Association between domestic water hardness, chlorine, and atopic dermatitis risk in early life: A population-based cross-sectional study

Background Domestic water hardness and chlorine have been suggested as important risk factors for atopic dermatitis (AD). Objective We sought to examine the link between domestic water calcium carbonate (CaCO 3 ) and chlorine concentrations, skin barrier dysfunction (increased transepidermal water loss), and AD in infancy. Methods We recruited 1303 three-month-old infants from the general population and gathered data on domestic water CaCO 3 (in milligrams per liter) and chlorine (Cl 2 ; in milligrams per liter) concentrations from local water suppliers. At enrollment, infants were examined for AD and screened for filaggrin (FLG) skin barrier gene mutation status. Transepidermal water loss was measured on unaffected forearm skin. Results CaCO 3 and chlorine levels were strongly correlated. A hybrid variable of greater than and less than median levels of CaCO 3 and total chlorine was constructed: a baseline group of low CaCO 3 /low total chlorine (CaL/ClL), high CaCO 3 /low total chlorine (CaH/ClL), low CaCO 3 /high total chlorine (CaL/ClH) and high CaCO 3 /high total chlorine (CaH/ClH). Visible AD was more common in all 3 groups versus the baseline group: adjusted odds ratio (AOR) of 1.87 (95% CI, 1.25-2.80; P  = .002) for the CaH/ClL group, AOR of 1.46 (95% CI, 0.97-2.21; P  = .07) for the CaL/ClH, and AOR of 1.61 (95% CI, 1.09-2.38; P  = .02) for the CaH/ClH group. The effect estimates were greater in children carrying FLG mutations, but formal interaction testing between water quality groups and filaggrin status was not statistically significant. Conclusions High domestic water CaCO 3 levels are associated with an increased risk of AD in infancy. The influence of increased total chlorine levels remains uncertain. An intervention trial is required to see whether installation of a domestic device to decrease CaCO 3 levels around the time of birth can reduce this risk.