Sushi: raw fish, freezing and salt

Almost everything known about sushi is about the raw fish in it, and most of the studies on this page measured raw fish in general rather than sushi itself. We found no trial that followed people eating sushi to see what it does to their heart, weight or long term health, so this page makes no health claim for it in either direction. What the evidence does cover is safety and salt. Parasites are handled by freezing. FDA guidance, written for seafood processors and retailers, says fish meant to be eaten raw needs -20°C for 7 days, or -35°C until solid and then 15 hours at that temperature. In a blind tasting in Japan, frozen mackerel sushi won 49.2 percent of pairs and fresh won 42.5 percent, a gap too small to count. Bacteria are a separate problem. Pooling 14,496 samples of fish in general, with sushi not counted separately, Listeria turned up in 5.8 percent of raw fish and 14.5 percent of ready-to-eat fish products. Finding the bacterium in a sample is not the same as someone falling ill from it. And one piece of sushi weighs about 30 g and carries about 128 mg of sodium before any soy sauce, so a six piece serving comes to roughly 770 mg.

How much is actually in it

• In a randomised dinner study in 9 adults with normal thyroid function, one sushi meal raised urinary iodine by about 40 µg/g, or about 400 µg/g with a fresh seaweed salad, followed by a brief rise in TSH and no adverse events. — 9 euthyroid adults in Greenland, one halibut maki meal with or without seaweed salad, followed for a week (n = 9) Evidence B source
  Urinary iodine up about 40 µg/g after sushi, about 400 µg/g with seaweed salad; brief rise in serum TSH; no adverse events
• Per 100 g, British vegetable sushi in CoFID has 521 mg of sodium and 3.4 g of protein, while salmon nigiri has 360 mg of sodium and 8.7 g of protein at almost the same 148-152 kcal. — UK composition table, three sushi types Evidence E source
  Sodium 521 (vegetable), 360 (salmon nigiri), 331 mg (tuna nigiri) per 100 g; energy 148-170 kcal

What a real portion looks like

• One piece of sushi weighs about 30 g in FNDDS and carries about 128 mg of sodium before any soy sauce, so a 6-piece serving gives roughly 770 mg. — US household portions, FNDDS Evidence E source
  1 piece 30 g, 128.4 mg sodium; 1 cup 150 g, 642 mg sodium

What your body absorbs

• In a randomised stable-isotope study in 35 men, selenium from fish was absorbed as well as selenate and retained better, and cooking did not change selenium absorption or retention from the fish. — 35 men aged 18-50 in Norwich and Zeist, labelled trout processed two ways, four meals over two days (n = 35) Evidence B source
  Fish selenium absorption similar to selenate, retention higher (P<0.001); no difference between processing methods

What cooking and storage change

• FDA guidance says fish meant to be eaten raw is made safe from parasites by freezing at -20°C (-4°F) or below for 7 days, or at -35°C (-31°F) until solid and then storing for 15 hours at that temperature. — US seafood processors and retailers Evidence E source
  -20°C for 7 days, or -35°C until solid plus 15 hours at -35°C, or plus 24 hours at -20°C

Safety, limits and interactions

• Pooling 14,496 samples, a meta-analysis found Listeria monocytogenes in 5.8% of raw fish and 14.5% of ready-to-eat fish products, in fish in general with sushi not counted separately. — published surveys of fish, fish products and processing environments, 14,496 samples (n = 14496) Evidence C source
  L. monocytogenes 5.8% (95% CI 4.7-6.9) in raw fish, 14.5% (11.1-18.0) in ready-to-eat products
• Japan recorded 2,009 cases of Kudoa septempunctata food poisoning from 2013 to 2023, and raw flounder served as sashimi or sushi was implicated in 99% of them. — national foodborne illness statistics, Japan, 2013-2023 (n = 2009) Evidence C source
  2,009 cases, peak 429 in 2014; flounder implicated in 99%

What people believe that the data does not support

• In a randomised double-blind tasting trial in Japan covering 240 pieces of mackerel sushi, frozen fish was judged better in 49.2% of pairs and fresh in 42.5%, so freezing to kill parasites did not spoil the taste. — Japanese adults in 120 paired discrimination rounds each for mackerel and squid sushi (n = 120 rounds) Evidence B source
  Mackerel: odds ratio 0.86 (95% CI 0.59-1.26) for preferring unfrozen; squid: 1.38 (0.93-2.05), neither significant

