Diabetic Retinopathy Risk

The relationship between tirzepatide and diabetic retinopathy is nuanced and depends critically on the patient’s baseline retinal status. Two seemingly contradictory messages from recent evidence can both be true simultaneously: tirzepatide may worsen retinopathy in patients with moderate-to-severe pre-existing NPDR, while simultaneously protecting against new-onset retinopathy in patients whose eyes are currently healthy.

Early worsening phenomenon: Early worsening of diabetic retinopathy (EWDR) with rapid glucose lowering is a well-documented class phenomenon, previously described with intensive insulin therapy, semaglutide, and other potent glucose-lowering agents. The proposed mechanism is that rapid HbA1c reduction creates a sudden change in retinal blood flow regulation that can destabilise moderately compromised retinal vasculature. A 2025 retrospective cohort study (PMID 40637847, n=3435 matched pairs) found that tirzepatide was associated with significantly increased odds of new-onset proliferative diabetic retinopathy (PDR) in patients who had moderate-to-severe NPDR (R2M0/R2M1) or mild NPDR with maculopathy (R1M1) at treatment initiation: OR 2.15 (95% CI 1.24–3.74), P<0.01. The absolute rate was 1.1% vs 0.5% — a 0.6 percentage point increase. Notably, this study population had relatively tight baseline glycaemic control (mean HbA1c 7.28%), suggesting the worsening risk is not solely confined to patients with previously poor control.

Overall ocular benefit: In contrast, a much larger retrospective study (PMID 41655764, n=102,590 matched patients, US nationwide EHR, 36-month follow-up) found that tirzepatide was associated with significantly lower risks of diabetic retinopathy (HR 0.79), diabetic macular oedema (HR 0.82), vitreous haemorrhage/retinal detachment (HR 0.66), need for vision-saving interventions (HR 0.65), and non-arteritic anterior ischaemic optic neuropathy (HR 0.45) compared to non-GIP GLP-1 receptor agonists. FAERS pharmacovigilance data (PMID 39141075) found no significantly increased DR risk vs GLP-1 RAs overall.

Patients without pre-existing retinopathy (R0M0): the matched cohort study showed tirzepatide was actually PROTECTIVE — OR 0.73 (95% CI 0.62–0.86) for new-onset retinopathy, consistent with better long-term glycaemic and cardiometabolic control reducing microvascular disease burden.

Practical implication: The key clinical decision point is baseline retinal status. Patients with moderate-to-severe NPDR who are about to start tirzepatide (or any potent glucose-lowering intensification) should have ophthalmological review and monitoring before and shortly after starting treatment. The FDA prescribing information for tirzepatide and GLP-1 RAs notes this risk. Patients with no retinopathy should have reassurance that tirzepatide likely reduces their long-term retinopathy risk.

Ontology Diabetic Retinopathy Risk [warns_about] Tirzepatide Diabetic Retinopathy Risk [relates] GLP-1 Receptor Agonism Diabetic Retinopathy Risk [relates] HbA1c Reduction Diabetic Retinopathy Risk [relates] Type 2 Diabetes

Numbers

PopulationOutcomeTirzepatideComparatorEffectSource
Moderate-severe NPDR (R2) at baselineNew-onset PDR1.1%0.5% (unexposed)OR 2.15 (1.24–3.74)PMID 40637847
No pre-existing DR (R0M0)New-onset DRLowerOR 0.73 (0.62–0.86)PMID 40637847
T2D patients (vs GLP-1 RA)DR riskLowerGLP-1 RAHR 0.79 (0.72–0.87)PMID 41655764
T2D patients (vs GLP-1 RA)DME riskLowerGLP-1 RAHR 0.82 (0.68–0.96)PMID 41655764
T2D patients (vs GLP-1 RA)VH/RDLowerGLP-1 RAHR 0.66 (0.45–0.95)PMID 41655764

Practical Interpretation

  • Who appears most at risk: Patients with moderate-to-severe NPDR (R2) or mild NPDR with maculopathy (R1M1) — early worsening risk on starting tirzepatide
  • Who appears protected: Patients without pre-existing DR (R0M0); long-term overall ocular outcomes better than GLP-1 RA comparators
  • Severity: PDR is sight-threatening; warrants specialist ophthalmological referral by ETDRS criteria
  • Mitigation: Retinal screening before starting; specialist review for moderate-severe NPDR; consider slower titration or alternative agent in high-risk retinopathy patients; monitor early after initiation

Connections

Sources