Detox Enzymes, DNA Damage and What Sprouts Actually Do

Brussels sprouts have an unusual place in nutrition research: they are one of the very few individual vegetables that were fed to human volunteers, in weighed amounts, under controlled conditions, with tissue biopsies taken afterwards. A small cluster of Dutch and Austrian trials from the 1990s and 2000s did exactly that, and they are the reason the “sprouts help your body detoxify” claim is not simply marketing. This page walks through what those trials measured, what they found, what they explicitly did not find, and how much of it survives contact with the amount of sprouts a person actually eats.

Table of Contents

  1. The Claim and What It Rests On
  2. What Is Actually in a Brussels Sprout
  3. What “Detoxification” Means Biochemically
  4. The Human Brussels Sprouts Trials
  5. The DNA-Damage Trials
  6. The Sex Difference Nobody Mentions
  7. What These Trials Did Not Show
  8. The Population Evidence
  9. Your Own Genes Change the Answer
  10. How Much, How Often
  11. An Honest Summary
  12. Key Research Papers
  13. Connections
  14. Featured Videos

The Claim and What It Rests On

The claim you meet everywhere is that cruciferous vegetables “support detoxification” and “protect DNA”. Both halves of that sentence describe something real, and both are routinely stretched past what the evidence supports. The useful version has three tiers, and it is worth keeping them separate in your head, because almost every argument about cruciferous vegetables is really an argument about which tier someone is standing on.

  1. Mechanism. Very strong. We know the compounds, the enzymes they act on, the transcription factor they work through, and the genes that get switched on. This is textbook biochemistry.
  2. Biomarkers in humans. Genuinely good, and unusually good for Brussels sprouts specifically. Controlled feeding trials show measurable changes in enzyme levels in human tissue and in urinary markers of oxidative DNA damage.
  3. Hard clinical endpoints. Weak. There is no randomised trial in which people ate Brussels sprouts for years and had fewer cancers than people who did not. What exists is observational epidemiology, which is suggestive and cannot settle causation.

Most disappointment about cruciferous vegetables comes from someone reading a tier-one or tier-two result and hearing a tier-three promise. The rest of this page keeps the tiers labelled.

Back to Table of Contents

What Is Actually in a Brussels Sprout

Sprouts are among the richest of the Brassica oleracea vegetables in total glucosinolates — the survey by Kushad and colleagues across the cole crops put them near the top — but the composition matters as much as the total, and this is where sprouts differ from their famous cousin.

None of these compounds does anything until the tissue is damaged and the enzyme myrosinase converts them. That conversion, and everything that destroys or preserves it, is the subject of the cooking page, and it is the single largest lever an ordinary cook controls.

Back to Table of Contents

What “Detoxification” Means Biochemically

The word has been so thoroughly hijacked by juice cleanses that it is worth restating what it means in a pharmacology textbook, because that meaning is precise and the vegetable really does act on it.

Your body handles foreign chemicals — drugs, combustion products, compounds formed when meat is browned — in two broad stages. Phase I enzymes, mostly cytochrome P450s, add a reactive handle to the molecule. This step is not automatically protective; for some compounds it is the step that turns a harmless molecule into one that can bind DNA. Phase II enzymes then attach a large water-soluble group — glutathione, glucuronic acid, a sulfate — which neutralises the reactive intermediate and makes it excretable. Glutathione S-transferases (GSTs) are the best-studied family here, and they are the ones the Brussels sprouts trials measured.

Isothiocyanates raise Phase II enzyme levels through a specific and now well-mapped route. In an unstressed cell the transcription factor Nrf2 is held in the cytoplasm by a partner protein, Keap1, and continuously destroyed. Keap1 carries reactive cysteine residues, and an isothiocyanate modifies them. Nrf2 escapes, moves into the nucleus, binds antioxidant response elements in DNA, and switches on a coordinated battery of protective genes — glutathione S-transferases, NAD(P)H quinone oxidoreductase 1, glutamate-cysteine ligase, heme oxygenase-1 and others. Kwak, Wakabayashi and Kensler laid this out as a chemoprevention mechanism, building on the work of Zhang and Talalay, who had isolated sulforaphane from broccoli in 1992 and identified it as a major inducer of these enzymes.

