A national survey of tetrabromobisphenol-A, hexabromocyclododecane and decabrominated diphenyl ether in human milk from China: Occurrence and exposure assessment

Science of the Total Environment 599–600 (2017) 237–245

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

A national survey of tetrabromobisphenol-A, hexabromocyclododecane and decabrominated diphenyl ether in human milk from China: Occurrence and exposure assessment Zhixiong Shi a,⁎, Lei Zhang b, Yunfeng Zhao b, Zhiwei Sun a, Xianqing Zhou a, Jingguang Li b,⁎, Yongning Wu b a b

School of Public Health and Beijing Key Laboratory of Environmental Toxicology, Capital Medical University, Beijing 100069, China The Key Laboratory of Food Safety Risk Assessment, Ministry of Health, and China National Center for Food Safety Risk Assessment, Beijing 100021, China

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• TBBPA, HBCD and BDE-209 were measured in a national survey of human milk. • The contamination levels of some nonPBDE BFRs were obviously higher than PBDEs. • Contamination levels of TBBPA and HBCD in human milk increased rapidly from 2007 to 2011.

a r t i c l e

i n f o

Article history: Received 15 February 2017 Received in revised form 29 April 2017 Accepted 29 April 2017 Available online xxxx Editor: Adrian Covaci Keywords: Tetrabromobisphenol-A Hexabromocyclododecane Decabrominated diphenyl ether Human milk Exposure assessment Margin of exposure

a b s t r a c t A national survey of three currently used brominated flame retardants (BFRs), tetrabromobisphenol A (TBBPA), hexabromocyclododecane (HBCD) and decabrominated diphenyl ether (BDE-209) in human milk was conducted in 2011. Human milk from 16 provinces of China were collected, pooled and measured. The estimated daily intake (EDI) via human milk ingestion for nursing infant and the related health risks were evaluated. The median levels of TBBPA, HBCD and BDE-209 were 1.21, 6.83 and 0.556 ng/g lipid weight (lw), respectively. Levels of BDE209 were lower than those of TBBPA, indicating that the production and application of deca-BDE in China has been below that of TBBPA after the restriction of PBDEs. Moreover, contamination levels of TBBPA and HBCD in this survey were higher than those observed in last national survey conducted in 2007, indicating an increase of TBBPA and HBCD in the environment from 2007 to 2011. The mean estimated daily intakes (EDIs) of TBBPA, HBCD and BDE-209 via human milk for 1–6 months old infant were 39.2, 51.7 and 3.65 ng/kg bw/day, respectively. For risk assessment, margin of exposure (MOE) was calculated by comparing the BMDL10 (benchmark dose lower confidence limit for a benchmark response of 10%) to the EDI of each BFR. Large MOEs indicates that the estimated dietary exposure to these three BFRs for nursing infant is unlikely to raise significant health concerns. Compared with some currently used novel BFRs which also measured in this survey, higher contamination levels were found in some non-PBDE BFRs, indicating that the consumption pattern of BFRs has shifted from PBDEs to non-PBDE BFRs in China. © 2017 Elsevier B.V. All rights reserved.

⁎ Corresponding authors. E-mail addresses: [email protected] (Z. Shi), [email protected] (J. Li).

http://dx.doi.org/10.1016/j.scitotenv.2017.04.237 0048-9697/© 2017 Elsevier B.V. All rights reserved.

238

Z. Shi et al. / Science of the Total Environment 599–600 (2017) 237–245

1. Introduction Brominated flame retardants (BFRs) are widely used in electronics, foams and padding materials, etc., to inhibit or slow the propagation of fire. Because of their long-term and widely used, BFRs have been found to be ubiquitous in various environmental and biota matrices (Fromme et al., 2016; Yu et al., 2016). As “The World's Factory”, BFRs are mainly produced and consumed in China, results in ubiquitous pollution (Yu et al., 2016). Furthermore, electrical and electronic waste (ewaste) from developed countries has brought new sources of BFRs to China. According to “The Global E-waste Monitor 2014” reported by United Nations University, huge amounts of e-waste is transported into China for recycling; hence lots of BFRs are inevitably emitted into the environment and present serious threats to the environment and human beings during the crude recycling of e-wastes (UNU, 2015). Polybrominated diphenyl ethers (PBDEs), hexabromocyclododecane (HBCD) and tetrabromobisphenol-A (TBBPA) are the three mainly used BFRs in China in the past decades, they have been used since the 1970s and are called “legacy BFRs” (Covaci et al., 2011; Yu et al., 2016). TBBPA could be used as a reactive or additive BFR. Since the European Union reported that the use of TBBPA in circuit boards and plastics is unlikely to raise health risks, the production and application of TBBPA greatly increased (Liu et al., 2016). However, a recent study reported that long-term exposure to TBBPA may lead to immunomodulatory changes that contribute to carcinogenic processes (Dunnick et al., 2017). PBDEs and HBCD are both additive BFRs and can therefore leach or volatilize from products and enter the environment. There are three types of commercial PBDEs: penta-BDE, octa-BDE and deca-BDE. Penta-BDE and octa-BDE have been phased out since they were listed as persistent organic pollutants (POPs) under the Stockholm Convention in 2009, whereas the deca-BDE, which is mainly composed of BDE-209, is still produced and used in China. Commercial HBCD is composed of three main isomers (α-, β- and γ-HBCD), in which γ-HBCD is predominant, accounting for 75–89% of the total HBCD (α-HBCD, 10– 15%; β-HBCD, 1–12%) (Du et al., 2012). Similar to penta-BDE and octaBDE, HBCD was added into the list of POPs in 2013 because of its persistence, bioaccumulation, and bio-toxicity (UNEP/POPS, 2013). According to an announcement declared by the Ministry of Environmental Protection (MEP) of China, the production, importation and application of HBCD have been banned in China since 2016, however, at the same time the MEP announced that the production and application of HBCD in two types of building materials, extruded polystyrene (XPS) and expanded polystyrene (EPS), is still permitted in China (www.chinasafety. gov.cn). Hence, the production and application of HBCD continues in China. Taken together, environmental monitoring and health risk assessment of these currently used BFRs are not only of scientific interest but also requirement from a management perspective. Compared to blood and urine, human milk has a far higher lipid content, hence it is utilized as a very suitable matrix for evaluating human internal exposure to lipophilic organic pollutants. Furthermore, human milk monitoring can also be utilized as a perfect non-invasive approach to assess the external exposure of nursing infant to environmental pollutants. In China, national survey of human milk is conducted as part of the Chinese Total Diet Study (TDS) in the past decades. The Chinese TDS is a continuous national study, which has been conducted five times since 1990, and it has become an important tool for monitoring dietary exposure of the Chinese population to chemicals and nutrients (Li et al., 2011). In the 4th Chinese TDS carried out in 2007, the national survey of human milk covered 12 provinces of China, and the occurrence of TBBPA, HBCD and PBDEs (not including BDE-209) was measured for the first time, showing the widespread presence of BFRs in Chinese human milk (Shi et al., 2009; Zhang et al., 2011). The 5th Chinese TDS was performed in 2011, and on this occasion, the national survey of human milk expanded from 12 provinces to 16 provinces. In this survey, a series of BFRs, including legacy BFRs (TBBPA, HBCD and BDE209), and some “novel” BFRs were measured. The results of the novel

