Does the gut microbiome causally influence depression, or only correlate with it?

The cleanest causal evidence in this whole question does not come from people. Move the gut community of a depressed human donor into a germ-free or antibiotic-treated rodent, and the behaviour tends to travel with it: the recipient animals show more of the immobility and reduced reward-seeking that stand in for low mood, while animals given a healthy donor's microbes do not. ◆3 families The effect is not pinned to a single bug; in the mouse work it tracks a broad shift in how the host handles carbohydrate and amino-acid metabolism, not one villain species. ◆2 families

What lifts this above a curiosity is that it has been seen more than once — in more than one species and more than one laboratory. ◆3 families In the council's language, the animal leg of this question is close to converged: under the controlled poverty of a gnotobiotic animal, the microbiome can be sufficient to move behaviour.

But sufficiency in a germ-free rodent is not the same as causation in a person, and this is exactly where an early over-reach was caught. One family boarded, at high confidence, that because transplant transfers the phenotype in rodents the human microbiome is therefore a causal driver of human depression. ◑contested The verification ladder let the animal claims through and held that one back: the rodent result does not entail the human one, and Gemini's methods critique — that germ-free physiology is not human physiology — was upheld.

In humans the evidence is real but quieter. The largest population cohorts find that people with depression are reliably missing some of the same microbes — butyrate-producers such as Coprococcus, along with Dialister — a signal strong enough to replicate across independent datasets. ◆3 families It is not simply an artefact of medication: the depletion survives when antidepressant use is statistically corrected for. ◆2 families The same faecal metagenomes carry microbial pathways for GABA and a dopamine metabolite that correlate with mental-health measures, which is where a plausible signalling story starts. ◆2 families

Two honest qualifications keep this from being more than it is. First, magnitude: the microbiome accounts for only a small slice of who is and isn't depressed — a genuine signal, not a diagnostic. ◆3 families Second, and decisively, these are observational cohorts. That people with depression carry a different microbiome does not tell you which came first, and the claim that the cohorts show the microbiome causally contributes to depression is the one the council marked unverified and left contested. ◑contested

If the microbiome causes depression, then feeding people the right microbes should treat it. Two small randomised trials are the ones everyone cites. In healthy adults, a four-week multispecies probiotic reduced cognitive reactivity to sad mood — less rumination, fewer aggressive thoughts. ◆2 families In adults already taking antidepressants for major depression, an add-on probiotic was well tolerated and nudged depression and anxiety scores in the right direction against placebo. ◆2 families

Both are small — tens of participants — and both describe themselves as pilots, not verdicts. ◆3 families They also lean on short follow-ups and self-report or surrogate outcomes rather than clinician-rated remission. ◆2 families Pool the wider trial literature and a modest reduction in depressive symptoms does appear, but it shrinks, and heterogeneity and publication-bias worries grow, once you restrict to people with a clinical diagnosis rather than stressed volunteers. ◑contested

So the trials neither confirm nor kill the causal story. What the council would not let stand was the leap from them to the clinic: the claim that this evidence is enough to recommend probiotics as a treatment for major depression was marked unverified, and Mistral's dissent — two pilot trials with surrogate endpoints do not clear that bar — was upheld. ◑contested

The hardest part of this question is everything that moves the microbiome and mood together. Diet is the obvious one: what you eat shapes the gut community directly, and a dietary-improvement trial (SMILES) cut depression symptoms sharply with no probiotic in sight. ◆3 families Its effect was large — a Cohen's d near 1.2, bigger than the pooled psychobiotic effect — which makes diet a heavyweight confounder rather than a footnote. ◆2 families

Treatment confounds too. Antidepressants themselves change gut bacteria, so a snapshot comparing depressed and non-depressed microbiomes is partly reading the drugs, not the disease. ◆3 families And the arrow may simply run backwards: depression's own appetite loss, inactivity, altered motility and stress hormones are a plausible route by which low mood reshapes the gut, rather than the other way round. ◑contested

The one tool that can, in principle, cut through confounding — Mendelian randomisation, which uses inherited genetic variants as a natural experiment — has so far given inconsistent and largely null answers here, with taxa changing sign between analyses. ◑contested Put the legs together and the honest heartwood is narrow but firm: the association is robust, a mechanism is plausible, and causation in humans remains unproven with confounding still in the room. ◆4 families