中国全科医学 ›› 2026, Vol. 29 ›› Issue (14): 1840-1848.DOI: 10.12114/j.issn.1007-9572.2025.0397

• 论著 • 上一篇    

基于中医证候及脉图参数构建原发性高血压伴左心室肥厚风险预测模型研究

王斯曼, 张梦楚, 李文, 许艾, 徐璡, 郭睿, 燕海霞*()   

  1. 201203 上海市,上海中医药大学上海市健康辨识与评估重点实验室
  • 收稿日期:2025-07-18 修回日期:2026-01-25 出版日期:2026-05-15 发布日期:2026-04-14
  • 通讯作者: 燕海霞

  • 作者贡献:

    王斯曼负责数据分析、论文撰写及修改;张梦楚负责数据整理及分析;李文、许艾负责数据收集;徐璡、郭睿负责研究思路指导;燕海霞负责研究设计构思、思路指导、数据整体分析及论文修改指导。

  • 基金资助:
    国家自然科学基金资助项目(81973749)

Research on the Construction of a Risk Prediction Model for Primary Hypertension with Left Ventricular Hypertrophy Based on TCM Syndromes and Pulse Graph Parameters

WANG Siman, ZHANG Mengchu, LI Wen, XU Ai, XU Jin, GUO Rui, YAN Haixia*()   

  1. Shanghai Key Laboratory of Health Identification and Assessment, Shanghai University of Traditional Chinese Medicine, Shanghai 201203, China
  • Received:2025-07-18 Revised:2026-01-25 Published:2026-05-15 Online:2026-04-14
  • Contact: YAN Haixia

摘要: 背景 原发性高血压伴左心室肥厚(EH-LVH)的早期风险评估对临床干预至关重要,但现有预测模型常忽视中医证候、脉图参数等中医临床信息。因此,整合上述特色指标构建EH-LVH风险预测模型,将为中西医结合的风险分层与临床决策提供新依据。 目的 基于中医证候和脉图参数构建EH-LVH风险预测列线图模型。 方法 选取2018年8月—2021年6月上海市中医医院、上海中医药大学附属上海市中西医结合医院、上海中医药大学附属曙光医院收治的201例原发性高血压住院患者,根据超声心动图结果将其分为原发性高血压伴左心室肥厚组(EH-LVH组)和原发性高血压无左心室肥厚组(EH-NLVH组)。收集两组患者一般资料、理化指标及中医问诊信息,通过SMART-Ⅰ型中医脉象分析仪检测脉图参数。采用多因素Logistic回归分析探讨与EH-LVH发生风险存在独立关联的因素。应用R软件4.1.1rms程序包,分别以脉图参数、中医证候、脉图参数+中医证候+一般资料为变量,建立3个列线图模型(依次为模型A、B、C),以受试者工作特征(ROC)曲线分析区分度,校正曲线评估准确度,Hosmer-Lemeshow检验验证校准度,临床决策曲线(DCA曲线)评价临床实用性,并进行模型比较。 结果 多因素Logistic回归分析结果显示,低密度脂蛋白胆固醇(OR=1.511,95%CI=1.709~2.115)、阴虚阳亢证(OR=2.493,95%CI=1.272~4.885)、气滞血瘀证(OR=7.866,95%CI=2.201~28.110)、T(OR=1.906,95%CI=1.278~2.842)、H3/H1(OR=1.549,95%CI=1.021~2.351)、W1/T(OR=2.129,95%CI=1.369~3.310)是与EH-LVH风险独立相关的影响因素(P<0.01)。列线图模型的ROC曲线下面积(AUC)分别为:模型A的AUC为0.642(95%CI=0.571~0.713),模型B的AUC为0.717(95%CI=0.646~0.788),模型C的AUC为0.784(95%CI=0.719~0.849)。校准度结果显示,模型A:χ2=0.133(P>0.05),模型B:χ2=4.316(P>0.05),模型C:χ2=3.754(P>0.05),提示3个模型校准曲线的预测概率与实际概率一致性良好。DCA曲线分析显示,当预测EH-LVH发生的阈值处于0.05~0.80(估算值)时,模型C的适用性达到最优,表明该模型的临床实用价值更为显著。 结论 本研究以中医证候及脉图参数为核心预测变量,成功构建EH-LVH风险预测列线图。该列线图的区分效能与预测精准度经验证表现良好,具备临床推广应用价值,可作为临床开展该疾病风险评估的参考依据。

