Identifying candidate gut microbiota indicators for Alzheimer's disease through integrated data

Alzheimer's disease (AD) is associated with the gut microbiota, and identifying reliable gut microbiota biomarkers enhances AD diagnosis. We aim to characterize the gut microbiota in AD patients by integrating data from multiple populations and identifying key candidate gut microbiota indicators with diagnostic value for AD. Public data from studies on AD and gut microbiota were collected, including participants from AD dementia, mild cognitive impairment (MCI), and normal control (NC) groups. Microbiota composition, diversity, and network analyses were used to characterize the gut microbiota of the three groups. Differential bacterial genera identified simultaneously by seven common methods served as candidate indicators. The study included 799 AD dementia, 170 MCI, and 731 NC participants. The AD dementia group demonstrated a lower relative abundance of Bacteroides and Faecalibacterium and lower α-diversity than the MCI and NC groups (P < 0.05). The microbial network density in the AD dementia group was reduced by 1.5% and 1.6% compared with the NC and MCI groups, respectively. We identified 35 bacterial genera as candidate indicators for AD, including first-time reports of RF39 and Oligella. Faecalibacterium was the most important candidate indicator in the overall population, Akkermansia in the Chinese population, Collinsella in the "Turkish and Kazakh" population, and Actinomyces in the "American and Canadian" population. Our findings contribute to the development of non-invasive biomarkers for AD diagnosis and targeted microbiota therapies and provide a valuable reference for selecting specific biomarkers for different application scenarios.

This study characterized the gut microbiota of Alzheimer's disease (AD) patients and identified candidate indicators for AD diagnosis using a large, multi-population data set. The AD dementia group consistently showed lower α-diversity and a sparser microbiota interaction network than the other groups. We identified 35 bacterial genera as candidate indicators for AD, including first-time reports of RF39 and Oligella. Faecalibacterium was the most important candidate indicator in the overall population, Akkermansia in the Chinese population, Collinsella in the "Turkish and Kazakh" population, and Actinomyces in the "American and Canadian" population. These findings provide a valuable reference for selecting biomarkers for different application scenarios.