Parker et al. 2022 — Human health risk assessment of As, Cd, Hg, and Pb in US baby foods
This Cardno ChemRisk (now Stantec) study purchased 36 commercially available US baby and toddler foods across four primary ingredient categories (fruit, grain, leguminous vegetable, root vegetable) and analyzed them for total As, Cd, Hg, and Pb by ICP-MS (FDA EAM Method 4.7). It then conducted a full deterministic health risk assessment across three age groups (birth to 1 year, 1 to 2 years, 2 to 3 years), calculating average daily doses, hazard indices, and lifetime cancer risks. The primary finding is that rice-based grain products drive all elevated risk signals: As in grain products exceeded non-cancer hazard quotients for all age groups and concentrations tested, and the two rice-based products in the grain category had mean As of approximately 120 µg/kg against a wheat-based product with much lower As. Mercury was not detected in any sample. Cadmium and Hg were not associated with non-cancer or cancer risks under any evaluated scenario. The study concludes that risks from heavy metals in baby foods are generally within acceptable limits under most scenarios, with As in rice-based grain products as the primary exception — consistent with the broader literature.
Conflict-of-interest disclosure: Four of five authors are or were employed by Cardno ChemRisk, a consulting firm providing scientific advice to corporations and law firms. The study was unfunded. This does not invalidate the methods or data, but the risk-assessment framing (including selection of conservative assumptions and characterization of exceedances as “likely overestimated”) reflects the professional orientation of the authoring organization. The concentration data and detection frequencies are the load-bearing evidence; the risk characterization is appropriately hedged but should be read alongside that institutional context.
Key numbers
Concentrations (µg/kg, total metals, as-consumed wet weight; n=9 per category; Table 4):
| Metal | Ingredient category | Min | Mean | Median | Max | Detection (%) |
|---|---|---|---|---|---|---|
| As | Fruit | 1.5 (ND) | 3.8 | 5.0 | 5.0 | 67 |
| As | Grain | 10.0 | 90.4 | 126.0 | 132.0 | 100 |
| As | Leguminous veg | 1.5 (ND) | 4.2 | 5.0 | 5.0 | 78 |
| As | Root vegetable | 5.0 | 10.8 | 12.0 | 22.0 | 100 |
| Cd | Fruit | 1.5 (ND) | 4.4 | 1.5 (ND) | 16.0 | 33 |
| Cd | Grain | 12.0 | 25.8 | 20.0 | 61.0 | 100 |
| Cd | Leguminous veg | 1.5 (ND) | 1.5 | 1.5 (ND) | 1.5 (ND) | 0 |
| Cd | Root vegetable | 1.5 (ND) | 3.8 | 5.0 | 5.0 | 67 |
| Hg | All categories | 1.5 (ND) | 1.5 | 1.5 (ND) | 1.5 (ND) | 0 (all ND) |
| Pb | Fruit | 1.5 (ND) | 2.7 | 1.5 (ND) | 5.0 | 33 |
| Pb | Grain | 5.0 | 9.7 | 5.0 (NQ) | 20.0 | 100 |
| Pb | Leguminous veg | 1.5 (ND) | 2.3 | 1.5 (ND) | 5.0 | 22 |
| Pb | Root vegetable | 1.5 (ND) | 15.8 | 5.0 (NQ) | 48.0 | 88 |
ND values replaced with 0.5 × LOD = 1.5 µg/kg. NQ (between LOD and LLOQ) replaced with 0.5 × LLOQ = 5.0 µg/kg. LOD: 3 µg/kg for all analytes. LLOQ: 10 µg/kg for As, Cd, Pb; 2.5 mg/kg for Hg.
Key contextual facts from the grain category: Of the three grain-based baby foods, two were rice-based and one was wheat-based. Rice-based products contained substantially higher As than wheat (Fig. 1B per text). Both rice-based products had As concentrations exceeding the FDA 100 µg/kg guidance for rice-based cereal and baby food. As concentrations in the grain category drove all elevated non-cancer hazard quotient and cancer risk findings.
Non-cancer risk (HQs and HIs, Table 7):
- As in grain products: HQs exceeded 1.0 for all age groups and all concentration scenarios (mean, median, maximum). HQs ranged from 1.18 (mean, <1 yr) to 2.82 (max, 1–3 yr). Cumulative HIs across all food categories were 1.47–3.25, driven by grain.
- Pb in root vegetable products: HQs exceeded 1.0 under mean and maximum concentration scenarios for children 1 to <3 years (HQs 1.05–3.5). Under median concentrations, HQs were below 1.0.
