Foods high in sulforaphane

Glucoraphanin, the precursor sulforaphane is made from, is a Brassicaceae glucosinolate, so the cabbage family is where the dietary sources are. Within that family the list is shorter than most published rankings suggest. Broccoli and broccoli sprouts carry the precursor that matters. Several vegetables that appear on every list carry a different precursor and yield a different compound.

There is a second filter, and it costs more than the choice of vegetable. No food holds sulforaphane ready-made. The plant stores glucoraphanin in one compartment and myrosinase, the enzyme that cuts it, in another, and damage to the tissue brings them together. Anything that destroys the enzyme before you eat the food takes most of the yield with it, which is why a bag of frozen broccoli florets and a fresh head of the same cultivar are not the same source.

Which foods actually contain the precursor?

Broccoli, among the vegetables people eat in quantity. Kushad and colleagues grew 65 accessions across five groups of Brassica oleracea under uniform cultural conditions and measured the glucosinolates in each. Glucoraphanin predominated in broccoli and ran from 0.8 µmol per gram dry weight in the accession EV6-1 to 21.7 in the cultivar Brigadier, a 27-fold spread with soil, climate and assay held constant.

Broccoli sprouts hold more of it per gram than the mature plant by a wide margin. Fahey and colleagues reported 3-day-old sprouts of certain broccoli and cauliflower cultivars at 10 to 100 times the glucoraphanin of the corresponding mature plants, and noted that sprouts of many broccoli cultivars hold negligible indole glucosinolates, which predominate in the mature vegetable. That is a glucoraphanin figure rather than a sulforaphane one, and how much of it converts is a separate question handled on how much broccoli sprouts per day.

Broccoli seeds are richer again. West and colleagues quantified glucoraphanin in the seeds of 33 broccoli cultivars plus four each of kohlrabi, radish, cauliflower, Brussels sprouts, kale and cabbage, and two of raab, and the paper is explicit about why: seeds concentrate glucoraphanin and are easier to process than vegetative parts. That is the feedstock for most sulforaphane supplements rather than a food. Seeds are not an edible portion of any of these plants, so a seed figure cannot be read as a serving.

Does frozen broccoli deliver sulforaphane?

Much less than fresh, and a human trial puts a number on the gap. Eighteen healthy volunteers ate soups made from lightly cooked fresh broccoli florets or lightly cooked commercial frozen florets, and sulforaphane thio-conjugates were measured in plasma and urine. Bioavailability was about tenfold higher from the fresh soup. The authors attributed the gap to destruction of myrosinase activity by the commercial blanching and freezing process, and concluded that the short blanch used to produce frozen broccoli substantially reduces sulforaphane bioavailability.

Read what that does and does not say. The frozen florets still contained glucoraphanin. What they had lost was the enzyme, so conversion fell to the gut, and the same paper showed that human colonic flora can produce sulforaphane, erucin and their nitriles from glucoraphanin. That route works and it is far less productive than the plant's own enzyme. The abstract names no analytical platform for the plasma and urine measurements and the full text sits behind a subscription, so no absolute concentration from it appears here.

The food-science side of the same question was worked out in a laboratory the following year, and it explains the blanch. Industrial blanching aims to inactivate peroxidase, while lipoxygenase does more of the damage during frozen storage. Dosz and Jeffery found that blanching at 86 °C or above inactivated peroxidase, lipoxygenase and myrosinase together, whereas 76 °C removed 92 percent of lipoxygenase activity at a cost of only 18 percent of myrosinase-dependent sulforaphane formation. A processor using the lower temperature would keep most of the enzyme. None of this was tested in people.

They also tried to rescue the finished product. Thawing frozen broccoli for 9 hours supported no sulforaphane formation on its own. The enzyme had been destroyed rather than merely kept apart from its substrate, so breaking the tissue open achieved nothing. Broccoli root turned out to be no more heat-stable a source of the enzyme than the floret. Daikon radish root did survive, including 10 minutes at 125 °C, and 0.25 percent daikon added to frozen broccoli as it thawed supported sulforaphane formation with no visible change to the broccoli.

What human trials measured, preparation against preparation

Four trials, all in people, each comparing two ways of eating the same precursor. Bioavailability percentages from different laboratories and different assays are not strictly comparable, so each row names its own design.

The pattern across all four is the same, and it is about the enzyme rather than the vegetable. Fresh beat frozen because the enzyme survived. Raw beat cooked because the enzyme survived. Sprouts beat capsules on a matched precursor dose because the enzyme came with them. The mustard trial is the one that isolates it: the vegetable, the dose and the cooking were held constant, and adding a gram of mustard powder was the only thing that changed. Its spread is wide enough to be worth seeing, though, at a standard deviation of 33.9 on a mean of 44.7, so the size of that gain in any one person is not settled. The full set of measured content figures, each with its weight basis, is in the sulforaphane content by food table.

Why kale, cabbage and Brussels sprouts are on the lists

Because they are rich in glucosinolates, and most lists stop there. Kushad's five groups make the distinction measurable: in Brussels sprouts, cabbage, cauliflower and kale the predominant glucosinolate was sinigrin, at 8.9, 7.8, 9.3 and 10.4 µmol per gram dry weight. Sinigrin yields allyl isothiocyanate, the compound behind the heat of mustard and horseradish. Glucoraphanin is the glucosinolate that yields sulforaphane, and among those five groups it predominated in broccoli alone. The paper reports no glucoraphanin figure for any of the other four.

The enzyme is not what these vegetables lack. Dosz and colleagues estimated myrosinase activity in unextracted fresh tissue from five broccoli and three kale cultivars and found that broccoli myrosinase shows no selectivity between sinigrin and glucoraphanin, with the two hydrolysis products correlating at r = 0.859 and p = 0.006. The enzyme will cut whichever glucosinolate is there. In kale what is there is mostly sinigrin, so what comes out is mostly allyl isothiocyanate.

Those vegetables are worth eating and the isothiocyanates they do produce have a literature of their own. They are a poor choice for raising sulforaphane specifically. Watercress, bok choy and radish appear on the same lists and no primary measurement with a stated assay has been found for their glucoraphanin content, so this site carries no figure for them rather than an estimate.

What follows for a shopping list

Fresh broccoli and fresh broccoli sprouts, prepared so the enzyme survives. That means raw or lightly cooked rather than boiled, and chopped or chewed properly, since the conversion needs the tissue broken. How to eat broccoli sprouts covers the four preparation steps that have been tested in people, each with its own trial.

If what you have is frozen or already cooked, the tested repair is to supply the enzyme from somewhere else. That is the mustard result in the table above, measured in 12 adults. Daikon radish did the same job in the laboratory study, but nobody has fed that version to a person. Neither turns frozen broccoli into fresh, and both beat doing nothing.

Doses used in trials, and how much of each one people actually absorbed, are on how much sulforaphane per day. For the chemistry connecting glucoraphanin to the compound, see glucoraphanin.

No pages in this section yet.