Which anti‑diabetic medications are associated with blindness?

No Anti-Diabetic Medications Directly Cause Blindness, But Some May Increase Risk of Specific Optic Nerve Disorders

The newer GLP-1 receptor agonists semaglutide and tirzepatide are associated with increased risk of nonarteritic anterior ischemic optic neuropathy (NAION) and other optic nerve disorders in patients with type 2 diabetes, though the absolute risk remains low (0.04% for NAION over 2 years). 1

Understanding the Context: Diabetic Retinopathy vs. Medication Effects

It's critical to distinguish between diabetic retinopathy (the disease process itself causing blindness) and medication-related optic nerve complications:

• Diabetic retinopathy is the leading cause of blindness in adults aged 20-74 years and results from the diabetes itself, not from medications 2, 3, 4
• The strongest predictors of blindness are duration of diabetes and poor glycemic control, not the medications used to treat it 5
• Intensive glycemic control with any medication reduces retinopathy risk by 54-76% 5

Medications Associated with Optic Nerve Complications

GLP-1 Receptor Agonists (Semaglutide/Tirzepatide)

Recent evidence shows:

• Hazard ratio of 1.76 for NAION (95% CI: 1.01-3.07) 1
• Hazard ratio of 1.65 for other optic nerve disorders (95% CI: 1.18-2.31) 1
• Case reports include sequential ischemic optic neuropathy, bilateral papillitis, and paracentral acute middle maculopathy 6

Important caveats:

• The absolute risk is extremely low (35 cases per 79,699 patients over 2 years) 1
• The mechanism may be rapid correction of hyperglycemia rather than direct drug toxicity 6
• These findings require close monitoring but should not prevent use when indicated

Insulin

Insulin is paradoxically associated with higher rates of clinically significant diabetic retinopathy requiring photocoagulation, but this reflects confounding by indication—patients on insulin have more severe, longer-duration diabetes 7:

• Intermediate-acting human insulin: OR 4.28 (95% CI: 1.69-10.8) 7
• This association reflects disease severity, not causation

Medications That May Actually Protect Against Vision Loss

Metformin

• Associated with lower risk of open-angle glaucoma (OR 0.18; 95% CI: 0.08-0.41) 8
• Cumulative lifetime glaucoma risk reduced to 1.5% vs. 7.2% in non-diabetics 8

DPP-4 Inhibitors

• Associated with reduced risk of primary open-angle glaucoma (aHR 0.53; 95% CI: 0.50-0.56) 9
• Also reduced normal tension glaucoma risk (aHR 0.55; 95% CI: 0.50-0.62) 9

Fenofibrate

• May slow retinopathy progression, particularly with mild nonproliferative diabetic retinopathy at baseline 3
• However, fenofibrate was also associated with incident clinically significant diabetic retinopathy (OR 1.96; 95% CI: 1.36-2.82), likely reflecting confounding by dyslipidemia severity 7

Clinical Approach

For patients starting semaglutide or tirzepatide:

• Obtain baseline dilated eye examination
• Monitor for sudden vision changes, particularly unilateral vision loss
• Educate patients about NAION symptoms (sudden, painless vision loss)
• Consider more gradual titration in patients with pre-existing optic nerve disease

For all diabetic patients:

• Annual dilated eye examinations remain the standard of care 3, 4
• Type 1 diabetes: begin screening 5 years after diagnosis 3, 4
• Type 2 diabetes: screen at diagnosis 3, 4
• Optimal glycemic and blood pressure control are the most effective strategies to prevent vision-threatening complications 2, 3, 10

Critical pitfall: Do not attribute diabetic retinopathy progression to medications when the underlying cause is inadequate glycemic control or disease duration. The medications themselves (except possibly GLP-1 agonists for optic nerve disorders) do not cause blindness—uncontrolled diabetes does.