Gut Bacteria Causally Linked to Brain White Matter Health via Genetic Analysis

White matter — the brain's communication highway made of myelinated axons — is increasingly recognized as a key target in neurological aging and disease. Damage to white matter, visible as white matter hyperintensities (WMHs) on MRI, is associated with cognitive decline, stroke, and dementia. Meanwhile, the gut microbiome has emerged as a powerful modulator of brain health via immune, neuroendocrine, and vagal pathways. Yet whether specific gut bacteria causally affect white matter integrity had never been rigorously tested — until this study.

The researchers conducted a two-sample Mendelian randomization (MR) analysis using GWAS summary statistics from four major datasets. Gut microbiota genetic data came from the MiBioGen consortium (N=18,340), covering 211 bacterial taxa. White matter hyperintensity data came from the CHARGE/UK Biobank consortium (N=50,970). White matter microstructure data were drawn from 31,356 UK Biobank participants using diffusion MRI metrics including DTI and NODDI. White matter connectivity data came from 26,333 UK Biobank participants. Thirteen neurological disease GWAS datasets were also included, covering stroke subtypes, MS, ALS, Alzheimer's, Parkinson's, and rare autoimmune conditions.

Four bacterial taxa showed statistically significant causal associations with WMH burden after Bonferroni correction (p<2.55×10⁻⁴): class Melainabacteria, order Gastranaerophilales, family Alcaligenaceae, and genus Ruminiclostridium 6. Three additional taxa demonstrated consistent effects across multiple white matter microstructure metrics, suggesting broad influence on axonal integrity and myelination. Twelve significant associations were identified between bacterial taxa and white matter connectivity pathways, with genes CPNE1, PIGU, MED22, SURF6, DOCK10, and COPS3 emerging as mediating candidates through transcriptomic MR and SMR/HEIDI analyses.

For neurological disease outcomes (Bonferroni threshold p<2.94×10⁻³), several striking findings emerged. Family Clostridiaceae 1 showed a protective effect against ischemic stroke. Genus Barnesiella was protective against both ischemic stroke and small vessel stroke, but paradoxically increased risk for neuromyelitis optica spectrum disorder (NMOSD) and its AQP4-IgG+ subtype. Order Desulfovibrionales and family Desulfovibrionaceae were protective against cardioembolic stroke. Genus Ruminococcus gnavus group showed a protective effect against ALS. These findings survived pleiotropy testing via MR-Egger and MR-PRESSO, and heterogeneity was addressed using random-effects models where Cochran's Q indicated significant variance.

The study's functional mapping layer adds mechanistic depth: by mapping significant SNPs to genes via FUMA and then running cis-eQTL-based transcriptomic MR using eQTLGEN data (31,684 blood samples), the authors identified specific genes through which gut bacteria may influence white matter connectivity. This multi-layered approach — from bacterial taxa to SNPs to gene expression to brain imaging phenotypes — represents a methodological advance over prior gut-brain MR studies. Limitations include reliance on mixed-sex European-ancestry cohorts without sex-stratified data, the inherent inability of MR to capture dynamic microbiome changes over time, and the fact that bacterial taxa GWAS were conducted at a single time point.