Circulatory dietary and gut-derived metabolites predict early cognitive decline

A key component of disease prevention is the identification of at-risk individuals. Microbial dysbiosis in the early stages of cognitive decline and Alzheimer's disease (AD) and can modulate the levels of microbe-derived metabolites (MDM), thought to contribute to neuroinflammation, blood‒brain barrier dysfunction, and neuronal degeneration. However, the precise role of MDM in this process, as well as their potential value as risk factors, remains poorly understood.

Mass spectrometry platforms determined the serum concentration of 33 metabolites (13 tryptophan-related compounds, 15 bile acid compounds, 3 TMAO-related metabolites and 2 cresol metabolites) from cognitively healthy subjects, subjective cognitive impairment (SCI) participants and mild cognitive impairment (MCI) participants (n = 50 per group, matched for age, BMI and sex). Multiple linear regression and machine learning techniques were applied to identify a metabolite panel capable of classifying early cognitive decline. 16S rRNA amplicon sequencing was employed to identify bacterial taxa associated with these metabolic changes.

Multiple linear regression modelling, adjusted for sex, BMI, age, albumin (for its role in metabolite transport), liver and kidney function, and background diet, identified key neuroprotective metabolites, namely choline, 5-hydroxyindole acetic acid, and indole propionic acid (IPA), as lower in SCI and MCI individuals compared to healthy controls. In contrast, the cytotoxic metabolite, indoxyl sulfate, and kynurenic acid were elevated. A random forest algorithm with multiclass classification further validated these findings, highlighting six metabolites (indoxyl sulfate, choline, 5-hydroxyindole acetic acid, IPA, kynurenic acid, and kynurenine) as classifiers of early cognitive decline, achieving an area under the curve (AUC) of 0.79.

These findings suggest that MDM may serve as putative composite biomarkers of early cognitive decline, offering potential clinical relevance for metabolic risk stratification and supporting the future development of minimally invasive screening tools.