Methodological approaches for the assessment of bisphenol A exposure
Sofia Almeida Costa,Milton Severo,Daniela Correia,Catarina Carvalho,Vânia Magalhães,Sofia Vilela,Sara Cunha,Susana Casal,Carla Lopes,Duarte Torres
DOI: https://doi.org/10.1016/j.foodres.2023.113251
Abstract:Bisphenol A (BPA) is an endocrine disruptor used in food contact materials, by the application of polycarbonate plastics and epoxy resins. The main objective of this study is to compare the estimate of daily BPA exposure at 13 years of age and in the adult Portuguese population, using different methodological approaches, and assess the associations between this exposure and sociodemographic characteristics. Methodology: Cross-sectional data of 13-years follow-up from a population-based birth cohort Generation XXI (GXXI) (n = 2804) and from the National Food, Nutrition and Physical Activity Survey (IAN-AF 2015-2016) (n = 3845, ≥18 years old) was used. Dietary information was collected through three food diaries for adolescents and two non-consecutive 24-hour-recalls for adults. To estimate the daily exposure to BPA, three methodological approaches were used. "Food groups attribution" merged the food consumption data with the concentration of BPA in food groups. "Regression tree model" and "random forest" combined food consumption information with urinary BPA, measured in a subsample of 24-hour urine (in adolescents n = 216, and in adults n = 82), both used to predict BPA exposure in the remaining sample. The fit-index of the methodologies was assessed through the root mean square error (RMSE), mean absolute error (MAE) and Spearman correlation coefficient (ρ). Associations between BPA exposure and sociodemographic variables were tested by linear regression models, adjusted for sex, age groups (in adults) and educational level. Tolerable Daily Intake (TDI) of 0.2 ng/kg body weight (bw), recently proposed by the European Food Safety Authority (EFSA), was used for the risk characterization of BPA exposure. Results: The "random forest" was found as the best methodology to estimate the daily BPA exposure (adolescents: RMSE = 0.989, MAE = 0.727, ρ = 0.168; adults: RMSE = 0.193, MAE = 0.147, ρ = 0.250). The median dietary BPA exposure, calculated by "food groups attribution", was 79.1 and 46.1 ng/kg bw/day for adolescents and adults, respectively, while "random forest" estimated a BPA exposure of 26.7 and 38.0 ng/kg bw/day. 99.9% of the Portuguese population presented a daily exposure above TDI. Male adolescents, females and higher educated adults, were those more exposed to BPA. Conclusions: The estimated daily BPA exposure strongly depends on the methodological approach. Food groups attribution may overestimate the exposure while the random forest appears to be a better methodological approach to estimate BPA exposure. Nevertheless, for all methods, the Portuguese population presented an unsafe BPA exposure by largely exceeding the safe levels proposed by EFSA.