单核细胞/高密度脂蛋白胆固醇比值联合系统炎症反应指数对老年2型糖尿病合并急性冠脉综合征患者心力衰竭再入院风险预测模型的价值

The predictive value of the monocyte-to-high density lipoprotein cholesterol ratio combined with systemic inflammation response index for heart failure readmission risk in elderly patients with type 2 diabetes mellitus complicated with acute coronary syndrome

摘要:

目的]探讨单核细胞/高密度脂蛋白胆固醇比值(MHR)联合系统炎症反应指数(SIRI)预测老年2型糖尿病(T2DM)合并急性冠脉综合征(ACS)患者心力衰竭(HF)再入院风险的价值,并构建风险预测模型。 [方法]采用回顾性队列研究方法,选取2023年1月—2025年10月于内蒙古医科大学附属医院住院治疗的老年T2DM合并ACS患者775例。以出院后至数据提取截止日期间本院首次因HF再入院为结局事件,按7∶3随机分为训练集542例和测试集233例。采用LASSO回归联合多因素Logistic回归筛选独立预测因子,构建风险预测模型,并采用受试者工作特征(ROC)曲线、校准曲线及决策曲线分析评价模型效能,绘制列线图。采用SHAP方法分析各预测因子对模型的贡献。 [结果]共筛选出9个独立预测因子,分别为性别、首次支架植入史、二次支架植入史、血小板计数(PLT)、低密度脂蛋白胆固醇(LDLC)、球蛋白(Glo)、左心室射血分数(LVEF)、MHR和SIRI。多因素Logistic回归分析显示,MHR水平升高(OR＝4.464,95%CI:2.972～6.704,P＜0.001)和SIRI水平升高(OR＝1.360,95%CI:1.015～1.823,P＝0.039)与HF再入院风险增加独立相关,而LVEF升高(OR＝0.320,95%CI:0.233～0.439,P＜0.001)与HF再入院风险降低独立相关。Logistic回归模型在测试集中的AUC为0.917,表现出良好的区分度和校准度。SHAP分析显示,MHR对模型预测贡献最大,其次为LVEF。 [结论]MHR联合SIRI可有效预测老年T2DM合并ACS患者HF再入院风险,基于Logistic回归构建的列线图具有较好的预测效能和潜在临床应用价值。

Abstract:

Aim To evaluate the value of the monocyte-to-high density lipoprotein cholesterol ratio (MHR) combined with systemic inflammation response index (SIRI) in predicting the risk of heart failure (HF) readmission in elderly patients with type 2 diabetes mellitus (T2DM) and acute coronary syndrome (ACS), and to establish a risk prediction model. Methods In this retrospective cohort study, 775 elderly inpatients with T2DM and ACS hospitalized in the Affiliated Hospital of Inner Mongolia Medical University from January 2023 to October 2025 were enrolled. The primary endpoint was the first readmission for HF in our hospital from discharge to the data cutoff date. All patients were randomly divided into a training set (n＝542) and a test set (n＝233) at a ratio of 7∶3. LASSO regression followed by multivariate Logistic regression was used to identify independent predictors and construct the prediction model. Model performance was assessed using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis, and a nomogram was established. The SHAP method was used to quantify the contribution of each predictor to the model. Results A total of nine independent predictors were identified:sex, history of primary stent implantation, history of secondary stent implantation, platelet count (PLT), low density lipoprotein cholesterol (LDLC), globulin (Glo), left ventricular ejection fraction (LVEF), MHR and SIRI. Multivariate Logistic regression showed that higher MHR (OR＝4.4,5%CI:2.972～6.704, P＜0.001) and higher SIRI (OR＝1.0,5%CI:1.015～1.823, P＝0.039) were independently associated with an increased risk of HF readmission, whereas higher LVEF (OR＝0.0,5%CI:0.233～0.439, P＜0.001) was independently associated with a reduced risk of HF readmission. In the test set, the Logistic regression model achieved an area under the curve (AUC) of 0.917, demonstrating good discrimination and calibration. SHAP analysis showed that MHR contributed most to the model prediction, followed by LVEF. ConclusionMHR combined with SIRI effectively predicts the risk of HF readmission in elderly patients with T2DM and ACS. The Logistic regression-based nomogram showed good predictive performance and potential clinical utility.