Development and validation of a nomogram for predicting stroke risk in patients with hypercholesterolemia: a study based on NHANES

Introduction We sought to identify independent risk factors for stroke in patients with hypercholesterolemia and develop a clinically applicable risk prediction nomogram using a nationally representative dataset. Methods Data from 1259 participants with hypercholesterolemia from the US National Health and Nutrition Examination Survey 2017 to 2018 and 2019 to March 2020 were analyzed. Participants were randomly assigned into training and validation sets in a 7:3 ratio (training set, n = 881; validation set, n = 378). Least absolute shrinkage and selection operator regression combined with multivariate logistic regression were employed to identify independent predictors and construct a nomogram model. The model was validated through receiver operating characteristic curve analysis, calibration curves, and decision curve analysis. Results Four independent predictors were identified: age, total cholesterol level, serum urea nitrogen level, and white blood cell count. The nomogram demonstrated excellent discriminatory capability, with areas under the curve of 0.811 (95% CI, 0.769-0.853) and 0.835 (95% CI, 0.765-0.902) in the training and validation sets, respectively. Calibration curves revealed high concordance between predicted probabilities and observed outcomes. Decision curve analysis confirmed substantial clinical utility across practical threshold probability ranges. Discussion We developed a novel nomogram for stroke risk prediction in patients with hypercholesterolemia that demonstrates robust performance and may facilitate personalized risk stratification and intervention strategies in clinical practice.