Milk Thistle and Silymarin Testing: Flavonolignan Assays, Mycotoxin Screening, and Label Accuracy for Brands

Milk thistle (Silybum marianum) occupies a peculiar position in the supplement market: nearly every product claims standardization to “80% silymarin,” yet the analytical meaning of that number varies across suppliers, reference standards, and HPLC methods. Silymarin is not a single compound — it is a complex of at least seven flavonolignans and one flavonoid, and the ratio between those components determines both biological activity and label defensibility.

At Qalitex, milk thistle extracts represent roughly 8% of our botanical workload — incoming raw extracts, finished capsules, and softgels across potency, identity, mycotoxin, pesticide, and elemental impurity panels. The recurring theme in audit findings is that brands treat silymarin as a single analyte when their labels and pharmacopeial monographs require a multi-component characterization.

The silymarin complex: composition and why individual markers matter

Silymarin is extracted from the achenes (fruits) of Silybum marianum using ethyl acetate or ethanol–water mixtures. The flavonolignans form by oxidative coupling of taxifolin with coniferyl alcohol. The major constituents in a well-characterized 80% extract:

For EU market entry and European regulatory compliance, Care Europe provides expert consulting from Paris.

Component	Approximate % of silymarin complex	Molecular weight (Da)	UV max (nm)
Silybin A	15–25%	482.4	288
Silybin B	20–30%	482.4	288
Isosilybin A	3–7%	482.4	288
Isosilybin B	2–5%	482.4	288
Silychristin A	10–18%	482.4	288
Silydianin	5–10%	482.4	288
Taxifolin	1–5%	304.3	290

Silybin A and silybin B are diastereomers — identical molecular formula, different stereochemistry at the 2,3-bond junction. Combined, they represent 35–55% of the silymarin complex and are the most pharmacologically studied fraction. Clinical trials reporting hepatoprotective activity (Ferenci et al., J Hepatol 1989; Rambaldi et al., Cochrane Database 2007) dosed silymarin but measured outcomes correlated to silybin blood levels. Brands reporting only a total silymarin number without silybin subtotals are selling a number disconnected from the clinical evidence.

What this means for your spec: If your label states “silymarin 80%” and your marketing references liver-health studies, your CoA should break out silybin A + B as a combined percentage within 35–55% of total silymarin and confirm that remaining flavonolignans are present at ratios consistent with genuine Silybum marianum extract.

HPLC method for silymarin quantitation: column, mobile phase, and detection

At Qalitex, we quantitate silymarin components by reversed-phase HPLC-UV following a method aligned with the USP monograph for Milk Thistle Extract (powdered). The critical parameters:

Parameter	Condition
Column	C18, 250 × 4.6 mm, 5 µm particle size
Mobile phase A	0.1% phosphoric acid in water
Mobile phase B	Acetonitrile
Gradient	25% B → 45% B over 40 min, then 45% B → 80% B over 5 min
Flow rate	1.0 mL/min
Detection	UV at 288 nm
Injection volume	10 µL
Column temperature	30 °C
Reference standards	USP Silybin RS (mixture of A + B), Silychristin RS, Silydianin RS, Taxifolin RS
Run time	55 min (including equilibration)

This method resolves silybin A from silybin B with baseline separation (resolution ≥ 2.0) and separates silychristin A from the isosilybin pair — critical because co-elution inflates reported totals by 5–12% in shorter-gradient methods. We validate per ICH Q2(R1) with linearity r² ≥ 0.999 across 50–150% of target, accuracy recovery of 97–103% at three spike levels, and repeatability RSD ≤ 2.0% across six injections.

Brands frequently ask why their supplier’s silymarin result (from a 20-minute isocratic run on a C8 column) disagrees with our gradient C18 result. The answer is selectivity: a faster, less resolving method groups co-eluting peaks into a single silymarin total, while our method integrates individual components and sums them. The difference routinely reaches 8–15% relative on splits of the same lot.

Identity testing: HPTLC fingerprinting and microscopic confirmation

Potency without identity is a compliance gap. Under 21 CFR 111.75(a)(1), manufacturers must confirm the identity of each component prior to use — confirming genuine Silybum marianum and detecting adulteration with cheaper botanical fillers.

