Relationship Between Microbiota of the Colonic Mucosa vs Feces and Symptoms, Colonic Transit, and Methane Production in Female Patients With Chronic Constipation

Abstract

Background & aims: In fecal samples from patients with chronic constipation, the microbiota differs from that of healthy subjects. However, the profiles of fecal microbiota only partially replicate those of the mucosal microbiota. It is not clear whether these differences are caused by variations in diet or colonic transit, or are associated with methane production (measured by breath tests). We compared the colonic mucosal and fecal microbiota in patients with chronic constipation and in healthy subjects to investigate the relationships between microbiota and other parameters.

Methods: Sigmoid colonic mucosal and fecal microbiota samples were collected from 25 healthy women (controls) and 25 women with chronic constipation and evaluated by 16S ribosomal RNA gene sequencing (average, 49,186 reads/sample). We assessed associations between microbiota (overall composition and operational taxonomic units) and demographic variables, diet, constipation status, colonic transit, and methane production (measured in breath samples after oral lactulose intake).

Results: Fourteen patients with chronic constipation had slow colonic transit. The profile of the colonic mucosal microbiota differed between constipated patients and controls (P < .05). The overall composition of the colonic mucosal microbiota was associated with constipation, independent of colonic transit (P < .05), and discriminated between patients with constipation and controls with 94% accuracy. Genera from Bacteroidetes were more abundant in the colonic mucosal microbiota of patients with constipation. The profile of the fecal microbiota was associated with colonic transit before adjusting for constipation, age, body mass index, and diet; genera from Firmicutes (Faecalibacterium, Lactococcus, and Roseburia) correlated with faster colonic transit. Methane production was associated with the composition of the fecal microbiota, but not with constipation or colonic transit.

Conclusions: After adjusting for diet and colonic transit, the profile of the microbiota in the colonic mucosa could discriminate patients with constipation from healthy individuals. The profile of the fecal microbiota was associated with colonic transit and methane production (measured in breath), but not constipation.

Keywords: Irritable Bowel Syndrome; Microbiome.

Figure 1. Relationship between stool microbiota and colonic transit

The analysis shows microbiota evaluated with the RF algorithm that was associated with colonic transit expressed as continuous (A) or dichotomous variables (B). In A, the bar graph represents the relative abundance of all genera in participants with GC24 less than (Low GC24) and ≥ (High GC24) the median value (1.8) in all participants. The asterisk represents genera that were also significant by single taxon analysis. In B, colonic transit is categorized as normal (GC24 ≥1.4 [10^th percentile value in healthy women]) and slow (GC24 <1.4) colonic transit. (C) ROC curves showing the utility of i) stool microbiota at the genus level; ii) constipation status, age, and diet; and iii) both i) and ii) for discriminating between normal and slow colonic transit.

Figure 2. Relationship between colonic mucosal microbiota and constipation

(A) Differential abundance of bacterial taxa in healthy participants and constipated patients. All bacterial taxa so identified by single taxon based analysis with FDR control are shown. (B) Colonic mucosal microbiota discriminates between healthy participants and constipated patients. The relative abundance of only those taxa that were found to be significantly different between the two groups was used in the principal components analysis. Taxa at the family level are shown in bold. Individual colonic mucosal samples of healthy controls are presented as ‘O’ and constipated patients are presented as ‘X’. (C) ROC curves showing the utility of i) colonic mucosal microbiota at the genus level; ii) colonic transit, and age; and iii) both i) and ii) for discriminating between healthy participants and constipated patients.

Figure 3. Relationship between stool microbiota and breath methane production

(A) Differential abundance of bacterial taxa in participants with breath methane excretion above (high) and below (low) the median value of breath methane excretion in all participants (Area Under Curve). All bacterial taxa so identified by single taxon-based analysis with FDR control are shown. (B) Stool microbiota discriminates between high and low breath methane producers. The relative abundances of only those taxa that were found to be significantly different between the two groups were used in the principal components analysis. Taxa at the family level are shown in bold. Individual stool samples of high methane producers are presented as ‘O’ and low methane producers are presented as ‘X’. (C) Stool microbiota at the genus level predicts breath methane production. The boxes represent the distribution of predictive mean squared error (PMSE) over 1,000 permutations for breath methane as a continuous variable, using overall microbiota composition at the genus level. The PMSE was lower compared to “guess,” signifying that the microbiota predicted breath methane excretion expressed as a continuous variable (by Friedman Rank Sum test).

Figure 4. Conceptual framework of interactions between fecal and colonic mucosal microbiota, colonic transit, breath methane excretion, and constipation

To emphasize, the lines denote associations, not causality a solid line represents an association that remained statistically significant in the multivariate analyses; the dotted line represents an association that was only univariately statistically significant (Table 2); the asterisk denotes relationships that are supported by previous literature. While the mucosal microbiota is associated with constipation, independent of colonic transit, the fecal microbiota is associated with colonic transit and breath methane production.