Heavy Metal Index

Comparison between pollutants found in breast milk and infant formula in the last decade: A review

Source access

Publication
Science of the Total Environment
Year
2023
Authors
Martín-Carrasco I; Carbonero-Aguilar P; Dahiri B; Moreno IM
Source type
peer reviewed review
License
CC BY NC ND 4.0

This is a narrative review by a research team at the University of Seville (Spain) summarizing detected pollutants in breast milk and infant formula across studies published between 2012 and 2022. The authors searched PubMed/MEDLINE and Scopus, identified 2412 breast-milk and 3053 infant-formula candidate records, screened and assessed for eligibility, and included 65 breast-milk studies and 73 infant-formula studies. The review groups contaminants into seven categories: metals and metalloids, chemical compounds derived from heat treatment (acrylamide, 3-MCPD, 2-MCPD, glycidol, PAHs, HMF, furan, dioxins and PCBs), pharmaceuticals (antibiotics, isoflavonoid phytoestrogens, veterinary drug residues), mycotoxins (AFM1, OTA, DON, ZEN, FBs, T2/HT-2), pesticides (OCPs, OPPs, carbamates, pyrethroids, neonicotinoids), packaging-derived contaminants (PFAS, bisphenols, phthalates), and other emerging pollutants (chlorinated paraffins, brominated flame retardants including PBDEs/HBCDD/TBBPA, parabens, benzophenones, mineral oil hydrocarbons). The authors conclude that breast milk most commonly carries metals and pesticides while infant formula most commonly carries metals, mycotoxins, and packaging-derived contaminants, and frame feeding choice as case-dependent on the maternal-environment exposure profile rather than a universal preference.

For the Heavy Metal Index, the load-bearing portions are Section 3.1 (metals) with Tables 3 (breast milk) and 4 (infant formula), Table 1 (regulatory limits in formula and calculated TDI-derived intakes by age band), and the metal-relevant discussion. The non-metal contaminant categories (Sections 3.2–3.7, Tables 5–16) are summarized at the description level so that future syntheses can locate the underlying primary studies through this review, but their numeric findings are out of HMI scope and are not reproduced in the Key numbers section below.

Key numbers

The review reproduces values from the constituent studies as the original authors reported them (means ± SD, medians and ranges, or detection-frequency summaries). Units in the source tables are inconsistent — Table 3 (breast milk) is mostly in µg/L, Table 4 (infant formula) is mostly in µg/kg as-placed-on-market — and that mixed-basis labeling is preserved here. All numbers below are exactly as printed by Martín-Carrasco et al.; they have not been re-normalized.

Regulatory and intake reference values (Table 1, condensed for metals; calculated TDI-derived intakes at 3-month/6-month/12-month body weights of 6.5/8/9.5 kg):

  • Pb: threshold in formula 20 µg/kg; TDI 3.57 µg/kg b.w.; calculated 23.21 / 28.56 / 33.92 µg.
  • Cd: threshold in formula 20 µg/kg; TDI 0.36 µg/kg b.w.; calculated 2.34 / 2.88 / 3.42 µg.
  • As (total): threshold in formula 2.1 µg/kg; TDI 0.3 µg/kg b.w.; calculated 1.95 / 2.4 / 2.85 µg.
  • Hg: TDI 4 µg/kg b.w. (maximum recommended; no explicit TDI); calculated 26 / 32 / 38 µg.
  • MeHg: TDI 1.6 µg/kg b.w.; calculated 10.4 / 12.8 / 15.2 µg.
  • Al: threshold in formula 286 µg/kg; TDI 1000 µg/kg b.w.; calculated 6500 / 8000 / 9500 µg.
  • Cr: threshold in formula 10 µg/100 kcal; TDI 2.8 µg/kg b.w.; calculated 18.2 / 22.4 / 26.6 µg.
  • Sn: threshold in formula 50,000 µg/kg; TDI 14,000 µg/kg b.w.; calculated 9100 / 11,200 / 13,300 µg.
  • Ni: threshold in formula 2.8 µg/kg; TDI 1.1 µg/kg b.w.; calculated 7.15 / 8.8 / 10.45 µg.

