基于主成分分析的缺血性心脑血管疾病共患危险因素研究

Study on risk factors for comorbidity of ischemic cardiovascular and cerebrovascular diseases based on principal component analysis

摘要:

目的]利用主成分分析法,系统全面地对缺血性心脑血管疾病共患危险因素进行分析,筛选影响心脑血管疾病共患的危险因素,构建预测模型,为共病防治提供筛查工具。 [方法]利用新疆医科大学第一附属医院冠心病防治一体化信息平台,回顾性纳入2009年12月1日—2020年6月30日首次诊断为冠心病的患者作为研究对象,根据纳排标准研究对象被分为缺血性心脑血管疾病共患组(简称共患组)和冠心病单患组(简称单患组)。收集患者入院期间的临床指标数据,根据7∶3的比例将样本随机分为建模集和验证集。在建模集中,利用主成分分析探索心脑共患与冠心病单患患者之间危险因素的分布差异。通过分析因子载荷贡献,识别出最重要的变量。利用Logistic回归模型构建预测模型,并且构建列线图。在验证集中评估模型的预测能力和稳健性。 [结果]共纳入11 808例受试者,其中2 781例(23.6%)为心脑共患者,9 027例(76.4%)为冠心病单患者。与单患组相比,共患组的平均年龄更大,女性占比较多,并且入院时收缩压水平较高(P＜0.001)。此外,共患组的血红蛋白、血小板分布宽度、低密度脂蛋白和血钠水平均显著较高(P＜0.05)。共有8 265例被随机分配到建模集,通过主成分分析筛选出因子载荷贡献大于5的7个关键因素,包括血钠水平、入院收缩压、入院舒张压、年龄、血红蛋白水平、血小板分布宽度和总胆红素。利用Logistic回归模型构建列线图,该列线图模型预测缺血性心脑血管疾病共患的受试者工作特征曲线下面积为0.630(95%CI:0.600~0.768,P＜0.001)。共有3 543例受试者被随机分配至验证集,预测模型在验证集曲线下面积为0.628。 [结论]高血钠水平、高入院收缩压和舒张压、年龄增高、高血红蛋白水平、高血小板分布宽度和低总胆红素水平是缺血性心脑血管疾病共患的危险因素,构建列线图用于筛查缺血性心脑血管疾病共患风险具有一定的临床价值。

Abstract:

Aim To systematically analyze the risk factors for comorbid ischemic cardiovascular and cerebrovascular diseases using principal component analysis. It seeks to identify key risk factors influencing comorbidity and to construct a predictive model to serve as a screening tool for the prevention and management of comorbidities. Methods This retrospective study included patients diagnosed with coronary artery disease from December 1,9 to June 0,0, at the First Affiliated Hospital of Xinjiang Medical University, using data from the hospital's integrated coronary heart disease prevention platform. Patients were divided into two groups:those with both cardiovascular and cerebrovascular diseases and those with only coronary artery disease based on inclusion and exclusion criteria. Clinical indicators during hospital admission were collected. The sample was randomly divided into a modeling set and a validation set in a 7∶3 ratio. In the modeling set, principal component analysis was used to explore the distribution differences in risk factors between the comorbid group and the single-disease group. By analyzing factor load contributions, the most significant variables were identified. Logistic regression was then used to construct a predictive model, and a nomogram was generated. The model's predictive ability and robustness were evaluated in the validation set. Results A total of 11 808 participants were included, with 2 781 (23.6%) in the comorbid group and 9 027 (76.4%) in the coronary heart disease-only group. Compared with the single-disease group, the comorbid group had a higher average age, a greater proportion of females, and higher systolic blood pressure levels at admission (P＜0.001). Additionally, the levels of hemoglobin, platelet distribution width, low density lipoprotein, and blood sodium were significantly higher in the comorbid group (P＜0.05). A total of 8 265 participants were randomly assigned to the modeling set. Principal component analysis identified seven key factors with factor load contributions greater than 5:sodium level, systolic blood pressure, diastolic blood pressure, age, hemoglobin concentration, platelet distribution width, and total bilirubin level. Using these factors, a nomogram was constructed via Logistic regression. The nomogram's area under the receiver operating characteristic curve for predicting comorbid ischemic cardiovascular and cerebrovascular diseases was 0.630(95%CI:0.600~0.768, P＜0.001). A total of 3 543 participants were randomly assigned to the validation set. In the validation set, the receiver operating characteristic curve was 0.628. Conclusion Elevated sodium level, higher systolic and diastolic blood pressure at admission, older age, increased hemoglobin concentration, higher platelet distribution width, and lower total bilirubin level are risk factors for comorbid ischemic cardiovascular and cerebrovascular diseases. The nomogram constructed has clinical value for screening patients with such comorbidities.