evipedia.ai
← Back

Sulforaphane for Health & Longevity

Evidence Review created on 11/01/2026 using AI4L / Opus 5.5

Also known as: SFN, Sulphoraphane, Sulforafan, 4-methylsulfinylbutyl isothiocyanate, 1-isothiocyanato-4-(methylsulfinyl)butane, SFX-01, Sulforadex

Motivation

Sulforaphane is a sulfur-containing plant compound that forms when broccoli, and especially young broccoli sprouts, are chopped or chewed. It draws interest because it does not act mainly as an antioxidant itself; instead, it switches on the body’s own defense genes that neutralize harmful chemicals and calm cellular stress. It is available as fresh sprouts, as broccoli seed and sprout extracts, and as stabilized supplement forms.

Since it was first isolated from broccoli in the early 1990s, sulforaphane has become one of the most studied food-derived compounds. Human trials have tested it in areas as different as autism-related behavior, blood sugar control and the clearance of airborne pollutants, and concentrated extracts are now widely marketed to health-focused adults as a daily longevity supplement.

This review examines the human evidence for and against taking sulforaphane to support health and longevity. It covers the expected benefits and the strength of their evidence, the risks and interactions, the differences between product forms, the dosing protocols used in trials, and how use can be monitored, for adults willing to follow a deliberate supplement protocol.

Benefits - Risks - Protocol - Conclusion

The following items give high-level overviews of sulforaphane from expert, industry and academic sources.

  • Sulforaphane - Rhonda Patrick

    A long-form overview of sulforaphane chemistry, bioavailability, activation of Nrf2 (a master switch for cellular defense genes) and human trials, including individual differences in response; enthusiastic in tone.

  • RHR: The Powerful Health Benefits of Sulforaphane - Chris Kresser

    A podcast episode framing sulforaphane as a detoxification aid, with a step-by-step sprout-growing tutorial and supplement guidance; the show notes link to a commercial sulforaphane product.

  • Optimize the Benefits of Broccoli - Richard Waterman

    Explains why cooking inactivates myrosinase (the plant enzyme that converts glucoraphanin, sulforaphane’s inactive precursor, into sulforaphane) and the rationale for adding mustard-seed myrosinase; published by a company selling such a product.

  • Broccoli or Sulforaphane: Is It the Source or Dose That Matters? - Yagishita et al., 2019

    A narrative review by researchers including the Johns Hopkins group that discovered sulforaphane, comparing plants, sprouts, beverages and supplements, and explaining how dose and delivery form shape trial results.

Four items are listed because no further source gave a qualifying in-depth overview. No directly relevant content was found from Peter Attia (site search returned “Nothing Found” and web searches found none) or Lifespan.io (only brief mentions inside monthly news digests), and Andrew Huberman’s most substantial coverage, an extended discussion in his episode with Rhonda Patrick (Huberman Lab episode with Rhonda Patrick), is not listed because her own overview is already included.

Grokipedia

  • Sulforaphane

    An AI-generated encyclopedic entry covering chemistry, dietary sources, Nrf2 activation, pharmacokinetics (how the body absorbs, distributes and clears a substance) and clinical research; useful for orientation, with claims best checked against primary studies.

Examine

  • Sulforaphane benefits, dosage, and side effects

    Examine’s dedicated supplement page, giving a brief overview, a dose estimate extrapolated from rat studies and a feed of recent research; its outcome-by-outcome summaries require an Examine+ subscription.

ConsumerLab

No dedicated ConsumerLab article, product review or product test of sulforaphane exists. Sulforaphane appears only as one of many ingredients discussed inside member-only answers on broader topics, such as memory, autism and blood sugar.

Systematic Reviews

The following systematic reviews and meta-analyses cover sulforaphane’s main human-tested effects and its principal suspected risk.

Mechanism of Action

  • Defense-gene activation: Sulforaphane is an isothiocyanate (a reactive sulfur-containing compound) released from glucoraphanin by plant myrosinase or by gut bacteria. It reacts with cysteine groups on Keap1 (the protein that normally marks Nrf2 for destruction), so Nrf2 accumulates in the nucleus and switches on “phase 2” defense genes, as a review describes (Kensler 2013). These include NQO1 (an enzyme that detoxifies reactive quinones) and glutathione S-transferases (GSTs; enzymes that attach glutathione, the main cellular antioxidant, to toxins) (Zhang 1992).
  • Gene-regulation effects: Sulforaphane also inhibits histone deacetylases (HDACs; enzymes that keep genes switched off), including in human blood cells after eating sprouts, according to a review (Dashwood 2007).
  • Pharmacology: It acts on many cysteine-rich proteins rather than one selective receptor. Plasma half-life is about 1.8 hours (Ye 2002). It is conjugated to glutathione by GSTs, with genotype of GSTM1 (the gene for one of these conjugating enzymes) altering its clearance (Gasper 2005), and excreted in urine as N-acetylcysteine conjugates through the mercapturic acid pathway (the route that trims glutathione conjugates into urine-ready forms), as shown in a trial by the developer TheraCryf (Clack 2025), and reaches tissues such as the prostate (Livingstone 2022). Absorption is about 70% from pre-formed sulforaphane but about 5% from glucoraphanin without active myrosinase (Egner 2011).
  • Competing view: In COPD (chronic obstructive pulmonary disease), four weeks of sulforaphane did not induce Nrf2 target genes in the airways (Wise 2016), and in mice oncogene-induced Nrf2 activity promotes tumor development (DeNicola 2011).