Selected breast-milk metals findings, Table 3 (as reported; mean ± SD or median:range; µg/L unless noted):

  • Morocco (Cherkani-Hassani et al. 2021b), n=70, ICP-MS: Pb 23.08 ± 62.75.
  • Nigeria (Ekeanyanwu et al. 2020), n=225 stratified by maternal age, HPLC-AAS: by age band 18–25 y (n=34) Cd 25 ± 13, Cr 20 ± 13, Cu 34 ± 11, Fe 51 ± 59, Pb 44 ± 13, Zn 7 ± 51; 26–30 y (n=118) Cd 31 ± 14, Pb 39 ± 12; 31–35 y (n=45) Cd 28 ± 13, Pb 39 ± 15; 36–40 y (n=28) Cd 27 ± 16, Pb 31 ± 12.
  • Zimbabwe (Bose-O’Reilly et al. 2020), n=120 stratified by Hg exposure, CV-AAS: control 0.5 µg/L; medium-exposure 1.10 µg/L; high-exposure 1.20 µg/L (no SD reported).
  • Brazil (Bastos et al. 2018), n=106, FIMS/ICP-OES: Al 2307 ± 2300, Mn 119 ± 170, As 63 ± 100, Hg 10 ± 10, Pb 531 ± 400, Cd 74 ± 10.
  • Brazil (Alves Peixoto et al. 2019), n=156, ICP-MS: Ba 11.6 (range 0.6–88.2); Cu 618.3 (95.4–954.5); Fe 380 (110.5–3594); Mn 4.5 (1.6–22.2); Mo 9.3 (0.1–39.1); Se 12.6 (3.0–70.6); Sr 43.5 (14.9–71.1); Zn 2614.4 (422.8–17,727.2).
  • South Korea (Park et al. 2018), n=207, AMA-AAS: Hg 1.19 ± 1.24 (day-15 colostrum); 0.79 ± 0.73 (day-30 milk).
  • Saudi Arabia (Al-Saleh et al. 2013), n=155, AAS/GC-MS/AMA: Hg 1.191 ± 0.764.
  • Turkey (Nazlican et al. 2022), n=34, ICP-MS: Cr 8.249 ± 11.69, As 1.64 ± 1.85, Cd 0.37 ± 0.47, Hg 2.59 ± 3.47, Pb 12.12 ± 9.08.
  • Cyprus (Kunter et al. 2017), n=50, ELISA/ICP-MS: Pb 1.19 ± 1.53, Cd 0.45 ± 0.23, As 0.73 ± 0.58, Hg 0 ± 0.20.
  • Norway (Vollset et al. 2019), n=300, ICP-MS/MS: Hg 0.2 ± 0.17 µg/L; Cd 0.08 ± 0.10 µg/kg; Pb not detected. (Source Table 3 prints Hg in µg/L and Cd in µg/kg for this row — unit inconsistency preserved as published.)
  • Poland (Winiarska-Mieczan 2014), n=320, GFAAS: Cd 2.114 ± 2.112, Pb 6.331 ± 4.614, Cu 0.137 ± 0.092, Zn 1.623 ± 1.763.
  • Spain (Motas et al. 2021), n=50, ICP-MS: Al 34.3 ± 133.0, Zn 1402.6 ± 1742.7, As 0.9 ± 2.71, Cd 0.4 ± 1.6, Pb 5.2 ± 16.7, Hg 5.6 ± 12.4, Cr 16.1 ± 63.6, Mn 10.7 ± 63.6, Fe 679.1 ± 1387.3, Ni 25.3 ± 33.8, Cu 368.5 ± 301.0, Se 44.5 ± 49.5.
  • Spain (Freire et al. 2022), n=242, ICP-MS: As 1.49 (range 0.56–3.50), Hg 0.26 (0.05–1.17), Pb 0.14 (0.10–0.65); Cd not detected.

Selected infant-formula metals findings, Table 4 (as reported; mean ± SD or median:range; µg/kg unless noted):

