Skip to content
HabitSeek
Back to results

Include fermented foods regularly

Eating fermented foods — live yoghurt, kefir, kimchi, sauerkraut, miso, tempeh and similar — as a regular part of the diet rather than as an occasional garnish.

At a glance

  • Evidence: A high-fermented-food diet increased microbiota diversity over 17 weeks, but the two randomised trials disagree on inflammatory markers and the primary outcome was unchanged.Microbiota diversity and circulating inflammatory markers · Limited certainty · Randomized intervention
  • What’s uncertain: Trials are small and measure microbiome or inflammatory markers; these secondary findings do not establish a clinical benefit.
  • First step: Start with one fermented food you will actually eat often — plain live yoghurt or kefir at breakfast is the easiest — rather than assembling a wide range of products you use once.
  • Before you try it: Food-safety risk is the most concrete harm here. All precautions
Read the full takeaway

What the intervention changed

This is a habit where the mechanism is better evidenced than the outcome. A 17-week randomised trial in 36 healthy adults found that a high-fermented-food diet steadily increased microbiota diversity and decreased inflammatory markers, while the high-fibre arm did not, which is a genuine randomised result rather than a marketing claim.

Limits of the trial evidence

Two things keep the certainty low. First, the trial's pre-specified primary outcome, a cytokine response score, was unchanged, so the headline result rests on secondary measures. Second, microbiota diversity and inflammatory markers are not themselves health outcomes, and no randomised trial has tested fermented foods against disease.

Other evidence and conclusion

The cohort evidence is thin and inconsistent: in 34,409 Dutch adults followed for 15 years, total fermented food intake was not associated with all-cause, cancer or cardiovascular mortality at all, and a systematic review of 52 cohorts found only modest inverse associations for fermented milk, chocolate and fermented soy that its authors judged insufficient to support a health claim. A meta-analysis of nine randomised trials found no benefit of probiotic yoghurt over conventional yoghurt for glucose control. Regularly eating live yoghurt, kefir and fermented vegetables is a reasonable, low-risk dietary change, but it should be chosen because you enjoy it and it displaces worse options, not because the outcome evidence justifies it.

Sauerkraut served with salmonDiscussed by2 peopleView statements ↓
NutritionLong-term healthUnder 10 min · Easy · Moderate
Reviewed Sep 23, 2026
Save

Scientific evidence

What the research shows

Outcome assessments based on research, including its uncertainties and limitations.

Microbiota diversity and circulating inflammatory markers

Finding: Mixed evidenceCertainty: Limitedstudies disagree, or the effect depends on who is studied

Evidence basis: randomised trials

A high-fermented-food diet increased microbiota diversity over 17 weeks, but the two randomised trials disagree on inflammatory markers and the primary outcome was unchanged.

Population: 36 healthy adults in a 17-week randomised study (18 per arm) in the Stanford study; 147 healthy adults in an 8-week randomised citizen-science trial with 21-week follow-up; and 472 adults with type 2 diabetes or obesity across nine randomised controlled trials in the probiotic yoghurt meta-analysis.

Research details

Summary

In a 17-week randomised prospective study with 18 adults per arm, a diet high in fermented foods steadily increased microbiota diversity and decreased inflammatory markers, whereas a high-fibre diet increased microbiome-encoded glycan-degrading carbohydrate active enzymes without changing microbial community diversity. The second randomised trial complicates the inflammatory half of this outcome: its fermented-food arm significantly increased IL-18R1, an inflammatory-signalling marker, alongside increases in T-cell activation markers (CD5, CD6, CD8A) and SIRT2, so the direction of the inflammatory-marker effect is not consistent between the two trials; the same trial also reported a modest shift in the gut microbiota of participants over 50 years toward a composition characteristic of younger participants. The cytokine response score, which was the primary outcome, was unchanged, and three distinct immunological trajectories among high-fibre consumers corresponded to their baseline microbiota diversity. An independent randomised controlled trial in 147 healthy adults found that microbial diversity increased significantly within the high-fermented-food group and the control group, with the increase in high-fermented-food participants over 50 reaching p=0.04, but the high-fermented-food group showed no difference in microbial diversity compared with the control group, while the high-fibre group showed a significant decrease. The most concrete human finding in this literature is negative: a meta-analysis of nine randomised controlled trials with 237 participants receiving probiotic yoghurt and 235 receiving mostly conventional yoghurt found no significant difference in HbA1c in type 2 diabetes (USMD -0.366; 95% CI -0.755 to 0.024, p=0.066) or obesity, and no effects on fasting blood glucose, fasting insulin or insulin resistance.

