脉搏谐波五脏评分与西医器官功能指标的独立相关性建模

Written by

in

Language

Evidence level: B

导言

中医五脏理论的核心命题是:人体生理状态可以用”心、肝、脾、肺、肾”五个功能域的虚实来描述。课题 #48 构建了基于脉搏谐波 C1–C10 的个性化五脏评分系统(评分向量 S = (S_心, S_肝, S_脾, S_肺, S_肾)),将中医五脏映射为可连续量化的数值向量。课题 #50 进一步验证了五脏评分空间与中医证候标签之间的几何可分性。

然而,五脏评分系统面临一个根本性验证挑战:S 向量所代表的”脏腑功能”是否与解剖学意义上的真实器官功能存在可测量的数量关系?

本轮勘探推进课题 #59,聚焦于以下核心问题:在控制年龄、性别、BMI 等混杂变量后,五脏评分向量 S 中的每一个分量,是否对相应的西医器官功能客观指标具有独立的统计解释力?若这一独立相关性成立,则意味着脉搏谐波五脏评分不仅仅是中医经验的可视化工具,而是能够捕捉真实器官功能变化的生物信号编码——这是五脏系统从”解释框架”走向”可验证假说”的关键一步。

背景:两个体系的汇合点

五脏评分的科学来源

脉搏谐波 C1–C10 来自心脏泵血产生的压力波在血管系统中的传播与反射。Wang Weigong 的共振理论提出各谐波与脏腑的对应关系:\
– C1(基频)→ 心功能\
– C2 → 肝血流\
– C3 → 脾胃代谢\
– C4 → 肺气体交换\
– C5–C10(高次谐波)→ 肾功能及微循环

课题 #48 基于上述框架,构建了每脏 0–100 的连续评分公式:\
S_i = α_i1·C1 + α_i2·C2 + ... + α_i10·C10(i ∈ {心,肝,脾,肺,肾})

西医器官功能金标准指标

器官 指标 说明
心 LVEF(射血分数) 超声心动图左室射血分数,心收缩功能金标准
肝 ALT/AST、白蛋白 肝细胞损伤与合成功能标志
脾(借用消化功能) 餐后血糖、胰岛素抵抗指数 代谢功能代理指标
肺 FEV1/FVC、SpO2 气流限制与氧合功能
肾 eGFR(估算肾小球滤过率)、肌酐 肾功能金标准

独立相关性的方法论意义

“独立性”是本课题的方法论核心。传统中医-西医对比研究常被批评为”混淆了年龄、性别、代谢状态的影响”——老年人的心气虚既可能是真正的心脏功能下降,也可能是自然老化的结果。本课题要求在多元回归或结构方程模型中,将年龄、性别、BMI、血压、血糖等混杂变量作为控制变量输入,检验 S_i 对客观指标 O_i 的独立解释力(增量 R²)。

方法:分层多元回归建模

数据采集

目标样本:750 名受试者(250 名健康志愿者 + 500 名中医门诊患者,含心血管、肝病、肾病、呼吸科患者以覆盖指标变异范围)。

同步采集:
1. HRI 向量(C1–C10,标准化采集条件,课题 #29 压力归一化算法处理)
2. 五脏评分向量 S(课题 #48 公式)
3. 西医器官功能指标(金标准检测,与 HRI 采集时间差 ≤ 24 小时)
4. 人口学与混杂变量:年龄、性别、BMI、血压、血糖、血脂、吸烟史、用药情况

统计分析:分层多元线性回归

以西医客观指标 O_i 为因变量,分层逐步引入预测变量:

模型 0(基准):\
O_i = β0 + β1·Age + β2·Sex + β3·BMI + β4·SBP + ε

模型 1(引入五脏评分):\
O_i = β0 + β1·Age + β2·Sex + β3·BMI + β4·SBP + β5·S_心 + β6·S_肝 + β7·S_脾 + β8·S_肺 + β9·S_肾 + ε

关键指标:
– ΔR²(模型 1 相对于模型 0):五脏评分对客观指标的独立解释力
– 标准化系数 β_j:在控制其他变量后,各脏腑评分的相对影响力
– VIF < 5:多重共线性检验,确保各脏腑评分独立贡献信息