Historical Context & Evolution

Sulforaphane was isolated in 1992 by Paul Talalay’s group at Johns Hopkins, which was screening vegetables for compounds that induce protective phase 2 enzymes; its original intended use was cancer prevention through diet, not the treatment of disease (Zhang 1992). In 1997 the same group reported that 3-day-old broccoli sprouts contain 10–100 times more glucoraphanin than mature broccoli and that sprout extracts reduced mammary tumors in carcinogen-treated rats (Fahey 1997). This finding led to commercial sprouts and seed extracts.

From the mid-2000s, trials in Qidong, China, gave broccoli sprout preparations to people heavily exposed to aflatoxin (a mold toxin linked to liver cancer) and to air pollution. The first found no difference in urinary aflatoxin-DNA adducts between arms, only lower levels in people excreting more sulforaphane metabolites (Kensler 2005); a later one found increased urinary excretion of benzene and acrolein metabolites (Egner 2014). Interest widened after a randomized trial in young men with autism reported behavioral improvement (Singh 2014) and a study linked sulforaphane to lower liver glucose output in type 2 diabetes (Axelsson 2017). Podcasters and supplement companies then promoted it to health-focused adults.

Scientific opinion has shifted as evidence accumulated. Later autism and prediabetes trials were partly null (Magner 2023; Dwibedi 2025); the long-standing worry that cruciferous (cabbage-family vegetable) compounds impair the thyroid was not confirmed in a 12-week trial (Chartoumpekis 2019); and the early cancer-prevention hopes still lack tests against cancer outcomes in people.

Expected Benefits

High 🟩 🟩 🟩

Autism Spectrum Disorder Symptoms ⚠️ Conflicted ⭕️ Not Central to Health & Longevity

Sulforaphane may ease autism-related behavior, a neurodevelopmental condition rather than adult health or longevity. Positive RCTs come from independent groups led by Singh in Boston (Singh 2014), Momtazmanesh in Tehran (Momtazmanesh 2020) and Ou in Changsha (Ou 2024), and a meta-analysis of six RCTs found a small effect (Wang 2025). A Czech RCT in children aged 3–7 found no improvement (Magner 2023), and a US pediatric RCT missed its primary outcome (Zimmerman 2021). Net reading: a small, replicated benefit that is weaker in young children.

Magnitude: Total symptom scores improved by a standardized mean difference (SMD; effect size in standard-deviation units) of −0.27 versus control, with a 95% confidence interval (CI; range likely to contain the true effect) of −0.42 to −0.12 (Wang 2025).

Medium 🟩 🟩

Processing Speed and Working Memory in Older Adults

In a 12-week RCT of 144 healthy older Japanese adults, sulforaphane improved processing speed and working memory compared with placebo, with no effect on other cognitive domains or mood (Nouchi 2021). It is a single trial co-authored by an employee of Kagome, a food company, and has not been independently replicated. A schizophrenia meta-analysis found no cognitive benefit, but in a different population (Kassar 2025).

Magnitude: Main effect of sulforaphane versus placebo of η² = 0.08 for processing speed and η² = 0.07 for working memory (eta squared; the share of score variation explained by treatment) (Nouchi 2021).

Depressive Symptoms

In 66 adults with mild to moderate depression after coronary stenting or bypass surgery, six weeks of sulforaphane improved HAM-D (Hamilton Depression Rating Scale, a clinician-rated score) results more than placebo (Ghazizadeh-Hashemi 2021). It is a single trial from one Tehran group, and side effects did not differ from placebo. No trial has tested depression in people without heart disease.

Magnitude: Treatment response (at least 50% HAM-D reduction) occurred in 30% versus 6.67% with placebo; remission occurred in 23.33% versus 3.33%, a difference that was not statistically significant (Ghazizadeh-Hashemi 2021).

Low 🟩

Lower Cholesterol and Triglycerides ⚠️ Conflicted

A meta-analysis (Kassar 2025) of four schizophrenia RCTs (Dickerson 2021; Hei 2022; Chen 2025; Huang 2025): sulforaphane added to antipsychotics lowered LDL (low-density lipoprotein) cholesterol and triglycerides; add-on use and whole broccoli (Armah 2015) are indirect. A prediabetes RCT found no LDL change (Dwibedi 2025). Net reading: indirect, unconfirmed signal.

Magnitude: Pooled mean differences versus placebo (units not stated in the review) were LDL −0.25 (95% CI −0.42 to −0.07), triglycerides −0.25 (95% CI −0.42 to −0.08) and total cholesterol −0.35 (95% CI −0.57 to −0.12) (Kassar 2025); in prediabetes, LDL change versus placebo was 0.0 mmol/L (95% CI −0.2 to 0.2) (Dwibedi 2025).

Fasting Glucose in Prediabetes and Type 2 Diabetes ⚠️ Conflicted

Sulforaphane suppresses liver glucose output (Axelsson 2017). In prediabetes, an RCT missed its primary target and HbA1c (glycated hemoglobin, a 3-month glucose average) did not change (Dwibedi 2025); a kidney-disease trial found no effect (Avesani 2026). Net reading: at most small subgroup effects.

Magnitude: Fasting glucose fell 0.2 mmol/L more than with placebo (95% CI −0.44 to −0.01); HbA1c difference −0.3 mmol/mol (95% CI −1.3 to 0.6) (Dwibedi 2025).

Liver Enzymes in Fatty Liver ⚠️ Conflicted

In 52 men with fatty liver, sprout extract lowered the liver enzymes ALT (alanine aminotransferase) and γ-GTP (gamma-glutamyl transpeptidase), but not versus placebo per the full text, unlike the abstract; Kagome, the maker, sponsored it (Kikuchi 2015). A small women’s RCT found no change (Ghorbanian 2026). Net reading: unconfirmed.

Magnitude: Mean ALT changed −10.7% with the extract versus −4.3% with placebo, and γ-GTP −8.9% versus −1.1%; the difference between groups was not significant, and no 95% CI is reported (Kikuchi 2015).

Slower Prostate Cancer Marker Rise After Surgery ⚠️ Conflicted

In 78 men with rising PSA (prostate-specific antigen, a prostate cancer marker) after prostatectomy (surgical removal of the prostate), 60 mg/day stabilized sulforaphane missed its primary endpoint, while secondary PSA measures favored treatment (Cipolla 2015). A systematic review found PSA results inconsistent (ElKhalifa 2023). Net reading: unproven.

Magnitude: Mean PSA rose 0.099 ng/mL versus 0.620 ng/mL with placebo over six months, and PSA doubling time was 28.9 versus 15.5 months (Cipolla 2015).

Lower Cancer Risk

A meta-analysis of observational studies associates higher broccoli intake with lower cancer risk (Baladia 2024); this evidence is indirect (whole food, not the purified compound). In former smokers, a year of sulforaphane lowered bronchial Ki-67 (a cell-proliferation marker) without changing tissue findings (Yuan 2025).

Magnitude: Pooled relative risk (RR; cancer risk with higher versus lower broccoli intake) of cancer 0.89 (95% CI 0.82 to 0.96) in cohort studies (following people over time) and odds ratio (OR; relative odds of exposure in cases versus controls) 0.64 (95% CI 0.58 to 0.70) in case-control studies (comparing people with and without cancer) (Baladia 2024).

Schizophrenia Symptoms ⚠️ Conflicted ⭕️ Not Central to Health & Longevity

It bears on add-on psychiatric treatment, not healthy longevity. In a meta-analysis of four RCTs, total and positive symptoms (hallucinations, delusions) did not improve, while negative symptoms (lost motivation and expression; 12 weeks only) and general psychopathology (overall psychiatric symptoms) improved slightly (Kassar 2025). Net reading: small sub-score gains only.

Magnitude: General psychopathology mean difference (MD; average score difference versus placebo) −1.5 points on the PANSS (Positive and Negative Syndrome Scale) subscale (95% CI −2.78 to −0.23); negative symptoms MD −1.06 at 12 weeks (95% CI −1.95 to −0.16), not maintained (Kassar 2025).

Airway and Allergic Inflammation ⚠️ Conflicted

In grass-pollen allergic rhinitis, sprout extract plus nasal steroid improved nasal airflow (Yusin 2021). In asthma, three days of whole sprouts did not reduce airway inflammation (Sudini 2016), and in COPD four weeks changed no inflammation or lung-function measure (Wise 2016). Net reading: no consistent benefit.

Magnitude: Change in the area under the curve (AUC; total response summed over the measurement period) of peak nasal inspiratory flow (maximum nasal airflow) after allergen challenge was 0.3362 with sprout extract plus nasal steroid versus −0.0538 with placebo tablet plus saline spray (p = 0.013; p-value, the probability of a difference this large arising by chance alone), and 0.1213 with steroid plus placebo tablet (Yusin 2021); the asthma and COPD trials found no difference from placebo.

Helicobacter pylori Colonization ⚠️ Conflicted

In 48 infected adults, eight weeks of 70 g/day broccoli sprouts (whole food, an indirect form) lowered breath-test and stool-antigen markers of Helicobacter pylori (a stomach bacterium linked to ulcers) (Yanaka 2009). A sprout-extract RCT found no breath-test change (Chang 2015). Net reading: unconfirmed.

Magnitude: Breath-test and stool-antigen values fell during eight weeks of broccoli sprouts but not with alfalfa-sprout placebo, and returned to baseline two months after stopping; the accessible literature reports no outcome figure (Yanaka 2009). With sprout extract, breath-test values did not change significantly (Chang 2015).

Speculative 🟨

Detoxification of Airborne Pollutants

Qidong RCTs found higher urinary excretion of benzene and acrolein conjugates with sprout beverages (Egner 2014; Chen 2019). These are unvalidated exposure markers, and no disease outcome was measured.

Lifespan Extension

Sulforaphane extended lifespan in Caenorhabditis elegans worms via insulin/IGF-1 signaling (a growth-hormone pathway whose dampening extends animal lifespan) (Qi 2021). No mammalian or human aging-outcome data exist; the basis is animal-only.

