Baroreceptor Sensitivity and Its Association with Diabetic Retinopathy Severity

Abstract

Diabetic retinopathy (DR) is the leading cause of preventable blindness in the working-age population, affecting hundreds of millions of individuals with diabetes worldwide. While glycemic control and retinal imaging remain the cornerstone of DR management, the role of the autonomic nervous system — particularly baroreceptor sensitivity (BRS) — in modulating DR severity has received limited systematic investigation. Baroreceptors are specialized mechanosensitive receptors in the aortic arch and carotid sinuses that regulate cardiovascular homeostasis through the baroreflex arc. Their progressive impairment in diabetes reflects underlying cardiovascular autonomic neuropathy (CAN), a condition driven by the same pathophysiological processes — oxidative stress, endothelial dysfunction, and advanced glycation end-products — that govern retinal microvascular damage. This paper presents a prospective, cross-sectional study involving 320 participants (80 healthy controls and 240 type 2 diabetic patients stratified across three DR severity grades) and introduces a novel AI-integrated diagnostic framework that jointly analyzes dual-method BRS indices and quantitative retinal vascular parameters. Using spectral BRS analysis (alpha-index) and the sequence method alongside retinal vascular biomarkers including the arteriolar-to-venular ratio, vessel fractal dimension, and central retinal arteriolar equivalent, a composite BRS-Retinal Vascular Index (BRVI) was formulated. A soft-voting ensemble of Random Forest, Support Vector Machine, and XGBoost classifiers achieves 93.4% accuracy, 91.8% sensitivity, 94.1% specificity, and AUC-ROC of 0.971 in four-class DR severity prediction. An IoT-cloud architecture enabling continuous outpatient BRS monitoring with automated clinical alerting is also proposed. Results confirm a statistically significant inverse correlation between BRS magnitude and DR severity (r = -0.78, p < 0.001), positioning BRS as a clinically accessible and complementary biomarker for DR stratification and early intervention.
Keywords — Baroreceptor Sensitivity, Diabetic Retinopathy, Autonomic Dysfunction, Machine Learning, Retinal Vascular Biomarker, IoT Health Monitoring, Cardiovascular Autonomic Neuropathy, Spectral Analysis, BRVI.