  • Egypt (Ghuniem et al. 2020b), n=83, ICP-OES/GFAAS: Cd 5–17, Fe 32,000–86,870, Cu 0–8760, Mn 0–7110, Zn 22,460–87,300.
  • Egypt (Ghuniem et al. 2020a), n=28, ICP-OES + UN: Mn 3300, Co/Ni/Pb not detected, Sn 47,400, Cd 20, Cu 1000, Zn 11,400.
  • Egypt (Ibrahim et al. 2020), n=60, ICP-MS, milk-based vs milk-cereal-based: milk-based Pb 424 ± 6, As 205 ± 3, Cd 14 ± 0.1, Hg 298 ± 7, Al 464 ± 29; milk-cereal-based Pb 145 ± 4, As 214 ± 8, Cd 14 ± 0.1, Hg 296 ± 14, Al 352 ± 9.
  • Nigeria (Igweze et al. 2019), n=26, AAS, milk-based vs cereal-based vs cereal-mix: milk-based Fe 4080 ± 1950, Zn 6710 ± 2100, Mn 150 ± 90, Cr 610 ± 700, Co 120 ± 320; cereal-based Fe 3710 ± 1090, Zn 6110 ± 1450, Mn 90 ± 60, Cr 360 ± 620, Co 10 ± 20; cereal-mix Fe 4490 ± 1560, Zn 7140 ± 2560, Mn 100 ± 80, Cr 400 ± 360, Co 10 ± 10.
  • Nigeria (Igweze et al. 2020a), n=26, AAS: Al 410–2470, As 20–1560, Hg 0–50.
  • Nigeria (Igweze et al. 2020b), n=26, AAS, by formula type: milk-based Pb 610–3500, Cd 10–550; cereal-based Pb 290–1950, Cd 20–370; cereal-mix Pb 470–2340, Cd 1–460.
  • Brazil (de Paiva et al. 2019), n=3, GFAAS: Cd 2, Al 1000, Cu 410, Zn 39,600, Ni 50; Pb not detected.
  • Brazil (de Mendonça Pereira et al. 2020), n=140, ICP-MS: As 0.007–0.01, Cd 0.23, Pb 0.90, Al 50, Cr 37.75, Ni 25.47.
  • United States (Gardener et al. 2019), n=91, ICP-MS: Pb 9.72; Cd 5.60.
  • China (Su et al. 2020), n=93, ICP-MS: Cr 2.51–83.80, As 0.89–7.87, Cd 0.13–3.58, Pb 0.36–5.57.
  • India (Pacquette and Anumula 2016), n=10, ICP-MS: As 0.018, Cd 0.014, Hg 0.013, Pb 0.016.
  • Iran (Dehcheshmeh et al. 2021), n=80, GFAAS/ELISA: Pb 12.57, Cd 4.97.
  • Jordan (Tahboub et al. 2021), n=22, ICP-MS, 24-element profile reported in full; toxic-element subset As 4.7 ± 3.9, Cd 8.0 ± 16.3, Pb 64.9 ± 39.9.
  • Lebanon (Elaridi et al. 2021), n=78, ICP-MS: Pb 371 ± 581, Cd 255 ± 179, As 88.7 ± 71.0.
  • Pakistan (Saeed et al. 2021), n=10, FAAS: Fe 5230–7530, Zn 3200–5370, Cu 304.67–487.0, Pb 290–1050, Cd 280–980, Al 1480–2050.
  • Pakistan (Lutfullah et al. 2014), n=46, AAS: Ca 3081 ± 277, Mg 540 ± 206, Cu 1 ± 0.76, Zn 8.97 ± 2.35, Fe 4.33 ± 2.22, Mn 0.0318 ± 0.02, Pb 0.0091 ± 0.01, Cd 0.1507 ± 0.28, Cr 0.0007 ± 0.00, Ni 0.0146 ± 0.01.
  • Italy (Astolfi et al. 2021), n=19, ICP-MS, 40-element profile in powdered formulas; toxic-element subset Sn 25 ± 2, Cd reported under non-essentials, Pb reported under non-essentials (see Table 4 for full per-element values).
  • Malta (Vella and Attard 2019), n=6, MP-AES, by age band: 0–6 mo Cr 290 ± 50, Cu 3330 ± 240, Ni 760 ± 0.00, Fe 18,340 ± 2510, Mn 2130 ± 410, Zn 27,240 ± 2770; 6–12 mo Cr 240 ± 30, Cu 3370 ± 210, Ni 820 ± 60, Fe 18,870 ± 3060, Mn 2050 ± 210, Zn 33,000 ± 950; Hg not detected in either band.
  • Portugal (Pereira et al. 2020), n=26, GFAAS: Ni 40.2.
  • Spain (Gómez-Nieto et al. 2020), n=2, HR-CS-GFAAS: Cu 3850; Cd not detected.
  • Spain (Paz 2017), n=30, ICP-OES, full mineral profile; toxic-element subset Cr 130 ± 20, Mn 1000 ± 230, Mo 140 ± 10, Ni 70 ± 40, Sr 2470 ± 950, Cd 10 ± 0.00, Pb 70 ± 20.