Protection Against Alzheimer’s Disease

Sulforaphane improved memory in an amyloid-injected mouse model (Kim 2013). It raised blood glutathione in healthy humans (Sedlak 2018), but no human dementia outcome data exist.

Benefit-Modifying Factors

  • Gut microbiome: Glucoraphanin-only products rely on gut bacteria for conversion. In prediabetes, glucose responders carried more Bacteroides with the BT2160 conversion gene and had higher serum sulforaphane (Dwibedi 2025).
  • Genetic polymorphisms: People lacking GSTM1 (a gene for a glutathione-conjugating enzyme) had higher plasma sulforaphane and faster urinary excretion (Gasper 2005); GSTT1 (a related conjugating-enzyme gene) carriers excreted more benzene conjugates, but equally with or without sulforaphane (Egner 2014).
  • Baseline biomarkers: In prediabetes, the largest glucose responses occurred in people with milder, age-related metabolic profiles and lower γ-GTP; those with severe insulin resistance did not respond (Dwibedi 2025).
  • Sex: No trial reports sex-specific differences in benefit. The strongest autism trial enrolled only males (Singh 2014), and the prediabetes trial was 64% male (Dwibedi 2025).
  • Pre-existing conditions: In type 2 diabetes, glucose lowering appeared in obese patients with poorly controlled disease (Axelsson 2017); in advanced kidney disease with diabetes, no glycemic effect was seen (Avesani 2026).
  • Age: Healthy older adults showed cognitive gains (Nouchi 2021), but a trial with mean age 74 found no glycemic benefit (Avesani 2026); in autism, effects were larger above age 10 (Ou 2024).
  • Food form and preparation: Raw broccoli gave 37% sulforaphane bioavailability versus 3.4% for cooked broccoli (Vermeulen 2008), so cooked crucifers deliver far less.

Potential Risks & Side Effects

High 🟥 🟥 🟥

Gastrointestinal Adverse Events ⚠️ Conflicted

Nausea, abdominal discomfort, loose stools and vomiting likely reflect gut irritation by this reactive compound. Trials from two independent groups found them more often than placebo: Clack’s phase 1 (first-in-human safety) trial of enteric-coated (dissolving in the intestine) stabilized sulforaphane at 260–520 µmol (micromoles)/day, by its developer (Clack 2025), and Dwibedi’s Swedish prediabetes trial of 150 µmol/day sprout extract (Dwibedi 2025). Reviews of autism and cancer trials found overall adverse event rates similar to placebo (Wang 2025; ElKhalifa 2023). Net reading: mostly mild effects, most frequent in the highest-dose trial.

Magnitude: Abdominal discomfort in 94% (17 of 18) and nausea in 56% (10 of 18) of participants on stabilized sulforaphane versus 0% (0 of 6) on placebo over seven days; all but two events were mild (Clack 2025).

Medium 🟥 🟥

No risk reaches Medium: the remaining human data are trials showing no thyroid change, a single pediatric trial without reported event rates, an uncontrolled asthma study and outbreak surveillance.

Low 🟥

Thyroid Function Disruption ⚠️ Conflicted

Isothiocyanate breakdown products may compete with iodine uptake, a goitrogenic (goiter-promoting) concern mainly with iodine deficiency. A 12-week RCT found no change in TSH (thyroid-stimulating hormone), free thyroxine or thyroid autoimmunity (Chartoumpekis 2019), consistent with a systematic review (Galanty 2024). Net reading: no human harm shown.

Magnitude: Not quantified in available studies. No human trial has detected a thyroid hormone change, so no harm figure exists.

Sleep Disturbance, Irritability and Taste Intolerance

In a pediatric autism RCT, rare insomnia, irritability and intolerance of the supplement’s taste and smell were reported, none serious (Zimmerman 2021). Rates were not given, and adult trials have not described these effects.

Magnitude: Not quantified in available studies. The trial described these events as rare without giving rates for either group.

Worsened Airway Responsiveness in Some People With Asthma

In 45 adults with moderate asthma taking 100 µmol/day for 14 days, sulforaphane aggravated methacholine-induced bronchoconstriction (drug-provoked airway narrowing) in 20% while lessening it in 60% (Brown 2015). The study had no placebo group.

Magnitude: Worsened bronchoconstrictor response in about 20% of 45 participants; no control group (Brown 2015).

Foodborne Infection From Raw Sprouts

When sprouts, rather than extracts, are the source, warm, moist sprouting conditions can carry Salmonella and Shiga toxin-producing Escherichia coli. US surveillance documented outbreaks from seed and bean sprouts despite seed-disinfection guidelines (Dechet 2014).

Magnitude: 33 outbreaks affecting 1,330 reported people in the US from 1998 to 2010; surveillance data, no control group (Dechet 2014).

Speculative 🟨

Possible Promotion of Existing Tumors

In mice, Nrf2 activity helped oncogene-driven tumors grow (DeNicola 2011). Low sulforaphane concentrations also increased human cell-line proliferation (Bao 2014); the basis is animal and cell data only.

