Agrimony for Diarrhoea: Astringency and Tannins

Agrimony (Agrimonia eupatoria) has one defining property, and almost everything claimed for the herb is either an expression of that property or an overreach beyond it. The property is astringency: the dry, puckering, tissue-tightening sensation you get from strong black tea, an unripe persimmon, or a mouthful of red wine tannin. It is the reason agrimony has been the European standby for mild diarrhoea for centuries, and it is also the reason the herb's plausible reach is narrower than the traditional literature implies.

This page treats astringency as the spine of the argument, because doing so lets us be genuinely useful in both directions. Where the mechanism supports a claim, we say so. Where the claim needs a mechanism astringency cannot supply — systemic anti-inflammatory action, liver support, treating an infection — the tannin story stops rather than stretches. And where the honest answer is that nobody has run the experiment, we write that down as a finding rather than filling the gap with tradition.

Table of Contents

  1. What Astringency Actually Is — and Is Not
  2. The Tannins in Agrimony, Including Agrimoniin
  3. What the Diarrhoea Tradition Actually Claims
  4. What a Traditional-Use Registration Means
  5. The Trials That Do Not Exist
  6. Tormentil: What a Real Astringent Trial Looks Like
  7. What Does Have Trials for Acute Diarrhoea
  8. Which Agrimony? A. eupatoria versus A. pilosa
  9. Where the Mechanism Runs Out
  10. Making an Infusion, Honestly
  11. Three Contradictions Worth Naming
  12. The Honest Bottom Line
  13. Key Research Papers
  14. Connections

What Astringency Actually Is — and Is Not

Astringency is not a taste and it is not a drug effect. It is a physical, mechanical event on a wet surface. Tannins are large polyphenol molecules with many hydroxyl groups arranged around a rigid core, and that architecture makes them promiscuous protein binders. When a tannin-rich liquid washes over the mucous membrane of your mouth or the lining of your gut, the tannins cross-link with proteins in the surface mucus and with the outermost proteins of the epithelial cells. Those complexes precipitate out of solution. What you feel as puckering is the mucus layer losing its slipperiness as its proteins are pulled together.

Three consequences follow, and they set the ceiling on what agrimony can do.

First, the action is local. It happens where the tannin touches. Large hydrolysable tannins are poorly absorbed from the gut — they are simply too big and too avidly bound to cross the intestinal wall intact. So a tannin's effects live in the lumen and on the mucosa, and effectively nowhere else. This is the single most clarifying fact about tannin-rich herbs: it makes the mucosal claims the strongest ones and the systemic claims the weakest, regardless of which are more popular.

Second, the action is dose-proportional and immediate. There is no receptor to saturate, no enzyme to induce, no build-up over weeks. A strong infusion is more astringent than a weak one, and the effect lasts as long as the tannin coating survives before being washed away or digested. Anyone who tells you agrimony needs six weeks to work has left the mechanism behind.

Third, the action is not selective. Tannins do not distinguish between the proteins you would like them to bind and the ones you would not. The same chemistry that firms up an irritated gut surface will bind dietary iron, bind some drug molecules, and — in quantity — irritate a stomach it was meant to settle. This is the pattern the doctrine behind these pages calls benefit and hazard as one property described twice, and agrimony is a textbook case. The safety article is not a list of unrelated cautions bolted onto this page — it is the same paragraph read from the other side.

If this shape feels familiar, it should. It is exactly the argument made for demulcent herbs such as marshmallow root and slippery elm, whose mucilage coats a mucous membrane physically rather than pharmacologically. Astringents and demulcents are chemical opposites — one tightens the surface, one lubricates it — but they are epistemically identical. Both are physical interventions on a surface, both are plausible for surface complaints, and both are almost entirely untested in controlled trials because there is no patentable molecule at the end of the story.

The Tannins in Agrimony, Including Agrimoniin

Agrimony's aerial parts — the leafy, flowering tops, dried — are dominated by polyphenols. The best-characterised of them is agrimoniin, an ellagitannin (a hydrolysable tannin built on ellagic acid units) that is distinctive enough to serve as a chemical signature for the genus. Alongside it sit other hydrolysable tannins, condensed tannins, and a substantial flavonoid fraction: quercetin, luteolin, apigenin and kaempferol derivatives, plus phenolic acids.

