Broccoli sprouts: the sulforaphane is real, the results are modest

Broccoli sprouts are not small broccoli. They are eaten for sulforaphane, and every study on this page is about the sprouted seedling rather than a head of broccoli. A systematic review of 17 human trials published between 2003 and 2013, with broccoli sprouts the most common intervention, found the steadiest improvements in blood glucose, lipid profile and markers of oxidative stress, and weaker evidence for inflammation, Helicobacter pylori and cancer. Nothing in that review was pooled, so it ranks areas by how consistent the results were and gives no size for any single effect, and the authors call the human evidence limited. The portions that were measured are specific ones. The trial that fed 70 g of sprouts a day delivered 420 micromol of the sulforaphane precursor glucoraphanin with it, about 6 micromol per gram of fresh sprouts. Another trial blended about 111 g of fresh sprouts into a 200 g drink holding about 100 micromol of sulforaphane itself, close to 0.9 micromol per gram, which counts the converted compound and not the precursor. Both figures come from one batch measured for one trial, and content swings with cultivar, seed lot and sprouting age. How you eat them counts too. Chewing fresh sprouts thoroughly raised urinary dithiocarbamate output to 42.4 micromol against 28.8 micromol when the same sprouts were swallowed whole, P = 0.049, in a small crossover from 2001 that measured absorption and not any health outcome.

How much is actually in it

• The daily portion used in the Helicobacter pylori trial was 70 g of broccoli sprouts, and that portion carried 420 micromol of the sulforaphane precursor glucoraphanin. — 48 Helicobacter pylori positive patients fed 70 g of broccoli sprouts a day for 8 weeks (n = 48) Evidence B source
  70 g of sprouts supplying 420 micromol of precursor, about 6 micromol per gram of fresh sprouts
• A trial drink built from about 111 g of fresh broccoli sprouts, blended to roughly 200 g with water, delivered about 100 micromol of sulforaphane in one dose. — adult volunteers in an influenza vaccine trial, one drink a day for four consecutive days (n = 29) Evidence B source
  About 100 micromol of sulforaphane per dose from about 111 g of fresh sprouts, close to 0.9 micromol per gram

What your body absorbs

• Fresh broccoli sprouts put about 3 times more sulforaphane into plasma and urine than a myrosinase-treated sprout extract at the same daily dose of 200 micromol. — 20 healthy adults aged 19 to 50, randomised to fresh sprouts or extract, 200 micromol sulforaphane equivalents daily (n = 20) Evidence B source
  About 3-fold higher total sulforaphane metabolites in plasma and urine with sprouts, no confidence interval reported

What cooking and storage change

• Chewing fresh broccoli sprouts thoroughly raised urinary dithiocarbamate output to 42.4 micromol against 28.8 micromol when the same sprouts were swallowed whole. — healthy adult volunteers in a crossover feeding study, dose for this comparison not stated in the abstract Evidence C source
  42.4 +/- 7.5 against 28.8 +/- 2.6 micromol excreted, P = 0.049
• Isothiocyanates from broccoli sprouts are about 6 times more available to the body than the glucosinolates they come from, and boiling the sprouts destroys the plant enzyme that makes the conversion. — healthy adult volunteers given uncooked fresh sprouts or homogenates of boiled sprouts Evidence C source
  About 6-fold difference in bioavailability between isothiocyanates and glucosinolates

