Kimchi: the effects are real and small, and the salt question is open

Kimchi is one of the few fermented foods with a meta-analysis of human trials behind it. Pooling five intervention trials and 205 adults, all of them Korean, fermented kimchi lowered fasting blood glucose by 1.93 mg/dL. The confidence interval ran from 3.82 lower to 0.03 lower, so the result is statistically significant and far too small to move anyone out of a diagnostic category. The same review reported triglycerides 28.9 mg/dL lower and systolic pressure 3.48 mmHg lower only after it dropped the trials that were driving its heterogeneity, which makes those the most favourable version of the pool rather than its main result. Fermentation itself appears to matter. Twenty-two overweight and obese adults ate fresh kimchi for four weeks and fermented kimchi for four weeks. The fermented phase came out ahead on blood pressure, percent body fat, fasting glucose and total cholesterol, though the abstract gives the direction without the size of each gap. Against all of that sits the salt. The single USDA record for kimchi is a survey entry with no measured sample spread. It puts sodium at 498 mg per 100 g, which works out to 747 mg in a 150 g cup. That is 37 percent of the WHO ceiling of 2000 mg a day in one side dish. The salt is not the only thing on the other side of the ledger: a meta-analysis of mostly case-control Korean studies put kimchi at 2.21 times the odds of gastric cancer (95% CI 1.29 to 3.77) and salt itself at 1.92, and an umbrella review of 16 meta-analyses graded that pickled-vegetable evidence as very low quality, which is the ceiling on how firmly the number can be read in either direction.

How much is actually in it

• Kimchi has no USDA Foundation record, only a survey entry, so the single set of numbers on offer carries no measured sample spread, and one cup of it holds 65 ug of vitamin K. — USDA FNDDS survey record Kimchi, food code 75502520 (n = 1 record) Evidence E source
  Vitamin K (phylloquinone) 43.6 ug per 100 g, 65.4 ug per 1 cup (150 g), no Foundation samples and no CoFID counterpart

What a real portion looks like

• A cup of kimchi weighs 150 g in the USDA survey database and carries 747 mg of sodium, which is 37% of the WHO daily ceiling in one side dish. — USDA FNDDS survey record Kimchi, food code 75502520 (n = 1 record) Evidence E source
  Sodium 498 mg per 100 g, 747 mg per 1 cup (150 g), 149 mg per the unspecified 30 g portion

What cooking and storage change

• In 22 overweight and obese adults who ate fresh and then fermented kimchi for 4 weeks each, the fermented version beat the fresh one on blood pressure, percent body fat, fasting glucose and total cholesterol, so ripeness changes what the same vegetable does. — 22 overweight and obese adults with BMI above 25 kg/m2, crossover, two 4-week phases with a 2-week washout (n = 22) Evidence B source
  Net differences in systolic and diastolic blood pressure, percent body fat, fasting glucose and total cholesterol significantly greater after fermented than after fresh kimchi
• A trial that gave 24 obese women fresh or fermented kimchi for eight weeks found the links between diet, gut bacteria and blood gene expression clearer in the fermented arm, which is the mechanistic side of the same fermentation difference. — 24 obese Korean women, 8 weeks of kimchi intervention (n = 24) Evidence B source
  Correlations between clinical change, gut microbial population and blood gene expression more evident with fermented than with fresh kimchi

What it does to your health

• Pooling five human intervention trials, fermented kimchi lowered fasting blood glucose by 1.93 mg/dL (95% CI 3.82 to 0.03 lower), an effect small enough that it would not move a person out of any diagnostic category. — Adults in 5 intervention trials, 205 participants in total (n = 205) Evidence A source
  Fasting blood glucose WMD -1.93 mg/dL (95% CI -3.82, -0.03), I2 17.4%
• After the same meta-analysis dropped the trials driving its heterogeneity, fermented kimchi was tied to triglycerides 28.9 mg/dL lower, systolic pressure 3.48 mmHg lower and diastolic pressure 2.68 mmHg lower, all from a pool of 5 trials and 205 people. — Adults in 5 intervention trials, 205 participants, plus 4 prospective cohorts with 42 455 people reviewed separately (n = 205 in the pooled trials) Evidence A source
  Triglycerides WMD -28.9 mg/dL (95% CI -53.2, -4.5), systolic BP -3.48 mmHg (95% CI -5.95, -1.01), diastolic BP -2.68 mmHg (95% CI -4.75, -0.62)
• In 100 young adults fed 15 g or 210 g of kimchi a day for 7 days in a dormitory, fasting blood glucose fell further on the high intake (P=.003), which is the only dose comparison of kimchi in the corpus. — 100 healthy young adult volunteers, 50 per group, housed together for 7 days (n = 100) Evidence B source
  Fasting blood glucose lower on 210 g/day than on 15 g/day (P=.003), total cholesterol and LDL-C fell in both groups
• A kimchi-derived Lactobacillus sakei capsule cut body fat mass by 0.8 kg against placebo in 114 Korean adults with obesity over 12 weeks. BMI and body weight did not move, and a capsule of one strain is not a serving of kimchi. — 114 Korean adults with obesity, BMI 25 kg/m2 or more, 12 weeks, randomised double-blind placebo-controlled (n = 114) Evidence B source
  Body fat mass -0.2 kg on the strain versus +0.6 kg on placebo, a 0.8 kg difference (P=0.018), waist circumference 0.8 cm smaller (P=0.013), BMI and weight unchanged
• In 61 761 Korean adults followed about five years, eating 2 to under 3 servings of cabbage kimchi a day tracked with total cholesterol 1.6 mg/dL lower and triglycerides 3.4 mg/dL lower in women, changes too small to feel. — 61 761 adults aged 40 to 69 in the HEXA arm of the Korean Genome and Epidemiology Study, mean 4.97 years between surveys (n = 61761) Evidence C source
  Total cholesterol beta -1.600 (95% CI -2.744, -0.456), triglycerides beta -3.372 (95% CI -5.414, -1.330), LDL-C beta -1.155 (95% CI -2.214, -0.095) in women at 2 to under 3 servings per day
• Across 115 726 Korean adults, men eating 1 to 3 servings of kimchi a day had lower odds of obesity than men eating under 1 serving (OR 0.875 and 0.893), and these are odds rather than risks. — 115 726 adults aged 40 to 69 in the Health Examinees study, 36 756 men and 78 970 women (n = 115726) Evidence C source
  Obesity OR 0.875 at 1 to 2 servings/day and OR 0.893 at 2 to 3 servings/day in men versus under 1 serving/day, about 10% lower odds for cabbage kimchi

Safety, limits and interactions

• The same 115 726-person analysis ends with a warning rather than an endorsement, because every association it found was J-shaped and the authors advise limiting excessive kimchi intake. — 115 726 Korean adults aged 40 to 69 (n = 115726) Evidence C source
  J-shaped association across all outcomes, with the favourable odds confined to 1 to 3 servings per day
• A meta-analysis of Korean studies put kimchi at 2.21 times the odds of gastric cancer (95% CI 1.29 to 3.77) and salt at 1.92, the strongest single reason not to treat kimchi as a food without a downside. — Korean population studies pooled by outcome, mostly case-control Evidence C source
  Gastric cancer OR 2.21 (95% CI 1.29 to 3.77) for kimchi and OR 1.92 (95% CI 1.52 to 2.43) for salt, odds rather than risk
• Across 60 studies of pickled vegetables the pooled odds of gastric cancer were 1.52 (95% CI 1.37 to 1.68), and the Korean subgroup was the highest at 1.89, so the signal is not unique to kimchi and not absent from it either. — 60 studies, 50 case-control and 10 prospective, users of pickled vegetables versus non-users or lowest category (n = 60 studies) Evidence C source
  Gastric cancer OR 1.52 (95% CI 1.37 to 1.68) overall, 1.32 (1.10 to 1.59) in cohorts, 1.89 (1.29 to 2.77) in Korean studies, 1.14 (0.96 to 1.35) elsewhere
• In a Korean workplace outbreak, 77 people met the case definition and kimchi eaters had 1.47 times the attack rate of non-eaters (95% CI 1.08 to 2.00), including 2 asymptomatic food handlers positive for enterotoxigenic E. coli O159. — Workers at a Korean workplace canteen, 77 cases including 2 asymptomatic food handlers, 27 laboratory-confirmed (n = 77 cases) Evidence C source
  Relative risk of illness 1.47 (95% CI 1.08 to 2.00) among kimchi consumers, incubation median 72 hours
• WHO tells adults to stay under 2000 mg of sodium a day, and one cup of kimchi carries 747 mg of it, so three cups spend the whole day's allowance. — Adults, general population recommendation Evidence E source
  Ceiling of less than 2000 mg sodium per day, equal to less than 5 g of salt
• The FDA-approved warfarin label asks patients to keep the amount of vitamin K in food steady, and a cup of kimchi brings 65 ug of it, so the point for warfarin users is consistency from week to week. — Patients taking warfarin sodium Evidence E source
  Vitamin K competitively antagonises warfarin, and the label instructs a consistent dietary amount rather than avoidance

