Extended Data Fig. 8: Control for the effect of additional medication taken by obese statin-medicated or non-statin-medicated individuals of the BMIS cohort (n = 888 participants) on the association between reduced Bact2 prevalence and statin intake.
From: Statin therapy is associated with lower prevalence of gut microbiota dysbiosis
a, List of drugs taken by non-statin-medicated and statin-medicated obese BMIS participants separated into 5 groups: those reporting no (co-)medication (beyond statin intake) (+0), and those reporting one (+1), two (+2), three (+3) and more than three (more) (co-)medications. The size and colour of the dots represent the fraction of the non-statin-medicated or statin-medicated obese BMIS participants falling within that group. b, Difference in prevalence of the Bact2 enterotype in statin-medicated compared with non-statin-medicated obese BMIS participants, with decreasing co-medication threshold for inclusion of participants. For ‘all’, the total number of statin-medicated and non-statin-medicated obese BMIS participants were included (n = 474 biologically independent samples); then only subjects reporting three or fewer (≤3; n = 419), two or fewer (≤2; n = 369), one or fewer (≤1; n = 296) or no (0; n = 226) (co-)medications were included. The relative risk and respective significance level associated with the prevalence of the Bact2 enterotype given statin intake is provided above the bar plots (Fisher’s exact test, two-sided, *P < 0.05, relative risk = P(Bact2|Statin = Yes)/P(Bact2|Statin = No)).
