Retinopathy Signs Improved Prediction and Reclassification of Cardiovascular Disease Risk in Diabetes: A prospective cohort study

Area of Science:

• Ophthalmology
• Cardiology
• Diabetology

Background:

• Cardiovascular disease (CVD) risk prediction in diabetic patients is suboptimal, as current models are primarily based on general population data.
• Diabetic retinopathy and retinal vascular caliber are potential indicators of systemic vascular health.

Purpose of the Study:

• To investigate if incorporating retinopathy and retinal vascular caliber measurements enhances CVD risk prediction in individuals with diabetes.
• To compare the predictive value of retinal parameters against established CVD risk factors and specific biomarkers.

Main Methods:

• Utilized data from the Singapore Malay Eye Study (SiMES) and Singapore Prospective Study Program (SP2).
• Developed Cox regression models incorporating established CVD risk factors, serum biomarkers (eGFR, hsCRP), and retinal parameters (retinopathy, arteriolar/venular caliber).
• Validated the models internally within SiMES and externally in SP2.

Main Results:

• Addition of retinal parameters significantly improved CVD risk prediction discrimination compared to adding biomarkers like eGFR and hsCRP alone.
• Combining retinal parameters with biomarkers further enhanced model performance (C-statistic increased from 0.721 to 0.774, p=0.013) and reclassification (NRI=17.0%, p=0.004).
• External validation confirmed these findings, showing consistent improvements in discrimination and reclassification.

Conclusions:

• Retinopathy and retinal microvascular parameters provide significant incremental value for CVD risk reclassification in diabetic individuals.
• These retinal indicators improve upon predictions made using established risk factors alone.
• Integrating retinal imaging into risk assessment could optimize CVD prevention strategies for the diabetic population.