Avens

What is Avens?

Avens is the dried root and rhizome, and sometimes the flowering herb, of Geum urbanum L., a yellow-flowered perennial in the rose family, Rosaceae, that grows to about 60 cm across most of Europe, North Africa's Atlas range, Turkey, the Caucasus and western Siberia. Dig the rhizome and it smells of cloves. That smell is the single most checkable thing about the plant: the root oil of Geum urbanum is 67 to 69.2 percent eugenol, the same phenylpropanoid that dominates clove bud oil, and it is released from a stored precursor rather than sitting there in free form. Owczarek, Gudej and Kicel identified 130 compounds across the oils of G. urbanum and G. rivale in Natural Product Communications in 2013 and put root eugenol at 69.2 percent for wood avens. Vollman and Schultze reported 67 percent in Flavour and Fragrance Journal in 1995. Everything else about avens, every claim printed on a tincture bottle, is far less certain than that number.

Common Names

• Wood avens, herb bennet, herb Bennet, colewort, clove root, city avens, way bennet, goldy star and wild rye, all Geum urbanum L.

• St. Benedict's herb, from the medieval Latin herba benedicta, which is where "bennet" comes from. Nelkenwurz in German, erba benedetta in Italian, nejlikrot in Swedish

• Water avens, purple avens, chocolate root and Indian chocolate, all Geum rivale L., a different species sold under the same trade name "avens"

• Mountain avens is Dryas octopetala, a different genus in the same family. It shares no chemistry with Geum and gave its name to the Younger Dryas cold period because of its pollen in ice-age cores

• Prairie smoke, old man's whiskers and three-flowered avens are Geum triflorum, a North American species

• Lan bu zheng is Geum aleppicum and Geum japonicum var. chinense in the Chinese herb trade, a separate botanical entity

• Older botanical synonyms still on some labels: Caryophyllata officinalis Moench, Caryophyllata urbana (L.) Scop., Geum caryophyllata Gilib., Streptilon odoratum Raf.

Primary Active Compounds

• Gein, the phenolic glycoside eugenol vicianoside. Vicianose is the disaccharide α-L-arabinopyranosyl-(1→6)-β-D-glucopyranose, C11H20O10. Psenák and colleagues isolated free vicianose from the rhizodermis and primary bark of G. urbanum roots and published it in Planta Medica in 1972, volume 22, page 93

• Eugenol, liberated from gein when the root is cut, bruised or dried. Grieve's A Modern Herbal puts the volatile oil yield at about 0.04 percent of the root

• Hydrolysable tannins: gallotannins and ellagitannins. Six ellagitannins were isolated from the root by Piwowarski and colleagues in Biochemical Systematics and Ecology in 2014, volume 53, page 46: gemin A, gemin G, pedunculagin, stachyurin, stenophyllanin A and casuarinin

• Ellagic acid and gallic acid. Owczarek and Olszewska measured total ellagic acid at 46.71 ± 0.51 mg/g in aerial parts and 32.19 ± 0.50 mg/g underground, with gallic acid at 8.35 ± 0.29 and 5.25 ± 0.11 mg/g

• Flavan-3-ols. In Polish populations, root catechin ran 402 to 600 mg/100 g and epicatechin 263 to 432 mg/100 g depending on collection region

• Triterpenoids and sterols. Ton That and colleagues isolated 11 triterpenoids, 2 steroids, 6 phenolic derivatives and one new dehydrodigallic acid derivative from the ethanol root extract in Natural Product Research in 2018, the first report of steroids and triterpenes in this species

Key Note

Total tannin content is the number most often quoted and least often measured. Grieve reported roughly one-eleventh of root weight, about 9 percent. A 2019 review put it at 10.5 percent. Retail pages routinely claim "up to 30 percent." When Maier and colleagues actually screened 47 European medicinal herbs by radial diffusion in Industrial Crops and Products in 2017, Geum urbanum was one of seven plants that came in below its literature value, while Potentilla erecta, Arctostaphylos uva-ursi and four others were confirmed. Neumann and colleagues, using European Pharmacopoeia colorimetry on 14 herbal drugs in Molecules in 2022, found tannin contents spanning 2.2 to 10.4 percent of dry weight across the set. The 30 percent figure is marketing, not assay.

Avens is an old European astringent with a distinctive volatile chemistry, a well-mapped set of polyphenols and, as the next section shows, essentially no human evidence behind any of it.

Primary Functions & Benefits

Every function listed below is traditional or preclinical. None has been demonstrated in a controlled human trial.

