Stanford Fermented Food Study (Cell, 2021): What It Showed and What It Doesn't

What the Stanford fermented food study found: the short answer

The Stanford fermented food study, published in Cell in 2021 by Wastyk and colleagues, compared two ten-week diets in healthy adults: one high in fermented foods and one high in fibre.^[1] In the fermented-food arm, gut microbiome diversity rose steadily and 19 inflammatory serum proteins, including interleukin 6, fell. In the fibre arm, diversity stayed stable on average. The trial’s primary immune outcome did not change in either arm.^[1]

So the study is good evidence that a diet rich in fermented foods can shift the gut microbiome and some immune markers in healthy adults. It is not evidence that any single product does the same, and it says nothing about treating or preventing disease. This explainer sets out what was tested, what was found, where the evidence stops and what that means for US brands developing fermented foods.

Why the Stanford fermented food study matters

Much of the earlier evidence on fermented foods and the gut came from observational studies or from single products tested in isolation. The Stanford trial, run by the groups of Justin Sonnenburg (microbiology) and Christopher Gardner (nutrition), instead assigned healthy adults to a whole-diet change and measured the stool microbiome and the immune system side by side over several months.^[1][2] That design is why brands quote it so often, and why it deserves a careful read.

It also opened a research programme. The same Stanford team has since enrolled 135 pregnant women in FeFiFo-MOMS, a four-arm trial of fermented foods, fibre, both, or no change,^[3] and a USDA Agricultural Research Service project in Davis, California, is testing fermented vegetables against acidified vegetables in older adults.^[4] For the microbes behind these foods, see FRL’s guide to microbes used in the food and beverage industry.

What the 2021 Cell study tested

Who took part

Of 381 people screened, 39 were assigned to a diet: 37 by simple randomisation and 2 placed in the high-fibre arm at their own request. One person dropped out and two were excluded after being prescribed antibiotics, leaving 36 healthy adults (25 women, 11 men) in the analysis, 18 per arm.^[1] The trial is registered as NCT03275662, the Fermented and Fiber-rich Food (FeFiFo) Study.^[2]

How the diets were delivered

After a three-week baseline, each arm had a four-week ramp-up and a six-week maintenance phase (the ten-week intervention), followed by a four-week choice phase in which participants ate as they wished. Stool and blood were sampled throughout.^[1] The fermented-food arm ate foods such as yoghurt, kefir, fermented cottage cheese, kombucha, vegetable brine drinks and fermented vegetables including kimchi, rising from 0.4 to 6.3 servings a day on average. The fibre arm added legumes, whole grains, seeds, nuts, vegetables and fruit, rising from 21.5 g to 45.1 g of fibre a day.^[1]

Fermented foods vs a high-fibre diet

The fibre arm is what makes the study informative. Fibre is the standard advice for feeding resident gut microbes, so the comparison asks whether eating live-microbe foods does something different. It is a comparison of two active diets, though: there was no group that kept eating as usual, and neither participants nor dietitians could be blinded to their arm.^[1]

Source: Wastyk et al., Cell (2021).^[1]

Headline findings on microbiome diversity and inflammatory markers

Microbiome diversity rose with fermented foods

In the fermented-food arm, microbiome alpha diversity increased steadily over the intervention and stayed higher through the choice phase. The number of servings eaten correlated with the rise, but less strongly than time on the diet did.^[1] Notably, the new microbial taxa that drove the increase were largely not the microbes found in the foods themselves, so the effect was not simply the foods’ cultures taking up residence.^[1]

Lower inflammatory proteins in the fermented-food arm

The fermented-food arm showed lower levels of 19 serum proteins (cytokines, chemokines and other inflammatory proteins), including interleukin 6, interleukin 10 and interleukin 12b. Immune cell signalling showed reduced activation in 14 of 60 cell-type-specific responses across four cell types.^[1] These are blood markers in healthy adults, not clinical outcomes. The trial’s prespecified primary outcome, a cytokine response score, did not change in either arm.^[1]

What happened in the high-fibre arm

Average microbiome diversity stayed stable, while microbiome-encoded carbohydrate-active enzymes that break down fibre increased. Immune responses split participants into three groups, one of which showed broad increases in immune activation.^[1] That does not make fibre unhelpful; it shows a different, more individual response over a short period.

