论文部分内容阅读
目的建立原发性高血压发病概率预测模型并对模型预测效果进行评价,为防治高血压发病提供依据。方法调查285例原发性高血压患者和260例对照病例,采用logistic回归分析的方法研究影响高血压发病的危险因素,建立发病概率预测模型并利用ROC曲线进行评价。结果影响高血压发病的危险因素包括家族史、BMI、文化程度、饮酒、蔬菜水果摄入、饮食偏咸、吃动物内脏、体育锻炼程度、血压差等,ROC曲线下面积为0.930,模型具有较好的预测价值。结论建立了合理的logistic回归概率预测模型,模型评价效果较好,能较为准确地预测高血压发病概率。
Objective To establish the predictive model for the incidence of essential hypertension and evaluate the predictive value of the model for the prevention and treatment of hypertension. Methods 285 patients with essential hypertension and 260 control cases were investigated. Logistic regression analysis was used to study the risk factors influencing the incidence of hypertension. The incidence prediction model was established and evaluated by ROC curve. Results The risk factors influencing the incidence of hypertension included family history, BMI, education level, drinking, intake of fruits and vegetables, salty diet, eating animal offal, physical activity and blood pressure difference. The area under the ROC curve was 0.930. Good forecast value. Conclusion The reasonable logistic regression probability prediction model is established, and the model evaluation effect is good, which can accurately predict the incidence of hypertension.