Allergic Reactions

A drug-reference monograph lists severe allergic reactions as a possible serious effect (WebMD sulforaphane monograph). A PubMed search found no case reports, and no trial reported such events.

Risk-Modifying Factors

  • Genetic polymorphisms: GSTM1 variants change how fast sulforaphane is conjugated and cleared (Gasper 2005); no study links this or other conjugating-enzyme genotypes to more adverse events.
  • Baseline biomarkers: Low iodine status or a high-normal baseline TSH is the setting in which the goitrogenic concern applies; the 12-week trial showing no thyroid change did not report participants’ iodine status (Chartoumpekis 2019).
  • Sex: No sex-specific adverse event data exist; the high-dose safety trial enrolled only men (Clack 2025), and the thyroid trial enrolled only women (Chartoumpekis 2019).
  • Pre-existing conditions: Asthma raises the chance of a worsened airway response (Brown 2015); a weakened immune system raises the consequences of sprout-borne infection (Dechet 2014); active cancer raises the theoretical tumor-protection concern.
  • Age: In type 2 diabetes with advanced kidney disease (mean age 74), 91 of 99 participants completed 12 weeks of escalating doses (Avesani 2026); older adults are more vulnerable to sprout-borne infection.
  • Dose and form: The highest gastrointestinal event rate came from the trial using the highest doses, 260–520 µmol/day of pre-formed sulforaphane, although that trial found no dose trend within this range (Clack 2025).

Key Interactions & Contraindications

Documented interactions:

  • CYP3A4 substrates (midazolam, rifampicin-induced metabolism): No interaction found: 450 µmol/day sulforaphane did not change midazolam clearance through CYP3A4 (the main drug-clearing liver enzyme) or block rifampicin’s induction of it (Poulton 2013).
  • Antipsychotics (risperidone and others): No interaction found: a meta-analysis of add-on trials reported good tolerability and fewer discontinuations with sulforaphane (Kassar 2025), and risperidone plus sulforaphane caused no excess adverse events (Momtazmanesh 2020).
  • Proton pump inhibitors (stomach-acid reducers; omeprazole, esomeprazole): In a human pilot study co-authored by Avmacol maker Nutramax, omeprazole increased sulforaphane yield from a myrosinase-containing extract, apparently by sparing myrosinase from acid (Fahey 2019), raising sulforaphane exposure, so dosing is reviewed when acid suppression changes.
  • Glucose-lowering drugs (metformin, glipizide, insulin): In a type 2 diabetes trial where nearly all participants took metformin, sprout extract further lowered fasting glucose in obese, poorly controlled patients (Axelsson 2017); possible hypoglycemia (low blood sugar) means glucose is checked more often when starting.
  • Myrosinase sources (mustard seed powder, myrosinase-containing supplements): Mustard-seed myrosinase doubled sulforaphane bioavailability from seed extract (Mastaloudis 2026) and raised it over fourfold from cooked broccoli (Okunade 2018); stronger effects, including gastrointestinal ones, are expected, so doses are reduced accordingly.
  • Vitamin K antagonists (anticoagulants that block vitamin K-dependent clotting; warfarin): The warfarin label advises steady vitamin K intake because vitamin K weakens it (DailyMed warfarin label); sprouts add vitamin K, so sprout intake is kept steady and INR (a blood-clotting test) is rechecked after changes.

Theoretical interactions (inferred from mechanism, not studied in humans):

  • CYP1A2 substrates (caffeine, theophylline, clozapine, tizanidine): Inferred from induction of CYP1A2 (a liver enzyme that clears caffeine and several drugs) by a 500 g/day whole-broccoli diet (Kall 1996), not tested with sulforaphane itself; drug levels could fall, so levels or response are rechecked after starting.
  • Glucuronidated drugs (acetaminophen, oxazepam): Inferred from faster glucuronidation (liver attachment of glucuronic acid for excretion) on a cabbage and Brussels sprouts diet (Pantuck 1984), not tested with sulforaphane itself; drug effect could weaken, so response is monitored after starting.
  • UGT1A1 substrates (irinotecan, bilirubin): Inferred from induction of UGT1A1 (the enzyme that clears bilirubin and irinotecan) by mixed cruciferous feeding, clearest in carriers of UGT1A1*28 (a common reduced-activity variant) (Navarro 2009), not tested with sulforaphane itself; irinotecan exposure could fall, so oncologists are informed.
  • Chemotherapy and radiotherapy (doxorubicin, cisplatin, radiation): Inferred from Nrf2’s protection of tumor cells against oxidative stress (DeNicola 2011); the consequence could be reduced treatment effect, so use is coordinated with the oncology team.
  • Other Nrf2 activators (curcumin, dimethyl fumarate, bardoxolone): Additive Nrf2 activation inferred from a shared target; the consequence could be amplified wanted and unwanted effects, so new combinations are introduced one at a time.
  • Broad-spectrum antibiotics (amoxicillin-clavulanate, ciprofloxacin): Inferred from gut-bacterial conversion of glucoraphanin (Dwibedi 2025); the consequence is lower sulforaphane yield from glucoraphanin-only products, avoided by products with active myrosinase.