What it does to your health

• In type 2 diabetes, 10 g a day of broccoli sprout powder for 4 weeks lowered serum insulin and HOMA-IR, while 5 g a day did not. — 81 patients with type 2 diabetes randomised to 10 g/day powder (27), 5 g/day (29) or placebo (25) for 4 weeks, 63 analysed (n = 63) Evidence B source
  Significant fall in serum insulin and HOMA-IR at 10 g/day, P = 0.05 for the treatment effect, no confidence interval reported
• In 52 Japanese men with fatty liver, 2 months of broccoli sprout extract moved median ALT from 54.0 to 48.5 IU/L, and placebo moved nothing. — 52 Japanese men with fatty liver, 24 on broccoli sprout capsules containing glucoraphanin and 28 on placebo, 2 months (n = 52) Evidence B source
  Median ALT 54.0 (34.5 to 79.0) before against 48.5 (33.3 to 65.3) IU/L after, P < 0.05, placebo unchanged
• Eating broccoli sprouts for 8 weeks lowered markers of Helicobacter pylori colonisation in 48 infected patients, and the markers returned to their starting levels 2 months after the sprouts stopped. — 48 Helicobacter pylori positive patients randomised to 70 g/day broccoli sprouts or an equal weight of alfalfa sprouts for 8 weeks (n = 48) Evidence B source
  Fall in urea breath test urease, stool antigen and serum pepsinogens I and II, full return to baseline by 2 months after stopping, no effect sizes given in the abstract
• Adding broccoli sprout extract to standard triple therapy did not raise Helicobacter pylori eradication in 183 patients, at 81.6% against 85.2% for therapy alone. — 183 patients with clarithromycin-sensitive Helicobacter pylori, 61 per arm, sprout extract capsules yielding 1,000 micrograms of sulforaphane three times daily for 4 weeks (n = 183) Evidence B source
  Eradication 81.6% with sulforaphane against 85.2% control by intention to treat, and 96.3% against 89.2% per protocol, neither difference significant
• A synthesis of 17 human trials of broccoli, glucoraphanin and sulforaphane found the steadiest improvements in blood glucose, lipid profile and oxidative stress markers, with weaker evidence for inflammation, Helicobacter pylori and cancer. — 17 randomised clinical trials in humans published 2003 to 2013, broccoli sprouts the most common intervention (n = 17 trials) Evidence A source
  Direction of benefit for glucose, lipids and oxidative stress markers, results not pooled so no summary estimate exists
• In 16 healthy volunteers, 3 days of broccoli sprout homogenate neither switched on antioxidant genes nor protected the airways from ozone-driven neutrophil inflammation. — 16 non-atopic non-smoking healthy volunteers aged 18 to 50, 200 g of broccoli sprout homogenate daily for 3 days, crossover against alfalfa sprout homogenate (n = 16) Evidence B source
  No significant difference in sputum neutrophils or in NRF2-regulated gene expression against placebo, despite a confirmed rise in blood sulforaphane

Safety, limits and interactions

• A phase I safety trial dosed broccoli sprout extract every 8 hours for 7 days, 21 doses in all, and found no consistent abnormality across 32 blood tests including liver and thyroid markers. — healthy adult inpatients in three cohorts of three treated and one placebo each, oral dosing every 8 hours for 7 days after 5 days on a crucifer-free diet (n = 12) Evidence B source
  No significant or consistent toxicity on 32 haematology and chemistry tests, including transaminases, TSH, total T3 and free T4
• Starting the proton pump inhibitor omeprazole changed how much sulforaphane people took up from a broccoli seed and sprout extract, with more sulforaphane and its metabolites excreted while on the drug. — adults given a glucoraphanin-rich broccoli seed and sprout extract with active myrosinase before and after starting omeprazole, pilot study Evidence C source
  More sulforaphane and metabolites excreted after omeprazole was started, magnitude not given in the abstract

What people believe that the data does not support

• The European Union register lists as non-authorised the claim that the cabbage family, broccoli and sprouts included, is particularly rich in compounds that protect cells and DNA from oxidative damage. — general EU population, claim submitted for the botanical family Brassicaceae Evidence E source
  —

Who this works differently for

• The US Food and Drug Administration advises people who are more vulnerable to foodborne illness to avoid raw or lightly cooked sprouts of any kind. — the page names children, older adults, pregnant women and people with weakened immune systems as the more vulnerable groups Evidence E source
  —
• In a pilot study, age, sex, ethnicity, body mass index and vegetable intake made little difference to how much glucoraphanin people converted to sulforaphane, while greater body mass went with lower conversion efficiency. — adults taking a broccoli seed and sprout extract in a pilot bioavailability study Evidence C source
  Correlation between greater body mass and reduced conversion efficiency, coefficient not given in the abstract

What is still unknown

• In 32 women with fatty liver disease, 1,000 mg a day of broccoli sprout extract for 8 weeks moved no liver enzyme, and the authors say the dose may have been too small to show anything. — 32 women aged 30 to 45 with non-alcoholic fatty liver disease in four arms, two 500 mg tablets of standardised broccoli sprout extract daily for 8 weeks with or without Pilates (n = 32) Evidence B source
  No significant difference between groups in aspartate aminotransferase, alanine aminotransferase or alkaline phosphatase at 8 weeks

The bottom line

Form changes the dose more than the label does. In 20 healthy adults aged 19 to 50 given the same 200 micromol of sulforaphane equivalents a day, fresh sprouts put about 3 times more sulforaphane metabolites into plasma and urine than a myrosinase-treated extract, with no confidence interval reported in a small single-centre trial. Isothiocyanates are about 6 times more available to the body than the glucosinolates they come from, and boiling the sprouts destroys the plant enzyme that does the conversion. Both of those numbers come from dithiocarbamate excretion, a marker of absorption rather than of benefit. Gut bacteria can hydrolyse glucosinolates when plant myrosinase is gone, and how well they do it varies between people. The health results are small, mixed, and they stop when you stop. In 48 Helicobacter pylori positive patients, 8 weeks of 70 g of sprouts a day lowered urea breath test urease, stool antigen and serum pepsinogens I and II against alfalfa sprouts, and the abstract gives the direction without magnitudes. Those are markers of colonisation rather than eradication, and every value was back to its starting level 2 months after the sprouts stopped. When sprout extract was added to standard triple therapy in 183 patients with clarithromycin-sensitive infection, eradication ran 81.6 percent against 85.2 percent for therapy alone, and the difference was not significant. That trial excluded clarithromycin-resistant patients, so it says nothing about the infections where extra help is most wanted. In type 2 diabetes, 10 g a day of broccoli sprout powder for 4 weeks lowered serum insulin and HOMA-IR while 5 g a day did not, at P = 0.05 for the treatment effect. That was one centre, 4 weeks, and 18 of the 81 randomised patients did not reach analysis. In 52 Japanese men with fatty liver, 2 months of capsules moved median ALT from 54.0 to 48.5 IU/L while placebo moved nothing, a within-group change that is small against the spread of the group. In 32 women aged 30 to 45 with the same diagnosis, 1,000 mg a day of extract for 8 weeks moved no liver enzyme at all. With eight women per arm the authors say themselves that the dose and the study may have been too small to show anything, so that is an underpowered trial rather than a clean negative. In 16 healthy volunteers, 200 g of sprout homogenate a day for 3 days did not switch on antioxidant genes or protect the airways from ozone-driven neutrophil inflammation, and blood sulforaphane did rise, which makes it a failure of effect rather than of dosing. The safety file is small and clean as far as it reaches. A phase I trial dosed the extract every 8 hours for 7 days, 21 doses in all, and found no consistent abnormality across 32 blood tests, including transaminases, TSH, total T3 and free T4. Twelve people dosed for 7 days can rule out common and fast harm, not rare or slow harm, and the trial used an extract rather than a plate of sprouts. Raw sprouts carry a separate problem that has nothing to do with sulforaphane. The US Food and Drug Administration advises people who are more vulnerable to foodborne illness to avoid raw or lightly cooked sprouts of any kind, and its page names children, older adults, pregnant women and people with weakened immune systems. If you take omeprazole, a pilot study with no control group found more sulforaphane and its metabolites excreted after the drug was started, with the size of the change left out of the abstract. No trial here tested broccoli sprouts in pregnancy or in children, so that class-wide warning about raw sprouts is all there is for those groups. The European Union register lists as non-authorised the claim that the cabbage family, broccoli and sprouts included, is particularly rich in compounds that protect cells and DNA from oxidative damage. Non-authorised means the evidence did not carry the claim when it was assessed, not that the effect is absent. One pilot study found age, sex, ethnicity, body mass index and vegetable intake made little difference to how much glucoraphanin a person converts to sulforaphane, while greater body mass went with lower conversion efficiency, which is an exploratory look and not a rule. We have no USDA entry for this pair in the cohort, so the only portion figures on this page are the ones trials weighed out for themselves. If you are not in one of the groups the FDA tells to avoid raw sprouts, buy them fresh, chew them properly, and expect a modest and reversible shift in markers rather than a cure.

