Soy sauce: almost everything tested was added to it

Most of what is known about soy sauce and health is about something put into it or pulled out of it. In China iron was added as NaFeEDTA. Pooling 16 randomised trials, the odds of anaemia fell to a quarter of the odds on unfortified sauce, an odds ratio of 0.25 (95% CI 0.19 to 0.35), and haemoglobin rose by 8.81 g/L. A second meta-analysis of the national programme from 2004 to 2013 put the anaemia rate among people using the fortified sauce at 27 percent of the rate among those who did not. In an 18-month trial in Bijie City, 14,000 people poured 16.4 ml a day, which carried 4.9 mg of iron. The bottle on a supermarket shelf outside that programme carries no such iron. The other human trials here used Shoyu polysaccharides, 600 mg a day of a fraction isolated from raw sauce and given as a supplement. Nobody reaches that by seasoning food. What the sauce itself brings to the table is sodium, and WHO names it among the condiments where sodium hides.

The numbers, per 100 g

What your body absorbs

• Fortifying soy sauce with NaFeEDTA iron did not significantly change how much zinc 60 children absorbed, with P = 0.5940 for zinc absorption between the three groups. — 60 children in China randomised to NaFeEDTA-fortified, FeSO4-fortified or unfortified soy sauce, all given 3 mg of zinc-67 and 1.2 mg of dysprosium daily (n = 60) Evidence B source
  Fractional zinc absorption P = 0.3895, dysprosium recovery P = 0.7498, zinc absorption P = 0.5940

What it does to your health

• Pooling 16 randomised trials in China, soy sauce fortified with NaFeEDTA cut the odds of anaemia to a quarter of the odds on unfortified soy sauce, an odds ratio of 0.25 (95% CI 0.19 to 0.35), and raised haemoglobin by 8.81 g/L. — Chinese populations at risk of iron deficiency, 16 randomised trials for anaemia rates and 12 for haemoglobin (n = 16 trials) Evidence A source
  Pooled odds ratio 0.25 (95% CI 0.19 to 0.35) for anaemia, weighted mean difference 8.81 g/L (95% CI 5.96 to 11.67) for haemoglobin
• A second meta-analysis, covering the Chinese programme from 2004 to 2013, put the anaemia rate among people using iron fortified soy sauce at 27% of the rate among those who did not use it. — Chinese population covered by the iron fortified soy sauce programme, more than 186 million people over ten years Evidence A source
  Anaemia rate 27% of the rate in non-consumers, an estimated 23.2% of the national decline attributed to the programme
• In an 18-month trial among 14,000 residents of Bijie City, the fortified group poured 16.4 ml of soy sauce a day, which carried 4.9 mg of iron, and the effect on haemoglobin became significant after six months. — 14,000 residents aged three years and older in Bijie City, Guizhou Province, China, 18 months, soy sauce at 29.6 mg iron per 100 ml (n = 14000) Evidence B source
  16.4 ml a day giving 4.9 mg iron, higher haemoglobin and ferritin and lower anaemia prevalence in all age and sex subgroups
• 600 mg a day of Shoyu polysaccharides, a fraction isolated from raw soy sauce, raised serum iron, haematocrit and haemoglobin in 45 healthy women over 4 weeks. — 45 healthy women, 22 on 600 mg of Shoyu polysaccharides and 23 on placebo, 4 weeks (n = 45) Evidence B source
  Serum iron, haematocrit and haemoglobin significantly higher than placebo at P < 0.05
• In 51 patients with mild seasonal allergic rhinitis, 600 mg a day of Shoyu polysaccharides improved sneezing, nasal stuffiness and daily life scores against placebo after 4 weeks at P < 0.05. — 51 patients with mild seasonal allergic rhinitis, 25 on 600 mg of Shoyu polysaccharides and 26 on placebo, 8 weeks (n = 51) Evidence B source
  Individual symptom scores differed at P < 0.05 from week 4, total symptom score from weeks 4 to 8
• In 30 men with triglycerides above 150 mg/dl, 600 mg a day of Shoyu polysaccharides for 4 weeks lowered triglycerides against their own baseline at P < 0.05, with no placebo contrast reported for that comparison. — 30 hyperlipidaemic men, 15 on 600 mg of Shoyu polysaccharides and 15 on placebo, 4 weeks (n = 30) Evidence B source
  Serum triglycerides lower than baseline at P < 0.05 in the treated group
• Sniffing water or a rose-like odour for 20 seconds changed a seated reach test in neither of two groups of 24 healthy adults, at P = 0.24 and P = 0.42, and the same trial reports an improvement only in the soy sauce group. — 72 healthy adults randomised in equal groups of 24 to sniff water, phenylethyl alcohol or Japanese dark soy sauce for 20 seconds (n = 72) Evidence B source
  No change with water (P = 0.24) or with phenylethyl alcohol (P = 0.42), improvement only with soy sauce

Safety, limits and interactions

• Drinking soy sauce in quantity can kill, and the documented case reached a serum sodium of 187 mmol/L with brain damage within hours. — One 55-year-old woman with depression who drank a large quantity of Japanese soy sauce (n = 1) Evidence C source
  Serum sodium 187 mmol/L, described as the highest ever documented, death from massive pulmonary oedema
• WHO advises adults to stay under 2000 mg of sodium a day and names soy sauce among the condiments where sodium hides. — Adults worldwide, WHO sodium reduction fact sheet Evidence E source
  Less than 2000 mg of sodium a day, equivalent to less than 5 g of salt

