Categories
Change Password!
Reset Password!
Coronary heart disease (CHD) and ischemic stroke (IS) share common risk factors and mechanisms, underscoring the need for novel biomarkers.
Elevated glucose-to-albumin ratio is a strong, independent indicator of stroke risk in coronary heart disease patients.
Coronary heart disease (CHD) and ischemic stroke (IS) share common risk factors and mechanisms, underscoring the need for novel biomarkers. This retrospective study estimated the glucose-to-albumin ratio (GAR) as a predictive marker for IS in CHD patients.
A total of 1,885 CHD sufferers were enrolled. Feature selection was executed via the Boruta algorithm. Multilayer perceptron model was developed to predict IS risk. Model performance was assessed through receiver operating characteristic curves and calibration plots, while shapley additive explanations (SHAP) values and partial dependence plots (PDP) provided interpretability to key predictors.
Patients with IS were older and had higher rates of hypertension and diabetes when compared to those without acute IS. Furthermore, the acute IS sufferers showed a greater prevalence of right coronary artery involvement and triple-vessel disease. The GAR was markedly elevated in the IS group, emerging as a key biomarker alongside hyperlipidemia and a history of hypertension identified through the Boruta algorithm. The SHAP analysis highlighted GAR as a strong risk determinant, and PDP analysis further validated its role as an independent predictor of IS, emphasizing its clinical relevance for risk stratification in CHD patients.
GAR proved to be an effective independent predictor of IS risk in patients with CHD, with elevated GAR levels showing a strong association with increased stroke risk.
BMC Cardiovascular Disorders
Association between glucose-to-albumin ratio and ischemic stroke risk in patients with coronary heart disease: a machine learning-based predictive model analysis
Ling Hou et al.
Comments (0)