Astringent action on mucous membranes. Hydrolysable tannins precipitate salivary and mucosal proteins, which is what "astringent" describes physically. This is the mechanism behind the traditional gargle and the traditional antidiarrhoeal use. It is the same mechanism credited to tormentil, Potentilla erecta, a close Rosaceae relative that does hold an EU herbal monograph.

Antibacterial activity in vitro. Dimitrova and colleagues reported minimum inhibitory concentrations in Chemistry Central Journal in 2017: ethyl acetate fraction at 0.078 mg/mL from aerial parts and 0.156 mg/mL from roots against Staphylococcus species, with the same values against Bacillus cereus. Bunse and colleagues, in Chemistry and Biodiversity in 2022, tested fresh and fermented root extracts against Staphylococcus aureus ATCC 6538 and Cutibacterium acnes and found the high-molecular-weight tannin fraction most active. These are agar and broth numbers, not infections in people.

Anti-biofilm and quorum-sensing interference. A 2025 paper in Antibiotics reported 72.4 to 90.5 percent inhibition of MRSA NBIMCC 8327 biofilm formation by the aerial-parts ethyl acetate extract at sub-inhibitory concentrations, versus 18.9 to 20.4 percent for the root extract. A companion paper in Plants the same year reported 84 to 85 percent inhibition of Pseudomonas aeruginosa biofilm by root extract at 1.56 to 6.25 mg/mL, with pyocyanin down 26 to 30 percent and swarming motility fully blocked at 3.12 mg/mL.

Anti-inflammatory activity at the neutrophil. Granica and colleagues, in the Journal of Ethnopharmacology in 2016, volume 188, showed that gemin A, the dominant root ellagitannin, cut CD11b surface expression, reactive oxygen species, elastase, MMP-9, IL-8 and IL-1β in stimulated human neutrophils, using quercetin as the positive control. Notably, gemin A raised TNF-α release rather than lowering it.

Antioxidant capacity. Owczarek and colleagues reported DPPH EC50 values of 26.57 ± 1.24 µg/mL for aerial parts and 26.92 ± 1.31 µg/mL for underground parts in Acta Poloniae Pharmaceutica in 2015. Paun and colleagues reported DPPH IC50 of 1.3 ± 0.1 µg/mL for an ethanol extract and 7.8 ± 0.5 µg/mL for an aqueous one. Antioxidant assay numbers are not clinical outcomes and never have been.

Forms & Standardization

There is no standardization. That is the honest headline, and it separates avens from almost every herb with a real market.

Cut and sifted root and rhizome. The default raw material. Suppliers sell it wild-crafted by the quarter pound and pound, typically 18 to 63 Canadian dollars per unit at one long-running Ontario herb house. No marker compound percentage is stated, because none is required and no compendial method has been adopted for the species.

Cut and sifted herb. The aerial parts, a chemically different product. Aerial parts carry higher ellagic acid, 46.71 mg/g versus 32.19 mg/g in the root, and a completely different volatile profile: (Z)-3-hexenol at 38.4 percent of the aerial oil versus eugenol at 69.2 percent of the root oil. Buying "avens" without specifying the part means you do not know which of these you received.

Tinctures at 1:2, 1:3 and 1:1. Practitioner suppliers sell 1:2 in 50 percent alcohol; other retailers list 1:3; the older British Herbal Pharmacopoeia figure quoted in Barnes, Anderson and Phillipson's Herbal Medicines, third edition, 2007, pages 75 to 76, is a 1:1 liquid extract in 25 percent alcohol. A 1:1 delivers roughly double the plant per millilitre of a 1:2 and triple that of a 1:3, so a dose in millilitres means nothing until you know the ratio.

Fresh versus dried root. This matters more here than for most herbs. Gein is converted to free eugenol by drying, cutting and hydrolysis, so a fresh-root tincture and a dried-root tincture are not the same preparation. Grieve records that the dried root loses much of its fragrance and should be kept whole. A powdered root that has sat in a warehouse has already vented much of its 0.04 percent volatile oil.

Fermented extracts. Used in anthroposophic pharmacy and characterised by Bunse in 2022, where rare ellagitannin sulfates were identified by LC-MSn. These are a distinct chemical product from a simple hydroethanolic extract.

What the studies actually used. No two of the preclinical papers used the same preparation. Dimitrova used methanol extracts fractionated into petroleum ether, ethyl acetate and n-butanol. Granica used aqueous, ethyl acetate and butanol fractions. Lobbens used ethanol. Günther used a library extract from the PECKISH collection. Zaharieva tested 12 extracts of differing polarity. None of them is what is in a retail bottle, and the ethyl acetate fraction that generates most of the impressive MIC numbers is not a form anyone sells for internal use.