Source: Wastyk et al., Cell (2021).^[1]

Reviews discuss microbial metabolites, such as short-chain fatty acids, and immune signalling as possible routes for effects like these,^[5] while noting that controlled human outcome data on fermented foods remain limited.^[6]

What the study does and does not prove

The authors list the limits themselves: 18 people per arm, which limits statistical power; no control arm; a short intervention, so durability is unknown; healthy adults only; and mechanisms and causality still to be established.^[1] Two more matter for brands. The fermented arm ate a mix of foods at about six servings a day, so the result cannot be pinned to one food, strain or dose. And the immune link is shown through markers, not through fewer infections or better health outcomes.

Formulator’s note. A trial that tests a diet is not a trial that tests your product. Before citing the study near your own product, define strain identity, culture count at the end of shelf life and serving size, and be clear that the study is background, not product evidence.

What it means for fermented food brands and claims

For fermented food product development, the study is context, not a substitute for product data. A brand can say that research on fermented foods and the gut microbiome is active; it cannot transfer the trial’s results onto its own product. In the US, FDA separates structure/function claims, which describe an effect on normal body structure or function and must be truthful and not misleading, from health claims about reducing the risk of a disease, which need FDA authorisation or enforcement discretion for a qualified claim. Claims to treat or cure disease are drug claims.^[7][8]

Two FDA points are easy to miss. Structure/function claims on conventional foods should focus on effects derived from nutritive value, and FDA is likely to draw the line between structure/function and disease claims for foods as it does for dietary supplements.^[8] In the EU, health claims must be authorised and substantiated under Regulation (EC) No 1924/2006, so US-style gut-health wording cannot simply be carried across.^[9] FRL’s explainer on probiotic claims in the EU and US covers the label wording in detail, and the cluster’s US rules for fermented foods and drinks covers alcohol limits and labelling.

The practical translation: use the study to justify why live-culture work is worth doing, then build product-specific evidence. The cluster pillar on formulating live-culture fermented foods covers strain, dose, matrix and survival, and FRL’s guide to traditional vs modern fermented beverage techniques covers controlled fermentation.

Open questions and follow-up research

Several questions remain open: which foods, strains and doses drive the effect; how long it lasts after intake falls; whether it holds in other groups; and whether marker changes translate into clinical benefit. Follow-up work is under way. FeFiFo-MOMS randomised 135 women in early pregnancy to more fibre, more fermented foods, both, or no dietary change, and follows mothers and infants to 18 months after birth.^[3] The USDA Davis project includes a pilot of fermented vegetables, which was paused in 2025 while the vegetable slurries were optimised to meet lactic acid bacteria and safety targets, before a planned crossover trial in adults aged 50 to 75.^[4] Until results land, treat the Stanford fermented food study as the foundation for good questions, not for claims.

For brands & formulators

If you are building a fermented or live-culture product for the US, start from what the Stanford evidence can and cannot carry, then generate data for your own product. FRL’s Microbiome Solutions team designs strain identity, culture count and shelf-life studies for fermented and probiotic products; we run testing or arrange it through accredited partner laboratories. Our new food product development team takes the formula from bench to pilot, and our regulatory and scientific research team reviews claim wording against FDA categories before launch.

• Claim review: which messages your evidence supports in the US, and what the EU would need.
• Product evidence plan: strain, count at end of shelf life, serving size and stability.
• Scale-up planning: see the next step in this series, fermentation scale-up for foods and drinks.

Talk to our Microbiome Solutions team or contact us with your product brief.

In this series

Fermented foods and live cultures (US): the study path.

