Epicatechin

Epicatechin at a Glance

(−)-Epicatechin is the flavan-3-ol behind the cocoa-flavanol cardiovascular story — the compound most often credited when dark chocolate is described as good for the heart.

The class-level evidence is real: cocoa flavanols improve endothelial function and modestly lower blood pressure across a substantial trial literature, with a plausible mechanism through nitric-oxide signalling.

It is also this site’s clearest isolate-versus-mixture case. When purified epicatechin has been tested on its own, it has not reproduced the whole-cocoa result — so the benefit appears to belong to the flavanol mixture rather than to this single molecule, which is the opposite of how supplements are sold.

What It's Used For

(−)-Epicatechin is the flavan-3-ol most credited as the “active” behind the cocoa-flavanol cardiovascular story — and it is a cautionary tale about the isolate-vs-mixture gap. Almost every positive human trial dosed a cocoa-flavanol mixture (or dark chocolate), in which epicatechin travels with (+)-catechin, procyanidins, theobromine and other flavanols; when investigators isolate pure (−)-epicatechin and test it against placebo, the vascular signal largely disappears ^{4,5Reference 4Dower JI et al. · 2015RCTEffects of the pure flavonoids epicatechin and quercetin on vascular function and cardiometabolic health — a randomised, double-blind, placebo-controlled crossover trialView study →Reference 5Kirch N et al. · 2018RCTA nutritive dose of pure (−)-epicatechin does not beneficially affect increased cardiometabolic risk factors in overweight-to-obese adults — a randomised, placebo-controlled, double-blind crossover studyView study →}.

So the honest picture is: strong evidence for the cocoa/flavan-3-ol class, weak-to-null evidence for the purified monomer. Keep it distinct throughout from (+)-catechin (its epimer), EGCG (the gallated green-tea catechin) and the proanthocyanidin polymers it helps build. Unusually for a flavonoid, (−)-epicatechin is well absorbed (~82% of a dose) — but it still circulates as conjugates and microbial metabolites, not the free aglycone ^{16Reference 16Ottaviani JI et al. · 2016The metabolome of [2-¹⁴C](−)-epicatechin in humans — implications for the assessment of efficacy, safety and mechanisms of action of polyphenolic bioactivesView study →}.

Evidence by IndicationStrength of support · Beta

Support is an experimental score I’m building — a composite weighted by study type (human > animal > in vitro > review) and study volume, tempered for isolate-vs-mixture match and bioavailability. It’s a beta: a fast way to rank strength of evidence at a glance, not a validated metric. The scores here are for the cocoa-flavanol class; the “Rests on” column flags where the pure monomer was null. Each application links to its write-up.

01Endothelial function & blood pressurePromising65%Cocoa-flavanol RCT meta-analyses (FMD +1.3%, BP ≈−1.6 mmHg) — but pure epicatechin RCTs were null ^{1,2,4Reference 1Lin X et al. · 2016Meta-analysisCocoa flavanol intake and biomarkers for cardiometabolic health — a systematic review and meta-analysisView study →Reference 2Lagou V et al. · 2025Meta-analysisImpact of flavan-3-ols on blood pressure and endothelial function in diverse populations — a systematic review and meta-analysis of randomised controlled trialsView study →Reference 4Dower JI et al. · 2015RCTEffects of the pure flavonoids epicatechin and quercetin on vascular function and cardiometabolic health — a randomised, double-blind, placebo-controlled crossover trialView study →}

02Cardiovascular events & mortalityPromising45%COSMOS (n=21,442): primary CVD endpoint null, secondary CV-death −27%; cohorts mixed ^{6,7,8Reference 6Sesso HD et al. · 2022RCTEffect of cocoa flavanol supplementation for the prevention of cardiovascular disease events — the COcoa Supplement and Multivitamin Outcomes Study (COSMOS) randomised clinical trialView study →Reference 7Dower JI et al. · 2016Dietary epicatechin intake and 25-y risk of cardiovascular mortality — the Zutphen Elderly StudyView study →Reference 8Vogiatzoglou A et al. · 2015ObservationalAssociations between flavan-3-ol intake and CVD risk in the Norfolk cohort of the European Prospective Investigation into Cancer (EPIC-Norfolk)View study →}

