Purpose: The study aims to develop and validate a novel nomogram for predicting diabetic retinopathy (DR) risk specifically in young and middle-aged patients with type 2 diabetes (T2DM). Methods: This retrospective cohort study analyzed 337 T2DM patients (Age 15– 59 years) admitted to Luoyang Central Hospital from July 2022 to January 2024, stratified by fundus examination into DR (n=155) and non-DR (n=182) groups. Demographic characteristics and relevant clinical parameters were systematically collected. A predictive nomogram for DR detection was constructed using significant variables identified through multivariate logistic regression analysis. The calibration, discrimination, and clinical utility of the nomogram were subsequently evaluated using calibration plots, receiver operating characteristic curves, and decision curves. Results: Multivariate analysis identified four independent predictors of diabetic retinopathy: diabetes duration (OR=1.125, 95% CI: 1.07– 1.182, P < 0.001), brachial-ankle pulse wave velocity (baPWV; OR=1.269, 95% CI: 1.133– 1.421, P < 0.001), blood urea nitrogen (BUN; OR=1.223, 95% CI: 1.052– 1.423, P =0.009), and age (OR=0.955, 95% CI: 0.922– 0.989, P =0.01), with the developed nomogram demonstrating excellent discrimination (AUC=0.75, 95% CI: 0.696– 0.800), significant improvement over individual predictors (ΔAUC+0.05 to+0.18), strong calibration (Bootstrap C-index=0.749), and clinical utility across 2– 85% threshold probabilities by decision curve analysis. Conclusion: This study presents the first nomogram for DR risk in young and middle-aged patients with T2DM. Integrating four routine clinical parameters (diabetes duration, baPWV, BUN, age), the model demonstrates robust predictive power (AUC=0.75) and clinical utility, enabling early risk stratification and timely intervention. Keywords: prediction models, line graphs, brachial-ankle pulse wave velocity, diabetic retinopathy
Ma et al. (Sun,) studied this question.