- Pb in fruit products: HQs marginally exceeded 1.0 (1.06–1.12) for children 1 to <3 years only at maximum concentrations, based entirely on NQ-imputed values — the authors flag this as likely overestimated.
- Cd: HQs did not exceed 1.0 for any category, age group, or concentration scenario.
- Hg: HQs not calculated (no detections); not applicable.
Cancer risk (lifetime cancer risks, Table 8):
- As: cumulative lifetime cancer risks from As across all ingredient categories ranged from 3.74 × 10⁻⁵ (mean concentrations) to 5.53 × 10⁻⁵ (maximum concentrations). These exceed the typical acceptable cancer risk threshold of 10⁻⁶; however, the authors note substantial conservatism in the assessment including treating all As as inorganic (iAs) when inorganic As is only a fraction of total As in grain products.
- Pb: cumulative lifetime cancer risks from Pb were 2.21 × 10⁻⁸ to 1.15 × 10⁻⁷ — well below 10⁻⁶.
Maximum daily Pb dose from any single food type: 1.77 µg/day (highest scenario). Authors note this is below the FDA interim limit of 3 µg/day for children.
Methods (brief)
Analytical: ICP-MS (heat-block acid digestion per FDA EAM Method 4.7, Version 1.1), conducted by AGQ Labs USA (Oxnard, CA; ISO-17025 accredited). Total As, Cd, Hg, and Pb measured; no speciation. All 36 samples analyzed in March 2019 (purchased Dec 2018 – Mar 2019). Three distinct lots per product type (n=3 products × 3 lots per category = 9 samples per category). Uncertainty: As ±6%, Cd ±15%, Pb ±8%, Hg ±15%.
Exposure parameters: age- and ingredient-specific consumer-only mean intake rates from EPA Exposure Factors Handbook; three age groups (birth to <1 yr, 1 to <2 yr, 2 to <3 yr). Body weights: 6.825, 11.4, and 13.8 kg respectively. Exposure duration per age group: 365 days each.
Toxicity criteria: iAs cancer — EPA IRIS OSF 1.5 (mg/kg/day)⁻¹; As non-cancer — EPA IRIS RfD 3×10⁻⁴ mg/kg/day; Cd non-cancer — EFSA TDI 1×10⁻³ mg/kg/day; Hg non-cancer — EFSA TDI (inorganic) 5.7×10⁻⁴ mg/kg/day; Pb non-cancer — Cal/OEHHA MADL 0.5 µg/day.
Critical limitation (for HMT&C): Total As was measured, not speciated. The risk assessment assumes all As is inorganic to produce a conservative (worst-case) estimate. This means the 90.4 µg/kg mean and 132 µg/kg maximum for grain products are total As, not iAs. The actual iAs concentrations in the grain products are unknown from this study. For the two rice-based products, FDA data suggest iAs can be 30–80% of total As in rice cereal; for wheat, the iAs fraction is typically lower. The Sadiq and Beauchemin (2021) study provides a speciated view of a comparable matrix.
Implications
Certification: This study’s concentration data for four ingredient categories supports HMT&C product-category risk framing. The grain category finding (100% detection of As and Pb, mean As 90.4 µg/kg tAs) confirms that rice-based grain products are the dominant risk driver in baby food. For root vegetables (sweet potato, carrot), Pb is detectable in 88% of samples with a mean of 15.8 µg/kg and maximum of 48.0 µg/kg — higher than most other categories — which is relevant for HMT&C product standards for root-vegetable-based baby foods. The absence of Hg detections across all 36 samples is informative context for the Hg/tHg component of HMT&C standards in non-seafood baby foods. Cadmium in grain baby foods (mean 25.8 µg/kg, max 61.0 µg/kg) is a non-trivial finding given the small sample size.
Courses: Good teaching example for how risk assessment methodology choices (treating total As as iAs, selection of intake rates, conservative non-detection handling) drive conclusions. The comparison of the paper’s grain intake rates (3.9 g/kg-day, representing total grains) to the FDA’s rice-specific rate (0.664–0.925 g/kg-day) illustrates how different exposure assumptions can produce dramatically different HQ values for the same concentration data.
App: The ingredient-category concentration data (Table 4) provides a useful multi-category reference for app contamination profile updates. Grain (rice-dominant), root vegetable, fruit, and leguminous vegetable baby food categories each get a clean distribution snapshot from this study.