关于「MET 分钟的分段对数线性:剂量—反应的拐点与天花板」,《最大摄氧量、二区与五禽戏》给出的定量结论是什么?

结论:第三章用六十六万人的合并数据拟合两个拐点的分段对数线性曲线,第一段每 MET 小时风险比为 0.9431。

数据口径:书中参数凡未注明出处者均为示意;全部计算由书中程序生成;本书不是医疗建议,不宣称任何活动对任何疾病有作用,自我试验须先咨询执业医师。 第六章为五禽戏估出示意的代谢当量,一轮约 43.75 MET 分钟,并排出周剂量。

English version

对照数据表

下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 0.9431

best knots: k15, k2 = 20 MET-h/week; chi2 = 3.04
HR per +1 MET-h/week below k10.9431
HR per +1 MET-h/week k1 to k20.9872
HR per +1 MET-h/week above k21.0001
dose observed fitted
3.750.80 0.803
11.250.69 0.688
18.750.63 0.625
31.250.61 0.616
57.500.61 0.617
90.000.68 0.618
fitted HR at 7.5 MET-h/week0.722
查看完整验算程序
# Chapter 3. p31: A piecewise log-linear dose-response for leisure-time
# activity. Data: Arem et al. (2015, JAMA Intern Med; 661,137 adults,
# 116,686 deaths), HRs against 0 MET-h/week for 0.1-<7.5, 7.5-<15,
# 15-<22.5, 22.5-<40, 40-<75 and 75+ MET-h/week. Doses are category
# midpoints (90 for the open top class, illustrative). Model:
#   log HR(x) = b1*min(x,k1) + b2*clip(x-k1, 0, k2-k1) + b3*max(x-k2, 0)
# with HR(0) = 1, fitted by inverse-variance weighted least squares over a
# grid of knots (k1, k2); the 3x3 normal equations are solved exactly.
import math
from fractions import Fraction
from statistics import NormalDist

Z = NormalDist().inv_cdf(0.975)
DATA = [(3.75, 0.80, 0.78, 0.82), (11.25, 0.69, 0.67, 0.70),
        (18.75, 0.63, 0.62, 0.65), (31.25, 0.61, 0.59, 0.62),
        (57.5, 0.61, 0.58, 0.64), (90.0, 0.68, 0.59, 0.78)]


def basis(x, k1, k2):
    return [min(x, k1), min(max(x - k1, 0.0), k2 - k1), max(x - k2, 0.0)]


def solve3(a, b):
    m = [[Fraction(v) for v in row] + [Fraction(r)]
         for row, r in zip(a, b)]
    for c in range(3):
        p = max(range(c, 3), key=lambda r: abs(m[r][c]))
        m[c], m[p] = m[p], m[c]
        for r in range(3):
            if r != c and m[c][c] != 0:
                f = m[r][c] / m[c][c]
                m[r] = [u - f * v for u, v in zip(m[r], m[c])]
    return [float(m[i][3] / m[i][i]) for i in range(3)]


def fit(k1, k2):
    rows = []
    for x, hr, lo, hi in DATA:
        se = (math.log(hi) - math.log(lo)) / (2 * Z)
        rows.append((basis(x, k1, k2), math.log(hr), 1 / se ** 2))
    a = [[math.fsum(w * f[i] * f[j] for f, y, w in rows)
          for j in range(3)] for i in range(3)]
    b = [math.fsum(w * f[i] * y for f, y, w in rows) for i in range(3)]
    beta = solve3(a, b)
    chi = math.fsum(w * (y - math.fsum(bi * fi for bi, fi in zip(beta, f)))
                    ** 2 for f, y, w in rows)
    return beta, chi


def loghr(x, beta, k1, k2):
    return math.fsum(b * f for b, f in zip(beta, basis(x, k1, k2)))


if __name__ == "__main__":
    best = None
    for i in range(2, 31):
        k1 = 1.0 * i
        for j in range(2, 18):
            k2 = 5.0 * j
            if k2 <= k1 + 5:
                continue
            beta, chi = fit(k1, k2)
            if best is None or chi < best[0]:
                best = (chi, k1, k2, beta)
    chi, k1, k2, beta = best
    print("best knots: k1 = %.0f, k2 = %.0f MET-h/week; chi2 = %.2f"
          % (k1, k2, chi))
    for n, b in zip(("below k1", "k1 to k2", "above k2"), beta):
        print("HR per +1 MET-h/week %-9s %.4f" % (n, math.exp(b)))
    print(" dose   observed   fitted")
    for x, hr, lo, hi in DATA:
        print("%5.2f   %8.2f   %6.3f"
              % (x, hr, math.exp(loghr(x, beta, k1, k2))))
    print("fitted HR at 7.5 MET-h/week: %.3f"
          % math.exp(loghr(7.5, beta, k1, k2)))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《最大摄氧量死亡率剂量响应、二区有氧乳酸阈与五禽戏间歇代谢当量母典》第 3 章「MET 分钟的分段对数线性:剂量—反应的拐点与天花板」(版本 1.0.0). 5D Library. DOI 待发布

DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-187 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026187,
  author    = {Fan, Yafu},
  title     = {The Mortality Dose-Response of VO2max, Zone-2 Lactate Thresholds and the Metabolic Equivalents of the Five-Animal Frolics},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论