Rahati et al. 2026 — Heavy Metals in Infant Formula and Baby Food, Iranian Market

This 2026 Journal of Health, Population and Nutrition study from Mashhad University of Medical Sciences measured 13 trace and toxic elements in 80 powdered infant formula and 27 baby food samples drawn from the Mashhad (Iran) retail market across three seasonal sampling rounds between June 2020 and June 2021, then applied deterministic and probabilistic (Monte Carlo, 10,000 iterations, 95th percentile) health risk assessment covering Hazard Quotient (HQ), Hazard Index (HI), and Cancer Risk (CR) for infants aged 0–24 months. Lead, cadmium, and mercury were below the instrument LOD in all samples; concentrations of aluminium, cobalt, chromium, copper, iron, and zinc exceeded FAO/WHO limits and drove all-age-group HI values above the EPA non-cancer threshold of 1.0. The authors summarize chromium cancer risks as ranging from 9.55 × 10⁻⁵ to 3.57 × 10⁻⁴ in the abstract and discussion, while Table 3 reports 3.75 × 10⁻⁴ for the 18–24 month infant-formula stratum.

Key numbers

Instrument: Inductive Coupled Plasma Emission Spectrometry (Spectro Arcoes-76004555, Germany). Acid digestion: 1 g sample + 5 mL HNO₃ + 5 mL HCl, Bain-Marie indirect heating (90 °C, 2 h after 24 h ambient), filtered through Whatman grade-42, made up to 25 mL in volumetric flask. Calibration r > 0.99 for all elements; recovery 80.46–100.55%; three iterations per sample plus a control. Reference standard materials (CRMs) used for quality control. All concentrations reported in mg/kg on the as-sampled (dry powder) basis.

LOD (Method validation, p. 3): Fe 0.06 mg/kg, Cu 0.2 mg/kg, Cr 0.14 mg/kg, Zn 0.02 mg/kg, Al 0.44 mg/kg, Cd 0.03 mg/kg, Co 0.0014 mg/kg, Hg 0.012 mg/kg, Pb 0.01 mg/kg, Se 0.11 mg/kg, Ba 0.13 mg/kg.

Below-LOD results in all samples (Table 2, p. 6):

• Pb: below LOD (0.01 mg/kg) in all 80 infant formula and 27 baby food samples across all three sampling rounds.
• Cd: below LOD (0.03 mg/kg) in all samples.
• Hg: below LOD (0.012 mg/kg) in all samples (the paper measures total mercury by ICP-OES; no MeHg speciation).

Infant formula total mean concentrations (N=80, Table 2):

• Al: 1 ± 1.4 mg/kg (FAO/WHO ML for infant formula is 1 mg/kg; mean equals the limit)
• Ba: 0.2 ± 0.1 mg/kg
• Co: 0.01 ± 0.00 mg/kg
• Cr: 0.5 ± 0.8 mg/kg (total Cr by ICP-OES, no Cr-VI speciation)
• Cu: 4.1 ± 1.2 mg/kg
• Mn: 0.6 ± 0.4 mg/kg
• Se: 0.4 ± 0.2 mg/kg
• Sr: 2.8 ± 1.5 mg/kg
• Zn: 44.12 ± 2.4 mg/kg
• Fe: 82.96 ± 4.7 mg/kg

Seasonal significance for infant formula (Table 2 p-values): Al p = 0.034, Cr p = 0.003, Se p = 0.001. Al and Cr concentrations were significantly higher in the wet (third) sampling round; Se was significantly higher in the dry (first) round.

Baby food total mean concentrations (N=27, Table 2):

• Al: 2.6 ± 1.6 mg/kg
• Ba: 0.4 ± 0.2 mg/kg
• Co: 0.08 ± 0.1 mg/kg
• Cr: 0.2 ± 0.2 mg/kg
• Cu: 2.3 ± 1.2 mg/kg
• Mn: 5.9 ± 9.5 mg/kg
• Se: 0.5 ± 0.4 mg/kg
• Sr: 2.3 ± 1.1 mg/kg
• Zn: 54.15 ± 3.1 mg/kg
• Fe: 78.27 ± 7.2 mg/kg

Seasonal significance for baby food (Table 2 p-values): Cr p = 0.003, Cu p = 0.016, Fe p = 0.026.