03Muscle & mitochondrial functionShaky30%The isolate’s best area (small RCTs) — but one RCT blunted aerobic adaptation ^{10,12Reference 10Mafi F et al. · 2019Improvement in skeletal muscle strength and plasma levels of follistatin and myostatin induced by an 8-week resistance training and epicatechin supplementation in sarcopenic older adultsView study →Reference 12Schwarz NA et al. · 2018Schwarz NA, Blahnik ZJ, Prahadeeswaran S, et al. (2018). (−)-Epicatechin supplementation inhibits aerobic adaptations to cycling exercise in humans. Frontiers in Nutrition, 5, 132. https://pubmed.ncbi.nlm.nih.gov/30622947/View study →}

04Metabolic & cognitive markersShaky25%Small cocoa metabolic shifts; COSMOS-Mind cognition null; pure-epicatechin metabolic null ^{5,13Reference 5Kirch N et al. · 2018RCTA nutritive dose of pure (−)-epicatechin does not beneficially affect increased cardiometabolic risk factors in overweight-to-obese adults — a randomised, placebo-controlled, double-blind crossover studyView study →Reference 13Baker LD et al. · 2023RCTEffects of cocoa extract and a multivitamin on cognitive function — a randomised clinical trial (COSMOS-Mind)View study →}

Mechanisms

The load-bearing caveat: circulating epicatechin is conjugates (SREMs) and gut-microbial γ-valerolactones, not free aglycone ^{16Reference 16Ottaviani JI et al. · 2016The metabolome of [2-¹⁴C](−)-epicatechin in humans — implications for the assessment of efficacy, safety and mechanisms of action of polyphenolic bioactivesView study →} — a likely reason the isolated aglycone underperforms a whole-cocoa matrix, and why inter-individual microbiome differences matter.

Pharmacokinetics

Unusually for a flavonoid, (−)-epicatechin is well absorbed. A definitive stable-isotope (¹⁴C) study in healthy men found ~82% of an oral dose was absorbed, with more than 20 metabolites and the gut microbiome as the dominant driver of overall metabolism ^{16Reference 16Ottaviani JI et al. · 2016The metabolome of [2-¹⁴C](−)-epicatechin in humans — implications for the assessment of efficacy, safety and mechanisms of action of polyphenolic bioactivesView study →}; jejunal-perfusion work confirmed efficient uptake in the proximal small intestine with substantial first-pass conjugation and biliary/intestinal efflux ^{17Reference 17Actis-Goretta L et al. · 2013Intestinal absorption, metabolism, and excretion of (−)-epicatechin in healthy humans assessed by using an intestinal perfusion techniqueView study →}. But “absorbed” does not mean “free aglycone in blood”: what crosses the enterocyte is near-completely conjugated (glucuronidation, sulfation, O-methylation) to the structurally related epicatechin metabolites (SREMs) — chiefly epicatechin-3′-glucuronide, epicatechin-3′-sulfate and methylated sulfates ^{18Reference 18Natsume M et al. · 2003AnimalStructures of (−)-epicatechin glucuronide identified from plasma and urine after oral ingestion of (−)-epicatechin — differences between human and ratView study →}. Total SREM peaks are low-micromolar-to-high-nanomolar at ~1–2 h, cleared with a ~2–3 h half-life. The fraction reaching the colon is cleaved by microbiota to 5-(3′,4′-dihydroxyphenyl)-γ-valerolactones, which appear later, persist longer, and vary between people with their microbiome ^{16Reference 16Ottaviani JI et al. · 2016The metabolome of [2-¹⁴C](−)-epicatechin in humans — implications for the assessment of efficacy, safety and mechanisms of action of polyphenolic bioactivesView study →}. This profile is specific to (−)-epicatechin — not EGCG (more poorly absorbed), not (+)-catechin (the epimer), and not the non-absorbed proanthocyanidin polymers.

Clinical trials

Epicatechin’s trial record is large but almost entirely cocoa-flavanol mixtures (endothelial function, BP, the COSMOS hard-outcome trial). Genuine pure-(−)-epicatechin efficacy RCTs are few — and their cardiovascular arms are null.

Cocoa-flavanol mixtures	Pure (−)-epicatechin	Hard-outcome	Cognition
Many RCTs + meta-analyses	Few (BP/FMD null; muscle mixed)	1 (COSMOS, primary null)	1 (COSMOS-Mind, null)

Last checked: July 2026.

