Evidence level: A
出生月份对疾病风险的贡献在群体层面真实存在,但在个体判别力上远不如家族史与多基因风险评分——这个头对头对比,恰好为”生辰能定个体命运”的断言划出了可证伪的量化边界。
正文
八字命理的核心假设是:一个人出生的时刻(精确到时辰)承载了足以判定其一生吉凶祸福的信息。上一篇我们已经论证,出生季节效应(OR≈1.05,AUC≈0.508)与命理判定所需的信息量之间存在 3-4 个数量级的鸿沟。但严谨的检验需要一个”对照组”:把出生时刻放进同一个坐标系,与医学界公认的两大风险因子——家族史和多基因风险评分(PRS)——做头对头对比。如果连这两个公认指标在个体判别上都只有中低水平,那么效应量远小于它们的”出生时刻”,凭什么宣称能判定个体命运?
家族史 vs PRS:现代遗传学的系统测量
先看两个公认强预测因子彼此如何比较。Mars 等人(2022)在 FinnGen 队列(N=306,418)中对 24 种常见疾病系统对比了一级/二级家族史与全基因组 PRS,结论非常明确:两者独立且互补,不可互换。PRS 平均只解释了家族史效应的 10%,家族史只解释 PRS 效应的 3%;在冠心病、青光眼、2 型糖尿病等大多数疾病中,高 PRS 可以完全”补偿”掉阳性家族史带来的风险。换句话说,即使在遗传学内部,两个公认指标也各管一段,谁都无法单凭自己锁定个体。
更直接的判别力数据来自北欧注册研究(Lu 等人,2018):在 5,959 例精神分裂症患者与 8,717 名对照中,GRS 与家族史的联合模型判别能力”接近临床可用性的最低阈值“——两强联手,也才刚刚摸到”勉强能用”的边。2026 年 BMC Medicine 的 All of Us 研究(N>120,000)进一步量化:PRS 只中介了家族史总效应的 13-16%,两者叠加才能把风险分层拉开到 2.5 倍以上。
出生季节效应在这个坐标系里的位置
现在把三方摆到一起:出生季节效应的 OR≈1.05-1.10,PRS 最高十分位人群的 OR≈3-4,一级家族史的 OR 依疾病在 1.5-2.5 之间。判别力(AUC)方面,精神分裂症 PRS 约 0.65-0.70,出生月份效应约 0.508。这个坐标系说明三件事:
第一,出生季节效应与遗传风险在机制上相互独立。Escott-Price 等人(2019)在 UK Biobank 136,538 人中直接测量了精神分裂症 PRS 与出生季节的相关性,结果完全不相关——出生季节效应不是基因-环境混杂的假象,而是真实的环境暴露效应。这为”出生时间承载环境信息”的直觉提供了科学支撑,但它的量级决定了它只配当群体流行病学线索,不配当个体判定的依据。
第二,从群体关联到个体判别,隔着一条信息量鸿沟。SeaWAS 扫描(Boland 2015,174.9 万人)发现 55 种疾病与出生月份显著相关,独立复制研究也证实了心血管疾病的风险峰值(Li 2016)——但”有相关”和”能判个体”是两码事。效应量每小一个数量级,判别力就断崖式下跌:OR 1.05 换算成 AUC 只有 0.508,比掷硬币好不到一个百分点,而 PRS 的 0.65 已经需要数十万人的 GWAS 才能堆出来。
第三,对命理学的检验含义:如果”出生时间影响命运”要走科学路线,它必须与家族史、PRS 站在同一个 AUC/净重分类指数(NRI)/净获益的擂台上,用同一个队列、同一套结局做头对头对比。命理体系若有真实的个体预测力,至少应当展现出与”一级家族史+高 PRS”相当的判别增量——而不是只停留在”冬天出生的人精神分裂症风险高 5%”这种群体水平。
可执行的实验设计
检验方案清晰可行:取一个含出生日期、基因型、家族史与多年随访结局的大型队列(UK Biobank 或 FinnGen 皆可),建立四个预测模型——(1) 仅出生季节/月份;(2) 家族史;(3) PRS;(4) 三者联合。对比 AUC 增量、NRI 与决策曲线净获益。预期结果:出生季节模型的 AUC 增量接近零,家族史与 PRS 各自贡献 0.01-0.02 的增量,联合模型达到最佳。若命理断言为真,其预测增量应显著超过随机——这就是可证伪的判据,也正是”生辰断命”从信仰走向科学的必经之路。
证据与局限
已验证: 出生季节效应为真实环境效应,与遗传风险相互独立(PMID 29501066);家族史与 PRS 在 24 种疾病中独立互补、不可互换(PMID 36347255);GRS+家族史联合模型判别力仅接近临床最低阈值(PMID 28942743);出生月份与 55 种疾病显著相关且可独立复制(PMID 26041386, 27624541);PRS 仅中介家族史效应的 13-16%(PMID 41629984)。
部分验证: 出生季节效应在心血管死亡中的微弱关联(春季 HR≈1.10)跨性别、跨地域稳定,但机制未明;PRS 跨祖先人群(如非洲裔)预测力显著衰减,出生季节效应的跨族群异质性尚缺系统评估。
待验证: ① 同一队列内”出生时刻(时辰粒度)×家族史×PRS”三方 AUC 头对头的正式研究(尚无文献直接执行);② 命理体系(如八字)作为预测因子加入临床模型后的 NRI/净获益增量检验;③ 出生季节效应在非欧洲人群中的量级是否一致。
相关阅读
证据等级: A(有PubMed直接支撑)
参考文献: PMID 36347255, PMID 29501066, PMID 28942743, PMID 26041386, PMID 27624541, PMID 41629984
免责说明:本文属传统知识与科学文献的整理、对照与研究设计讨论,仅供理解与参考,不构成医疗建议,不能替代专业诊断与治疗。文中涉及的效应量、频率区间与暴露结论多为研究报告层面的推断,个体情况差异很大;如有身体不适或健康疑虑,请就医并由专业医师判断。
Birth Season Effect vs Family History/Polygenic Risk Score: Head-to-Head Comparison of Individual Discrimination
Evidence level: A
