progression regardless of glucose control. 1, 4

Monitoring Protocol

Patients Without Baseline Retinopathy

• If no retinopathy present and glycemia well-controlled: screening every 1-2 years 4
• Annual dilated examinations remain standard for type 2 diabetes 4

Patients With Pre-Existing Retinopathy

Increase ophthalmologic monitoring frequency to every 6 months rather than annually in high-risk patients. 1 This includes:

• Patients with any level of pre-existing diabetic retinopathy 4
• Patients with baseline HbA1c >9% 1
• Patients on concurrent insulin therapy 1

If retinopathy is progressing or sight-threatening, refer immediately to ophthalmology for more frequent examinations. 4

Contradictory Evidence and Clinical Context

While earlier trials (SUSTAIN-6) showed increased diabetic retinopathy complications with semaglutide 3, 5, the most recent 2025 real-world cohort study of 185,066 patients demonstrates GLP-1 RAs reduce sight-threatening complications: 2

• 26% reduction in vitreous hemorrhages (HR 0.74) 2
• 22% reduction in neovascular glaucoma (HR 0.78) 2
• 23% reduction in blindness (HR 0.77) 2

This protective effect occurred even in the 32,695 patients with pre-existing diabetic retinopathy. 2 The modest 7% increase in incident diabetic retinopathy (HR 1.07) was offset by substantial reductions in vision-threatening outcomes. 2

Risk-Benefit Decision Algorithm

For patients with established cardiovascular disease, the 13% reduction in major adverse cardiovascular events outweighs optic nerve risks in most clinical scenarios. 1 The decision framework:

1. If no proliferative retinopathy: Proceed with GLP-1 RA after baseline eye exam, using standard titration 1

2. If proliferative retinopathy present: Consider SGLT2 inhibitor first, or if GLP-1 RA chosen, use slower titration with reduced insulin/sulfonylurea doses 1

3. If cardiovascular disease present: GLP-1 RA benefits likely outweigh retinopathy risks given protective effects against blindness 1, 2

Critical Pitfalls to Avoid

Never initiate GLP-1 therapy without current ophthalmologic assessment in patients whose last exam exceeds 12 months. 1 This is the most common error in clinical practice.

Do not use aggressive titration schedules in patients with established retinopathy or HbA1c >9%. 1 The rapid A1C reduction is the mechanism of harm, not the medication itself.

Document the pre-treatment ophthalmologic assessment and informed consent discussion about retinopathy risks in the medical record. 1 This is both a clinical and medicolegal requirement.