Chrysin

Best-in-class: Chrysin

Sold as a natural aromatase inhibitor. It is a potent aromatase inhibitor — in a test tube. Orally, it barely gets into you.

How Chrysin actually works

A flavonoid from passionflower that is a genuinely potent aromatase inhibitor in vitro — the cell-culture data is strong enough that it is frequently cited as a natural alternative to anastrozole.

Where to get Chrysin

The evidence for Chrysin

Graded by what exists behind each claim.

✅ Clinically validated

• Human trials show no meaningful effect on testosterone or estradiol. This is one of the clearest negative results in the sports supplement literature.

📊 Correlative data

• No meaningful observational or traditional-use literature — a synthesized or isolated compound whose entire evidence base is trials and mechanism. Worth stating rather than leaving blank: it means there is no population-level signal either supporting or contradicting what the trials show.

🧪 Theoretical / extrapolated benefits

• Potent in-vitro aromatase inhibition with essentially no oral bioavailability — extensive first-pass glucuronidation removes it before it reaches circulation. A textbook case of in-vitro potency not surviving contact with pharmacokinetics.

How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →

What Chrysin actually does

Chrysin is 5,7-dihydroxyflavone, and the enzyme it is sold against is aromatase — CYP19A1, the cytochrome P450 that carries out three successive oxidations on the steroid A-ring, ending by expelling carbon 19 as formate and leaving an aromatic ring. That is the reaction that turns androstenedione into estrone and testosterone into estradiol. There is no other route to estrogen in a human body. Block that enzyme and estradiol falls; this is not a folk claim, it is how anastrozole and letrozole work.

The 1984 experiment is real and it is better than the marketing describes. Kellis and Vickery incubated human placental and ovarian microsomes with androstenedione and testosterone and measured aromatization against a panel of flavones. In order of decreasing potency: 7,8-benzoflavone, then chrysin, then apigenin, flavone, flavanone and quercetin — and 5,6-benzoflavone did nothing at all Kellis 1984. Chrysin was not a weak member of a class. It was second of the six.

And the kind of inhibition was identified, which matters more than the ranking. 7,8-benzoflavone and chrysin were competitive inhibitors, and both produced a spectral change in the aromatase cytochrome P-450 indicative of substrate displacement Kellis 1984. In plain terms: chrysin sits in the pocket where androstenedione is supposed to sit, close enough to the heme iron to change how the P450 absorbs light. The planar chromone with hydroxyls at positions 5 and 7 is a passable imitation of a steroid A-ring, and the enzyme is fooled by it. That 5,6-benzoflavone — the same atoms in a different arrangement — was inert is the proof that this is geometry rather than generic antioxidant hand-waving.

The second enzyme is the one nobody prints, and it is the same chemical feature that causes both. Those two free phenolic hydroxyls that let chrysin mimic a steroid also make it one of the better substrates known for UDP-glucuronosyltransferase. In human Caco-2 intestinal cells and Hep G2 liver cells, chrysin is metabolized extensively — conjugated to glucuronide and sulfate during passage rather than after it Galijatovic 1999. The molecule is not degraded; it is decorated with a sugar acid and thereby made invisible to the enzyme it was going to inhibit.

Then a transporter finishes the job. In a Caco-2 monolayer, chrysin and its conjugated metabolites were followed across the cell, and the conjugates are moved back toward the apical side — toward the gut lumen you swallowed them into Walle 1999. That is a futile cycle built into the intestinal wall: absorb, conjugate, export, repeat. Aromatase inhibition and enterocyte glucuronidation are not two facts about chrysin that happen to sit side by side. They are the same molecular feature, read by two different proteins, and the second protein wins because it gets there first.

Cell, rodent, human — and where it stops

The dish: human enzyme, human microsomes, a clean result. Human placental and ovarian microsomal aromatase, androstenedione and testosterone as substrates, chrysin as a competitive inhibitor second only to 7,8-benzoflavone Kellis 1984. Where a number is attached to a flavone of this shape in a modern human recombinant aromatase assay, it lands in the low micromolar range — damiana's pinocembrin at an IC₅₀ of 10.8 µM and acacetin at 18.7 µM are the closest measured relatives Zhao 2008. Take roughly 1–20 µM as the honest working range for flavone aromatase inhibition; that is an extrapolation from adjacent molecules and it is labeled as one.