Cosmetic-grade Geum urbanum extract. A separate supply chain. Skincare ingredient databases list it as an INCI-named botanical extract, and the 2025 Antibiotics paper reported a primary irritation index of zero for the aerial-parts ethyl acetate extract at ten times the minimum inhibitory concentration. That is a topical tolerance finding, not evidence of a cosmetic benefit.

Food Sources

Avens is a foraged food as well as a drug plant, and the food story is more defensible than the supplement story.

• The rhizome was used as a clove substitute and a flavouring for ale. Grieve records that Augsburg Ale took its character from avens root added in small bags to each cask, and that medieval practice fixed 25 March as the day to lift the root when it was most fragrant.

• Kosmala and colleagues, in Molecules in 2025, measured Geum urbanum at 2.1 g of phenolics per 100 g of fresh matter, 8.2 g of dietary fibre per 100 g and 0.1 g of vitamin C per 100 g, alongside Sanguisorba officinalis at 3.0 g and Agrimonia procera at 3.4 g of phenolics. That is several times the phenolic density of berries.

• Geum rivale root is the "chocolate root" of North American and British foraging, simmered in milk as a cocoa substitute. Its root oil is 53.3 percent cis-myrtanal and only marginally eugenol, so it does not taste of cloves.

• Young avens leaves go into spring salads and soups. Rosette leaves are the richest aerial fraction for ellagic acid and epicatechin gallate.

• Seeds of both species carry linoleic and linolenic acid as dominant fatty acids, plus derivatives of asiatic acid and madecassic acid reported for the first time in the genus by Bunse in Plants in 2021.

The practical verdict: at culinary quantities, a pinch of root in a cask of ale or a handful of leaves in a salad, avens is food with an interesting polyphenol profile. There is no dose of food-form avens that has been shown to do anything therapeutic, and no reason to think a capsule improves on the plant.

Who Should Take Avens

Almost nobody has a reason to, and the honest list is short.

• Someone with a specific interest in traditional European bitter and astringent preparations who wants a short trial of a gargle or a strong root tea for gum irritation, understanding that the entire rationale is tannin chemistry plus 500 years of habit.

• Someone using it as a flavouring, in home brewing or as a clove substitute, where the quantities are small and the purpose is culinary.

• Foragers and gardeners who already have the plant and want to use it rather than compost it.

Nobody should take avens expecting a documented clinical effect on diarrhoea, gingivitis, ulcerative colitis, fever, blood sugar or anything else. WebMD's monograph rates it "Insufficient Evidence" for every listed use, which is unusually blunt for that source and happens to be accurate.

Who Should AVOID or Use Caution

Contraindications

• Pregnancy. Avens is classed as possibly unsafe in pregnancy on the basis of reported effects on the menstrual cycle. There is no controlled reproductive toxicology for the species, which is itself the reason to avoid it.

• Breastfeeding. No data at all. Not a reassurance, an absence.

• Children. The only paediatric exposure documented in the literature is the 62-child anthroposophic case series, which used homeopathic and low-dose anthroposophic dilutions rather than a herbal tincture dose, and was not designed to detect harm.

• Known clove or eugenol sensitivity. Eugenol is a recognised contact allergen, familiar from dental zinc-oxide eugenol cements.

Use Caution

• Iron deficiency or borderline ferritin. Hydrolysable tannins are potent inhibitors of non-haem iron absorption. In a British Journal of Nutrition study from 1999, 20 to 50 mg of total polyphenols reduced iron absorption from a bread meal by 50 to 70 percent, and 100 to 400 mg reduced it by 60 to 90 percent, with black tea at 79 to 94 percent and camomile at 47 percent. A strong avens root infusion sits squarely in that range.

• Anyone on chronic anticoagulation or antiplatelet therapy, on the general principle that eugenol-rich preparations affect platelet function and no interaction study exists here.

• Gastritis, reflux or an irritated gut. Concentrated tannins are locally irritating on an empty stomach.

• Anyone taking oral medication with a narrow therapeutic index. Tannins bind alkaloids and basic drugs in the gut lumen.

Critical Safety Point

The eugenol in avens is a genuine hepatotoxin at gram quantities, which is a fact about the molecule rather than about traditional tea. LiverTox records paediatric clove oil ingestions of 8 to 20 mL producing hepatic failure within 12 to 24 hours with ALT above 10,000 U/L, driven by glutathione depletion, with N-acetylcysteine used as antidote in published cases including a 2005 European Journal of Pediatrics series. A root yielding 0.04 percent volatile oil cannot reach those quantities in any realistic preparation, but concentrated essential oils sold beside the herb absolutely can, and they must never be swallowed. JECFA set an acceptable daily intake for eugenol of 0 to 2.5 mg per kg body weight, roughly 175 mg for a 70 kg adult.