The most useful chemical work here is analytical rather than pharmacological. A 2013 paper in the Journal of Pharmaceutical and Biomedical Analysis by Granica and colleagues developed and validated an HPLC method specifically to standardise the polyphenols in Agrimoniae eupatoriae herba — which tells you something important in itself. Analytical chemists build standardisation methods for materials whose content varies. Agrimony's polyphenol load is not a fixed number; it moves with harvest timing, plant part, drying, and geography. Earlier profiling by Correia and colleagues in Biomedical Chromatography (2006) mapped the same flavonoid and phenolic landscape.

We are deliberately not printing a tannin percentage or a milligrams-per-cup figure on this page. Secondary herbal sources quote such numbers freely and they do not agree with each other, and the extraction step adds another unquantified variable: how much of the tannin in 2 g of dried herb actually leaves the leaf and enters your cup depends on water temperature, steep time and agitation. A confident figure would be a guess dressed up as data. What we can say without inventing anything is directional and still useful: a longer, hotter steep of more herb yields a more astringent, more tannin-loaded cup, and you can verify that yourself on your tongue in about four minutes. Taste is a genuinely calibrated instrument for this one property.

What the Diarrhoea Tradition Actually Claims

Read the older European herbals carefully and the diarrhoea claim is narrower than modern marketing suggests. The traditional indication is mild, acute, ordinary diarrhoea — the self-limiting kind that follows a dubious meal or accompanies a viral illness. Agrimony infusion appears as a comfort measure taken by the cupful over a day or two, expected to firm the stool, not to cure an infection.

That is a coherent claim on the mechanism. An irritated, hypersecretory gut surface with a compromised mucus layer is precisely the setting where a protein-precipitating coating could plausibly do something: reduce weeping from an inflamed surface, blunt the sensory irritation driving urgency, and add a little structure to liquid stool. Nothing in that requires the tannin to be absorbed, act on a receptor, or reach a target organ.

The tradition also travels. In Traditional Chinese Medicine the related East Asian species Agrimonia pilosa, known as Xian He Cao, is classified among the haemostatic or "stop bleeding" herbs and is used for chronic diarrhoea and dysentery — the same astringent logic reaching a similar indication independently. Convergent traditional use across unconnected medical systems is a real signal about a plant's sensory properties, and a weak one about clinical outcomes. It tells you that people in many places noticed the same puckering effect. It does not tell you that anyone got better faster.

What a Traditional-Use Registration Means

You will see it asserted that agrimony is "approved in Europe" for diarrhoea and for sore throat. That is broadly accurate and almost universally misread, so it is worth explaining the instrument rather than quoting it.

European herbal medicines can reach the market by more than one route. The demanding route requires the same efficacy evidence any other medicine needs: adequate, well-controlled clinical trials. The other route is traditional-use registration, and its entry requirements are of a completely different kind. A product qualifies on the strength of documented use over a long period — conventionally at least thirty years, including a stretch within the European Union — together with a plausible pharmacological rationale and, crucially, an acceptable safety record at the traditional dose. Assessors read the historical literature, examine the chemistry, and satisfy themselves that the preparation is not dangerous and that its claimed action is not absurd.

What such a registration explicitly does not require is proof that the product works. That is not a loophole or an oversight; it is the stated design of the category. The label wording follows from it, and the wording is the giveaway: traditional-use products are labelled as traditional herbal medicinal products used for a purpose, based exclusively on long-standing use. Regulators wrote that sentence precisely so the registration could not be mistaken for an efficacy finding.

So the correct reading is: Europe has certified that agrimony infusion has been used for mild diarrhoea and as a gargle for a very long time, that this use is chemically plausible, and that the herb appears acceptably safe at those doses. That is genuinely worth knowing — it is a real regulatory assessment of a real safety record, not a rubber stamp. It is simply not evidence of benefit, and anyone citing it as such has substituted one kind of claim for another. The same instrument, applied to other herbs, produces apparent paradoxes that dissolve once you understand it: a regulator can grant a traditional-use registration to a plant that another regulator has found not generally recognised as effective, and both bodies can be entirely correct, because they answered different questions.

The Trials That Do Not Exist

Here is the part most agrimony pages leave out. Setting out the missing evidence as an explicit list turns "we do not know" into a stated result rather than a silence the reader has to notice for themselves.