• Stanford fermented food study explained (You are here): what the Cell 2021 trial showed and what it does not.
• Fermentation scale-up for foods and drinks: seed trains, vessel control and keeping cultures viable from bench to plant.
• Formulating live-culture fermented foods (pillar): strain, dose, matrix and survival.
• US rules for fermented foods and drinks: alcohol limits, labelling and claims.
• Live-culture survival trial in a fermented beverage: a worked example of CFU and pH through shelf life.
• Fermented foods and drinks FAQ: formulation, regulation and shelf-life questions answered.
• Service: Microbiome Solutions.

1. Wastyk, H. C., Fragiadakis, G. K., Perelman, D., Dahan, D., Merrill, B. D., Yu, F. B., Topf, M., Gonzalez, C. G., Van Treuren, W., Han, S., Robinson, J. L., Elias, J. E., Sonnenburg, E. D., Gardner, C. D., & Sonnenburg, J. L. (2021). Gut-microbiota-targeted diets modulate human immune status. Cell, 184(16), 4137–4153.e14. https://doi.org/10.1016/j.cell.2021.06.019 (free author manuscript: PMC9020749) (cited for enrolment and analysis numbers, the diet phases, intake changes, diversity, inflammatory protein, immune signalling and CAZyme results, the unchanged primary outcome, the origin of new taxa and the authors’ limitations).
2. Stanford University. (2017, updated 2023). Fermented and Fiber-rich Food (FeFiFo) Study (NCT03275662). ClinicalTrials.gov. https://clinicaltrials.gov/study/NCT03275662 (cited for the trial registration, title and completed status).
3. Ward, C. P., Perelman, D., Durand, L. R., Robinson, J. L., Cunanan, K. M., Sudakaran, S., Sabetan, R., Madrigal-Moeller, M. J., Dant, C., Sonnenburg, E. D., Sonnenburg, J. L., & Gardner, C. D. (2025). Effects of fermented and fiber-rich foods on maternal & offspring microbiome study (FeFiFo-MOMS): Study design and methods. Contemporary Clinical Trials, 150, 107834. https://doi.org/10.1016/j.cct.2025.107834 (cited for 135 women enrolled in early pregnancy, the four diet arms and follow-up to 18 months after birth).
4. USDA Agricultural Research Service. (2025). Diet, Microbiome and Immunity Research, Davis, CA: Effect of diet on gut microbiome, gastrointestinal health, and immune function (project 2032-10700-002-000D), FY 2025 progress report. https://www.ars.usda.gov/research/project?accnNo=446030&fy=2025 (cited for the planned fermented vs acidified vegetable crossover trial in adults aged 50–75 and the paused fermented vegetable pilot).
5. Park, I., & Mannaa, M. (2025). Fermented foods as functional systems: Microbial communities and metabolites influencing gut health and systemic outcomes. Foods, 14(13), 2292. https://doi.org/10.3390/foods14132292 (cited for proposed mechanisms, including short-chain fatty acids and immune regulation).
6. Sampsell, K., Marcolla, C. S., Tapping, S., Fan, Y., Sánchez-Lafuente, C. L., Willing, B. P., Reimer, R. A., & Burton, J. P. (2025). Current research in fermented foods: Bridging tradition and science. Advances in Nutrition, 16(12), 100554. https://doi.org/10.1016/j.advnut.2025.100554 (cited for methodological limits and the need for well-powered clinical trials).
7. U.S. Food and Drug Administration. (n.d.). Label claims for conventional foods and dietary supplements. https://www.fda.gov/food/nutrition-food-labeling-and-critical-foods/label-claims-conventional-foods-and-dietary-supplements (cited for the definitions of health claims, qualified health claims, structure/function claims and dietary guidance).
8. U.S. Food and Drug Administration. (n.d.). Structure/function claims. https://www.fda.gov/food/nutrition-food-labeling-and-critical-foods/structurefunction-claims (cited for the nutritive-value focus of conventional-food structure/function claims and the structure/function vs disease claim line).
9. European Parliament and Council. (2006). Regulation (EC) No 1924/2006 on nutrition and health claims made on foods. Official Journal of the European Union, L 404, 9–25. https://eur-lex.europa.eu/eli/reg/2006/1924/oj (cited for the requirement that health claims in the EU be authorised and substantiated).