Causal relationships between gut metabolites and Alzheimer's disease: a bidirectional Mendelian randomization study

Abstract

Observational studies have shown that gut microbiota–dependent metabolites are associated with the risk of Alzheimer's disease (AD). However, whether such association reflects a causality remains unclear. We conducted a bidirectional Mendelian randomization analysis to examine the causal relationships between gut microbiota–dependent metabolites trimethylamine N-oxide (TMAO) or its predecessors and AD. We observed that genetically predicted TMAO (odds ratio: 0.99, 95% confidence interval: 0.89 to 1.09 per 10 units, p = 0.775) or its predecessors including betaine (1.06, 1.00 to 1.12 per 10 units, p = 0.056), carnitine (1.05, 0.98 to 1.12 per 10 units, p = 0.178), and choline (1.01, 0.92 to 1.10 per 10 units, p = 0.905) were not associated with the risk of AD. Our Mendelian randomization estimates from AD to metabolites showed that genetically predicted higher risk of AD was also not causally associated with TMAO, betaine, carnitine, and choline levels. Our findings support that gut microbiota–dependent metabolites TMAO or its predecessors do not play causal roles in the development of AD.