The cell layer, where the trouble starts. Caco-2 and Hep G2 cells metabolize chrysin extensively Galijatovic 1999, and a Caco-2 monolayer exports the conjugates back to the apical compartment Walle 1999. No rodent step is needed to see the problem, which is why this page skips one.

The human pharmacokinetic study, with every number it reported. Seven volunteers took 400 mg of chrysin by mouth. Peak plasma chrysin concentrations were 3–16 ng/mL, with AUCs of 5–193 ng/mL·h. Plasma chrysin sulfate ran 30-fold higher, AUC 450–4220 ng/mL·h. Urinary recovery over the collection was 0.2–3.1 mg of chrysin and 2–26 mg of chrysin glucuronide Walle 2001.

Now do the arithmetic, because the arithmetic is the whole page. Chrysin's molecular weight is 254.2 g/mol. A peak of 16 ng/mL is 16 µg/L ÷ 254.2 = 63 nmol/L; the low end, 3 ng/mL, is 12 nmol/L. So the highest free-chrysin concentration ever measured in a human after a 400 mg dose is 63 nM, against a working requirement of 1–20 µM at the enzyme. That is a shortfall of roughly 16-fold at the most generous reading and 300-fold at the realistic one. And the 30-fold-larger plasma pool is the sulfate Walle 2001, which has had the 7-hydroxyl capped — the group that was doing the mimicry.

The dose you would need, stated so it can be checked. Holding dose-linearity, which almost certainly fails, reaching 10 µM (2,540 ng/mL) from a 400 mg dose that produces 16 ng/mL requires multiplying by about 159 — roughly 64 grams a day. No capsule contains it, no trial has given it, and the mass balance says why: of 400 mg swallowed, at most about 7% came back in urine and nearly all of that was already conjugated Walle 2001.

The human outcome study, described accurately for once. The one human hormone study did not give chrysin. It gave propolis and honey containing chrysin, at the doses usually taken as oral supplementation, for 21 days, and measured urinary testosterone at 7, 14 and 21 days against baseline and against controls. Nothing moved Gambelunghe 2003. That is a genuine negative and it points the right way — but it is one small study, of a bee product rather than a capsule, of a urine metabolite rather than serum, and it never measured estradiol at all. The claim on the shelf is an estrogen claim, and no human trial has ever drawn the number.

So the specific obstacle is not the usual one. It is not species: the 1984 enzyme was human Kellis 1984. It is not dose: the trial dose and the shelf dose are the same order. It is the intestinal wall, and it is a conjugation-plus-efflux problem that a review of chrysin's whole ADME literature treats as the central obstacle to developing it as anything Gao 2021.

Chrysin — which form, and does it matter

Start with what the bottle says it is. The card on this site describes chrysin as passionflower-derived, and passionflower does contain it — but the human study that produced the only hormone data used propolis and honey Gambelunghe 2003, while retail chrysin is commonly isolated from Oroxylum indicum seed or made synthetically. All three are the same molecule, 5,7-dihydroxyflavone, so the botanical source changes nothing about the pharmacokinetics. It changes what else is in the capsule, which matters for allergy and not for effect.

The forms that would actually change the answer, and their evidence. Every serious attempt to rescue chrysin is an attempt to get past the enterocyte: piperine co-dosing to inhibit glucuronidation, phospholipid complexes, nanoparticle and liposomal carriers, prodrugs with the 5- and 7-hydroxyls temporarily blocked. The ADME literature that catalogs them treats bioavailability as the unsolved problem rather than a solved one Gao 2021, and no published human pharmacokinetic study shows any chrysin product reaching micromolar plasma concentrations. Until one does, a ‘high-absorption’ chrysin is a claim about a capsule, not a measurement in a person.