Isolate vs. Plant Studies

Epicatechin is the textbook example of a molecule whose reputation rides on a mixture. The cardiovascular evidence is cocoa, not epicatechin: the FMD and blood-pressure benefits come from cocoa-flavanol extracts and dark chocolate, where epicatechin is only the presumed active among many flavanols and procyanidins ^{1,6Reference 1Lin X et al. · 2016Meta-analysisCocoa flavanol intake and biomarkers for cardiometabolic health — a systematic review and meta-analysisView study →Reference 6Sesso HD et al. · 2022RCTEffect of cocoa flavanol supplementation for the prevention of cardiovascular disease events — the COcoa Supplement and Multivitamin Outcomes Study (COSMOS) randomised clinical trialView study →}. The purified monomer, tested alone, has failed to reproduce those vascular effects ^{4,5Reference 4Dower JI et al. · 2015RCTEffects of the pure flavonoids epicatechin and quercetin on vascular function and cardiometabolic health — a randomised, double-blind, placebo-controlled crossover trialView study →Reference 5Kirch N et al. · 2018RCTA nutritive dose of pure (−)-epicatechin does not beneficially affect increased cardiometabolic risk factors in overweight-to-obese adults — a randomised, placebo-controlled, double-blind crossover studyView study →}. Two mechanistic reasons are worth stating: the circulating species after any dose are conjugates and microbial γ-valerolactones rather than free epicatechin, and a whole-cocoa matrix may deliver co-factors (other flavanols, or effects on the microbiome) the isolate lacks ^{16Reference 16Ottaviani JI et al. · 2016The metabolome of [2-¹⁴C](−)-epicatechin in humans — implications for the assessment of efficacy, safety and mechanisms of action of polyphenolic bioactivesView study →}. Within this database epicatechin is named across many herbs — cacao (its flagship source), tea, sarsaparilla, kratom and others — usually as an antioxidant co-constituent, and often bundled with (+)-catechin or as part of the proanthocyanidin fraction. A housekeeping note: it must not be conflated with (+)-catechin (separate page), EGCG (green-tea gallated catechin), or the proanthocyanidins (polymers) — and unlike (+)-catechin, epicatechin carries no cianidanol-style hemolysis history.

Prevalence in Nature

(−)-Epicatechin is one of the two dominant flavan-3-ol monomers in the plant kingdom (the 2,3-cis diastereomer of (+)-catechin) and a principal building block of proanthocyanidins ^{22,25Reference 22Neveu V et al. · 2010Phenol-Explorer — an online comprehensive database on polyphenol contents in foods ((−)-epicatechin, food values)View study →Reference 25Xie DY et al. · 2003Role of anthocyanidin reductase, encoded by BANYULS, in plant flavonoid biosynthesis (ANR → (−)-epicatechin)View study →}. Its richest common source by a wide margin is cocoa and dark chocolate — natural cocoa powder averages ~158 mg/100 g (range 63–330), dark chocolate ~70 mg/100 g, versus only ~15 mg/100 g in milk chocolate ^{22Reference 22Neveu V et al. · 2010Phenol-Explorer — an online comprehensive database on polyphenol contents in foods ((−)-epicatechin, food values)View study →}. It is also abundant in broad beans, apples (especially the skin and cider varieties), grapes and red wine, berries (blackberry, raspberry) and stone fruit, and it is a minor but consistent component of green and black tea (~7.9 and ~3.9 mg/100 mL respectively; black tea is lower because oxidative “fermentation” polymerizes monomers into theaflavins), alongside red wine (~3.8 mg/100 mL) ^{22Reference 22Neveu V et al. · 2010Phenol-Explorer — an online comprehensive database on polyphenol contents in foods ((−)-epicatechin, food values)View study →}.

Processing destroys it — the critical caveat for cocoa. Raw beans are extremely rich (~30 mg/g defatted bean), but the chain to a chocolate bar is destructive at every step: fermentation and drying polymerize the monomer, roasting degrades it non-linearly (~8% loss under mild roasting rising to essentially complete destruction at 190–200 °C) ^{23Reference 23Fernández-Romero E et al. · 2020The kinetics of total phenolic content and monomeric flavan-3-ols during the roasting process of Criollo cocoaView study →}, and alkalization (“Dutching”) collapses it further ^{24Reference 24Miller KB et al. · 2008Impact of alkalisation on the antioxidant and flavanol content of commercial cocoa powdersView study →}. This is why high-flavanol dark chocolate and minimally processed, non-alkalized cocoa are the headline sources while most confectionery is poor.

Biosynthetically, (−)-epicatechin comes from the anthocyanidin branch: leucoanthocyanidin → anthocyanidin → (via anthocyanidin reductase, ANR/BANYULS) → (−)-epicatechin (2,3-cis) — parallel to the leucoanthocyanidin-reductase (LAR) route that makes (+)-catechin (2,3-trans) ^{25Reference 25Xie DY et al. · 2003Role of anthocyanidin reductase, encoded by BANYULS, in plant flavonoid biosynthesis (ANR → (−)-epicatechin)View study →}. Both feed the proanthocyanidin polymers. There is essentially no non-plant source.

Discovery & Synthesis

(−)-Epicatechin shares the historical “catechin”/catechu naming lineage — originally from Acacia catechu / gambier (the same source that named its epimer and, in turn, the catechol functional group). Structurally it is a flavan-3-ol: a flavan with a 3-hydroxyl and a catechol B-ring (3,3′,4′,5,7-pentahydroxyflavan), and — crucially — no 4-oxo group and no C2–C3 double bond, so the C-ring is saturated and non-planar, with two stereocenters. The defining relationship is with catechin: (−)-epicatechin is the 2,3-cis epimer [(2R,3R)], (+)-catechin the 2,3-trans [(2R,3S)], and these two are by far the most abundant natural flavan-3-ol monomers. The cis geometry makes epicatechin prone to epimerize to catechin on heating — one reason cocoa and tea processing shift the isomer balance.

Commercially it is obtained by extraction from its rich sources (cocoa above all, plus tea, apples, grapes) and is specifically the epicatechin the cocoa-flavanol field studies; it also serves as the predominant extension unit of the procyanidins. Enantioselective total synthesis of flavan-3-ols exists (setting the two stereocenters by asymmetric dihydroxylation/epoxidation) and is used chiefly for isotopically-labeled standards — the ¹⁴C material behind the definitive PK study was made this way — but plant extraction remains the practical source ^{21Reference 21Yang Z et al. · 2019Asymmetric synthesis of chiral flavan-3-olsView study →}.

Patents: not yet researched (future patent-loop pass).

Toxicity & Safety

As a dietary flavan-3-ol consumed for millennia in cocoa, tea, apples and grapes, (−)-epicatechin has very low intrinsic toxicity and is well tolerated in trials — cocoa-flavanol extracts delivering up to ~1,000 mg flavanols twice daily for up to three months were tolerated without adverse changes in blood pressure, platelets or lipids, and isolated-epicatechin dosing in small studies showed no serious safety signals. There is no established toxic dose for the monomer.

The best-established interaction is the flavan-3-ol class effect on non-heme iron: these catechol polyphenols chelate iron in the gut and dose-dependently reduce its absorption, so high-flavanol foods/supplements are best separated from iron-rich meals or iron supplements, particularly in iron deficiency ^{19Reference 19Hurrell RF et al. · 1999Inhibition of non-haem iron absorption in man by polyphenolic-containing beveragesView study →}. CYP/transporter signals exist in vitro but are weak-to-unestablished for epicatechin given how little free aglycone circulates — treat as theoretical. Two important non-interactions to state plainly: (−)-epicatechin does not carry the immune-hemolysis history of the (+)-catechin drug cianidanol (a different stereoisomer, at gram-scale drug doses) ^{20Reference 20Rotoli B et al. · 1985Immune-mediated acute intravascular haemolysis caused by cianidanol (Catergen) — the (+)-catechin drug, cited only to distinguish it from (−)-epicatechinView study →}, and the hepatotoxicity concern attached to concentrated green-tea/EGCG extracts is EGCG-driven and should not be generalized to dietary epicatechin.

Dosage

There is no established dose of isolated (−)-epicatechin, and the cardiovascular trials that “work” dose cocoa-flavanol mixtures — typically ~400–600 mg/day of flavan-3-ols from foods in the dietary-bioactive guideline (a food-pattern target, delivering on the order of 80 mg/day epicatechin), not a dose of the monomer ^{9Reference 9Crowe-White KM et al. · 2022Flavan-3-ols and cardiometabolic health — first ever dietary bioactive guidelineView study →}. Pure-epicatechin RCTs used 25–100 mg/day, and their cardiovascular arms were null ^{4,5Reference 4Dower JI et al. · 2015RCTEffects of the pure flavonoids epicatechin and quercetin on vascular function and cardiometabolic health — a randomised, double-blind, placebo-controlled crossover trialView study →Reference 5Kirch N et al. · 2018RCTA nutritive dose of pure (−)-epicatechin does not beneficially affect increased cardiometabolic risk factors in overweight-to-obese adults — a randomised, placebo-controlled, double-blind crossover studyView study →}.