Populations who should avoid Sulforaphane:

  • Pregnancy and breastfeeding: Inferred from the absence of human safety data and from transfer of the metabolite sulforaphane-N-acetylcysteine into human breast milk, which exposes the infant (Shore 2024); the drug-reference monograph states effects on pregnancy are unknown (WebMD sulforaphane monograph).
  • Known allergy to sulforaphane, glucoraphanin or product ingredients: Inferred from immune hypersensitivity to the ingredient itself, which the drug-reference monograph lists as a reason not to use it (WebMD sulforaphane monograph); no human case reports exist.
  • Active cancer during chemotherapy or radiotherapy, outside a trial: Inferred from Nrf2’s protection of tumor cells in mice (DeNicola 2011); no human data.
  • Untreated hypothyroidism with iodine deficiency: Inferred from the goitrogenic mechanism; a 12-week trial in healthy Chinese women, whose iodine status was not reported, showed no thyroid change (Chartoumpekis 2019).
  • Poorly controlled asthma: Inferred from an uncontrolled study in which airway responsiveness worsened in about 20% of adults with moderate asthma (Brown 2015); not studied in poorly controlled asthma.
  • Raw sprouts for people with weakened immune systems: Inferred from sprout-borne Salmonella and Escherichia coli outbreaks in the general population (Dechet 2014), whose consequences are more severe with weakened immunity; extracts avoid this route.

Risk Mitigation Strategies

Doses and timings below follow common practice unless cited.

  • Low start with slow titration: Common practice starts around 25–50 µmol/day and increases over 2–4 weeks toward trial doses, as the kidney-disease trial escalated from 50 to 150 µmol/day (Avesani 2026); this mitigates gastrointestinal adverse events.
  • Dose with food: Taking each dose with a meal mitigates nausea and abdominal discomfort, the most frequent adverse events at higher doses (Clack 2025).
  • Limiting routine use of high-dose pre-formed sulforaphane: Doses of 260 µmol/day or more caused abdominal discomfort in nearly all participants (Clack 2025); staying at 50–150 µmol/day mitigates gastrointestinal effects.
  • Iodine sufficiency and TSH check: Adequate dietary iodine and a TSH test at baseline and after 3 months, in people with thyroid disease, mitigate the goitrogenic concern.
  • Asthma symptom tracking: People with asthma track symptoms, rescue-inhaler use and peak flow for the first 2 weeks and stop if they worsen, mitigating worsened airway responsiveness (Brown 2015).
  • Glucose checks with glucose-lowering drugs: Fasting glucose checked several times weekly for the first 4 weeks mitigates hypoglycemia from additive glucose lowering.
  • Drug-level checks for narrow-margin CYP1A2 drugs: Clozapine or theophylline levels rechecked 1–2 weeks after starting mitigate under-dosing from enzyme induction (Kall 1996).
  • Sprout hygiene: Sprouts caused repeated Salmonella and Escherichia coli outbreaks (Dechet 2014); seeds sold for sprouting, refrigeration at or below 4 °C, and discarding slimy or musty sprouts mitigate this infection risk.
  • Pausing during cancer treatment: Stopping during chemotherapy or radiotherapy unless enrolled in a trial mitigates the theoretical tumor-protection concern.

Therapeutic Protocol

Doses below are cited to their sources; other parameters without a citation (timing, titration steps, dose splitting) reflect common practice.

  • Dose units: Doses are best expressed in µmol of sulforaphane or glucoraphanin; 1 mg sulforaphane is about 5.6 µmol, and 60 mg stabilized sulforaphane was used in prostate cancer (Cipolla 2015).
  • Extract approach (trial doses): 50–150 µmol/day by body weight in autism (Singh 2014), 150 µmol/day in prediabetes (Dwibedi 2025), and 95 µmol/day for 12 months in former smokers (Yuan 2025).
  • Stabilized sulforaphane approach: 60 mg/day for six months in recurrent prostate cancer (Cipolla 2015) and 30 mg/day for four weeks in young men doing resistance exercise (Sato 2021).
  • Whole-sprout approach: 70 g/day of fresh broccoli sprouts, providing 420 µmol glucoraphanin (Yanaka 2009); promoted for home use by Chris Kresser, whose episode includes a sprout-growing tutorial (Kresser RHR episode), and discussed by Rhonda Patrick (FoundMyFitness sulforaphane overview).
  • Best time of day: No trial compared timing; most dosed once daily, and morning dosing with breakfast is common practice to reduce nausea and any sleep disturbance.
  • Half-life: Plasma half-life is about 1.8 hours (Ye 2002), and urinary excretion half-life 2.4–2.6 hours (Vermeulen 2008); glucoraphanin-only products release sulforaphane more slowly (Egner 2011).
  • Single or split doses: Two 100 µmol doses 12 hours apart kept plasma levels higher at later time points than one 200 µmol dose (Atwell 2015); most trials used once-daily dosing.
  • Genetic polymorphisms: GSTM1-null people clear sulforaphane faster (Gasper 2005), but no genotype-guided dosing exists; gut-microbial conversion capacity matters most for glucoraphanin-only products (Dwibedi 2025).
  • Sex: No sex-specific dosing data exist; weight-based dosing was used in the all-male autism trial (Singh 2014).
  • Age: Older adults with kidney disease tolerated escalation from 50 to 150 µmol/day (Avesani 2026); slower titration is common practice at older ages.
  • Baseline biomarkers: In prediabetes, glucose response was largest with milder age-related profiles and low γ-GTP (Dwibedi 2025), so baseline glucose and liver enzymes help set expectations.
  • Pre-existing conditions: People with asthma start low and track airway symptoms (Brown 2015); people on glucose-lowering drugs monitor glucose during titration.

Discontinuation & Cycling

  • Duration of use: Sulforaphane is taken as an ongoing daily supplement; trials lasted from 3 days (Sudini 2016) to 12 months (Yuan 2025), and no lifelong data exist.
  • Withdrawal effects: None reported, but benefits fade: autism scores rose toward baseline within four weeks off (Singh 2014), and Helicobacter pylori markers returned within two months (Yanaka 2009).
  • Tapering: Not needed; no rebound effect has been reported, and trials stopped abruptly (Cipolla 2015).
  • Cycling: No trial tested cycling; bioavailability did not decline over 12 weeks of daily dosing (Egner 2014), so loss of effect from continuous use is not documented.

Sourcing and Quality

  • Formulation types: Products contain glucoraphanin with active myrosinase, glucoraphanin alone (relying on gut bacteria), or stabilized sulforaphane; absorption was about 70% from pre-formed sulforaphane versus 5% from a glucoraphanin-rich beverage (Egner 2011).
  • Active myrosinase: Labels naming a myrosinase source (broccoli seed, sprout or mustard seed) matter; mustard-seed myrosinase doubled bioavailability from seed extract (Mastaloudis 2026).
  • Label accuracy: Sulforaphane is unstable, products may contain less than labeled, and most “sulforaphane glucosinolate” products lack myrosinase (WebMD sulforaphane monograph).
  • Third-party testing: Certification by USP (United States Pharmacopeia) or NSF International, or published batch testing stating glucoraphanin and sulforaphane yield in µmol, helps confirm content.
  • Reputable brands: Avmacol (glucoraphanin plus myrosinase) is the product tested in a National Cancer Institute trial (NCT05121051); Life Extension sells a seed extract with mustard-seed myrosinase (Life Extension article).
  • Sprouting seeds: Seeds sold specifically for sprouting, from suppliers that test for pathogens, reduce contamination risk (Dechet 2014).

Practical Considerations

  • Time to effect: Detoxification markers rose within days (Egner 2014); behavioral and depression effects appeared over 6–18 weeks (Singh 2014; Ghazizadeh-Hashemi 2021); PSA effects emerged after three months (Cipolla 2015).
  • Common pitfalls: Cooking broccoli cut sulforaphane bioavailability from 37% to 3.4% (Vermeulen 2008); glucoraphanin-only supplements convert poorly; doses labeled in mg of “broccoli extract” do not state sulforaphane yield.
  • Regulatory status: In the US, sulforaphane products are sold as dietary supplements, which the US Food and Drug Administration (FDA) is not authorized to approve for safety and effectiveness before marketing (FDA consumer information on dietary supplements).
  • Cost and accessibility: Not exceptionally expensive or hard to access; home-grown sprouts are the cheapest source.

Interaction with Foundational Habits

  • Sleep: No direct interaction is documented in adults; rare insomnia was reported in a pediatric trial (Zimmerman 2021). Morning dosing is common practice when sleep disruption is noticed.
  • Nutrition: Potentiating: 1 g of mustard seed powder supplies myrosinase and multiplied sulforaphane yield from cooked broccoli over fourfold (Okunade 2018); raw crucifers also supply myrosinase. Cooking blunts yield (Vermeulen 2008); adequate iodine intake addresses the thyroid concern.
  • Exercise: Indirect and unclear: sulforaphane lowered creatine kinase (a muscle-damage marker) after resistance exercise in 10 men (Sato 2021), but glucoraphanin did not speed strength recovery after eccentric exercise (muscle lengthening under load) (Cesanelli 2026). Whether it blunts training adaptation is untested.
  • Stress management: No direct interaction with cortisol or the stress response has been studied. An indirect link exists through mood: sulforaphane improved depressive symptoms in heart patients (Ghazizadeh-Hashemi 2021).

Monitoring Protocol & Defining Success

Before starting, a baseline panel records the markers sulforaphane is expected to change and the safety markers tied to its known concerns: fasting glucose, a fasting lipid panel and liver enzymes for everyone; TSH for people with thyroid disease or low iodine intake; and PSA for men under prostate cancer surveillance. People taking clozapine or theophylline also record a baseline drug level.

Ongoing monitoring repeats the panel at 12 weeks, the length of most trials, and then every 6–12 months while use continues. People on glucose-lowering drugs check fasting glucose several times a week for the first 4 weeks, people with asthma track symptoms and peak flow during the first 2 weeks, and drug levels are rechecked 1–2 weeks after starting. Success is defined against the individual’s own baseline.