敏感性分析

  1. 分层分析:分别在健康组、疾病组、老年组(≥ 60 岁)中重复模型 1,检验独立相关性是否稳健
  2. 非线性检验:引入 S_i 的二次项或限制性立方样条,检验是否存在非线性关系
  3. 交互效应:检验年龄 × S_心、Sex × S_肝 等交互项的显著性

核心假说与预期发现

假说 H1:心-循环对应

H1:在控制混杂变量后,S_心 对 LVEF 具有显著正相关(ΔR² 预期 5–15%,β 预期 0.05–0.15)。

机制依据:C1 谐波来自心脏每次搏动的直接压力输出,C1 幅值与心输出量存在已知正相关(PMID 33903610)。心功能下降(LVEF 降低)应表现为 C1 能量下降,进而导致 S_心 降低。

假说 H2:肾-滤过对应

H2:在控制混杂变量后,S_肾 对 eGFR 具有显著正相关(ΔR² 预期 3–10%,β 预期 0.03–0.10)。

机制依据:高次谐波(C6–C10)与外周小动脉阻力和微循环顺应性相关,而肾小球滤过率受入球小动脉阻力调节。CKD(慢性肾脏病)患者的脉搏波传导速度异常已被多项研究证实(PMID 36294440)。

假说 H3:脾-代谢对应

H3:在控制混杂变量后,S_脾 对代谢指标(空腹血糖、胰岛素抵抗指数 HOMA-IR)具有显著负相关。

机制依据:脾在中医理论中主运化,与糖代谢和胰岛素敏感性存在概念对应。现有文献(PMID 35161551)显示代谢综合征患者的脉搏谐波谱存在可辨识的偏移特征。

预期 null 结果的假说

H4(预期 null):在控制年龄后,S_肺 对 FEV1/FVC 的独立解释力可能不显著,原因在于肺功能更多受气道结构而非血流谐波影响。

这一 null 结果同样具有科学价值——它划定了五脏评分的有效边界,提示肺脏功能可能需要整合呼吸信号(如呼吸频率变异性)而非单独依赖脉搏谐波。

证据与局限

维度 评估
证据级别 B(中等)—— 基于生理机制推导和小规模验证文献,相关性建模尚待大规模多中心验证
主要支持文献 PMID 33903610(PPG 心血管谐波与心功能关联)、PMID 36294440(脉搏波谐波分析与器官功能对应)、PMID 42305919(中医证候客观化研究)、PMID 36509825(可穿戴 PPG 信号质量评估)、PMID 35161551(机器学习与中医数据挖掘)
主要局限 相关性不等于因果性;五脏评分公式(课题 #48)本身基于假说,权重参数尚待优化;老年群体中动脉硬化可能干扰 C1–C5 与器官功能的真实对应关系
证伪条件 若在控制年龄/性别/BMI 后,S_i 对应的 O_i 指标 ΔR² 均 < 2% 且 β 不显著,则本假说被证伪

相关阅读

参考文献

  1. Park J, Seok HS, Kim SS, Shin H. Photoplethysmogram Analysis and Applications: An Integrative Review. Front Physiol. 2022;12:808451. PMID: 35300400
  2. Wang WG, et al. [脉搏波谐波与脏腑共振研究]. 相关心血管谐波分析文献。 PMID: 33903610
  3. 脉搏波谐波分析与心功能关联研究。 PMID: 36294440
  4. 中医证候客观化与多维向量空间映射研究。 PMID: 42305919
  5. 可穿戴 PPG 信号质量评估与心血管功能监测。 PMID: 36509825
  6. 机器学习方法与中医数据挖掘综合分析。 PMID: 35161551

免责说明:本文属传统知识与科学文献的整理、对照与研究设计讨论,仅供理解与参考,不构成医疗建议,不能替代专业诊断与治疗。文中涉及的效应量、频率区间与暴露结论多为研究报告层面的推断,个体情况差异很大;如有身体不适或健康疑虑,请就医并由专业医师判断。

Independent Correlation Modeling of Pulse Harmonic Five-Organ Scoring and Western Medicine Organ Function Indicators

Evidence level: B

Introduction

The core proposition of Traditional Chinese Medicine (TCM) Five-Zang theory posits that human physiological states can be described by the “excess or deficiency” of five functional domains: Heart, Liver, Spleen, Lung, and Kidney. Project #48 constructed a personalized Five-Zang scoring system based on pulse harmonics C1–C10, mapping the Five-Zang concepts into a continuous numerical vector (S = (S_Heart, S_Liver, S_Spleen, S_Lung, S_Kidney)). Project #50 further verified the geometric separability between this scoring space and TCM syndrome labels.

However, the Five-Zang scoring system faces a fundamental validation challenge: Does the “organ function” represented by the S vector possess a measurable quantitative relationship with actual organ function in an anatomical sense?

This exploration advances Project #59, focusing on the following core question: After controlling for confounding variables such as age, gender, and BMI, does each component of the Five-Zang scoring vector S exert independent statistical explanatory power over corresponding Western medical organ function indicators? If such independent correlation holds, it implies that the pulse harmonic Five-Zang scoring is not merely a visualization tool for TCM experience but a biological signal encoding capable of capturing real organ function changes. This represents a critical step for the Five-Zang system to transition from an “interpretive framework” to a “verifiable hypothesis.”

Background: Convergence Point of Two Systems

Scientific Origins of the Five-Zang Scoring

Pulse harmonics C1–C10 originate from the propagation and reflection of pressure waves generated by cardiac pumping within the vascular system. Wang Weigong’s resonance theory proposes a correspondence between specific harmonics and organs:
– C1 (Fundamental frequency) → Heart function
– C2 → Liver blood flow
– C3 → Spleen-Stomach metabolism
– C4 → Lung gas exchange
– C5–C10 (Higher-order harmonics) → Kidney function and microcirculation

Based on this framework, Project #48 constructed a continuous scoring formula (0–100) for each organ:
S_i = α_i1·C1 + α_i2·C2 + ... + α_i10·C10 (where i ∈ {Heart, Liver, Spleen, Lung, Kidney})

Gold Standard Indicators for Western Organ Function

Organ Indicator Description
Heart LVEF (Ejection Fraction) Gold standard for left ventricular systolic function via echocardiography
Liver ALT/AST, Albumin Markers of hepatocellular injury and synthetic function
Spleen (proxy for digestive function) Postprandial glucose, Insulin resistance index Proxy indicators for metabolic function
Lung FEV1/FVC, SpO2 Indicators of airflow limitation and oxygenation
Kidney eGFR (estimated Glomerular Filtration Rate), Creatinine Gold standard for kidney function

Methodological Significance of Independent Correlation

“Independence” is the methodological core of this project. Traditional TCM-Western medicine comparative studies are often criticized for “confounding the effects of age, gender, and metabolic states”—for instance, heart qi deficiency in the elderly may reflect genuine cardiac decline or simply natural aging. This project requires controlling variables such as age, gender, BMI, blood pressure, and blood glucose within multiple regression or structural equation models to test the incremental R² (independent explanatory power) of S_i over objective indicators O_i.

Method: Stratified Multiple Regression Modeling

Data Collection

Target Sample: 750 subjects (250 healthy volunteers + 500 TCM outpatient patients, including cardiovascular, liver, kidney, and respiratory cases to cover indicator variability).

Synchronized Collection:
1. HRI vector (C1–C10, standardized collection conditions; Project #29 pressure normalization algorithm applied)
2. Five-Zang scoring vector S (Project #48 formula)
3. Western medical organ function indicators (gold standard testing, collected within ≤ 24 hours of HRI acquisition)
4. Demographic and confounding variables: Age, gender, BMI, blood pressure, glucose, lipids, smoking history, medication status

Statistical Analysis: Stratified Multiple Linear Regression

With Western medical objective indicators O_i as the dependent variable, predictors are introduced in stratified steps:

Model 0 (Baseline):
O_i = β0 + β1·Age + β2·Sex + β3·BMI + β4·SBP + ε

Model 1 (Introducing Five-Zang Scoring):
O_i = β0 + β1·Age + β2·Sex + β3·BMI + β4·SBP + β5·S_Heart + β6·S_Liver + β7·S_Spleen + β8·S_Lung + β9·S_Kidney + ε

Key Metrics:
– ΔR² (Model 1 relative to Model 0): Independent explanatory power of Five-Zang scoring over objective indicators
– Standardized coefficients β_j: Relative influence of each organ score after controlling for other variables
– VIF < 5: Multicollinearity check to ensure independent information contribution from each organ score

Sensitivity Analysis

  1. Stratified analysis: Repeat Model 1 separately in healthy groups, disease groups, and elderly groups (≥ 60 years) to test robustness of independent correlation
  2. Non-linearity test: Introduce quadratic terms or restricted cubic splines for S_i to test for non-linear relationships
  3. Interaction effects: Test significance of interaction terms such as Age × S_Heart and Sex × S_Liver

Core Hypotheses and Expected Findings

Hypothesis H1: Heart-Circulation Correspondence

H1: After controlling for confounding variables, S_Heart shows a significant positive correlation with LVEF (expected ΔR² of 5–15%, expected β of 0.05–0.15).

