Evidence level: B
脉诊现代化的直觉是”用光电容积描记测脉搏变异性(PRV)替代心电变异性(HRV)”——但三质判定恰恰依赖频域指标,而 PRV 与 HRV 在频域的一致性最差:静息时高度可换,一旦进入应激与反应性窗口,两者分道扬镳,偏离到某个阈值之外就不可互换。
正文
上一篇文章我们建立了三质频率测量协议的四关(设备、操作者、模态、时长)。这一篇专攻其中”模态”一关最隐蔽的陷阱:PRV 与 HRV 的等效性边界。瓦塔、皮塔、卡法三质的判定依赖 LF/HF 等频域指标(PMID 36106418),而频域恰恰是 PRV 与 HRV 偏差最大的地带——这意味着”用什么测”本身就是三质分型里最大的潜在混杂因子。
静息态:PRV 高度接近 HRV,这是真的。 17 名年轻受试者倾斜试验的经典研究里,静息时 PRV 与 HRV 各指标相关高达 0.97,时频谱相关 0.99,LF/HF 频带时频相干分别达 0.97 与 0.89(PMID 20702919)。72 名儿童(含睡眠呼吸障碍者)同步记录也显示时域指标(SDNN、RMSSD)相关 r>0.90、一致性良好(PMID 24111246)。所以在”静息仰卧 5 分钟”这个三质筛查的默认窗口内,用指腹式 PPG 测 PRV 做筛查,科学上有立足点——这也是 #178 确立的时长协议能继续用的前提。
但一进入非静息状态,等效性就破裂。 76 人 Finometer 研究直接比较静息仰卧(SU)与直立倾斜(HUT)两种状态:整体一致性在 HUT 时明显下降,偏差最大的两个指标恰恰是带功率比(LF/HF)与条件熵——也就是说,体质判定最依赖的比值型频域指标,恰是应激状态下 PRV 最不可靠的指标(PMID 30441608)。20 人全身冷暴露研究更彻底:PRV 对外周血管收缩的反应与 HRV 截然不同,尤其指尖与脚趾等外周部位,Bland-Altman 显示两者的关系被冷暴露直接改写,核心血管在暴露结束后恢复更快(PMID 32792970)。一篇系统综述给出保守结论:静息时 PRV 足够准确,但运动与精神应激时一致性恶化到不可接受(PMID 22809539)。
机制层给出了为什么——PRV 从来就不是劣质 HRV,而是另一个信号。 从心电 R 波到 PPG 脉搏波要经过电机耦合、压力-容积、脉冲传导等六步物理转换,每一步都有自己的延迟与频率特性。日本团队用起搏器患者做了一次”思想实验”般的演示:在 HRV 完全为零(固定起搏频率、频谱平坦)的人身上,PRV 依然存在——其来源是搏动传递时间(PTT)的呼吸性振荡,而 PTT 振荡在 0.3Hz 处会被放大 2.0 倍(PMID 32811571)。这就是为什么站立与运动时 PRV 的 HF 功率会系统性膨胀:呼吸频率升高,PTT 振荡被传递函数放大,PRV 里混入了 HRV 没有的成分。结论是革命性的:PRV 与 HRV 的偏离不是测量误差,而是外周血管、PTT 与呼吸耦合携带的独立生理信息——它们不是同一个东西的两次测量,而是两个东西的各一次测量。
这对三质判定意味着什么? 第一,静息分类任务里,用 HRV 还是 PRV 必须二选一并全程固定——因为两者的系统偏差虽小但真实存在,混用等于给体质差异叠加一层模态噪声。第二,一旦引入反应性窗口(如 #337 的直立负荷 3 分钟反应性 HRV),必须改用 ECG 心电模态,因为应激态正是 PRV 失效区。第三,可穿戴设备做体质趋势追踪可以接受(同一设备同一条件纵向比较),但跨设备、跨状态的绝对数值对比不可行——可穿戴 PPG 文献连统一的准确度报告标准都还没有(PMID 36288566)。第四,积极面:如果 PRV 携带外周血管张力与 PTT 的独立信息,那”PRV-HRV 差值”本身可能就是一个新的体质判别通道——传统脉诊”搭脉搏”测的正是血管壁上的脉搏波,而非心电,这个”差值通道”说不定比纯 HRV 更贴近阿育吠陀原意。
落地协议:在同一批三质人群上,ECG 与多部位 PPG(指/腕/耳)同步采集,在静息/站立/轻度应激三状态下,逐指标做 Bland-Altman 百分比偏差标定,定义”不可互换阈值”——即偏差超过某界限时三质分类 AUC 跌破临床可用线。阈值标定出来之前,三质研究默认用 HRV,用 PRV 必须声明模态并附一致性证据。
证据与局限
已验证: 静息态 PRV 与 HRV 时域指标高度一致、相关>0.9(PMID 24111246, 20702919);应激/体位/冷暴露下 PRV 与 HRV 一致性恶化、频域比值型指标偏差最大(PMID 30441608, 32792970, 22809539);PRV 与 HRV 的偏离源于 PTT 呼吸性振荡经传递函数放大,PRV 是独立生物标志物(PMID 32811571)。
部分验证: PRV 外周血管信息(冷暴露反应、PTT)确实与 HRV 不同,但其”独立信息量”能否用于三质判别尚无实证;可穿戴 PPG 的纵向个体内追踪被认为可用,但报告标准缺失(PMID 36288566)。
待验证: 三质判定场景下 PRV 与 HRV 的”不可互换阈值”(Bland-Altman 偏差百分比 + 分类 AUC 双终点)尚无直接标定;”PRV-HRV 差值通道”是否携带独立于 HRV 的体质量信息;反应性窗口(HUT/应激)中 PRV 偏差超出可接受界限的定量边界。
相关阅读
证据等级: B
参考文献: PMID 20702919, PMID 24111246, PMID 30441608, PMID 32792970, PMID 22809539, PMID 32811571, PMID 36288566
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。
Equivalence Thresholds of PRV and HRV in Three-Phase Determination: When the Pulse Cannot Impersonate the Heartbeat
Evidence level: B
The intuitive modernization of pulse diagnosis is the proposal to “replace Heart Rate Variability (HRV) with Pulse Rate Variability (PRV) measured via photoplethysmography”—yet the three-body qualification relies precisely on frequency-domain metrics, where the consistency between PRV and HRV is poorest. They are highly interchangeable at rest; however, they diverge significantly once entering stress and reactivity windows. Beyond a specific threshold, they become non-interchangeable.