BFRs have been published (Shi et al., 2016). The objective of the present study is to measure the contamination levels of TBBPA, HBCD and BDE209 in the national human milk survey. Based on occurrence measurement, the dietary intakes of the three BFRs via human milk ingestion for nursing infants were subsequently calculated for risk assessment. In addition, since this is the second time we measured TBBPA and HBCD in national human milk survey, a comparison between TBBPA and HBCD in 2007 and 2011 is presented. In order to assess the current consumption pattern of BFRs in China, a comparison between the legacy and novel BFRs is also presented. 2. Materials and methods 2.1. Human milk collection As part of the 5th Chinese TDS, the national human milk survey in this study was conducted in 2011. Individual human milk samples were collected from healthy volunteer donors living in 16 provinces of China. These provinces are Heilongjiang (HLJ), Jilin (JL), Neimenggu (NM), Hebei (HeB), Henan (HeN), Shanxi (SX), Ningxia (NX), Qinghai (QH), Jiangxi (JX), Fujian (FJ), Shanghai (SH), Hubei (HuB), Sichuan (SC), Guangxi (GX), Zhejiang (ZJ), Guangdong (GD), respectively. These 16 provinces cover approximately 48% of the land area of China and 56% of the Chinese population. The locations of these provinces are shown in Fig. 1. In each province, 50–60 rural donors and 50 urban donors were recruited according to their living area; each had resided in her place of residence for at least 5 years. A questionnaire was employed to collect information on the donors, and the data from the questionnaires revealed no evidence of abnormally high occupational exposure to BFRs among these donors. All donors were primiparous and non-smokers. After being told the purpose of the study and signing the consent form and questionnaire, human milk was collected within 3–8 weeks after delivery. An individual sample was collected either using a pump or by hand directly expressing the milk into a polypropylene jar, which was pre-washed and provided to the donors. The samples were stored at − 18 °C, after all the human milk sample in a province were collected, all the samples were put into ice box, packed with drikold and shipped to our lab immediately. Subsequently, for each province, 10 mL of milk from each urban individual human milk sample was pooled to form one composite sample (urban sample), and each rural sample was pooled to form another composite sample (rural sample). All the pooled samples were stored at −40 °C until analysis. That is, a total of 32 pooled human milk samples from 16 provinces were subjected to analysis. However, only 29 pooled samples were tested in the present study because in earlier study these pooled samples have been

Fig. 1. Sampling provinces in the Chinese national human milk survey.

Z. Shi et al. / Science of the Total Environment 599–600 (2017) 237–245

used to test lots of chemicals or nutrients (i.e. dioxin, PCB, PFC, etc.), when these pooled samples were used to test BFRs, the sample volume of three pooled samples, including rural sample from Sichuan and the urban samples from Henan and Liaoning, was insufficient, thus, the detection of these three pooled samples had to be excluded. In addition, it should be noted that in pooled sample where a highly contaminated individual sample may increase the result of the pool, thus, results from pooled human milk samples are not ideal for making comparisons because it will skew towards an outlier (Toms et al., 2012). 2.2. Analysis of human milk 2.2.1. Reagents and chemicals All organic solvents (acetone, n-hexane, etc.) were obtained from Merck (Darmstadt, Germany); concentrated sulfuric acid (98%) was obtained from Beijing Chemical Factory (Beijing, China); LC-Si SPE cartridge (3 mL/500 mg) was obtained from Supelco (Bellefonte, PA, USA). Analytical standards (α-, β- and γ-HBCD, TBBPA and BDE-209) and isotopic internal standards (ISs) (13C12-α-, β- and γ-HBCD, 13C12TBBPA and 13C12-BDE-209) were obtained from Wellington Laboratories (Andover, MA, USA). 2.2.2. Sample preparation and instrumental analysis A slightly modified method described elsewhere was used for sample preparation (Shi et al., 2013b). In brief, human milk (20–25 mL) was freeze-dried, after spiking with ISs (5 ng each of the 13C12-labeled IS) and equilibrating, the dry powder was Soxhlet extracted for 20 h with 150 mL of n-hexane/acetone (1:1, v/v). And then the Soxhlet extract was evaporated to constant weight. After the lipid content was calculated by gravimetry, the extract was reconstituted in cyclohexane/ ethyl acetate (1:1 v/v) for following cleanup conducted by gel permeation chromatography (GPC) coupled to concentrated sulfuric acid treatment. The GPC cleanup was performed on an AccuPrep MPS GPC system (J2 Scientific Inc., USA) with a low-pressure column (BioBeads S-X3, 2 cm i.d. × 50 cm) to remove the bulk of lipid. After the extract was injected into the GPC system, fraction of 19 to 40 min was collected using cyclohexane/ethyl acetate (1:1 v/v) as the mobile phase with a flow rate of 5 mL/min. The extract was then evaporated to dryness and reconstituted in 2 mL of hexane, shaken with 1 mL of concentrated sulfuric acid to digest remaining lipid. Subsequently, the extract, which contained all the analytes, was employed to the fractionation of BDE-209 and TBBPA/HBCD. The extract was loaded into a LC-Si SPE cartridge prewashed with hexane, and then the first fraction containing BDE-209 was eluted with 6 mL of hexane, after the column dried by suction, and the second fraction containing HBCD and TBBPA was eluted with 6 mL of acetone. Finally, the hexane eluent was dried and reconstituted in 100 μL of n-hexane for instrumental analysis, whereas the acetone eluent was dried and reconstituted in 200 μL of methanol for analysis. BDE-209 was detected on an Agilent 5977A-7890B GC–MS (Palo Alto, CA, USA) with a 15 m DB-5MS column (0.25 mm i.d. and 0.1 μm film, J&W Scientific, Folsom, CA, USA). Purity helium was utilized as the mobile phase with a flow of 3 mL/min; the temperature program was as follow: initial at 100 °C for 1 min, increased to 300 °C with 20 °C/min and then held for 7 min. The mass detector was operated in NCI mode using methane as the reagent gas. The ion source and interface temperatures of the mass spectrometer were set at 200 °C and 300 °C, respectively. The selected ion monitoring (SIM) mode was used for quantitative analysis. The ions, m/z 492.5 and 494.5 for 13C12BDE-209 and m/z 486.7 and 488.6 for BDE-209, were selected for monitoring and quantitation. TBBPA and HBCD isomers were separated and detected on an Agilent 1290–6490 UHPLC-MS/MS with a 100 mm Acquity UPLC BEH C18 column (2.1 mm i.d. and 1.7 μm particle size, Waters, MA, USA); the elution flow rate was maintained at 0.3 mL/min. The initial solvent composition was set at 70:30 (water (A):methanol (B), v/v) for 1 min, changed