That last sentence contains the important nuance. The vegetable is not supplying antioxidants that mop up damage. It is supplying a mild irritant that makes the cell build its own defences — the cellular equivalent of exercise. It is why the effect outlasts the meal, and why isothiocyanates are described as indirect antioxidants.

Back to Table of Contents

The Human Brussels Sprouts Trials

Between 1994 and 1997 a group of Dutch researchers — TNO Nutrition and Food Research in Zeist working with the Department of Gastroenterology at the University Hospital St Radboud in Nijmegen — ran a short series of controlled feeding studies with an almost identical design. Healthy non-smoking volunteers ate 300 g of cooked Brussels sprouts per day, substituted for 300 g of glucosinolate-free vegetables, for one to three weeks, usually in a crossover so each person served as their own control.

Three hundred grams is about ten and a half ounces. It is a very large serving, roughly two full side portions every day, and remembering that figure is the key to reading everything below sensibly.

  1. Bogaards and colleagues, 1994. Consumption of Brussels sprouts raised α-class glutathione S-transferase levels in human blood plasma — the first demonstration in people that the vegetable moved a Phase II enzyme marker.
  2. Nijhoff and colleagues, 1995 (plasma and urine). Five men and five women, crossover, one week per arm. Plasma class-α GST rose about 1.5-fold — in the men only. Plasma class-π GST and urinary GST were unchanged. The authors concluded that the plasma α signal probably originates in the liver, and that urinary GST is a poor biomarker for this purpose.
  3. Nijhoff and colleagues, 1995 (gut tissue). The most invasive and most informative of the set: ten volunteers, crossover, one week per arm, with duodenal and rectal biopsies taken at the end of each period. Total GST enzyme activity did not change in any tissue. But the isozyme levels did: rectal GST-α was about 30 percent higher and rectal GST-π about 15 percent higher at the end of the sprouts week. The distribution of isozymes turned out to differ sharply between duodenum, rectum and lymphocytes, which is itself worth knowing — a blood measurement is not a proxy for what is happening in the colon.

Notice how modest and specific those results are. Activity did not move; particular isozymes in particular tissues moved by tens of percent, not multiples. That is what a real dietary effect looks like, and it is a long way from the language used to sell supplements.

Back to Table of Contents

The DNA-Damage Trials

The second strand asked a more direct question: does eating sprouts reduce measurable damage to DNA?

Verhagen and colleagues, 1995. Ten healthy male non-smokers. After a three-week run-in, five continued on a cruciferous-free diet and five ate 300 g of cooked Brussels sprouts daily for three weeks. The team measured 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxodG) in 24-hour urine — a standard marker of oxidative damage to DNA, produced when guanine is oxidised and the repair machinery excretes the damaged base. In the control group nothing changed. In the sprouts group, 8-oxodG excretion fell by 28 percent (P = 0.039).

Hoelzl and colleagues, 2008. A group at the Medical University of Vienna took a different approach: they fed eight volunteers 300 g of sprouts daily and then challenged their lymphocytes in the test tube with known DNA-damaging agents, reading the damage with the comet assay. Damage caused by PhIP — a heterocyclic amine formed when meat is cooked at high temperature — was strongly reduced after the sprouts period, while damage from a different heterocyclic amine, Trp-P-2, was not. Damage caused by hydrogen peroxide was 39 percent lower. Endogenous oxidised bases decreased. Serum vitamin C rose by 37 percent.

Two details in that paper deserve emphasis because they are more interesting than the headline. First, the protection against PhIP appears to work through inhibition of sulfotransferase 1A1 — a Phase II enzyme that, in PhIP's case, is the step that activates the carcinogen rather than disposing of it. So the vegetable helped by turning an enzyme down, which is the opposite of the simple “boost your detox enzymes” story. Second, the antioxidant effect was not explained by induction of glutathione peroxidase or superoxide dismutase; the authors' in vitro work pointed instead to compounds in sprouts acting as direct radical scavengers. Real biology is messier than the slogan, and the specificity — PhIP yes, Trp-P-2 no — is exactly the kind of detail that separates a measured effect from a marketing claim.

