BackgroundExposure to endocrine disrupting chemicals (EDCs) is being rigorously studied in associations with various health outcomes, however less attention has been paid to its socio-economic determinants. This study investigated how EDCs exposure levels in pregnant women could differ on individual- and area- level income in Taiwan. MethodsUrinary measurements of phthalates, nonylphenol, bisphenol A, parabens and individual socio-economic variables (income, education, etc.) of pregnant women from Taiwan Maternal and Infant Cohort Study (TMICS) were linked via residence to area-level average annual household income. EDCs concentrations were compared between the four main Taiwan regions (North, Central, South and East) and between individual income groups. Lorenz curves were plotted to describe inequalities in EDCs exposure, and EDCs exposure related to income. Concentration indexes (CIx) were calculated and compared between the four regions. ResultsNo significant differences between EDCs concentrations adjusted for molar mass and creatinine across individual-level income groups were detected. Exposure inequalities were highest for parabens (CIx = 73.6 %), BPA (CIx = 73.6) and low molecular weight (LMW) phthalates (CIx = 63.3 %). Lorenz curve for LMW phthalates distribution across area-level income was significantly above the equality line (CIx = -21.9 %, p-value < 0.05). Stratification showed significant differences in inequality (p = 0.046) in LMW phthalates exposure across area-level income between South (CIx = -23.1 %), East (CIx = -35.3 %), North (CIx = -3.3 %) and Central (CIx = -6.5 %) regions. ConclusionInequalities in LMW phthalates exposure related to area-level income in East and South regions suggest target areas for public health interventions in terms of environmental regulations and health promotion. Using Lorenz curves and CIx to describe inequalities can be useful in case of inadequate sample sizes in costly human biomonitoring studies, however this method needs to be further developed and validated using human biomonitoring datasets in more diverse populations with larger socio-economic inequalities.
Read full abstract