239

linearly to 20:80 (A:B, v/v) over 3 min, and then changed linearly to 10:90 (A:B, v/v) over 3.5, finally changed to 100% methanol, maintained for 1 min, and then changed to the initial composition. The MS/MS was operated in electrospray ionization (ESI) negative ion mode using nitrogen as the collision gas. Data were collected in multi-reaction monitoring (MRM) mode. The transitions, m/z 542.6 → 80.9 and 542.6 → 78.9 for TBBPA, m/z 554.6 → 80.9 and 554.6 → 78.9 for 13C12-TBBPA, m/z 640.7 → 80.9 and 640.7 → 78.9 for α-, β-, γ-HBCD, and m/z 652.7 → 80.9 and 652.7 → 78.9 for 13C12-α-, β-, γ-HBCD, respectively, were selected for monitoring and quantitation. Quantitative analysis of all the analytes was performed by isotope dilution approach with the corresponding 13C12-labeled IS. 2.3. Quality assurance/quality control Method blank samples were run every 10 samples to check for interference or contamination from solvents or glassware. Levels of analyte in the blank samples were all satisfactory (b5% of the typical analyte concentration in the samples). Hence the blank values were not subtracted from the sample results. Matrix spiking tests were conducted for recovery tests. Cow milk was used for recovery tests at two spiked levels (1 and 10 ng/g). Recoveries of the analytes ranged from 80% to 120% with RSDs b15% (n = 5). The concentrations reported in this study were not corrected according to the results of the recovery tests. The LODs of TBBPA, α-, β-, γ-HBCD and BDE-209 were 5, 8, 5, 10, 10 pg/g wet weight, respectively. The laboratory performance was validated by participating in the Bi-ennial Global Interlaboratory Assessment on POPs organized by the United Nations Environment Programme (UNEP) in 2016; fish and human milk samples were measured for various POPs including HBCD and PBDEs. Data from our laboratory were within acceptable range of the consensus values. All statistical analyses were performed with the SPSS software package (version 20; IBM). Because neither the concentrations nor the logtransformed concentrations of the individual BFRs were normally distributed (Kolmogorov-Smirnov-Lilliefors test, p b 0.05). Therefore, non-parametric methods were used to assess significant differences in the concentrations of various BFRs between the rural and urban human milk samples (Mann-Whitney test), the significance level was 5% unless stated otherwise. 3. Results and discussion 3.1. BFRs in human milk Levels of BFRs in human milk are listed in Table 1. 3.1.1. TBBPA TBBPA was found in all human milk samples, with mean and median levels of 7.58 and 1.21 ng/g lw, respectively. The levels of TBBPA in 70% of the samples were no N2 ng/g lw, however, far higher levels of TBBPA were found in some regions, including the urban sample from HLJ and the rural sample from LN, with levels higher than 60 ng/g lw. Because what we tested is pooled milk samples, the high TBBPA levels in this pooled samples might be due to the small number of individual samples in the pool, some highly contaminated individual samples might dramatically increase the result of the pool. In addition, high TBBPA levels in some individual samples might indicate that the donors were exposed to elevated level of TBBPA, although the questionnaires revealed no evidence of abnormally high occupational exposure to BFRs among the donors, we inferred that the regions where they lived or worked is close to some industrial sites where TBBPA was produced or used, or heavily emitted. Levels of TBBPA were found to be higher than that of BDE-209 in our study. This finding is quite different to some other studies. In 120 human milk samples collected in UK between 2010 and 2011, levels of HBCD was found to be higher than those of TBBPA and total PBDEs, whereas levels of BDE-209 were higher than those of TBBPA

Z. Shi et al. / Science of the Total Environment 599–600 (2017) 237–245

240

Table 1 Concentrations (ng/g lw) of BFRs in human milk from China. Province

Heilongjiang Liaoning Jilin Hebei Neimenggu Qinghai Shanxi Ningxia Henan Shanghai Jiangxi Hubei Sichuan Guangxi Guangdong Fujian Mean Median SD

α-HBCD

TBBPA

β-HBCD

γ-HBCD

HBCD

BDE-209

Rural

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

Urban

1.13 62.9 13.9 1.27 0.199 4.46 0.14 1.49 1.35 0.861 21.7 2.36 – 1.43 2.23 16.8 7.58 1.21 16.8

66.8 – 0.431 0.174 0.277 5.16 0.278 0.847 – 0.712 0.343 1.17 9.33 0.592 0.335 1.21

8.85 7.6 12.6 7.05 3.89 2.5 2.94 2.93 4.74 2.72 2.06 1.91 – 0.682 3.51 1.48 3.88 2.93 2.87

8.2 – 7.9 2.41 4.43 4.29 1.52 0.816 – 1.04 4.01 2.39 1.99 1.81 1.47 4.68

0.058 0.08 nd nd 0.187 0.18 0.067 1.4 0.613 nd 0.11 0.466 – 0.338 0.126 0.098 0.329 0.121 0.627

0.391 – 0.079 0.1 0.084 0.074 0.121 3.27 – 0.405 nd 0.094 0.301 0.145 0.122 0.385

2.57 2.33 0.906 1.82 0.681 0.752 1.02 13.3 3.6 8.84 1.21 13.5 – nd 4.16 0.953 5.86 1.82 13.5

72.5 – 1.650 0.455 0.591 1.42 0.676 16.4 – 2.67 2.62 0.801 4.54 6.07 0.632 3.08

11.5 10 13.5 8.87 4.75 3.44 4.03 17.6 8.95 11.9 3.38 15.9 – 1.02 7.8 2.53 10.1 6.83 14.5

81.1 – 9.63 2.97 5.11 5.78 2.32 20.5 – 4.12 6.83 3.28 6.83 8.03 2.22 8.15

1.58 1.17 0.617 0.492 1.66 0.664 0.293 0.903 1.9 0.281 0.468 0.435 – 0.87 0.385 0.369 0.799 0.556 0.56

0.454 – 1.957 0.34 0.524 0.366 1.34 0.354 – 0.276 1.08 0.758 2.06 0.555 0.658 0.352

The concentrations below LOD were treated as 1/2LOD for arithmetic mean; nd, not detected; -, no sample for this group was analysed.

(Harrad and Abdallah, 2015). Similar results were also observed in 11 pooled human milk samples collected in Ireland, in which levels of BDE-209 were approximately half those of HBCD, but were 30 times

higher than those of TBBPA (Pratt et al., 2013). The current production capacity of TBBPA in China is unknown, but an article published in 2007 reported that the production capacities of TBBPA and PBDEs in

Table 2 Levels (ng/g lw) of TBBPA, HBCD and BDE-209 in human milk from various regions. Region

Year

n

DF (%)

Mean/median

Range

LOD

Reference

TBBPA China UK UK Ireland Japan USA Czech

2011 2011 2010–2011 2013 2005–2006 2004–2005 2010

29 36 120 11b 9 43 50

100 36 61 18 100 35 30

7.58/1.21 0.06/b0.04 0.06/0.04 0.05/1.04/0.72 -/-/-

0.14–66.8 b0.04–0.65 0.03–0.17a b0.29–0.17 0.39–2.22 b0.03–0.55 b2–688

0.06 0.04 0.015 0.29 0.2c 0.03 2

This study Abdallah and Harrad, 2011 Harrad and Abdallah, 2015 Pratt et al., 2013 Fujii et al., 2014 Carignan et al., 2012 Lankova et al., 2013

HBCD China UK UK Ireland USA France Denmark Finland Belgium Sweden Norway South Africa Ghana India Philippines Vietnam USA Canada.