Why piperine is the most plausible of them and still not enough. The rate-limiting step is UGT conjugation in the gut wall Galijatovic 1999 followed by apical export of the conjugate Walle 1999. Piperine inhibits glucuronidation, so the mechanism is at least aimed at the right enzyme — which is more than can be said for most absorption claims. But the required improvement is not 2-fold or 5-fold. It is two to three orders of magnitude, from 63 nM to somewhere above 1 µM Walle 2001, and no enzyme inhibitor delivers that.

The one form with a straightforward mechanistic case is the one nobody sells for this. If the target enzyme is in tissue you can reach without a gut — skin, for instance — the entire argument changes, because the first-pass step is skipped. That is a reason topical flavone work exists. It is not a reason to swallow one, and no topical chrysin has a human hormone endpoint behind it.

What a label would have to carry for this to be a considered purchase: a plasma concentration–time curve in humans, measured by mass spectrometry, showing free (unconjugated) chrysin above 1 µM. Not a milligram number, not an extract ratio, not a patent. One curve. It does not exist for any product on the market.

What would have to be true, and how you would know it was not

1. The hormone panel, which is the claim, and the expected answer is nothing. Draw sensitive estradiol by mass spectrometry and total-testosterone before starting and again at eight weeks on any retail dose. Predict both unchanged within assay variation. The prediction is not a hunch: 63 nM peak plasma against a 1–20 µM requirement Walle 2001 Zhao 2008 leaves no concentration at which the enzyme could be occupied.

2. shbg is the confirmatory negative, and it is the one people forget. A real aromatase inhibitor lowers estradiol and shifts SHBG with it, because hepatic SHBG output is estrogen-responsive. So predict SHBG flat as well. Two independent hormone read-outs both flat is a much stronger negative than estradiol alone, and it costs one extra line on the same requisition.

3. The prediction that would prove this page wrong, stated so somebody can go and win the argument. It is not a hormone. Take your product, dose it, and have free (unconjugated) plasma chrysin measured by LC-MS/MS at 1 and 2 hours. If any formulation returns a value above 254 ng/mL — 1 µM — the pharmacokinetic objection collapses and the hormone question becomes worth asking again. Nobody has published that number for any product, and it is the cheapest experiment on this page.

4. The extrapolated prediction, labeled as extrapolation. Chrysin induces UDP-glucuronosyltransferase in human intestinal and liver cell lines Galijatovic 1999, and UGT1A1 is the enzyme that conjugates bilirubin. So the marker most likely to move on chrysin is not a sex hormone at all — it is bilirubin, and the predicted direction is down, slightly, on a liver-function panel, with alt and ast unchanged. That prediction has never been tested in a person; it follows from the cell work rather than from any trial, and a flat bilirubin would tell you the induction does not reach the liver either.

5. Timing, so the test is fair. Eight weeks, same lab, same assay, morning draw both times, and if the subject menstruates, the same cycle day. Estradiol swings across a cycle by more than any aromatase inhibitor would move it, which is how an uncontrolled before-and-after manufactures a result out of nothing.

What nobody has tested yet

Nobody has ever drawn chrysin and estradiol from the same person. The pharmacokinetics were measured in seven people with no hormone endpoint Walle 2001; the hormone study measured urinary testosterone with no plasma chrysin Gambelunghe 2003. Twelve volunteers, one dose, serial plasma for free and conjugated chrysin by mass spectrometry, sensitive estradiol drawn at the same time points: one day of clinic time answers a question that has been open since 1984.

Nobody has looked in the tissue where the enzyme actually is. In men, most aromatase activity sits in subcutaneous adipose rather than in plasma. A lipophilic flavone could in principle concentrate in fat while plasma reads 63 nM, and if it did, the whole pharmacokinetic objection would need rewriting. Nobody has taken a fat biopsy after a chrysin dose and measured what is in it. This is the one genuinely open route by which the marketed mechanism could still be true, and it is stated here because it has not been closed, not because it is likely.