Main Text
In the previous article, we established a four-stage protocol for three-body frequency measurement (equipment, operator, modality, duration). This article focuses on the most subtle trap within the “modality” stage: the equivalence boundary between PRV and HRV. The qualification of the three bodies (Vata, Pitta, Kapha) relies on frequency-domain indices such as LF/HF (PMID 36106418). Since the frequency domain is precisely where the deviation between PRV and HRV is greatest, this means “what is measured” constitutes the largest potential confounding factor within the three-body typology.
Resting State: PRV closely mirrors HRV; this is true. In a classic study involving 17 young subjects undergoing tilt testing, correlations between PRV and HRV indices at rest reached as high as 0.97; time-frequency correlations were 0.99, and band-specific time-frequency coherence for LF/HF reached 0.97 and 0.89 respectively (PMID 20702919). A concurrent recording study of 72 children (including those with sleep-disordered breathing) also demonstrated strong consistency in time-domain indices (SDNN, RMSSD), with correlations $r > 0.90$ (PMID 24111246). Therefore, within the default screening window of “supine rest for 5 minutes” used in three-body screening, using finger-plethysmography to measure PRV for screening has scientific grounding—this is also a prerequisite for the duration protocol established in #178 to remain valid.
However, once entering a non-resting state, equivalence breaks down. A study of 76 individuals using Finometer directly compared two states: supine rest (SU) and upright tilt (HUT). Overall consistency decreased significantly during HUT, with the two indices showing the largest deviation being the power ratio (LF/HF) and conditional entropy. This means that the ratio-based frequency-domain indices most relied upon for body constitution qualification are precisely the least reliable indicators for PRV under stress conditions (PMID 30441608). A more comprehensive study involving 20 subjects exposed to whole-body cold revealed a distinct difference: PRV reacts differently to peripheral vasoconstriction compared to HRV, particularly at peripheral sites like fingertips and toes. Bland-Altman analysis showed that the relationship between the two was directly rewritten by cold exposure, with core vessels recovering faster after exposure ended (PMID 32792970). A systematic review offered a conservative conclusion: PRV is sufficiently accurate at rest, but consistency deteriorates to unacceptable levels during exercise and mental stress (PMID 22809539).