Recommended Dosages

No regulator anywhere has set a dose for avens. Every figure below is a traditional or trade convention with no clinical trial behind it.

Dried root or herb, infusion or decoction. 1 to 4 g three times daily, per the British Herbal Pharmacopoeia figure carried in Barnes 2007. A common retail preparation is half to one teaspoon of crushed root steeped in cold water, brought to a boil and simmered five minutes, up to three cups a day.

Liquid extract 1:1 in 25 percent alcohol. 1 to 4 mL three times daily, the same traditional source.

Tincture 1:2 or 1:3. Practitioner sites list 2 to 4 mL of a 1:3 three times daily, or 1 to 3 mL diluted in water three times daily. Note that these ranges overlap despite representing different amounts of plant, which is a good illustration of how loose the tradition is.

Mouthwash or gargle. A cold infusion of one teaspoon of coarsely powdered root, brought briefly to the boil, steeped ten minutes and strained, used undiluted as a rinse. This is the traditional gingivitis preparation and the one with the most plausible mechanism, since tannins act at the surface and nothing needs to be absorbed.

Duration

Traditional practice caps internal tannin-rich astringents at short courses, commonly two to four weeks, and multiple herbal sources warn explicitly against excessive doses or prolonged use of wood avens. There is no long-term safety data of any kind, and no study has followed anyone taking avens for any period. If diarrhoea has not settled within 48 hours or gum inflammation within about two weeks, the answer is a clinician, not a longer course.

Timing & Administration

Take internal preparations with or just after food. A concentrated tannin dose hitting an empty stomach is the most reliable way to provoke the nausea the herb is traditionally given to settle.

Separate avens from iron supplements and from iron-rich meals by at least two hours, for the reasons given above. The same spacing applies to alkaloid drugs and to any medication you have been told to keep away from tea.

For the gargle, hold and swish for 30 to 60 seconds and expectorate. The astringent effect is a surface interaction with mucosal protein and does not require swallowing.

If you are using dried root for the eugenol character rather than the tannins, buy it whole and grind immediately before use. The volatile fraction is 0.04 percent of the root at best and it leaves as soon as the surface area increases.

Timeline of Effects

There are no trial durations to cite, because there are no trials. What follows is bounded by mechanism and by what comparable astringents have shown.

Immediate, seconds to minutes. The astringent sensation in the mouth. This is protein precipitation and it is real, measurable and instantaneous. It is also the only effect of avens anyone can verify at home.

Hours. Any luminal antidiarrhoeal effect from tannins would act within the same digestive transit as the dose, so a first pass would be visible within 6 to 12 hours. If nothing has changed within 48 hours, the traditional preparation has failed and continuing is a decision to keep taking an unstudied plant.

One to two days. The ellagitannin story shifts to the colon. Piwowarski and colleagues showed in the Journal of Ethnopharmacology in 2014 that Geum urbanum root and rhizome extract is converted by human faecal microbiota into urolithins A, B and C, and that urolithin A inhibited TNF-α production by 29.2 ± 6.4 percent at 0.625 µM in THP-1 derived macrophages, with urolithin C inhibiting IL-6 by 13.9 ± 2.2 percent at the same concentration. Urolithin appearance in humans peaks roughly a day after ellagitannin intake.

Never, for a meaningful fraction of users. Urolithin production is a metabotype. Across intervention trials, 25 to 80 percent of volunteers produced urolithin A, 10 to 50 percent produced isourolithin A or urolithin B, and 5 to 25 percent produced no detectable urolithins at all. A 2018 Food and Function analysis of 839 Caucasians aged 5 to 90 found aging the main determinant. If you are in the non-producing group, the systemic anti-inflammatory rationale for avens does not apply to you and no label will tell you which group you are in.

What counts as a fair trial. Two weeks for a mouth rinse, 48 hours for acute diarrhoea, and no fair trial at all for anything systemic, because there is no endpoint that has ever been measured in a person.

Benefits of Taking Avens

Stated at the level the evidence supports, which is low.

• A demonstrable astringent effect on oral and gut mucosa, mechanistically identical to tormentil and oak bark, both of which have EU herbal monographs and avens does not.

