Do Gut Bacteria Decide Whether Supplements Help Your Mood?

Your gut may decide whether a supplement reaches your brain

Biohackers like clean rules: take fiber, take vitamin C, expect a mental upgrade. The better rule from the current evidence is messier. A supplement can change the gut and still do nothing noticeable for mood or cognition unless the person starts in the right biological state.

Two randomized controlled trials point in that direction. In one, Leyrolle et al. 2021 gave obese adults 16 g/day of native inulin or maltodextrin for three months, alongside calorie restriction and dietary counseling. In the other, Sim et al. 2025 gave young adults with suboptimal vitamin C status 500 mg vitamin C twice daily or placebo for four weeks. Both trials reported psychological benefits only in a subset of participants, and in both cases the subgroup signal tracked the gut microbiome.

That does not mean a stool test can tell you which supplement to buy. It means the old idea that an evidence-backed supplement should work about the same way in everyone looks weaker than it seemed.

The inulin trial: same prebiotic, different outcomes

The more provocative study is the inulin trial. Leyrolle and colleagues randomized 106 obese adults in a single-blind design to inulin or maltodextrin for three months while everyone also received calorie restriction and dietary counseling. At the group level, the inulin group showed moderate improvements in emotional competence and cognitive flexibility. The problem is that the available record does not report the effect sizes, confidence intervals, or p-values for those psychological outcomes, so you cannot judge how large or reliable the average benefit was from the abstract alone.

The headline result came later, in an exploratory subgroup analysis. Participants with higher baseline abundance of Coprococcus, a butyrate-producing genus, appeared more likely to benefit emotionally from inulin. That makes biological sense. Inulin is a fermentable fiber, and bacteria such as Coprococcus can turn fermentable carbohydrates into short-chain fatty acids, especially butyrate. Reviews such as Korenblik et al. 2026, Śliwka et al. 2025, and Marano et al. 2026 lay out why that matters: butyrate can strengthen gut barrier function, reduce lipopolysaccharide translocation, alter gene expression through histone deacetylase inhibition, and affect microglial activity. Those are plausible routes from gut fermentation to brain-relevant inflammation and neurotrophic signaling.

But plausible is not the same as proven. The Coprococcus finding was post hoc, not pre-specified. The report does not provide the subgroup sample size, the mood effect estimate, or the p-value for responders versus non-responders. It also did not correct for multiple comparisons across the full 16S microbiome matrix. In plain terms, the study searched many microbial features after the fact and found one interesting pattern. That is how useful hypotheses are born. It is also how false positives happen.

There is another complication. The apparent responders did not just have more Coprococcus. They also started with worse inflammatory and metabolic profiles, including higher IL-8, greater insulin resistance, and higher adiposity. Since the publication does not report a multivariable model that separates those factors, the microbiome signal may be tagging a broader inflammatory state rather than acting as an independent predictor. If so, the story is less "this bacterium predicts your response" and more "people with a certain dysbiotic, inflamed profile may have more room to improve." That is still interesting. It is just less precise.

The vitamin C trial: deficiency first, microbiome second

The vitamin C trial is cleaner in one sense and narrower in another. Sim and colleagues randomized 40 adults ages 20 to 39 with serum vitamin C below 50 μM to vitamin C at 500 mg twice daily or placebo for four weeks in a double-blind design. The study reported that supplementation increased the relative abundance of Bacillaceae and Anaerotruncus and reduced Desulfovibrio, a lipopolysaccharide-generating bacterium. Greater reduction in Desulfovibrio correlated with better performance on the Stroop interference task, and the responder group also showed higher BDNF, lower serum LPS, lower IL-10, and fewer classical monocytes.

The proposed chain is coherent: correct vitamin C deficiency, suppress Desulfovibrio, reduce circulating LPS, ease neuroimmune stress, then improve BDNF-linked and dopamine-linked cognitive and motivational outcomes. It is a sensible model, and it fits broader gut-brain work on LPS-driven inflammation described in Marano et al. 2026.

