Sulforaphane Benefits: Human Evidence Ranked by Strength (2026)

Sulforaphane benefits are real in some human endpoints, mixed in others, and unproven in several of the claims most often used to sell supplements. Our Sulforaphane Data Center now tracks 37 papers across bioavailability, safety, pollutant metabolism, glucose regulation, autism, schizophrenia, respiratory biology, H. pylori and cancer-related research.

The correct way to interpret this literature is not to count how many studies sound positive. Ask five questions: Was meaningful sulforaphane exposure achieved? What endpoint changed? Was it a biomarker or a clinical outcome? Was the prespecified primary outcome met? And does the studied formulation match the claim being made?

Sulforaphane benefits ranked by strength of human evidence

1. The strongest human evidence is about delivery

This is the most robust part of the sulforaphane literature because every downstream biological claim depends on exposure. In a crossover study, a pre-hydrolyzed sulforaphane-rich beverage produced about 70% average urinary bioavailability compared with about 5% from a glucoraphanin-rich beverage.^[1]

Fresh broccoli sprouts produced substantially higher and faster exposure than a tested broccoli supplement whose plant myrosinase had been inactivated.^[2] In a 2026 randomized crossover trial, adding mustard-seed myrosinase increased average sulforaphane bioavailability from 18.6% to 39.8%.^[3]

Evidence verdict: strong. Formulation materially changes delivered sulforaphane. Higher exposure still does not prove a better clinical outcome. See the full Sulforaphane Bioavailability guide.

2. Detoxification evidence is real—but narrow

In a 12-week randomized trial of 291 adults exposed to substantial air pollution, a broccoli sprout beverage increased urinary excretion of a benzene-derived mercapturic acid by 61% and an acrolein-derived metabolite by 23%, while the crotonaldehyde marker did not significantly change.^[4]

A later dose-response study found a significant benzene-biomarker effect only at the highest tested beverage concentration.^[5] In current smokers, a randomized crossover trial found that higher-dose Avmacol increased urinary detoxification biomarkers for benzene, acrolein and crotonaldehyde.^[6]

Evidence verdict: moderate to strong for selected xenobiotic-metabolism biomarkers. This is not evidence for a generic cleanse, liver repair or disease prevention. See Sulforaphane Detoxification.

3. NRF2 biology works in some human settings and fails in others

In healthy adults, broccoli-sprout-derived sulforaphane increased expression of several Phase II stress-response genes in nasal airway cells.^[7] But an 89-participant randomized COPD trial found no significant increase in its prespecified NRF2 target genes and no significant improvement in inflammatory or pulmonary-function endpoints despite measurable absorption.^[8]

Evidence verdict: biologically active but context-dependent. A mechanism is not a guaranteed health benefit. See Sulforaphane & NRF2.

4. Blood sugar evidence is mixed across three modern randomized trials

A 2017 randomized trial in 97 people with type 2 diabetes reported its strongest glycemic signal in participants with obesity and dysregulated diabetes.^[9]

A 2025 prediabetes trial in 74 drug-naive adults did not meet its prespecified 0.3 mmol/L fasting-glucose reduction target overall, although exploratory analyses suggested a stronger responder subgroup with distinct metabolic and microbiome characteristics.^[10]

A 2026 multicenter trial in people with type 2 diabetes and stage 3b–4 chronic kidney disease then found no significant improvement in fasting glucose, insulin or HbA1c.^[11]

Evidence verdict: mixed and phenotype-dependent, not a general glucose-lowering effect. See Sulforaphane & Blood Sugar.

5. Autism evidence includes five RCTs and two 2025 meta-analyses

The autism literature is one of the clearest examples of why positive, mixed and null trials—and competing syntheses—must be read together.

