可穿戴心率变异性参数与急性期入院心衰患者NT-proBNP水平的关联性分析

Association between wearable heart rate variability parameters and NT-proBNP levels in acute heart failure patients

  • 摘要: 背景 心率变异性(heart rate variability,HRV)作为评估自主神经系统活性的重要指标,已被广泛用作心力衰竭(heart failure,HF)患者预后的评估工具。N末端B型利钠肽原(NT-proBNP)作为一种标志物,在心衰的诊断、风险分层及预后评估中发挥着至关重要的作用,但其检测依赖侵入性操作且难以动态监测。目的 探讨可穿戴设备监测的HRV参数与NT-proBNP 水平的相关性,并验证其作为无创风险分层工具的可行性。方法 本研究利用可穿戴设备收集了2021 年1 月至2022 年12 月四川大学华西医院心内科的急性心衰患者入院、出院时的24 h 心电数据,经过预处理后提取了HRV参数,以NT-proBNP≥3 500 pmol/L 为高风险截断值,采用Spearman 相关分析HRV与NT-proBNP的关联性,评估逻辑回归、K最近邻(KNN)、支持向量机(SVM)、随机森林和XGBoost五种机器学习模型预测不良预后(NT-proBNP≥3 500 pmol/L)效能,并通过SHAP值量化特征贡献度。结果 共纳入51 名心衰患者,收集心电数据87 份。HRV参数中,SDNN、SD2、VLF等10 项指标与NT-proBNP 呈显著负相关(r=-0.371 至-0.390,P<0.001)。NT-proBNP 水平≥3 500 pmol/L 的患者HRV 参数SDNNM(IQR): 51.10(38.50 ~ 67.20) ms vs 77.95 (54.45 ~ 95.50) ms, P<0.001, SD2M(IQR): 68.30(52.90 ~ 93.90) ms vs 108.00(76.20 ~ 132.47) ms,P=0.003,VLFM(IQR):18.82(5.84 ~ 59.61) mHz vs 59.36 (33.70 ~ 116.90) mHz,P=0.002,ULFM(IQR):6.30(1.99 ~ 18.02) mHz vs 18.60 (10.05 ~ 34.09) mHz,P=0.001均显著低于NT-proBNP<3 500 pmol/L 组,差异具有统计学意义。机器学习模型中,Logistic 回归分类性能最优(AUC=0.830,95% CI:0.760 ~ 0.890),SHAP分析显示SD2 和LF/HF贡献度较高。结论 可穿戴HRV参数与NT-proBNP 水平显著相关,且能通过机器学习模型有效区分NT-proBNP 高低水平组,为心衰患者的无创动态监测及急性失代偿风险预测提供了新策略。

     

    Abstract: Background Heart rate variability (HRV), an indicator of autonomic nervous system activity, is widely recognized for assessing heart failure (HF) prognosis. N-terminal pro-B-type natriuretic peptide (NT-proBNP) plays a crucial role in the diagnosis, risk stratification, and prognosis evaluation of HF, but its measurement relies on invasive procedures and lacks dynamic monitoring capabilities. Objective To explore the correlation between wearable HRV parameters and NT-proBNP levels and validate their feasibility as a non-invasive risk stratification tool.Methods The 24-hour wearable device physiological data and clinical indicators were collected from acute HF patients admitted to the Department of Cardiology, West China Hospital, Sichuan University, from January 2021 to December 2022. Post preprocessing, HRV parameters were extracted, and a threshold of 3500 pmol/L was established for NT-proBNP levels. Spearman correlation was used to analyze the association between HRV and NT proBNP, and the efficacy of five machine learning models including Logistic Regression, K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Random Forest and XGBoost to predict poor prognosis (NT proBNP ≥ 3500 pmol/l) was evaluated, and the characteristic contribution was quantified by the snap value. Results A total of 51 heart failure patients were included, with 87 electrocardiographic datasets collected. Among the HRV parameters, SDNN, SD2, VLF, and 7 others showed significant negative correlations with NT-proBNP levels (r=-0.371 to -0.390, P<0.001). Patients with NT-proBNP≥3500 pmol/L had significantly lower HRV parameters, including SDNN (M(IQR): 51.10 38.50 - 67.20 ms vs 77.95 54.45 - 95.50 ms, P<0.001), SD2 (M(IQR): 68.30 52.90 - 93.90 ms vs 108.00 76.20 - 132.47 ms, P=0.003), VLF (M(IQR): 18.82 5.84 - 59.61 mHz vs 59.36 33.70 - 116.90 mHz, P=0.002), ULF (M(IQR): 6.30 1.99 - 18.02 mHz vs 18.60 10.05 - 34.09 mHz, P=0.001). Among the machine learning models, Logistic Regression demonstrated the best classification performance (AUC=0.830, 95% CI: 0.760 - 0.890). SHAP analysis revealed that SD2 and LF/HF contributed the most to the model's classification. Conclusion Wearable HRV parameters are significantly correlated with NT-proBNP levels and can effectively differentiate high and low NT-proBNP groups using machine learning models, offering a new strategy for non-invasive dynamic monitoring and acute decompensation risk prediction in heart failure patients.

     

/

返回文章
返回