Identification of Candidate Biomarkers of Fiber-Rich Foods Captured by Metabolomics Analyses

Objectives: The physiological impact of dietary fiber (DF) consumption is difficult to ascertain given its wide range of sources and properties. This review aims to identify emerging candidate metabolite biomarkers capturing different sources of DF intake. Methods: We searched four electronic databases (PubMed, Web of Science, BIOSIS, and CINAHL) using a combination of keywords related to DF (79), biomarkers (9), intake (9), and biofluid (urine, blood, and fecal). Search criteria included 2+ arm parallel studies, crossovers, and observational studies. We excluded studies in non-adult participants or those with type 2 diabetes, cardiovascular disease, or cancer. Positive, negative, and null associations between metabolites and DF sources (cereal, fruit, legume, vegetable, starchy foods, or isolated fiber supplements) were captured for all studies. Only metabolites with a positive association with a DF source in at least two studies were considered candidate metabolite biomarkers. A unique scoring system was developed and assigned to each candidate metabolite biomarker. This score accounted for both the number of positive, negative, and null associations consistent for a specific DF to capture degree of replication, as well the number of positive associations to other DF sources to capture degree of specificity. Results: A total of 900 unique metabolites were identified from the 114 publications (15% observational and 85% RCTs). With respect to the DF sources, fifty-nine (including 19 alkylresorcinols) candidate metabolite biomarkers were identified for cereal fiber, 16 for fruit fiber, and 9 for legume fiber. The top scoring metabolites in urine and blood were ferulic acid 4-O-sulfate and pipecolic acid betaine for cereal fiber; hydroxyhippuric acid and N-methyl-L-proline for fruit fiber; and dimethylglycine and methylcysteine for legume fiber, respectively. Conclusions: We identified candidate biomarkers of cereal, fruit, and legume fiber that warrant further evaluation. These biomarkers may be derived from consumption of these specific fibers or the downstream metabolic consequences of their consumption, but also may reflect intakes of correlated nutrients or foods. Funding Sources: Jean Hankin Nutrition Epidemiology Research Grant.