Research Question: What is the relationship between exposure to a) contaminants and b) contaminants specifically from foods during pregnancy and lactation and contaminant concentration in human milk?

Approach: Systematic Review   

Publication: Arsenic, cadmium, lead, mercury, and PFAS exposure during pregnancy or lactation and respective concentrations in human milk: Systematic review. DOI: 10.1016/j.envres.2025.123433

Graphical Abstract: 

Background:

• This systematic review synthesizes the peer-reviewed literature on relationships between arsenic, cadmium, lead, mercury, and PFAS exposure during pregnancy or lactation and respective concentrations in human milk.
• Contaminant exposure was measured as the concentration in biological samples collected during pregnancy or lactation.

Key Findings:

• 46 studies conducted in 25 countries met the inclusion criteria.
• Higher exposure to lead, mercury, and specific PFAS chemicals (PFOA and PFOS) during pregnancy or lactation correlated with higher concentrations of each respective contaminant in human milk. Certainty of evidence was moderate for lead, PFOA, and PFOS, and low for mercury. Evidence for arsenic and cadmium was limited and inconclusive.
  • While causality cannot be determined from these correlations, evidence suggests exposure to certain contaminants during pregnancy or lactation may result in contaminant transfer to human milk and subsequent ingestion by the infant.

Addressing Research Gaps:

• Future research may consider alternative and indirect approaches to assess contaminant exposure specifically from food, such as:
  • Measuring contaminants in foods reported in diets during pregnancy or lactation and assessing those associations with concentrations in milk. Additionally, to reduce measurement error in self-reported dietary data, prospective research studies that use “-omics” technologies can identify objective markers of food exposures for pregnant and lactating individuals. For example, metabolomics – the study of how cell processes result in various substances – can be used to identify how those substances could be measured in human samples, like blood or urine.
  • Standardizing how related data sets are structured and strengthening the linkages between them, such as national food, nutrient, and contaminant datasets. Examples of data sets include dietary data collected from the National Health and Nutrition Examination Survey, nutrient composition, manufacturer data from FoodData Central, and contaminant data from the Total Diet Study.
• Intervention studies targeting reduction of contaminant exposures before or during pregnancy or lactation could potentially reduce concentrations in human milk, subsequently potentially reducing infant exposure to contaminants.
• Studies assessing relationships between exposure to total and speciated forms of contaminants, such as inorganic arsenic and methylmercury and emerging PFAS isomers, during pregnancy or lactation can contribute to a more robust understanding about their respective concentrations in human milk.

Research Questions:

• What is the bioavailability of contaminants from human milk and infant formula when consumed by the infant?
• What factors impact the bioavailability of contaminants from human milk and infant formula when consumed by the infant?
• What is the relationship between contaminants and the bioavailability of other components in human milk and infant formula?
• What are the observed levels of contaminants in infants in the United States consuming human milk and/or infant formula?

Approach: Systematic Review

Publication: Bioavailability of and US infant exposure to arsenic, cadmium, lead, mercury, and per- and polyfluoroalkyls from human milk and infant formula: Results from a series of systematic reviews. DOI: 10.1016/j.advnut.2026.100628

Background:

• This series of systematic reviews synthesizes the peer-reviewed literature on the observed levels of arsenic, cadmium, lead, mercury, and per- and polyfluoroalkyls in biological samples from infants and the bioavailability of these contaminants from human milk or infant formula when consumed by the infant.

Key Findings:

• 7 articles from 4 studies met inclusion criteria.
• These studies measured concentrations of arsenic, cadmium, lead, and mercury in biological samples from infants in the U.S in relation to their consumption of human milk and/or infant formula. No studies measured PFAS in biological samples from infants in the U.S. in relation to their consumption of human milk and/or infant formula.
  • Most evidence was from one cohort or published 30+ years ago and lacked demographic and geographic diversity. Therefore, conclusions could not be made about infants’ exposure to arsenic, cadmium, lead, or mercury from human milk and/or infant formula in the United States.
• No studies reported on the bioavailability of these contaminants from human milk or infant formula when consumed by the infant.

Addressing Research Gaps:

• Conduct biomonitoring in existing national surveillance systems (e.g., the U.S. Centers for Disease Control and Prevention’s NHANES and National Biomonitoring Program) in the U.S. to estimate infant contaminant exposure from all sources accounting for factors such as location, demographics, and socioeconomic status.
  • These existing surveillance systems as well as research partnerships with pediatric practices and local health departments could also facilitate sample collection of human milk and infant formula to estimate bioavailability ratios.
• Nationally-representative surveillance studies of mother-infant pairs with maternal dietary data, human milk, and maternal and infant biological samples could inform how contaminants transfer from maternal food exposure to human milk and, finally, from human milk to the infant. Additional sample collection of home-prepared infant formula samples could inform how contaminants transfer from infant formula to the infant.
• Bioavailability experiments for infant populations are unlikely to be conducted because of the ethical considerations of conducting experiments on infants and practical challenges such as the amount of time and repeated measures that would be needed from the infant. Observational evidence may offer useful insight if the results can be corroborated by mechanistic study designs, such as animal or in vitro models.
• Systematic reviews to synthesize findings from animal or in vitro digestion models that indirectly assess bioavailability of contaminants from human milk or infant formula are needed.