Extended Data Fig. 1: Microbiome variation in the BMIS cohort (n = 888 participants).

a, Percentage of subjects in the BMIS cohort taking medication of the stated drug classes. ACE inhibitors, angiotensin converting enzyme inhibitors; ARB, angiotensin II receptor blockers; ASA, acetylsalicylic acid; PPI, proton-pump inhibitors. b, Best model explaining inter-individual microbiome variation based on obesity-defining and metabolic-syndrome-defining variables as well as on most frequently disclosed therapeutics (taken by more than 10% of participants; Supplementary Table 4). Explanatory power of the included variables are reported for the variables taken individually (black bars; n = 888 biologically independent samples, univariate dbRDA)) or in a multivariate model (grey bars; n = 888 biologically independent samples, multivariate dbRDA). c, Principal coordinates analysis of inter-individual differences (genus level Bray–Curtis dissimilarity) in the microbiome profiles of the BMIS cohort (n = 888 biologically independent samples, data points coloured by enterotypes (Extended Data Fig. 4)) with the rest of the MetaCardis dataset in the background (n = 1,134, grey dots). Full and open circles corresponding to statin-medicated (Stat(+)) and non-statin-medicated participants (Stat(−)), respectively. Arrows represent the effect sizes of a post hoc fit of significant microbiome covariates identified in the multivariate model in b. d, Same principal coordinates analysis as in c, with the statin intake variable split into the separate statin classes (n = 888 biologically independent samples, simvastatin (n = 51), atorvastatin (n = 33) and other statins (n = 22); Supplementary Table 4). In c, d, the body of the box plot represents the first and third quartiles of the distribution, the line represents the median, and the whiskers extend from the quartiles to the last data point within 1.5× the interquartile range (IQR), with outliers beyond.