Lead, cadmium and mercury in food supplements
What does the Regulation 2023/915 limit apply to: product, daily portion or raw material?

The limit is expressed in mg per kg of the finished supplement as placed on the market. It does not apply per daily portion or per raw material. Commission Regulation (EU) 2023/915 sets three values for food supplements: lead 3.0 mg/kg (Annex I point 3.1.28); cadmium 1.0 mg/kg (point 3.2.21.1), or 3.0 mg/kg for supplements consisting at least of 80% from dried seaweed, from products derived from seaweed or from dried bivalve molluscs (point 3.2.21.2); and mercury 0.10 mg/kg (point 3.3.2). None of these entries carries a remark that changes the reference point, so the general rule in Article 2(3) applies: the food as placed on the market and its edible part.

1. The three entries and their exact wording

The consolidated text of 19 August 2026 (after eleven amending acts, most recently Reg. 2026/1825) keeps these values unchanged. The entries cover every format: tablets, capsules, powders and liquids. The higher seaweed and bivalve limit exists for cadmium only. For lead and mercury, a seaweed supplement is held to the same 3.0 and 0.10 mg/kg as any other supplement.

2. What “as placed on the market” means in practice

Article 2(3) provides that the maximum levels, unless otherwise specified in Annex I, apply to food as placed on the market and to its edible part. For supplements this has three practical consequences.

• The unit is product mass, not dose. A 500 mg capsule and a 10 g sachet have the same limit in mg/kg. The recommended daily portion does not change the compliance assessment.
• The capsule shell is part of the sample. Gelatine or HPMC is an edible part of the product as placed on the market. The blister and the bottle are not. Regulation 333/2007 (point C.2.1) requires the whole part to which the limit applies to be homogenised, and for products other than fish all sample material received by the laboratory to be used.
• Article 3 does not recalculate the limit. The concentration, dilution and ingredient-proportion factors of Article 3(1)–(2) apply where no specific Union maximum level exists for the dried, diluted, processed or compound food. Food supplements have one. The argument that “the extract is only 5% of the capsule, so scale it proportionally” does not change the limit for the finished product.

The Regulation spells out the same reference point for pyrrolizidine alkaloids in botanical supplements (point 2.4.10: “applies to the food supplements as placed on the market”). For metals, it follows from the general rule. More on PAs: pyrrolizidine alkaloids in botanical food supplements.

3. Raw materials with their own limit

The finished-supplement limit is not the only threshold in the chain. Article 2(1) prohibits placing on the market food listed in Annex I, and using it as a raw material or ingredient, where it exceeds its maximum level. If a raw material is itself an Annex I category, it must meet its own threshold before it reaches the blender. Examples from the current text:

• dried root and rhizome spices (as categorised in Annex I to Reg. 396/2005, e.g. turmeric powder): lead 1.50 mg/kg (point 3.1.12.4);
• dried seaweed: nickel 30 mg/kg, wakame 40 mg/kg in the product as placed on the market (point 3.6.8, applicable from 1 July 2025). No nickel limit has been set for food supplements as such.

A botanical extract or mineral blend that is not an Annex I category has no metal limit of its own. That does not mean anything goes. Article 2(2) of Reg. 315/93 requires contaminants to be kept as low as can reasonably be achieved by following good practice (ALARA), and the finished product must still meet 3.0 / 1.0 / 0.10 mg/kg.

4. The 80% threshold for cadmium

The higher cadmium limit (3.0 mg/kg) applies to supplements consisting at least of 80% from dried seaweed, seaweed-derived products or dried bivalve molluscs. The earlier version under Reg. 1881/2006 referred to products consisting “exclusively or mainly” of these materials. The current text replaces that judgement call with a number. Consequences:

• a capsule with 60% kelp powder, magnesium and a filler falls under 1.0 mg/kg;
• the text does not specify the basis for calculating the share. With a heavy shell or carrier, a product designed as “pure algae” can fall below 80% of the finished capsule mass. Document the mass balance of the whole product;
• iodine is not covered by Reg. 2023/915. Commission Recommendation (EU) 2018/464 provides for monitoring of metals and iodine in seaweed but sets no limits.

5. No dilution, no detoxification

Food complying with the Annex I maximum levels must not be mixed with food that exceeds them (Article 2(2)). A supplement batch above 3.0 mg/kg lead cannot be “rescued” by blending it with a clean batch. Article 4 also prohibits deliberate detoxification by chemical treatment of food containing contaminants listed in Annex I. Any correction has to happen earlier, at supplier and raw-material qualification.

6. How the authority samples and assesses

Sampling and analysis for lead, cadmium and mercury are governed by Reg. 333/2007, which has a dedicated scheme for supplements. The aggregate sample must be at least 100 g or 100 ml (point B.2.2), and the number of packages is set by Table 4b:

The laboratory method must meet the Table 5 criteria. For limits ≥ 0.1 mg/kg, the limit of quantification (LOQ) must not exceed one fifth of the limit. That gives LOQ ≤ 0.6 mg/kg for lead, ≤ 0.2 mg/kg for cadmium at the 1.0 limit (≤ 0.6 at 3.0) and ≤ 0.02 mg/kg for mercury. Results are reported as x ± U, where U is the expanded uncertainty (k = 2). For metals, where there is no extraction step, the result may be reported uncorrected for recovery if a certified reference material shows the method is not biased (point D.1.2).

Decision rule (point D.2): a lot is accepted if the result does not exceed the limit, taking into account the expanded uncertainty. It is rejected if it exceeds the limit “beyond reasonable doubt”, which in practice means x − U > limit. Example: lead 3.4 ± 0.6 mg/kg gives 2.8 mg/kg, so the lot is compliant in official control. A result of 3.9 ± 0.6 mg/kg gives 3.3 mg/kg, so the lot is non-compliant. The rule applies to the enforcement sample. For defence or reference analysis, national rules apply (point D.2.3). We described the same mechanism for pesticides here: Regulation 2026/765 and supplement sampling.

7. mg/kg is not a dose: what the science says

A limit in mg/kg is a market-surveillance tool, not a safety assessment of a particular formula. The EFSA reference values are:

• Lead: the CONTAM Panel concluded that the former PTWI of 25 µg/kg bw is no longer appropriate, because there is no evidence of a threshold. The reference points are BMDL₀₁ 0.50 µg/kg bw/day (developmental neurotoxicity in young children), and in adults BMDL₀₁ 1.50 (systolic blood pressure) and BMDL₁₀ 0.63 µg/kg bw/day (chronic kidney disease) (EFSA Journal 2010;8(4):1570).
• Cadmium: tolerable weekly intake (TWI) of 2.5 µg/kg bw (EFSA 2009, confirmed in 2011: EFSA Journal 2011;9(2):1975).
• Mercury: TWI of 4 µg/kg bw for inorganic mercury and 1.3 µg/kg bw for methylmercury, both expressed as mercury (EFSA Journal 2012;10(12):2985).

Scenario B is formally compliant, yet a single product takes up a large share of the reference point for a substance with no threshold. That comes on top of dietary exposure, which EFSA estimated at 0.36–1.24 µg/kg bw/day for average adult consumers. The law answers this in two ways. Compliance with specific provisions creates a presumption of safety only for the aspects those provisions cover (Article 14(7) of Reg. 178/2002). And the authority may restrict marketing or require withdrawal where there are reasons to suspect the food is unsafe despite compliance (Article 14(8)). For products with large portion sizes, set the specification well below the limit rather than right at it.

The figures in the table are a worked example for one product, not a risk assessment. They ignore other sources, vulnerable groups and mercury speciation. Mercury in a wider context: EFSA and ATSDR on mercury limits.

8. What the Regulation does not decide

• Arsenic: the inorganic-arsenic limits in point 3.4 cover rice and rice products, food for infants, juices and (from 8 October 2025) fish, crustaceans, bivalves and cephalopods. There is no separate entry for supplements.
• Nickel in the finished supplement: entry 3.6 does not cover supplements. The seaweed limit (3.6.8) binds algal raw material used as an ingredient. Whether a capsule of 100% dried kelp is itself assessed against point 3.6.8 is not settled by the text.
• The GIS notification of a supplement placed on the Polish market does not verify metal content. Responsibility for compliance lies with the food business operator.
• Communication: meeting the limit is an obligation for every product. Presenting it as a special feature (“guaranteed to meet heavy-metal limits”) can suggest a distinction that does not exist (Art. 7(1)(c) of Reg. 1169/2011). An absolute “0% heavy metals” is hard to defend with a finite LOQ.

9. Checklist for QA and regulatory teams

1. Assign the entry: 3.1.28 / 3.2.21.1 or 3.2.21.2 / 3.3.2. For cadmium, calculate the seaweed or bivalve share of the finished-product mass (80% threshold).
2. Check raw materials with their own limit (spices, seaweed, fishery products, etc.) and require certificates of analysis (CoA) against those thresholds (Art. 2(1)).
3. Test the finished product including the shell, with a method meeting the Table 5 LOQ of Reg. 333/2007, and a report giving x ± U.
4. Set an internal batch-release limit below the legal limit that accounts for U, especially for large daily portions.
5. Do not blend non-compliant with compliant batches, and do not use chemical detoxification (Arts 2(2), 4).
6. Calculate exposure from the daily portion against the EFSA reference values, and document ALARA (Reg. 315/93 Art. 2).
7. Keep the file ready for inspection: mass balance, raw-material CoAs, results with uncertainty, and the reasoning for the entry chosen.

Glossary

Maximum level (ML)

A value in Annex I to Reg. 2023/915. Food exceeding it must not be placed on the market or used as an ingredient (Art. 2(1)).

“As placed on the market”

The reference point under Art. 2(3): the product as sold, its edible part, including the capsule shell.

Expanded uncertainty U

Measurement uncertainty with coverage factor 2 (about 95%). The authority rejects a lot when the result exceeds the limit beyond reasonable doubt.

BMDL / TWI

Lower confidence limit of the benchmark dose (a reference point, not a “safe dose”), and tolerable weekly intake.

ALARA

The principle of keeping contaminants as low as reasonably achievable (Reg. 315/93 Art. 2(2)).

Conclusion

For lead, cadmium and mercury in a food supplement, the answer is mg per kg of the finished product as sold, shell included, with no Article 3 recalculation and no reference to dose. The higher seaweed limit exists only for cadmium and only from an 80% share. The authority assesses the result with the expanded uncertainty and samples according to Table 4b. Compliance alone does not close the safety analysis for products taken in large portions, because exposure depends on the mass of the daily portion.

Primary sources

1. Commission Regulation (EU) 2023/915: Arts 2–4, Art. 10; Annex I points 3.1.12, 3.1.28, 3.2.21, 3.3.2, 3.4, 3.6.8; consolidated text 02023R0915-20260819.
2. Commission Regulation (EC) No 333/2007: Annex points B.2.2 and Table 4b, C.2.1, C.3.3.1 and Table 5, D.1–D.2; consolidated text 02007R0333-20240430 (after Implementing Reg. 2024/1045).
3. Council Regulation (EEC) No 315/93: Art. 2 (ALARA); CELEX 31993R0315.
4. Regulation (EC) No 178/2002: Art. 14(7)–(8); CELEX 32002R0178.
5. Commission Recommendation (EU) 2018/464 on the monitoring of metals and iodine in seaweed, halophytes and products based on seaweed; CELEX 32018H0464.
6. EFSA CONTAM, Scientific Opinion on Lead in Food, EFSA Journal 2010;8(4):1570.
7. EFSA CONTAM, Statement on tolerable weekly intake for cadmium, EFSA Journal 2011;9(2):1975 (cf. EFSA Journal 2009;7(3):980).
8. EFSA CONTAM, Scientific Opinion on the risk for public health related to the presence of mercury and methylmercury in food, EFSA Journal 2012;10(12):2985.