Authors’ exceedance summary (p. 5, Results and discussion): the source text states that mean concentrations of Al, Cd, Co, Cr, Cu, Fe, and Zn in infant formula and baby food exceeded permissible limits, but the same paragraph and Table 2 report Cd below LOD in all samples. This page treats Cd as below-LOD and flags the Cd exceedance wording as a source-side contradiction. Ba and Sr were below permissible levels. The paper’s Introduction (p. 2) cites JECFA/Codex MLs for infant formula as Cd ≤ 0.003 mg/kg, Pb ≤ 0.01 mg/kg, Hg ≤ 0.005 mg/kg, Al ≤ 1 mg/kg, Cr 0.001–0.0001 mg/kg; for baby food Cd 0.001–0.04 mg/kg, Pb ≤ 0.05 mg/kg, Zn ≤ 7 mg/day, Fe ≤ 11 mg/day, Cu ≤ 1 mg/day, Al ≤ 1 mg/day, Se ≤ 0.06 mg/day, Mn ≥ 1.2 mg/day.

Deterministic non-cancer Hazard Index (HI) — infant formula (Table 3, p. 7):

• Age 1–2 mo: HI 3.92
• Age 2–3 mo: HI 3.54
• Age 3–6 mo: HI 3.73
• Age 6–12 mo: HI 3.62
• Age 12–18 mo: HI 3.81
• Age 18–24 mo: HI 3.6
• Mean HQ row: HI 3.70

All six age strata exceed the EPA non-cancer threshold of 1.0. Mean HQ contributions: Cr 0.95, Fe 0.86, Zn 0.74, Se 0.54, Cu 0.41, Co 0.06, Mn 0.06, Sr 0.03, Al 0.00, Ba 0.

Deterministic non-cancer HI — baby food (Table 4, p. 8):

• Age >6 mo: HI 3.97
• Age >8 mo: HI 3.77
• Age 12–18 mo: HI 3.73
• Age 18–24 mo: HI 3.82
• Mean HQ row: HI 3.82

All four age strata exceed the EPA non-cancer threshold. Mean HQ contributions: Zn 0.94, Cr 0.84, Cu 0.66, Fe 0.58, Se 0.39, Mn 0.14, Ba 0.152, Co 0.03, Sr 0.03, Al 0.027.

Deterministic cancer risk (CR) via Cr — infant formula (Table 3, p. 7):

• Age 1–2 mo: 4.50 × 10⁻⁵
• Age 2–3 mo: 3.48 × 10⁻⁵
• Age 3–6 mo: 4.89 × 10⁻⁵
• Age 6–12 mo: 3.60 × 10⁻⁵
• Age 12–18 mo: 3.30 × 10⁻⁵
• Age 18–24 mo: 3.75 × 10⁻⁴ (highest age-stratum CR; at the EPA high-risk threshold of 10⁻⁴)
• Mean: 9.55 × 10⁻⁵

Deterministic cancer risk (CR) via Cr — baby food (Table 4, p. 8):

• Age >6 mo: 1.30 × 10⁻⁴
• Age >8 mo: 6.00 × 10⁻⁵
• Age 12–18 mo: 3.30 × 10⁻⁵
• Age 18–24 mo: 6.00 × 10⁻⁵
• Mean: 7.08 × 10⁻⁵

The paper computed CR for Cr, Pb, and Cd; because Pb and Cd were below LOD in all samples, only Cr-driven CR is reported.

Probabilistic (Monte Carlo, 10,000 iterations, 95% CI) — infant formula HI (Fig. 1, p. 9):

• HI 0–3 mo: 5th percentile 2.27 / 95th 4.45
• HI 3–6 mo: 5th 2.49 / 95th 4.85
• HI 6–12 mo: 5th 2.33 / 95th 4.56
• HI 12–18 mo: 5th 2.17 / 95th 4.19
• HI 18–24 mo: 5th 2.61 / 95th 5.05

Probabilistic CR — infant formula (Fig. 2, p. 10):

• CR 0–3 mo: 5th 7.27 × 10⁻⁶ / 95th 5.03 × 10⁻⁴
• CR 3–6 mo: 5th 7.82 × 10⁻⁶ / 95th 5.52 × 10⁻⁴
• CR 6–12 mo: 5th 7.44 × 10⁻⁶ / 95th 5.15 × 10⁻⁴
• CR 12–18 mo: 5th 6.84 × 10⁻⁶ / 95th 4.75 × 10⁻⁴
• CR 18–24 mo: 5th 8.23 × 10⁻⁶ / 95th 5.77 × 10⁻⁴

Probabilistic HI and CR — baby food (Fig. 3, p. 11):