Back to Table of Contents

The Sex Difference Nobody Mentions

This is the part of the Brussels sprouts literature that almost never survives into popular summaries, and it should.

Verhagen's group repeated the DNA-damage study in 1997 with a mixed group of five men and five women in a one-week crossover. In four of the five men, urinary 8-oxodG fell. In the fifth man it rose sharply, having already been high in the control period. In the women, no effect on 8-oxodG was seen at all. The same pattern appeared in the enzyme work: the 1.5-fold rise in plasma class-α GST in the 1995 study was statistically significant in males and not in females, and the authors explicitly concluded that men appeared more susceptible to induction of hepatic GSTs than women.

What to make of this is genuinely unsettled. The groups were tiny — five people per sex — so the sensible reading is not “sprouts do not work in women” but “these studies were far too small to say, and the one sex difference they did see went unreplicated and unexplained”. Baseline hormone status, differences in body composition and dilution, and differences in baseline enzyme expression have all been proposed. The honest position is that the classic Brussels sprouts biomarker results were obtained mostly in young men, and that anyone quoting them at a general audience should say so.

Back to Table of Contents

What These Trials Did Not Show

Four things, and they matter more than the positive findings.

  1. Nobody got fewer cancers. Every study above measured a biomarker over days or weeks. Not one followed anyone to a clinical outcome. A urinary marker moving by 28 percent for three weeks is a plausible-looking arrow pointing at reduced risk; it is not reduced risk.
  2. The dose is not a normal dinner. Three hundred grams a day, every day, is far more than most people eat. Effects on biomarkers are generally dose-related — Kristensen and colleagues showed a clear dose–response between cruciferous vegetable intake and urinary isothiocyanate excretion — so a portion of sprouts twice a week does something smaller than these trials measured, and how much smaller is unknown.
  3. Not every marker moved. Total GST activity did not change. Plasma GST-π did not change. Urinary GST did not change. Trp-P-2-induced damage did not change. Reporting only the markers that moved is how a modest result becomes a headline.
  4. The samples were tiny. Five, eight, ten people. These are mechanistic pilot studies, and they are extremely good ones for what they are. They are not evidence that would justify a health claim on a package.

Back to Table of Contents

The Population Evidence

Above the trials sits observational epidemiology, which has the opposite strengths and weaknesses: huge numbers, real endpoints, and no way to rule out that people who eat more vegetables differ in a hundred other ways.

For colorectal cancer, a meta-analysis of observational studies by Wu and colleagues found higher cruciferous vegetable intake associated with modestly lower risk. The association is consistent in direction and small in size, and it is the kind of finding that fits the mechanism without proving it.

For cardiovascular outcomes, the most interesting work comes from a long-running Western Australian cohort of older women. Blekkenhorst and colleagues reported that cruciferous and allium vegetable intakes were inversely associated with fifteen-year atherosclerotic vascular disease deaths, and in a later analysis that higher cruciferous intake was associated with less extensive abdominal aortic calcification. Notably, the associations for cruciferous vegetables looked stronger than for total vegetable intake, which is at least suggestive of something specific to the family rather than to eating vegetables in general.

And at the broadest level, the dose–response meta-analysis by Aune and colleagues found fruit and vegetable intake inversely associated with cardiovascular disease, total cancer and all-cause mortality, with most of the benefit accruing up to around 800 g a day. Brussels sprouts are a good way to reach that total; they are not magic within it.

Back to Table of Contents

Your Own Genes Change the Answer

Two genetic wrinkles alter how much any of this applies to you personally.

GST genotype. Large fractions of the population carry complete deletions of the GSTM1 or GSTT1 genes and produce none of those enzymes at all. Because GSTs are one of the main routes by which isothiocyanates are conjugated and cleared, people lacking them excrete isothiocyanates more slowly — which, counter-intuitively, means the compounds hang around longer and may act on tissues for longer. Several studies of cruciferous vegetables and cancer risk have found stronger associations in people with the null genotypes. The evidence is not fully consistent, and it is not a reason to get genotyped, but it is a good explanation for why identical diets give different measurements in different people.

Your gut bacteria. When cooking has destroyed the plant's own myrosinase, the job of converting glucosinolates to isothiocyanates falls to bacteria in the colon, and people differ substantially in how well their microbiota does it. That is the main reason two people can eat the same boiled sprouts and excrete very different amounts of isothiocyanate, and it is covered further on the cooking and tolerance page.

Back to Table of Contents

How Much, How Often

Nobody can give you an evidence-based number, because no trial has compared doses in humans with the intention of finding a threshold. What can be said is bounded and useful.

Back to Table of Contents

An Honest Summary

Brussels sprouts contain compounds that measurably change human biochemistry. In controlled feeding studies at a large daily dose, they raised specific glutathione S-transferase isozymes in rectal tissue and in the plasma of male volunteers, cut urinary excretion of an oxidative DNA-damage marker by about a quarter, and made lymphocytes markedly more resistant to damage from a meat-derived carcinogen and from hydrogen peroxide. That is a genuinely strong showing for a single vegetable, and very few foods have anything like it.

What has never been shown is that any of this translates into a person living longer or avoiding a disease they would otherwise have had. The population data are consistent with it and cannot establish it. Anyone who tells you Brussels sprouts prevent cancer has skipped the step where that gets demonstrated.

The reasonable conclusion is unglamorous and worth acting on anyway: sprouts are a cheap, high-fibre, nutrient-dense vegetable with an unusually well-characterised set of bioactive compounds and an unusually good mechanistic case. Eat them because they are good food and the biology is plausible, cook them so the compounds survive, and treat anyone selling the extract with suspicion.

