Gut bacteria before a diet predicted who stayed well

The Crohn's disease exclusion diet is an emerging dietary therapy used to induce remission in Crohn's disease, but it does not work for everyone, and there is currently no way to tell in advance who will respond. Most of the microbiome evidence for the diet comes from children, leaving its effects in adults poorly described. If the composition of a patient's gut bacteria before starting the diet predicted whether their remission would last, that would open the door to choosing dietary therapy for the people most likely to benefit. This study asked whether such a microbial signature exists.

The investigators analysed microbiome data from a randomised controlled trial in adults with mild-to-moderate Crohn's disease, sampled at baseline and weeks 6, 12 and 24, and compared it with data from a paediatric randomised trial sampled at baseline and weeks 6 and 12, plus healthy paediatric controls. They divided adult patients into those achieving sustained clinical remission at both week 12 and week 24 and those who did not, then compared baseline microbial composition, diversity and functional potential between the groups. Functional profiling used gene ortholog annotations, linear discriminant analysis and metabolite inference.

Baseline microbial and functional profiles differed between patients who went on to sustain remission and those who did not. Sustained remission was associated with lower alpha diversity, higher relative abundance of Alistipes and Faecalibacterium, and increased flagellin gene expression. Functionally, the remission group was enriched for genes involved in redox balance, fatty acid metabolism and DNA repair, whereas the non-remission group showed elevated NAD biosynthesis, bacterial adhesion and pro-inflammatory pathways. Haemophilus and Prevotella were negatively linked to sustained remission. During diet-induced remission, both compositional and functional analyses showed the microbiome shifting toward a profile more closely resembling healthy paediatric controls.

This extends to adults a line of work largely established in children, and it complicates the common assumption that higher microbial diversity is always the healthier state — here, lower baseline alpha diversity tracked with better outcomes. The convergence of the treated microbiome toward healthy control profiles supports the idea that the diet is doing something biologically meaningful rather than merely restricting symptoms. But this is a secondary analysis of relative abundances, so the bacteria identified could be markers rather than drivers, and the comparison group for the microbiome shift is paediatric rather than adult. Prospective testing of these signatures as a predictive tool has not been done.

The interesting idea here is that response to a dietary therapy might be partly written into the gut microbiome before the diet even starts — a step toward personalised nutrition in inflammatory bowel disease rather than trial and error. It also punctures the tidy notion that more bacterial diversity is automatically better, since the opposite pattern appeared in those who did best. This is exploratory work in a specific clinical population using a medically supervised exclusion diet, not a general eating strategy. Nobody should read a list of bacterial genera here as something to chase with off-the-shelf products.