While the contribution of birth month to disease risk is genuinely present at the population level, it falls far short of family history and polygenic risk scores (PRS) in terms of individual discriminatory power. This head-to-head comparison effectively establishes a falsifiable quantitative boundary for the assertion that “birth date determines individual fate.”
Main Text
The core assumption of Bazi (Four Pillars) is that the precise moment of birth carries sufficient information to judge a person’s life trajectory, including fortune and misfortune. As argued previously, the effect size of birth season effects (OR≈1.05, AUC≈0.508) is separated from the information volume required for Bazi judgment by a gap of 3–4 orders of magnitude. However, rigorous validation requires a “control group”: placing the birth moment within the same coordinate system and comparing it against two widely recognized risk factors in medicine—family history and PRS. If even these established indicators only demonstrate moderate-to-low discriminatory power at the individual level, then how can a birth moment with an even smaller effect size claim to determine individual fate?
Family History vs. PRS: Systematic Measurement in Modern Genetics
First, let us compare how these two recognized strong predictors relate to one another. Mars et al. (2022) systematically compared first- and second-degree family history against whole-genome PRS across 24 common diseases in the FinnGen cohort (N=306,418). The conclusion is unequivocal: the two are independent and complementary, not interchangeable. PRS accounts for only about 10% of the family history effect, while family history accounts for only 3% of the PRS effect. In most diseases, such as coronary heart disease, glaucoma, and type 2 diabetes, a high PRS can completely “offset” the risk brought by a positive family history. In other words, even within genetics, each recognized indicator manages its own segment, and neither can lock in an individual on its own.
More direct data on discriminatory power comes from the Nordic Register Study (Lu et al., 2018): in a comparison of 5,959 patients with schizophrenia and 8,717 controls, the combined model of GRS and family history achieved a discriminatory ability “just reaching the lowest threshold for clinical utility“—even two strong factors together barely scraped the edge of “marginally usable.” The 2026 All of Us study published in BMC Medicine (N>120,000) further quantified this: PRS mediated only 13–16% of the total family history effect; only by combining both could risk stratification be pulled apart by more than 2.5 times.
