Human biomonitoring of phosphate flame retardants and alternative plasticizers

Ondertitel:from method development to exposure assessment

Organophosphate flame retardants (PFRs) and alternative plasticizers (APs) are two classes of emerging contaminants widely used in commercial products, such as textiles, plastics, electronic devices, food contact materials, paints and furniture. PFRs and APs were introduced as alternatives to regulated chemicals such as brominated flame retardants (BFRs) and phthalates, respectively, but toxicological studies have demonstrated potential adverse effects such as endocrine disruption of these replacement chemicals. Moreover, the release of PFRs and APs from products has resulted in the contamination of the indoor and outdoor environment, which could lead to significant exposure to humans as a result of dust or food ingestion, inhalation or dermal absorption. Although human exposure to PFRs and APs has been studied to some extent in these external matrices, data on the internal dose of human populations are scarce. As such, more information is needed regarding the associated risk of these type of exposures for humans. This thesis aimed 1) to develop reliable methods for the determination of biomarkers in human matrices, 2) to investigate the distribution and temporal variability of these biomarkers in urine, and 3) to answer relevant exposure-, effect- or risk-related questions using population data. In the first part of this thesis, two multi-target analytical methods were developed and validated according to a similar workflow. Candidate biomarkers were identified based on literature in vitro and in vivo metabolism studies. The selection consisted of fourteen metabolites (diaryl/alkyl phosphates and hydroxylated metabolites) originating from seven PFRs, and eight oxidative metabolites of three APs. Solid phase extraction proved to be a simple but effective sample preparation technique, while LC-MS/MS analysis provided the required sensitivity to detect the different metabolites at trace levels in human samples (i.e., ng/mL level). Both methods were validated in terms of recovery, matrix effects, linearity, within- and between-run precision and accuracy, and results were satisfactory for all compounds, except for bis(2-chloroethyl) phosphate (BCEP) and mono(2-ethylhexyl) adipate (MEHA). Additional quality control was assured for several biomarkers by participation in several interlaboratory comparison exercises (ICIs). Excellent results were obtained for all compounds which showed that the analytical performance was accurate and reliable, also in the human biomonitoring studies of this thesis. The distribution of PFR and AP biomarkers was studied by determining reference values for Japanese children (7-12 years old) and Flemish adolescents (14-15 years old). Results showed that several hydroxylated PFR metabolites, such as 1-hydroxy-2-propyl bis(1-chloro-2-propyl) phosphate (BCIPHIPP) and 2-hydroxyethyl bis(2-butoxyethyl) phosphate (BBOEHEP), could be detected with high frequencies in urine, complementary to other DAP metabolites originating from the same parent compound. The exposure profile of PFR metabolites in Flemish adolescents differed from the one observed for Japanese children. Increased levels and detection frequencies of all three tris(2- butoxylethyl) phosphate (TBOEP) metabolites were observed for children from Japan, likely as a result of dust ingestion because significant correlations were found between TBOEP in house dust and its metabolites in urine. Flemish adolescents, on the other hand, had comparably higher levels of both 2-ethylhexyldiphenyl phosphate (EHDPHP) metabolites in urine as a consequence of food intake. EHDPHP was the major contributor to the total dietary exposure of PFRs in food basket studies from Belgium, Sweden and Norway, possibly due to its migration from packaging material. However, no significant exposure determinants could be identified from questionnaire data in the Flemish study for biomarkers of EHDPHP or other PFRs to confirm this association. AP metabolites were also frequently detected in Flemish adolescents indicating widespread exposure. Although the levels of several phthalate metabolites have decreased significantly from 2008 to 2018, phthalates were still the predominant compounds compared to APs, as shown by the calculation of estimated daily intakes (EDIs) based on metabolite levels in urine. The use of population data and questionnaires made it possible to identify several determinants of exposure to PFRs and APs. In general, physiological parameters such as sex or BMI did not impact metabolite levels. Specific product use was associated with exposure to PFRs, but not to APs. Japanese children had higher TBOEP exposure due to wooden floors treated with floor wax containing TBOEP, while the introduction of new decoration, such as curtains, carpets and wall papers was associated with increased levels of bis(1,3-dichloro-2-propyl) phosphate (BDCIPP) in Flemish adolescents. Product use is likely also related to the socio-economic status (SES), because exposure to TBOEP was higher in families with higher income (in Japan) or educational level (in Flanders). Seasonality was observed in the levels of multiple PFR metabolites with peak concentrations in summer (for both Japan and Flanders). Temperature determines the emission rate of PFRs from products into the surrounding environment (e.g., dust) and the partitioning of PFRs between air and dust, which could result in more frequent inhalation exposure to more volatile PFRs, such as tris(1-chloro-isopropyl) phosphate (TCIPP) in warmer months. However, it is also possible that the association with season is partly related to the ventilation habits of the participants because more frequent use of ventilation was consistently associated with higher levels of PFR metabolites in Japanese children's urine. However, a significant trend in the opposite direction in Flemish adolescents suggests that the mechanism behind this association is not well understood. Furthermore, exposure to APs, such as di(isononyl)cyclohexane-1,2-dicarboxylate (DINCH) and di(2-ethylhexyl) terephthalate (DEHTP), was more continuous throughout the year, possibly due to consistent exposure by food ingestion. The third part of the thesis investigated the short-term temporal variability (within and between individuals) of PFR and AP biomarkers. Changes in physiological status or external exposure might result in significant variation, which, in turn, could introduce misclassification bias and a poor estimate of exposure. AP metabolites had weak reproducibility with high within-day variance in samples collected from the same individual during 5 consecutive days (ICCs < 0.4). This is probably related to exposure through food ingestion which is more heterogeneous from day-to-day. Similar observations were made for the PFR metabolites, diphenyl phosphate (DPHP) and 2- ethylhexyl phenyl phosphate (EHPHP), while PFR metabolites with more continuous and stable sources had fair-to-good reproducibility (ICCs > 0.4). Recommendations on sampling strategies for future studies were formulated based on the discussion of these results. Due to the lack of quantitative toxicokinetic data, several assumptions had to be made in the calculation of estimated daily intakes (EDIs) of PFR and AP exposure. Median EDIs were all well below the reference doses (only one participant exceeded the RfD for TCIPP). None of the Flemish adolescents had concentrations higher than the healthbased guidance HBM-values for the sum of DINCH metabolites and the sum of DEHP metabolites, while only a very low proportion exceeded the limit of mono-iso-butyl phthalate (MiBP) and mono-n-butyl phthalate (MnBP). However, in the study population of Japanese children, higher levels of tris(1,3-dichloro-isopropyl)-phosphate (TDCIPP) in 10 dust and of TDCIPP, TBOEP, and TCIPP metabolites in urine were associated with allergy symptoms, such as eczema and rhino-conjunctivitis. Moreover, metabolite concentrations of DPHP, EHPHP and bis(2-butoxyethyl) phosphate (BBOEP) were also associated with increased levels of oxidative stress biomarkers. Overall, these results indicated that the estimated risk from PFR exposure was low for Flemish adolescents if compared with available guidance values. However in the final chapter, similar urinary PFR metabolite concentrations were positively associated with allergic symptoms and oxidative stress in Japanese children. Further experimental and epidemiological studies are needed to elucidate the association between continuous PFR and AP exposure in non-occupationally exposed humans and potential health effects.

Jaar van publicatie:2020

Trefwoorden:Doctoral thesis

Toegankelijkheid:Closed