The Association Between Maternal Dietary Intake and the Risk of Heavy Metals in Human Breast Milk in Korea

This study measured Pb, Cd, tHg, and tAs in breast milk from 209 Korean lactating mothers, assessed potential infant health risks using estimated daily intake (EDI) and hazard quotient (HQ), and identified maternal dietary factors associated with heavy metal concentrations via multiple linear regression. Cadmium was the most frequently detected metal (99% of samples), followed by Hg (97%), As (89%), and Pb (79%). The authors employed age-specific infant body weight and milk consumption to refine the exposure model beyond single representative values. Arsenic was identified as the largest contributor to potential health risk. Dietary associations: Pb with legume and seaweed consumption; Cd with vegetables and seaweed; tHg with sugar intake; tAs with meat intake.

Key numbers

All concentrations in µg/L (breast milk, wet-weight equivalent — milk reported in volume basis). n = 209 mother–infant pairs sampled January–March 2023. Analyzed by ICP-MS (PerkinElmer Nexion 2000 B in DRC mode) for Pb, Cd, and tAs; direct mercury analyzer (DMA; Milestones DMA, Sorisole, Italy) by thermal decomposition/amalgamation/AAS for Hg. LODs: Pb 0.017 µg/L, Cd 0.030 µg/L, Hg 0.047 µg/L, As 0.054 µg/L. Values < LOD imputed as LOD/√2.

Concentrations — Table 3 (µg/L):

EDI (µg/kg body weight/day; median, GM (SD), min–max):

• Pb: 0.017, 0.016 (0.030), 0.002–0.150
• Cd: 0.014, 0.014 (0.012), 0.003–0.051
• Hg: 0.021, 0.019 (0.015), 0.004–0.070
• As (tAs): 0.130, 0.059 (0.238), 0.003–1.503

Mean infant body weight 5.71 kg (range 2.99–9.40); absolute breastfed-infant daily milk intake (n = 40) 883.54 mL/day (range 420–1260).

Health-based guidance reference values used (Table 1):

• Pb: JECFA 0.6 µg/kg bw/day; K-MFDS BMDL₀.₁ 0.5 µg/kg bw/day
• Cd: JECFA PTMI 25 µg/kg bw/month (≈ 0.8 µg/kg bw/day); K-MFDS PTMI 25 µg/kg bw/month
• Hg (inorganic): JECFA PTWI 4.0 µg/kg bw/week (≈ 0.6 µg/kg bw/day); K-MFDS TWI 3.7 µg/kg bw/week (≈ 0.5 µg/kg bw/day)
• As (inorganic): JECFA BMDL₀.₅ 3.0 µg/kg bw/day; K-MFDS PTWI 9.0 µg/kg bw/week (≈ 1.3 µg/kg bw/day)

Health risk (HQ): As was the largest HQ contributor among the four analytes. One infant’s As EDI exceeded the Korea-MFDS health-based guidance value (HQ > 1 for As, MFDS basis). HQ for Pb, Cd, and Hg remained below 1 across the cohort at median EDI. Among absolutely breastfed infants, the ordering of HQ contribution was As > Hg > Cd > Pb (consistent with both JECFA and MFDS reference bases).

Dietary associations — Table 4 (multiple linear regression β coefficients, 95% CI; significant rows only):

• Pb ↑: Legumes β = 0.281 (95% CI 0.002–0.029); Seaweed β = 0.286 (95% CI 0.001–0.049)
• Cd ↑: Vegetables β = 0.218 (95% CI 0.004–0.036); Seaweed β = 0.191 (95% CI 0.001–0.021)
• Hg ↑: Sugars β = 0.351 (95% CI 0.008–0.059)
• As ↑: Meat β = 0.191 (95% CI 0.002–0.100)

Pine mushroom intake contributed most to total HM concentration across all four analytes (Figure 3a). Seaweed contributed most to Cd and As. By food-category aggregate (Figure 3b), agricultural products dominated Pb, Cd, and Hg contributions; seafood dominated As contribution.

Methods (brief)

ICP-MS (not specified by model). Sample collection: trained nurses used hand-pressing method to minimize contamination; 20 mL breast milk into pre-cleaned glass bottles, stored at −70°C. Sample prep: ~0.1–0.2 g diluted with alkaline solution. LODs: Pb 0.017 µg/L, Cd 0.030 µg/L, Hg 0.047 µg/L, As 0.054 µg/L. Linearity: R² > 0.99 for all. Dietary assessment: 24-hour recall method administered by trained nurses. Ethics: approved by IRB of Kyung Hee University (KHSIRB-24-218(NA) and KHSIRB-21-598(NA)).

Speciation note: tAs reported; no distinction between inorganic and organic As species. tHg reported; no MeHg speciation. The tAs figures include organic arsenicals from seafood and possibly feed additives in meat, which means the toxicological significance is different from iAs-dominated exposure scenarios.

Implications

Certification: This is the largest and most methodologically rigorous recent study of heavy metals in Korean breast milk, with a nationally representative prospective cohort design. The As GM of 1.16 µg/L and detection rate of 89% make tAs the primary concern. The association of tAs with meat intake (rather than rice or seafood) is unexpected and warrants attention — possible explanation is residual organoarsenic compounds from animal feed additives.

Courses: The dietary association data (legumes → Pb; seaweed → Pb and Cd; vegetables → Cd) provide practical nutritional guidance for lactating mothers trying to reduce heavy metal transfer to infants. The tHg → sugar intake association is unusual and should be flagged as a correlation-not-causation finding.

App: Breast milk is not in the app’s ingredient-list model. This source informs the health effects sections for breastmilk and arsenic.

Microbiome: High tAs exposure to the infant gut during the 0–6 month colonization window, combined with Cd and Pb, has implications for gut microbiome development; cross-link when relevant pages exist.

Wiki pages updated on ingest

• breastmilk
• lead
• cadmium
• mercury
• arsenic

Verification notes

2026-05-19 (Claude Opus 4.7, manual-fetch enhancement): Repaired truncated raw_path filename (was cut at “Risk of H.pdf”; now reflects the actual on-disk filename). Expanded Key numbers to include the full Table 3 percentile distribution (P5 / P25 / P50 / P75 / P95) for Pb, Cd, Hg, tAs, the complete As EDI quartet (median, GM, SD, range), the four health-based guidance reference values reproduced from Table 1, and the regression β coefficients with 95% CIs from Table 4. Added instrument-model detail for both ICP-MS (PerkinElmer Nexion 2000 B, DRC mode) and DMA (Milestones, Sorisole, Italy). Documented LOD/√2 censoring rule applied to sub-LOD values. No content claims were softened or strengthened relative to the literature; this enhancement adds precision that was previously summarized.

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.