Why this certainty

Rated Limited. The evidence has genuine randomised design, which is more than most gut-health claims in this catalogue have, and the direction is consistent with the mechanistic expectation that live microbes and fermentation products reach the colon. Certainty is held at Limited because the positive trial is very small, its primary outcome was null, the larger independent trial did not replicate the diversity difference against its control, and none of the measured endpoints is a clinical outcome, so the size of any resulting health benefit cannot be estimated from these data. The outcome is recorded as Mixed evidence rather than Benefit supported because the larger independent trial not only failed to separate fermented food from control on diversity but moved one inflammatory-signalling marker in the opposite direction to the first trial.

Limitations

The randomised evidence is small: 18 participants per arm in the trial that produced the positive result, which is enough to detect large shifts in microbiome composition but not to estimate clinical effects. Microbiota diversity and inflammatory markers are intermediate measures, not health outcomes, and the trial that reported them also found its pre-specified primary outcome unchanged, so the positive finding is a secondary endpoint. The larger independent trial failed to separate the fermented-food group from the control group on diversity, which weakens the replication, and its duration was 8 weeks against the 17 weeks of the earlier trial. Diet composition changed alongside the intervention — the high-fermented-food arm in the Stanford trial reached several servings a day and also increased overall diet quality — so the effects cannot be attributed to fermentation alone rather than to the broader dietary pattern. In the intervention that came closest to a clinical question, probiotic yoghurt did not beat conventional yoghurt for any glycaemic marker, which suggests that adding specific live cultures to an already fermented food does not add measurable benefit. The two randomised trials also disagree on the direction of the inflammatory-marker change, so this outcome combines a positive microbiome result with a conflicting immune-marker result rather than a single consistent effect.

Sources for this outcome:

Incident type 2 diabetes and death from any cause

Finding: Mixed evidenceCertainty: Very limitedstudies disagree, or the effect depends on who is studied

Evidence basis: observational studies (association, not proof of cause)

Total fermented food intake was not associated with mortality in 34,409 adults followed for 15 years, though yoghurt intake was associated with about 14 percent lower type 2 diabetes risk.

Population: 34,409 Dutch adults in the EPIC-Netherlands cohort followed a mean of 15 years with 2,436 deaths; 13 prospective studies in the yoghurt and diabetes review; and 52 prospective cohort studies in the systematic review of fermented food and mortality.

Research details

Summary

In the European Prospective Investigation into Cancer and Nutrition-Netherlands cohort, 34,409 men and women aged 20 to 70 and free of cardiovascular disease or cancer at baseline were followed for a mean of 15 years, during which 2,436 deaths occurred. After adjustment for age, sex, total energy intake, physical activity, education level, hypertension, smoking habit, body mass index and intakes of fruit, vegetables and alcohol, total fermented food intake was not associated with mortality from all causes (hazard ratio upper versus lowest quartile 1.00; 95% CI 0.88 to 1.13), cancer (1.02; 95% CI 0.86 to 1.21) or cardiovascular disease (1.04; 95% CI 0.83 to 1.30), and bacterial fermented foods in that cohort were mainly fermented dairy (78 percent) and cheese (16 percent). For type 2 diabetes specifically, a review of 13 prospective studies found that most showed an inverse association between yoghurt consumption and diabetes risk, with the most recent meta-analysis reporting a 14 percent lower risk at 80-125 g/day compared with no yoghurt consumption. A systematic review of 52 cohort studies concluded that fermented milk products including yoghurt, chocolate and fermented soy products, particularly natto, suggested modest inverse associations with all-cause and cardiovascular mortality, that cheese was inconsistent for cardiovascular mortality, that the cancer mortality evidence was weaker, and that the current evidence remains insufficient to support EFSA-approved health claims.

Why this certainty

Rated Very limited. The single largest retrieved cohort found no association between total fermented food intake and mortality of any kind, so the observational evidence does not converge, and the positive yoghurt signal is confined to type 2 diabetes and to a meta-analysis reported second-hand in a narrative review rather than retrieved directly as a pooled estimate with its own confidence interval. Certainty would be higher if the cohorts agreed on direction or if any randomised trial had measured a clinical endpoint, but no such trial was retrieved.