Biomarker Optimal Functional Range Why Measure It? Context/Notes
Fasting glucose 70–99 mg/dL (standard reference range) Expected to change: may fall slightly; safety check with glucose-lowering drugs Fast 8–12 hours; morning draw. A small fall was seen in prediabetes (Dwibedi 2025)
LDL cholesterol Below 100 mg/dL (NCEP ATP III “optimal” category) Expected to change: may fall modestly Source: NCEP ATP III guideline (Adult Treatment Panel III) of the National Cholesterol Education Program (NCEP), a US federal program without a revenue-earning membership. Conventional lab reference range: typically below 100–130 mg/dL depending on the lab. Fasting lipid panel
Triglycerides Below 150 mg/dL (NCEP ATP III “normal” category) Expected to change: may fall modestly Source: NCEP ATP III guideline. Conventional lab reference range: typically below 150 mg/dL. Requires a 9–12 hour fast
ALT 7–56 U/L (standard reference range) Expected to change: may fall in fatty liver; also a safety check Lab ranges vary by laboratory and sex; pair with γ-GTP
TSH 0.4–4.0 mIU/L (standard reference range) Safety check: a rise prompts a review of iodine intake and use Only for thyroid disease or low iodine intake; morning draw, consistent timing
PSA (men under prostate cancer surveillance) No established target; track the individual’s PSA slope and doubling time Expected to change: rise may slow Same laboratory each time; avoid ejaculation and cycling 48 hours before the draw
Clozapine or theophylline level No established target for this use; track change from the individual’s own pre-supplement level Safety check: a fall prompts a dose review Trough level (lowest level, just before a dose); CYP1A2 induction by broccoli shown in Kall 1996

Qualitative markers tracked alongside the panel:

  • Digestive tolerance: nausea, bloating or loose stools after doses
  • Sleep quality: sleep onset and night waking, especially with evening doses
  • Mood: day-to-day mood and motivation
  • Mental clarity: processing speed and working memory in daily tasks
  • Breathing (people with asthma): symptoms and rescue-inhaler use

Emerging Research

  • Psychosis prevention (DROPS): Phase 3 (large confirmatory) trial of 300 people at clinical high risk of psychosis; primary outcome 2-year conversion to psychosis; active, completion listed December 2026 (NCT03932136). A positive result would give the first hard-outcome psychiatric evidence; a null result would weaken the neuropsychiatric case.
  • Detoxification in heavy smokers: National Cancer Institute phase 2 (mid-stage) trial of Avmacol ES in 135 heavy smokers; primary outcome benzene and acrolein detoxification; recruiting (NCT05121051). A positive result would extend the pollutant-detoxification signal to Western smokers; a null result would weaken that speculative item.
  • Carcinogen detoxification in firefighters: Phase 2 trial in 72 firefighters measuring urinary acetaminophen mercapturate; active, not recruiting, completion listed November 2027 (NCT06009926). A positive result would support a detoxification effect under occupational exposure; a null result would weaken it.
  • Melanoma prevention: ECOG-ACRIN (a US cancer-research cooperative group) phase 2 trial in 120 people with atypical moles and prior melanoma; primary outcome change in mole area; recruiting (NCT07040280). A positive result would provide tissue-level cancer-prevention evidence; a null result would weaken the cancer-risk item.
  • Heart protection during doxorubicin chemotherapy: Phase 1/2 trial in 70 people with breast cancer; primary outcome cardiac function; recruiting (NCT03934905). A positive result would support protective use during chemotherapy; a null or harmful result would reinforce caution about combining it with cancer treatment.
  • Cognitive function: 100-participant trial of broccoli sprout extract in Korea; primary outcome 12-week computerized cognitive score; recruiting, completion listed December 2026 (NCT07334366). A positive result would independently replicate the cognition finding; a null result would downgrade it.
  • Microbiome-guided response: Gut-bacterial conversion genes predicted glucose response in prediabetes (Dwibedi 2025); confirmation could identify likely responders, while failure would leave the glucose signal at subgroup level.
  • Long-term Nrf2 safety: Animal data that Nrf2 promotes oncogene-driven tumor development (DeNicola 2011) and the human lung-proliferation finding (Yuan 2025) point in opposite directions; long-term human cancer-incidence data would resolve which dominates.

Conclusion

Sulforaphane is a broccoli-derived compound that switches on the body’s own chemical-defense genes, taken as fresh sprouts, seed and sprout extracts, or stabilized supplements. Its human evidence is broad but thin. The most consistent clinical signal is a small improvement in autism-related behavior, which does not bear on healthy aging in adults. For health-focused adults the more relevant findings are modest and less certain: better mental processing speed and mood in single trials, a cholesterol-lowering signal seen mainly in patients on antipsychotic drugs, and at most a slight blood sugar effect in some people with early diabetes. Benefits for cancer, pollutant clearance, brain aging and lifespan rest on lab markers, studies of people who eat broccoli, or animal work rather than on outcomes in people.

Safety is reassuring at the doses used in most trials. The main problems are dose-related digestive discomfort, occasional worsening of airway responses in asthma, and the infection hazard of raw sprouts, while the long-standing thyroid concern has not appeared in human trials. A theoretical concern is that the same defense system may also shield existing tumors. Results depend strongly on product form and on each person’s gut bacteria, and several studies and promotional sources involve a supplement maker, a food company or a drug developer with a stake in the compound. Overall, sulforaphane is low-risk and inexpensive, but its value for longevity remains unproven.

Top - Benefits - Risks - Protocol