中国神经再生研究(英文版) ›› 2026, Vol. 21 ›› Issue (9): 4474-4480.doi: 10.4103/NRR.NRR-D-25-01711

• 原著:神经损伤修复保护与再生 • 上一篇    

肠道微生物群与炎症蛋白的双样本孟德尔随机化分析:偏头痛预测性及预防性和个性化治疗

  

  • 出版日期:2026-09-15 发布日期:2026-05-22

Two-sample Mendelian randomization study of gut microbiota and inflammatory proteins: Predictive, preventive, and personalized treatment for migraine

Ruolan Ma, Haiyan An, Yi Feng*   

  1. Department of Anesthesiology, Peking University People’s Hospital, The Second Clinical College of Peking University, Beijing, China
  • Online:2026-09-15 Published:2026-05-22
  • Contact: Yi Feng, MD, PhD, doctor_yifeng@sina.com.

摘要:

人类肠道微生物群作为偏头痛发病机制的重要因素日益受到关注,其作用可能通过炎症通路实现。识别与偏头痛相关的特定人类肠道微生物群成分,并结合特定炎症蛋白的研究,对推进偏头痛的一级预测、靶向预防及个性化治疗(PPPM)策略至关重要。此研究采用公开的GWAS汇总数据开展双样本孟德尔随机化(MR)分析。从芬兰国家健康与福利研究所开展的全国健康调查(FINRISK,n=5,959名欧洲受试者)中获取473种人类肠道微生物群分类群数据。通过Olink Target 96炎症检测面板,从11个队列共14,824名参与者中获取了91种循环炎症蛋白的GWAS数据(https://www.ebi.ac.uk/gwas/)。偏头痛结局数据源自FinnGen R12数据集,病例依据ICD-10编码G43定义。所有GWAS分析均对性别、年龄、基因分型批次及10个遗传主成分进行校正以控制人群分层(基因组膨胀因子:1.00-1.05)。逆方差加权(IVW)MR作为主要分析方法,辅以MR-Egger检验、加权中位数法及模态法进行敏感性分析。两步MR中介分析量化了肠道微生物组对偏离基因位点介导的偏头痛影响比例。IVW方法发现37个细菌属与偏头痛相关,其中18个属呈负相关,19个属呈正相关。此外,8种肠道代谢产物被证实可增加偏头痛风险。在高丰度微生物群中,4种被观察到能降低代谢产物水平,另4种则与代谢产物水平升高相关。两项孟德尔随机化分析均识别出5种肠道微生物群通过代谢产物影响偏头痛。具体而言,放线菌、短螺旋菌科、CAG-269 sp001915995及冰原菌属通过肠道代谢物影响偏头痛结局,其介导比例分别为12%、19%、15.5%和6.7%。Lawsonibacter sp002161175被证实通过Oncostatin-M和SLAM影响偏头痛风险,其介导比例分别为15.6%和11.3%。本研究阐明了特定肠道微生物组改变在偏头痛发病机制中的作用,并突出了炎症蛋白的介导效应。针对这些特定肠道微生物组改变,为偏头痛管理中的预测性、预防性和个性化医疗提供了有前景的策略,有望实现重大临床进展。


https://orcid.org/0000-0002-5025-2818 (Yi Feng)

关键词: 肠道微生物组l偏头痛l炎症蛋白l孟德尔随机化l介导比例l炎症l微生物组l全基因组关联研究(GWAS)l肠脑轴l因果推断

Abstract:

The human gut microbiota is increasingly recognized as a significant factor in the pathogenesis of migraine, potentially via inflammatory pathways. Identifying specific human gut microbiota components associated with migraines, along with the investigation of particular inflammatory proteins, is essential for advancing primary prediction, targeted prevention, and personalized treatment strategies for migraines. We conducted a two-sample Mendelian randomization study using publicly available summary statistics from genome-wide association studies. Data for 473 human gut microbiota taxa were obtained from the Finnish national health survey conducted by the National Institute for Health and Welfare study (FINRISK, n = 5959 European participants). Genome-wide association study data (https://www.ebi.ac.uk/gwas/) for 91 circulating inflammatory proteins were obtained from 14,824 participants across 11 cohorts using the Olink Target 96 Inflammation panel. Migraine outcome data were obtained from the FinnGen R12 release, with cases defined using ICD-10 code G43. All genome-wide association study analyses were adjusted for sex, age, genotyping batch, and 10 genetic principal components to control population stratification (genomic inflation factors: 1.00–1.05). Inverse variance-weighted Mendelian randomization was the primary analysis method, with Mendelian randomization-Egger, weighted median, and mode-based methods as sensitivity analyses. Two-step Mendelian randomization mediation analysis quantified the proportion of the effects of human gut microbiota on migraine that are mediated through inflammatory proteins. Thirty-seven bacterial genera were found to be associated with migraine using the inverse variance-weighted method. Of these, 18 genera exhibited a negative association, while 19 genera demonstrated a positive association with migraine risk. Additionally, eight inflammatory proteins were found to increase the risk of migraine. Among human gut microbiota, four were observed to reduce inflammatory protein levels, whereas another four were associated with increased inflammatory protein levels. Additionally, five gut microbiota were identified to influence migraine through inflammatory proteins in both Mendelian randomization analyses. Specifically, Actinobacteria, Brachyspiraceae, CAG-269 sp001915995, and Paraglaciecola were found to affect migraine outcomes via inflammatory proteins, with mediation proportions of 12%, 19%, 15.5%, and 6.7%, respectively. Lawsonibacter sp002161175 was identified to influence migraine risk through Oncostatin-M and SLAM, with mediation proportions of 15.6% and 11.3%, respectively. Our study elucidated the role of specific human gut microbiota alterations in the pathogenesis of migraine and highlighted the mediating effects of inflammatory proteins. Targeting these particular human gut microbiota alterations offers a promising strategy for predictive, preventive, and personalized medicine in migraine management, resulting in substantial clinical advancements.

Key words: causalityl cytokinesl gut–brain axisl human gut microbiotal inflammation proteinl mediating proportionl mediationl mendelian randomizationl microbiomel migraine