• Real polyphenol density. At 2.1 g of phenolics per 100 g fresh matter, avens is a richer ellagitannin source than most berries, and ellagitannins have a defensible general nutritional case as urolithin precursors in people who produce urolithins.

• A distinctive eugenol-dominant flavour with genuine culinary use, from Augsburg Ale to a clove substitute in cooking.

• Laboratory antibacterial activity that is consistent across independent groups in Bulgaria, Germany and Poland, including MICs as low as 0.078 mg/mL and biofilm inhibition above 80 percent. This is a legitimate reason for further research and not a reason to take the herb.

• A clean acute toxicology signal in the one animal study that looked. Zaharieva and colleagues, following OECD 423, found no signs of acute toxicity in liver or kidney specimens of H-albino mice at doses up to 210 mg/kg.

What is not on this list: any demonstrated effect on diarrhoea, gingivitis, ulcerative colitis, Crohn's disease, peptic ulcer, nausea, fever, haemorrhoids, gout, rheumatoid arthritis or nerve pain in a human being. Those indications are all in circulation on retail pages. None of them is supported.

Potential Negatives & Side Effects

• Nausea and gastric discomfort, the standard consequence of concentrated hydrolysable tannins taken without food.

• Constipation on prolonged use, the mirror image of the antidiarrhoeal action.

• Reduced iron status with regular use, per the 50 to 90 percent absorption inhibition figures for polyphenol-rich beverages.

• Contact or oral mucosal sensitivity to eugenol in susceptible people, including stomatitis from repeated high-eugenol exposure familiar from dental practice.

• Reported effects on the menstrual cycle, which is the specific basis for the pregnancy warning and which has never been characterised properly.

• Unknown long-term consequences. Not "probably fine." Unknown. Nobody has taken avens under observation for a defined period and reported what happened.

The honest hazard with this plant is not a dramatic adverse event. It is paying for a preparation with no assay, no marker compound and no trial, then attributing to it whatever your gut was going to do anyway.

Deficiency Symptoms

Avens is not a nutrient. The human body has no requirement for Geum urbanum, no transport protein for gein, no storage pool of eugenol and no deficiency state. Nothing goes wrong if you never encounter it.

What avens addresses

Traditional practice used it for three things: loose stools, inflamed gums and mouth, and fevers. The first two rest on tannin chemistry that is real but generic to the whole astringent Rosaceae group. The third, the febrifuge reputation, has no mechanism and no data. External traditional uses included poultices for frostbite, haemorrhoids and skin complaints, and homeopathic practice used it for bladder and urinary tract inflammation.

Bottom line

Contrast this with an actual essential nutrient, where deficiency produces a named syndrome with a measurable biomarker and a reproducible response to repletion. Iodine deficiency gives goitre and a measurable urinary iodine concentration. Avens deficiency gives nothing, because there is no such thing.

Toxicity Symptoms

At high intake

The two toxicological concerns are tannin load and eugenol. Large quantities of hydrolysable tannins produce nausea, vomiting, abdominal pain and, chronically, impaired mineral absorption. Eugenol at gram doses is directly hepatotoxic through glutathione depletion, with the documented cases coming from clove essential oil rather than from Geum preparations. Avens root at 0.04 percent volatile oil cannot deliver a hepatotoxic eugenol dose in tea or tincture form.

Signs to reduce or stop

Persistent nausea, epigastric pain, new constipation, mouth or lip irritation, or any unexpected fatigue or pallor that could reflect falling iron status. Stop immediately for jaundice, dark urine or right upper quadrant pain, and get liver enzymes checked, particularly if a concentrated oil was involved.

General note

The mouse work under OECD 423 found no liver or kidney damage at 210 mg/kg of the ethyl acetate aerial extract, and JECFA's eugenol ADI of 0 to 2.5 mg/kg body weight leaves considerable headroom relative to what a root infusion supplies. That is reassurance about a narrow question in a narrow model. It says nothing about repeated human dosing over months, which nobody has studied.

How Avens Works

Two mechanisms carry the plant, and they operate in different places.

Tannins act at surfaces, not in the bloodstream. Gallotannins and ellagitannins are large, polar and poorly absorbed intact. They cross-link proteins on contact, which tightens mucosal surfaces, reduces exudation and gives the puckering sensation. This is why the mouthwash use is the most defensible traditional application: nothing has to be absorbed for it to work. It is also why systemic claims for tannin-rich herbs usually fail, since the molecule that produces the local effect is not the molecule that reaches tissue.

The systemic story runs through gut bacteria. Ellagitannins such as gemin A and pedunculagin release ellagic acid, which colonic microbiota convert stepwise into urolithins A, B and C. Urolithins are absorbable and are the compounds that show anti-inflammatory activity in macrophage models. This step is not under the label's control and varies enormously between individuals, from a full urolithin A producer to a non-producer.

Gein is a storage form, not an active. The intact eugenol vicianoside has no reported activity of its own. Damage to root tissue and drying hydrolyse the glycosidic bond and free eugenol, which is why the smell appears when you cut the root and fades as it ages. Eugenol has local anaesthetic and antiseptic properties well characterised from dentistry, and those properties plausibly contribute to the traditional mouth uses alongside the tannins.

Everything downstream of that is laboratory work. Gemin A suppresses neutrophil CD11b, ROS, elastase and MMP-9. Extracts inhibit acetylcholinesterase, with an ethanol extract IC50 of 0.513 ± 0.03 mg/mL against 2.293 ± 0.14 mg/mL for aqueous. An extract inhibits α-synuclein fibrillation in a thioflavin T assay and partly disassembles preformed fibrils. Another inhibits α-glucosidase and DPP-4 and lowers glucose in a hen's egg model and triglycerides in starch-fed Drosophila melanogaster. Extracts push adipose-derived human mesenchymal stem cells toward cardiac markers at 50 to 100 µg/mL. A 20 percent aqueous decoction injected intravenously lowered blood pressure in cats, reported by Petkov in the American Journal of Chinese Medicine in 1979. Not one of these findings has been carried into a person.

Synergistic Supplements

Synergy claims for avens are entirely theoretical. These are the pairings that at least have a rationale.

• Vitamin C, if you are taking avens regularly and care about iron. Ascorbic acid counteracts the dose-dependent inhibition of non-haem iron absorption by polyphenols, which is a documented interaction rather than a marketing pairing.

• Other astringent Rosaceae, specifically Potentilla erecta rhizome and Agrimonia species, which share the ellagitannin chemistry and, in tormentil's case, actually hold a finalised EU herbal monograph adopted on 20 January 2011 for mild diarrhoea and minor mouth inflammation. If you want the astringent effect with regulatory scrutiny attached, tormentil is the version to buy.

• Chamomile or sage as a mouth rinse partner, both long used for oral inflammation and both better documented than avens.

• A urolithin-producing gut, which is not a supplement but is the actual prerequisite for the systemic ellagitannin rationale. Nothing on the market reliably converts a metabotype 0 individual into a producer.

Avoid stacking avens with other high-tannin products such as strong black tea, oak bark or large doses of green tea extract. The effects on mineral absorption are additive and the benefit is not.

Interactions & What NOT to Take

No formal interaction study has ever been performed with Geum urbanum. Everything here is inferred from tannin and eugenol chemistry.

• Iron salts. Ferrous sulfate, ferrous fumarate, ferrous gluconate and ferrous bisglycinate. Separate by at least two hours. The polyphenol effect on non-haem iron runs from 47 percent inhibition for a weak infusion up to 94 percent for strong black tea.

• Warfarin, clopidogrel, apixaban, rivaroxaban and aspirin. Eugenol affects platelet function, and a plant with no interaction data has no business alongside anticoagulation without a clinician's sign-off.

• Alkaloid drugs. Tannins precipitate basic nitrogenous compounds in the gut lumen. This is a general and long-recognised class effect that applies to any strong tannin preparation.

• Any narrow-therapeutic-index oral drug, including levothyroxine, digoxin and lithium. Space dosing by two hours or more.

• Other hepatotoxic exposures, most obviously clove essential oil, high-dose paracetamol or heavy alcohol, given eugenol's glutathione-depleting mechanism.

• Zinc, calcium and magnesium supplements. The same chelation logic that applies to iron applies here, with less quantification behind it.

Quality, Testing & Adulteration

This is the section where avens fails hardest, and it fails for a structural reason: no compendial standard exists to fail against.

No pharmacopoeial monograph. The European Medicines Agency's herbal medicines dataset lists 203 herbal substance entries, 184 of them with assessment finalised. Geum urbanum appears in none of them. The MHRA's list of granted traditional herbal registrations, dated 27 October 2025 and running to 34 pages, contains 155 registration numbers and not one mentions avens or Geum. The plant is not on FDA's 21 CFR 172.510 list of natural flavouring substances either, a list that does name yarrow, Achillea millefolium, "in beverages only" and sweet woodruff, Asperula odorata, "in alcoholic beverages only." The Council of Europe listed avens as a natural source of food flavouring, category N2, and that is close to the sum of its regulatory recognition.

