OBJECTIVE

To analyze the risk and influencing factors for coronary heart disease (CHD) in patients with diabetes (DM), and to develop and validate a nomogram prediction model, providing a basis for the early diagnosis and individualized intervention in patients with DM and CHD.

METHODS

This study was based on data from the National Health and Nutrition Examination Survey (NHANES). A total of 2,141 diabetic patients from 2011 to 2020 were included, randomly divided into a training set (n = 1,499) and a validation set (n = 642) at a 7:3 ratio. The least absolute shrinkage and selection operator (Lasso) regression analysis was used to screen risk factors, and a multivariate logistic regression model was developed to construct the DM-CHD nomogram prediction model. Model performance was internally validated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).The Hosmer-Lemeshow Test was utilized to evaluate the overall goodness of fit of the nomogram.

RESULTS

Univariate analysis identified 15 factors as risk factors for DM-related CHD. Lasso regression further selected 7 key predictors: Age (OR 1.06, CI 1.05-1.08, P < 0.001), Gender (OR 0.47, CI 0.36-0.63, P < 0.001), Hypertension (OR 1.85, CI 1.33-2.57, P < 0.001), Weight Adjusted Waist Index (OR 1.50, CI 1.25-1.81, P < 0.001), Neutrophils (OR 1.09, CI 1.02-1.17, P = 0.009), Platelets (OR 0.99, CI 0.99-0.99, P < 0.001), and Triglycerides (OR 1.18, CI 1.08-1.30, P < 0.001). The area under the ROC curve (AUC) for the nomogram model was 0.758 (95% CI 0.728-0.789) in the training set and 0.747 (95% CI 0.699-0.796) in the validation set. Calibration curves and DCA indicated that the model exhibited satisfactory predictive performance. The model's reliability and clinical net benefit were further validated.

CONCLUSION

The nomogram model developed in this study, based on multiple clinical indicators (Age, Gender, Hypertension, Weight Adjusted Waist Index, Neutrophils, Platelets, and Triglycerides), demonstrated adequate calibration and clinical net benefit in the validation cohort. The model demonstrates moderate but clinically useful discrimination ability, providing scientific guidance for early diagnosis and personalized interventions in patients with DM complicated by CHD, and may help reduce CHD risk in diabetic patients.