关于「可训练性与衰老斜率:最大摄氧量的个体差异与测量噪声」,《最大摄氧量、二区与五禽戏》给出的定量结论是什么?

结论:第五章说明单次测试会把训练无反应者的比例从 0.107 夸大到 0.196。

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: 0.107, 0.196

teststrue nonseen nontruly nonSD obsSD ctrlSD recr
10.1070.1960.3572.892.131.960.685
20.1070.1600.4712.501.502.010.799
查看完整验算程序
# Chapter 5. p52: How many "non-responders" are measurement noise? True
# 20-week gains in VO2max are drawn from N(3.5, 2.0) mL/kg/min, and every
# test adds an independent error with SD 1.5 (both illustrative; the
# HERITAGE family study, Bouchard et al. 1999, motivates a wide true
# spread). A person counts as a non-responder when the gain is below 1.0.
# We compare the truth with one test before and after, and with two tests
# averaged at each end, and recover the true spread from a control arm.
# Fixed seed; 50,000 trained people and 50,000 controls. HERITAGE found
# 2.6 times more variance between families than within and a maximal
# heritability of 47 %. Pre-registered bar (Chapter 5): a true share of
# non-responders at most 0.15; a share above 0.2 refutes it.
import math
import random
import statistics

rng = random.Random(20261112)
N_SIM = 50000
MU, SD_TRUE, TE, CUT = 3.5, 2.0, 1.5, 1.0


def run(k_tests):
    true, obs, ctrl = [], [], []
    for _ in range(N_SIM):
        g = rng.gauss(MU, SD_TRUE)
        pre = statistics.fmean(rng.gauss(0, TE) for _ in range(k_tests))
        post = statistics.fmean(rng.gauss(0, TE) for _ in range(k_tests))
        true.append(g)
        obs.append(g + post - pre)
        c_pre = statistics.fmean(rng.gauss(0, TE) for _ in range(k_tests))
        c_post = statistics.fmean(rng.gauss(0, TE) for _ in range(k_tests))
        ctrl.append(c_post - c_pre)
    t_non = len([g for g in true if g < CUT]) / N_SIM
    o_non = len([g for g in obs if g < CUT]) / N_SIM
    both = len([1 for g, o in zip(true, obs) if g < CUT and o < CUT])
    ppv = both / max(1, len([o for o in obs if o < CUT]))
    sd_o, sd_c = statistics.pstdev(obs), statistics.pstdev(ctrl)
    rec = math.sqrt(max(0.0, sd_o ** 2 - sd_c ** 2))
    r = statistics.correlation(true, obs)
    return t_non, o_non, ppv, sd_o, sd_c, rec, r


if __name__ == "__main__":
    print("tests  true non  seen non  truly non  SD obs  SD ctrl  SD rec"
          "  r")
    for k in (1, 2):
        t, o, p, so, sc, rc, r = run(k)
        print("%5d  %8.3f  %8.3f  %9.3f  %6.2f  %7.2f  %6.2f  %.3f"
              % (k, t, o, p, so, sc, rc, r))

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 5 章「可训练性与衰老斜率:最大摄氧量的个体差异与测量噪声」(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