🎧 Listen to this article:
Schematic links between cocoa polyphenols, the gut microbiome and mood
Fig. 1. Possible biological links. Arrows represent hypotheses and mechanisms, not evidence that chocolate treats mood disorders.

The most defensible conclusion is narrower than a press headline: in a small three-week trial, 30 g of 85% cocoa chocolate per day was associated with lower negative affect and microbiome changes. That is an interesting signal, not evidence of therapy or a universal recipe for better mood.

1. What was actually studied?

The starting point is the randomized trial by Shin and colleagues, published in the Journal of Nutritional Biochemistry in 2022 (PMID 34530112). Healthy adults aged 20–30 were assigned to one of three groups: 18 participants ate 30 g of 85% cocoa chocolate daily, 16 received a 70% cocoa product and 14 ate no chocolate. The intervention lasted three weeks.

Mood was measured with the PANAS questionnaire. Stool samples from the 85% cocoa and control groups were also analysed using 16S rRNA sequencing. The 85% group showed lower negative affect and greater microbiome diversity than the no-chocolate group. The relative abundance of selected bacteria, including Blautia obeum and Faecalibacterium prausnitzii, also changed.

How to read this result: a correlation between mood scores and bacterial changes does not establish causality. We do not know whether cocoa changed the microbiome, whether the microbiome influenced mood, or how expectations, diet and other behaviours contributed.

2. Why might cocoa reach the gut?

Cocoa contains flavanols such as epicatechin, catechin and procyanidins, as well as fibre, theobromine and a small amount of caffeine. Bean processing changes their proportions. Fermentation, drying, roasting and alkalization can lower some polyphenol levels, so the stated cocoa percentage is not equivalent to a measured flavanol dose.

A substantial proportion of polyphenols is not absorbed in its original form in the small intestine. In the colon, microorganisms can break larger molecules into smaller metabolites that may interact with the intestinal lining, immune system and metabolism. This is biologically plausible, but “can be transformed” does not mean “produces a clinical benefit”. Microbiome composition differs between people, as does the response to the same cocoa portion.

Evidence ladder for claims about chocolate, the microbiome and mood
Fig. 2. From mechanism to clinical conclusion: each level needs its own evidence.

3. Earlier studies support the hypothesis, but do not settle it

2011 crossover trial. In 22 healthy volunteers, a high-cocoa-flavanol drink (494 mg/day for four weeks) increased bifidobacteria and lactobacilli and reduced clostridia compared with a low-flavanol drink. This is informative, but it involved a standardized drink, not any dark chocolate bar.
2019 stress study. Twenty-six adults consumed 25 g/day of high-polyphenol or control chocolate for four weeks. Selected salivary cortisol measures fell in the high-polyphenol group, but there was no significant between-group difference in overall positive or negative affect. The small sample makes this exploratory evidence.
2024 study. In a randomized crossover study, only eight healthy adults received 900 mg/day of cocoa-derived flavanols or placebo for eight days. There was no difference in the abundance of Bifidobacterium or Lactobacillus; alpha diversity was lower after the intervention. The authors noted that this does not support a simple model in which flavanols always increase “good bacteria”.

This inconsistency is scientifically useful. It shows that effects may depend on the food matrix, dose, duration, measurement method, baseline microbiome and overall diet. The 85% chocolate trial tested a food, while the 2024 trial tested an isolated flavanol preparation; these are not interchangeable interventions.

4. What did the 2025 studies add?

Newer publications strengthen the cautious conclusion about the gut, but they do not provide a large clinical study that changes the interpretation of the mood evidence. They also examined different products, populations and biological levels; their findings cannot be combined as though they were one intervention.

Japanese study: constipation and the microbiome. In a randomized study, 31 women with constipation (15 in the white-chocolate group and 16 in the dark-chocolate group) consumed 25 g per day for two weeks of a product containing 72% cocoa or a product without cocoa proteins. In the dark-chocolate group, bowel-movement frequency and estimated stool amount increased more than after white chocolate. Faecal alpha diversity increased, as did the relative abundance of Faecalibacterium and Megamonas; increases in Anaerostipes, Butyricicoccus and Roseburia were reported relative to baseline. The authors linked these genera with butyrate production, but the study did not measure butyrate production directly. The result concerns women with constipation, one product and a short intervention; it does not show an effect on mood or a universal laxative effect.
Study in people with MASH. In a crossover study, 19 patients consumed 40 g per day of dark chocolate containing more than 85% cocoa or 40 g per day of milk chocolate containing less than 35% cocoa for 14 days. After dark chocolate, circulating LPS and zonulin levels were lower than before the intervention, and treatment-period differences were significant. This is a signal about endotoxaemia and markers associated with the gut barrier, not proof that the barrier was repaired, MASH was treated or well-being improved. The sample was small and the intervention lasted only two weeks.
Enterotype-dependent response — in-vitro study. In an incubation model using faecal samples and cocoa, the effect depended on the starting microbiome profile. In the Bacteroides group, Roseburia, Faecalibacterium and Agathobacter increased among other taxa, while the response in the Prevotella group was weaker. This is an interesting hypothesis about individual response, but a laboratory model is not a human feeding trial and does not demonstrate a clinical outcome.
Animal data. In a mouse study, cocoa polyphenols changed the microbiome in a site-specific manner: effects were stronger in intestinal contents and faeces than in mucosa-associated communities. The finding shows that a stool sample may not represent the entire gut ecosystem, but it cannot be transferred directly to humans.