No third-party verification programme. USP, NSF and ConsumerLab all build their testing around ingredients with reference standards and established assay methods. There is no USP monograph for avens, no reference material and no validated marker assay in general use.

What a certificate of analysis should show and usually will not. Botanical identity confirmed to species by macroscopic, microscopic or DNA method, distinguishing Geum urbanum from Geum rivale and from other Geum; plant part stated as root and rhizome or aerial parts, not "herb"; loss on drying; total ash and acid-insoluble ash; heavy metals; pesticide residues; microbial limits; and a tannin figure by a stated method, ideally the European Pharmacopoeia colorimetric assay that Neumann's group used. If a supplier cannot give you the species, the part and a tannin number with a method attached, they do not know what is in the bag.

The substitution risk is specific and easy. Geum urbanum and Geum rivale grow together, hybridise freely, are both called avens and both have astringent roots. Their volatile chemistry is entirely different: 69.2 percent eugenol in G. urbanum root oil against 53.3 percent cis-myrtanal in G. rivale. A buyer paying for clove-scented wood avens root can receive water avens root and have no way to tell from the label. There is also a straightforward sensory check available here that most herbs do not offer: real wood avens root smells of cloves. If it does not, question it.

The wider trade record. On 3 February 2015 the New York Attorney General ordered Walmart, Walgreens, GNC and Target to stop selling certain store-brand herbal supplements after 390 DNA tests on 78 samples found 79 percent failed to show the labelled plant or contained unlisted fillers, with Walmart's Spring Valley line at 4 percent and Target's Up and Up at 41 percent. Note also what happened to the study most often cited on this topic: Newmaster's 2013 DNA barcoding paper in BMC Medicine, which reported 32 percent substitution across 44 products, was retracted by the journal on 4 July 2024. Cite the attorney general's numbers, not the retracted paper.

A literature-level error worth knowing. A widely circulated 2019 review lists the sesquiterpene lactone cnicin among the constituents of Geum urbanum. Cnicin is the bitter principle of Cnicus benedictus, blessed thistle, a completely different plant in the Asteraceae. The two got confused because both carry Benedict's name, avens as herba benedicta and blessed thistle as Cnicus benedictus. A 2021 pharmacology paper in Frontiers in Pharmacology, meanwhile, printed its common names backwards, calling Geum urbanum "roseroot" and Rhodiola rosea "avens root." When the primary literature on a plant is this loosely edited, treat every downstream retail claim accordingly.

Special Considerations

Wildcrafting and sustainability. Geum urbanum is common to the point of being weedy across Europe and is naturalised in Australia and New Zealand. It is not a conservation concern, which is a genuine advantage over many wild-harvested roots. Kosmala's group noted the plant's resistance to fungal disease means commercial material tends to be free of pesticide residues.

Harvest timing changes the material. Polish work on accumulation in the plant's organs found tannin content in underground organs highest in first-year plants collected at the start of dormancy, while eugenol and nopinone levels were largely unaffected by plant age or harvest date. Regional variation is substantial: root catechin differed by roughly 50 percent between central and eastern Polish populations, 600.46 against 402.44 mg/100 g.

Anthroposophic medicine is a separate context. Geum urbanum is used in that system, often at potencies well below herbal dosing, and the 62-patient Herdecke case series sits inside that tradition. Findings from homeopathic or anthroposophic dilutions do not transfer to a 1:2 tincture and should not be quoted as if they do.

The genus is not interchangeable. Geum aleppicum and Geum japonicum var. chinense circulate in Chinese herbal commerce as lan bu zheng with a different traditional indication set. Geum rivale has its own separate WebMD monograph. Dryas octopetala, mountain avens, is a different genus with no shared chemistry. Species names on the label are not a formality here.

The retail claim gap is the thing to notice. Practitioner-facing pages describe avens as a herb that will "stimulate digestive secretions and improve gastric and intestinal function," tone the bowel wall, and support "the body's natural fever response," recommending 2 to 3 cups of tea or 2 to 4 mL of tincture daily. Read those sentences next to a PubMed clinical-trial count of zero and the distance between the two is the whole story.

Research Status & Evidence Quality

Strong Evidence For

Nothing in humans. Not one indication. The only claims that reach "strong" are chemical rather than clinical: the root oil is eugenol-dominant at 67 to 69.2 percent, the root contains hydrolysable tannins including gemin A, pedunculagin, stachyurin, stenophyllanin A, casuarinin and gemin G, and human gut microbiota convert those to urolithins.

Moderate Evidence For

In vitro antibacterial activity, replicated across independent laboratories with concordant MICs in the 0.078 to 0.156 mg/mL range for ethyl acetate fractions against Gram-positive organisms, plus biofilm inhibition of 72 to 90 percent against MRSA and 84 to 85 percent against Pseudomonas aeruginosa at sub-inhibitory concentrations. Also moderate: antioxidant capacity, with DPPH EC50 values clustered around 26 to 27 µg/mL for crude extracts across separate groups. Both are dish-level findings with strong internal consistency and no clinical translation.

Emerging / Preliminary Evidence For

Anti-inflammatory action at the neutrophil via gemin A. α-glucosidase and DPP-4 inhibition with glucose lowering in the hen's egg test and triglyceride lowering in Drosophila, from a target-based screen of 111 root extracts published in Frontiers in Pharmacology in 2021. Inhibition of α-synuclein fibrillation in vitro. Antineoplastic activity against T-24 and BC-3C bladder cancer lines at IC50 21.33 to 25.28 µg/mL and against A-375 melanoma and A-431 skin cancer lines at 6.7 to 14.68 µg/mL. Moderate antiviral activity against human adenovirus type 5 and herpes simplex virus type 1. Cardiogenic marker upregulation in human mesenchymal stem cells at 50 to 100 µg/mL. Acetylcholinesterase inhibition at IC50 0.513 mg/mL. Every one of these is cell culture, egg or insect.

Research Limitations

The core limitation is the one stated at the top: PubMed indexes 48 records for the species and none carries the Clinical Trial publication type, while ClinicalTrials.gov has no registered Geum study. Beyond that, the preclinical work is fragmented. No two papers use the same extract, so MIC values from an ethyl acetate fraction cannot be applied to a hydroethanolic tincture. Effective in vitro concentrations for the quorum-sensing work sit at 1.56 to 6.25 mg/mL, which is a concentration no oral dose achieves in tissue. There is one animal toxicology study, one 1979 cat blood pressure report and no pharmacokinetics in humans at all, meaning nobody knows what plasma concentration of anything an avens dose produces. The regulatory picture confirms the gap rather than contradicting it: under Directive 2004/24/EC, adopted 31 March 2004 with a transition period running to 30 April 2011, a traditional herbal registration requires only 30 years of medicinal use, 15 of them within the EU, and efficacy that is "plausible on the basis of long-standing use and experience." No clinical trial is needed. Avens has not been granted even that.

Summary & Key Takeaways

Avens is a real plant with real chemistry and an evidence file that is close to empty. The root oil is 67 to 69.2 percent eugenol, released from the glycoside gein when the root is cut or dried. The root carries hydrolysable tannins somewhere between 2 and 10.5 percent by validated assay, not the 30 percent that retail copy claims. Those tannins do what tannins do, tightening mucosal surfaces on contact, and that is the one effect a person can verify. Beyond it, the file consists of radical scavenging assays, agar plate MICs, neutrophil chemiluminescence, a hen's egg, some fruit flies and one intravenous cat experiment from 1979.

Bottom Line

If you want the astringent effect that avens is sold for, buy tormentil instead. Potentilla erecta rhizome has the same ellagitannin chemistry and a finalised EU herbal monograph adopted on 20 January 2011 covering mild diarrhoea and minor mouth inflammation, which means a regulator has at least looked at the safety file. Avens has no monograph among the EMA's 203 herbal substance entries, no place among the 155 UK traditional herbal registrations current on 27 October 2025, no entry in FDA's flavouring lists, no USP standard and no clinical trial. Keep it as a foraged flavouring and a mouth rinse. Do not buy it as a treatment.

Key Safety Points

• Avoid in pregnancy and while breastfeeding. The pregnancy warning is based on reported menstrual effects and the absence of any reproductive toxicology.

• Separate from iron by at least two hours. Polyphenol-rich preparations cut non-haem iron absorption by 50 to 90 percent.

• Never swallow clove essential oil as a substitute or adjunct. Paediatric ingestions of 8 to 20 mL have produced hepatic failure with ALT above 10,000 U/L.

• Short courses only, and no internal use beyond two to four weeks. There is no long-term data of any kind.

• Confirm the species. Wood avens root smells of cloves. Water avens root does not.

Special Note

There is a pattern here that repeats across the whole traditional-herb shelf. A plant with a striking sensory signature, in this case a root that smells of cloves, accumulates centuries of use and a long list of indications, and the sensory signature gets silently treated as evidence that the indications are real. They are separate facts. The eugenol is verifiable in a gas chromatograph. The claim that avens settles ulcerative colitis is verifiable nowhere. When a herb's marketing leans on how distinctive it is, check whether anyone ever tested what it does.