首都医科大学学报 ›› 2024, Vol. 45 ›› Issue (2): 356-362.doi: 10.3969/j.issn.1006-7795.2024.02.025

• 临床研究 • 上一篇    下一篇

北京市功能单位体检人群的7年高血压病风险评估与预测模型

潘  援1,4,  金  瑞1,2,  刘相佟1,2,  吕世云1,2,  陈  硕3,  郭秀花1,2*   

  1. 1.首都医科大学公共卫生学院,北京  100069;2.临床流行病学北京市重点实验室,北京 100069;3.北京市体检中心,北京 100050;4 首都医科大学教务处,北京  100069
  • 收稿日期:2023-05-16 出版日期:2024-04-21 发布日期:2024-04-25
  • 通讯作者: 郭秀花 E-mail:guoxiuh@ccmu.edu.cn

Seven-year risk assessment and prediction model of hypertension among people who underwent physical examination in functional units in Beijing

Pan Yuan1,4,Jin Rui1,2,Liu Xiangtong1,2,Lü Shiyun1,2,Chen Shuo3,Guo Xiuhua1,2*   

  1. 1.School of Public Health, Capital Medical University, Beijing 100069, China; 2.Beijing Key Laboratory of Clinical Epidemiology, Beijing 100069, China; 3.Beijing Physical Examination Center, Beijing 100050, China; 4.Teaching Affairs Office, Capital Medical University, Beijing 100069, China
  • Received:2023-05-16 Online:2024-04-21 Published:2024-04-25

摘要: 目的  构建北京市功能单位体检人群7年高血压病风险评估与预测模型,为高血压防控提供科学依据。 方法  本研究数据来自北京市健康管理队列研究中的健康体检人群,最终纳入2020年1月至2020年12月在北京市体检中心进行体检的29 209人作为研究对象,依据2014年-2020年间人群中高血压发病率与人口学、行为习惯等信息,采用单因素分析筛选风险因子,通过Logistic回归分析建立高血压病风险评估模型,并绘制受试者工作特征(receiver operating characteristic,ROC)曲线验证模型预测能力。 结果  高血压7年新发病例为2 346人,发病率为8.03%;建立Logistic回归模型,最终模型纳入了高血压家族史、体质量指数(body mass index,BMI)、性别、婚姻、锻炼强度、年龄、饮酒情况、吸烟情况,ROC曲线下面积为0.736。 结论  北京市功能单位人群7年高血压发病率为8.03%,发病风险较高,高血压家族史、BMI、性别、婚姻、锻炼强度、年龄、饮酒情况、吸烟情况为高血压发病的危险因素,据此构建了风险评估预测模型。

关键词: 高血压, 风险评估, Logistic回归, ROC曲线, 列线图

Abstract: Objective  To predict the 7-year risk of hypertension among people who underwent physical examination in functional units in Beijing, and to provide foundation data for prevention of hypertension. Methods  A total of 29 209 people in health management cohort in Beijing in 2020 were selected as research subjects. Based on the information of hypertension incidence, demography, behavior habits and blood index from 2014 to 2020, we use monofactor analysis to choose the risk factors, and establish the risk assessment model for hypertension by Logistic regression, then draw the receiver operating characteristic (ROC) curve to verify the predictive ability of the model. Results  The number of new cases of  hypertension in 7 years were 2 346, the incidence of hypertension was 8.03%; the variables were used to establish a Logistic regression model, and the final model included six variables including parental history of hypertension, body mass index, gender, marriage, exercise intensity and milk drinking frequency. The area under the ROC curve was 0.736. Conclusion  The 7-year risk of hypertension is higher in functional unit population in Beijing, and parental history of hypertension, body mass index (BMI), gender, marriage, exercise intensity and milk drinking frequency are the risk factors for hypertension. Based on this, the risk assessment and prediction model is constructed. 

Key words: hypertension, risk assessment, Logistic regression, ROC curve, nomogram

中图分类号: