# Stool and Plasma Metabolites and Cognitive Function in the Boston Puerto Rican Health Study

## TL;DR

This study explores how gut and blood metabolites relate to cognitive function in Puerto Rican adults, finding several metabolites linked to better or worse cognition.

## Contribution

The study is the first to concurrently examine stool and plasma metabolites in relation to cognition in a Latino/Hispanic population.

## Key findings

- 14 stool and 257 plasma metabolites were associated with cognitive function, including palmitoleoyl ethanolamide and gulonate.
- Metabolite sets like tyrosine metabolism in plasma and tocopherol metabolism in stool were more frequently linked to cognition.
- Correlations between stool and plasma metabolites were limited, mostly involving caffeine-related compounds.

## Abstract

Several studies have reported alterations in the gut microbiome and plasma metabolome in dementia, but no studies have concurrently examined metabolites in stool and plasma. Evidence is especially lacking in Latinos/Hispanics who have unique metabolic and microbiome profiles and are at an increased risk of dementia.

Our study focused on 314 Boston Puerto Rican Health Study participants (median age 68 y), who provided concurrent stool and blood samples (untargeted metabolomic profiling, Metabolon, Inc) and underwent a comprehensive cognitive assessment, summarized as a global cognitive score (GCS). Multivariate analyses in MaAsLin2 were used to identify stool and plasma metabolites associated with GCS. Overrepresentation analysis (ORA) was used to identify metabolite sets associated with GCS. An FDRp cutoff of < 0.25 was considered significant.

We identified 14 stool and 257 plasma metabolites (89 at FDRp < 0.05) associated with GCS. In stool, palmitoleoyl ethanolamide (pFDR = 0.15) and steviol (pFDR = 0.20) were associated with better cognition. Lactobacillic acid (pFDR = 0.09), a microbially derived metabolite, stigmasterol (pFDR = 0.09), and other metabolites were associated with worse cognition. In plasma, gulonate (pFDR= 0.00007), N.acetylneuraminate (FDR= 0.0001) and many others were associated with lower GCS. Ergothioneine, an antioxidant, (FDR = 0.02), among other metabolites, was associated with higher GCS. ORA analyses identified that metabolites within the tyrosine metabolism group pFDR = 0.2145) in plasma, and long chain saturated fatty acid (pFDR = 0.0003) and tocopherol metabolism (pFDR = 0.0018) groups in stool, were associated with GCS more frequently than expected. Correlations between stool and plasma metabolites were limited and mostly restricted to caffeine‐related metabolites.

Our findings of an association between the plasma metabolome and GCS are consistent with prior studies. More work is needed to examine the stool and plasma metabolomes in dementia.