At Qalitex, we run HPTLC identity per the AHP (American Herbal Pharmacopoeia) monograph for milk thistle fruit. The developing solvent is ethyl acetate–methanol–water (77:13:10 v/v/v) with natural products/PEG derivatization and UV 366 nm visualization. Genuine extract produces intense orange-yellow silybin zones at Rf 0.45–0.55, a silychristin zone near Rf 0.30, and a taxifolin zone at Rf 0.70–0.75. Material adulterated with artichoke (Cynara scolymus) or safflower (Carthamus tinctorius) shows additional chlorogenic acid bands and absent or reduced silybin zones.

For incoming whole-fruit or powdered-fruit lots, we add microscopic ID per USP <561>. Silybum marianum achenes show thick-walled palisade sclereids in the pericarp, elongated endosperm cells with oil droplets, and the absence of starch granules — markers that close the identity loop chromatography alone cannot.

Mycotoxin risk: aflatoxins in milk thistle supply chains

Milk thistle achenes stored in bulk across Eastern Europe, South America, and China are vulnerable to post-harvest moisture that promotes Aspergillus flavus and A. parasiticus growth. In 2019, RASFF (the EU Rapid Alert System for Food and Feed) logged 14 notifications for aflatoxin in milk thistle–based food supplements — the highest count for any single botanical that year.

Aflatoxins B1, B2, G1, and G2 are the four analytes that matter. At Qalitex, we screen by LC-MS/MS with isotope-dilution quantitation, achieving a limit of quantitation (LOQ) of 0.1 µg/kg for each aflatoxin. The regulatory limits brands must meet depend on market:

Aflatoxin	EU limit (food supplements)	USP <561> advisory limit	California Prop 65 NSRL
B1	2.0 µg/kg	5.0 µg/kg	0.15 µg/day (calculated from intake)
B2	— (included in total)	— (included in total)	—
G1	— (included in total)	— (included in total)	—
G2	— (included in total)	— (included in total)	—
Total (B1+B2+G1+G2)	4.0 µg/kg	20 µg/kg	—

The EU limits are the binding constraint for brands selling internationally. A lot passing the USP advisory limit of 20 µg/kg total aflatoxins can fail the EU total limit by 5x. We have documented this gap on incoming extract lots from two of the largest global milk thistle suppliers.

Prop 65 exposure math: California’s NSRL (No Significant Risk Level) for aflatoxin B1 is 0.15 µg/day. For a product delivering 500 mg of milk thistle extract per serving, the extract must contain less than 0.30 µg/kg (0.30 ppb) of aflatoxin B1 to stay below the NSRL without a warning label. That threshold is below the EU limit and requires LC-MS/MS sensitivity that immunoaffinity ELISA kits cannot reliably achieve.

Pesticide residues: multi-residue screening for agricultural botanicals

Pesticide inputs vary by origin: Egyptian and Argentinian material may carry organophosphate and pyrethroid residues; Eastern European sources show fungicide residues (azoxystrobin, boscalid) applied during flowering. At Qalitex, we run a 400+ compound multi-residue screen by GC-MS/MS and LC-MS/MS aligned with USP <561> and the EU MRL database.

For brands selling into the EU or through retailers with “clean” positioning, we recommend testing against the tighter of USP and EU MRLs for each detected compound. The most frequent positives we encounter in milk thistle lots:

• Chlorpyrifos: detected in 12–15% of Egyptian-origin seed lots, typically at 0.02–0.10 mg/kg (EU MRL for herbal infusions: 0.01 mg/kg)
• Boscalid: detected in 8–10% of Eastern European extracts, typically at 0.05–0.20 mg/kg
• Cypermethrin: sporadic detections at low ppb levels in Chinese-origin material

Residue levels can pass one regulatory framework while failing another. Screen every incoming lot through a multi-residue pesticide panel and build pass/fail decisions from the most restrictive limit applicable to your distribution markets.

Heavy metals: concentration effects in extraction

Extraction concentrates elements along with target compounds. A 30:1 extract ratio — common for milk thistle 80% silymarin products — means that a raw seed lot at 0.10 ppm lead can yield extract at 1.5–3.0 ppm lead after accounting for incomplete co-extraction and variable partitioning. The practical limits we recommend for milk thistle extracts:

Element	Incoming extract limit	Rationale
Lead (Pb)	≤ 0.5 ppm	Prop 65 MADL 0.5 µg/day ÷ 500 mg serving = 1.0 ppm max; use 0.5 ppm for safety margin
Cadmium (Cd)	≤ 0.3 ppm	Prop 65 MADL 4.1 µg/day provides headroom, but EU limits are tighter for food supplements
Arsenic (As)	≤ 1.5 ppm total	Speciate if total exceeds 1.0 ppm — inorganic As is the regulatory concern
Mercury (Hg)	≤ 0.2 ppm	Rarely elevated in milk thistle; include as a standard panel element

Run ICP-MS per USP <233> with microwave-assisted acid digestion. Do not accept supplier data generated by ICP-OES without confirming the LOQ is sufficient — OES lacks sensitivity for sub-ppm lead quantitation in complex botanical matrices.

For more detail on elemental impurity methods and Prop 65 compliance, see our heavy metal testing services.

Softgel and lipid-matrix extraction: the recovery problem

Approximately 40% of the milk thistle products we test at Qalitex arrive as softgels — extract suspended in soybean oil, MCT, or sunflower lecithin. The flavonolignans must be partitioned away from the lipid matrix before HPLC injection; incomplete partitioning produces artificially low potency results.

Our validated softgel extraction protocol:

1. Shell removal: Puncture softgel, express fill into a tared flask, rinse shell with extraction solvent, and combine. Discard shell weight.
2. Lipid dissolution and partitioning: Dissolve fill mass in n-hexane, then extract flavonolignans into methanol–water (70:30 v/v) by liquid–liquid partitioning. Repeat partitioning three times to achieve ≥ 95% silybin recovery.
3. Filtration and dilution: Filter the combined methanol–water fraction through a 0.45 µm PVDF membrane, dilute to volume, and inject.

Recovery across three oil types (soybean, MCT, sunflower lecithin) ranged from 96.2% to 101.8% with RSD ≤ 1.8%. Brands that skip hexane partitioning and attempt direct methanol dissolution typically see 10–20% under-recovery, which reads as a potency failure on the CoA and triggers unnecessary investigations.

Stability considerations for silymarin products

Silymarin flavonolignans are susceptible to oxidative degradation, particularly in softgel formats where dissolved oxygen and lipid peroxides accelerate breakdown. At Qalitex, accelerated stability studies (40 °C / 75% RH, 6 months per ICH Q1A) on milk thistle softgels showed silybin A + B decreases of 8–14% — enough to fail label claim if overages were set below 15%.

Track individual flavonolignans on stability, not just total silymarin. Silybin B degrades faster than silybin A under oxidative conditions, and silydianin is the least stable component. Monitoring only the total obscures which peaks are declining and whether the profile indicates a formulation defect such as inadequate nitrogen blanketing during encapsulation.

Practical checklist for your next milk thistle lot

1. Require component-level silymarin data. Your CoA should report silybin A, silybin B, isosilybin A + B, silychristin, silydianin, and taxifolin individually — not just a total silymarin percentage.

2. Align your HPLC method to your label claim. If your label says “80% silymarin,” confirm whether that number is the sum of individual flavonolignans by HPLC-UV at 288 nm or a UV-spectrophotometric total. The two methods can disagree by 10–15%.

3. Screen every lot for aflatoxins by LC-MS/MS. Set your internal limit at ≤ 2.0 µg/kg for aflatoxin B1 and ≤ 4.0 µg/kg total if you sell into the EU. For Prop 65 compliance at a 500 mg serving, you need ≤ 0.30 ppb B1.

4. Run multi-residue pesticide screening on incoming seed or extract. Use a 400+ compound panel and evaluate results against both USP and EU MRLs.

5. Run HPTLC identity on every new supplier lot. Compare banding pattern against an authenticated Silybum marianum reference — do not rely on a supplier CoA alone for 21 CFR 111 identity compliance.

6. Validate softgel extraction recovery before release testing. If your contract lab cannot demonstrate ≥ 95% silybin recovery from your specific softgel matrix, the potency number is suspect.

7. Set stability overages based on data, not convention. Plan for 10–15% silybin loss over 24 months in softgel formats and adjust initial fill accordingly.

8. Calculate elemental limits from serving size. Use ICP-MS per USP <233> and work backward from Prop 65 MADLs to set incoming material limits that protect the finished product.

A single “silymarin 80%” assay is not a testing program — it is one data point. The full system includes component-level potency, botanical identity, mycotoxin screening, pesticide residues, and elemental impurities. Brands that build the complete panel into specifications avoid reactive investigations that cost more in holds and retesting than the upfront work.

For a broader view of how these panels fit together, see our botanical supplement testing services and the complete brand guide to lab testing for supplements.

Editorial scope

This article summarizes common lab-testing considerations for brands and is not a substitute for product-specific regulatory or legal advice. Method availability and accreditation scope vary by project — confirm with Qalitex before relying on a test menu for release or registration.