关键词: 原发性高血压, 左心室肥厚, 脉图参数, 中医证候, 预测模型

Abstract:

Background

Early risk assessment of essential hypertension with left ventricular hypertrophy (EH-LVH) is crucial for clinical intervention, but existing predictive models often overlook Traditional Chinese Medicine (TCM) syndromes, pulse graph parameters, and other TCM clinical information. Therefore, integrating the aforementioned characteristic indicators to construct an EH-LVH risk prediction model will provide new evidence for risk stratification and clinical decision-making in integrated Traditional Chinese and Western Medicine.

Objective

To develop a nomogram model for predicting the risk of EH-LVH based on TCM syndromes and pulse graph parameters.

Methods

A total of 201 inpatients with essential hypertension admitted to Shanghai Municipal Hospital of Traditional Chinese Medicine, Shanghai Integrated Traditional Chinese and Western Medicine Hospital Affiliated to Shanghai University of Traditional Chinese Medicine, and Shuguang Hospital Affiliated to Shanghai University of Traditional Chinese Medicine from August 2018 to June 2021 were selected. Based on echocardiographic results, they were divided into the essential hypertension with left ventricular hypertrophy group (EH-LVH group) and the essential hypertension without left ventricular hypertrophy group (EH-NLVH group). General information, physicochemical indicators, and TCM inquiry data of the two groups were collected, and pulse graph parameters were detected using the SMART-Ⅰ TCM pulse analysis instrument. Multivariate Logistic regression analysis was used to explore factors independently associated with the risk of EH-LVH. Using the rms package in R software version 4.1.1, three nomogram models (Models A, B, and C) were established with pulse graph parameters, TCM syndromes, and pulse graph parameters + TCM syndromes + general information as variables, respectively. Receiver operating characteristic (ROC) curves were used to assess discrimination, calibration curves to evaluate accuracy, Hosmer-Lemeshow test to verify calibration, and decision curve analysis (DCA) to evaluate clinical utility, followed by model comparisons.

Results

Multivariate Logistic regression analysis showed that low-density lipoprotein cholesterol (OR=1.511, 95%CI=1.709-2.115), Yin deficiency with Yang hyperactivity syndrome (OR=2.493, 95%CI=1.272-4.885), Qi stagnation and blood stasis syndrome (OR=7.866, 95%CI=2.201-28.110), T (OR=1.906, 95%CI=1.278-2.842), H3/H1 (OR=1.549, 95%CI=1.021-2.351), and W1/T (OR=2.129, 95%CI=1.369-3.310) were factors independently associated with the risk of EH-LVH (P<0.05). The areas under the ROC curves (AUC) of the nomogram models were as follows: the AUC of Model A was 0.642 (95%CI=0.571-0.713), that of Model B was 0.717 (95%CI=0.646-0.788), and that of Model C was 0.784 (95%CI=0.719-0.849). The calibration results showed that for Model A: χ2=0.133 (P>0.05), for Model B: χ2=4.316 (P>0.05), and for Model C: χ2=3.754 (P>0.05), indicating good agreement between the predicted probabilities and actual probabilities in the calibration curves of all three models. DCA curve analysis showed that when the threshold for predicting EH-LVH occurrence was between 0.05 and 0.80 (estimated value), Model C achieved optimal applicability, indicating its more significant clinical utility value.

Conclusion

This study successfully construct a nomogram for predicting the risk of EH-LVH, with TCM syndromes and pulse graph parameters as core predictive variables. The discrimination performance and predictive accuracy of this nomogram were validated to be favorable, demonstrating value for clinical promotion and application, and it can serve as a reference for clinical risk assessment of this disease.

Key words: Essential hypertension, Left ventricular hypertrophy, Pulse wave parameters, TCM syndromes, Prediction model