2011 2011 2010–2011 2013 2004–2005 2011–2014 1997–2002 1997–2002 2009–2010 2010 2006 2004 2009 2009 2008 2007 2004–2005 2005

29 36 120 11b 43 41 438 22 1b 30 193 14 16 30 9 43 34

100 100 100 100 100 95.1 98.6 95.5 100 100 67.9 93 100 100 100 100 76

10.1/6.83 5.95/3.83 5.27/4.16 3.45/ -/-/1.47 -/0.31 -/0.31 3.8 -/0.22 1.1/0.54 0.55/0.34 0.8/0.62 2.2/0.61 0.21/0.19 -/2.0 1.02/1.8/0.7

1.02–81.1 1.04–22.37 1.64–15.1a 1.67–5.94 b0.03–0.55 0.45–15.3 0.02–28.7 0.03–2.19 0.07–1.0 0.1–31 b0.23–1.4 0.03–3.2 1.2–13 b0.01–0.91 1.4–7.6 0.36–8.1 0.1–28.2

0.28 0.011 0.2–0.3 0.03 0.08 0.23 0.01 0.05 0.01 0.036 -

This study Abdallah and Harrad, 2011 Harrad and Abdallah, 2015 Pratt et al., 2013 Carignan et al., 2012 Antignac et al., 2016 Antignac et al., 2016 Antignac et al., 2016 Croes et al., 2012 Darnerud et al., 2015 Eggesbo et al., 2011 Darnerud et al., 2011 Asante et al., 2011 Devanathan et al., 2012 Malarvannan et al., 2013 Tue et al., 2010 Carignan et al., 2012 Ryan and Rawn, 2014

BDE-209 China France Denmark UK Sweden Norway New Zealand India Philippines Spain Ireland USA

2011 2011–2014 1997–2002 2010–2011 2010 2006 2010 2009 2008 2004–2008 2013 2003–2005

29 66 199 120 30 46 33 30 290 11b 82

100 100 100 63 100 76.1 97 100 85.5 100 100

0.779/0.555 -/0.21 -/0.34 0.14/0.08 -/0.06 0.5/0.25 0.376/0.191 0.32/0.39 0.5/b0.05 1.03/1.02 1.5/0.77 3.67/1.41

0.276–2.06 0.01–11.6 0.01–14.16 0.05–0.39a 0.02–0.18 0.005–5.8 0.065–3.14 nd-0.75 b0.05–3.4 0.04–6.49 0.37–7.19 0.28–55.3

0.12 0.02 0.05 0.2 -

This study Antignac et al., 2016 Antignac et al., 2016 Harrad and Abdallah, 2015 Darnerud et al., 2015 Eggesbo et al., 2011 Coakley et al., 2013 Devanathan et al., 2012 Malarvannan et al., 2013 Gascon et al., 2012 Pratt et al., 2013 Park et al., 2011

a: 5%ile–95%ile, -: no data; b: pooled sample; c: ng/mL; nd: not detected.

Z. Shi et al. / Science of the Total Environment 599–600 (2017) 237–245

China are approximately 18 000 and 36 000 t per year, respectively (Jiang, 2007). However, since the restriction of octa- and penta-BDE in China, demand for TBBPA sharply increases in recent years. Thus, we infer that the current production capacity of TBBPA has come to surpass that of PBDEs. Unlike PBDEs, data on the occurrence of TBBPA in human milk is limited. In general, the concentration and detection frequency of TBBPA in human body fluid are relatively low, because the bioaccumulation potential of TBBPA is relatively low, and it is normally used as a reactive flame retardant. Levels of TBBPA in human milk from various regions were compared in Table 2. In a study conducted in Birmingham, UK, the levels of TBBPA were detected above the LOD in only 36% of 34 human milk samples with a mean value of 0.06 ng/g lw, which were far lower than our results (Abdallah and Harrad, 2011), another study conducted in Birmingham also showed similar results, in 120 samples from 10 donors, only half of which were found above the LOD, with a median TBBPA level of 0.04 ng/g lw (Harrad and Abdallah, 2015). TBBPA levels in our study were also higher than those in studies conducted in Ireland and Japan. In a study conducted in Boston, USA, low detection frequency (35%) and low levels of TBBPA (range ND0.55 ng/g lw) in 45 human milk samples were reported, this is the only study from North America (Carignan et al., 2012). In 50 human milk samples collected in Czech Republic, TBBPA was found in only 30% of the samples with a wide concentration range (b2–688 ng/g lw), however, mean and median levels were not reported (Lankova et al., 2013). No other studies on TBBPA in human milk have been found in the literature since 2010. Based on the limited data, it is obvious that TBBPA levels in our study were higher than those in studies conducted in other countries. TBBPA is a reactive FR in most circuit boards but is additive in plastic. When using as a reactive BFR, TBBPA is chemically bound to the polymer structure and thus the leaching/release into the environment is limited. In addition, as a phenolic substance, TBBPA can be rapidly conjugated in the human liver and subsequently excreted in the bile (Schauer et al., 2006). Furthermore, the biological half-life of TBBPA is relatively short with approximately only 2 days (Jakobsson et al., 2002). Taken together, TBBPA may accumulate in humans, but continuous exposure is required to maintain a detectable level in a human subject. Thus, the contamination levels of TBBPA in human milk most likely reflect recent rather than past exposure. That is to say, the relatively high TBBPA levels in our study may indicate that the Chinese population is continuously exposed to relatively high levels of TBBPA. Liu et al. (2016) concluded that China is one the most polluted regions affected by TBBPA due to the “irregular and uncontrolled use”. Therefore, further investigations of the sources, fates and health effects of TBBPA in China is a huge and urgent task, more importantly, critical limits of TBBPA should be set to restrict unnecessary release of this pollutant to the environment. 3.1.2. HBCD and isomer profile HBCD (sum of α-, β- and γ-HBCD) was detected above the LOD in all samples with levels ranged from 1.02 to 81.1 ng/g lw, and with mean and median levels of 10.1 and 6.83 ng/g lw, respectively, which were higher than those of TBBPA as well as BDE-209, even if the production and application quantities of HBCD were lower than TBBPA and PBDEs in China (Jiang, 2007). This finding is thought to be a result of the chemical properties and bioaccumulation potential of HBCD. HBCD is used as an additive BFR, and it is hydrophobic, lipophilic and bioaccumulative (average half-life of 64 days in non-occupationally exposed humans) with a much longer half-life than TBBPA (Schecter et al., 2012). Similar results were also found in some other studies (Harrad and Abdallah, 2015; Pratt et al., 2013), in these studies HBCD was also found to be the predominant present in human milk. In commercial HBCD and environmental samples (sludge, soil, etc.), γ-HBCD is normally the predominant isomer, followed by α-HBCD and β-HBCD. However, in human bodily fluids (breast milk, serum, etc.) and animal tissues (muscle, liver, etc.), α-HBCD is usually found to be

241

predominant, followed by γ-HBCD, and β-HBCD was always the lowest (Barghi et al., 2016; Fernandes et al., 2016; Shi et al., 2013a). The predominance of α-HBCD in biota samples is most likely due to the selective metabolism or biotransformation of the three isomers (Du et al., 2012; Erratico et al., 2016). In the 29 samples in our study, α-HBCD comprises 4% to 93% of total HBCDs with an mean of 56% (γ-HBCD, 39%; β-HBCD, 5%). However, exceptions were observed in several samples, such as the rural Heilongjiang sample, the urban Ningxia sample, etc. In these samples, γ-HBCD is the predominant isomer. Particularly, the urban Heilongjiang sample showed the highest contamination level among all samples (81.1 ng/g lw). In addition, in this sample the levels of the three isomer, α-, β- and γ-HBCD, were 8.2, 0.391 and 72.6 ng/g lw, respectively. This isomer pattern was different from most other samples but similar to environmental sample and manufactured product. As we mentioned above, the urban sample of Heilongjiang also showed the highest TBBPA level, hence the source of such a high level of BFRs in human milk from Heilongjiang and the reason for the unique pattern should be deep investigated. We inferred that some highly contaminated individual samples in this pooled sample may dramatically increase the result of the pool. In addition, high BFRs levels in some individual samples may indicate that the donors were exposed to elevated level of BFRs. Compared to recent studies (Table 2), HBCD levels in our study are higher than those reported for many other countries, including Denmark, Finland, France, Belgium, Sweden, Norway, Ireland, South Africa, Ghana, dumping sites in India, Philippines, Vietnam, the USA and Canada. The comparison with other studies indicated that various matrices (food, dust, etc.) in China contain elevated levels of HBCD, leading to high contamination levels of HBCD in human body. Exceptionally, our values are similar to or lower than those in studies reported for the UK and Australia. 3.1.3. BDE-209 After the restriction of penta-BDE and octa-BDE in China, deca-BDE has become the only produced and used commercial PBDE at present. In commercial deca-BDE, BDE-209 is the predominant component with a contribution of at least 95%. This is the first time we measured BDE-209 in national human milk survey. Levels of BDE-209 ranged from 0.276 to 2.06 ng/g lw, with mean and median levels of 0.799 and 0.566 ng/g lw, respectively. The highest level of BDE-209 was found in the urban Sichuan sample. In some previous studies which testing PBDEs in non-e-waste regions, high PBDE concentrations were normally found in samples from large cities with rapid industrial development and high economic level, release of PBDEs from domestic products and/or office equipment containing PBDEs was thought to be a main source (Yu et al., 2016). However, the levels of BDE-209 as well as those of TBBPA and HBCD in the present study do not necessarily follow the above-mentioned principles. No obvious spatial distribution pattern in China was found in this study. One possibility is that the pooling approach would result in averages which are influenced by outliers. In addition, based on Mann-Whitney test, a comparison between rural samples and urban samples were conducted (Supplementary Fig. A1), although median levels of urban samples were normally lower than rural samples, no significant difference in concentrations was found between the urban and rural samples (by SPSS 20, p b 0.05). In comparison, the levels of BDE-209 in human milk were lower than those of both HBCD and TBBPA in this study. Similar to TBBPA, because of the short half-life (15 days) of BDE-209 in human blood (Thuresson et al., 2006), the contamination level of BDE-209 in human tissues has long been seen as an indicator of recent exposure to deca-BDE (Sudaryanto et al., 2008). Higher BDE-209 levels than TBBPA in this study may suggest that the production and application of deca-BDE in China were below that of TBBPA in the past several years, resulted in lower levels of BDE-209 in human body. With the restriction of PBDEs, we predict that the production and application of TBBPA and HBCD in China may continue to increase, thereupon the contamination level

242

Z. Shi et al. / Science of the Total Environment 599–600 (2017) 237–245

would also continue to increase, and we should pay more attention to the environmental fates and health effects of TBBPA and HBCD. Although the BDE-209 levels were lower than TBBPA and HBCD in this study, when comparing with recent studies from other countries (Table 2), BDE-209 levels in our study were found to be higher than those reported for many other countries, including Denmark, France, UK, Sweden, Norway, New Zealand, India and Philippines. However, our results were slightly lower than Spain and Ireland. Data of BDE209 from USA is extremely limited, compared to BDE-209 levels in 82 human milk samples collected in California (mean: 3.67 ng/g lw, range: 0.28–55.3 ng/g lw), our results were lower than those in the USA study; moreover, in the USA study, BDE-209 only accounted for approximately 3% of the total PBDEs, obviously, due to the high market demand for PBDEs in the USA, the contamination levels of PBDEs in USA were much higher than other countries (Park et al., 2011). In summary, the contamination levels of TBBPA, HBCD and BDE-209 in human milk in our study were higher than those in most other studies conducted in general populations worldwide, pooling approach used in the present study might be one of the sources of high contamination level in our samples, whereas our results might also indicate that China has become one the most polluted regions affected by these BFRs. Bi et al. (2006) inferred that the domestic demand for BFRs increased by 8% annually with the rapid development of industry in China in the past decade. Apart from mass production and application, e-waste also brought serious BFRs pollution to China. According to the comparisons between our results and those from other countries, we could arrive at a conclusion that China is now facing serious environmental problems with BFRs, hence the monitoring of BFRs should be continued for a long time, and what's more, an intervention should be performed as soon as possible. 3.2. Estimated daily intake (EDI) and risk assessment 3.2.1. EDIs via human milk for nursing infants EDIs for nursing infants (1–6 months old) were calculated based on BFR levels in human milk (ng/mL), infant ingestion data (750 mL/day) and body weight data from the Chinese Dietary Reference Intakes (Edition 2013) suggested by the Chinese Nutrition Society, assuming that human milk is the only food source for a 1–6 months old nursing infant. As the levels of β- and γ-HBCD in certain samples were below the LOD, bLOD was replaced by zero when calculating the lower bound (LB) EDI. The results are listed in Table 3. The mean lower bound EDI of TBBPA, HBCD and BDE-209 were 235, 310 and 21.9 ng/day, respectively. As the 50th percentile body weight for a 1–6 months old infant is 6 kg, the mean lower bound EDI normalized by body weight for TBBPA, HBCD and BDE-209 were 39.2, 51.7 and 3.65 ng/kg bw/day, respectively, which were much higher than the EDIs via foods for adults (b 2 ng/kg bw/day) in TDS 2011, suggesting a far higher BFR body burden for nursing infants. When calculating the upper bound EDI, 3th percentile of infant body weight at 6 weeks was used (3.94 kg), the max upper bound EDIs for TBBPA, HBCD and BDE-209 were 570, 692 and 11.8 ng/kg bw/day, respectively (Table 3). Table 3 EDIs of each BFR via human milk in China (ng/kg bw/day). TBBPA Mean ± SDa Mediana Rangea Mean ± SDb Medianb Rangeb a

39.2 ± 88

α-HBCD

19.4 ± 16 5.57 13.7 0.823–374 4.24–74.8 59.6 29.5 ± 130 ± 24 8.48 20.9 1.25–570 6.46–114

β-HBCD

γ-HBCD

HBCD

BDE-209

1.66 ± 3.2 0.688 0–17 2.53 ± 4.9 1.05 0–25.9

30.7 ± 76 7.29 0–406 46.8 ± 120 11.1 0–619

51.7 ± 82 26.3 6.49–454 78.8 ± 120 40 9.89–692

3.65 ± 2.2 2.77 1.26–9.61 5.56 ± 3.3 4.22 1.92–11.8

Lower bound EDI, 50th percentile infant body weight (6 kg) was used for calculating. Upper bound EDI, 3th percentile of infant body weight at 6 weeks (3.94 kg) was used for calculating. b