Who this works differently for

• A systematic review of Anisakis found the highest rates of allergic anisakiasis in Portugal and Norway, at 18.45-22.50% among the people examined in those studies, and sensitisation there is not the same as illness. — people examined in published studies of allergic anisakiasis, Portugal and Norway Evidence C source
  Allergic anisakiasis 18.45-22.50% among those examined in Portugal and Norway, the highest rate in the review
• In a survey of 81 pregnant Australian women, 96.0% of those aware of listeriosis advice cut down or avoided raw fish, one of the high-risk listeriosis foods, compared with 69.2% of those unaware. — 81 pregnant women at antenatal clinics, 10-23 weeks gestation (n = 81) Evidence C source
  Raw fish reduced or avoided by 96.0% of aware versus 69.2% of unaware women (p=0.046)
• Among pregnant women in Shanghai eating 66 g of fish and shellfish a day, only 1.4% exceeded the methylmercury reference dose while 81.3% fell short of 250 mg of EPA plus DHA, in local species that do not tell us about tuna for sushi. — pregnant women in the Shanghai Diet and Health Survey 2016-2017, 59 local fish species analysed Evidence C source
  1.4% above MeHg reference dose of 0.1 µg/kg bw/day; 81.3% below 250 mg EPA+DHA

What is still unknown

• A 2021 systematic review on warfarin listed sushi among foods linked to higher bleeding risk, next to cranberry and green tea, but the abstract gives no mechanism or amount for sushi. — published reports on foods, herbs and supplements taken with warfarin Evidence D source
  Sushi named among items with increased bleeding risk; no effect size

The bottom line

If you like sushi, the simplest change is in what you pick. In the British food composition table, vegetable sushi has 521 mg of sodium and 3.4 g of protein per 100 g, and salmon nigiri has 360 mg of sodium and 8.7 g of protein, at almost the same energy. Soy sauce comes on top of either. Seaweed brings iodine. In a randomised study of 9 adults in Greenland with normal thyroid function, one sushi meal raised iodine in the urine by about 40 µg/g, and by about 400 µg/g when a fresh seaweed salad came with it. TSH, the hormone that tells the thyroid to work, rose briefly, and the meal caused no adverse events. Nobody with thyroid disease was in that study, so we cannot tell you how it goes for them. Selenium from fish is well absorbed. In 35 men eating labelled trout, fish selenium was absorbed as well as the supplement form selenate and retained better, and cooking changed neither. That study used trout, and we have no such study on sushi. Some people need more care with raw fish. Raw fish is listed among the high-risk foods for listeriosis, an infection that is dangerous in pregnancy, and in a survey of 81 pregnant women in Australia, 96.0 percent of those who knew that advice cut down or avoided raw fish, against 69.2 percent of those who did not. Mercury is the other worry, and one Chinese survey found the bigger shortfall ran the other way. Pregnant women in Shanghai ate about 66 g of fish and shellfish a day. Only 1.4 percent went over the reference dose for methylmercury, while 81.3 percent fell short of 250 mg a day of the omega-3 fats EPA and DHA. Those were local Shanghai species. Tuna, common in sushi, was not measured in our sources, so this survey cannot tell you whether tuna sushi is low in mercury. If you are pregnant, the raw part and the fish type are the questions to take to your midwife or doctor. In one review, the highest rates of allergy to Anisakis, a worm found in fish, came from Portugal and Norway, at 18.45 to 22.50 percent of the people examined in those studies. That is not a share of everyone living there, and a positive allergy test is not the same as being ill. In Japan, 2,009 cases of food poisoning from the parasite Kudoa were recorded from 2013 to 2023, and raw flounder served as sashimi or sushi was behind 99 percent of them. One question is open. A review of reports on warfarin listed sushi among foods linked to a higher risk of bleeding, next to cranberry and green tea, and gave no mechanism and no amount. That is a flag without an explanation, and nobody in our sources has measured it. If you take warfarin, mention sushi to the person who manages your dose. We also have no USDA entry for sushi in this cohort, no data on the rice itself, and no study of the long term health of people who eat sushi.

