缺血性脑卒中危险因素的多元回归分析

目的探讨缺血性脑卒中患者发病的影响因素,建立发病概率预测模型,为加强防治缺血性脑卒中提供依据。方法分析136例缺血性脑卒中病例及145例对照资料,进行统计学分析,研究影响缺血性脑卒中发病的危险因素,采用Logistic回归分析的方法建立发病概率预测模型并利用ROC曲线(受试者工作特征曲线receiver operating characteristic curve,ROC)进行评价。结果缺血性脑卒中组的教育程度相对较低,三酰甘油(TG)和胆固醇(TC)含量较对照组显著高,而高密度脂蛋白(HDL-C)显著低于对照组;多因素分析结果显示吸烟、糖尿病、TG及TC等为影响缺血性脑卒中发病的危险因素;ROC曲线下面积为0.96,模型具有较好的预测价值。结论吸烟、糖尿病、TG、TC是缺血性脑卒中的危险因素,Logistic回归预测概率模型评价效果较好,能较为准确地预测缺血性脑卒中发病概率。

Logistic regression analysis of risk factors of ischemic stroke

Objective To explore the risk factors of ischemic stroke patients and build up predictive model of incidence probability,then provide measures for prevention and cure of ischemic stroke.Methods Retrospectivly analyzed 136 cases of ischemic stroke and 145 cases of control for understanding the major risk factors of ischemic stroke,then used Logistic regression method to study the risk factors of stroke.Results Comparing with controls,ischemic stroke patients tended to be less educated,higher TG and TC.The main risk factors of ischemic stroke included smoking,diabetes,TG and TC.The area of ROC curve was 0.96,which indicated that the predictive value was very good.Conclusion Smoking,diabetes,TG and TC are the risk factors of ischemic stroke.We have built up a good predictive model by Logistic regression method.It is good for predicting the incidence probabiltiy of ischemic stroke.