Allergy and Sensitization during Childhood Associated with Prenatal and Lactational Exposure to Marine Pollutants
Grandjean, Philippe, Poulsen, Lars K., Heilmann, Carsten, Steuerwald, Ulrike, Weihe, Pal, Environmental Health Perspectives
BACKGROUND: Breast-feeding may affect the risk of developing allergy during childhood and may also cause exposure to immunotoxicants, such as polychlorinated biphenyls (PCBs), which are of concern as marine pollutants in the Faroe Islands and the Arctic region.
OBJECTIVES: The objective was to assess whether sensitization and development of allergic disease is associated with duration of breast-feeding and prenatal or postnatal exposures to PCBs and methylmercury.
METHODS: A cohort of 656 singleton births was formed in the Faroe Islands during 1999-2001. Duration of breast-feeding and history of asthma and atopic dermatitis were recorded at clinical examinations at 5 and 7 years of age. PCB and mercury concentrations were determined in blood samples obtained at parturition and at follow-up. Serum from 464 children (71%) at 7 years of age was analyzed for total immunoglobulin E (IgE) and grass-specific IgE.
RESULTS: The total IgE concentration in serum at 7 years of age was positively associated both with the concomitant serum PCB concentration and with the duration of breast-feeding. However, the effect only of the latter was substantially attenuated in a multivariate analysis. A raised grass-specific IgE concentration compatible with sensitization was positively associated with the duration of breast-feeding and inversely associated with prenatal methylmercury exposure. However, a history of asthma or atopic dermatitis was not associated with the duration of breast-feeding, although children with atopic dermatitis had lower prenatal PCB exposures than did nonallergic children.
CONCLUSIONS: These findings suggest that developmental exposure to immunotoxicants may both increase and decrease the risk of allergic disease and that associations between breast-feeding and subsequent allergic disease in children may, at least in part, reflect lactational exposure to immunotoxic food contaminants.
KEY WORDS: allergy, breast-feeding, developmental toxicity, environmental exposure, immunotoxicity, methylmercury, polychlorinated biphenyls. Environ Health Perspect 118:1429-1433 (2010). doi:10.1289/ehp.1002289 [Online 20 June 2010]
Exposures to marine contaminants are of much concern to populations that rely on seafood for their livelihood. Among main contaminants resulting in increased exposures, methylmercury and polychlorinated biphenyls (PCBs) share immunotoxic potentials (Belles-Isles et al. 2002; Bilrha et al. 2003; Heilmann et al. 2006). Immunoroxicity is of particular concern when the exposures happen during the development of the immune system. Important windows of vulnerability are the intrauterine and the early postnatal periods, when unique immune maturational events take place (Dietert 2008). Breast-feeding is thought to play an important role for the infant's immune system development, but the evidence is equivocal in regard to the extent of possible protection against allergic disease (Bergmann et al. 2002; Kirsten 2009; Kramer et al. 2007; van Odijk et al. 2003).
As illustrated by studies on 2,3,7,8-tetra-chlorodibenzo-p-dioxin (TCDD) (Kerkvliet 2009), an immunotoxicant may disrupt different immune maturational processes, depending upon the specific developmental timing of exposure (Dietert 2008). Because development of T-helper cell type 2 (Th2) functions is favored prenatally, whereas acquisition of Th1 functional capacities happens postnatally, the effects may depend on the age at peak exposure. For substances such as methylmercury, the peak exposure occurs during prenatal development, when the fetus shares the contaminant from the mother's diet; human milk is not an important exposure pathway for this substance (Grandjean et al. 1995). However, lipophilic contaminants, such as PCBs, accumulate postnatally, so longer breast-feeding periods will result in higher body burdens in the child (Barr et al. 2006; Patandin et al. 1997). …