Relationships between Habitual Polyphenol Consumption and Gut Microbiota in the INCLD Health Cohort

Abstract

While polyphenol consumption is often associated with an increased abundance of beneficial microbes and decreased opportunistic pathogens, these relationships are not completely described for polyphenols consumed via habitual diet, including culinary herb and spice consumption. This analysis of the International Cohort on Lifestyle Determinants of Health (INCLD Health) cohort uses a dietary questionnaire and 16s microbiome data to examine relationships between habitual polyphenol consumption and gut microbiota in healthy adults (n = 96). In this exploratory analysis, microbial taxa, but not diversity measures, differed by levels of dietary polyphenol consumption. Taxa identified as exploratory biomarkers of daily polyphenol consumption (mg/day) included Lactobacillus, Bacteroides, Enterococcus, Eubacterium ventriosum group, Ruminococcus torques group, and Sutterella. Taxa identified as exploratory biomarkers of the frequency of polyphenol-weighted herb and spice use included Lachnospiraceae UCG-001, Lachnospiraceae UCG-004, Methanobrevibacter, Lachnoclostridium, and Lachnotalea. Several of the differentiating taxa carry out activities important for human health, although out of these taxa, those with previously described pro-inflammatory qualities in certain contexts displayed inverse relationships with polyphenol consumption. Our results suggest that higher quantities of habitual polyphenol consumption may support an intestinal environment where opportunistic and pro-inflammatory bacteria are represented in a lower relative abundance compared to those with less potentially virulent qualities.

Keywords: culinary herbs and spices; microbiome; phytochemicals; polyphenols.

Figure 1

Average herb and spice use. Shown here are the average frequencies with which each herb and spice was used by study participants. Each herb and spice are color labeled by the category of its total polyphenol contents in milligrams (mg) of total polyphenols per kilogram (kg) dry weight of the herb/spice (mg/kg DW).

Figure 2

Community profiling by polyphenol consumption categories. The relative abundances of participants’ (A) phyla and (B) genera are stratified by low, medium, and high consumers of dietary polyphenols.

Figure 3

Diversity does not differ by estimated polyphenol consumption. Two measures of alpha diversity, Shannon Index (A) and observed richness (B), are stratified by low, medium, and high consumers of dietary polyphenols; no significant features were detected. Bray—Curtis dissimilarity metrics (C), a measure of beta diversity, are plotted between low, medium, and high consumers of dietary polyphenols; PERMANOVA revealed that no significant features were detected.

Figure 4

Taxa abundance differs by estimated polyphenol consumption. Shown are taxa at the genus level (A) which were identified as biomarkers of polyphenol consumption and differ by low, medium, and high consumers; only taxa with p < 0.1 are displayed. A phylogenetic heat tree (B) displaying a differential abundance of genera (p-value < 0.1) was detected by Wilcoxon Rank Sum comparing the low and high consumers.

Figure 5

Diversity does not differ by frequency of polyphenol-weighted herb and spice use. Two measures of alpha diversity, Shannon Index (A) and observed richness (B), are stratified by low-, medium-, and high-frequency use of polyphenol-weighted herbs and spices; no significant features were detected. Bray–Curtis dissimilarity metrics (C), a measure of beta diversity, are plotted between low-, medium-, and high-frequency use of polyphenol-weighted herbs and spices; PERMANOVA revealed that no significant features were detected.

Figure 6

Taxa abundance differs by frequency of polyphenol-weighted herb and spice use. Shown are taxa at the genus level (A) which were identified as biomarkers of polyphenol-weighted herb and spice use and differ by low-, medium-, and high-frequency users; only taxa with p < 0.1 are displayed. A phylogenetic heat tree (B) displaying a differential abundance of genera (p-value < 0.1) was detected by Wilcoxon Rank Sum comparing the low- and high-frequency users.

Figure 7

Spearman’s rank correlations between taxa abundance and estimated daily consumption of major polyphenol classes. Microbial taxa identified as biomarkers of polyphenol exposure were further explored by correlating the relative abundance of each with estimated daily polyphenol consumption (“Total”; mg/day), and the five major polyphenol classes. Shown is a heat map displaying the directionality (e.g., positive or inverse) and significance of the correlation coefficients. * p < 0.05; ** p < 0.01; *** p < 0.001; t < 0.1.