关于「耐受—致癌悖论:一条单调上升的剂量曲线」,《红脸基因与宴请血醇曲线》给出的定量结论是什么?
结论:第二章正面陈述本书要推翻的常识,给出按饮酒层合并的单调梯度,每周每 10 个标准杯把比值比乘以 1.891,重度层的 7.12 对应的 E 值为 13.72;同时并陈反方证据:不饮酒的携带者没有可检出的超额风险。
数据口径:本书用纯标准库的程序把每一步写成可以核对的算术:参数凡未注明出处者均为示意,所有数字由书中程序生成。 本书不是医疗建议,不提供任何形式的安全饮用量;书中的血醇曲线用于安排用车与住宿,绝不可用于判断能否驾驶——饮酒之后不驾车。
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: 1.891, 10, 7.12
ALDH2*1/*2 vs *1/*1, oesophageal squamous cell carcinoma stratum units/wk OR low high logOR SE never or rare 0.0 1.28 0.91 1.80 0.247 0.174 moderate 13.5 3.12 1.95 5.01 1.138 0.241 heavy 27.0 7.12 4.67 10.86 1.963 0.215 weighted log-linear trend in weekly units slope 0.06368 per unit, SE 0.01020, z 6.25 odds multiplier per 10 units a week: 1.891 (1.548 to 2.309) two-sided p for no trend: 4.217e-10 the same carrier status in people who do not drink never or rare drinkers OR 1.28, 95 pct CI 0.91 to 1.80 z against the null 1.42; the interval contains 1 Lewis and Davey Smith pooled heterozygote OR 3.19 (1.86 to 5.47) their *2/*2 homozygotes, who can hardly drink: 0.36 (0.16 to 0.80) Yang 2010 heterozygote overall, crude: 2.52 (1.76 to 3.61) Yang 2010 *2/*2 who do drink: 4.42 (1.72 to 11.36) the two *2/*2 figures disagree and are shown together Brooks 2009 adjusted heterozygote ORs: 3.7 to 18.1
查看完整验算程序
# Chapter 2. p21: The published dose-response for carriers who drink.
# Pooled odds ratios for oesophageal squamous cell carcinoma in
# ALDH2*1/*2 heterozygotes against ALDH2*1/*1, by drinking stratum,
# as reported in the meta-analysis of Yang and colleagues (2010,
# World J Gastroenterol 16(33):4210); the overall heterozygote
# estimate of Lewis and Davey Smith (2005, Cancer Epidemiol
# Biomarkers Prev 14(8):1967) is shown for comparison. Standard
# errors are recovered from the published confidence intervals on
# the log scale. The weekly drink midpoints are illustrative labels
# for the strata, not figures from either paper.
import math
from statistics import NormalDist
Z = NormalDist().inv_cdf(0.975)
STRATA = [("never or rare", 0.0, 1.28, 0.91, 1.80),
("moderate", 13.5, 3.12, 1.95, 5.01),
("heavy", 27.0, 7.12, 4.67, 10.86)]
N = NormalDist()
print("ALDH2*1/*2 vs *1/*1, oesophageal squamous cell carcinoma")
print("stratum units/wk OR low high logOR SE")
for name, dose, orv, lo, hi in STRATA:
se = (math.log(hi) - math.log(lo)) / (2.0 * Z)
print("{:14s} {:8.1f} {:6.2f} {:6.2f} {:6.2f} {:6.3f} {:6.3f}"
.format(name, dose, orv, lo, hi, math.log(orv), se))
# inverse-variance weighted line through log OR against weekly units
xs = [s[1] for s in STRATA]
ys = [math.log(s[2]) for s in STRATA]
ws = []
for _, _, orv, lo, hi in STRATA:
se = (math.log(hi) - math.log(lo)) / (2.0 * Z)
ws.append(1.0 / se ** 2)
sw = math.fsum(ws)
mx = math.fsum([w * x for w, x in zip(ws, xs)]) / sw
my = math.fsum([w * y for w, y in zip(ws, ys)]) / sw
num = math.fsum([w * (x - mx) * (y - my) for w, x, y in zip(ws, xs, ys)])
den = math.fsum([w * (x - mx) ** 2 for w, x in zip(ws, xs)])
b = num / den
se_b = math.sqrt(1.0 / den)
print("")
print("weighted log-linear trend in weekly units")
print("slope {:.5f} per unit, SE {:.5f}, z {:.2f}"
.format(b, se_b, b / se_b))
print("odds multiplier per 10 units a week: {:.3f} ({:.3f} to {:.3f})"
.format(math.exp(10.0 * b), math.exp(10.0 * (b - Z * se_b)),
math.exp(10.0 * (b + Z * se_b))))
print("two-sided p for no trend: {:.3e}"
.format(2.0 * (1.0 - N.cdf(abs(b / se_b)))))
print("")
print("the same carrier status in people who do not drink")
se0 = (math.log(1.80) - math.log(0.91)) / (2.0 * Z)
print("never or rare drinkers OR {:.2f}, 95 pct CI {:.2f} to {:.2f}"
.format(1.28, 0.91, 1.80))
print("z against the null {:.2f}; the interval contains 1"
.format(math.log(1.28) / se0))
print("Lewis and Davey Smith pooled heterozygote OR 3.19 "
"(1.86 to 5.47)")
print("their *2/*2 homozygotes, who can hardly drink: 0.36 "
"(0.16 to 0.80)")
print("Yang 2010 heterozygote overall, crude: 2.52 (1.76 to 3.61)")
print("Yang 2010 *2/*2 who do drink: 4.42 (1.72 to 11.36)")
print("the two *2/*2 figures disagree and are shown together")
print("Brooks 2009 adjusted heterozygote ORs: 3.7 to 18.1")唯一数据源与引用
范雅福(Fan, Y.). (2026). 《乙醛脱氢酶二型红脸基因、酒量耐受致癌悖论与商务宴请血醇曲线母典》第 2 章「耐受—致癌悖论:一条单调上升的剂量曲线」(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-191 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026191,
author = {Fan, Yafu},
title = {ALDH2 Flushing Genetics, the Tolerance-Carcinogenesis Paradox and Blood-Alcohol Curves at Business Banquets},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}