Limitations

The largest single cohort found no association at all between total fermented food intake and any mortality outcome, which is the most direct test available and is null. In the same cohort, cheese was the one subtype moderately inversely associated with cardiovascular and particularly stroke mortality (HR 0.59, 95% CI 0.38 to 0.92), so 'null' applies to total fermented food intake rather than to every subtype. That cohort's fermented food intake was dominated by fermented dairy and cheese, so it does not describe kimchi, sauerkraut, miso or tempeh, and the retrieved evidence for fermented vegetables and fermented meat is limited to single cohorts. Yoghurt and diabetes findings are observational, and people who eat yoghurt regularly differ in income, education and overall diet in ways adjustment cannot remove. The systematic review of 52 cohorts found the associations modest and inconsistent by food and outcome, and its authors stated the evidence is insufficient to support a health claim; the FDA's 2024 qualified health claim for yoghurt and type 2 diabetes risk reduction was itself framed as resting on limited scientific evidence (editorial context: this regulatory statement has no retrieved source record in this entry and is not attributed to any stored source).

Effect estimate

All-cause mortality hazard ratio 1.00 (95% CI 0.88 to 1.13) comparing the highest with the lowest quartile of total fermented food intake over a mean 15 years in 34,409 adults.

Sources for this outcome:

What people say

Who discusses this?

Public statements, with positions specific to each source and topic. These views do not determine the scientific evidence assessment.

What the research shows ↑

Practical steps and precautions

How to try it

  1. Start with one fermented food you will actually eat often — plain live yoghurt or kefir at breakfast is the easiest — rather than assembling a wide range of products you use once. In the randomised trial, the benefit came from sustained daily intake over 17 weeks, not from variety alone.
  2. Increase the number of servings per day rather than the number of different products, since the trial that showed an effect asked participants to eat several servings daily and measured adherence by intake rather than by product count.
  3. Check the label for live cultures and, for yoghurt, for unsweetened plain rather than flavoured products; sweetened flavoured yoghurt carries added sugar that works against the rest of the diet.
  4. Treat fermented foods as a swap, not an addition. Replacing a sweetened snack or a processed meat portion with plain yoghurt or fermented vegetables is what makes the change coherent with the rest of the catalogue.

Precautions

  • Food-safety risk is the most concrete harm here. Unpasteurised fermented products and home-fermented vegetables can carry harmful bacteria, and people who are pregnant, immunocompromised, very young or elderly should follow national food-safety advice about unpasteurised dairy and raw fermented products rather than general gut-health guidance.
  • Fermented foods are frequently high in salt — sauerkraut, kimchi, miso, soy sauce and many cheeses especially — which may work against the existing sodium-reduction entry; check the salt content if you eat them daily.
  • Live yoghurt and kefir are dairy, and fermented soy products are soy; people with dairy allergy or soy allergy should not treat fermentation as making these foods suitable.
  • If you have histamine intolerance or a mast cell condition, aged and fermented foods are a common trigger for symptoms, and increasing them is more likely to cause problems than benefits.

References

Gut-microbiota-targeted diets modulate human immune status.

Wastyk HC, Fragiadakis GK, Perelman D, Dahan D, Merrill BD, Yu FB, Topf M, Gonzalez CG, Van Treuren W, Han S, Robinson JL, Elias JE, Sonnenburg ED, Gardner CD, Sonnenburg JL · Cell · 2021

Randomized controlled dietary trial with longitudinal microbiome and immune profilingVerified from full text
Open original source

Distinct modulatory effects of high-fiber and fermented-food diets on gut microbiota, immune function, transit time, and sleep quality in a citizen science randomized controlled trial

van den Belt M, van de Put M, Yüksel E, Lee D, van Eeden F, van Straaten JP, Pintar N, Balen B, Andelic M, Rijnaarts I, Besseling-van der Vaart I, Rook M, Molenaar D, Levin E, de Vos WM, de Wit NJ, Kort R · medRxiv · 2025

Randomized placebo-controlled dietary trial (preprint)Verified from abstract
Open original source

The Effect of Probiotic Yogurt on Glycemic Control in Type 2 Diabetes or Obesity: A Meta-Analysis of Nine Randomized Controlled Trials.