Authors’ qualitative conclusion on metals (Section 3.1): more metals were detected in infant formula than in breast milk across Africa, Asia, and Europe; the Americas were the only region where the reverse held for the studies reviewed. The authors note that exceedances of EFSA-derived TDI-equivalent values were reported for at least one metal in studies from Morocco, Nigeria, Tanzania, Jordan, Pakistan, Italy, Malta, Spain, Egypt, Lebanon, Iran, India, China, Brazil, and the United States; for breast milk the most-exceedance metals across the reviewed studies were Zn, Cd, Cr, Pb, Ni, As, and Hg depending on cohort, and for infant formula the most-exceedance metals were Al, Mn, Fe, Cr, Co, As, Pb, Cd, and Ni.

Methods (brief)

Narrative review. Databases: PubMed/MEDLINE and Scopus. Search strings (per Section 2.1):

  • Breast milk: (breast milk OR breastfeeding OR human breast milk) AND (contaminants OR metals OR BPA OR heavy AND metals OR pesticides OR POPs OR nitrogenous OR phthalates OR mycotoxins), 2012–2022.
  • Infant formula: (infant AND formula OR infant AND formulas OR infant AND formulae OR infant AND food OR baby AND foods) AND (contaminants OR metals OR BPA OR heavy AND metals OR pesticides OR POPs OR nitrogenous OR phthalates OR mycotoxins), 2012–2022.

PubMed returned 2109 (breast milk) and 2605 (infant formula); Scopus returned 303 and 448. Records screening reduced these to 1022 and 984; full-text eligibility assessment to 354 and 258; final inclusion to 65 breast-milk studies and 73 infant-formula studies (Fig. 1 PRISMA-style flow diagram). The authors state that selection considered both abstract and full text and that additional studies were drawn in by cross-referencing.

Data treatment (Section 2.2): two parallel databases were assembled, each with six sections (location, sample size, contaminant/residue, methods, main outcomes, references). No meta-analysis, no quality appraisal, no PROSPERO registration mentioned; this is a narrative review of inclusion-then-tabulation form rather than a PRISMA-compliant systematic review. The constituent studies’ analytical methods (ICP-MS, ICP-OES, AAS, GFAAS, AMA, CV-AAS, AAS, ELISA, GC-MS, LC-MS, etc.) are recorded per row in Tables 3, 4, and 7–16; the review itself does not perform additional measurements.

Implications

  • Certification: this review is a navigation index into the 2012–2022 primary literature on breast milk and infant formula. For HMTc threshold work the load-bearing artifacts are the constituent studies referenced in Tables 3 and 4 (each of which is a candidate for its own source page if not already present in the wiki). The review’s own contribution to threshold setting is to establish that across many world regions, both breast milk and infant formula carry detectable concentrations of Pb, Cd, As, Hg, Al, Cr, Mn, Ni, Cu, Zn, Fe, and Sn at levels that, in some cohorts, exceed EFSA-derived TDI-equivalent intake values.
  • Courses: the regulatory-intake table (Table 1) is a useful teaching artifact for showing how EFSA TDI values translate into per-infant intake at 3-, 6-, and 12-month body weights, and for showing how thresholds in formula relate to those intakes. The review’s seven-category contaminant breakdown is a useful course structure for an infant-exposure module that extends beyond metals.
  • App: breast milk is not an app ingredient; infant formula products are. The review supports app-side health-effects context for infant exposure via formula but does not provide product-occurrence values suitable for direct app ingestion without going back to the constituent studies.
  • Microbiome: the review notes infant gut microbiota colonization timing but does not study microbiome outcomes directly.