Who this works differently for

• Among 21 people with prediabetes, 33.3% improved their glucose tolerance on fermented kimchi against 9.5% on fresh kimchi over 8 weeks each, the only kimchi trial run in a group that is already metabolically off track. — 21 adults with prediabetes, crossover, 8 weeks on each type with a 4-week washout (n = 21) Evidence B source
  Improved glucose tolerance in 33.3% on fermented kimchi versus 9.5% on fresh kimchi, with insulin resistance down and insulin sensitivity up only on the fermented type
• The full text limits that 0.8 kg to adults aged 20 to 65 with a BMI of 25 kg/m2 or more. Compliance meant taking 80% to 120% of the allocated doses. — 114 men and women aged 20 to 65 with BMI 25 kg/m2 or more, screened from 131 candidates (n = 114) Evidence B source
  Inclusion limited to BMI 25 kg/m2 or more, compliance window 80% to 120% of allocated doses

What is still unknown

• An umbrella review of 16 meta-analyses graded the pickled-vegetable link with gastric cancer as very low quality evidence, which is the honest ceiling on both the scary and the reassuring readings of it. — 16 meta-analyses of prospective cohort studies covering 57 diet and gastric cancer associations (n = 16 meta-analyses) Evidence C source
  Pickled vegetables and citrus fruit graded very low quality, only alcohol reached high quality evidence (RR 1.19, 95% CI 1.06 to 1.34)

The bottom line

How much you eat, and how sour it is, matter more than whether you eat it at all. One hundred healthy young adults spent a week in a dormitory on identical meals apart from the kimchi, 15 g a day for half of them and 210 g a day for the rest. Fasting glucose fell further on the larger amount (P=.003), over seven days, in young people with nothing wrong with them. In 21 adults with prediabetes, glucose tolerance improved in 33.3 percent during eight weeks of fermented kimchi against 9.5 percent during eight weeks of fresh kimchi, percentages that rest on seven people and two people. A Lactobacillus sakei strain isolated from kimchi was given as a capsule for 12 weeks to 114 Korean adults aged 20 to 65 with a BMI of 25 kg/m2 or more. Body fat mass ended 0.8 kg lower than on placebo (P=0.018) and waist circumference 0.8 cm smaller (P=0.013), while BMI and body weight did not change. A capsule of one strain is not a bowl of kimchi. Observational data link kimchi with modest numbers in both directions. In 61 761 Korean adults resurveyed after about five years, women eating 2 to under 3 servings of cabbage kimchi a day had total cholesterol 1.6 mg/dL lower and triglycerides 3.4 mg/dL lower than women eating under 1 serving. Intake there was self-reported on a questionnaire, and differences of that size are invisible to the person carrying them. Across 115 726 adults measured at a single visit, men eating 1 to 3 servings a day had lower odds of obesity than men eating under 1 serving (OR 0.875 and 0.893), and odds measured on one day are not risks. The authors of that analysis close by advising people to limit excessive kimchi intake, because every association they found was J-shaped and the favourable odds sat only between 1 and 3 servings a day. The cancer question deserves the same plain treatment. A meta-analysis of mostly case-control studies in Korea put kimchi at 2.21 times the odds of gastric cancer (95% CI 1.29 to 3.77) and salt itself at 1.92, and asking sick people to recall how much of a daily staple they used to eat is a known weak point of that design. A wider pool of 60 studies of pickled vegetables found odds of 1.52 overall, 1.89 in Korean studies and 1.14 elsewhere, where the interval crossed 1 and the signal faded. An umbrella review of 16 meta-analyses graded the pickled-vegetable association as very low quality evidence, and that grade is the honest ceiling on both the alarming and the reassuring reading of these numbers. Two smaller cautions are concrete. In one Korean workplace outbreak, 77 people met the case definition and kimchi eaters fell ill 1.47 times as often as non-eaters (95% CI 1.08 to 2.00), which is a reminder that an unheated fermented vegetable is an uncooked food. The FDA-approved warfarin label asks for a consistent amount of vitamin K in the diet rather than for avoidance, and a cup of kimchi carries 65 ug of it by the same single survey record, so steadiness from week to week is the point for anyone on that drug. We found no trial and no cohort on kimchi in pregnancy, in children or in kidney disease, and given a sodium load of 747 mg per cup that is the emptiest part of this page.

