Fig. 4

A Spearman’s rank correlation of bacterial load at baseline with alpha diversity (Simpson Index), the gray area around the regression line indicates the confidence interval at a confidence level of 95%. B Spearman’s rank correlation of bacterial load at baseline with leukocytes in the study and replication cohorts. Colors depict cohorts with red being specific to the study cohort and blue to the replication cohort. Bacterial quantity is standardized within cohorts. The gray area around the regression line indicates the confidence interval at a confidence level of 95%. C Blood bacterial in load in T2D vs nonT2D in both study and replication cohorts. Bacterial quantity was standardized prior to statistical analysis. Boxplots are shown with Tukey-whiskers and mean (◆) as well as median. T2D was compared with nonT2D using Wilcoxon signed-rank test after pooling both cohorts. D Bacterial quantity in pg per μg extracted DNA over time. The three timepoints are compared using the Kruskal-Wallis test and results are validated via Friedman’s test (Kruskal-Wallis p-value is depicted). Paired samples Wilcoxon signed-rank test is used to compare two groups at once. Only samples with available triplicates in blood bacterial load are used for the comparisons (n = 39). E Bacterial quantity dynamics in good and poor responders over time (n = 21 good responders, n = 13 poor responders). The three timepoints are compared using the Kruskal-Wallis test and results are validated via Friedman’s test (Kruskal-Wallis p-value is depicted). Paired samples Wilcoxon signed-rank test is used to compare two groups at once. F Diversity dynamics in good and poor responders over time (n = 24 good responders, n = 13 poor responders). The three timepoints are compared using the Kruskal-Wallis test and results are validated via Friedman’s test (Kruskal-Wallis p-value is depicted). Paired samples Wilcoxon signed-rank test is used to compare two groups at once