3.2.2. Risk assessment The margin of exposure (MOE) approach was recommended by European Food Safety Authority (EFSA) for evaluating potential health risks from dietary intake of BFRs. MOE was calculated by comparing the BMDL10 (benchmark dose lower confidence limit for a benchmark response of 10%) to the EDI of each BFR. The EFSA Panel on Contaminants in the Food Chain identified effects on thyroid hormones as the critical endpoint for TBBPA (EFSA, 2011c) and neurodevelopment as the critical endpoint for HBCD and BDE-209 (EFSA, 2011a; EFSA, 2011b). Based on chronic, oral exposure associated with body burdens, EFSA suggested BMDLs10 of 16, 0.79 and 1.7 mg/kg bw/day for TBBPA, HBCD and BDE-209, respectively (EFSA, 2011a; EFSA, 2011b; EFSA, 2011c). For TBBPA, when mean (39.2 ng/kg bw/day) EDI of lower bound and max (570 ng/kg bw/day) EDI of upper bound were compared with BMDL10 to determine the MOEs, the results were 4.1 × 105 and 2.8 × 104, respectively. Although the threshold MOE of 100 suggested by EFSA is insufficient to account for uncertainties and variability due to deficiencies in the toxicological database for TBBPA (EFSA, 2011c), such high MOEs estimated in our study indicated that exposure to TBBPA via human milk ingestion does not raise a health concern for nursing infants. Similar conclusions were reported by some other studies. In EFSA's report, EDIs for high milk consuming infants ranged from 0.41 to 257 ng/kg bw/day, resulting in MOEs of 4 × 107 to 6 × 104, which was also unlikely to raise significant health risk (EFSA, 2011c). MOEs estimated in a study conducted in USA were 8.7 × 105 to 1.1 × 108, which were much higher than those in our results, but equally such high MOEs indicated a low level of health concern. For HBCD, mean and max MOEs were 1.5 × 104 and 1.1 × 103, respectively, which is also unlikely to pose a health risk, as the lowest estimated MOE is still several orders of magnitude higher than the threshold of 8 set by the EFSA (EFSA, 2011a). For BDE-209, mean and max MOEs were 4.7 × 105 and 1.4 × 105, compared with threshold of 2.5 set by EFSA (EFSA, 2011b), it is obviously that intake of BDE-209 via human milk is also unlikely to raise a health concern. In summary, based on the observed MOEs in the present study, it is unlikely that the dietary exposure of the Chinese nursing infants to TBBPA, HBCD and BDE-209 would pose a health risk, even if China has become one the most polluted regions affected by BFRs. 3.3. Temporal trends of TBBPA and HBCD from 2007 to 2011 This is the second time we measured TBBPA and HBCD in the national human milk survey, the first survey was conducted in 2007. However, the first survey was performed in only 12 provinces (HLJ, HeB, LN, SX, HeN, NX, FJ, SH, JX, SC, HuB and GX) (Shi et al., 2009). When the median levels were used for comparison, median levels of TBBPA and HBCD in the whole samples increased from 0.482 and 0.993 ng/g lw in 2007 to 1.21 and 6.83 ng/kg bw/day in 2011, respectively. That is, contamination levels of TBBPA and HBCD in human milk in 2011 were found to be approximately 3 to 7 times higher than those in 2007, showing a distinct increase of TBBPA and HBCD in human body. However, when the levels of TBBPA and HBCD in the same 12 provinces between 2007 and 2011 were compared (Fig. 2 and Supplementary Fig. A2), the temporal trends of TBBPA and HBCD were quite different. In detail, for TBBPA, the levels in some provinces, including HLJ, LN, FJ, JX and SC, jumped sharply from 2007 to 2011; in some provinces, such HuB and NX, a slightly increase from 2007 to 2011 was observed; whereas in other provinces, the contamination levels between 2007 and 2011 were similar, or even decreased from 2007 to 2011. In fact, the median of 2011 was only 2.5 times higher than that of 2007. Obviously, in the national survey of 2011, high contamination levels of TBBPA were observed in some areas of China, indicating that these areas were heavily contaminated with TBBPA during 2007 and 2011, whereas in some areas the contamination levels were stable through this period. For HBCD, the temporal trend was another case, rising HBCD levels were observed in each of the 12 provinces from 2007 to 2011, the mean and median levels of

Z. Shi et al. / Science of the Total Environment 599–600 (2017) 237–245

243

Fig. 2. Temporal trends of TBBPA and HBCD levels in the same 12 provinces from 2007 to 2011.

HBCD in 2011 were both far higher than those of 2007, suggested a rapid increase in production and/or application for HBCD between 2007 and 2011 throughout China. In summary, since the legislative focus is still firmly on PBDEs in China, TBBPA has received little attention until now; especially after the EU official approval of the use of TBBPA as a safe FR, the production and application of TBBPA obviously increased, resulted in rapidly increase of TBBPA levels in some areas of China (Liu et al., 2016). Thus, we suggest that the source of TBBPA in these high polluted areas should be further studied, and based on the source analysis, critical limits should be set to restrict unnecessary release of TBBPA. HBCD was listed as a POP in 2013, whereas the samples tested in this study were collected in 2011, and thus HBCD was still in use during sample collection. Therefore, we conclude that the production and application of HBCD also continued to increase between 2007 and 2011, resulting in higher levels of HBCD in 2011. However, although the application of HBCD has been banned in China since 2016, the use of HBCD in building materials is still permitted (www.chinasafety.gov.cn). That is to say, the production and application of HBCD continues in China. Due to its high lipophilicity and long half-life, and the rising contamination levels observed in this national survey, we do suggest that an overall restriction of HBCD should be performed in China immediately.

milk samples. Levels of DBDPE were not only several orders of magnitude higher than those of other novel BFRs, but also significant higher than that of BDE-209. Compared to TBBPA and HBCD, median level of DBDPE was slightly lower than that of HBCD but obvious higher than TBBPA, indicating that the production and application of DBDPE has outperformed most of the currently used BFRs, resulting in high contamination level in human milk. DBDPE was introduced into market as a replacement for PBDEs since the end of last century, and it was produced and used in increasing amounts (Covaci et al., 2011). In China, DPDPE has been produced only since 2005 but with production increasing 80% per year (www.polymer.cn). Currently, DBDPE is the most popular NBFR in Chinese market with a production capacity of 25,000 t in 2012; this volume might have outperformed that of other BFRs to make DBDPE the most produced and used BFR at present (Zhang and

3.4. A comparison to novel BFRs In this study, six currently used novel BFRs (NBFRs), including 1,2bis(2,4,6-tribromophenoxy)-ethane (BTBPE), 2,3-dibromopropyl2,4,6-tribromophenyl ether (DPTE), hexabromobenzene (HBB), pentabromotoluene (PBT), pentabromoethylbenzene (PBEB) and decabromodiphenyl ethane (DBDPE), were also measured in the same 29 pooled samples (Shi et al., 2016). A comparison between the legacy BFRs and the novel BFRs is shown in Fig. 3. Obviously, the mean and median levels of the three legacy BFRs, TBBPA, HBCD and BDE-209, were still far higher than those of the novel BFRs (except for DBDPE). Particularly, levels of DPDPE, with mean and median of 8.06 and 5.8 ng/g lw, respectively, showed high contamination levels in these human

Fig. 3. A comparison between legacy and novel BFRs in the national human milk survey in 2011.

