The Gut–Brain Axis: Part One: Five Routes, Not One

Abstract

The gut–brain axis is the most enthusiastically marketed idea in nutritional science and one of the better documented at the mechanistic level. Both are true, which is what makes it hard to read honestly.

This is the first of three parts, one per review. It follows the 2024 synthesis in Signal Transduction and Targeted Therapy to establish what physically connects the intestine to the brain — and there is not one connection, there are five.

It also contains the most useful sentence we have found in this literature, which is a warning the authors issue about their own field's methods.

1.The Scale of the System

The human gastrointestinal tract holds the most diverse microbial community in the body — bacteria, fungi, viruses and archaea. Around 2,000 bacterial species have been identified, and the collective gene content of the gut microbiota vastly exceeds that of the human genome.

That gene count is why the field exists. A community with that much metabolic capacity produces compounds the human body cannot make for itself, and some of those compounds are biologically active well beyond the intestine.

The review by Jian Sheng Loh, Wen Qi Mak and colleagues makes an argument about how to study this that is worth carrying forward: examining the gut microbiota through a functional metabolic lens — what it produces — is more valuable than focusing on taxonomic or phylogenetic composition. What matters is the chemistry, not the census.

2.Five Routes, Not One

Popular accounts collapse this into a single gut–brain connection, usually the vagus nerve. The review describes bidirectional communication through five distinct systems: the immune system, the vagus nerve, the enteric nervous system, endocrine signalling, and circulating microbial metabolites.

The distinction matters because these routes differ in speed, in reversibility and in how easily they can be targeted. A signal travelling along a vagal afferent is a different kind of event from a metabolite circulating in blood for hours, and an immune signal is different again.

A product claiming to act on the gut–brain axis is claiming to act on one of these, and almost never says which.

3.Three Classes of Messenger

Short-chain fatty acids are the best-characterised. Microbial fermentation of non-digestible dietary fibre produces acetate, propionate and butyrate, which together comprise around 95 per cent of the total short-chain fatty acid pool. They act as endogenous ligands for G-protein-coupled receptors and as inhibitors of histone deacetylases — a direct route from bacterial metabolism to gene expression. Part Three covers them in detail.

Trimethylamine N-oxide runs the other way. Derived from microbial metabolism of dietary choline and carnitine, it promotes microglial activation, neuroinflammation, and amyloid and tau pathology. The same system that produces protective messengers produces this one.

Bile acids form the third class. Ursodeoxycholic acid and its taurine conjugate act on the nuclear farnesoid X receptor and on the membrane receptor TGR5, both present on microglia and neurons, inhibiting neuroinflammation through direct and indirect pathways.

These are named molecules acting on named receptors. Whatever is unresolved about clinical application, the mechanism is not vague.

4.The Sentence the Field Should Quote More Often

The dominant experimental design in this area transfers human microbiota into rodents and measures behavioural or neurological outcomes. It is powerful, because it permits manipulation impossible in people.

The review flags a problem with the resulting literature more bluntly than we have seen stated elsewhere: the excessively high rate — 95 per cent — of positive results and causal claims in human microbiota-associated rodents demands caution against overinterpreting and overstating the causal implications of these findings.

A field in which ninety-five per cent of experiments confirm the hypothesis has not found a very robust effect. It has a publication and analysis problem, and these authors are saying so about their own method, in a review of their own field.

They add the more general difficulty. Establishing a definitive causal relationship between an altered microbiome and disease remains challenging, because it is hard to determine whether observed alterations are causative, consequential, or merely a bystander response to the disease.

That is three possibilities for every microbiome association ever reported, and most reporting considers only the first.

5.Where the Interventions Stand

Clinical translation is under way but early, and mostly not in the areas the consumer market addresses. The trials cited include tauroursodeoxycholic acid combinations in Alzheimer's disease and in amyotrophic lateral sclerosis — pharmaceutical development against defined diseases, not supplementation in healthy adults.

On probiotics the review is unsentimental. Their notable limitation is suboptimal therapeutic efficacy, attributed primarily to vulnerability to low gastric pH and digestive enzymes, resulting in inactivation and impaired delivery.

It also lists recurring methodological weaknesses in the underlying studies: absent metabolomic analysis, absent microbiome profiling before and after intervention, and risk of bias from relative-abundance measures.

A trial that changes a microbiome without measuring the metabolites it produces has not tested the mechanism it invokes.

Editorial Comment

MindHeaven® makes no claim relating to the microbiome, and none of our formulations is designed to act on it. We have no probiotic product and this series is not a preface to one.

We open with the anatomy because the marketing skips it. "Supports the gut–brain axis" is a phrase that survives precisely by not specifying which of five routes, which messenger, or in whom — and a claim that cannot be made specific cannot be tested.

Part Two takes up the metabolites themselves, and a problem the mechanism cannot solve: whether a molecule found in the blood of sick people is causing their illness, marking it, or merely accompanying it.