Clostridium scindens is the named gut species at the centre of the apparent Crohn’s-disease (CD) enrichment paradox in the IBD metagenome re-analyses 1. The central digest describes it as a protective species: a secondary bile-acid producer via 7α-dehydroxylation, a TGR5 activator, present at ~79 % prevalence in healthy individuals, and an inhibitor of C. difficile 2. It is also one of the named organisms in the ecotype feature-leakage analysis 3.
Aliases
- Aliases: C. scindens, Clostridium scindens 2
Bile-acid metabolism
The report cites the bai operon of Clostridium scindens-like bacteria as the mechanism of bile-acid 7α-dehydroxylation, and treats it as the literature anchor for the NB09c bile-acid network finding 1. See bile-acid-7alpha-dehydroxylation.
The CD-enrichment call and its re-analysis
The v2 preliminary IBD report (2026-03-28) called C. scindens CD-enriched at log₂FC = +2.67 from pooled Mann-Whitney differential abundance (DA — testing whether a taxon’s abundance differs between groups), a direct conflict with its documented protective role and ~79 % healthy prevalence; three explanations were considered possible at that stage — compositional artifact, strain heterogeneity, and ecotype mixing in the pooled analysis 2,1. See mann-whitney-u-test.
The preliminary interpretation spelled the three candidate mechanisms out as reasons the call was false: (1) compositional / relative-abundance artifact — Ruminococcaceae loss in severe patients inflates the relative abundance of surviving C. scindens strains; (2) strain heterogeneity — not all C. scindens strains carry the bai operon, so species-level presence ≠ functional presence; (3) ecotype mixing — pooled analysis averages across patient subgroups with different microbiomes 2.
At that point compositional-aware DA (ANCOM-BC / MaAsLin2 / linda) run within patient ecotypes was only the proposed fix, not a reported validation 2. The preliminary false-call framing is in tension with the later within-substudy result below, which found the species genuinely CD↑; the tension is not resolved by averaging the two, and what the digest later retracts is the H2c claim that stratification resolved the paradox, not this analysis proposal 2. Relevant concepts: compositional-robustness-of-differential-abundance-calls, gut-microbiome-ecotypes-as-patient-strata.
Compositional correction alone did not remove the signal — a null result for explanation (1): C. scindens remained CD-enriched under both methods (raw log₂FC +5.66; CLR Δ +1.25, where CLR is the centred log-ratio transform used to blunt compositional bias), leaving strain heterogeneity and ecotype mixing live at that stage 1.
The report’s own implication is that pooled Mann-Whitney on relative abundance is systematically under-sensitive for protective-species depletion, an artifact of compositional bias plus pooled-cohort heterogeneity, and that C. scindens is not resolvable at that level; resolution required a design that eliminates study confounding, namely the within-IBD-substudy CD-vs-nonIBD meta-analysis 1.
Retraction of the “paradox resolved” claim
Before the repair work, NB04 reported a 33-species within-ecotype Tier-A list (18 E1, 15 E3) with the H2c C. scindens paradox marked “RESOLVED by stratification”; two rigor-repair notebooks (NB04b + NB04c) then applied three evidence filters to every candidate 2.
H2c is retracted. Under the confound-free within-IBD-substudy CD-vs-nonIBD contrast, C. scindens is genuinely CD↑ with pooled CLR-Δ = +1.18, FDR = 1e-8 (FDR — false discovery rate, the multiple-testing-adjusted error rate) and 4/4 sign concordance across sub-studies; the NB04 within-ecotype “n.s.” call was a feature-leakage artifact of clustering samples on taxon abundances and then testing the same taxa within cluster, and under leave-one-species-out (LOO) refit C. scindens is CD↑ in both E1 and E3 once it is not part of the clustering input — this contradicts the earlier paradox-resolution claim rather than qualifying it 2,1. See selection-on-outcome-leakage and ecotype-clustering-validity.
The pitfalls digest records the reversal with intervals: C. scindens went from NB04 “n.s. within both ecotypes” (the basis for the “paradox RESOLVED” claim) to CD↑ in both E1 (CI +0.68, +0.87) and E3 (CI +1.13, +1.71) under LOO, making the within-ecotype n.s. call a self-selection effect 3.
The same confound-free analysis reversed several other NB04 within-ecotype calls — F. prausnitzii, R. hominis and L. eligens had been called CD↑ within both E1 and E3 and are CD↓ under the confound-free design — so the C. scindens reversal is part of a systematic artifact of feature leakage plus the substudy × diagnosis confound, not an isolated species-level correction; critically, both compositional-bias-aware DA methods tried on the within-ecotype subsets (CLR-MW and LinDA) share that bias, so agreement among within-ecotype methods alone does not resolve the direction 1.
Status as an intervention target
Despite being CD↑ under the confound-free design, C. scindens was excluded at the A4 screening step on the curated protective list, alongside Anaerostipes hadrus (−0.32 confound-free effect) and Roseburia faecis (−2.74 effect) 1. This bears on ecological-cost-of-microbiome-target-depletion.