244

Z. Shi et al. / Science of the Total Environment 599–600 (2017) 237–245

Gu, 2013; Zhang and Lu, 2011). In summary, the comparison between the legacy and novel BFRs indicated an obvious shift in the consumption pattern between PBDE and non-PBDE BFRs in China. 4. Conclusion A national survey of three currently used BFRs, TBBPA, HBCD and BDE-209, in human milk was presented in this report. Based on the measurement of human milk collected from various areas of China, occurrence, estimated daily intake and related health concern of the three BFRs were assessed. The levels of BDE-209 were found to be lower than those of TBBPA and HBCD, as well as DBDPE, suggesting an obvious shift in the consumption pattern between PBDE and non-PBDE BFRs in China. Furthermore, the levels of TBBPA and HBCD were found to increase from 2007 to 2011, resulting in a rising contamination levels of these two BFRs. Due to the phase out of PBDEs, the demands for TBBPA and HBCD increased rapidly, both levels and temporal trends for TBBPA and HBCD in this study confirm their relatively high current use in flame retardant market. Since as of now there is no restriction on TBBPA and the use of HBCD in building materials is still permitted in China, we predict that the contamination levels of TBBPA and HBCD in China may continue to increase in the future. In comparison, the levels of the three BFRs were higher than most other studies worldwide, suggesting that China is now facing serious environmental problems with BFRs because of the mass production and application, along with the improper recycling of e-wastes. Therefore, exposure and risk assessment for BFRs should be continued, especially for those with high and rising contamination levels, including TBBPA, HBCD and DBDPE. However, based on the large MOEs observed in the present study, it is unlikely that the current dietary intake of TBBPA, HBCD and BDE-209 via human milk for Chinese nursing infants would pose a health risk. Conflict of interest statements The authors have no conflict of interest to declare. Acknowledgements The authors thank all the human milk donors. This study was supported by the National Natural Science Foundation of China (21477083, 21537001, 21507018); Beijing Natural Science Foundation (7122022); the Importation and Development of High-Caliber Talents Project of Beijing (CIT&TCD201404190); the Development of Outstanding Talents Project of Beijing (2013D005018000008). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2017.04.237. References Abdallah, M.A., Harrad, S., 2011. Tetrabromobisphenol-A, hexabromocyclododecane and its degradation products in UK human milk: relationship to external exposure. Environ. Int. 37, 443–448. Antignac, J.P., Main, K.M., Virtanen, H.E., Boquien, C.Y., Marchand, P., Venisseau, A., et al., 2016. Country-specific chemical signatures of persistent organic pollutants (POPs) in breast milk of French, Danish and Finnish women. Environ. Pollut. 218, 728–738. Asante, K.A., Adu-Kumi, S., Nakahiro, K., Takahashi, S., Isobe, T., Sudaryanto, A., et al., 2011. Human exposure to PCBs, PBDEs and HBCDs in Ghana: temporal variation, sources of exposure and estimation of daily intakes by infants. Environ. Int. 37, 921–928. Barghi, M., Shin, E.S., Son, M.H., Choi, S.D., Pyo, H., Chang, Y.S., 2016. Hexabromocyclododecane (HBCD) in the Korean food basket and estimation of dietary exposure. Environ. Pollut. 213, 268–277. Bi, X., Qu, W., Sheng, G., Zhang, W., Mai, B., Chen, D., et al., 2006. Polybrominated diphenyl ethers in South China maternal and fetal blood and breast milk. Environ. Pollut. 144, 1024–1030. Carignan, C.C., Abdallah, M.A., Wu, N., Heiger-Bernays, W., McClean, M.D., Harrad, S., et al., 2012. Predictors of tetrabromobisphenol-A (TBBP-A) and hexabromocyclododecanes (HBCD) in milk from Boston mothers. Environ. Sci. Technol. 46 (21), 12146–12153.

Coakley, J.D., Harrad, S.J., Goosey, E., Ali, N., Dirtu, A.C., Van den Eede, N., et al., 2013. Concentrations of polybrominated diphenyl ethers in matched samples of indoor dust and breast milk in New Zealand. Environ. Int. 59, 255–261. Covaci, A., Harrad, S., Abdallah, M.A., Ali, N., Law, R.J., Herzke, D., et al., 2011. Novel brominated flame retardants: a review of their analysis, environmental fate and behaviour. Environ. Int. 37, 532–556. Croes, K., Colles, A., Koppen, G., Govarts, E., Bruckers, L., Van de Mieroop, E., et al., 2012. Persistent organic pollutants (POPs) in human milk: a biomonitoring study in rural areas of Flanders (Belgium). Chemosphere 89, 988–994. Darnerud, P.O., Aune, M., Larsson, L., Lignell, S., Mutshatshi, T., Okonkwo, J., et al., 2011. Levels of brominated flame retardants and other pesistent organic pollutants in breast milk samples from Limpopo province, South Africa. Sci. Total Environ. 409, 4048–4053. Darnerud, P.O., Lignell, S., Aune, M., Isaksson, M., Cantillana, T., Redeby, J., et al., 2015. Time trends of polybrominated diphenylether (PBDE) congeners in serum of Swedish mothers and comparisons to breast milk data. Environ. Res. 138, 352–360. Devanathan, G., Subramanian, A., Sudaryanto, A., Takahashi, S., Isobe, T., Tanabe, S., 2012. Brominated flame retardants and polychlorinated biphenyls in human breast milk from several locations in India: potential contaminant sources in a municipal dumping site. Environ. Int. 39, 87–95. Du, M., Lin, L., Yan, C., Zhang, X., 2012. Diastereoisomer- and enantiomer-specific accumulation, depuration, and bioisomerization of hexabromocyclododecanes in zebrafish (Danio rerio). Environ. Sci. Technol. 46, 11040–11046. Dunnick, J.K., Morgan, D.L., Elmore, S.A., Gerrish, K., Pandiri, A., Ton, T.V., et al., 2017. Tetrabromobisphenol A activates the hepatic interferon pathway in rats. Toxicol. Lett. 266, 32–41. EFSA, 2011a. Scientific opinion on hexabromocyclododecanes (HBCDDs) in food. EFSA J. 9, 2296 (118pp). EFSA, 2011b. Scientific opinion on polybrominated diphenyl ethers (PBDEs) in food. EFSA J. 9, 2156 (274pp). EFSA, 2011c. Scientific opinion on tetrabromobisphenol A (TBBPA) and its derivatives in food. EFSA J. 9, 2477 (67pp). Eggesbo, M., Thomsen, C., Jorgensen, J.V., Becher, G., Odland, J.O., Longnecker, M.P., 2011. Associations between brominated flame retardants in human milk and thyroidstimulating hormone (TSH) in neonates. Environ. Res. 111, 737–743. Erratico, C., Zheng, X., van den Eede, N., Tomy, G., Covaci, A., 2016. Stereoselective metabolism of alpha-, beta-, and gamma-hexabromocyclododecanes (HBCDs) by human liver microsomes and CYP3A4. Environ. Sci. Technol. 50, 8263–8273. Fernandes, A.R., Mortimer, D., Rose, M., Smith, F., Panton, S., Garcia-Lopez, M., 2016. Bromine content and brominated flame retardants in food and animal feed from the UK. Chemosphere 150, 472–478. Fromme, H., Becher, G., Hilger, B., Volkel, W., 2016. Brominated flame retardants - exposure and risk assessment for the general population. J Hyg Environ Health]–>Int. J. Hyg. Environ. Health 219, 1–23. Fujii, Y., Nishimura, E., Kato, Y., Harada, K.H., Koizumi, A., Haraguchi, K., 2014. Dietary exposure to phenolic and methoxylated organohalogen contaminants in relation to their concentrations in breast milk and serum in Japan. Environ. Int. 63, 19–25. Gascon, M., Fort, M., Martinez, D., Carsin, A.E., Forns, J., Grimalt, J.O., et al., 2012. Polybrominated diphenyl ethers (PBDEs) in breast milk and neuropsychological development in infants. Environ. Health Perspect. 120, 1760–1765. Harrad, S., Abdallah, M.A., 2015. Concentrations of polybrominated diphenyl ethers, hexabromocyclododecanes and tetrabromobisphenol-A in breast milk from United Kingdom women do not decrease over twelve months of lactation. Environ. Sci. Technol. 49, 13899–13903. Jakobsson, K., Thuresson, K., Rylander, L., Sjödin, A., Hagmar, L., Bergman, Å., 2002. Exposure to polybrominated diphenyl ethers and tetrabromobisphenol A among computer technicians. Chemosphere 46, 709–716. Jiang, Y., 2007. The status quo and trend of brominated flame retardants. Chin. J. Flame Retardant 2, 1–7. Lankova, D., Lacina, O., Pulkrabova, J., Hajslova, J., 2013. The determination of perfluoroalkyl substances, brominated flame retardants and their metabolites in human breast milk and infant formula. Talanta 117, 318–325. Li, X., Wu, Y., Chen, J., 2011. The development and evolution of Chinese Total Diet Study. Chin. J. Epidemiol. 32, 456. Liu, K., Li, J., Yan, S., Zhang, W., Li, Y., Han, D., 2016. A review of status of tetrabromobisphenol A (TBBPA) in China. Chemosphere 148, 8–20. Malarvannan, G., Isobe, T., Covaci, A., Prudente, M., Tanabe, S., 2013. Accumulation of brominated flame retardants and polychlorinated biphenyls in human breast milk and scalp hair from the Philippines: levels, distribution and profiles. Sci. Total Environ. 442, 366–379. Park, J.S., She, J., Holden, A., Sharp, M., Gephart, R., Souders-Mason, G., et al., 2011. High postnatal exposures to polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls (PCBs) via breast milk in California: does BDE-209 transfer to breast milk? Environ. Sci. Technol. 45, 4579–4585. Pratt, I., Anderson, W., Crowley, D., Daly, S., Evans, R., Fernandes, A., et al., 2013. Brominated and fluorinated organic pollutants in the breast milk of first-time Irish mothers: is there a relationship to levels in food? Food Addit. Contam. Part A Chem. Anal. Control Expo. Risk Assess. 30, 1788–1798. Ryan, J.J., Rawn, D.F., 2014. The brominated flame retardants, PBDEs and HBCD, in Canadian human milk samples collected from 1992 to 2005; concentrations and trends. Environ. Int. 70, 1–8. Schauer, U.M., Volkel, W., Dekant, W., 2006. Toxicokinetics of tetrabromobisphenol A in humans and rats after oral administration. Toxicol. Sci. 91, 49–58. Schecter, A., Szabo, D.T., Miller, J., Gent, T.L., Malik-Bass, N., Petersen, M., et al., 2012. Hexabromocyclododecane (HBCD) stereoisomers in U.S. food from Dallas, Texas. Environ. Health Perspect. 120, 1260–1264.