The common message is that the response depends on the product, dose, duration, food matrix, baseline microbiome and health status. None of these publications establishes a health claim about improved mood or justifies calling ordinary chocolate a “prebiotic” without qualification.

5. What about the brain?

The gut–brain axis involves several routes: the vagus nerve, microbial metabolites, immune signals, hormones and diet-related metabolism. This does not mean that one bacterium produces a “happiness hormone” that travels straight to the brain after chocolate is eaten. That is a popular but overly simple shortcut.

Flavanol research also suggests possible vascular effects and direct actions of metabolites on neural cells. Cognitive findings remain mixed. In the COSMOS clinical subcohort, 573 older adults received cocoa extract providing 500 mg flavanols per day or placebo; after two years, there was no significant improvement in global cognition, episodic memory or executive function. This does not exclude short-term effects, but it weakens the claim that cocoa supplementation durably “boosts the brain”.

Keep the levels separate: a change in a mood questionnaire, a metabolite, microbiome composition and a diagnosed disease are four different levels of evidence. Articles and advertisements often collapse them into one claim.

6. How can consumers avoid false certainty?

  • Do not treat cocoa percentage as a laboratory result. 85% describes the recipe, not the flavanol dose or microbiome effect.
  • Check sugar and portion size. Chocolate is energy-dense; possible polyphenol benefits do not cancel excess sugar and calories.
  • Do not assume “raw” or “non-alkalized” automatically means better. Safety, quality and the actual composition also matter.
  • Treat 30 g as the dose in one study, not medical advice. The Shin trial did not establish an optimal dose for the population.
  • Consider caffeine and theobromine. People who are sensitive, pregnant or breastfeeding, or who have reflux, migraine or heart disease should discuss personal tolerance with a clinician.

7. What may be said on a label?

In the European Union, the authorized health claim for cocoa flavanols concerns maintaining the elasticity of blood vessels, which contributes to normal blood flow. Commission Regulation (EU) 2015/539 sets a condition of 200 mg cocoa flavanols per day, with a degree of polymerisation of 1–10, for specified product categories. It does not authorize claims that chocolate “improves mood”, “repairs the microbiome”, “treats the gut” or “protects the brain”.

An EFSA opinion is a scientific assessment, not a free-standing permission to use any marketing wording. Each statement must be assessed against its wording, product category and the conditions in food-claims law. A single 85% cocoa trial does not create a new authorized health claim.

Conclusion

High-cocoa chocolate is an interesting carrier of polyphenols, and research suggests possible relationships between flavanols, the microbiome and some gut functions. The Shin trial remains the key small signal concerning mood and the microbiome. Data published in 2025 add signals concerning bowel movements, endotoxaemia markers, and microbiome responses that vary by person and gut location, but they also come from small studies, in-vitro models and animals. They do not amount to evidence for therapy, a “prebiotic” effect of ordinary chocolate or guaranteed “brain nourishment”.

Educational material, not medical advice. Do not change treatment or replace professional care with chocolate. Persistent mood or gastrointestinal symptoms need proper clinical assessment.

Sources and further reading

  1. Shin JH et al., Journal of Nutritional Biochemistry (2022), PMID 34530112 — randomized 85% and 70% cocoa trial.
  2. Tzounis X et al., American Journal of Clinical Nutrition (2011), PMID 21068351 — high- and low-flavanol crossover trial.
  3. Suther C et al., Journal of Nutritional Science (2024), PMC11704926 — short-term cocoa-derived flavanols and the microbiome.
  4. Tsang C et al., Nutrients (2019), PMC6616509 — high-polyphenol chocolate, cortisol and mood.
  5. Sorrenti V et al., Nutrients (2020), PMC7400387 — cocoa polyphenol metabolism and microbiome interactions.
  6. Vyas CM et al., American Journal of Clinical Nutrition (2024), PMC11347806 — COSMOS and two-year cocoa-extract cognition results.
  7. Ito H et al., Bioscience of Microbiota, Food and Health (2025), PMID 40636155 — dark chocolate, constipation and the microbiome in Japanese women.
  8. Pannunzio A et al., Nutrition (2025), PMID 39693929, DOI 10.1016/j.nut.2024.112643 — dark chocolate, LPS and zonulin in patients with MASH.
  9. Kim J et al., Current Issues in Molecular Biology (2025), 47, 414 — in vitro enterotype-dependent responses to cocoa.
  10. Hayden M et al., Nutrients (2025), 17, 2876, PMID 40944264 — mouse study of site-specific microbiome changes.
  11. Commission Regulation (EU) 2015/539 — authorized cocoa-flavanol claim and conditions of use.
  12. The Epoch Times: How Chocolate Nourishes the Gut and Mind — journalistic starting point only; this article is an independent synthesis based on primary publications.