Wiki pages this source may touch

Verification notes

  • 2026-06-09 (Manual Fetch Discovery, fresh ingest via /ingest-next-manual-fetch-pdf v2 skill, Claude Opus 4.7): page written from the source PDF end-to-end. DOI identity check, raw_handle check, and cite-key check against wiki/sources/ all returned no prior page, so this is a NEW path ingest, not a merge-enhance.
  • Scope discipline: the review covers seven contaminant categories; only metals are reproduced in Key numbers because the wiki is HMI-scope. The non-metal categories are summarized in the description so future synthesis can locate the underlying primary studies via this review. No values for mycotoxins, pesticides, packaging, etc., are pulled into the wiki page.
  • Brand firewall (Part 12): the source does not name brands in the metals tables (Tables 3 and 4 identify studies only by country, sample size, and citation); no brand-naming risk in the metals subset reproduced here. The non-metal tables (Tables 5–16) include some product-form descriptors and dietary-supplement / cosmetic / pharmaceutical descriptors but no brand-by-brand contamination rankings; this is not reproduced on this wiki page in any case.
  • Speciation discipline (Part 14): the review reports total mercury (Hg) in most constituent studies and explicitly distinguishes MeHg only in Table 1’s regulatory section; the metals frontmatter therefore lists both tHg and MeHg to reflect that the source uses both labels in different contexts. Arsenic similarly: the review’s Table 1 distinguishes “As” (with TDI 0.3 µg/kg b.w., which matches EFSA inorganic-arsenic guidance) from regulatory context where iAs is the species of concern, while Tables 3 and 4 report total As from the constituent studies without speciation. tAs and iAs are both included in the metals array to preserve that distinction; the per-cell numbers reproduced under Key numbers are total As as reported in the constituent studies.
  • Unit handling: Table 3 (breast milk) values are reported predominantly as µg/L; Table 4 (infant formula) values predominantly as µg/kg or µg/100 kcal. No conversions applied. Within-row unit inconsistencies in the source (e.g., Vollset et al. 2019 reports Hg in µg/L and Cd in µg/kg in the same row of Table 3) are preserved as printed.
  • Wiki/HMTc firewall (Part 2): no threshold proposals, no synthesis claims comparing this review to other literature, no consumer guidance. The regulatory intake table is reproduced as the source presents it (EFSA TDI values translated to age-banded intakes); no HMTc-side commentary added.
  • Jurisdiction coverage: jurisdictions: lists country codes for cohorts named in Tables 3 and 4 plus a few non-metals tables where the cohort country is salient. This is a navigation aid for downstream routing; the underlying constituent studies are the authoritative sources for any per-country claim.
  • 2026-06-09 (fresh-context audit application, Claude Opus 4.7): audit subagent (general-purpose) returned REVISE verdict. Applied:
    • ❌ Malta Vella & Attard 2019, 0–6 mo Ni: audit caught a real transposition — wiki had “Ni 18,340 ± 0.00” with a fabricated table-layout-artifact rationale; the source (Table 4, p11) actually prints “Ni: 760 ± 0.00 µg/kg” and “Fe: 18,340 ± 2510 µg/kg” as two unambiguously separate rows. Corrected Ni to 760 ± 0.00 and removed the fabricated artifact note.
    • ⚠️ Norway Vollset 2019 Hg unit: audit caught µg/kg where the source prints µg/L. Corrected to µg/L; added the within-row unit-inconsistency note to acknowledge the source’s own labeling.
    • ⚠️ iAs in metals: frontmatter: per Part 14 strict reading, removed iAs from frontmatter. The constituent studies in Tables 3 and 4 measure total As; Table 1’s reference to EFSA iAs TDI is a regulatory citation, not a measurement claim by this paper, so keeping iAs in frontmatter was over-reach. tAs retained.
    • ⚠️ MeHg in metals: frontmatter: retained. Table 1 explicitly separates MeHg (TDI 1.6 µg/kg b.w.) from total Hg (4 µg/kg b.w.) as distinct regulatory rows, so the paper does distinguish the species in regulatory context even though Tables 3 and 4 report constituent-study total Hg.
    • ⚠️ [[metals/selenium]]: audit caught a missing slug — selenium has no page in wiki/metals/. Replaced with [[metals/antimony]] because Sb is in the metals frontmatter array (the source reports Sb in Ghuniem 2020a Table 4), antimony is a valid metals slug, and selenium values from constituent studies were the original justification but cannot route without a target page. Se remains in the metals frontmatter abbreviation array because the source does report it; the routing layer will reach wiki/metals/selenium if and when that page is created.

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.

CommitDateDescription
4039d202026-06-10scope: broaden ingest to the full upstream+downstream literature (marine, atmospheric, attribution, exposure, toxicology) — inclusion is the default