四象体质的孟德尔随机化结果能否在中医九种体质上跨体系复现:痰湿与太阴的心源性栓塞卒中因果链

Written by

in

Language

Evidence level: B

韩国四象体质研究用孟德尔随机化发现”太阴体质升高、少阴体质降低心源性栓塞卒中风险”,这条因果链是否能在中医九种体质的痰湿-平和对照上复现,是检验两大体质体系”同构性”的关键实验。

正文

韩国四象医学(Sasang Constitutional Medicine)把人分为太阴、少阴、少阳、太阳四种体质。2024 年韩国韩医学研究院的一项研究,用孟德尔随机化(MR)对四象体质与卒中亚型做了因果推断:携带太阴体质相关遗传信号的个体,心源性栓塞性卒中(cardioembolic stroke, CES)风险显著升高(β=4.19, p=0.035);而少阴体质的信号则显著降低该风险(β=−9.34, p=3.87×10⁻⁵)(PMID 38927518)。这是第一次有人用遗传工具变量给”某种体质更容易得某类中风”提供了因果级证据,而不是普通的相关性。

问题在于:这是韩国体系内的结论。中医的九种体质(平和、气虚、阳虚、阴虚、痰湿、湿热、血瘀、气郁、特禀)里,哪个体质对应太阴?学界普遍认为太阴体质的”痰湿偏盛、腹型肥胖、代谢紊乱”表型,与九种体质里的痰湿体质高度重合。痰湿体质在流行病学上同样是代谢综合征和脑卒中的高危人群——一篇纳入 41 项研究、11211 名受试者的 Meta 分析显示,痰湿体质在缺血性脑卒中患者中占 23%,其患病比例是普通人群的 2.34 倍(OR=2.34, 95%CI 1.39-3.94)(PMID 34257683)。也就是说,表型层面”太阴≈痰湿”的假设有支撑。

如果两大体系真的在描述同一套生物底层,那么”太阴↑CES”这条 MR 因果链就应该能在中医体质上复现:痰湿体质的遗传风险位点,应当与心源性栓塞性卒中的风险呈同向因果效应。这不是拍脑袋能判的,需要三步:第一步,用中医体质量表给大样本人群分型并做 GWAS,拿到痰湿体质的汇总统计量;第二步,用卒中 GWAS(MEGASTROKE、GIGASTROKE 等国际联盟的公开数据)做双样本 MR;第三步,把效应方向和量级与 PMID 38927518 的韩国结果直接对比。

值得注意的一点是,韩国研究本身的可重复设计并不完美:它使用的体质诊断工具(KS-15 问卷)Cronbach α 只有 0.630,体质分型本身就带着测量噪声;而且工具变量来自体质 GWAS 的关联 SNP,样本量仅 5797 人,统计功效偏弱。但即便如此,太阴/少阴在 CES 上的效应方向相反且少阴侧达到 3.87×10⁻⁵,这个信号值得认真对待。

佐证太阴-痰湿”同构”的还有代谢通路层面的证据。四象体质的全基因组通路分析显示,太阴体质与不饱和脂肪酸生物合成、心肌收缩等通路显著相关(PMID 22889377);流行病学上太阴体质是代谢综合征各亚元素(腹型肥胖、高血糖、高甘油三酯、低 HDL、高血压)的独立危险因素(PMID 26914220);早在 2003 年,一项病例对照研究就发现太阴体质在脑梗死患者中占比高达 59.6%,显著高于对照人群的 36.9%(χ²=41.2, p<0.001)(PMID 14641479)。中医侧的数据同样指向痰湿:痰湿体质的基因组学研究发现了 4 个上调和 6 个下调基因,功能分析提示易感高脂血症和糖尿病(PMID 23548117);11 项研究 12890 人的 Meta 分析显示痰湿体质是血脂异常的危险因素(OR=2.50, 95%CI 2.22-2.80)(PMID 29234371);2025 年一项 1029 人的体检队列也确认痰湿体质在 BMI、血压、甘油三酯等指标上显著偏离平和体质(PMID 39364874)。

两条证据链在”代谢紊乱-血管风险”这个节点上高度收敛:太阴和痰湿都携带代谢性遗传负荷,而代谢紊乱正是心源性栓塞性卒中的上游。这提示”太阴↑CES”若能在痰湿体质上复现,就说明两套体质分类捕捉到了同一组遗传-代谢-血管因果通路;若不能复现,则说明两体系至少在脑血管风险维度上并不同构,需要重新审视”太阴≈痰湿”的映射。

从方法学看,这条复现路径最干净的执行方案是:以中医体质 GWAS 汇总统计量(暴露)与卒中亚型 GWAS(结局)做双样本 MR,选用 F 统计量达标的独立工具变量,并用 MR-Egger 截距和 MR-PRESSO 检验水平多效性;随后将效应方向、量级与 PMID 38927518 对比,最好加上跨人群的留一法敏感性分析。母文(tcm-constitution-mendelian-randomization-exposure)已经指出中医体质 MR 的可行性前提,姊妹篇(tcm-constitution-gwas-instrument-strength-audit)则警示了”F>10 就能跑 MR”的误区——九种体质 GWAS 目前样本量仅千人级,独立工具变量密度可能不足以支撑稳健的 MR。因此这条复现实验的真实瓶颈,很可能不是统计方法,而是体质 GWAS 的样本规模。

