Cheonggukjang benefits: three trials exist — but the two recent ones have no placebo arm

Key benefits

1. 01Early research

   All three groups improved — because all three groups ate cheonggukjang

   2025, Nutrients — 60 postmenopausal women, eight weeks.

   • Three groups: high-microorganism (HTC), low-microorganism (LTC), commercial (CC) cheonggukjang
   • 3.3 g of cheonggukjang tablets daily
   • Kupperman index, metabolic parameters, stool microbiome analysis

   Outcome	Result
   Kupperman index (menopausal symptoms)	significantly decreased in all three groups, greatest in HTC
   BMI, weight	no significant change
   Lipid profile	no significant change
   Fasting glucose	decreased significantly only in HTC
   Beneficial bacteria	increased in HTC and CC
   Firmicutes/Bacteroidetes ratio	decreased in HTC and CC, not significant

   Look at the design. All three arms were cheonggukjang. The comparison was not cheonggukjang versus none but cheonggukjang A versus B versus C.

   That limits what "symptoms improved in all three groups" can mean. Menopausal symptom scores are a questionnaire, and taking something diligently for eight weeks while recording symptoms commonly lowers such scores. Some improvement happens simply with time. A placebo arm is what separates those out, and this trial has none.

   The between-group difference survives. That HTC fell most is a comparison among the three products, so it holds. But that says a cheonggukjang richer in microorganisms beat a poorer one, not cheonggukjang beat not eating it.

   The 2023 Foods trial has the same structure. All three groups (19, 20, 20) took cheonggukjang pills, and the abstract concludes:

   No changes in obesity and inflammation-related indicators were observed before and after Cheonggukjang consumption. The Firmicutes/Bacteroidetes ratio decreased in all three groups, but no statistical significance was indicated.

   What that trial actually achieved lies elsewhere. It confirmed that the biogenic amines present in cheonggukjang did not produce adverse effects in participants. It is closer to a safety study than an efficacy one.

   2011, Journal of Medicinal Food — this one has a control.

   • 45 people with impaired fasting glucose, three groups: control (starch 2 g/day), cheonggukjang 20 g/day, red ginseng cheonggukjang 20 g/day, eight weeks
   • Total cholesterol fell about 30 mg/dL in the cheonggukjang group
   • LDL cholesterol fell 29.66%
   • The apolipoprotein B to A-1 ratio fell from 3.0 to 2.5
   • Fasting glucose was lower in both active groups

   But look at what these were compared against. The abstract repeatedly says "compared to the initial value". Not compared to the control group. A within-group before-and-after comparison and a between-group comparison are different claims.

   One note on the abstract's units. The fall in total cholesterol is printed as "30.0 mg/mL", which would be 3,000 mg/dL — impossible in a person. It appears to be a misprint for mg/dL.

   The honest conclusion:

   • Three Korean trials on a Korean food is rare and welcome in itself
   • But the two recent ones have no placebo arm. All three groups ate cheonggukjang
   • The 2023 trial found no change at all in obesity or inflammation markers
   • The 2011 improvements are measured against baseline, not against the control
   • The grade here is "early research"

   [1][2][3]
2. 02Nutrition data

   19.4 g of protein, 217 mg of calcium, 8.6 mg of iron — but this table is natto's

   First, what this table is. The figures below are the USDA entry for natto. Cheonggukjang is hard to find as a separate entry in the public composition databases.

   The two are close. Both are whole cooked soybeans fermented with Bacillus subtilis, unlike doenjang or miso, where the beans are crushed and held in brine for a long time.

   They are not the same, though. Cheonggukjang ferments for two or three days and is usually salted and simmered into a stew; natto is eaten as it is with soy sauce. Read the numbers as indicative, not exact.

   Per 100 g of natto:

   • 211 kcal, protein 19.4 g, fat 11.0 g, carbohydrate 12.7 g, fibre 5.4 g
   • Calcium 217 mg, iron 8.6 mg, potassium 729 mg, magnesium 115 mg
   • Zinc 3.03 mg, manganese 1.53 mg, copper 0.667 mg, selenium 8.8 µg
   • Vitamin K 23.1 µg, vitamin C 13.0 mg
   • Sodium 7 mg

   Per tablespoon (15 g):

   	100 g	A tablespoon (15 g)
   Energy	211 kcal	32 kcal
   Protein	19.4 g	2.9 g
   Calcium	217 mg	33 mg
   Iron	8.6 mg	1.3 mg

   The 19.4 g of protein is the centre of this food, high among fermented soy products because the whole bean is used.

   Iron at 8.6 mg also stands out, a rare figure for a plant food. But plant iron (non-haem) is absorbed less readily than animal iron. Vitamin C helps, and this table happens to carry 13 mg of it.

   Calcium at 217 mg is above milk's 123 mg per 100 g.

   Do not take the 7 mg of sodium at face value. That is unseasoned natto. Cheonggukjang stew is another food entirely — salt, soy sauce and doenjang go in, sometimes kimchi or salted seafood. What you actually eat is decided by the cooking.

   One qualification on the 23.1 µg of vitamin K. That figure counts phylloquinone (vitamin K1) only. The menaquinones (vitamin K2) that the fermenting bacteria produce are not in that number. That soybean foods fermented with B. subtilis contain vitamin K2 is known; how much cannot be read from this table.

   There is a case where this genuinely matters — warfarin. The next section continues.

3. 03Not established

   The trials fed tablets, not stew — and a word about warfarin

   This entry is about actually eating it.

   First, the form. Look again at what the trials gave:

   • 2025 — 3.3 g of cheonggukjang tablets a day
   • 2023 — cheonggukjang pills
   • 2011 — cheonggukjang 20 g/day

   3.3 g a day is less than a quarter of a tablespoon. That is a different scale from what goes into a bowl of stew. When carrying these results across to the stew, carry the difference in form and amount with them.

   And stew comes with salt. Cheonggukjang itself is low in sodium, but simmering it into a stew adds salt, soy sauce and doenjang. If you eat it often as a health food, the thing to count is not the paste but the broth. This site's low-sodium doenjang stew recipe makes the same point.

   Now warfarin.

   Soybean foods fermented with Bacillus subtilis contain vitamin K2 (menaquinone). Vitamin K is involved in blood clotting, and warfarin works by blocking that action. A sudden rise in vitamin K intake can blunt the drug.

   In Japan it is standard to advise patients on warfarin to avoid natto. Cheonggukjang is fermented with the same organism, so if you take warfarin, speak to your doctor before eating it. This site does not hold data on how much vitamin K2 cheonggukjang contains, and not knowing an amount is not the same as it being safe.

   One common claim, handled plainly.

   The enzyme the bacteria produce, nattokinase, is widely said to dissolve blood clots. This site has not published a trial feeding cheonggukjang or natto and measuring clot-related clinical outcomes. That an enzyme breaks down fibrin in a test tube and that eating the food reduces clots in people are different claims. It will be added as its own entry once the evidence is checked.

   With a soy allergy, cheonggukjang counts.

   The smell is what fermentation does. Ammonia-family compounds cause it, and it is not the smell of spoilage. But prolonged boiling kills the bacteria. If you want them alive, take the pan off the heat and stir it in at the end — though, as above, what that buys you is not yet clear.

How to eat it well

What to watch out for