Tannins

Structurally the group splits in two, and the split predicts both behaviour and safety:

Hydrolysable tannins

a sugar core esterified with gallic or ellagic acid, cleaved by acid, alkali or enzymes. Gallotannins and ellagitannins.

Condensed tannins (proanthocyanidins)

flavan-3-ol polymers joined by carbon–carbon bonds, and not hydrolysable.

Hydrolysable tannins release small, absorbable phenolics on breakdown, which is where most of their systemic activity and most of their toxicity risk comes from. Condensed tannins largely stay in the gut, acting on surfaces rather than entering circulation.

A practical consequence: when a herb is described as tannin-rich, the useful question is which family. Barbatimao bark and purple prairie clover are condensed-tannin plants and behave as contact agents; red-osier dogwood and pomegranate are hydrolysable-tannin plants with a systemic dimension and a real ceiling on prolonged high intake.

Within the condensed tannins there is a further distinction that matters more than its obscurity suggests. B-type proanthocyanidins — the ordinary kind, and the great majority — are joined by a single carbon–carbon bond. A-type proanthocyanidins carry an additional ether bridge, giving a more rigid, doubly linked structure.

The distinction earned its reputation through cranberry, where A-type proanthocyanidins are credited with interfering with the adhesion of uropathogenic bacteria to the urinary epithelium rather than killing them. That is a genuinely different mode of action from astringency, and it does not follow from tannin content in general.

It is also where careful reading is required. Salal berries have been shown to contain procyanidin A2, and the presence of that molecule is solid analytical data — but demonstrating that a plant contains an A-type proanthocyanidin is not the same as demonstrating anti-adhesion activity, which has not been tested for salal. The compound is real; the extrapolation is not evidence.

Barks, galls, unripe fruit and astringent leaves — with some conspicuous concentrations:

Rosaceae is worth naming as a family habit: the astringent members run on ellagitannins, and in silverweed those tannins are unusually well characterised — agrimoniin and potentillin identified by name rather than reported as an unresolved tannin fraction. Most herbs described as tannin-rich have never had their tannins individually identified at all.

Astringency is the therapeutic action, and it is a physical one. On inflamed or weeping tissue, tannins precipitate surface proteins into a thin protective layer that reduces exudate, seals minor damage and calms irritation. That single mechanism underlies oak-bark and witch-hazel washes, tannin-rich gargles for sore throat, agrimony and tormentil for diarrhoea, and the barbatimao bark preparations used across Brazil for wounds, ulcers and burns.

Because the effect is contact-dependent, it is strongest exactly where the herb touches tissue — skin, gum, throat, gut lumen. This is why the astringent herbs have such a consistent record topically and such a thin one systemically, and why a tannin result obtained on a wound should never be read as a claim about circulation.

It also sets a natural limit on the evidence available. Astringency is so well established as a physical phenomenon that it is rarely the subject of a controlled trial, which leaves many astringent herbs in an awkward position: a coherent, well-understood mechanism and a long traditional record, with no efficacy study behind either. Red-osier dogwood is a clean example — its antidiarrhoeal use rests on confirmed high hydrolysable-tannin content rather than on any trial of the bark.

Tannins have a second, less familiar action: they are directly antibacterial, and the mechanism is an extension of protein binding. By precipitating proteins at the bacterial surface they increase outer-membrane permeability, disrupt the membrane and cause cells to aggregate.

The best-worked example comes from an unexpected direction — agricultural research on forage legumes. Condensed tannins from purple prairie clover act bacteriostatically against Escherichia coli O157:H7 at low microgram concentrations, and comparison with sainfoin tannins showed that greater protein-precipitating capacity tracked four-to-six-fold lower minimum inhibitory concentrations. The physical property and the antibacterial potency move together, which is a stronger form of evidence than either alone.

It carried into live animals: over a three-year grazing study, cattle on pastures containing purple prairie clover shed significantly less faecal E. coli than controls. Red-osier dogwood polyphenols show the same pattern across feeding trials in pigs and poultry. This is in-vivo evidence at a scale most herbal research never reaches — obtained because the funding was aimed at livestock gut health rather than human medicine, and it should be read as veterinary and food-safety work rather than as clinical data.

Intact tannins are, for the most part, too large to be absorbed. This is easy to lose sight of, because so much is written about the antioxidant capacity of tannin-rich plants — a property measured in a test tube, where absorption is not a consideration.

The ellagitannins are the important exception, and the route is indirect. Gut bacteria convert them to urolithins, and it is the urolithins — small, absorbable molecules the plant never made — that account for most of the systemic activity attributed to pomegranate and to other ellagitannin sources. The active agent is a metabolite, not a constituent.

Two consequences follow. First, urolithin production varies substantially between individuals depending on their gut microbiota, so the same dose does not deliver the same exposure to different people. Second, a high measured tannin content predicts astringency reliably and systemic effect poorly. When a herb monograph reports strong antioxidant assay results for a tannin-rich extract, that finding describes the extract in a cuvette, and the step from there to the bloodstream is the one that usually fails.

Tannins are water-soluble, which makes the traditional forms — decoctions, infusions, washes and gargles — chemically appropriate rather than merely customary. Hot water extracts them efficiently, and for barks a simmered decoction rather than a steeped infusion is the form that matters.

Freshness can invert the effect. Red-osier dogwood bark is traditionally purgative when fresh, and drying removes this — which is why traditional preparation specifies dried bark, and why the distinction is a preparation instruction rather than a piece of folklore. Where a tradition is specific about the state of the plant material, it is usually specific for a reason.

Standardisation is the other practical issue. The human and veterinary evidence for barbatimao is for topical bark extracts standardised to tannin content, and those results do not transfer to an unstandardised oral decoction. Across this class, tannin content varies enormously with species, plant part, season and preparation, so a dose expressed as grams of dried herb carries much less information than it appears to.

Reference

Tannins — Reference Notes