Environ. Sci. Technol. 2009, 43, 3535–3541
Factors Influencing Concentrations of Polybrominated Diphenyl Ethers (PBDEs) in Students from Antwerp, Belgium LAURENCE ROOSENS,† M O H A M E D A B O U - E L W A F A A B D A L L A H , ‡,§ STUART HARRAD,‡ HUGO NEELS,† AND A D R I A N C O V A C I * ,†,| Toxicological Centre, Department of Pharmaceutical Sciences, University of Antwerp, Universiteitsplein 1, 2610 Wilrijk, Belgium, Division of Environmental Health and Risk Management, Public Health Building, School of Geography, Earth, and Environmental Sciences, University of Birmingham, Birmingham, B15 2TT, United Kingdom, Department of Analytical Chemistry, Faculty of Pharmacy, Assiut University, 71526 Assiut, Egypt, and Laboratory for Ecophysiology, Biochemistry and Toxicology, Department of Biology, University of Antwerp, Groenenborgerlaan 171, 2020 Antwerp, Belgium
Received February 22, 2009. Revised manuscript received April 02, 2009. Accepted April 03, 2009.
Human exposure to polybrominated diphenyl ethers (PBDEs) through food and indoor dust ingestion was assessed for 19 Belgian adults. The intake of PBDEs (Σtri-hepta BDEs and BDE 209) in the studied population is influenced mainly by diet. Dietary intakes of Σtri-hepta BDEs (BDEs 28, 47, 99, 100, 153, 154, and 183) were 5.9-22.0 ng/day (median 10.3), while those via dust ingestion were 0.1-1.4 ng/day (median 0.25) or 0.3-3.5 ng/day (average 0.6), assuming dust ingestion rates of 20 and 50 mg/day, respectively. Dietary intakes of BDE 209 were 50-238 ng/day (median 95), whereas those via dust ingestion were 0.4-11 ng/day (median 1.8) or 1.0-29 ng/day (median 4.6) for dust ingestion rates of 20 and 50 mg/day, respectively. It is important to acknowledge the uncertainty associated with the dust ingestion rates. Concentrations of Σtri-hepta BDEs measured in blood serum were 0.9-7.2 ng/g lipid weight (lw) (median 1.9). This is similar to other European populations, but lower than for nonoccupationally exposed Americans (average of 19 ng/g lw). When compared with estimates of exposure via both dietary and indoor dust ingestion for Americans, the exposures reported here are consistent with the hypothesis that the difference between European and American body burdens of PBDEs is attributable primarily to greater exposure via dust ingestion for Americans. The total intake of PBDEs through food and dust for each participant could not be correlated with the corresponding serum concentration. Instead, it is hypothesized that past and episodic current higher intakes of PBDEs are more important determinants of body burden than continuous background exposures at the low levels measured in this study. * Corresponding author e-mail:
[email protected]. † Department of Pharmaceutical Sciences, University of Antwerp. ‡ University of Birmingham. § Assiut University. | Department of Biology, University of Antwerp. 10.1021/es900571h CCC: $40.75
Published on Web 04/14/2009
2009 American Chemical Society
Introduction Polybrominated diphenyl ethers (PBDEs), have been used in consumer products, such as soft furnishings, carpets, and casings for electronic equipment (1). As they are not chemically bound to the treated matrix, PBDEs have migrated from these products into the environment, contaminating indoor air and dust (2), and as a result of their hydrophobicity and resistance to degradation/metabolism, foodstuffs (3, 4). This has led to their detection in humans (5, 6). Concerns regarding the human health implications of such exposure (7) have led to bans and other restrictions within several jurisdictions on the manufacture and new use of the three commercial PBDE formulations: penta-BDE, octa-BDE, and deca-BDE. Although such regulatory measures in the European Union have stabilized concentrations during the last years (8), it has been suggested that it will take at least a decade before resulting in lower dietary intakes (9). Nonoccupational exposure is considered to occur via inhalation of indoor air, and via ingestion of food and indoor dust (2, 10). While inhalation of indoor air is considered a minor pathway (2, 11), the relative significance of the other two pathways depends both on the geographical location and on the age group considered. Specifically, while European and North American dietary exposure is similar, exposure via indoor dust ingestion is substantially greater for Americans, with the result that, while dust ingestion is considered the principal exposure pathway for most Americans (11), diet appears more important for Britons (2) and Europeans in general. Furthermore, dust ingestion is a more important exposure pathway for toddlers than adults, owing to the higher dust ingestion rates of young children (10). Indeed, higher PBDE serum concentrations have been recently acknowledged in Australian infants (12) or U.S. adolescents (13). Despite this, there are to date very few studies that have examined the relationship between an individual’s body burden and their exposure via either diet or indoor dust. While one study from the Greater Boston area (U.S.) demonstrated a positive correlation between concentrations of PBDEs in human milk and those in indoor dust, as well as the consumption of dairy products and meat (except for BDE-209, which was not detected in human milk) (5), another study conducted at different locations across the United States found no such associations (14). Against this background and the apparent differences in the relative importance of dust ingestion and dietary exposure for Europeans and North Americans, this study examines the empirical relationship between individual body burden and exposure via these pathways for a group of Belgian adults. Furthermore, while there are consistent reports of the elevated presence of BDE-209 in indoor dust (15, 16), the literature is conflicting regarding dietary exposure to this congener, with UK dietary exposure (17) exceeding considerably estimates for North America (11). The present study therefore quantifies also dietary exposure to BDE-209. To achieve these aims, PBDE concentrations and profiles in food, dust, and serum will be compared, the contribution of dietary intake and dust ingestion estimated, and concentrations of individual PBDEs measured in blood serum of 19 Belgian adults will be correlated with dust samples from their dormitories and with duplicates of their dietary intake over a period of 1 week.