• HI <12 mo: 5th 2.55 / 95th 4.91
• HI 12–18 mo: 5th 2.83 / 95th 5.21
• HI 18–24 mo: 5th 3.41 / 95th 6.29
• CR <12 mo: 5th 1.44 × 10⁻⁵ / 95th 1.39 × 10⁻⁴
• CR 12–18 mo: 5th 1.34 × 10⁻⁵ / 95th 1.24 × 10⁻⁴
• CR 18–24 mo: 5th 1.61 × 10⁻⁵ / 95th 1.55 × 10⁻⁴

Sensitivity analysis — infant formula HI (Fig. 4, p. 12): Cr concentration contribution ranged 30.0–33.7% across age strata; Fe 27.2–29.4%; Zn 15.0–15.8%; Se 9.9–10.3%; Cu 8.7–9.6%; body weight (BW) −1.3 to −8.8%; ingestion rate (IR) ≤ 0.3%; Sr, Ba, Mn, Al each ≤ 0.3%.

Sensitivity analysis — baby food HI (Fig. 6, p. 14): Zn concentration contribution ranged 35.8–38.5% across age strata; Cr 27.2–29.7%; Fe 12.0–13.6%; Cu 0–8.7%; Mn 6.4–6.7%; Se 6.3–6.9%; BW −2.6 to −4.1%.

Sensitivity analysis — CR (Figs. 5 and 6, pp. 13–14): Cr concentration dominated carcinogenic risk variance: 97.3–99.2% for infant formula across age strata; 99.8–99.9% for baby food. BW contributed −0.1 to −2.7%; IR essentially 0%. Effectively all carcinogenic-risk uncertainty in this study reduces to uncertainty in the Cr concentration estimate.

EDI descending order (p. 13): Fe > Zn > Cu > Sr > Al > Mn > Ba > Se > Cr > Co across all age groups.

Methods (brief)

Cross-sectional market survey at the Mashhad retail level (supermarkets and pharmacies across the metropolitan area), three seasonal collection rounds at four-month intervals between June 2020 and June 2021 (two dry rounds, one wet round). Samples included both domestically produced and imported products from the dominant commercial brands available to Iranian consumers. Samples stored in pre-sterilized polyethylene bottles at 4 °C until analysis.

Acid digestion (1 g sample + 5 mL concentrated HNO₃ + 5 mL HCl, Bain-Marie indirect heating; container cap closed only after primary vapors exited; H₂O₂ added gradually; 24 h ambient followed by 90 °C × 2 h; Whatman grade-42 ashless filtration; reconstituted to 25 mL). Glassware and plastic bottles pre-washed in dilute HNO₃ and deionized water. Quantification: Inductive Coupled Plasma Emission Spectrometry (Spectro Arcoes-76004555, Germany) using multi-element calibration standards (Merck KGaA reagents) with calibration r > 0.99 for all elements. Quality control via three replicate measurements per sample, blank/control sample inclusion, and Certified Reference Materials (CRMs); reported recovery range 80.46–100.55%. All analyses performed at the central laboratory of Ferdowsi University of Mashhad.

Risk assessment followed the USEPA framework. EDI = (C × IR × EF × ED) / (BW × AT), with body weight (BW) and ingestion rate (IR) drawn from age-specific tables (Table S1). HQ = EDI / RfD. HI = ΣHQ. CR = EDI × OSF. RfD values used (Table 1, p. 4): Cd 0.001, Al 0.7, Co 0.0004, Cu 0.04, Fe 0.7, Hg 0.0004, Zn 0.3, Cr 0.003, Ni 0.02, Pb 0.0035, Ba 0.07, Mn 0.14, Se 0.005, Sr 0.6 mg/kg/day. OSF values used: Cr 0.021, Pb 0.0085, Cd 0.38 (mg/kg/day)⁻¹.

Monte Carlo simulation via Crystal Ball v11.1.34190 (10,000 iterations) at 95% confidence with sensitivity analysis of the contribution of each input variable to HI and CR variance. Statistical comparison: SPSS v26; Kolmogorov-Smirnov for normality, one-way ANOVA for normally distributed groups, Kruskal-Wallis otherwise; α = 0.05.

Limitations explicitly named by the authors (pp. 14–15): (1) BW assumed from EPA standard; (2) the oral-route cancer slope factor (CSF) for Cr is not established, so the authors substituted the respiratory-route CSF — the authors state explicitly that “caution should be exercised when interpreting the results related to the risk of carcinogenesis”; (3) CSF assumed constant across individuals despite known inter-individual variability; (4) risk assessment based only on heavy metal exposure, not on other contaminants; (5) single-city (Mashhad) sampling frame may not capture geographic variability across Iran; (6) one-year sampling period limits inference to longer-term trends.

