Semaglutide and diabetic retinopathy: what real-world data shows

Concerns about GLP-1 medications and vision complications surfaced in 2023 when diabetic retinopathy signals appeared in some clinical trial populations. The concern was specific to semaglutide - early data suggested the drug might increase the risk of retinopathy worsening in people with pre-existing diabetes. A massive real-world analysis published in BMJ Open Diabetes Research & Care in November 2025 examined 810,390 semaglutide users across 14 databases (administrative claims and electronic health records) and found no increased risk for either proliferative diabetic retinopathy or diabetic macular edema. The reassuring real-world data contrasts with some earlier clinical trial signals, and suggests the underlying risk profile is much lower than initial concerns indicated.

TL;DR

An OHDSI network analysis of 810,390 semaglutide users found no elevated risk of proliferative diabetic retinopathy or treatment-requiring diabetic macular edema compared to other GLP-1 drugs or non-GLP-1 diabetes medications. Incidence rates in semaglutide users were similar or lower than comparators. This real-world evidence contrasts with some clinical trial signals and provides reassurance on vision safety.

The concern: where the retinopathy signal came from

The SUSTAIN-6 trial of semaglutide in type 2 diabetes, published in 2016, reported a higher rate of diabetic retinopathy complications versus placebo (HR 1.76). The PIONEER-6 trial of oral semaglutide also suggested elevated retinopathy development. These findings were concerning enough that they made it into FDA labelling and drove clinical caution about semaglutide in people with established diabetes, particularly those with pre-existing retinopathy.

However, these signals were inconsistent with GLP-1 theory - GLP-1 medications lower blood sugar and reduce systemic inflammation, both of which should theoretically protect retinas. Mechanistically, the signal did not make sense.

What the real-world data revealed

The OHDSI network study, led by Cindy Xinji Cai at Johns Hopkins with 60+ collaborators across multiple institutions, analysed 810,390 semaglutide users across 14 electronic medical record and administrative claims databases from December 2017 through December 2023.

Primary outcomes:

• Proliferative diabetic retinopathy (PDR) - the sight-threatening form of eye disease
• Treatment-requiring diabetic retinopathy or diabetic macular edema (DR/DME)

Comparators: Semaglutide was compared to other GLP-1 drugs (dulaglutide, liraglutide) and non-GLP-1 medications (empagliflozin SGLT2 inhibitor, sitagliptin DPP-4 inhibitor, glipizide sulfonylurea).

Key findings:

• PDR: No significantly elevated risk comparing semaglutide to dulaglutide (HR 0.81, p=0.51), empagliflozin (HR 0.83, p=0.41), sitagliptin (HR 0.83, p=0.57), or glipizide (HR 0.59, p=0.01 - actually lower with semaglutide)
• Treatment-requiring DR/DME: Semaglutide showed similar or lower risk versus most comparators
• Incidence rates in semaglutide users: PDR 11.2 per 100,000 persons (22.6 per 100,000 person-years); treatment-requiring DR/DME 5.4 per 100,000 persons (11.0 per 100,000 person-years)

The authors concluded: "We did not identify increased risk for either PDR or treatment-requiring DR/DME comparing semaglutide with other GLP-1RAs or non-GLP-1RAs."

Why clinical trials and real-world data diverged

The discrepancy between the SUSTAIN-6/PIONEER-6 clinical trial signals and this real-world safety analysis is worth examining. Several mechanisms could explain it:

Trial selection bias

Patients enrolled in clinical trials are selected for specific characteristics - often including people with longer diabetes duration and more advanced complications. The trial populations may have had higher baseline retinopathy risk than the real-world semaglutide-treated population broadly.

Rapid glycaemic control effect

Paradoxically, very rapid improvements in blood sugar control can transiently worsen retinopathy in people with pre-existing disease - a phenomenon called "rebound retinopathy." This occurs because the sudden normalisation of glucose creates osmotic stress in the retina. This effect is temporary and resolves, but it could have been captured in clinical trials as a signal without indicating a long-term safety problem.

Real-world adherence and dosing

Clinical trial participants receive intensive management and standardised dosing. Real-world semaglutide use is more variable - some patients may titrate more slowly, some may not reach maximal doses. The real-world population may include individuals with milder disease or shorter diabetes duration, diluting the retinopathy signal if it is specific to high-risk subgroups.

What the evidence now supports

With this large real-world analysis, the evidence base on semaglutide and retinopathy has shifted. The current picture is:

• Real-world data (810,000+ users) shows no increased retinopathy risk
• Clinical trial signals (SUSTAIN-6, PIONEER-6) were observed in specific trial populations but have not been confirmed in broader real-world populations
• Mechanistically, semaglutide's glucose control and anti-inflammatory effects should be protective against retinopathy
• Monitoring for retinopathy changes remains important (as with any diabetes medication), but semaglutide should not be avoided based on retinopathy concerns

What you should do if you are on semaglutide and have diabetes

• Annual eye exams with an optometrist or ophthalmologist remain essential for anyone with diabetes, regardless of GLP-1 use
• If you have pre-existing diabetic retinopathy, discuss semaglutide safety explicitly with your ophthalmologist, but the real-world data suggests it is safe
• If your blood sugar control improves rapidly on semaglutide, transient blurring or worsening of retinopathy can occur (rebound effect) - this is temporary and should be reported to your eye care provider but is not a reason to stop the medication
• Maintaining excellent blood sugar control (which semaglutide supports) is the most important modifiable risk factor for preventing retinopathy progression

Eye health and nutrient support on GLP-1

Several nutrients support retinal health and are depleted by reduced food intake on GLP-1 medications: lutein and zeaxanthin (carotenoids that protect the macula), zinc (structural component of retinal proteins), and vitamin C (antioxidant that protects against light-induced retinal damage). While GLP-1 Shield does not contain high-dose targeted eye nutrients, ensuring baseline micronutrient sufficiency supports the foundational health that protects eyes during diabetes management.

GLP-1 Shield is formulated around the nutrient gaps that GLP-1 medications create, including the micronutrients that support tissue health and antioxidant defence.

Frequently asked questions

Does semaglutide cause diabetic retinopathy?

Real-world data from 810,390 semaglutide users found no increased retinopathy risk compared to other diabetes medications. Earlier clinical trials (SUSTAIN-6, PIONEER-6) reported retinopathy signals, but these were not confirmed in broader real-world populations. Current evidence supports semaglutide's safety for vision.

Is it safe to use Ozempic if I already have diabetic retinopathy?

Yes, based on current evidence. The real-world analysis found no increased risk in semaglutide users with pre-existing retinopathy. If you have retinopathy, discuss semaglutide with your ophthalmologist, but the safety data does not contraindicate its use.

Why did the clinical trials show a retinopathy signal?

The mechanisms are unclear. Possible explanations include: trial populations had higher baseline retinopathy risk; transient retinopathy worsening from rapid blood sugar improvement (rebound effect); or differences between trial and real-world populations. The signal has not been reproduced in real-world data.

Should I get eye exams more frequently on a GLP-1 medication?

No change from standard diabetes recommendations is needed based on current evidence. Annual eye exams for anyone with diabetes remain appropriate. If your blood sugar control improves rapidly, report any vision changes to your ophthalmologist, but this is not specific to GLP-1 medications.