• gastrointestinal physiology
• intestinal microbiology

We read with great interest the study by Asnicar and colleagues, describing the exploration of gut microbiota composition in relation to gut transit time using the ‘blue dye’ method.1 In agreement with previous research, the authors provided convincing evidence that longer gut transit times are linked to increasing relative abundances of microbial species (ie, Akkermansia muciniphila, Bacteroides spp. and Alistipes spp.).1–3 However, despite practical advantages, the ‘blue dye’ method does not allow to distinguish between segmental transit times.

In a recent study, we investigated segmental transit time and passage-related variation in pressure and pH in 22 individuals (11 subjects with normal weight and 11 subjects with obesity) using a wireless motility capsule (SmartPill^©).4 During capsule passage, participants were asked to collect a faecal sample. Samples were subjected to 16S rRNA gene amplicon sequencing according to Valles-Colomer et al.5 Associations between clinical and SmartPill^© variables and faecal microbiota community variation were assessed using single and stepwise multivariate distance-based redundancy analyses (dbRDA). Next, we explored associations between the microbiome covariates identified and participants’ enterotypes using Kruskal-Wallis test with post-hoc pairwise Dunn’s test.6 Finally, correlations between relative abundances of genera and significant variables were assessed using Spearman correlations. Two-sided p values were adjusted for multiple testing using the Benjamini-Hochberg method.

As single explanatory variables, we identified a significant association between faecal community variation and body mass index (BMI), small-intestinal transit, small-intestinal pH and colonic transit (single dbRDA, p<0.05; figure 1A; table 1). However, only BMI and small-intestinal pH were observed to provide a significant, non-redundant contribution to genus-level microbiome diversification of 11.78% and 3.36%, respectively (stepwise dbRDA, false discovery rate (FDR) <0.05). All significant covariates of community variation were found to be associated with enterotype classification of faecal samples (Kruskal-Wallis test, FDR<0.05; figure 1B). More specifically, pairwise comparisons indicated that Ruminococcaceae (Rum)-enterotyped samples were associated with longer colonic transit than their Prevotella (Prev)-counterparts (Dunn, FDR=0.04). Additionally, individuals hosting a Bacteroides2 (Bact2)-enterotyped microbiota were characterised by a higher BMI (Dunn, FDR=0.02) and shorter small-intestinal transit (Dunn, FDR=0.02) than Bacteroides1 (Bact1)-carriers.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Faecal microbiota community variation among the SmartPill^© cohort. (A) Visualisation of the faecal microbiota community variation among the SmartPill^© cohort coloured per enterotype and projected on the FGFP background. Clinical and SmartPill^© variables that have a significant explanatory power on community variation are depicted by arrows. The size of the arrow represents their effect size (single dbRDA). (B) Visualisation of the association between the significant covariates and enterotype classification of the faecal samples. (C) Visualisation of the association between the significant covariates and genera relative abundances. The colour gradient represents the Spearman Rho correlation coefficient, while marker size is linked to the significance level (large marker: FDR<0.05; small marker FDR>0.05. Bact1, Bacteroides1; Bact2, Bacteroides2; BMI, body mass index; CTT, colonic transit time; dbRDA, distance-based redundancy analyses; FDR, false discovery rate; FGFP, Flemish Gut Flora Project; Prev, Prevotella; Rum, Ruminococcaceae; SmartP, SmartPill^©; SI pH, small-intestinal pH; SITT, small-intestinal transit time.

Sixteen genera were observed to be significantly associated with the covariates of microbiota community composition identified (Spearman, FDR<0.05; figure 1C). While relative abundances of 12 genera correlated with BMI, two were observed to be negatively associated with small-intestinal pH (Bacteroides, FDR=0.03 and Flavonifractor, FDR=0.03). While no taxa correlated significantly with small-intestinal transit, six genera were linked to colonic transit, five positively (Alistipes, FDR=0.03; Clostridium IV, FDR=0.003; Faecalicoccus, FDR=0.02; Methanobrevibacter, FDR=0.01; Phascolarctobacterium, FDR=0.03) and one negatively (Parasutterella, FDR=0.03). Among the former, Methanobrevibacter and Clostridium IV were additionally found to be associated with a lower BMI.

In conclusion, we observed that faecal microbiota community variation was linked to BMI and gastrointestinal (GI) conditions. In our cohort, a shorter small-intestinal transit was associated with the Bact2-enterotype, while a longer colonic transit was associated with the Rum-community type. Moreover, our findings revealed two genera to be related to small-intestinal pH and six genera to colonic transit. These findings are in agreement with the findings of Asnicar and colleagues, and prior publications.1 3 However, while they link gut microbiota to whole-gut transit, our findings provide insights into the link with segmental transit time. Moreover, these findings emphasise the importance of other GI conditions including small-intestinal pH.