Skip to main content
. 2023 Aug 8;11:178. doi: 10.1186/s40168-023-01604-z

Fig. 2.

Fig. 2

Microbiome compositional differences of responders, partial-responders, and non-responders to lifestyle interventions. a Bray–Curtis dissimilarity of longitudinal samples in subjects from no and lifestyle interventions. The two longitudinal no-intervention cohorts were combined in the first box. Bray–Curtis indexes dot colors indicate the microbiome response classification group by coral, blue, and green for non-responders, partial-responders, and responders, respectively. The two red dash lines represent the mean + SD and mean + 2*SD of the Bray–Curtis dissimilarities in the no-intervention cohorts as cutoffs to differentiate significant microbiome compositional changes from normal microbiome fluctuation. (CTL: study with no intervention; I_MWP: intervention study with multidisciplinary weight-loss program; I_LCD: intervention study with low-carbohydrate diet; I_HFD: intervention study with high-fiber diet; I_HRS: intervention study with high-resistant starch; I_ETP: intervention study with exercise training program). b Principal coordinate analysis of Bray–Curtis dissimilarity in non-responders, partial-responders, and responders to lifestyle interventions. c Relative abundances of the significant species using ordinal regression among non-responders, partial-responders, and responders groups (p < 0.05)