Fig. 1

Enterotypes differ in stool energy density, intestinal transit time, microbial alpha-diversity, and body weight. a The study included baseline measurements of 85 overweight subjects. Prior to collection of the stool and urine samples used in the study, habitual dietary intake was estimated based on 4-day dietary registrations and intestinal transit time was estimated using radio-opaque markers from day 1 to 6 where participants maintained their habitual diet and lifestyle. The collected stool sample was used to estimate dry stool energy density as a measure of gut microbial energy extraction, bacterial cell counts, the gut microbiome community structure, and short-chain fatty acids. Microbial-derived metabolites were measured in the urine samples. b Principal coordinate analysis plot using Bray-Curtis distance of bacterial relative abundance on the genus level as distance metric. Symbols are samples, with shape/colour indicating assigned enterotype (red circles: Bacteroides (B-type), n = 35; yellow diamonds: Prevotella (P-type), n = 16; green squares: Ruminococcaceae (R-type), n = 34). Relative abundance of the taxa used for enterotype assignment (black arrows) and values for dry energy, Shannon index and transit time (purple arrows) were plotted supplementary (i.e. projected after ordination). Horizontal and vertical axis explain 20% and 12% of variation, respectively. Subjects stratified into three enterotypes differed in c stool energy density (n = 77), d intestinal transit time (n = 85), microbiome alpha-diversity as reflected by e Shannon Index and f observed richness (n = 85), as well (g) body weight (n = 85). Differences between enterotypes were detected using the Mann-Whitney U test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001