Sources

1. bspr-r3-16 · randomised controlled trial
   Forty-eight H. pylori-infected patients were randomly assigned to feeding of broccoli sprouts (70 g/d; containing 420 micromol of SF precursor) for 8 weeks or to consumption of an equal weight of alfalfa sprouts (not containing SF) as placebo
   Cancer Prev Res (Phila), 2009 · checked 2026-09-24
2. bspr-r3-17 · randomised double-blind trial
   a daily portion of BSH shake was about 200g (containing about 111g of fresh broccoli sprouts (Brassica Protection Products LLC) and water)...One dose of BSH contains about 100μmol of SFN.
   PLoS One, 2016 · checked 2026-09-24
3. bspr-r3-01 · randomised crossover trial
   3 x higher SFN metabolite levels in plasma and urine of sprout consumers, indicating enhanced SFN absorption from sprouts
   Mol Nutr Food Res, 2015 · checked 2026-09-24
4. bspr-r3-02 · crossover feeding study
   Dithiocarbamate excretion was higher when intact sprouts were chewed thoroughly rather than swallowed whole (42.4 +/- 7.5 and 28.8 +/- 2.6 micromol; P = 0.049).
   Cancer Epidemiol Biomarkers Prev, 2001 · checked 2026-09-24
5. bspr-r3-03 · crossover feeding study
   Dosing preparations included uncooked fresh sprouts (with active myrosinase) as well as homogenates of boiled sprouts that were devoid of myrosinase activity and contained either glucosinolates only or isothiocyanates only...These studies indicate that isothiocyanates are about six times more bioavailable than glucosinolates, which must first be hydrolyzed.
   Cancer Epidemiol Biomarkers Prev, 2001 · checked 2026-09-24
6. bspr-r3-04 · randomised controlled trial
   After 4 weeks, consumption of 10 g/d BSP resulted in a significant decrease in serum insulin concentration and HOMA-IR (p = 0.05 for treatment effect).
   Int J Food Sci Nutr, 2012 · checked 2026-09-24
7. bspr-r3-05 · randomised controlled trial
   Dietary supplementation with BS extract containing SF precursor GR for 2 mo significantly decreased serum levels of liver function markers, ALT [median (interquartile range), before: 54.0 (34.5-79.0) vs after supplementation: 48.5 (33.3-65.3) IU/L, P < 0.05]
   World J Gastroenterol, 2015 · checked 2026-09-24
8. bspr-r3-06 · randomised controlled trial
   Intervention with broccoli sprouts, but not with placebo, decreased the levels of urease measured by the urea breath test and H. pylori stool antigen (both biomarkers of H. pylori colonization) and serum pepsinogens I and II (biomarkers of gastric inflammation). Values recovered to their original levels 2 months after treatment was discontinued.
   Cancer Prev Res (Phila), 2009 · checked 2026-09-24
9. bspr-r3-07 · randomised controlled trial
   (A = 85.2%, B = 89.6%, and C = 81.6%) and per-protocol (A = 89.2%, B = 86.8%, and C = 96.3%) analyses
   Korean J Intern Med, 2020 · checked 2026-09-24
10. bspr-r3-12 · systematic review
    Seventeen studies were selected, and the predominant type of intervention was broccoli sprouts. More consistent results were obtained for the clinical parameters blood glucose and lipid profile and for molecular parameters of oxidative stress, indicating that there was an improvement in these parameters after intervention. Less solid evidence was found with regard to decreased inflammation, Helicobacter pylori colonization, and protection against cancer.
    Nutr Hosp, 2014 · checked 2026-09-24
11. bspr-r3-13 · randomised crossover trial
    In this proof-of-concept clinical study, we show that supplementation of SFN with broccoli sprout homogenate in healthy human subjects did not induce expression of antioxidant genes or protect against neutrophilic airway inflammation in an ozone-exposure model.
    Respir Res, 2016 · checked 2026-09-24
12. bspr-r3-08 · phase I safety trial
    Thirty-two types of hematology or chemistry tests were done before, during, and after the treatment period. Indicators of liver (transaminases) and thyroid [thyroid-stimulating hormone, total triiodothyronine (T3), and free thyroxine (T4)] function were examined in detail. No significant or consistent subjective or objective abnormal events (toxicities) associated with any of the sprout extract ingestions were observed.
    Nutr Cancer, 2006 · checked 2026-09-24
13. bspr-r3-09 · pilot clinical study
    A broccoli seed and sprout extract (BSE) rich in GR and active myrosinase was delivered before and after participants began taking the anti-acid omeprazole, a potent proton pump inhibitor. Gastric acidity appears to attenuate GR bioavailability, as evidenced by more SF and its metabolites being excreted after participants started taking omeprazole.
    Nutrients, 2019 · checked 2026-09-24
14. bspr-r3-11 · regulatory decision
    POL-HC-6819 Brassicaceae (Cruciferae) (Common Name : Botanical family that include broccoli, couliflower, cabbage, Bruxelles sprouts etc.) Are particularly rich of protective compounds that protect cells and DNA from oxidative damage/are particularly rich of compounds that help our body to eliminate toxic substances Non-authorised
    EU Register on nutrition and health claims, claim POL-HC-6819, snapshot 2026-09-21 · checked 2026-09-24
15. bspr-r3-10 · agency guidance
    People who are more vulnerable to foodborne illness (as identified above) should avoid eating raw or lightly cooked sprouts of any kind (including onion, alfalfa, clover, radish, and mung bean sprouts).
    U.S. Food and Drug Administration, Selecting and Serving Produce Safely · checked 2026-09-24
16. bspr-r3-14 · pilot clinical study
    There were negligible effects of age, sex, ethnicity, BMI, vegetable consumption, and bowel movement frequency and quality. Greater body mass correlated with reduced conversion efficiency.
    Nutrients, 2019 · checked 2026-09-24
17. bspr-r3-15 · randomised controlled trial
    After 8 weeks, no statistically significant differences were observed in enzyme levels among the groups...The modest exercise intensity and supplement dosage may have been insufficient to elicit measurable hepatic effects.
    Plant Foods Hum Nutr, 2026 · checked 2026-09-24