Implications

Certification: Contributes Iranian-market occurrence data for the infant-formula-powder and infant-and-child-foods-master product categories, with full age-stratified HQ/HI/CR breakdowns suitable for probabilistic-versus-deterministic comparison in HMTc methodology work (Part 19). The Cr-driven CR signal in the 18–24 month infant formula stratum (3.75 × 10⁻⁴ deterministic, 5.77 × 10⁻⁴ probabilistic 95th percentile) is the most certification-relevant finding, but the authors’ use of an inhalation-route CSF as proxy for oral Cr carcinogenicity is a recognized methodological weakness and should not be propagated to threshold derivation without species-resolved chromium data. Pb, Cd, and Hg below LOD across the full 107-sample frame are an occurrence signal in their own right for Iranian-market 2020–2021 conditions; downstream synthesis should treat the LOD values as the relevant censoring threshold rather than imputing zero.

Courses: Useful as a worked example of probabilistic versus deterministic infant-food risk assessment using the EPA framework, including the explicit limitations panel naming the oral-Cr CSF gap and the inhalation-route substitution. The sensitivity analysis showing Cr concentration alone driving >97% of CR variance is a pedagogically clean illustration of input-variable dominance.

App: Adds Iranian-market 2020–2021 occurrence data for Al, Cr (total), Cu, Fe, Zn in infant formula powder and generic baby food. Pb, Cd, tHg all below LOD across all 107 samples. Useful as corroborating evidence in a multi-country app aggregation; not sufficient as standalone distribution data given single-city sampling.

Wiki pages this source may touch

• infant-formula-powder
• infant-and-child-foods-master
• chromium
• aluminum
• lead
• cadmium
• mercury-total
• infant-formula-powder