The bottom line

Treat soy sauce as salt with flavour, and season with it the way you would season with salt. WHO advises adults to stay under 2000 mg of sodium a day, about 5 g of salt, and puts soy sauce on its short list of condiments people forget to count. Our own USDA table holds a single record for this entity, a reduced sodium sauce made from hydrolysed vegetable protein, so we cannot tell you here what an ordinary spoonful of brewed sauce carries. Read the ingredient list on the bottle, and the sodium figure with it. Quantity is where this food turns dangerous, and the case on record is grim. A 55-year-old woman drank a large amount of Japanese soy sauce and reached a serum sodium of 187 mmol/L, the highest ever documented, with brain damage within hours and death from massive pulmonary oedema. Soy sauce is not a cleanse, a dare or a home remedy. That is one case, so it says what can happen rather than how often it happens. One worry about the fortified sauce was tested and came back quiet. Fortifying soy sauce with NaFeEDTA iron did not significantly change how much zinc 60 children in China absorbed, with P = 0.5940 between the three groups. Iron and zinc compete, so this was worth checking. With 20 children in each arm, though, a null of this size rules out only a large interference. Then there is the supplement fraction, and here the honest word is unsettled. Six hundred milligrams a day of Shoyu polysaccharides raised serum iron, haematocrit and haemoglobin in 45 healthy women over 4 weeks. In 51 people with mild seasonal allergic rhinitis the same dose improved sneezing, stuffiness and daily life scores against placebo from week 4. In 30 men with triglycerides above 150 mg/dl it lowered triglycerides against their own baseline, and for that comparison no placebo contrast was reported, which is the weakest way to show an effect. These are small groups, four to eight weeks long, on an isolated fraction that comes in a capsule. Until someone repeats them on a larger sample and against placebo throughout, there is nothing here to act on. The oddest entry is a single experiment in which 72 adults sniffed water, a rose-like odour or dark soy sauce for 20 seconds, and only the soy sauce improved a seated reach test, while water and the rose odour changed nothing at P = 0.24 and P = 0.42. One small study, one lab, and nobody has repeated it. What we could not tell you matters as much. We hold no card on soy allergy, none on the wheat in brewed sauce and coeliac disease, none on tyramine and the drugs that interact with it, and none on blood pressure measured in people who cut soy sauce specifically. We also have nothing on how brewing differs from chemical hydrolysis in the finished bottle. Those questions are open on this page because the evidence we trust has not answered them, and we would rather say so than fill the space.

Sources

1. soys-r4-04 · randomised trial
   There were no significant differences in fractional Zn absorption (FZA) (P = 0.3895), dysprosium recovery (P = 0.7498) and Zn absorption (P = 0.5940) among the three groups.
   Food Funct, 2015 · checked 2026-09-23
2. soys-r4-01 · meta-analysis of RCTs
   For anemia rates, data from 16 studies could be pooled, and the pooled estimate odds ratio was 0.25 (95% CI 0.19-0.35).
   Biomed Environ Sci, 2015 · checked 2026-09-23
3. soys-r4-02 · meta-analysis
   By Meta analysis, the anemia rate of the population consumed IFSS was 27% compared with the anemia rate of the population non-consumed IFSS.
   Wei Sheng Yan Jiu, 2017 · checked 2026-09-23
4. soys-r4-03 · randomised trial
   The average daily soy sauce consumption of the group consuming the fortified product was 16.4 ml per person, which provided 4.9 mg of iron from NaFeEDTA.
   Food Nutr Bull, 2005 · checked 2026-09-23
5. soys-r4-05 · randomised trial
   After the 4 weeks, serum levels of iron, hematocrit, and hemoglobin were significantly higher (P < 0.05) in the SPS-treated than in the placebo-treated group.
   Int J Mol Med, 2006 · checked 2026-09-23
6. soys-r4-06 · randomised trial
   After 4 weeks of treatment with SPS, symptom scores such as sneezing, nasal stuffiness, and hindrance of daily life were significantly different (P<0.05) from those in placebo-treated groups.
   Int J Mol Med, 2005 · checked 2026-09-23
7. soys-r4-07 · randomised trial
   After 4 weeks, serum TG levels in the SPS-treated group were significantly (P<0.05) lower than the baseline (0 week).
   Int J Mol Med, 2008 · checked 2026-09-23
8. soys-r4-08 · randomised trial
   Sniffing water or PEA had no effect on the mFRT score (P = 0.24 for water, P = 0.42 for PEA, n = 24, Wilcoxon signed-rank test; Figure 1).
   Occup Ther Int, 2019 · checked 2026-09-23
9. soys-r4-09 · case report
   She presented with the highest ever documented serum sodium level of 187 mmol/L.
   J Forensic Leg Med, 2011 · checked 2026-09-23
10. soys-r4-10 · fact sheet
    For adults, WHO recommends less than 2000 mg/day of sodium (equivalent to less than 5 g/day salt)
    WHO, Sodium reduction fact sheet · checked 2026-09-23