关于「三种机制的竞争:可辨识性、敏感度与人群账」,《照度剂量与屈光发育》给出的定量结论是什么?
结论:当亮、远、开阔总是同时出现时,观察数据只能确认户外有用,不能指出哪一部分有用;模拟显示,在观察世界里错误的预测量也能解释可观的方差,而在四臂设计中,两种真相留下几乎不可混淆的指纹。人群账对覆盖率最敏感:按广州效应,每千名儿童三年少发 20.8 到 91.7 例,取决于实际剂量的覆盖;单一干预无法把高发地区的发病率压到 0.30 以下。下一章回到两部原典,看古人对眼与光说对了什么。
数据口径:书中儿童、日程与灯具参数为示意;全部计算由书中程序生成;本书不是医疗建议,不对任何个人作效果承诺,自我试验须先咨询执业医师。 第三章写出眼轴生长的线性混合模型,模拟中纵向估计每次都能检出照度效应,横断面估计只有 0.110 的机会。
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。 与结论吻合的打印值: 0.30, 20.8, 91.7
| cases of new myopia averted per 1000 children over 3 years | |
| coverage RR 0.768 (2015) IRR 0.84 (2022) | |
| 0.2273 | 20.8 14.4 |
| 0.4979 | 45.7 31.5 |
| 0.7500 | 68.8 47.4 |
| 1.0000 | 91.7 63.2 |
| relative cut at full execution with RR 0.768 | 0.232 |
| relative cut needed to hold 3-year incidence at 0.30 | 0.241 |
| coverage needed for that with RR 0.768 | 1.039 |
| if the 2015 relative cut applied to every incident case, the | |
| 2050 projection 0.498 would fall to about 0.382 |
查看完整验算程序
# Chapter 6. p62: Population arithmetic and a sensitivity scan. Control
# 3-year incidence 0.3953 and risk ratio 0.768 come from p11 (He et al.
# 2015); the incidence rate ratio 0.84 is the 40-minute arm of He et al.
# (2022). Global prevalence projected for 2050 is 0.498 (Holden et al.
# 2016). Coverage is the share of children whose days really reach the
# illuminance dose; the Taiwan trial (p23) reached 0.4979. We assume the
# effect scales linearly with coverage (illustrative). Holden et al.
# projected 4758 million people with myopia in 2050 (0.498 of the
# world) and 938 million with high myopia (0.098), against 1406
# million (0.229) in 2000.
BASE = 0.3953
EFFECTS = (("RR 0.768 (2015)", 0.768), ("IRR 0.84 (2022)", 0.84))
COHORT = 1000
print("cases of new myopia averted per {} children over 3 years"
.format(COHORT))
print(" coverage " + " ".join("{:>15s}".format(n) for n, _ in EFFECTS))
for cov in (0.2273, 0.4979, 0.75, 1.0):
cells = []
for _, rr in EFFECTS:
eff_rr = 1 - cov * (1 - rr)
cells.append(COHORT * BASE * (1 - eff_rr))
print(" {:8.4f} ".format(cov) + " ".join(
"{:15.1f}".format(c) for c in cells))
print("relative cut at full execution with RR 0.768: {:.3f}".format(
1 - 0.768))
print("relative cut needed to hold 3-year incidence at 0.30: {:.3f}"
.format(1 - 0.30 / BASE))
print("coverage needed for that with RR 0.768: {:.3f}".format(
(1 - 0.30 / BASE) / (1 - 0.768)))
p2050 = 0.498
print("if the 2015 relative cut applied to every incident case, the")
print("2050 projection {:.3f} would fall to about {:.3f}".format(
p2050, p2050 * 0.768))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《户外照度剂量近视抑制、调节滞后与周边离焦光学的屈光发育母典》第 6 章「三种机制的竞争:可辨识性、敏感度与人群账」(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-189 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026189,
author = {Fan, Yafu},
title = {Outdoor Illuminance Dose, Accommodative Lag and Peripheral-Defocus Optics in Refractive Development},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}