Nobody has run the piperine study properly. Chrysin plus a glucuronidation inhibitor is sold widely; the human plasma curve for that combination has not been published Gao 2021. It is the obvious study, it is small, and the reason it does not exist is worth sitting with.

And nobody has measured the gut wall. The enterocyte is exposed to concentrations five orders of magnitude above plasma; whether chronic dosing changes intestinal UGT1A1 activity in a living human — which would be a drug-interaction finding, not a hormone one — has never been tested.

Chrysin — its own safety story, not its category's

The unusual safety fact about chrysin is that almost nothing reaches you, and that is not the same as it being inert. A 500 mg capsule is 1.97 mmol of chrysin; dissolved in a couple of hundred milliliters of gut contents that is on the order of 8 mmol/L at the intestinal surface, against 63 nmol/L in plasma Walle 2001. The gut wall sees roughly a hundred thousand times what your bloodstream does. Any real effect of this molecule is an effect on the enterocyte.

Which points at the interaction nobody warns about. Chrysin induces UDP-glucuronosyltransferase in human intestinal cells Galijatovic 1999, and intestinal UGT1A1 conjugates bilirubin, ethinylestradiol, raloxifene and SN-38, the active metabolite of irinotecan. An oral UGT inducer sitting at millimolar concentrations against the enterocyte is a plausible way to lower exposure to drugs cleared that way — and the one that matters to most readers is the estrogen in a combined oral contraceptive. This has never been tested in a person and is stated at mechanism level, which is exactly how it should be weighed: not a documented interaction, and not a reassurance either.

The real harm here is the treatment not taken. Anyone using chrysin in place of an aromatase inhibitor prescribed after a hormone-receptor-positive breast cancer is substituting a molecule with no demonstrated human exposure for one with a measured effect on recurrence. Anyone stacking it alongside anastrozole, letrozole or exemestane is adding nothing and cannot know that without testing estradiol. That is the sentence on this page that could matter to somebody.

Two smaller, real ones. Propolis is a recognized allergen in people sensitized to bee products, and propolis-sourced chrysin is a propolis product — the only human study of the compound used it in that form Gambelunghe 2003. And this is not a pregnancy molecule: placental aromatase is required for estrogen synthesis throughout gestation Kellis 1984, and a compound whose entire sales pitch is inhibiting that enzyme has no business being taken on the assumption that it does not work.

Sources read for this page

• Kellis JT Jr, et al. Inhibition of human estrogen synthetase (aromatase) by flavones. Science 1984 · PMID 6474163
• Walle T, et al. Disposition and metabolism of the flavonoid chrysin in normal volunteers. British Journal of Clinical Pharmacology 2001 · PMID 11259985
• Galijatovic A, et al. Extensive metabolism of the flavonoid chrysin by human Caco-2 and Hep G2 cells. Xenobiotica 1999 · PMID 10647910
• Walle UK, et al. Transport of the flavonoid chrysin and its conjugated metabolites by the human intestinal cell line Caco-2. Biochemical Pharmacology 1999 · PMID 10424761
• Gambelunghe C, et al. Effects of chrysin on urinary testosterone levels in human males. Journal of Medicinal Food 2003 · PMID 14977449
• Gao S, Siddiqui N, Etim I, Du T, Zhang Y, Liang D. Developing nutritional component chrysin as a therapeutic agent: Bioavailability and pharmacokinetics consideration, and ADME mechanisms. Biomedicine & Pharmacotherapy 2021 · PMID 34449320
• Zhao J, et al. Anti-aromatase activity of the constituents from damiana (Turnera diffusa). Journal of Ethnopharmacology 2008 · PMID 18948180

How you would know if it worked

Chrysin — safety & side effects

Bloodwork to run alongside Chrysin

Baseline first, then again at 8–12 weeks.

The Basics — Start Here panel covers these in one order — 4 markers, $32.40 with the discount applied.

Check results you already have → · All 103 markers A–Z

Chrysin — frequently asked questions

What Chrysin is used for

Chrysin appears under 1 goal in the goal router.

Related Hormonal & Wellness supplements

← Browse the full Supplement Vault

↑ Back to on this page