临床荟萃 ›› 2025, Vol. 40 ›› Issue (3): 205-210.doi: 10.3969/j.issn.1004-583X.2025.03.002

• 论著 • 上一篇    下一篇

中青年人群中隐匿性高血压患病危险因素与预测模型

赵光艳1a,2, 韩拓1a, 梁西颖1a, 王倩1b, 张岩1a, 王聪霞1a()   

  1. 1.西安交通大学第二附属医院 a.心血管内科;b.健康管理部,陕西 西安 710004
    2.榆林市子洲县人民医院,陕西 榆林 718400
  • 收稿日期:2024-09-04 出版日期:2025-03-20 发布日期:2025-03-24
  • 通讯作者: 王聪霞 E-mail:wcx622@163.com
  • 基金资助:
    国家自然科学基金青年项目——STING-IRF3信号轴介导细胞焦亡在心肌缺血/再灌注损伤中的作用及机制研究(82100369);陕西省自然科学基础研究计划青年项目——PI3K信号通路诱导自噬在重组人ACE2调控CMVEC中的机制(2020JQ-552)

Risk factors and predictive model for masked hypertension in young and middle-aged adults

Zhao Guangyan1a,2, Han Tuo1a, Liang Xiying1a, Wang Qian1b, Zhang Yan1a, Wang Congxia1a()   

  1. 1a. Department of Cardiology; b. Department of Health Management, the Second Affiliated Hospital of Xi'an Jiaotong University, Xi'an 710004, China
    2. Zizhou People's Hospital, Yulin 718400, China
  • Received:2024-09-04 Online:2025-03-20 Published:2025-03-24
  • Contact: Wang Congxia E-mail:wcx622@163.com

摘要:

目的 探究中青年人群中隐匿性高血压(masked hypertension, MH)的危险因素,并构建诊断预测模型。方法 选取2021年1月-2023年12月就诊于西安交通大学第二附属医院心血管门诊或同期健康体检的中青年人群,进行问卷调查、体格检查、生化指标检测及诊室血压测量与动态血压监测。采用单因素-多因素logistic回归探究MH患病危险因素,并建立疾病诊断模型及列线图,受试者工作特征(receiver operator characteristic, ROC)曲线评估模型预测性能。采用Bootstrap自助法(n=1 000次)进行模型内部验证,计算C统计量及校准曲线、决策曲线评估模型优劣。结果 最终纳入805例门诊与健康体检患者,其中MH患者152例,患病率为18.9%。与正常血压组相比,MH患者平均年龄、体质量指数(BMI)、男性占比更高,总胆固醇、甘油三酯(TG)、空腹血糖与血肌酐水平更高。Logistic回归分析显示,年龄(OR=1.09, 95%CI: 1.07~1.11)、BMI (OR=1.25, 95%CI: 1.15~1.35)、log2(TG) (OR=1.29, 95%CI: 1.09~1.52)、空腹血糖(OR=1.28, 95%CI: 1.01~1.62)可能是MH发生的独立危险因素。基于上述结果建立中青年人群中MH预测模型与诊断列线图,ROC曲线下面积为0.821。通过Bootstrap(n=1 000次)进行内部验证,模型的区分度与校准度均表现良好。决策曲线显示在MH患病概率介于1%~65%时,该模型均能带来临床净获益。结论 年龄、BMI、空腹血糖与TG是中青年MH的独立危险因素,基于上述危险因素构建的预测模型具有较好的区分度、校准度与临床实际效用,有助力于门诊MH患者的早期精准识别。

关键词: 隐匿性高血压, 人体质量指数, 甘油三酯, 空腹血糖, 预测模型

Abstract:

Objective To investigate the risk factors of masked hypertension (MH) in young and middle-aged population and to develop a predictive model for MH diagnosis. Methods From January 2021 to December 2023, young and middle-aged adults who visited the cardiovascular outpatient clinic or underwent health check-up during the same period were recruited for questionnaire survey, physical examination, biochemical tests, waiting room blood pressure measurement and ambulatory blood pressure monitoring. The univariate and multifactorial logistic regression were performed to evaluate the risk factors of MH. Then a diagnostic prediction model was developed and a nomogram was created. The predictive performance of the model was evaluated using the receiver operator characteristic (ROC) curve. The bootstrap method (n=1000 times) was used for internally verification, and C-statistics, calibration curve and decision curve were created to evaluate the model. Results A total of 805 outpatients and health check-ups were included, with a prevalence rate of 18.9%(152/805). Compared to the normotensives, MH patients were elder, more male, and had significantly higher levels of body mass index (BMI), total cholesterol, triglycerides (TG), fasting blood glucose(FBG), and serum creatinine(SCR).Logistic regression revealed that age (OR=1.09, 95%CI: 1.07-1.11), BMI (OR=1.25, 95%CI: 1.15-1.35), log2(TG) (OR=1.29, 95%CI: 1.09-1.52), FBG(OR=1.28, 95%CI: 1.01-1.62) could be the independent risk factors for MH. Based on the above results, an MH prediction model and diagnostic nomogram were constructed for young and middle-aged adults, and the area under the ROC curve was 0.821. As for the internal verification by 1000 times bootstrap, the differentiation and calibration of the model were excellent. The decision curve showd that the model could yield a net benefit when the probability of MH was between 1% and 65%. Conclusion Age, BMI, FBG and TG are independent risk factors for MH in young and middle-aged adults. The prediction model based on the above risk factors has good differentiation, calibration and clinical benefit, which can contribute to the early and precisive identification of MH in outpatients.

Key words: masked hypertension, body mass index, triglycerides, fasting blood glucose, prediction model

中图分类号: