Extended Data Fig. 4: rUTI dysbiosis is not driven by antibiotic use during the study.
We grouped rUTI women according to their antibiotic exposures at any point during the UMB study; (i) ciprofloxacin (n = 6) (ii) non-ciprofloxacin antibiotics (n = 6); (iii) no antibiotics (n = 3); (iv) any antibiotics (n = 12). Groups were compared against each other and against the control cohort (n = 16) for (a) overall microbial richness and (b) relative abundance of butyrate producers. Crosses represent mean values for individuals, boxplots denote the IQR and 95% central quantiles for each group. Wilcoxon rank sum tests (two-sided) were applied to group pairs to derive p-values. (c) Temporal trends of microbial richness (black) and relative abundance of butyrate producers (red) in all rUTI participants using antibiotics during the study. For each individual, linear models were fit to observations (points) over time; fitted trends are shown, with coefficients & p values reported at the top of each panel. Dashed vertical lines denote antibiotic usage. Participant mean values are represented by horizontal lines.