Sources

1. kimc-r2-18 · USDA FNDDS 2021-2023, dataset
   Kimchi (FDC 2710077, FNDDS, food code 75502520): vitamin K (phylloquinone) 43.6 ug per 100 g; 1 cup = 150 g = 65.4 ug; Foundation samples 0; variability_search no record; CoFID match none
   USDA FoodData Central, FNDDS, FDC 2710077 · checked 2026-09-24
2. kimc-r2-17 · USDA FNDDS 2021-2023, dataset
   Kimchi (FDC 2710077, FNDDS, food code 75502520, WWEIA category Olives, pickles, pickled vegetables): sodium 498 mg per 100 g; portions 1 cup = 150 g = 747 mg sodium, quantity not specified = 30 g = 149.4 mg sodium
   USDA FoodData Central, FNDDS, FDC 2710077 · checked 2026-09-24
3. kimc-r2-04 · randomised crossover feeding trial
   Twenty-two overweight and obese patients with body mass indexes greater than 25 kg/m(2) were randomly assigned to two 4-week diet phases separated by a 2-week washout period (crossover design)...Net differences in the systolic blood pressure, diastolic blood pressure, percent body fat, fasting glucose, and total cholesterol in the fermented kimchi group were significantly greater than those in the fresh kimchi group.
   Nutr Res, 2011 · checked 2026-09-24
4. kimc-r2-05 · randomised controlled trial, pyrosequencing and microarray
   Twenty-four obese women were randomly assigned to either fresh or fermented kimchi group for eight weeks of kimchi intervention...Correlations of these effects with changes in blood gene expression and gut microbial population were more evident in the fermented kimchi group than the fresh kimchi group.
   Mol Nutr Food Res, 2015 · checked 2026-09-24
5. kimc-r2-01 · systematic review and meta-analysis of intervention trials
   A meta-analysis of the intervention studies revealed a significant reduction in fasting blood glucose (WMD: -1.93 mg/dL; 95% CI: -3.82, -0.03; I2 = 17.4%) following the consumption of fermented kimchi.
   Nutr Rev, 2025 · checked 2026-09-24
6. kimc-r2-02 · systematic review and meta-analysis of intervention trials
   Five intervention studies (205 participants) and 4 prospective cohort studies (42 455 participants) were selected...After excluding studies that contributed to increased heterogeneity, significant inverse associations were observed between fermented kimchi consumption and triglycerides (WMD: -28.9 mg/dL; 95% CI: -53.2, -4.5; I2 = 0.0%), systolic blood pressure (WMD: -3.48 mmHg; 95% CI: -5.95, -1.01, I2 = 0.0%), and diastolic blood pressure (WMD: -2.68 mmHg; 95% CI: -4.75, -0.62; I2 = 0.0%).
   Nutr Rev, 2025 · checked 2026-09-24
7. kimc-r2-03 · randomised clinical trial, controlled diet
   For the study, 100 volunteers were assigned to 2 dietary groups, low (15 g/day, n=50) and high (210 g/day, n=50) kimchi intake, and were housed together in a dormitory for 7 days...FBG was significantly decreased in the high kimchi intake as compared to the low intake group (P=.003).
   J Med Food, 2013 · checked 2026-09-24
8. kimc-r2-07 · randomised controlled trial, probiotic capsule
   Body fat mass decreased by 0.2 kg in the CJLS03 group and increased by 0.6 kg in the placebo group (0.8 kg difference, P=0.018)...BMI and body weight did not change after the 12 weeks.
   Endocrinol Metab (Seoul), 2020 · checked 2026-09-24
9. kimc-r2-09 · cohort study, HEXA/KoGES
   We conducted an epidemiological study on participants (aged 40-69 years old) selected from the Health Examinees (HEXA) cohort study (n = 61,761)...In this study, compared to the lowest category (< 1 serving/day), Baechu kimchi intake (2- < 3 servings/day) had more negative correlations with the change in values of total cholesterol (β: -1.600, 95% confidence interval [CI, -2.744, -0.456]), triglycerides (β: -3.372, 95% CI [-5.414, -1.330]), and low-density lipoprotein cholesterol (β: -1.155, 95% CI [-2.214, -0.095]) in women.
