New AI-Driven Blood Test Predicts Diabetic Retinal Damage Early

A groundbreaking study published in PLOS Medicine introduces a predictive model, Pro-DRN, capable of identifying diabetic retinal neurodegeneration (DRN) before clinical symptoms manifest. DRN is a serious complication of diabetes that leads to vision loss and is often considered a broader indicator of systemic neural degradation, including cognitive decline and peripheral nerve damage. Currently, medical intervention for DRN is reactive, occurring only after irreversible damage has already taken hold.

Researchers developed this tool by analyzing blood plasma from over 1,400 patients, identifying 71 specific proteins linked to inflammation and cellular maintenance. By utilizing machine learning to process these protein levels, the Pro-DRN model successfully predicted the onset of retinal nerve damage with 26% greater accuracy than existing methods. This shift from reactive to proactive diagnostics represents a significant advancement in how clinicians can monitor patients with type 2 diabetes.

The implications for patient care are substantial. By integrating this blood test into routine check-ups, healthcare providers could identify high-risk individuals years earlier than currently possible. This early warning system would allow for timely interventions—ranging from lifestyle modifications to neuroprotective therapies—potentially preserving vision and preventing broader neurological complications. As the Pro-DRN model is now available for clinical use, it marks a promising step toward personalized, preventative care for the millions of people living with diabetes worldwide.