Verification notes

• 2026-05-17 merge-enhance pass (Claude Code, manual-fetch-kimi ingest cycle, /ingest-next-manual-fetch-pdf skill v2.0): predecessor page (updated 2026-05-13) had the following defects, all fixed in this revision against the source PDF:
  • raw_handle: manual-fetch-kimi (generic placeholder) → MFK_monte-carlo-simulation-approach-for-health-risk-an (per the MFK_ convention).
  • raw_path truncated to Monte Carlo simulation approach for health risk analysis of .pdf → full filename Monte Carlo simulation approach for health risk analysis of heavy metals' contamination in infant formula and food on the Iranian market.pdf.
  • access_url added (DOI URL).
  • near_duplicates: [] added.
  • metals: [Pb, Cd, tHg, Al, Cr] extended to include the additional bare-symbol elements the paper measures (Co, Cu, Fe, Mn, Zn, Ba, Sr, Se) per the source-page metal-vocabulary convention.
  • ingredients: [[[ingredients/rice]], [[ingredients/wheat]]] — paper does not measure rice or wheat as ingredients; the mentions in the Discussion are general agricultural-context commentary, not direct measurements on this study’s samples. Replaced with [[ingredients/infant-formula-powder]] (the only ingredient-page-vocabulary slug that maps to the sampled commodity).
  • products: [[[products/infant-formula-powder-non-soy]]] was overspecified — the paper does not differentiate soy versus non-soy formula. Replaced with the umbrella [[products/infant-formula-powder]] per Part 5b broad-scope rule; the routing layer fans this out to soy and non-soy sibling pages.
  • matrices: [infant-formula, baby-food, dry-weight] — dry-weight is a basis, not a matrix; removed. Native basis (as-sampled dry powder) clarified in Key numbers.
  • Numerical fidelity fixes against Table 2, p. 6:
    • Co (infant formula total mean): predecessor reported 0.08 ± 0.1 mg/kg; Table 2 shows infant formula total mean Co = 0.01 ± 0.00 mg/kg (the 0.08 ± 0.1 is the baby food row). Corrected.
    • Se (infant formula total mean): predecessor reported SD as ± 0.2; Table 2 shows 0.4 ± 0.2. Retained.
    • Baby food row: predecessor listed Cr 2.3 ± 1.2, Cu 5.9 ± 9.5, Mn 5.9 ± 9.8, Se 2.3 ± 1.1. Table 2 actual baby food total mean is Cr 0.2 ± 0.2, Cu 2.3 ± 1.2, Mn 5.9 ± 9.5, Se 0.5 ± 0.4 — the predecessor had a column-shift transposition across Cr / Cu / Mn / Se (each value attributed to the next-leftward metal). Corrected and the missing Ba (0.4 ± 0.2) and Co (0.08 ± 0.1) added.
    • Cu infant formula total mean (4.1 ± 1.2 mg/kg) and Mn infant formula total mean (0.6 ± 0.4 mg/kg) were absent from predecessor Key numbers; added.
    • Probabilistic CR infant formula 5th-percentile lower bound: predecessor reported 6.86 × 10⁻⁶; Fig. 2 (CR-12_18) shows 6.84 × 10⁻⁶. Corrected.
  • All deterministic HI, HQ, and CR values for the six infant-formula and four baby-food age strata re-verified line-by-line against Tables 3 and 4 (pp. 7–8); the predecessor values were correct.
  • All Monte Carlo 5th/95th percentile values for HI and CR re-verified against Figs. 1, 2, and 3 (pp. 9–11); the predecessor’s collapsed range statements (e.g., “Probabilistic HI 2.17–5.05”) replaced with explicit per-age-stratum 5th and 95th percentile values for downstream synthesis precision.
  • Sensitivity analysis percentages for HI (Figs. 4 and 6) and CR (Figs. 5 and 6) added in full at the per-element level; predecessor had only collapsed-range statements.
  • ## Wiki pages updated on ingest heading (legacy) → ## Wiki pages this source may touch (current v2 template), with [[metals/aluminium]] → [[metals/aluminum]] (correct taxonomy spelling), invalid [[products/baby-food-grain-based]] removed (not in current product vocabulary), and [[metals/mercury-total]] and [[ingredients/infant-formula-powder]] added.
• Cross-vendor audit (Codex, 2026-05-17) corrected Table 2 standard-deviation transcriptions for infant formula Cu/Mn/Se and baby food Ba/Co/Cr/Se against Table 2, p. 6; it also made explicit the source-side contradiction where the abstract/discussion summarize Cr cancer risk to 3.57 × 10⁻⁴ while Table 3 reports 3.75 × 10⁻⁴ for the 18–24 month infant-formula stratum, and where Results text lists Cd among exceeded elements despite Table 2 and the same paragraph reporting Cd below LOD.
• Paper-internal note: the Results-and-discussion paragraph on p. 5 lists cadmium among elements exceeding FAO/WHO permissible limits (“aluminium, cadmium, cobalt, chromium, copper, iron, and zinc”), but Table 2 on p. 6 reports cadmium as nd (below the 0.03 mg/kg LOD) in all 80 infant formula and all 27 baby food samples, and the abstract also omits Cd from the exceedance list. The wiki’s exceedance summary follows Table 2 and the abstract (Al, Co, Cr, Cu, Fe, Zn) rather than the contradicted p. 5 sentence.
• Paper-internal note: the authors substitute the inhalation-route Cancer Slope Factor for Cr in place of the (non-established) oral-route CSF (limitation 2, p. 14). This is explicitly named in the source; downstream Cr threshold work should treat the CR values as a methodological-uncertainty signal, not as direct evidence for an oral Cr-VI carcinogenicity endpoint.
• 2026-05-17 audit subagent (general-purpose, Phase 2 of /ingest-next-manual-fetch-pdf v2.0) returned PROMOTE on all five checks against the source PDF. Optional enhancement applied: the p. 5 / Table 2 Cd inconsistency note above. No corrections required; no false positives recorded.
• 2026-05-17 second-pass merge-enhance (Claude Code, /ingest-next-manual-fetch-pdf skill v2.0, re-process of the originally-named PDF): identified a Co/Hg LOD transposition that both the prior subagent PROMOTE pass and the Codex REVISE pass missed. Source PDF p. 3 body text states verbatim “0.0014mg/kg for cobalt, 0.012mg/kg for mercury”; the predecessor revision (and both subsequent audits) recorded “Co 0.012 mg/kg, Hg 0.0014 mg/kg” — i.e., the Co and Hg LOD values were swapped. The downstream consequence is that the Hg below-LOD censoring threshold reported on this page was off by an order of magnitude (0.0014 vs the correct 0.012 mg/kg). Corrected both the LOD list and the Hg below-LOD statement. No other Table 2 / Table 3 / Table 4 values touched on this pass; the Codex SD audit is preserved as-is.

Page history

The five most recent substantive edits to this page. The full version history lives in git; when DOI minting comes online (see schema docs), each entry below will also link to a version-pinned DataCite DOI.