   Eur J Nutr, 2024 · checked 2026-09-24
10. kimc-r2-10 · cross-sectional study, food frequency questionnaire
    In men, total kimchi consumption of 1-3 servings/day was related to a lower prevalence of obesity (OR: 0.875 in 1-2 servings/day and OR: 0.893 in 2-3 servings/day) compared with total kimchi consumption of <1 serving/day.
    BMJ Open, 2024 · checked 2026-09-24
11. kimc-r2-11 · cross-sectional study, food frequency questionnaire
    However, since all results showed a 'J-shaped' association, it is recommended to limit excessive kimchi intake.
    BMJ Open, 2024 · checked 2026-09-24
12. kimc-r2-12 · meta-analysis of observational studies
    Salt and Kimchi were associated with an increased gastric cancer risk [OR (95% CI): 1.92 (1.52-2.43) and 2.21 (1.29-3.77), respectively].
    Asian Pac J Cancer Prev, 2014 · checked 2026-09-24
13. kimc-r2-13 · systematic review and meta-analysis of observational studies
    The OR (95% CI) was 1.52 (1.37-1.68) for the overall association, 1.56 (1.39-1.75) for case-control, and 1.32 (1.10-1.59) for cohort studies...The OR (95% CI) was 1.89 (1.29-2.77) in Korean, 1.86 (1.61-2.15) in Chinese, and 1.16 (1.04-1.29) in Japanese studies, and 1.14 (0.96-1.35) in studies from other countries.
    Cancer Epidemiol Biomarkers Prev, 2012 · checked 2026-09-24
14. kimc-r2-15 · outbreak investigation, retrospective cohort
    A total of 77 individuals met the case definition, including 75 symptomatic cases and 2 asymptomatic food handlers who tested positive for enterotoxigenic Escherichia coli (ETEC) O159...The attack rate was significantly higher among kimchi consumers than among non-consumers (RR, 1.47; 95% CI, 1.08 to 2.00).
    Epidemiol Health, 2026 · checked 2026-09-24
15. kimc-r2-16 · fact sheet of an international agency
    For adults, WHO recommends less than 2000 mg/day of sodium (equivalent to less than 5 g/day salt), or just under a teaspoon.
    WHO fact sheet, Salt reduction, who.int · checked 2026-09-24
16. kimc-r2-19 · drug label, manufacturer text approved by FDA
    The amount of vitamin K in food may affect therapy with warfarin sodium. Advise patients taking warfarin sodium to eat a normal, balanced diet maintaining a consistent amount of vitamin K.
    Warfarin Sodium, FDA-approved label via DailyMed, effective 2026-06-25 · checked 2026-09-24
17. kimc-r2-06 · randomised crossover feeding trial in prediabetes
    A total of 21 participants with prediabetes were enrolled...The percentages of participants who showed improved glucose tolerance were 9.5 and 33.3% in the fresh and fermented kimchi groups, respectively.
    Ann Nutr Metab, 2013 · checked 2026-09-24
18. kimc-r2-08 · randomised controlled trial, methods section
    The study included 114 men and women aged 20 to 65 years with a BMI ≥25 kg/m2 who understood the content of this study and agreed to participate in this clinical trial...Good compliance with the treatment was defined as taking 80% to 120% of the allocation.
    Endocrinol Metab (Seoul), 2020 · checked 2026-09-24
19. kimc-r2-14 · umbrella review of observational evidence
    Three associations (total fruit, vitamin E, and carotenoids) were determined to be supported by low-quality evidence, and two (pickled vegetables/foods and citrus fruit) were supported by very low-quality.
    J Cancer Res Clin Oncol, 2022 · checked 2026-09-24