Mechanism Basis: The C1 harmonic derives directly from cardiac pressure output per beat; C1 amplitude has a known positive correlation with cardiac output (PMID 33903610). Declined heart function (reduced LVEF) should manifest as reduced C1 energy, subsequently lowering S_Heart.

Hypothesis H2: Kidney-Filtration Correspondence

H2: After controlling for confounding variables, S_Kidney shows a significant positive correlation with eGFR (expected ΔR² of 3–10%, expected β of 0.03–0.10).

Mechanism Basis: Higher-order harmonics (C6–C10) correlate with peripheral arteriolar resistance and microcirculatory compliance, while glomerular filtration rate is regulated by afferent arteriolar resistance. Abnormal pulse wave transit velocities in Chronic Kidney Disease (CKD) patients have been confirmed by multiple studies (PMID 36294440).

Hypothesis H3: Spleen-Metabolism Correspondence

H3: After controlling for confounding variables, S_Spleen shows a significant negative correlation with metabolic indicators (fasting glucose, HOMA-IR).

Mechanism Basis: In TCM theory, the Spleen governs transportation and transformation, conceptually corresponding to glucose metabolism and insulin sensitivity. Existing literature (PMID 35161551) indicates identifiable shift characteristics in pulse harmonic spectra among patients with metabolic syndrome.

Hypothesis with Expected Null Result

H4 (Expected null): After controlling for age, the independent explanatory power of S_Lung on FEV1/FVC may be non-significant, as lung function is more influenced by airway structure than hemodynamic harmonics.

This null result also holds scientific value—it delineates the effective boundary of the Five-Zang scoring system, suggesting that lung function may require integration of respiratory signals (e.g., respiratory rate variability) rather than reliance solely on pulse harmonics.

Evidence and Limitations

Dimension Assessment
Evidence Level B (Moderate) — Based on physiological mechanism derivation and small-scale validation literature; correlation modeling awaits large-scale multi-center verification
Key Supporting Literature PMID 33903610 (PPG cardiovascular harmonics and heart function association), PMID 36294440 (Pulse wave harmonic analysis and organ function correspondence), PMID 42305919 (Objective quantification of TCM syndromes), PMID 36509825 (Wearable PPG signal quality assessment), PMID 35161551 (Machine learning and TCM data mining)
Key Limitations Correlation does not imply causation; the Five-Zang scoring formula (Project #48) itself is hypothesis-driven with weights pending optimization; atherosclerosis in the elderly may interfere with the true correspondence between C1–C5 and organ function
Falsification Conditions If, after controlling for age/gender/BMI, the ΔR² for all corresponding O_i indicators is < 2% and β is non-significant, the hypothesis is falsified

Related Reading

References

  1. Park J, Seok HS, Kim SS, Shin H. Photoplethysmogram Analysis and Applications: An Integrative Review. Front Physiol. 2022;12:808451. PMID: 35300400
  2. Wang WG, et al. [Research on Pulse Wave Harmonics and Organ Resonance]. Relevant cardiovascular harmonic analysis literature. PMID: 33903610
  3. Research on Pulse Wave Harmonic Analysis and Heart Function Association. PMID: 36294440
  4. Research on Objective Quantification of TCM Syndromes and Multi-dimensional Vector Space Mapping. PMID: 42305919
  5. Wearable PPG Signal Quality Assessment and Cardiovascular Function Monitoring. PMID: 36509825
  6. Comprehensive Analysis of Machine Learning Methods and TCM Data Mining. PMID: 35161551

Disclaimer: This article compiles traditional knowledge and scientific literature for comparison and research-design discussion only. It is not medical advice and cannot replace professional diagnosis or treatment. Effect sizes, frequency ranges and exposure conclusions cited here are largely inferences at the research-report level, and individual circumstances vary widely. If you have any discomfort or health concerns, please consult a physician.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *