Pandelova 2012 — EU Baby Food And Formula Element Baskets
Summary
This peer-reviewed CASCADE project study measured ten elements in designed European infant-formula and solid-food/beverage market baskets. It is relevant to Category 1 because it distinguishes milk-based, soy-based, and hypoallergenic infant formula baskets and reports higher element levels in soy-based infant formulae, but its pooled-basket design means it does not provide brand-level or sample-level percentiles.
Study Scope
| Field | Value |
|---|---|
| Product scope | Infant formulae and commercial solid foods/beverages for non-breast-fed infants through 9 months |
| EU infant formula basket | 42 infant formula products from six countries: France, Germany, Italy, Portugal, Sweden, and UK |
| EU solid foods and beverages basket | 22 products from five countries: Germany, Italy, France, Spain, and UK |
| National infant formula baskets | 24 products from Italy, Spain, Slovakia, and Sweden |
| National SFB5 baskets | 13 solid food products representing fifth-month diets in Italy, Spain, Slovakia, and Sweden |
| Basket design | Pooled market baskets weighted by market share or energy contribution to infant diet |
| Analytical methods | ICP-AES for Ca, Cu, Fe, Mn, Ni, Se, Zn; ICP-sf-MS for Pb, Hg, Cd |
| Important unit note | Table 3 unit labels were checked against the rendered PDF image on 2026-05-07. Cd, Hg, Pb, and Se are reported as ug/kg fw; Cu, Fe, Mn, Ni, and Zn as mg/kg fw; Ca as g/kg fw. Structured evidence rows normalize these source units deterministically to ppb while preserving the source unit in source-page tables and structured evidence notes. |
EU Basket Formula Concentration Findings
Trace element units are preserved from Table 3 as extracted below. For structured evidence, the Table 3 PDF image was checked and units were normalized deterministically to ppb. These values should still be treated as pooled-basket concentrations, not individual-product distributions.
| Analyte | Starting milk formula | Starting soy formula | Starting hypoallergenic formula | Follow-on milk formula | Follow-on soy formula | Follow-on hypoallergenic formula | Table |
|---|---|---|---|---|---|---|---|
| Cadmium | 3.3 | 15.8 | 4.2 | 4.5 | 18.3 | 3.9 | Table 3 |
| Total mercury | <0.5 | <0.5 | <0.5 | 113 | 29.3 | 14.4 | Table 3 |
| Nickel | <0.5 | <0.5 | <0.5 | 0.1 | 1.3 | 0.1 | Table 3 |
| Lead | 8.2 | 30.5 | 13.2 | 43.9 | 20.1 | 9.3 | Table 3 |
EU Basket Solid Food And Beverage Findings
| Analyte | SFB5 | SFB6 | SFB7 | SFB8 | SFB9 | Table |
|---|---|---|---|---|---|---|
| Cadmium | 2.8 | 8.3 | 9.6 | 9.5 | 9.1 | Table 3 |
| Total mercury | 2.7 | 3.0 | 4.5 | 29.9 | 5.9 | Table 3 |
| Nickel | <0.2 | 0.1 | 0.2 | 0.1 | 0.1 | Table 3 |
| Lead | 7.9 | 8.7 | 8.3 | 17.2 | 11.0 | Table 3 |
Intake Findings
| Finding | Value | Table/section |
|---|---|---|
| EU basket cadmium intake from milk infant formula, 0 to 1 month | 0.6, unit extracted as mg/kg bw/week but likely microgram-scale in source context | Table 4 |
| EU basket cadmium intake from soy infant formula, 0 to 1 month | 2.8, same unit caveat | Table 4 |
| EU basket cadmium intake from SFB, 5 to 9 months | 3.2 to 4.9 across SFB6 to SFB9, same unit caveat | Table 4 |
| EU basket lead intake from SFB, 5 to 9 months | 3.3 to 8.5 across SFB6 to SFB9, same unit caveat | Table 4 |
| EU basket mercury intake from follow-on milk formula, 4 to 5 months | 10.2, same unit caveat | Table 4 and conclusion |
| Worst-case mercury exposure in scenario combining follow-on milk formula and fish-based commercial infant solid foods | 18.2, same unit caveat | Section 3.4.9 |
Interpretation Notes
The authors conclude that essential and non-essential element levels in starting soy-based infant formulae were higher than values in milk-based or hypoallergenic powder formulations. They also report elevated mercury and lead in the designed follow-on milk formula sample. Cadmium exposure during the weaning period exceeded EFSA’s cadmium TWI in many scenarios and nearly attained the JECFA PTWI.
Distribution Limits
This source is a pooled market-basket study. It is useful for formula-type comparisons and exposure modeling, but it cannot provide p10, p90, p95, or p100 values for individual products. The source should be used as basket evidence, not as a finished-product market distribution.
Limitations
The extracted text has unit-symbol OCR ambiguity, with microgram symbols often rendered as m; Table 3 unit labels were therefore checked against the PDF image before structured normalization. The study uses pooled baskets weighted by market share or diet contribution, so sample-level variability is not available.
Implications
- Certification: Useful A-tier evidence for soy-versus-milk formula contrast and EU infant-food basket exposure, but not for percentile calculations.
- Courses: Useful example of why market-basket studies and product-distribution studies must be kept separate.
- App: Supports formula-type and weaning-stage risk features after unit normalization.
- Microbiome: No direct microbiome endpoint.