Z. Shi et al. / Science of the Total Environment 599–600 (2017) 237–245 Shi, Z.X., Wu, Y.N., Li, J.G., Zhao, Y.F., Feng, J.F., 2009. Dietary exposure assessment of Chinese adults and nursing infants to tetrabromobisphenol-A and hexabromocyclododecanes: occurrence measurements in foods and human milk. Environ. Sci. Technol. 43, 4314–4319. Shi, Z., Jiao, Y., Hu, Y., Sun, Z., Zhou, X., Feng, J., et al., 2013a. Levels of tetrabromobisphenol A, hexabromocyclododecanes and polybrominated diphenyl ethers in human milk from the general population in Beijing, China. Sci. Total Environ. 452-453, 10–18. Shi, Z., Wang, Y., Niu, P., Wang, J., Sun, Z., Zhang, S., et al., 2013b. Concurrent extraction, clean-up, and analysis of polybrominated diphenyl ethers, hexabromocyclododecane isomers, and tetrabromobisphenol A in human milk and serum. J. Sep. Sci. 36, 3402–3410. Shi, Z., Zhang, L., Li, J., Zhao, Y., Sun, Z., Zhou, X., et al., 2016. Novel brominated flame retardants in food composites and human milk from the Chinese Total Diet Study in 2011: concentrations and a dietary exposure assessment. Environ. Int. 96, 82–90. Sudaryanto, A., Kajiwara, N., Takahashi, S., Muawanah, Tanabe S., 2008. Geographical distribution and accumulation features of PBDEs in human breast milk from Indonesia. Environ. Pollut. 151, 130–138. Thuresson, K., Hoglund, P., Hagmar, L., Sjodin, A., Bergman, A., Jakobsson, K., 2006. Apparent half-lives of hepta- to decabrominated diphenyl ethers in human serum as determined in occupationally exposed workers. Environ. Health Perspect. 114, 176–181. Toms, L.-M.L., Guerra, P., Eljarrat, E., Barceló, D., Harden, F.A., Hobson, P., et al., 2012. Brominated flame retardants in the Australian population: 1993–2009. Chemosphere 89, 398–403.

245

Tue, N.M., Sudaryanto, A., Minh, T.B., Isobe, T., Takahashi, S., Viet, P.H., et al., 2010. Accumulation of polychlorinated biphenyls and brominated flame retardants in breast milk from women living in Vietnamese e-waste recycling sites. Sci. Total Environ. 408, 2155–2162. UNEP/POPS, 2013. Report of the Conference of the Parties to the Stockholm Convention on Persistent Organic Pollutants on the Work of Its Sixth Meeting. http://chm.pops. int/TheConvention/ConferenceoftheParties/ReportsandDecisions/tabid/208/Default. aspx. UNU, 2015. The Global E-waste Monitor 2014: Quantities, Flows and Resources. http://i. unu.edu/media/unu.edu/news/52624/UNU-1stGlobal-E-Waste-Monitor-2014-small. pdf. Yu, G., Bu, Q., Cao, Z., Du, X., Xia, J., Wu, M., et al., 2016. Brominated flame retardants (BFRs): a review on environmental contamination in China. Chemosphere 150, 479–490. Zhang, S., Gu, X.Y., 2013. Analysis and forcast of flame retardant decabromdiphenyl ethame in market. Chin. J. Plast. Addit. 12–15. Zhang, X.Y., Lu, Q.Y., 2011. Production status and developing prospects of flame retardants. China Plast. Ind. 39, 1–5. Zhang, L., Li, J., Zhao, Y., Li, X., Yang, X., Wen, S., et al., 2011. A national survey of polybrominated diphenyl ethers (PBDEs) and indicator polychlorinated biphenyls (PCBs) in Chinese mothers' milk. Chemosphere 84, 625–633.