Sources

1. sush-b4-01 · RCT
   Nine euthyroid participants were randomized into three groups ... A sushi meal increased urinary iodine excretion by 40 µg/g, or 400 µg/g if a newly harvested seaweed salad was added. An ensuing rise in serum TSH was brief, and a single sushi meal with seaweed salad did not cause any adverse events.
   J Endocrinol Invest, 2020 · checked 2026-09-23
2. sush-b4-11 · reference dataset
   CoFID 15-617 Sushi, vegetable · Energy 148 kcal · Protein 3.4 g · Carbohydrate 32.8 g · Sodium 521 mg · Iodine 20 µg || CoFID 16-361 Sushi, salmon nigiri · 152 kcal · 8.7 g · 25.2 g · 360 mg · 4 µg || CoFID 16-362 Sushi, tuna nigiri · 170 kcal · 8.4 g · 30.3 g · 331 mg · 4 µg
   CoFID 2021, Public Health England / OGL v3.0 · checked 2026-09-23
3. sush-b4-12 · reference dataset
   FNDDS 58151100 Sushi, NFS (fdc 2708959) · Sodium, Na 428 mg per 100 g · 1 piece 30 g 128.4 mg · 1 cup 150 g 642.0 mg
   USDA FNDDS, Sushi NFS, food code 58151100 · checked 2026-09-23
4. sush-b4-07 · RCT
   In all, 35 male volunteers aged 18-50 y were recruited ... Apparent absorption of selenium from fish was similar to selenate and there was no difference between the two processing methods used. However, retention of fish selenium was significantly higher than selenate (P<0.001). ... Cooking did not affect selenium apparent absorption or retention from fish.
   Eur J Clin Nutr, 2004 · checked 2026-09-23
5. sush-b4-03 · agency guidance
   Freezing and storing at an ambient temperature of -4°F (-20°C) or below for 7 days (total time), or freezing at an ambient temperature of -31°F (-35°C) or below until solid and storing at an ambient temperature of -31°F (-35°C) or below for 15 hours, or freezing at an ambient temperature of -31°F (-35°C) or below until solid and storing at an ambient temperature of -4°F (-20°C) or below for 24 hours are sufficient to kill parasites.
   FDA, Fish and Fishery Products Hazards and Controls Guidance, Chapter 5: Parasites · checked 2026-09-23
6. sush-b4-04 · meta-analysis of prevalence studies
   Data from a total of 14,496 samples were analysed. Pooled prevalence of L. monocytogenes and Listeria sp. in raw fish was 5.8 % (95 % CI: 4.7-6.9) and 12.2 % (95 % CI: 8.9-15.5) respectively. Significantly higher results were observed for ready-to-eat products, occurrence of L. monocytogenes and Listeria sp. was 14.5 % (95 % CI: 11.1-18.0)
   Sci Total Environ, 2024 · checked 2026-09-23
7. sush-b4-05 · retrospective cohort
   Reported cases totaled 2009, reaching a peak in 2014 (429 cases) then declining to < 100 cases during the COVID-19 pandemic. ... Flounder, particularly sashimi and sushi, were implicated in 99% of cases.
   Sci Rep, 2026 · checked 2026-09-23
8. sush-b4-02 · RCT
   Among 120 rounds of discrimination testing involving the consumption of 240 pieces of mackerel sushi, unfrozen sushi was believed to taste better in 42.5% (51 dishes) of cases, frozen sushi was thought to taste better in 49.2% (59 dishes) ... Freezing raw fish did not ruin sushi's taste.
   Clin Infect Dis, 2015 · checked 2026-09-23
9. sush-b4-06 · systematic review of observational studies
   The highest rate of allergic anisakiasis was in Portugal and Norway with the prevalence rate of 18.45-22.50%.
   Parasitol Res, 2020 · checked 2026-09-23
10. sush-b4-08 · cross-sectional
    The proportion of women who decreased or avoided consumption of certain high-risk Listeriosis foods were higher in those who were aware compared with those who were unaware of Listeriosis recommendations [raw fish (96.0 vs 69.2%, p=0.046)
    Women Birth, 2019 · checked 2026-09-23
11. sush-b4-09 · cross-sectional
    The average consumption of fish and shellfish was 66.24 g/d among pregnant women in Shanghai. ... Only 1.4% of the population exceeded the MeHg reference dose of 0.1 μg/kg·bw/d, whereas 81.3% of those who did not meet the recommended daily intakes of 250 mg EPA + DHA.
    Food Chem Toxicol, 2023 · checked 2026-09-23
12. sush-b4-10 · systematic review of case reports
    healthcare professionals should be aware of the increased risk of bleeding when taking several food and herbs. These include Chinese wolfberry, chamomile tea, cannabis, cranberry, chitosan, green tea, Ginkgo biloba, ginger, spinach, St. John's Wort, sushi and smoking tobacco.
    Br J Clin Pharmacol, 2021 · checked 2026-09-23