Crab: strong on zinc and B12, and never tested in a randomised trial

No randomised trial has ever fed crab to people and followed what happened. The one study that put crab on a plate, in 1990, was not randomised and ran in 18 men. The two randomised trials behind this page fed something else. One gave 796 children aged 0 to 6 with food allergies six months of a compound probiotic with zinc and B vitamins, and the prevalence of several allergies including crab fell, though the thing tested was the supplement and the abstract reports no effect size for crab. The other fed 400 g of cod a week to 137 pregnant Norwegian women for 16 weeks. Median hair mercury ended at 554 µg/kg against 485 µg/kg in controls, a gap that did not reach significance at p = 0.186, in a species that was not crab. What crab is measurably good at is minerals. Crab is also a shellfish, and boiling does not destroy its two major allergens. A cooked 3-ounce portion of blue crab carries 3.2 mg of zinc, 29 percent of the Daily Value, taken from an NIH fact sheet table rather than from a measurement of the animal in front of you. Vitamin B12 is wildly inconsistent, running 0.57 to 11.5 µg per 100 g across 12 USDA records, and the UK table gives 12.9 µg for cooked crab where the USDA reference record gives 3.116 µg, because two national tables measured different animals by different methods.

The numbers, per 100 g

How much is actually in it

• Crab composition swings by species and by cut. Vitamin B12 runs 0.57 to 11.5 µg per 100 g across 12 USDA records, and UK CoFID gives 12.9 µg for cooked crab against 3.116 µg in the USDA reference record. — — (n = 12 USDA records, 8 Foundation samples, 5 CoFID records) Evidence E source
  Vitamin B12 0.57-11.5 µg per 100 g, copper 0.032-1.182 mg, selenium 22.3-47.6 µg across USDA records

What a real portion looks like

• A cooked 3-ounce portion of blue crab carries 3.2 mg of zinc, which is 29% of the Daily Value. — — Evidence E source
  3.2 mg zinc per 3 ounces of cooked blue crab, 29% DV

What cooking and storage change

• Boiling cut the number of crab protein bands, yet the 36 kDa tropomyosin and the 41 kDa arginine kinase stayed the major allergens for sera from 50 crab-allergic patients. — sera from 50 patients with crab allergy, raw and boiled extracts of Charybdis feriatus (n = 50) Evidence D source
  Boiled extract produced fewer protein bands while the 36 kDa and 41 kDa allergens were still recognized
• Heating costs crab little of its vitamin B12. USDA retention factors for the finfish and shellfish group keep 95% of vitamin B12 and 100% of copper when the frozen product is heated or fried, and selenium is absent from Release 6 altogether. — — (n = 2 retention codes) Evidence E source
  Vitamin B12 95% retained, copper 100% retained, selenium not covered by the table

What it does to your health

• In 18 normolipidemic men fed six shellfish species in turn, the crab diet lowered VLDL triglycerides and cholesterol together with LDL and total cholesterol, while squid and shrimp diets left blood lipids unchanged. — 18 normolipidemic men, each shellfish fed so that its protein equalled that of animal foods in the normal diet (n = 18) Evidence C source
  Lower VLDL triglycerides and cholesterol, lower LDL and total cholesterol, sizes and intervals not reported in the abstract

Safety, limits and interactions

• Spoiled cooked crab held 47.08 mg of histamine per 100 g and caused fatal histamine poisoning in a 37-year-old woman who ate it. — a 37-year-old woman and her 14-year-old son who ate cooked crabs (Scylla serrata) (n = 2) Evidence C source
  Histamine 47.08 mg/100 g in the cooked crab and 22.54 mg/100 g in intestinal content, one death
• Pooled across 109 data reports, crab muscle held 4.490 mg/kg wet weight of nickel and 1.891 mg/kg of lead, and the review placed adults and children in many countries at risk of non-cancer health effects from these elements. — 108 papers with 109 data reports on crab muscle worldwide (n = 109 data reports) Evidence D source
  Pooled means Ni 4.490, Pb 1.891, As 1.601, Cd 1.101 mg/kg wet weight
• In mouse models, oral immunotherapy with slightly processed crab allergen lowered the anaphylactic score by 0.63 points (p < 0.01). — mouse models of crustacean allergy, 5 case-control studies (n = 5 mouse studies) Evidence D source
  Mean difference -0.63 for crab species (p < 0.01, I2 = 0)

What people believe that the data does not support

• Self-reported shellfish allergy ran from 0% to 10.3% across 61 population studies, while food challenge put it at 0% to 0.9%. — population-based cross-sectional and cohort samples worldwide, all ages (n = 61 studies out of 7,333 screened articles) Evidence C source
  0% to 10.3% by any method of diagnosis versus 0% to 0.9% by food challenge