The mechanism layer explains why—PRV is never inferior HRV, but rather a distinct signal. From the ECG R-wave to the PPG pulse wave involves six physical conversion steps including electromechanical coupling, pressure-volume dynamics, and pulse transmission, each with its own latency and frequency characteristics. A Japanese team conducted a demonstration akin to a “thought experiment” using pacemaker patients: in individuals where HRV was completely zero (fixed pacing frequency, flat spectrum), PRV still existed. Its source was the respiratory oscillation of Pulse Transit Time (PTT), which is amplified 2.0 times at 0.3Hz (PMID 32811571). This explains why PRV HF power systematically expands during standing and exercise: increased respiratory frequency amplifies the PTT oscillation via the transfer function, introducing components into PRV that HRV does not contain. The conclusion is revolutionary: the deviation between PRV and HRV is not measurement error, but independent physiological information carried by peripheral vasculature, PTT, and respiratory coupling—they are not two measurements of the same thing, but rather one measurement each of two different things.
What does this mean for three-body qualification? First, in resting classification tasks, one must choose either HRV or PRV and maintain consistency throughout, as their systematic biases are small but real; mixing them adds a modality noise layer to constitution differences. Second, once a reactivity window is introduced (e.g., the 3-minute reactivity HRV under upright load in #337), ECG modality must be switched, as the stress state is precisely where PRV fails. Third, wearable devices can acceptably track constitution trends longitudinally (comparisons under the same device and conditions), but absolute numerical comparisons across devices and states are not feasible—wearable PPG literature lacks unified accuracy reporting standards (PMID 36288566). Fourth, a positive aspect: if PRV carries independent information regarding peripheral vascular tone and PTT, then the “PRV-HRV difference” itself may constitute a new constitution discrimination channel. Traditional pulse diagnosis (“palpating the pulse”) measures the pulse wave on the vessel wall, not the ECG; this “difference channel” might actually be closer to the original Ayurvedic intent than pure HRV.
Implementation Protocol: On the same cohort of three-body subjects, collect data synchronously using ECG and multi-site PPG (finger/wrist/ear) across three states: rest, standing, and mild stress. Perform Bland-Altman percentage bias calibration for each index to define an “non-interchangeable threshold”—defined as the point where deviation exceeds a specific limit and the three-body classification AUC drops below the clinically usable line. Before this threshold is calibrated, three-body studies should default to HRV; if PRV is used, the modality must be declared and consistency evidence provided.
Evidence and Limitations
Verified: PRV and HRV show high consistency in time-domain indices at rest with correlations >0.9 (PMID 24111246, 20702919); under stress/posture/cold exposure, consistency deteriorates, with ratio-based frequency-domain indices showing the largest deviation (PMID 30441608, 32792970, 22809539); the deviation between PRV and HRV stems from the amplification of respiratory PTT oscillations via a transfer function, establishing PRV as an independent biomarker (PMID 32811571).
Partially Verified: PRV does indeed carry peripheral vascular information (cold exposure response, PTT) distinct from HRV, but empirical evidence is lacking regarding whether this “independent information volume” can be utilized for three-body discrimination; longitudinal within-subject tracking via wearable PPG is considered feasible but lacks reporting standards (PMID 36288566).
To Be Verified: The “non-interchangeable threshold” for PRV vs. HRV in three-body qualification scenarios (dual endpoints of Bland-Altman percentage bias + classification AUC) has not been directly calibrated; whether the “PRV-HRV difference channel” carries body mass information independent of HRV; and the quantitative boundary for PRV deviation exceeding acceptable limits within reactivity windows (HUT/stress).
Related Reading
- Multicenter Reproducibility of the Three-Body Frequency Measurement Protocol
- Shortest Valid Recording Duration Protocol for Three-Body Frequency Measurement
- Generalization Validation of Three-Body HRV Spectral Features in Non-Indian Populations
- Research Pool: Cross-Cultural Frequency Cognition
Evidence Level: B
References: PMID 20702919, PMID 24111246, PMID 30441608, PMID 32792970, PMID 22809539, PMID 32811571, PMID 36288566
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.
Leave a Reply