Extended Data Fig. 7: Comparisons of microbial compositions between baseline and follow-up.

(a) aPCoA of the Bray‒Curtis distance plot for baseline and follow-up after adjusting for hypertension, medication use, and alcohol and tea drinking. (b) α-Diversity differences between baseline and follow-up for stratified health statuses (HB, n = 21; HF, n = 15; LHB, n = 19; LHF, n = 27). (c) Composition of the gut microbiome at the phylum level between baseline and follow-up for stratified health statuses. (d) LEfSe between baseline and follow-up stratified by health status (two-sided Kruskal‒Wallis test between classes; P < 0.05, LDA > 2). Healthy baseline (HB, n = 21), healthy follow-up (HF, n = 15), less healthy baseline (LHB, n = 19) and less healthy follow-up (LHF, n = 27) samples were examined. The Wilcoxon rank-sum test (two-sided) was used for panel b. Boxes in boxplots represent the 25th percentile, median, and 75th percentile and whiskers stretch to 1.5 times the interquartile range from the corresponding hinge.