Who this works differently for

• FDA and EPA list crab among Best Choices seafood and advise 2 to 3 servings a week, 4 ounces per serving, for people who are pregnant or breastfeeding. — people who are pregnant or breastfeeding and children ages 1 to 11 years Evidence E source
  2 to 3 servings a week from the Best Choices list, 1 serving is 4 ounces for pregnancy and breastfeeding
• In 796 children aged 0 to 6 years, 6 months of compound probiotics with zinc and B vitamins lowered the prevalence of crab allergy compared with the control group. — 796 children aged 0-6 years with food allergies, 6-month intervention and 6-month follow-up (n = 796) Evidence B source
  Lower prevalence of specific food allergies including crab after treatment and after follow-up, effect sizes not given in the abstract
• Adding 400 g a week of a low-mercury seafood species for 16 weeks of pregnancy left median hair mercury at 554 µg/kg against 485 µg/kg in controls (p = 0.186), under the US EPA reference dose of 1,000 µg/kg. — 137 pregnant Norwegian women, gestational weeks 20 to 36, randomised to 400 g of cod fillets per week or habitual diet (n = 137) Evidence B source
  Median hair mercury 554 (392-805) µg/kg in the intervention group versus 485 (341-740) µg/kg in controls, p = 0.186

What is still unknown

• For lead carried by seafood eaten in pregnancy or lactation, a 2025 systematic review found 4 articles from 3 cohorts and concluded that evidence to link it with child neurodevelopment was not available. — healthy pregnant or lactating people in high human development index countries, children aged 12 to 60 months (n = 4 articles from 3 prospective cohort studies) Evidence C source
  No usable estimate, certainty of evidence rated very low by GRADE

The bottom line

The one study that actually put crab on a plate is from 1990. Eighteen normolipidemic men ate six shellfish species in turn, each species fed so that its protein matched the animal foods in their usual diet. The crab diet lowered VLDL triglycerides and cholesterol along with LDL and total cholesterol, while the squid and shrimp diets left blood lipids unchanged. The abstract gives no effect sizes and no intervals, and the men were not randomised, so this is a direction without a magnitude. How common shellfish allergy is depends on how it was measured. Across 61 population studies, it ran from 0 to 10.3 percent by whatever method each study used, and from 0 to 0.9 percent where a food challenge was done, and most of the included studies relied on questionnaires. That gap is between methods of diagnosis across populations, not a verdict on any one person, and a food challenge is done in a clinic. Real crab allergy does not cook away. Boiling produced fewer protein bands in extracts of Charybdis feriatus, yet sera from 50 crab-allergic patients still recognised the 36 kDa tropomyosin and the 41 kDa arginine kinase, which is IgE binding in a dish and not a challenge in a clinic. The oral immunotherapy that lowered anaphylactic scores by 0.63 points was done in mice. The hazard that has actually killed someone here is storage. A 37-year-old woman died of histamine poisoning after eating cooked crab that held 47.08 mg of histamine per 100 g, one forensic case with no denominator and no dose-response, and histamine builds up in mishandled crab rather than in fresh crab. Pooled across 109 data reports on crab muscle worldwide, the meat held 4.490 mg/kg wet weight of nickel and 1.891 mg/kg of lead, and the review placed adults and children in many countries at risk of non-cancer effects from these elements. Those are composition measurements pooled from very different waters, so a local crab can sit far from either mean. If you are pregnant or breastfeeding, FDA and EPA put crab on the Best Choices list and advise 2 to 3 servings a week at 4 ounces a serving, an agency position built on mercury by species that says nothing about the other elements. On the lead in seafood, a 2025 systematic review found 4 articles from 3 cohorts and concluded that evidence linking it in pregnancy to child neurodevelopment was not available, at a certainty GRADE rated very low. Cooking costs the B12 little, 95 percent retained and copper 100 percent, though those factors were set for the whole finfish and shellfish group and selenium is missing from that table altogether.