Materials and Methods Participants. Nineteen students (8 males and 11 females aged between 20 and 25 years) residing in university housing VOL. 43, NO. 10, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3535
were recruited. The study was approved by the Ethics Committee of the University of Antwerp and informed consent obtained. To minimize confounding due to previous exposures, participants were required to have resided in university housing for at least three years prior to the study, and to have been resident in Belgium since childhood. Each participant completed a questionnaire providing potentially relevant information about their contact with PBDE sources at the university residence (location, furnishing, electronic/ electrical appliances, etc) and lifestyle information such as smoking, and transportation habits. Sample Collection. Duplicate Diet. These were collected between May and June 2007. Participants were instructed not to alter their usual dietary habits and provided, at the end of each day, a duplicate of each breakfast, lunch, and dinner consumed over a period of one week. Duplicate food samples were collected on a daily basis, homogenized, frozen at -20 °C and freeze-dried. The water content of each sample was determined gravimetrically for determination of concentrations on a wet weight (ww) basis. PBDE concentrations (ng/g ww) in each sample were multiplied by the mass of that sample consumed, to provide an estimate of dietary intake. Indoor Dust. Dust samples were collected on the last day of the duplicate diet, according to a standardized protocol (14). Four m2 of bare floor was vacuumed for 4 min in the student’s room (single room with no division in living, bedroom, or kitchen). Samples were collected using nylon sampling socks which were mounted in the furniture attachment of the vacuum cleaner. After sampling, socks were closed with a twist tie and sealed in a plastic container. Before and after sampling, the furniture attachment was cleaned thoroughly using soap and water and a hexaneimpregnated disposable wipe. Samples were sieved through a 500 µm mesh to ensure particle homogeneity prior to extraction. Blood Serum. Following acquisition of the diet and dust samples, each participant donated 10 mL of blood which was centrifuged in order to obtain serum. During analysis, one serum sample was lost. An aliquot (150 µL) of the samples was sent to a clinical laboratory for the determination of triglycerides and total cholesterol content. The total lipid content was calculated using the formula of Phillips et al. (18) and varied between 2.95 and 10.10 g/L. Remaining serum (3-4.5 mL) was stored at -20 °C until analysis. Sample Analysis. Full information and details on the procedures followed are given in the Supporting Information (SI), however, brief summaries are provided here. The method used for the analysis of food samples is based on that described previously (4). An accurately weighed aliquot (3-8 g) of the freeze-dried sample was spiked with internal standards (2 ng BDE-77 and 5 ng 13C-BDE-209) and Soxhlet extracted. An aliquot of the extract was used for gravimetric lipid determination, while the rest was purified via acid silica cartridges. For dust, internal standards (2 ng BDE-77 and 10 ng 13C BDE-209) were added to an accurately weighed amount of dust (typically 0.2 g) followed by Soxhlet extraction and purification as previously described (14). Preparation, extraction, and cleanup of serum samples were as described by Covaci and Voorspoels (19). Internal standards (1 ng BDE77 and 5 ng 13C-BDE-209) were added, samples were mixed with formic acid for protein denaturation and diluted. After sample loading onto OASIS HLB cartridges, PBDEs were eluted with dichloromethane and purified further on acid silica cartridges. All purified extracts were evaporated to dryness and dissolved in 100 µL iso-octane prior to GC/ECNIMS analysis (for details see SI). Quality control was achieved by regular analysis of procedural blanks, certified materials and through regular participation in interlaboratory comparison exercises for 3536
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 10, 2009
matrices, such as serum, dust and biological samples (see SI). For each analyte, the mean procedural blank value was used for subtraction. After blank subtraction, the limit of quantification (LOQ) was set at 3 times the standard deviation of the procedural blank, which ensures >99% certainty that the reported value originated from the sample. LOQs in food were 0.2-5 pg/g wet weight (ww) for Σtri-hepta-BDEs and 43-169 pg/g ww for BDE-209. Those in serum were 0.08-0.3 ng/g lipid weight (lw) for Σtri-hepta-BDEs and 5 ng/g lw for BDE-209, while those in dust were 0.04-0.5 ng/g dry weight (dw) and 20 ng/g dw for Σtri-hepta-BDEs and BDE-209, respectively. Where concentrations were below LOQ, they were reported as f*LOQ with f being the fraction of samples above LOQ. Statistical Analysis. PBDE concentrations in food, dust and serum were checked for outliers using Graphpad “Quick calcs” software performing a Grubbs’ test or extreme studentized deviate method with a standard significance level of 0.05. Correlation analysis was performed using Graphpad Instat 3.