Back to Table of Contents

Key Research Papers

  1. Verhagen H, Poulsen HE, Loft S, van Poppel G, Willems MI, van Bladeren PJ. Reduction of oxidative DNA-damage in humans by Brussels sprouts. Carcinogenesis. 1995;16(4):969-970. — doi:10.1093/carcin/16.4.969
  2. Verhagen H, de Vries A, Nijhoff WA, Schouten A, van Poppel G, Peters WHM, van den Berg H. Effect of Brussels sprouts on oxidative DNA-damage in man. Cancer Letters. 1997;114(1-2):127-130. — doi:10.1016/s0304-3835(97)04641-7
  3. Nijhoff WA, Grubben MJAL, Nagengast FM, Jansen JBMJ, Verhagen H, van Poppel G, Peters WHM. Effects of consumption of Brussels sprouts on intestinal and lymphocytic glutathione S-transferases in humans. Carcinogenesis. 1995;16(9):2125-2128. — doi:10.1093/carcin/16.9.2125
  4. Nijhoff WA, Mulder TPJ, Verhagen H, van Poppel G, Peters WHM. Effects of consumption of Brussels sprouts on plasma and urinary glutathione S-transferase class-α and -π in humans. Carcinogenesis. 1995;16(4):955-957. — doi:10.1093/carcin/16.4.955
  5. Bogaards JJP, Verhagen H, Willems MI, van Poppel G, van Bladeren PJ. Consumption of Brussels sprouts results in elevated α-class glutathione S-transferase levels in human blood plasma. Carcinogenesis. 1994;15(5):1073-1075. — doi:10.1093/carcin/15.5.1073
  6. Hoelzl C, Glatt H, Meinl W, Sontag G, Haidinger G, Kundi M, Simic T, Chakraborty A, Bichler J, Ferk F, Angelis K, Nersesyan A, Knasmüller S. Consumption of Brussels sprouts protects peripheral human lymphocytes against PhIP and oxidative DNA-damage: results of a controlled human intervention trial. Molecular Nutrition & Food Research. 2008;52(3):330-341. — doi:10.1002/mnfr.200700406
  7. Bogaards JJP, van Ommen B, Falke HE, Willems MI, van Bladeren PJ. Glutathione S-transferase subunit induction patterns of Brussels sprouts, allyl isothiocyanate and goitrin in rat liver and small intestinal mucosa. Food and Chemical Toxicology. 1990;28(2):81-88. — doi:10.1016/0278-6915(90)90014-e
  8. Zhang Y, Talalay P, Cho CG, Posner GH. A major inducer of anticarcinogenic protective enzymes from broccoli: isolation and elucidation of structure. Proceedings of the National Academy of Sciences. 1992;89(6):2399-2403. — doi:10.1073/pnas.89.6.2399
  9. Kwak MK, Wakabayashi N, Kensler TW. Chemoprevention through the Keap1–Nrf2 signaling pathway by phase 2 enzyme inducers. Mutation Research. 2004;555(1-2):133-148. — doi:10.1016/j.mrfmmm.2004.06.041
  10. Fahey JW, Zalcmann AT, Talalay P. The chemical diversity and distribution of glucosinolates and isothiocyanates among plants. Phytochemistry. 2001;56(1):5-51. — doi:10.1016/s0031-9422(00)00316-2
  11. Steinkellner H, Rabot S, Freywald C, Nobis E, Scharf G, Chabicovsky M, Knasmüller S, Kassie F. Effects of cruciferous vegetables and their constituents on drug metabolizing enzymes involved in the bioactivation of DNA-reactive dietary carcinogens. Mutation Research. 2001;480-481:285-297. — doi:10.1016/s0027-5107(01)00188-9
  12. Hwang ES, Jeffery EH. Evaluation of urinary N-acetyl cysteinyl allyl isothiocyanate as a biomarker for intake and bioactivity of Brussels sprouts. Food and Chemical Toxicology. 2003;41(12):1817-1825. — doi:10.1016/s0278-6915(03)00235-7
  13. Kristensen M, Krogholm KS, Frederiksen H, Bügel SH, Rasmussen SE. Urinary excretion of total isothiocyanates from cruciferous vegetables shows high dose–response relationship and may be a useful biomarker. European Journal of Nutrition. 2007;46(7):377-382. — doi:10.1007/s00394-007-0676-5
  14. Traka M, Mithen R. Glucosinolates, isothiocyanates and human health. Phytochemistry Reviews. 2008;8(1):269-282. — doi:10.1007/s11101-008-9103-7
  15. Wu QJ, Yang Y, Vogtmann E, Wang J, Han LH, Li HL, Xiang YB. Cruciferous vegetables intake and the risk of colorectal cancer: a meta-analysis of observational studies. Annals of Oncology. 2013;24(4):1079-1087. — doi:10.1093/annonc/mds601
  16. Blekkenhorst LC, Bondonno CP, Lewis JR, Devine A, Zhu K, Lim WH, Woodman RJ, Beilin LJ, Prince RL, Hodgson JM. Cruciferous and allium vegetable intakes are inversely associated with 15-year atherosclerotic vascular disease deaths in older adult women. Journal of the American Heart Association. 2017;6(10). — doi:10.1161/jaha.117.006558
  17. Blekkenhorst LC, Sim M, Radavelli-Bagatini S, Bondonno NP, Bondonno CP, Devine A, Schousboe JT, Lim WH, Kiel DP, Woodman RJ, Hodgson JM, Prince RL, Lewis JR. Cruciferous vegetable intake is inversely associated with extensive abdominal aortic calcification in elderly women. British Journal of Nutrition. 2020;125(3):337-345. — doi:10.1017/s0007114520002706
  18. Aune D, Giovannucci E, Boffetta P, Fadnes LT, Keum N, Norat T, Greenwood DC, Riboli E, Vatten LJ, Tonstad S. Fruit and vegetable intake and the risk of cardiovascular disease, total cancer and all-cause mortality — a systematic review and dose-response meta-analysis. International Journal of Epidemiology. 2017;46(3):1029-1056. — doi:10.1093/ije/dyw319
  19. Kushad MM, Brown AF, Kurilich AC, Juvik JA, Klein BP, Wallig MA, Jeffery EH. Variation of glucosinolates in vegetable crops of Brassica oleracea. Journal of Agricultural and Food Chemistry. 1999;47(4):1541-1548. — doi:10.1021/jf980985s
  20. Cartea ME, Velasco P. Glucosinolates in Brassica foods: bioavailability in food and significance for human health. Phytochemistry Reviews. 2007;7(2):213-229. — doi:10.1007/s11101-007-9072-2

Back to Table of Contents

Connections

Back to Table of Contents