Rishi Singh, MD, hosted Andrew J. Barkmeier, MD, professor of ophthalmology at the Mayo Clinic, on Evidence Based Retina to discuss whether newer glucose-lowering medications affect diabetic retinopathy risk.
Dr. Barkmeier has published extensively on GLP-1 receptor agonists and SGLT2 inhibitors, including a large comparative database study built to answer the question directly.
“These are incredibly powerful medications—they're literally life-saving medications. Patients taking them have a 13% lower risk of all-cause death.”
Glucose-lowering medications and diabetic retinopathy fast facts
- GLP-1 receptor agonists and SGLT2 inhibitors reduce major adverse cardiovascular events—13% with liraglutide in the LEADER trial and 14% with empagliflozin in EMPA-REG OUTCOME—in patients with type 2 diabetes at high cardiovascular risk.1,2
- The 2016 SUSTAIN-6 trial found semaglutide was associated with a 76% higher relative risk of retinopathy complications versus placebo (3.0% vs. 1.8%; hazard ratio 1.76; P=0.02).3
- Barkmeier et al. analyzed 371,698 patients across four diabetes drug classes using the Optum Labs database, weighting groups statistically to emulate an idealized target trial.4
- GLP-1 receptor agonists showed no increased risk of treatment-requiring proliferative diabetic retinopathy or diabetic macular edema at short or long durations of use.4
- SGLT2 inhibitors showed a 27% lower risk of sight-threatening retinopathy versus GLP-1 agents, 21% lower than DPP-4 inhibitors, and 39% lower than sulfonylureas.4
- Early retinopathy worsening with rapid glycemic improvement was documented in the Diabetes Control and Complications Trial (DCCT; 1982-1993): 13.1% of patients on intensive insulin therapy versus 7.6% on conventional treatment.5
What the comparative effectiveness data show
Dr. Barkmeier calls the SUSTAIN-6 finding a “record scratch” moment: Semaglutide carried a 76% higher relative risk of retinopathy complications than placebo—50 patients (3.0%) versus 29 (1.8%).3 Endocrinologists, primary care doctors, and patients started calling retina specialists directly.
But SUSTAIN-6 wasn't built to answer the ophthalmic question well—its retinal outcomes came from a mix of clinical exams or fundus photographs from local ophthalmologists, optometrists, or other healthcare professionals, rather than a standardized reading center protocol, so Dr. Barkmeier's group set out to answer that question using a much larger real-world dataset.
Using the Optum Labs administrative database, they tracked 371,698 adults with type 2 diabetes and moderate cardiovascular risk who started a GLP-1 receptor agonist, SGLT2 inhibitor, DPP-4 inhibitor, or sulfonylurea between 2014 and 2021, weighting the groups statistically to emulate an idealized target trial.4
The primary outcome was blunt and clinically meaningful: new treatment for proliferative diabetic retinopathy or diabetic macular edema. GLP-1 use showed no increased risk of treatment-requiring retinopathy compared with older drug classes, at either short or long durations of use.4
Additional findings from the study
SGLT2 inhibitors performed better than everything else in the analysis—a 27% lower risk than GLP-1 agents, 21% lower than DPP-4 inhibitors, and 39% lower than sulfonylureas.4 Dr. Barkmeier's group also compared individual GLP-1 drugs against each other and found semaglutide carried no elevated risk relative to the rest of its class.5
“Sometimes we'll see a patient six months later and they're almost unrecognizable—major health changes. We will often watch those patients a bit more closely.”
The early-worsening phenomenon Dr. Barkmeier references is real and well documented, and it's separate from long-term risk. In the original DCCT, 13.1% of patients on intensive insulin therapy had early retinopathy worsening within a year, compared with 7.6% on conventional treatment—a pattern associated with rapid, large drops in glycated hemoglobin, not by drug class.6
His advice is simple:
“You may consider watching patients more closely after a new GLP-1 prescription, bariatric surgery, or pregnancy—any major systemic change. Don't avoid the medications.”
Key takeaways
- Dr. Barkmeier's real-world analysis of more than 371,000 patients found no increased risk of treatment-requiring diabetic retinopathy with GLP-1 receptor agonists compared with older diabetes drug classes.4
- SGLT2 inhibitors showed a lower risk of sight-threatening retinopathy than GLP-1 agents, DPP-4 inhibitors, and sulfonylureas alike.4
- The SUSTAIN-6 signal that raised early concern was likely driven by rapid glycemic improvement, a well-documented phenomenon independent of GLP-1 drugs specifically.3
- Patients undergoing rapid systemic health changes, whether from a new diabetes medication, bariatric surgery, or pregnancy, warrant closer retinal monitoring regardless of drug class.
- Retina specialists can reassure endocrinologists and primary care colleagues that GLP-1 and SGLT2 medications carry substantial systemic benefits without an added retinopathy tradeoff.
Final thoughts
The evidence has moved a long way since SUSTAIN-6 raised the alarm in 2016. For more on how Dr. Barkmeier's team built their analysis and what it means for patients starting these medications, watch the full conversation.