证据与局限

已验证: ① 韩国四象体质 MR:太阴↑CES(β=4.19, p=0.035)、少阴↓CES(β=−9.34, p=3.87×10⁻⁵)(PMID 38927518);② 太阴体质在脑梗死患者中显著富集(59.6% vs 36.9%)(PMID 14641479);③ 太阴体质是代谢综合征独立危险因素(PMID 26914220);④ 痰湿体质是血脂异常(OR=2.50)与缺血性卒中(OR=2.34)的危险因素(PMID 29234371, 34257683)。

部分验证: 太阴与痰湿在代谢表型(腹型肥胖、血脂、血糖)上的相似性有独立证据支持(PMID 22889377, 23548117, 39364874),但”两体系指涉同一遗传结构”仍是待检验假设。

待验证: ① 中医痰湿体质 GWAS 位点与心源性栓塞性卒中的双样本 MR 因果效应方向/量级,能否与 PMID 38927518 对齐;② 若复现失败,需判定是”映射错误”还是”体质 GWAS 功效不足”;③ 少阴体质的 CES 保护效应在中医哪类体质(阳虚?阴虚?)上对应。

相关阅读

证据等级: B
参考文献: PMID 38927518, PMID 34257683, PMID 29234371, PMID 14641479, PMID 26914220, PMID 22889377, PMID 23548117, PMID 39364874


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。

Can Mendelian Randomization Results for the Four Constituent Body Types Be Replicated Across Traditional Chinese Medicine’s Nine Body Types: The Causal Chain Between Phlegm-Damp…

Evidence level: B

A Korean study using Sasang Constitutional Medicine found via Mendelian randomization (MR) that “Taiyin constitution increases and Shaoyin constitution decreases the risk of cardioembolic stroke.” Whether this causal chain can be replicated in the comparison between Phlegm-Dampness and Peiping (Normal) among Traditional Chinese Medicine’s (TCM) nine constitutions is a critical experiment testing the “isomorphism” between the two constitution systems.

Main Text

Sasang Constitutional Medicine divides humans into four constitutions: Taiyin, Shaoyin, Shaoyang, and Taiyang. A 2024 study by the Korean Academy of Traditional Korean Medicine utilized Mendelian randomization (MR) to infer causality between Sasang constitution and stroke subtypes: individuals carrying genetic signals associated with Taiyin constitution showed a significantly increased risk of cardioembolic stroke (CES) ($\beta=4.19, p=0.035$), whereas signals associated with Shaoyin constitution significantly decreased this risk ($\beta=-9.34, p=3.87\times10^{-5}$) (PMID 38927518). This represents the first provision of causal-level evidence using genetic instrumental variables for “a specific constitution predisposes to a certain type of stroke,” rather than mere correlation.

The issue lies in that this conclusion is specific to the Korean system. Among TCM’s nine constitutions (Peiping, Qi-deficiency, Yang-deficiency, Yin-deficiency, Phlegm-Dampness, Damp-Heat, Blood Stasis, Qi Stagnation, and Special Predisposition), which constitution corresponds to Taiyin? The academic consensus generally holds that the Taiyin phenotype of “excess Phlegm-Dampness, abdominal obesity, and metabolic disorder” highly overlaps with the Phlegm-Dampness constitution in the nine-category system. Epidemiologically, Phlegm-Dampness is also a high-risk group for metabolic syndrome and brain stroke—a meta-analysis including 41 studies and 11,211 participants showed that Phlegm-Dampness constituted 23% of ischemic stroke patients, with a prevalence ratio 2.34 times that of the general population (OR=2.34, 95%CI 1.39-3.94) (PMID 34257683). Thus, the hypothesis that “Taiyin $\approx$ Phlegm-Dampness” at the phenotypic level is supported by evidence.

If both systems truly describe the same underlying biological mechanisms, then the MR causal chain “Taiyin $\uparrow$ CES” should replicate in TCM constitutions: genetic risk loci for Phlegm-Dampness constitution should exhibit a causal effect in the same direction as cardioembolic stroke risk. This cannot be judged arbitrarily; it requires three steps: First, phenotype the large population using TCM constitution scales and conduct a GWAS to obtain summary statistics for Phlegm-Dampness constitution; Second, perform two-sample MR using stroke GWAS data (public data from international consortia such as MEGASTROKE and GIGASTROKE); Third, directly compare the effect direction and magnitude with the Korean results in PMID 38927518.

