Diabetic Retinopathy Grading and Concurrent Cardiorenal Biomarker Abnormalities in Type 2 Diabetes: Albuminuria and Elevated NT-proBNP-A Retrospective Cross-Sectional Study.

Background

Diabetic retinopathy (DR) is the most prevalent microvascular complication of type 2 diabetes mellitus (T2DM) and a recognised marker of systemic vascular injury. Whether DR-graded independently by two experienced ophthalmologists as part of routine institutional care-is independently associated with concurrent nephropathy (urine albumin-to-creatinine ratio [UACR] ≥ 30 mg/g) and subclinical cardiac stress (N-terminal pro-B-type natriuretic peptide [NT-proBNP] ≥ 125 pg/mL) remains insufficiently examined.

Methods

Retrospective cross-sectional study of 401 T2DM adults who underwent dilated fundus examination independently graded by two board-certified ophthalmologists (each ≥ 10 years' experience in diabetic eye disease) using ETDRS-based classification. Inter-physician agreement was quantified by Cohen's weighted kappa. The primary composite outcome was concurrent nephropathy and subclinical cardiac stress. Multivariable logistic regression adjusted for age, sex, HbA1c, diabetes duration, eGFR, hypertension, and pharmacological treatments.

Results

The composite outcome was present in 120 (64.2%) DR-positive vs. 14 (6.5%) DR-negative patients (crude OR: 25.59 (13.78-47.51); p < 0.001; fully adjusted OR: 21.49 (11.02-41.81); p < 0.001). Inter-physician agreement was high (Cohen's weighted κ = 0.88; 95% CI: 0.83-0.93). Nephropathy alone was found in 164 (87.7%) vs. 76 (35.5%) patients (OR: 12.95 (7.71-21.75); p < 0.001), and elevated NT-proBNP in 137 (73.3%) vs. 39 (18.2%) patients (OR: 12.29 (7.65-19.76); p < 0.001). A significant trend across DR severity grades was observed (p for trend < 0.001), although the gradient was not strictly monotonic. ROC AUC = 0.837 (95% CI: 0.784-0.860).

Conclusions

Ophthalmologist-graded DR was independently associated with the concurrent presence of albuminuria and elevated NT-proBNP in T2DM. Because the design is cross-sectional, these findings describe association rather than prediction or causation. They generate the hypothesis that retinal grading may help flag patients in whom cardiorenal biomarker assessment is worth considering, but prospective validation is required before any screening application.