Still, the evidence is much thinner than the mechanism diagram. The total sample was 40. That is small for a randomized trial and especially small for subgroup inference. The available abstract does not report the correlation coefficients, fold changes, confidence intervals, or p-values for the microbiome-cognition links or for the responder versus non-responder comparisons. So the trial supports a hypothesis with a real biological backbone, not a ready-made clinical rule.

It does support one practical point more strongly than the inulin paper does: the entire signal came from people with suboptimal vitamin C status. That matters. This was deficiency-correction biology, not proof that vitamin C sharpens cognition in already replete adults. If your serum vitamin C is adequate, the trial gives you no reason to expect the same brain effect.

Why other trials matter just as much as the flashy ones

If you stop at the two positive RCTs, the story sounds almost personalized. Eat the right thing, feed the right bugs, feel better. The broader literature is less romantic and more useful.

Vaghef-Mehrabani et al. 2023 found that inulin reduced inflammatory biomarkers tied to depression in obese women with major depressive disorder who were also on a calorie-restricted diet. That supports the idea that inulin can help in metabolically and psychologically stressed populations. But the study did not stratify by microbiome, so it cannot confirm that Coprococcus was the reason.

Now look at the trials that did not show a mental payoff. Puhlmann et al. 2025 found that inulin changed the microbiome and improved bowel habits in adults with functional constipation. That is a real effect. It just was not a mood story. Ng et al. 2025 increased fiber intake in healthy adults with low dietary fiber intake using defatted rice bran-fortified bread. The microbiome changed, but wellbeing did not improve.

Those null or narrower findings are not awkward side notes. They help define the boundary of the claim. A gut change is not enough by itself. The host context matters: obesity, depression, inflammatory burden, nutritional deficiency, or some combination of them may be the real precondition for a psychological effect.

That same problem shows up in observational work. A 2025 cohort study by Ni et al. linked Mediterranean diet adherence, microbiota patterns, and cognitive decline in older adults with overweight, obesity, and metabolic syndrome, but individual cognitive responses varied. A 2021 observational study by Liu et al. associated higher fiber intake with lower depression and anxiety incidence in people with essential hypertension. Those studies support the general diet-microbiota-mental health link. They do not tell you which individual will respond to which supplement.

What the evidence actually supports right now

The gut-brain axis is real. Short-chain fatty acids, barrier integrity, lipopolysaccharide signaling, and neuroimmune pathways are not fringe ideas. The reviews cited above converge on that much. What remains unsettled is the stronger claim that baseline microbiome composition can prospectively tell you whether a given supplement will improve your mood or cognition.

Right now, that stronger claim rests on two small subgroup stories. One came from a single-blind inulin trial with post-hoc microbiome mining and missing subgroup statistics. The other came from a 40-person vitamin C trial in a deficiency-selected population, with an elegant mechanism but limited precision and incomplete quantitative reporting in the available abstract. Neither finding has independent replication.

That leaves a practical verdict less sexy than a personalized supplement stack and more defensible. If you want cognitive benefit from vitamin C, first confirm that your vitamin C status is actually low. If you are already replete, this evidence does not justify taking it for brain performance. If you want mood benefit from inulin, the evidence does not support treating a commercial microbiome report as a prescription pad. The interesting pattern is that people with more dysbiosis, more inflammation, or more deficiency seem more likely to respond. Healthy people chasing an extra edge should lower their expectations.

The current data support a baseline-first principle, not a bacteria-first shopping list.

The next step is obvious. Researchers need prospective trials that pre-specify microbial responder thresholds, power the subgroup analyses properly, report effect sizes and uncertainty, and replicate the result in independent cohorts. Until then, "your gut bacteria decide whether supplements work" is a promising hypothesis with some human evidence behind it. It is not yet a tool you can use with confidence, and it is definitely not a reason to confuse a sequencing report with an oracle. Your microbiome may be complicated enough already.