These are descriptive research figures, not a recommendation — the class evidence is a food-pattern recommendation (cocoa, tea, apples, berries), and the isolated monomer has not been shown to reproduce the cardiovascular effects.

References

1. Lin X, Zhang I, Li A, et al. (2016). Cocoa flavanol intake and biomarkers for cardiometabolic health — a systematic review and meta-analysis. The Journal of Nutrition, 146(9), 1734S–1745S. https://pubmed.ncbi.nlm.nih.gov/27683874/
2. Lagou V, Greyling A, Ferruzzi MG, et al. (2025). Impact of flavan-3-ols on blood pressure and endothelial function in diverse populations — a systematic review and meta-analysis of randomised controlled trials. European Journal of Preventive Cardiology. https://pubmed.ncbi.nlm.nih.gov/40126033/
3. Ellinger S, Reusch A, Stehle P, Helfrich HP. (2012). Epicatechin ingested via cocoa products reduces blood pressure in humans — a nonlinear regression model with a Bayesian approach. The American Journal of Clinical Nutrition, 95(6), 1365–1377. https://pubmed.ncbi.nlm.nih.gov/22552030/
4. Dower JI, Geleijnse JM, Gijsbers L, et al. (2015). Effects of the pure flavonoids epicatechin and quercetin on vascular function and cardiometabolic health — a randomised, double-blind, placebo-controlled crossover trial. The American Journal of Clinical Nutrition, 101(5), 914–921. https://pubmed.ncbi.nlm.nih.gov/25934864/
5. Kirch N, Berk L, Liegl Y, et al. (2018). A nutritive dose of pure (−)-epicatechin does not beneficially affect increased cardiometabolic risk factors in overweight-to-obese adults — a randomised, placebo-controlled, double-blind crossover study. The American Journal of Clinical Nutrition, 107(4), 540–548. https://pubmed.ncbi.nlm.nih.gov/29868915/
6. Sesso HD, Manson JE, Aragaki AK, et al. (2022). Effect of cocoa flavanol supplementation for the prevention of cardiovascular disease events — the COcoa Supplement and Multivitamin Outcomes Study (COSMOS) randomised clinical trial. The American Journal of Clinical Nutrition, 115(6), 1490–1500. https://pubmed.ncbi.nlm.nih.gov/35294962/
7. Dower JI, Geleijnse JM, Hollman PCh, et al. (2016). Dietary epicatechin intake and 25-y risk of cardiovascular mortality — the Zutphen Elderly Study. The American Journal of Clinical Nutrition, 104(1), 58–64. https://pubmed.ncbi.nlm.nih.gov/27225434/
8. Vogiatzoglou A, Mulligan AA, Bhaniani A, et al. (2015). Associations between flavan-3-ol intake and CVD risk in the Norfolk cohort of the European Prospective Investigation into Cancer (EPIC-Norfolk). Free Radical Biology & Medicine, 84, 1–10. https://pubmed.ncbi.nlm.nih.gov/25795512/
9. Crowe-White KM, Evans LW, Kuhnle GGC, et al. (2022). Flavan-3-ols and cardiometabolic health — first ever dietary bioactive guideline. Advances in Nutrition, 13(6), 2070–2083. https://pubmed.ncbi.nlm.nih.gov/36190328/
10. Mafi F, Biglari S, Ghardashi Afousi A, Gaeini AA. (2019). Improvement in skeletal muscle strength and plasma levels of follistatin and myostatin induced by an 8-week resistance training and epicatechin supplementation in sarcopenic older adults. Journal of Aging and Physical Activity, 27(3), 384–391. https://pubmed.ncbi.nlm.nih.gov/30299198/
11. McDonald CM, Ramirez-Sanchez I, Oskarsson B, et al. (2021). (−)-Epicatechin induces mitochondrial biogenesis and markers of muscle regeneration in adults with Becker muscular dystrophy. Muscle & Nerve, 63(2), 239–249. https://pubmed.ncbi.nlm.nih.gov/33125736/