Barengolts E, Smith ED, Reutrakul S, Tonucci L, Anothaisintawee T · Nutrients · 2019

Systematic review and meta-analysis of randomized controlled trialsVerified from full text
Open original source

The relationship between fermented food intake and mortality risk in the European Prospective Investigation into Cancer and Nutrition-Netherlands cohort.

Praagman J, Dalmeijer GW, van der Schouw YT, Soedamah-Muthu SS, Monique Verschuren WM, Bas Bueno-de-Mesquita H, Geleijnse JM, Beulens JW · The British journal of nutrition · 2015

Prospective cohort studyVerified from abstract
Open original source

Yogurt and Diabetes: Overview of Recent Observational Studies.

Salas-Salvadó J, Guasch-Ferré M, Díaz-López A, Babio N · The Journal of nutrition · 2017

Narrative review of prospective observational studiesVerified from abstract
Open original source

A systematic review of prospective evidence linking non-alcoholic fermented food consumption with lower mortality risk.

Paveljšek D, Pertziger E, Fardet A, Panagiotakos DB, Savary-Auzeloux I, Adamberg S, Peñas E, Frias J, Ntantou A, Diamantoglou I, Domínguez-Soberanes J, Louis S, Chassard C, Praćer S, Vergères G, Matalas A · Frontiers in nutrition · 2025

Systematic review of prospective cohort studies with narrative evidence gradingVerified from full text
Open original source

Community

Discussion

Have you tried this? What helped or made it difficult?

Reader experiences are separate from the scientific evidence. Explore all discussions

No discussions yet.

Be the first to ask a question or share your experience.

Sign in to vote or join the discussion.

Related habits

Explore other habits in a shared category. Each has its own evidence and precautions.

Editorial history and review method

Revision 3 · Approved for publication · Codex (publication approval explicitly authorized by site owner; category migration)

Prepared and checked with AI. Publication approval does not imply independent expert review. Request a correction.

Applied the audit corrections for fermented-foods. (1) The default card outcome microbiome-and-inflammatory-markers moves from 'Benefit supported' to 'Mixed evidence': the entry previously reported only the inflammatory half of its evidence, and the second randomised trial (van den Belt 2025) states that its fermented-food arm significantly increased IL-18R1, an inflammatory-signalling marker, alongside T-cell activation markers and SIRT2. That conflict is now in the summary, limitations and certainty rationale, and the trial's positive findings (the immune-marker increases and the shift toward a younger microbiota composition in participants over 50) are now recorded alongside the negative high-fibre result. (2) The type-2-diabetes-and-mortality limitations now carry the EPIC-Netherlands subtype result that cuts against the 'no association with any mortality outcome' framing: cheese was moderately inversely associated with cardiovascular and particularly stroke mortality (HR 0.59, 95% CI 0.38 to 0.92), which is also stored as a new verified fact on praagman-2015. (3) The truncated van-den-belt-2025 stored fact has been repaired to the full retrieved sentence. (4) The FDA 2024 qualified-claim sentence is retained but explicitly marked as editorial context rather than as a sourced finding, because the FDA constituent update has no DOI and cannot be retrieved through scripts/epmc-search.mjs. Second reviewer must re-check: (a) that Mixed evidence is not read as a downgrade of the trial's microbiome result, which is still positive; (b) that the van den Belt trial is still labelled a medRxiv preprint downstream, since its numbers may change on peer review; (c) whether the FDA sentence should instead be deleted; (d) that the Wastyk 2021 cytokine response score null is still present in all three narrative fields. Publication release 2026-10-04: site owner authorized a bounded AI audit and release of this exact revision. See docs/release-audit-2026-10-04.md for source-check findings and limits. No independent health-professional or peer review is claimed. Category migration 2026-10-04: removed the separate goals field; categories: Nutrition, Long-term health → Nutrition, Long-term health. Research assessments, sources, guidance, precautions, and evidence review dates are unchanged. This revision awaits publication approval; no new evidence review is claimed. Publication approval 2026-10-04: site owner explicitly accepted all current category-migration revisions for deployment and requested that remaining limitations be retained in docs/content-review-backlog.md. Approval applies to this exact content revision. No new evidence review, independent expert review, or clinical validation is claimed; existing precautions and disclosures remain in effect.