Heterogeneous associations of gut microbiota with Crohn’s disease activity

ABSTRACT

The multi-factorial involvement of gut microbiota with Crohn’s disease (CD) necessitates robust analysis to uncover possible associations with particular microbes. CD has been linked to specific bacteria, but reported associations vary widely across studies. This inconsistency may result from heterogeneous associations across individual patients, resulting in no apparent or only weak relationships with the means of bacterial abundances. We investigated the relationship between bacterial relative abundances and disease activity in a longitudinal cohort of CD patients (n = 57) and healthy controls (n = 15). We applied quantile regression, a statistical technique that allows investigation of possible relationships outside the mean response. We found several significant and mostly negative associations with CD, especially in lower quantiles of relative abundance on family or genus level. Associations found by quantile regression deviated from the mean response in relative abundances of Coriobacteriaceae, Pasteurellaceae, Peptostreptococcaceae, Prevotellaceae, and Ruminococcaceae. For the family Streptococcaceae we found a significant elevation in relative abundance for patients experiencing an exacerbation relative to those who remained without self-reported symptoms or measurable inflammation. Our analysis suggests that specific bacterial families are related to CD and exacerbation, but associations vary between patients due to heterogeneity in disease course, medication history, therapy response, gut microbiota composition and historical contingency. Our study underscores that microbial diversity is reduced in the gut of CD patients, but suggests that the process of diversity loss is rather irregular with respect to specific taxonomic groups. This novel insight may advance our ecological understanding of this complex disease.

KEYWORDS:

• Microbiota
• Crohn’s disease
• disease activity
• quantile regression
• microbial ecology

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data sharing statement

The study protocol of the South Limburg IBD cohort in which the current study was embedded has been published previously: https://academic.oup.com/ije/article/46/2/e7/3038105. All study subjects gave written informed consent prior to participation. Both studies have been approved by the Medical Ethics Committee of Maastricht University Medical Center and have been registered in the US National Library of Medicine (http://www.clinicaltrials.gov): NCT02130349 and NCT00775060, respectively. The 16S rRNA gene sequencing data are released in the European Nucleotide Archive. The accession number is: PRJEB62578 ERP147674. Example code of the LQMM and LME models can be found on the Github repository: https://github.com/susannepinto/Quantile-Regression-CD.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19490976.2023.2292239

Additional information

Funding

The Dutch Organization for Scientific Research (NWO) financed this research through the program Complexity in Health and Nutrition (NWO grant 645.001.002; www.nwo.nl/onderzoeksprogrammas/complexiteit), with co-funding by the National Institute for Public Health and the Environment (RIVM) of the Netherlands.