12. Schwarz NA, Blahnik ZJ, Prahadeeswaran S, et al. (2018). (−)-Epicatechin supplementation inhibits aerobic adaptations to cycling exercise in humans. Frontiers in Nutrition, 5, 132. https://pubmed.ncbi.nlm.nih.gov/30622947/
13. Baker LD, Manson JE, Rapp SR, et al. (2023). Effects of cocoa extract and a multivitamin on cognitive function — a randomised clinical trial (COSMOS-Mind). Alzheimer’s & Dementia, 19(4), 1308–1319. https://pubmed.ncbi.nlm.nih.gov/36102337/
14. Ramirez-Sanchez I, Maya L, Ceballos G, Villarreal F. (2010). (−)-Epicatechin activation of endothelial cell endothelial nitric oxide synthase, nitric oxide, and related signaling pathways. Hypertension, 55(6), 1398–1405. https://pubmed.ncbi.nlm.nih.gov/20404222/
15. Brossette T, Hundsdörfer C, Kröncke KD, et al. (2011). (−)-Epicatechin induces calcium and translocation-independent eNOS activation and increases NO in human endothelial cells. European Journal of Nutrition. https://pubmed.ncbi.nlm.nih.gov/21327831/
16. Ottaviani JI, Borges G, Momma TY, et al. (2016). The metabolome of [2-¹⁴C](−)-epicatechin in humans — implications for the assessment of efficacy, safety and mechanisms of action of polyphenolic bioactives. Scientific Reports, 6, 29034. https://pubmed.ncbi.nlm.nih.gov/27363516/
17. Actis-Goretta L, Lévèques A, Rein M, et al. (2013). Intestinal absorption, metabolism, and excretion of (−)-epicatechin in healthy humans assessed by using an intestinal perfusion technique. The American Journal of Clinical Nutrition, 98(4), 924–933. https://pubmed.ncbi.nlm.nih.gov/23864538/
18. Natsume M, Osakabe N, Oyama M, et al. (2003). Structures of (−)-epicatechin glucuronide identified from plasma and urine after oral ingestion of (−)-epicatechin — differences between human and rat. Free Radical Biology & Medicine, 34(7), 840–849. https://pubmed.ncbi.nlm.nih.gov/12654472/
19. Hurrell RF, Reddy M, Cook JD. (1999). Inhibition of non-haem iron absorption in man by polyphenolic-containing beverages. The British Journal of Nutrition, 81(4), 289–295. https://pubmed.ncbi.nlm.nih.gov/10999016/
20. Rotoli B, Giglio F, Bile M, Formisano S. (1985). Immune-mediated acute intravascular haemolysis caused by cianidanol (Catergen) — the (+)-catechin drug, cited only to distinguish it from (−)-epicatechin. Haematologica, 70(6), 495–499. https://pubmed.ncbi.nlm.nih.gov/3938427/
21. Yang Z, et al. (2019). Asymmetric synthesis of chiral flavan-3-ols. Natural Product Research. https://pubmed.ncbi.nlm.nih.gov/30521402/
22. Neveu V, Perez-Jiménez J, Vos F, et al. (2010). Phenol-Explorer — an online comprehensive database on polyphenol contents in foods ((−)-epicatechin, food values). Database (Oxford), 2010, bap024. https://pubmed.ncbi.nlm.nih.gov/20428313/
23. Fernández-Romero E, Chavez-Quintana SG, Siche R, et al. (2020). The kinetics of total phenolic content and monomeric flavan-3-ols during the roasting process of Criollo cocoa. Antioxidants, 9(2), 146. https://pubmed.ncbi.nlm.nih.gov/32050504/
24. Miller KB, Hurst WJ, Payne MJ, et al. (2008). Impact of alkalisation on the antioxidant and flavanol content of commercial cocoa powders. Journal of Agricultural and Food Chemistry, 56(18), 8527–8533. https://pubmed.ncbi.nlm.nih.gov/18710243/
25. Xie DY, Sharma SB, Paiva NL, Ferreira D, Dixon RA. (2003). Role of anthocyanidin reductase, encoded by BANYULS, in plant flavonoid biosynthesis (ANR → (−)-epicatechin). Science, 299(5605), 396–399. https://pubmed.ncbi.nlm.nih.gov/12532018/

View all 25 references▾