• Singh 2014: several behavioral measures improved during treatment, with scores moving toward baseline after withdrawal.^[12]
• Momtazmanesh 2020: adjunctive sulforaphane improved irritability and hyperactivity/noncompliance alongside risperidone, while several other domains did not improve.^[13]
• Zimmerman 2021: the primary overall clinical outcome was not significantly improved in the randomized phase; selected secondary and biomarker outcomes showed signals.^[14]
• Ou multicenter trial: caregiver-rated scales did not significantly differ while clinician-rated measures improved and a responder subset emerged.^[15]
• Magner 2023: no significant between-group improvement on ABC, SRS-2 or ADOS-2 outcomes.^[16]

The two 2025 meta-analyses do not give the same headline answer. Guo and colleagues found no significant pooled improvement in ABC total score, with irritability and hyperactivity signals in duration-based analyses.^[17] Wang and colleagues, pooling six randomized trials, reported small improvements across several symptom domains while also finding nonsignificant outcomes and meaningful heterogeneity.^[18]

Evidence verdict: mixed, symptom- and analysis-dependent; not an established autism treatment. See Sulforaphane & Autism.

6. Schizophrenia has a promising product-specific randomized signal

A 2025 randomized double-blind trial tested Avmacol Extra Strength as an add-on to antipsychotic treatment in adults with schizophrenia. Negative-symptom scores improved more than placebo over 24 weeks.^[19]

Evidence verdict: promising and product-specific, but in need of replication. It does not validate every sulforaphane supplement.

7. H. pylori research shows a temporary food-based effect

In a randomized feeding study, people with H. pylori infection consumed sulforaphane-rich broccoli sprouts or alfalfa sprouts for eight weeks. The broccoli-sprout group showed reductions in breath-test, stool-antigen and gastric-inflammation-related markers during treatment.^[20]

The markers moved back toward baseline after the intervention stopped. Evidence verdict: interesting temporary effects, not eradication therapy.

8. Cancer research is real—but dominated by surrogate endpoints

A recurrent-prostate-cancer Phase II study missed its primary ≥50% PSA-decline endpoint, although PSA doubling time changed during treatment.^[21] The ESCAPE trial found dose-related prostate-tissue gene-expression changes without being designed to prove reduced progression.^[22]

A 2023 systematic review identified eight randomized cancer trials and concluded that heterogeneous molecular and clinical findings were insufficient for therapeutic efficacy claims.^[23] A 2025 lung-risk trial in former smokers reduced Ki-67 but did not significantly improve bronchial histopathology or apoptosis endpoints.^[24]

A broader 2026 systematic review synthesized 118 clinical, observational and preclinical studies. It strengthens the rationale for continued chemoprevention research, but much of the evidence remains mechanistic or biomarker-based and the authors still call for larger clinical-endpoint prevention trials.^[25]

Evidence verdict: active human research, not established cancer treatment or prevention. See Sulforaphane & Cancer.

What about longevity and anti-aging?

NRF2, redox biology, xenobiotic metabolism and cellular stress responses make sulforaphane relevant to longevity research. The current human corpus does not demonstrate lifespan extension, reduced biological age or a validated broad anti-aging outcome.

How much confidence should we place in the literature?

Confidence should be endpoint-specific. Pharmacokinetic findings are unusually strong. Selected pollutant biomarkers are replicated. Disease-specific clinical outcomes are much less consistent. Formulations, dose units, populations and outcome definitions vary substantially. A 2025 clinical-trial landscape review identified 84 eligible interventional trials but only 39 peer-reviewed publications, so publication bias is a meaningful constraint.^[26]

Bottom line

The strongest defensible conclusion is that sulforaphane is biologically active in humans, systemic exposure is highly formulation-dependent, and several human endpoints are promising—but clinical benefits remain uneven, disease-specific and sometimes null.

For the formulation problem, start with Sulforaphane Bioavailability. For product decisions, use the Sulforaphane Supplement Guide and Best Sulforaphane Supplements. The complete evidence library is indexed in the Sulforaphane Data Center.