  1. No adequate controlled human trial of agrimony for acute diarrhoea. Searching the clinical literature for a randomised comparison of agrimony infusion against placebo or against standard care, with stool frequency or duration of illness as an endpoint, returns nothing of the kind. This is the central fact about agrimony and diarrhoea, and it is a finding, not a gap.
  2. No dose-finding study. There is no work establishing how much dried herb, steeped how long, produces a measurable mucosal effect — which means every dose recommendation in circulation, including the traditional teaspoon-or-two per cup, is inherited rather than derived.
  3. No human pharmacokinetic data worth the name. We reason that agrimony's hydrolysable tannins stay in the lumen because that is how large hydrolysable tannins generally behave, not because anyone has measured agrimoniin in human plasma after a cup of tea.
  4. No quantified interaction study. That agrimony's tannins inhibit non-haem iron absorption is a confident inference from a large body of human work on tea and other polyphenol beverages. How much a cup of agrimony inhibits it has never been measured. See the safety article, where this matters most.
  5. No paediatric data at all. Which is the point at which the absence stops being academic, because childhood diarrhoea is where diarrhoea actually kills people.
  6. No formal toxicology at repeated high doses. The safety record supporting the traditional-use registration is a record of use, not a dossier of studies. That is meaningful reassurance for ordinary tea amounts and no reassurance whatsoever for concentrated extracts taken daily for months.

Note the shape of these statements. Each one says a claim is unevaluated, not that it has been refuted. Those are different epistemic states and conflating them misleads in both directions — it lets enthusiasts read absence as vindication and lets sceptics read it as disproof. Agrimony for mild diarrhoea is untested. That is all.

Tormentil: What a Real Astringent Trial Looks Like

The most useful thing on this page is not about agrimony at all. It is a demonstration that the untested state of agrimony is not because nobody trials traditional astringents.

Tormentil root (Potentilla erecta) is agrimony's close analogue: another tannin-rich member of the rose family, another European folk remedy for diarrhoea, working by the same protein-precipitating mechanism. And tormentil has a randomised, double-blind, controlled trial. Subbotina and colleagues tested a tormentil root extract in children with rotavirus diarrhoea and published the results in the Pediatric Infectious Disease Journal in 2003, reporting a shortened duration of diarrhoea in the treated group.

Two conclusions follow, and they pull in opposite directions, which is why both need stating.

The encouraging one: the mechanism is not inherently untestable. A tannin astringent can be put in a double-blind trial against a hard endpoint in the population that matters most, and when that was done the result was not null. That makes the traditional astringent story more credible in general.

The disciplining one: this is borrowed evidence, and it is not agrimony's. Different genus, different tannin profile, different extract, different preparation, single small trial. Under the borrowed-evidence rules these pages follow, species substitution is the most common way a herb page overstates itself, and it does not become acceptable because the species are related and the mechanism is shared. The honest formulation is: tormentil shows the hypothesis is testable and once tested positive; agrimony remains untested. If you want a tannin astringent with actual trial support for childhood diarrhoea, the answer today is tormentil, not agrimony — and a single trial is a thin foundation even there.

What Does Have Trials for Acute Diarrhoea

Naming the interventions that do have evidence is more respectful of the reader than any warning, because it converts a discouraging finding into a decision they can act on. It also pre-empts the reflexive defence that trials are simply unavailable in this territory. For acute diarrhoea they are abundant.

Against that background, agrimony's position is clear and not shameful: it is a traditional comfort measure with a plausible local mechanism, sitting outside the evidence base rather than in a weak position within it.

Which Agrimony? A. eupatoria versus A. pilosa

A large share of the research literature filed under "agrimony" is not about the plant in a European teabag, and this quietly inflates the herb's apparent evidence base.

Agrimonia eupatoria is the European species — church steeples, sticklewort — and it is what a British or continental herbal infusion contains. Agrimonia pilosa is the East Asian species used as Xian He Cao in Chinese medicine. They are related, share the ellagitannin chemistry, and are sometimes treated as interchangeable in secondary sources. They are not the same plant, their traditional indications differ in emphasis (haemostasis is far more central for A. pilosa), and a good deal of the more exciting pharmacology — particularly the older work on agrimoniin's immunostimulant and antitumour activity in mice — was done on A. pilosa material.

When you see "agrimony has been shown to have antitumour activity," check the species before anything else. A finding in A. pilosa is evidence about A. pilosa. The European species has its own literature and it is smaller. A third species, Agrimonia procera (fragrant agrimony), turns up as a substitute in commerce; radical-scavenging work by Venskutonis and colleagues in Fitoterapia (2007) examined both A. eupatoria and A. procera and found scavenging activity in each, so the substitution is probably chemically benign — but it is still a substitution, and if you cannot tell which species you bought, you cannot tell which literature applies to it.

Where the Mechanism Runs Out

Astringency explains the diarrhoea and gargle traditions. It does not stretch to cover everything agrimony is sold for, and the honest move is to say where it stops.

Systemic antioxidant benefit. Agrimony extracts are genuinely powerful radical scavengers in the test tube — Correia, Batista and Dinis reported activity against reactive species in BioFactors (2007), and Ivanova and colleagues ranked agrimony among high-polyphenol Bulgarian medicinal plants in the Journal of Ethnopharmacology (2005). But a tube is not a bloodstream. The very poor absorption that makes the mucosal argument strong makes the systemic antioxidant argument weak: a compound that stays in the gut lumen cannot scavenge radicals in your tissues. Antioxidant assay results are a statement about chemistry, not about physiology.

Antimicrobial action against the cause of your diarrhoea. Extracts do inhibit bacteria in vitro. Muruzović and colleagues documented antioxidant, antimicrobial and antibiofilm activity in the Journal of Food and Drug Analysis (2016); Cwikla and colleagues tested phytotherapeutics including agrimony against Helicobacter pylori and Campylobacter jejuni in Phytotherapy Research (2010). This is more interesting than most in-vitro antimicrobial work because a gut pathogen is one of the few targets a poorly absorbed compound can actually reach at concentration — the lumen is the one compartment where an infusion's real concentration might approach a laboratory one. It is still a long way from showing that drinking agrimony clears an infection, and the tannins' non-selectivity cuts here too: protein-precipitating compounds inhibit a great many things in a broth, including things you would rather keep.

Liver and urinary "tonic" use. These are traditional and require systemic delivery to make sense. The mechanism does not supply it. Treat them as history.

Blood sugar. This one is different, and it belongs in the safety article rather than the benefits column. Gray and Flatt reported effects of agrimony on hyperglycaemia, cellular glucose metabolism and insulin secretion in the British Journal of Nutrition (1998), following earlier work by Swanston-Flatt and colleagues in Diabetologia (1990) on traditional plant treatments in diabetic mice. Animal hypoglycaemic activity is not a human benefit claim — but it is a reason for anyone on glucose-lowering medication to pay attention, which is exactly how we treat it.

Making an Infusion, Honestly

If you are going to drink agrimony for a mild upset stomach, here is the practical picture with its uncertainties intact.

The form is an infusion of the dried aerial parts. Traditional practice is roughly a teaspoon or two of dried herb per cup of freshly boiled water, covered, for several minutes. We are reporting that as inherited practice, not as a validated dose — nobody has run the study that would justify a number, and item 2 of the findings list above is exactly this problem. A longer steep extracts more tannin, so steep time is your intensity control, and your tongue is the readout.

A few things follow from the mechanism rather than from tradition, and they are the parts most worth acting on:

Do not use an astringent instead of medical care when diarrhoea is bloody, is accompanied by high fever, has lasted more than a couple of days in an adult, occurs in an infant or small child, occurs in an older or frail person, or comes with signs of dehydration. Fluid replacement is not optional and an astringent is not a substitute for it. Persistent diarrhoea has causes that need identifying — see chronic diarrhoea and irritable bowel syndrome for what that workup involves.

Three Contradictions Worth Naming

Unresolved tensions are more informative than a tidy summary, so here they are rather than smoothed over.

1. Haemostatic or anticoagulant? The Chinese tradition classifies agrimony among the herbs that stop bleeding. Western caution lists warn about additive effects with anticoagulant and antiplatelet drugs — that is, about increased bleeding risk. These point in opposite directions and both are in circulation. Candidate reconciliations: the two claims may concern different species (A. pilosa for the haemostatic tradition); a local, surface, contact effect on small vessels in an open wound is a genuinely different phenomenon from a systemic effect on coagulation; and the Western warning may be a category inference from "polyphenol-rich plant" rather than an observation about agrimony specifically. We cannot resolve it from the available evidence and we are not going to pretend otherwise. The same contradiction has been noted independently for yarrow, which is also claimed both to arrest bleeding and to interact with antiplatelet drugs.

2. Astringent for the gut, irritant to the gut. Agrimony is recommended to settle an upset stomach, and tannins in quantity cause nausea, gastric irritation and constipation. This is not really a contradiction so much as a dose curve with no measured axis — but because the axis is unmeasured, nobody can tell you where the crossover is, and "more if it is not working" is precisely the wrong instinct.

3. Iron in the herb, iron blocked by the herb. Secondary sources sometimes list minerals among agrimony's constituents as a nutritional plus. Whatever iron a 2 g infusion contributes is trivially small, and the same cup's tannins inhibit non-haem iron absorption from everything else you consume near it. Net, an astringent infusion is plausibly negative for iron status, not positive. Exactly the same arithmetic applies to red raspberry leaf and to haritaki, and it is one of the more consequential errors in popular herbal writing.

The Honest Bottom Line

Agrimony's astringency is real, mechanistically well understood, and immediately verifiable on your own tongue. It makes mild acute diarrhoea the herb's most plausible internal use, and it is why European regulators were willing to register the traditional use on grounds of documented long use plus chemical plausibility. That registration is a statement about tradition and safety, not about efficacy, and it should never be quoted as though it were.

No adequate controlled human trial of agrimony for diarrhoea exists. The closely analogous tannin astringent tormentil has one, which shows the question is answerable and has not been asked of agrimony. Meanwhile the interventions that actually determine how acute diarrhoea goes — fluids and electrolytes above all — have a deep trial literature and no glamour at all.

Used as intended — a cup or two of infusion for a day or two of ordinary gut upset, taken away from meals and medicines, in an adult who is otherwise well — agrimony is a reasonable traditional comfort measure with a coherent rationale and an acceptable safety record. Used as a treatment, or used heavily and continuously, it leaves the evidence entirely and takes the tannin cautions with it.

Back to Table of Contents

Key Research Papers

Citations below name the paper, journal and year in the text and link a PubMed search rather than a record number, so you can see the full result set and its context rather than a single identifier that may or may not be the right one.

  1. Granica S and colleagues (2013) — development and validation of an HPLC method for qualitative and quantitative standardisation of polyphenols in Agrimoniae eupatoriae herba, Journal of Pharmaceutical and Biomedical Analysis. The reason we do not quote a fixed tannin percentage. PubMed search
  2. Correia HS and colleagues (2006) — polyphenolic profile characterisation of Agrimonia eupatoria by HPLC with different detection devices, Biomedical Chromatography. PubMed search
  3. Correia H, Batista MT, Dinis TC (2007) — activity of an extract and fraction of Agrimonia eupatoria against reactive species, BioFactors. In-vitro antioxidant chemistry; note the absorption limit discussed above. PubMed search
  4. Ivanova D and colleagues (2005) — polyphenols and antioxidant capacity of Bulgarian medicinal plants, Journal of Ethnopharmacology, which ranks agrimony among the high-polyphenol traditional herbs. PubMed search
  5. Venskutonis PR and colleagues (2007) — radical scavenging capacity of Agrimonia eupatoria and Agrimonia procera, Fitoterapia. Relevant to the species-substitution question. PubMed search
  6. Muruzović MŽ and colleagues (2016) — extracts of Agrimonia eupatoria evaluated for antioxidant, antimicrobial and antibiofilm activity, Journal of Food and Drug Analysis. PubMed search
  7. Cwikla C and colleagues (2010) — antibacterial activities of phytotherapeutics against Helicobacter pylori and Campylobacter jejuni, Phytotherapy Research. The one antimicrobial target a poorly absorbed tannin can plausibly reach. PubMed search
  8. Subbotina MD and colleagues (2003) — randomised, double-blind, controlled trial of tormentil root extract in children with rotavirus diarrhoea, Pediatric Infectious Disease Journal. Borrowed evidence: Potentilla erecta, not agrimony. PubMed search
  9. Lazzerini M, Wanzira H — oral zinc for treating diarrhoea in children, Cochrane Database of Systematic Reviews. A comparator with real trials. PubMed search
  10. Collinson S and colleagues (2020) — probiotics for treating acute infectious diarrhoea, Cochrane Database of Systematic Reviews; notable for downgrading earlier effect estimates as trial quality improved. PubMed search
  11. Reduced-osmolarity oral rehydration solution for dehydration caused by acute diarrhoea in children, Cochrane Database of Systematic Reviews. The intervention that addresses the danger rather than the symptom. PubMed search
  12. Gray AM, Flatt PR (1998) — actions of the traditional antidiabetic plant agrimony on hyperglycaemia, cellular glucose metabolism and insulin secretion, British Journal of Nutrition; and Swanston-Flatt SK and colleagues (1990), traditional plant treatments for diabetes in normal and streptozotocin-diabetic mice, Diabetologia. Animal data, read here as a caution rather than a claim. PubMed search
  13. Paluch Z and colleagues (2020) — a review of the therapeutic effects of Agrimonia eupatoria, Physiological Research. The most convenient single overview; read it noting how much of the cited pharmacology is preclinical. PubMed search
  14. Topic searches — Agrimonia eupatoria · agrimoniin · Agrimonia pilosa · astringent tannins in diarrhoea · hydrolysable tannin bioavailability

Back to Table of Contents

Connections

Back to Table of Contents