It is worth noting that the Korean study itself has imperfect design for reproducibility: the constitution diagnosis tool (KS-15 questionnaire) had a Cronbach $\alpha$ of only 0.630, meaning the constitution phenotyping itself carries measurement noise; furthermore, the instrumental variables are SNPs from the constitution GWAS association with a sample size of only 5,797 people, resulting in weak statistical power. Nevertheless, the effect directions for Taiyin/Shaoyin on CES are opposite, and the Shaoyin side reaches a significance of $3.87\times10^{-5}$, a signal worthy of serious attention.

Further evidence supporting the “isomorphism” between Taiyin and Phlegm-Dampness exists at the level of metabolic pathways. Whole-genome pathway analysis in Sasang constitution showed significant associations between Taiyin constitution and pathways such as unsaturated fatty acid biosynthesis and myocardial contraction (PMID 22889377); epidemiologically, Taiyin constitution is an independent risk factor for all sub-elements of metabolic syndrome (abdominal obesity, hyperglycemia, hypertriglyceridemia, low HDL, hypertension) (PMID 26914220); as early as 2003, a case-control study found that Taiyin constitution accounted for 59.6% of patients with cerebral infarction, significantly higher than the 36.9% in the control population ($\chi^2=41.2, p<0.001$) (PMID 14641479). Data from the TCM side also point to Phlegm-Dampness: genomic studies on Phlegm-Dampness constitution identified 4 upregulated and 6 downregulated genes, with functional analysis suggesting susceptibility to hyperlipidemia and diabetes (PMID 23548117); a meta-analysis of 11 studies with 12,890 participants showed Phlegm-Dampness constitution is a risk factor for lipid abnormalities (OR=2.50, 95%CI 2.22-2.80) (PMID 29234371); and a 2025 cohort study of 1,029 individuals confirmed that Phlegm-Dampness constitution significantly deviates from Peiping constitution in indicators such as BMI, blood pressure, and triglycerides (PMID 39364874).

The two evidence chains converge highly at the node of “metabolic disorder-vascular risk”: both Taiyin and Phlegm-Dampness carry a metabolic genetic load, while metabolic disorder is upstream of cardioembolic stroke. This suggests that if “Taiyin $\uparrow$ CES” can be replicated in Phlegm-Dampness constitution, it indicates that both constitution classification systems capture the same group of genetic-metabolic-vascular causal pathways; if replication fails, it indicates that the two systems are not isomorphic at least in the dimension of cerebrovascular risk, requiring a re-examination of the mapping “Taiyin $\approx$ Phlegm-Dampness”.

From a methodological perspective, the cleanest execution plan for this replication path is: perform two-sample MR using TCM constitution GWAS summary statistics (exposure) and stroke subtype GWAS (outcome), selecting independent instrumental variables meeting the F-statistic threshold, and applying MR-Egger intercept and MR-PRESSO tests for horizontal pleiotropy; subsequently compare the effect direction and magnitude with PMID 38927518, preferably adding leave-one-out sensitivity analysis across populations. The original paper (tcm-constitution-mendelian-randomization-exposure) already pointed out the feasibility prerequisites for TCM constitution MR, while the sister paper (tcm-constitution-gwas-instrument-strength-audit) warns against the misconception that “F>10 allows running MR”—the nine TCM constitution GWAS currently have sample sizes in the thousands, and the density of independent instrumental variables may be insufficient to support robust MR. Therefore, the real bottleneck for this replication experiment is likely not the statistical method but the sample size of the constitution GWAS.

Evidence and Limitations

Verified: ① Korean Sasang MR: Taiyin $\uparrow$ CES ($\beta=4.19, p=0.035$), Shaoyin $\downarrow$ CES ($\beta=-9.34, p=3.87\times10^{-5}$) (PMID 38927518); ② Significant enrichment of Taiyin constitution in cerebral infarction patients (59.6% vs 36.9%) (PMID 14641479); ③ Taiyin constitution is an independent risk factor for metabolic syndrome (PMID 26914220); ④ Phlegm-Dampness constitution is a risk factor for lipid abnormalities (OR=2.50) and ischemic stroke (OR=2.34) (PMID 29234371, 34257683).

Partially Verified: Similarity between Taiyin and Phlegm-Dampness on metabolic phenotypes (abdominal obesity, lipids, blood sugar) is supported by independent evidence (PMID 22889377, 23548117, 39364874), but “referring to the same genetic structure” remains a hypothesis awaiting verification.

To Be Verified: ① Whether the causal effect direction/magnitude of two-sample MR between TCM Phlegm-Dampness constitution GWAS loci and cardioembolic stroke aligns with PMID 38927518; ② If replication fails, determine whether it is a “mapping error” or “insufficient power of constitution GWAS”; ③ Which TCM constitution corresponds to the protective effect of Shaoyin on CES (Yang-deficiency? Yin-deficiency?).

Related Reading

Evidence Level: B
References: PMID 38927518, PMID 34257683, PMID 29234371, PMID 14641479, PMID 26914220, PMID 22889377, PMID 23548117, PMID 39364874


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.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *