5D Library › 自主神经交感迷走平衡、六经辨证状态空间与多模态标志物反演 · English

第一章 心率变异谱与「交感迷走平衡」的局限

从心跳间期到频谱

心率变异性(HRV)研究相邻心跳间期(RR 间期)的起伏。1996 年的国际标准把五分钟记录的频谱分成低频(LF,0.04 至 0.15 Hz)与高频(HF,0.15 至 0.40 Hz),并定义了 RMSSD(相邻间期差的均方根)等时域指标 (Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology 1996; Shaffer and Ginsberg 2017)。高频成分主要来自呼吸窦性心律不齐,由迷走神经介导;低频成分的来源则复杂得多。

下面用合成数据把这些量算一遍。每一段五分钟的心跳由平均间期、随呼吸起伏的迷走波、0.1 Hz 的压力反射(迈尔)波与随机噪声组成,再以 4 Hz 重采样、做离散傅里叶变换。

推导

# Chapter 1. p01: Synthetic RR intervals and their HRV spectrum. Five
# minutes of beats are built from a mean RR, a respiratory (vagal) wave at
# the breathing rate, a 0.1 Hz baroreflex (Mayer) wave and white noise.
# The RR series is resampled at 4 Hz and its power is read with a plain
# DFT in the LF band (0.04-0.15 Hz) and HF band (0.15-0.40 Hz).
# RMSSD is the root mean square of successive RR differences. Seed 86.
import random
from math import cos, pi, sin, sqrt

def beats(rr0, a_resp, f_resp, a_mayer, noise, rng, secs=300.0):
    t, out = 0.0, []
    while t < secs:
        rr = (rr0 + a_resp * sin(2 * pi * f_resp * t)
              + a_mayer * sin(2 * pi * 0.1 * t) + rng.gauss(0, noise))
        out.append((t, rr))
        t += rr / 1000.0
    return out

def resample(series, fs=4.0):
    ts, rr = [t for t, _ in series], [r for _, r in series]
    n, j, out = int(ts[-1] * fs), 0, []
    for k in range(n):
        x = k / fs
        while j < len(ts) - 2 and ts[j + 1] <= x:
            j += 1
        w = (x - ts[j]) / (ts[j + 1] - ts[j])
        out.append(rr[j] + w * (rr[j + 1] - rr[j]))
    m = sum(out) / len(out)
    return [v - m for v in out]

def band_power(x, fs, lo, hi):
    n, p = len(x), 0.0
    for k in range(1, n // 2):
        f = k * fs / n
        if lo <= f < hi:
            re = sum(x[i] * cos(2 * pi * k * i / n) for i in range(n))
            im = sum(x[i] * sin(2 * pi * k * i / n) for i in range(n))
            p += 2 * (re * re + im * im) / (n * n)
    return p

def summary(series):
    rr = [r for _, r in series]
    rmssd = sqrt(sum((b - a) ** 2 for a, b in zip(rr, rr[1:]))
                 / (len(rr) - 1))
    x = resample(series)
    lf, hf = band_power(x, 4.0, 0.04, 0.15), band_power(x, 4.0, 0.15, 0.40)
    return 60000 / (sum(rr) / len(rr)), rmssd, lf, hf

CASES = (("rest, 15 breaths/min", 900, 30, 0.25, 15),
         ("stress: vagal wave cut", 700, 8, 0.25, 15),
         ("slow breathing 6/min", 900, 30, 0.10, 15),
         ("weak baroreflex wave", 900, 30, 0.25, 4))

if __name__ == "__main__":
    rng = random.Random(86)
    print("   case                     HR   RMSSD   LF ms2   HF ms2  LF/HF")
    for name, rr0, ar, fr, am in CASES:
        hr, rmssd, lf, hf = summary(beats(rr0, ar, fr, am, 6.0, rng))
        print("   {:<23} {:4.0f}  {:6.1f}  {:7.1f}  {:7.1f}  {:5.2f}".format(
            name, hr, rmssd, lf, hf, lf / hf))
#    case                     HR   RMSSD   LF ms2   HF ms2  LF/HF
#    rest, 15 breaths/min      67    29.3    112.1    326.7   0.34
#    stress: vagal wave cut    86    11.6    121.2     36.7   3.30
#    slow breathing 6/min      67    19.8    959.9     11.9  80.66
#    weak baroreflex wave      67    28.7     13.5    331.6   0.04

静息、每分钟呼吸 15 次时,心率 67 次每分钟,RMSSD 29.3 毫秒,LF/HF 为 0.34。削弱迷走波、心率升到 86 次时,RMSSD 降到 11.6 毫秒,LF/HF 升到 3.30。这两行符合教科书的「交感迷走平衡」读法。后两行则不然:自主神经设定完全不变,只把呼吸放慢到每分钟 6 次,LF/HF 就升到 80.66;只削弱压力反射波,LF/HF 降到 0.04。

LF/HF 不是交感迷走平衡

推导

# Chapter 1. p02: Same heart, same nerves, different breathing. The
# autonomic settings of p01's resting case are held fixed (mean RR 900 ms,
# respiratory wave 30 ms, Mayer wave 15 ms) and only the breathing rate
# changes. Below 9 breaths a minute the respiratory wave falls inside the
# LF band, so LF/HF rises although no sympathetic input changed. Each rate
# is averaged over 5 recordings; seed 86.
import importlib.util
import random

def load(name):
    spec = importlib.util.spec_from_file_location(name, name + ".py")
    m = importlib.util.module_from_spec(spec)
    spec.loader.exec_module(m)
    return m

p01 = load("p01_rr_spectrum")

if __name__ == "__main__":
    rng = random.Random(86)
    print("   breaths/min  Hz     LF ms2   HF ms2  LF/HF  RMSSD")
    for bpm in (6, 7, 8, 9, 10, 12, 15):
        f = bpm / 60.0
        acc = [0.0, 0.0, 0.0]
        for _ in range(5):
            _, rmssd, lf, hf = p01.summary(
                p01.beats(900, 30, f, 15, 6.0, rng))
            acc = [acc[0] + lf / 5, acc[1] + hf / 5, acc[2] + rmssd / 5]
        print("   {:11d}  {:.3f}  {:7.1f}  {:7.1f}  {:5.2f}  {:5.1f}".format(
            bpm, f, acc[0], acc[1], acc[0] / acc[1], acc[2]))
#    breaths/min  Hz     LF ms2   HF ms2  LF/HF  RMSSD
#              6  0.100    969.9     11.4  85.14   19.8
#              7  0.117    535.9     10.2  52.57   17.1
#              8  0.133    522.7     12.1  43.37   18.8
#              9  0.150    118.2    410.6   0.29   20.4
#             10  0.167    116.2    401.6   0.29   21.9
#             12  0.200    117.4    381.7   0.31   25.3
#             15  0.250    111.7    339.7   0.33   29.7

把呼吸频率从每分钟 15 次逐步降到 6 次,其余一切不变:每分钟 9 次以上时 LF/HF 稳定在 0.29 至 0.33;降到 8 次时,呼吸波的频率(0.133 Hz)落入 LF 频带,LF/HF 跳到 43.37;每分钟 6 次时是 85.14。这正是 LF/HF 受到批评的原因:低频功率主要反映压力反射对心脏自主输出的调制,而不是心脏交感张力本身 (Goldstein et al. 2011; Billman 2013)。

自主神经设定不变时,LF/HF 随呼吸频率的变化(p02)。虚线为每分钟 9 次。

注

读 HRV 之前先记呼吸。 同一个人,慢呼吸、说话、打哈欠都会把能量移入 LF 频带。没有同步记录呼吸频率的 LF/HF,不能解读为交感或迷走的强弱;RMSSD 与 HF 对迷走调制的反映较为直接,但同样受呼吸深度影响。

小结

在示意模型中,不改变任何自主神经设定,只把呼吸放慢,LF/HF 就从 0.33 升到 85.14。LF/HF 不能单独代表「交感迷走平衡」;本书以下章节把心率、RMSSD 与呼吸频率一起使用,并把所有自主神经的解读都当作需要检验的模型。

参考文献

Billman, George E. 2013. “The LF/HF Ratio Does Not Accurately Measure Cardiac Sympatho-Vagal Balance.” Frontiers in Physiology 4: 26.
Goldstein, David S., Oladi Bentho, Mee-Yeong Park, and Yehonatan Sharabi. 2011. “Low-Frequency Power of Heart Rate Variability Is Not a Measure of Cardiac Sympathetic Tone but May Be a Measure of Modulation of Cardiac Autonomic Outflows by Baroreflexes.” Experimental Physiology 96 (12): 1255 to 1261.
Shaffer, Fred, and J. P. Ginsberg. 2017. “An Overview of Heart Rate Variability Metrics and Norms.” Frontiers in Public Health 5: 258.
Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. 1996. “Heart Rate Variability: Standards of Measurement, Physiological Interpretation and Clinical Use.” Circulation 93 (5): 1043 to 1065.

Fan, Y. (2026). Autonomic Sympathovagal Balance, Six-Meridian State-Space Models and Multimodal Biomarker Inversion: Heart-Rate Variability Spectral Decomposition, Thermoregulatory Phase Drift and Bayesian Syndrome Classification [自主神经交感迷走平衡、六经辨证状态空间与多模态标志物反演] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090227 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).