Gut microbiome‐related metabolites in plasma are associated with general cognition

• Published in: Alzheimer's & dementia Vol. 19; no. S15
• Main Authors: Kastenmuller, Gabi, Wu, Tong, Arnold, Matthias, Hankemeier, Thomas, Ghanbari, Mohsen, Uitterlinden, André G., Kraaij, Robert, Van Duijn, Cornelia M, Kaddurah‐Daouk, Rima F., Ikram, M. Arfan
• Format: Journal Article
• Language: English
• Published: 01.12.2023
• ISSN: 1552-5260; 1552-5279
• DOI: 10.1002/alz.078299

Background Increasing evidence suggests a role of metabolism in neurological disorders including Alzheimer’s disease (AD). The human metabolome is determined by genes and modifiable factors such as gut microbiota, diet, medication, and other lifestyle choices. We hypothesized that metabolites related to endophenotypes of brain health maybe derived or influenced by these modifiable factors. Identifying the determinants of metabolites associated with AD‐related endophenotypes may help to discover new targets for drugs and prevention. Method We profiled plasma samples of 1082 participants of the Rotterdam study for 1387 metabolites using a non‐targeted metabolomics platform. After preprocessing and quality control, plasma levels of 991 metabolites were available for analysis of 1068 cognitively normal participants. We assessed the role of metabolites in general cognition (G‐factor) and magnetic resonance imaging (MRI) markers using linear regression. Replication analysis was performed for general cognition in an independent dataset. To identify the source of brain health‐associated metabolites, we estimated the explained variance (EV) of metabolites by genes, gut‐microbiome, medication use, and clinical features using gradient boosting decision tree (GBDT). Multiple testing correction was performed based on the false discovery rate (FDR) by Benjamini Hochberg (FDR<0.05). Result Among the 991 analyzed metabolites, 14 showed significant associations with general cognition, 21 metabolites with total brain volume, and one with total white matter lesions. In the replication analysis of metabolites associated with cognition, 9 metabolites showed consistent association with G‐factor. While genetics explained more than 5% of the variance of all associated metabolites, for three cognition‐associated metabolites (ergothioneine, 3‐hydroxy‐2‐methylpyridine sulfate, 4‐vinylguaiacol sulfate), modifiable lifestyle factors, medication and microbiota additionally explained up to 15% of their variance, with lifestyle factors showing the largest influence. Among the metabolites associated with brain imaging phenotypes, clinical factors and medication explained more than 5% of the variance of four metabolites (3‐hydroxysebacate, 3‐hydroxybutyrylcarnitine, sphingomyelin (d18:2/24:2), unnamed metabolite). Conclusion Our study provides evidence for the association between circulating metabolites and cognition or brain imaging phenotypes in participants without cognitive impairment. The levels of these metabolites are influenced by genetics and in part by modifiable factors. Further, studies are warranted to elucidate the complex relationship between gut microbiota, diet, genome, metabolome, and cognition.