Impacts of host genetics on gut microbiome composition in Alzheimer's diseaseOriginal paper
What was studied?
This study asked how host genetics shape gut microbial changes in Alzheimer's disease (AD). Researchers profiled both the gut metagenome and blood whole-genome sequencing in the same people. They used a latent Dirichlet allocation topic model to define microbial subgroups, called enterosignatures. They then ran a microbiome genome-wide association study on the abundance of one AD-linked enterosignature. This combined design links specific gut microbes to human genetic variants and to AD risk.
Who was studied?
The primary analysis profiled 252 Chinese adults with varying degrees of cognitive impairment. Participants were sorted into five groups: normal control, subjective cognitive symptoms, subjective cognitive decline, mild cognitive impairment, and Alzheimer's disease. Groups were similar in gender and body mass index, but differed in age and cognitive test scores. Findings were checked in an independent Japanese cohort and other external human datasets.
What were the most important findings?
One enterosignature, dominated by the genus Anaerostipes (ES-Ana), was lower in people with worse cognition. Its abundance rose with higher MMSE and MoCA-B scores and fell in the AD group (beta = -0.378, p = 0.042). Many of its key genera are butyrate producers, and ES-Ana tracked butyrate synthesis pathways. The genome-wide analysis found 41 lead genetic variants and 174 associated genes. These genes were enriched in AD-related functions and were underexpressed in glial cells during brain aging.
What are the greatest implications of this study?
The work suggests host genetics help set the abundance of a butyrate-linked gut signature tied to cognition. Higher ES-Ana abundance was negatively correlated with the polygenic risk score for AD, consistent with a protective role. This points to gut-brain axis mechanisms in AD and possible non-invasive markers. The associations are correlational, and the moderate classification accuracy and cohort heterogeneity mean causation is not established.