The Position of Birth Season Effects in This Coordinate System
Now, let us place all three factors side by side: the OR for birth season effects is ≈1.05–1.10, while the OR for the highest decile of PRS is ≈3–4, and family history ORs range between 1.5 and 2.5 depending on the disease. Regarding discriminatory power (AUC), PRS for schizophrenia is approximately 0.65–0.70, whereas the birth month effect is ≈0.508. This coordinate system illustrates three points:
First, birth season effects and genetic risks are mechanistically independent. Escott-Price et al. (2019) directly measured the correlation between schizophrenia PRS and birth season in 136,538 individuals from the UK Biobank, finding no correlation whatsoever. The birth season effect is not a spurious gene-environment confounding artifact but a genuine environmental exposure effect. This provides scientific support for the intuition that “birth time carries environmental information,” but its magnitude dictates that it serves only as a clue for population epidemiology, not a basis for individual judgment.
Second, there is an information gap between population association and individual discrimination. SeaWAS scanning (Boland 2015, N=174.9 million) found significant associations between 55 diseases and birth month, and independent replication studies confirmed the risk peak for cardiovascular disease (Li 2016). However, “having an association” is distinct from “being able to judge an individual.” For every order of magnitude the effect size decreases, discriminatory power drops precipitously: an OR of 1.05 translates to an AUC of only 0.508, which is barely better than flipping a coin by one percentage point, whereas PRS’s 0.65 requires hundreds of thousands of GWAS participants to accumulate.
Third, implications for Bazi studies: If “birth time influences fate” is to follow a scientific route, it must stand on the same platform as family history and PRS regarding AUC, Net Reclassification Index (NRI), or Net Benefit, using the same cohort and identical outcomes for head-to-head comparison. If a Bazi system possesses genuine individual predictive power, it should demonstrate at least a discriminatory increment comparable to “first-degree family history + high PRS”—rather than remaining at the population level, such as stating “people born in winter have a 5% higher risk of schizophrenia.”
Executable Experimental Design
The testing protocol is clear and feasible: select a large cohort containing birth dates, genotypes, family history, and multi-year follow-up outcomes (either UK Biobank or FinnGen). Construct four prediction models: (1) birth season/month only; (2) family history only; (3) PRS only; (4) all three combined. Compare AUC increments, NRI, and decision curve net benefit. Expected results: the AUC increment from the birth season model will be near zero; family history and PRS will each contribute an increment of 0.01–0.02; the combined model will reach optimal performance. If the Bazi assertion holds true, its predictive increment should significantly exceed random chance—this constitutes a falsifiable criterion and represents the necessary path for “birth date fortune-telling” to move from belief to science.
Evidence and Limitations
Verified: Birth season effects represent a genuine environmental effect independent of genetic risk (PMID 29501066); family history and PRS are independent and complementary across 24 diseases, not interchangeable (PMID 36347255); the combined GRS+family history model reaches only the lowest threshold for clinical utility (PMID 28942743); birth month shows significant associations with 55 diseases that can be independently replicated (PMID 26041386, 27624541); PRS mediates only 13–16% of the total family history effect (PMID 41629984).
Partially Verified: The weak association of birth season effects with cardiovascular mortality (spring HR≈1.10) is stable across sexes and regions, though mechanisms remain unclear; PRS predictive power significantly attenuates across ancestral groups (e.g., African ancestry), and systematic assessment of cross-ethnic heterogeneity in birth season effects is still lacking.
To Be Verified: ① Formal head-to-head AUC studies within the same cohort comparing “birth moment (hourly granularity) × family history × PRS” (no literature directly addresses this yet); ② Testing NRI/net benefit increments when Bazi systems are added as predictors in clinical models; ③ Whether the magnitude of birth season effects remains consistent in non-European populations.
Related Reading
- Discriminative Power of Birth Season Effects: The Gap from Statistical Significance to Individual Fate
- Quantification of the Information Gap in Birth Moment
- Blind Test of Bazi: Chinese Execution of the Carlson Protocol
- Research Pool: Spatiotemporal Frequency Topology
Evidence Level: A (Direct PubMed support)
References: PMID 36347255, PMID 29501066, PMID 28942743, PMID 26041386, PMID 27624541, PMID 41629984
Disclaimer: This article compiles traditional knowledge and scientific literature for comparison and research-design discussion only. It is not medical advice and cannot replace professional diagnosis or treatment. Effect sizes, frequency ranges and exposure conclusions cited here are largely inferences at the research-report level, and individual circumstances vary widely. If you have any discomfort or health concerns, please consult a physician.
Leave a Reply