Supplements

Sulforaphane

Updated · 21 sources

Sulforaphane is a sulfur-containing compound (an isothiocyanate) formed when broccoli and other cruciferous vegetables are chopped or chewed. It is one of the most potent known natural activators of Nrf2, a master switch for the body’s antioxidant and detoxification genes, and it is sold as broccoli sprout and broccoli seed extract supplements.

The most important facts are about delivery and evidence. Plants store an inactive precursor, glucoraphanin, and only a small and highly variable fraction becomes sulforaphane unless the plant enzyme myrosinase is still active. Human trials exist in autism, type 2 diabetes, air-pollution detoxification, schizophrenia and COPD, but they are mostly small, and several of the larger or later trials missed their primary outcomes. Supplement doses are hard to verify: researchers in the field warn that labeled “doses” often cannot be trusted.

This page summarizes what the clinical trials, pharmacokinetic studies and reviews report. It is for education only and is not medical advice.

Property Detail
Class Isothiocyanate; Nrf2 activator (acts through the Keap1 sensor protein)
Chemical name 1-isothiocyanato-4-(methylsulfinyl)butane; C₆H₁₁NOS₂, 177.3 g/mol
Precursor Glucoraphanin, a glucosinolate converted by the enzyme myrosinase (in the plant or in gut bacteria)
Richest food sources Broccoli seeds and 3-day-old broccoli sprouts, which contain 10–100 times more glucoraphanin than mature broccoli
Route Oral (sprouts, sprout or seed extracts, capsules, stabilized drug formulations)
Doses in trials Sulforaphane 9.9–847 µmol/day (median 100 µmol, about 18 mg); glucoraphanin 25–800 µmol/day (median 190 µmol)
Fraction recovered in urine 70–90% of a sulforaphane dose; typically 2–15% of a glucoraphanin dose without active myrosinase
US status No FDA-approved drug; sold as dietary supplements (more than 600 labels in the NIH database)
WADA status Not named on the 2026 Prohibited List

What it is

In intact broccoli, glucoraphanin and the enzyme myrosinase are kept in separate compartments. When the plant tissue is damaged by chopping or chewing, myrosinase meets glucoraphanin and converts it into sulforaphane (Yagishita et al., Molecules 2019). If myrosinase has been destroyed, for example by cooking or by heat during extract processing, gut bacteria can still perform the conversion, but less efficiently and very differently from person to person.

Sulforaphane was first isolated from broccoli and shown to protect rats against chemical carcinogens in the early 1990s (Yagishita et al. 2019). In 1997 the Talalay laboratory reported that 3-day-old sprouts of certain broccoli cultivars contain 10–100 times more glucoraphanin than mature plants, and that sprout extracts reduced mammary tumors in rats exposed to a chemical carcinogen (Fahey et al., PNAS 1997). Many of the human trials since then were run by the same group and its collaborators.

Mechanism: Keap1 and Nrf2

Sulforaphane’s best-characterized target is Keap1, a sulfur-rich protein that normally tags the transcription factor Nrf2 for destruction. Sulforaphane reacts with Keap1, which lets Nrf2 accumulate in the cell nucleus and switch on a set of protective genes, including detoxification enzymes such as glutathione S-transferases and NQO1 (Kensler et al., Top Curr Chem 2013). Inside cells, sulforaphane is itself conjugated with glutathione and excreted in urine as mercapturic acid metabolites (Yagishita et al. 2019).

Sulforaphane also has other effects in laboratory studies, such as inhibiting histone deacetylases (HDACs), enzymes that control gene expression (Hei et al., Schizophr Bull Open 2022). Showing that Nrf2 is actually switched on in human target tissues has been harder (see the COPD trial below).

Related antioxidant-pathway supplements with pages in this library include N-acetylcysteine (NAC) and glutathione.

Pharmacokinetics and bioavailability

Bioavailability is the central practical problem, and it depends on which form is taken.

Form taken What reaches the body Source
Sulforaphane (pre-formed) Rapidly absorbed and eliminated; 70–90% of the dose typically recovered as urinary metabolites, with small differences between people Yagishita et al. 2019
Glucoraphanin without active myrosinase Slow, variable conversion by gut bacteria; typically 2–15% of the dose recovered, 1–45% at the extremes Yagishita et al. 2019
Glucoraphanin with active plant myrosinase (fresh sprouts, unextracted seeds, myrosinase-added products) Sulforaphane 3–4 times more bioavailable than from glucoraphanin delivered without active myrosinase Fahey et al., PLoS One 2015

Each format has a drawback. Pure sulforaphane is unstable, which has led to stabilized forms such as an alpha-cyclodextrin complex and Prostaphane (Yagishita et al. 2019). Glucoraphanin is stable but converts poorly without the enzyme. For these reasons the review authors argue that the amount of sulforaphane metabolites excreted in urine, not the amount stated for the preparation, is the more reliable measure of the dose a person actually received.

Gut bacteria matter. In a 2025 prediabetes trial, people who responded to broccoli sprout extract had more of a Bacteroides gene cluster needed to convert glucoraphanin to sulforaphane, and its abundance correlated with blood sulforaphane levels (Dwibedi et al., Nat Microbiol 2025).

A drug developer has also made SFX-01 (Sulforadex), a synthetic sulforaphane stabilized in alpha-cyclodextrin. In a 7-day phase 1 study in healthy men, enteric-coated tablets delivering 46.2 or 92.4 mg of sulforaphane per dose were absorbed with peak blood levels 3–6 hours after dosing, and urinary excretion ranged from under 1% to 41% (Clack et al., Adv Ther 2025).

Converting doses

Trials report doses in micromoles (µmol). Using sulforaphane’s molecular weight of 177.3 g/mol (PubChem), 100 µmol of sulforaphane is about 17.7 mg. Across published trials, sulforaphane doses ranged from 9.9 to 847 µmol per day (median 100 µmol) and glucoraphanin doses from 25 to 800 µmol per day (median 190 µmol) (Yagishita et al. 2019).

Regulatory status

Sulforaphane is not an FDA-approved drug; a Drugs@FDA search returned no product with sulforaphane as an active ingredient (openFDA, accessed November 2026). It is sold in the US as a dietary supplement ingredient, and the NIH Dietary Supplement Label Database returned 633 labels mentioning it in November 2026 (NIH DSLD). Supplements do not need FDA review of efficacy or label accuracy before sale.

What the research shows

Autism spectrum disorder

Autism is the most publicized use, and the results have become less impressive as trials grew.

Trial Participants Dose, duration Result
Singh et al., PNAS 2014 44 young men aged 13–27 with moderate to severe autism (29 sulforaphane, 15 placebo) 50–150 µmol/day for 18 weeks Aberrant Behavior Checklist improved 34% and Social Responsiveness Scale 17% vs under 3.3% change on placebo; scores rose back toward baseline after stopping
Zimmerman et al., Mol Autism 2021 57 children aged 3–12 (45 analyzed) 15-week double-blind phase, then open label Primary outcome (OACIS clinical impression) not significant; the caregiver-rated Aberrant Behavior Checklist improved at 15 weeks, the Social Responsiveness Scale did not
Magner et al., Nutrients 2023 40 children aged 3–7 randomized, 28 completed 36 weeks No significant difference from placebo on ADOS-2, SRS-2 or the Aberrant Behavior Checklist
Ou et al., J Autism Dev Disord 2024 108 children in China 12 weeks No difference on caregiver-rated scales; clinician-rated scales improved; the authors called for more trials

A 2025 meta-analysis of six randomized trials found small improvements in total symptoms (SMD −0.27, 95% CI −0.42 to −0.12), aberrant behavior and social interaction, but not in irritability, anxiety or several social subscales. Adverse events were similar to placebo. The authors cautioned that results may be influenced by assessment tools, outcome assessors and treatment duration (Wang et al., EXCLI J 2025). SMD (standardized mean difference) is an effect size; −0.27 is small.

Type 2 diabetes and prediabetes

Trial Participants Intervention Result
Axelsson et al., Sci Transl Med 2017 97 people with type 2 diabetes Concentrated broccoli sprout extract providing about 153 µmol sulforaphane/day (dose per Yagishita et al. 2019) Lowered fasting blood glucose and HbA1c, but the benefit was reported in obese patients with poorly controlled (dysregulated) diabetes
Dwibedi et al., Nat Microbiol 2025 74 drug-naive adults with prediabetes (35 extract, 39 placebo) Broccoli sprout extract once daily for 12 weeks Did not meet the prespecified primary outcome (a 0.3 mmol/L fall in fasting glucose); the overall reduction was 0.2 mmol/L (95% CI −0.44 to −0.01). A subgroup labeled “responders” was identified in exploratory analyses

The 2017 study first identified sulforaphane by computer screening of drug gene-expression “signatures” and showed in liver cells that it suppressed glucose production through Nrf2. In diabetic animals, the effect on glucose production was similar in size to that of metformin (Axelsson et al. 2017). That comparison was made in animals, not in people.

Air pollution and detoxification

Researchers ran a series of trials in Qidong, China, a region with high exposure to air- and food-borne carcinogens (Kensler et al. 2013). In the largest, 291 adults drank a broccoli sprout beverage providing 600 µmol glucoraphanin and 40 µmol sulforaphane daily, or placebo, for 12 weeks (Egner et al., Cancer Prev Res 2014):

  • Urinary excretion of the detoxified form of benzene rose 61% and of acrolein 23% compared with placebo; crotonaldehyde did not change.
  • The effect appeared rapidly and was sustained, and bioavailability did not decline over the 12 weeks.

These are biomarker results. The trials did not measure cancer or other disease outcomes.

COPD

In a phase 2 trial at three US centers, 89 patients with chronic obstructive pulmonary disease took placebo, 25 µmol or 150 µmol of sulforaphane daily for 4 weeks. Sulforaphane was absorbed, but it did not increase expression of Nrf2 target genes in lung immune cells or airway lining cells, and inflammation and lung function did not differ between groups (Wise et al., PLoS One 2016).

Schizophrenia

In a 22-week trial of 172 patients with first-episode schizophrenia in China, sulforaphane had no effect on the primary outcome (the MATRICS composite cognitive score) or on symptom scores, though some cognitive subdomains improved (Hei et al., Schizophr Bull Open 2022). A 2025 meta-analysis of four trials (369 patients) found no improvement in total or positive symptom scores, a modest improvement in negative symptoms at 12 weeks that was not maintained, and lower LDL cholesterol and triglycerides (Kassar et al., BMC Psychiatry 2025).

Exercise and muscle

We did not find human trials that support sulforaphane for strength, muscle growth or endurance performance. The rationale offered in marketing comes from Nrf2 biology and animal studies, not from athletes.

Side effects and risks

  • Generally well tolerated in trials. The autism and COPD trials reported good tolerability at doses up to 150 µmol/day (Singh et al. 2014; Wise et al. 2016), and a meta-analysis found adverse events similar to placebo (Wang et al. 2025).
  • Digestive effects. Gastrointestinal side effects were the main adverse events in the prediabetes trial (Dwibedi et al. 2025). In the SFX-01 phase 1 study, 94% of participants on the drug had treatment-emergent adverse events, mostly mild gastrointestinal events (Clack et al. 2025).
  • Taste and behavior. Rare insomnia, irritability and intolerance of the taste and smell were reported in children (Zimmerman et al. 2021). Higher doses of pre-formed sulforaphane cause a harsh, burning taste at the back of the throat (Yagishita et al. 2019).
  • Thyroid. Cruciferous products have long been suspected of affecting the thyroid. In 45 women who drank a sulforaphane- and glucoraphanin-rich beverage for 84 days, thyroid-stimulating hormone, free thyroxine, thyroglobulin and thyroid autoimmunity did not change (Chartoumpekis et al., Food Chem Toxicol 2019).
  • High doses in animals. Very high injected doses in mice caused sedation, low body temperature and impaired coordination (Yagishita et al. 2019). These doses are far above those used in people.

Long-term safety beyond the few months studied in trials, and safety in pregnancy, are not established in the sources reviewed here.

Quality and labeling

Supplement variability is a recurring theme in this field. Fahey, Kensler and colleagues, whose group ran many of the trials, wrote that measuring broccoli-derived ingredients “can be a source of mystery, obfuscation, and confusion, that is readily exploited by the supplement industry,” and that “one cannot for the foreseeable future trust that what appears on the label is an accurate representation of ‘dose’” (Yagishita et al., Molecules 2019). In their table of clinical studies they flagged examples from their own testing:

  • a broccoli sprout powder used in diabetes trials and described as rich in sulforaphane, which they had found was not;
  • a broccoli seed extract supplement in which they had found the myrosinase inactive and the glucoraphanin content not as represented.

Practical consequences: a product labeled as “glucoraphanin” without active myrosinase may yield only a few percent of its stated amount as sulforaphane, while one labeled “sulforaphane” may degrade unless stabilized.

Sulforaphane is not named on the WADA 2026 Prohibited List (WADA 2026 Prohibited List). As with any supplement, contamination or undeclared ingredients are the main anti-doping risk.

Key takeaways

  • A potent Nrf2 activator from broccoli. Sulforaphane switches on antioxidant and detoxification genes by reacting with the Keap1 sensor protein.
  • Form decides the dose. Pre-formed sulforaphane is well absorbed; glucoraphanin without active myrosinase typically yields only 2–15%, and active myrosinase raises bioavailability 3–4 times.
  • Autism evidence weakened with larger trials. A 2014 trial was strongly positive, but later trials in children missed primary or caregiver-rated outcomes, and a 2025 meta-analysis found small effects.
  • Metabolic results are mixed. Glucose fell in obese patients with poorly controlled type 2 diabetes, but a 2025 prediabetes trial missed its primary endpoint.
  • Biomarker effects are not disease outcomes. Broccoli sprout drinks increased excretion of benzene and acrolein in polluted Qidong, but no trial has shown fewer cancers.
  • Labels are unreliable. Researchers in the field warn that labeled doses often cannot be trusted, so measured content matters.

Sources

  1. National Center for Biotechnology Information. Sulforaphane (CID 5350). PubChem. PubChem.
  2. Yagishita Y, Fahey JW, Dinkova-Kostova AT, Kensler TW. Broccoli or sulforaphane: is it the source or dose that matters? Molecules. 2019;24(19):3593. PubMed.
  3. Fahey JW, Zhang Y, Talalay P. Broccoli sprouts: an exceptionally rich source of inducers of enzymes that protect against chemical carcinogens. Proc Natl Acad Sci U S A. 1997;94(19):10367–10372. PubMed.
  4. Kensler TW, Egner PA, Agyeman AS, et al. Keap1-nrf2 signaling: a target for cancer prevention by sulforaphane. Top Curr Chem. 2013;329:163–177. PubMed.
  5. Fahey JW, Holtzclaw WD, Wehage SL, Wade KL, Stephenson KK, Talalay P. Sulforaphane bioavailability from glucoraphanin-rich broccoli: control by active endogenous myrosinase. PLoS One. 2015;10(11):e0140963. PubMed.
  6. Dwibedi C, Axelsson AS, Abrahamsson B, et al. Effect of broccoli sprout extract and baseline gut microbiota on fasting blood glucose in prediabetes: a randomized, placebo-controlled trial. Nat Microbiol. 2025;10(3):681–693. PubMed.
  7. Clack G, Moore C, Ruston L, et al. A phase 1 randomized, placebo-controlled study evaluating the safety, tolerability, and pharmacokinetics of enteric-coated stabilized sulforaphane (SFX-01) in male participants. Adv Ther. 2025;42(1):216–232. PubMed.
  8. Hei G, Smith RC, Li R, et al. Sulforaphane effects on cognition and symptoms in first and early episode schizophrenia: a randomized double-blind trial. Schizophr Bull Open. 2022;3(1):sgac024. PubMed.
  9. US Food and Drug Administration. openFDA Drugs@FDA dataset: search for active ingredient sulforaphane (no matching records; accessed November 2026). openFDA.
  10. National Institutes of Health, Office of Dietary Supplements. Dietary Supplement Label Database: search for “sulforaphane” (accessed November 2026). NIH DSLD.
  11. Singh K, Connors SL, Macklin EA, et al. Sulforaphane treatment of autism spectrum disorder (ASD). Proc Natl Acad Sci U S A. 2014;111(43):15550–15555. PubMed.
  12. Zimmerman AW, Singh K, Connors SL, et al. Randomized controlled trial of sulforaphane and metabolite discovery in children with autism spectrum disorder. Mol Autism. 2021;12(1):38. PubMed.
  13. Magner M, Thorová K, Župová V, et al. Sulforaphane treatment in children with autism: a prospective randomized double-blind study. Nutrients. 2023;15(3):718. PubMed.
  14. Ou J, Smith RC, Tobe RH, et al. Efficacy of sulforaphane in treatment of children with autism spectrum disorder: a randomized double-blind placebo-controlled multi-center trial. J Autism Dev Disord. 2024;54(2):628–641. PubMed.
  15. Wang R, Ren Z, Li Y. The effect of sulforaphane on autism spectrum disorder: systematic review and meta-analysis. EXCLI J. 2025;24:542–557. PubMed.
  16. Axelsson AS, Tubbs E, Mecham B, et al. Sulforaphane reduces hepatic glucose production and improves glucose control in patients with type 2 diabetes. Sci Transl Med. 2017;9(394):eaah4477. PubMed.
  17. Egner PA, Chen JG, Zarth AT, et al. Rapid and sustainable detoxication of airborne pollutants by broccoli sprout beverage: results of a randomized clinical trial in China. Cancer Prev Res (Phila). 2014;7(8):813–823. PubMed.
  18. Wise RA, Holbrook JT, Criner G, et al. Lack of effect of oral sulforaphane administration on Nrf2 expression in COPD: a randomized, double-blind, placebo controlled trial. PLoS One. 2016;11(11):e0163716. PubMed.
  19. Kassar O, Mansour MEM, Farag N, et al. Efficacy and safety of sulforaphane in schizophrenia: a systematic review and meta-analysis of randomized controlled trials. BMC Psychiatry. 2025;25(1):1045. PubMed.
  20. Chartoumpekis DV, Ziros PG, Chen JG, Groopman JD, Kensler TW, Sykiotis GP. Broccoli sprout beverage is safe for thyroid hormonal and autoimmune status: results of a 12-week randomized trial. Food Chem Toxicol. 2019;126:1–6. PubMed.
  21. World Anti-Doping Agency. The 2026 Prohibited List. In force January 1, 2026. WADA.