Results and Discussion Profiles and Concentrations of Tri- To Hepta-PBDEs in Food, Dust, and Serum. Food. While the frequency of detection of these congeners in food samples was in the order BDE-153 > BDE-99 > BDE-47, BDE-47 was the major contributing congener to concentrations of Σtri-hepta-BDEs in most samples after f*LOQ correction, followed by BDE-99 and BDE-153. Concentrations of Σtri-hepta-BDEs in duplicate diet samples ranged between 1 and 128 pg/g ww with an average of 14 and a median of 10 pg/g ww. These concentrations are lower than our earlier Belgian market basket study which reported concentrations of Σtri-hepta-BDEs of between 20 and 1600 pg/g ww in fish, meat, and dairy products (4). However, our earlier study did not account for the potential decrease in PBDE content caused by food preparation techniques, like broiling (20) and by inclusion of plant-based comestibles, which contain lower concentrations of PBDEs (3). Concentrations in food samples from this study are also markedly lower than those reported previously for North America where those in fish, meat, and dairy products fell between 7 and 3700 pg/g ww (21). Dust. Samples were characterized by a high detection frequency of PBDE congeners, between 60 and 100% for individual congeners. BDE-209 contributed substantially to the total PBDE content in every sample (45-93%, Figure 1). Sum tri-hepta-BDEs were 5.3 - 69.7 ng/g dw, median 11.9 ng/g dw, with one outlier (69.7 ng/g dw) detected. Table 1 summarizes the descriptive statistics for the present and related studies. Concentrations of ΣPBDEs, including BDE209, are similar to previous data for Belgium and Italy (22), but much lower than reported for the UK (16) (Table 1). In addition, PBDE concentrations are an order of magnitude lower than that reported for the United States (16) and Canada (23). We note that PBDE concentrations in U.S. dust samples are composed equally of BDE-47, BDE-99, and BDE-209 (11), whereas European samples (and UK dust in particular) are composed mainly of BDE-209 (16). Serum. The PBDE profile is comparable to other European studies with BDE-153 > BDE-99 > BDE-47 as the most abundant congeners. However, the relative congener abundance in serum differs to that in food (BDE-47 > BDE-99 > BDE-153). This might indicate preferential metabolism or bioaccumulation of lower brominated congeners (28) and may explain why BDE-28 was not detected in any serum samples, yet was present above LOQ in the majority of food and dust samples. Concentrations of Σtri-hepta-BDEs ranged between 0.9 and 7.2 ng/g lw, with a median value of 1.9 ng/g lw (Table 1). Concentrations of Σtri-hepta-BDEs are lower or similar compared to other Belgian studies (19, 25), but are
FIGURE 1. Contribution (expressed as percentage) of PBDE congeners to the studied dust samples.
TABLE 1. Descriptive Statistics of Σtri-hepta-BDEs (Congeners 28, 47, 99, 100, 153, 154, and 183) and BDE 209 Concentrations in Food (pg/g ww), Dust (ng/g dw), and Serum (ng/g lw) from the Present and Related Studies
duplicate diet(pg/g ww)
country
compounds
median
average
range
reference
Belgium
Σtri-heptaa BDE 209
10 139
14 305
17.6 903
< 1-128 < 40-7 750
present study
Σtri-heptaa BDE 209 ΣBDEsc ΣBDEsc Σtri-hexad BDE 209 Σtri-heptaa BDE 209 Σtri-heptaa BDE 209 Σtri-heptaa BDE 209 Σtri-heptae BDE 209
11.9 106 125 286 1170 630 620 560 59 2 800 1600 1300
