Milk thistle quality: how to recognise it

You cannot see the quality of a milk thistle preparation — a powder of ground herb looks exactly like a high-grade fruit extract once it is inside a capsule. What can be checked are four statements on the pack: which plant part was processed, how the extract is described (ratio and solvent), how much silymarin per daily portion it contains, and whether the manufacturer can produce a batch analysis. If one of these is missing, what is missing is not just a number but the very possibility of comparing the product with another one.

That this is no theoretical question with milk thistle has been shown by more than one market survey. Let us start there.

What market surveys found

A group around Débora Frommenwiler examined 31 milk thistle products from the British market in 2022 in Frontiers in Pharmacology (13:925298) — tablets, capsules, tinctures and liquid extracts, bought online, in shops and in pharmacies. Testing was done by high-performance thin-layer chromatography (HPTLC) following the method the European Pharmacopoeia prescribes for milk thistle fruit.

Regulatory category	Examined	Fingerprint unremarkable	Conspicuous
Traditional herbal medicinal product	10	10	0
Food supplement	21	10	11 (52.4 %)

The eleven conspicuous samples broke down as follows. In nine of them the zone of the flavonoid taxifolin was missing, which the identity test of the pharmacopoeia monograph requires — the silymarin zones were there, the rest of the fruit pattern was not. In two samples every zone typical of milk thistle fruit or its extract was absent; the authors read that as evidence that the plant was not present in those products. Eight samples additionally showed a chlorophyll zone, of the kind typical of green plant parts, which does not occur in the fruit or in the fruit extract.

A second study, Fenclová and colleagues in 2019 in Scientific Reports (9:11118), chemically measured 26 milk thistle food supplements from the US and Czech markets. The result: the measured flavonolignan content frequently deviated from the declaration — in four preparations it was below half of the stated silymarin value. Between different batches from the same manufacturer, the content differed by up to 277 mg per recommended daily dose. Mycotoxins, pesticide residues and microbiological load were found in all preparations tested.

Both studies concern other markets and cannot be transferred one to one to German products. What they are good for is the question of what to look at when comparing — and that they answer very clearly.

The plant part: fruit, not herb

The European Pharmacopoeia describes as starting material the ripe fruit without the pappus of Silybum marianum, with at least 1.5 % silymarin, calculated as silibinin (Ph. Eur. 01/2014:1860). Not the leaves, not the herb. The difference is no detail: the flavonolignans sit in the fruit, whereas the herb shows only two very faint green zones in the fingerprint and a strong chlorophyll band instead.

German packs usually say “milk thistle seeds” — botanically these are fruits, not seeds. Why the terms diverge is explained in Milk thistle: plant, origin and use. All that matters here is this: if the ingredient list says “milk thistle herb”, or simply “milk thistle” without a plant part, it is left open what was processed.

Identity: what a chromatogram shows

The identity test according to the pharmacopoeia is a thin-layer chromatography (Ph. Eur. 2.8.25). An extract of the sample is applied; in the developed fingerprint, four green zones appear after staining, three of which go back to silybin, silydianin and silychristin, plus an orange zone from taxifolin. Ground fruit additionally shows a blue zone that is absent in the extract — which is how powder can be told apart from extract.

This is not an elaborate procedure. The authors of the British study put the cost of such an analysis at around 21 US dollars per sample if only a single sample runs on a plate; with several samples per plate the figure drops considerably. A manufacturer who does not test the identity of the raw material saves little by it.

What has to be on the label — and what does not

The German Food Supplements Regulation (NemV) sets out in § 4 what has to appear on the pack: the designation “food supplement”, the names of the characterising categories of substances, the recommended daily intake in portions, the warning that this amount must not be exceeded, the note on a balanced diet, and the note to store the product out of the reach of young children. On top of that, the amount of the active substances per stated daily portion — expressly as an average value based on the manufacturer’s analysis of the product.

What is not required there is the more interesting list:

Statement	Mandatory?	What it tells you
Amount of silymarin per daily portion	yes (as an average from the manufacturer’s own analysis)	the only directly comparable value
Plant part (fruit/herb)	no	whether the flavonolignan-rich material was processed at all
Drug-extract ratio	no	how much raw material is in one part of extract
Solvent of the extraction	no	which substance pattern is to be expected
Measurement method (HPLC or photometric)	no	whether two milligram figures are comparable
Batch number with certificate of analysis	no (traceability yes)	whether the batch was tested

The lower half of this table is exactly what distinguishes one product from the next — and exactly what is voluntary. A manufacturer who states it makes itself comparable. How extract ratio and percentage figure relate to each other, and why photometrically determined values come out systematically higher than values measured by HPLC, is covered in Milk thistle 20:1 or 80 %: what the label means.

Residues: where limits apply and where they do not

For food supplements, the maximum levels of Regulation (EU) 2023/915 on contaminants in food apply. For herbal preparations, these entries are the relevant ones:

Substance	Maximum level for food supplements	Reference
Lead	3.0 mg/kg	Annex I, 3.1.28
Cadmium	1.0 mg/kg	Annex I, 3.2.21.1
Mercury	0.10 mg/kg	Annex I, 3.3.2
Pyrrolizidine alkaloids (sum)	400 µg/kg	Annex I, 2.4.10
Benzo(a)pyrene	10.0 µg/kg	Annex I, 5.1.16
Sum of four PAHs	50.0 µg/kg	Annex I, 5.1.16

The gap is conspicuous: for mycotoxins, the regulation contains no separate entry for food supplements containing botanical preparations. Aflatoxins, ochratoxin A and the Fusarium toxins are regulated for nuts, cereals, dried fruit and dried spices, not for botanical preparations — as the Czech group also points out. Their findings lie precisely in that gap.

How large the problem actually is is shown by a study from the US Food and Drug Administration (Tournas et al., Food Additives & Contaminants 2012; 29:994–999): in 83 milk thistle samples from the US market, aflatoxins were detected in 19 %, at levels from 0.04 to 2.0 µg/kg. These are low values — they lie below what would be permissible for oilseeds sold to the end consumer. The finding is therefore less an alarm signal than an indication that the raw material has a mould issue: the milk thistle fruit contains 20 to 30 % fatty oil, and how it is dried and stored after harvest is part of what decides what ends up in the extract later.

What a certificate of analysis should contain

A certificate of analysis (CoA) refers to a particular batch and usually names: identity of the raw material, silymarin content with the method used, heavy metals, pesticide residues, microbiological purity and — for extracts — residual solvents. The last point has a concrete background: for the standardised milk thistle dry extract, the pharmacopoeia (Ph. Eur. 01/2019:2071) permits ethyl acetate, acetone, ethanol and methanol, each also as a mixture with water. Which of these was used determines which residues have to be tested for at all.

A CoA is not mandatory, and no manufacturer has to hand one out unasked. Asking for it is nevertheless the most informative question you can put — the answer, in experience, comes out unambiguous.

Our own preparation as an example

Our tub carries the statements discussed above: a milk thistle seed extract 20:1 with 80 % silymarin, giving 100 mg of silymarin per tablet, plus 100 mg of inulin from chicory root, one tablet daily after a meal.

How a price per month can be worked out from this, and what makes two tubs comparable, is set out in the milk thistle buying guide.

So plant part, ratio, content and milligrams per daily portion are stated — not stated are the solvent and the measurement method, and the value of 80 % lies above the pharmacopoeia range of 30 to 65 %, which fits a photometric determination. Further preparations in the same context are in the liver category.

For milk thistle and silymarin there is no health claim authorised under Regulation (EC) No 1924/2006; the applications concerning botanical substances have been on hold for years. A milk thistle product therefore describes what it contains — and quality here means nothing other than that this description is accurate. Food supplements are not a substitute for a balanced and varied diet and a healthy lifestyle.

Frequently asked questions

How can I tell whether there is any milk thistle in it at all? From the label, not at all — that was the point of the British study, in which two out of 21 food supplements showed not a single zone typical of milk thistle. What you can check is the density of the statements: plant part, ratio, milligrams of silymarin per daily portion. Anyone stating all three has measured them.

Is a medicinal product automatically better than a food supplement? It is subject to different requirements. In the study cited, all ten products authorised as traditional herbal medicinal products were unremarkable, while every second food supplement was not. That is a difference in testing obligations, not a statement about individual preparations — a food supplement with clean analytics is in no way worse off.

Do I need to worry about mould toxins? The aflatoxin levels measured in the FDA study were low. What is more disquieting is that there is no separate maximum level for herbal food supplements — so testing only happens where the manufacturer arranges it. That is one of the points a certificate of analysis can be asked about.

What does the batch number get me? It is the reference point for every analysis. A content value without a batch is an advertising statement, a content value with a batch is a measurement — and because the content demonstrably varies between batches from the same manufacturer, the difference is not academic.

Sources: Frommenwiler DA et al., Front Pharmacol 2022;13:925298 (HPTLC study of 31 milk thistle products) · Fenclová M et al., Sci Rep 2019;9:11118 (26 preparations, content and contaminants) · Tournas VH et al., Food Addit Contam Part A 2012;29:994–999 (aflatoxins in 83 samples) · European Medicines Agency, assessment report on Silybum marianum (L.) Gaertn., fructus, EMA/HMPC/294188/2013 (containing the references to Ph. Eur. 01/2014:1860 and 01/2019:2071) · Regulation (EU) 2023/915 on maximum levels for certain contaminants in food · German Food Supplements Regulation (NemV) § 4 · Regulation (EC) No 1924/2006.