Sources

1. crab-r5-12 · USDA FoodData Central and CoFID 2021, dataset
   Crustaceans, crab, blue swimming, lump, pasteurized, refrigerated (FDC 2684446, Foundation Foods, Finfish and Shellfish Products): vitamin B12 3.116 ug per 100 g = 130% DV, copper 1.088 mg = 121% DV, selenium 45.79 ug = 83% DV. Across 12 USDA records vitamin B12 0.57-11.5 ug, copper 0.032-1.182 mg, selenium 22.3-47.6 ug. Foundation samples n=8: vitamin B12 2.44-3.56 ug, copper 0.964-1.38 mg, selenium 28.6-52.7 ug. CoFID 2021 Crab, purchased cooked [16-441]: vitamin B12 12.9 ug, copper 1.72 mg, selenium 156 ug.
   USDA FoodData Central, Foundation Foods, FDC 2684446, release 2026-04-30, and CoFID 2021 (Public Health England, OGL v3.0) · checked 2026-09-24
2. crab-r5-11 · fact sheet
   Blue crab, cooked, 3 ounces 3.2 29
   NIH Office of Dietary Supplements, Zinc fact sheet for health professionals · checked 2026-09-24
3. crab-r5-05 · in vitro allergen characterisation
   At least 20 protein bands between 13 to 250 kDa were detected in the SDS-PAGE gel of raw extract, while boiled extract procuced fewer protein bands. Proteins of 36 kDa and 41 kDa were recognized as the major allergens of the crab.
   Asian Pac J Allergy Immunol, 2012 · checked 2026-09-24
4. crab-r5-13 · USDA Table of Nutrient Retention Factors, dataset
   USDA-RETN-R6:2605 FINFISH(FROZEN),HEATED, group Finfish and Shellfish Products: vitamin B12 95% retained, copper 100% retained. USDA-RETN-R6:2601 FINFISH(FROZEN),FRIED: vitamin B12 95% retained, copper 100% retained. Selenium, vitamin K, vitamin E, vitamin D, manganese and iodine are not in Release 6 at all.
   USDA Table of Nutrient Retention Factors, Release 6, 2007, Retn_Codes 2601 and 2605 · checked 2026-09-24
5. crab-r5-07 · controlled feeding trial, non-randomised
   Oyster, clam, crab, and mussel diets, low in cholesterol and high in n-3 fatty acids, lowered VLDL triglycerides and cholesterol and, except for the mussel diet, LDL and total cholesterol.
   Am J Clin Nutr, 1990 · checked 2026-09-24
6. crab-r5-01 · case report
   Toxicological analysis revealed that histamine concentrations were very high in the cooked crab (47.08 mg/100 g) and intestinal content (22.54 mg/100 g).
   J Forensic Sci, 2018 · checked 2026-09-24
7. crab-r5-02 · meta-analysis of food composition measurements
   The rank order of PTEs based on pooled (mean) concentration in the muscle of crabs was Ni (4.490 mg/kg-ww) > Pb (1.891 mg/kg-ww) >As (1.601 mg/kg-ww) > Cd (1.101 mg/kg-ww).
   Int J Environ Health Res, 2024 · checked 2026-09-24
8. crab-r5-06 · meta-analysis of animal studies
   Subgroup analyses with low-level heterogeneities provided more reliable results for crab species (MD = -0.63, p < 0.01, I2 = 0), arginine kinase allergen (MD = -0.83, p < 0.01, I2 = 0), and Maillard reaction processing method (MD = -0.65, p < 0.01, I2 = 29%), respectively.
   Nutrients, 2024 · checked 2026-09-24
9. crab-r5-04 · systematic review of observational studies
   The prevalence of fish allergy ranged from 0% to 7% and the prevalence of shellfish allergy from 0% to 10.3%, depending on the method of diagnosis. Where food challenges were used, the prevalence for fish allergy was found to be 0% to 0.3% and for shellfish allergy was 0% to 0.9%.
   Ann Allergy Asthma Immunol, 2016 · checked 2026-09-24
10. crab-r5-03 · federal consumer advice
    Eat 2 to 3 servings a week from the "Best Choices" list (OR 1 serving from the "Good Choices" list).
    FDA and EPA, Advice about Eating Fish, updated 2024-03-05 · checked 2026-09-24
11. crab-r5-08 · randomised controlled trial
    Furthermore, the prevalence of specific food allergies, including egg, milk, crab, shrimp, cashew, beef, mango, and shellfish, was significantly lower in the experimental group posttreatment and post-follow-up, highlighting the therapeutic potential of the intervention.
    Allergol Immunopathol (Madr), 2026 · checked 2026-09-24
12. crab-r5-09 · randomised controlled trial
    Post-intervention, the intervention group had median (inter-quartile range) THHg concentrations of 554 (392-805) µg/kg, and the control group 485 (341-740) µg/kg (p = 0.186).
    Environ Int, 2020 · checked 2026-09-24
13. crab-r5-10 · systematic review of observational studies
    Overall, evidence was not available to address a direct association between Pb exposure from seafood intake during PL and child neurodevelopment.
    Adv Nutr, 2025 · checked 2026-09-24