Prepared by: Prepared for: September 09, 2013

INFANT/TODDLER HEALTH RISKS FROM EXPOSURE TO POLYBROMINATED DIPHENYL ETHERS (PBDES) IN CAR SEATS AND AUTOMOTIVE UPHOLSTERY

Prepared by: Jeff Fowles Tox-Logic Consulting, LLC Petaluma, CA 94954 David Morgott Pennsport Consulting, LLC Philadelphia, PA 19147

Prepared for: Matthew Ashworth Institute of Environmental Science and Research, Ltd Christchurch, New Zealand

September 09, 2013 Report No FW13051

Table of Contents FOREWORD .......................................................................................................... 6 SUMMARY ............................................................................................................. 7 1.

BACKGROUND .............................................................................................. 9 1.1.

PBDES IN NEW ZEALAND.......................................................................... 11

2.

FRAMEWORK .............................................................................................. 15

3.

LITERATURE SEARCH ............................................................................... 17 1.2. 1.3. 1.4.

DUST LEVELS ........................................................................................... 18 AIR LEVELS .............................................................................................. 22 TRAVEL TIME ............................................................................................ 24

4.

BODY BURDEN............................................................................................ 27

5.

HAZARD ASSESSMENT ............................................................................. 32 5.1. EXPERIMENTAL ANIMAL DATA .................................................................... 32 5.1.1. Neurodevelopment ............................................................................. 32 5.1.2. Endocrine effects ................................................................................ 35 5.1.3. Carcinogenicity ................................................................................... 36 5.2. HUMAN DATA ............................................................................................ 38 5.3. THE SIGNIFICANCE OF BDE 209 ................................................................ 40 5.4. SELECTION OF HAZARD VALUES FOR RISK ASSESSMENT ............................... 41

6.

RISK DETERMINATION ............................................................................... 44

7.

REFERENCES .............................................................................................. 51

APPENDIX A ....................................................................................................... 66 APPENDIX B ....................................................................................................... 72 APPENDIX C ....................................................................................................... 75 APPENDIX D ....................................................................................................... 77

ACKNOWLEDGMENTS ESR acknowledge the work done by Jeff Fowles (ToxLogic Consulting, LLC) and David Morgott (Pennsport Consulting, LLC) in the research and preparation of this report.

List of Tables Table 1

Market demand for PBDEs in 2001. ............................................................10

Table 2

Percentage of car seats free from bromine and brominated flame retardants......................................................................................................13

Table 3

Characteristics of the PBDE congener categories. ...................................15

Table 4

Results from the Web of Science search for published exposure information specific to a car interior. .........................................................17

Table 5

Descriptive statistics for the range of reported dust concentrations inside cars. ...................................................................................................22

Table 6

Descriptive statistics for the range of reported air levels inside cars. ....24

Table 7

Studies evaluating the amount of time an infant or child spends inside a car. ..................................................................................................25

Table 8

Mean and 95% confidence limits for the range of reported values for amount of time a child spends in an automobile. .....................................26

Table 9

Physiological and pharmacokinetic parameters used to determine whole body uptake. ......................................................................................28

Table 10 Uptake of PBDE congeners by the inhalation route. .................................30 Table 11 Uptake of PBDE congeners by the oral route. ...........................................30 Table 12 Uptake of PBDE congeners by the dermal route. ......................................31 Table 13 Reported key neurodevelopmental effects in animal models. .................34 Table 14 Summary of tumour findings in NTP carcinogenicity study on BDE 209. ................................................................................................................37 Table 15 Blood PBDE data (ng/g lipid) from the U.S. and Canadian populations. .39 Table 16 Summary of key PBDE hazard values. .......................................................42 Table 17 Hazard values used for screening level risk assessment. .......................44 Table 18 Compilation of studies investigating the uptake of PBDEs by various routes. ...........................................................................................................46 Table 19 Total uptake and risk from exposure to PBDE congener categories found inside automobiles. ...........................................................................48

Table 20 Adjusted uptake and risk from car seat exposure to eight PBDE congener categories. ...................................................................................49

List of Figures Figure 1

General Structure of PBDEs. .........................................................................9

Figure 2

Total PBDE levels (ng/g) in house dust from different countries. ...........12

Figure 3

Tiered approach to the determination of health risk from exposures to chemical mixtures (WHO, 2009a). ...............................................................16

Figure 4

Example distribution plot for published dust concentrations for hexaBDE. ......................................................................................................21

PDBE Risk for Infants/Toddlers

Page 5

Foreword This report was commissioned by ESR for the Ministry of Health under the Hazardous Substances Prioritisation Project during the 2012–2013 financial year. International interest in polybrominated diphenyl ethers has recently grown due to their persistent nature and their potential biological activity. Particular attention has been paid to the exposure routes for infants and children. This report examines the health risk which may be attributed to PBDE exposure of children from car safety seats and automotive upholstery. This report was prepared for ESR by Tox-Logic Consulting, LLC and Pennsport Consulting, LLC.


Page 6

Summary Polybrominated diphenyl ethers (PBDEs) are commercially produced halogenated compounds that are added to plastics and polyurethane foams to prevent fires. Their occurrence in household dust is well documented, and recent studies also show that they are found in the dust within cars as well. Because of the unique susceptibility of children to toxicants that affect development, a tier 1 risk assessment was performed that focused on oral, dermal, and inhalation exposures of infants/toddlers exposed to PBDEs while restrained in a car seat. The risk analysis was patterned from a scheme recently issued by the International Programme on Chemical Safety and the World Health Organization for investigating health risks from exposures to complex chemical mixtures. For the purposes of this assessment, PBDEs were subdivided into eight congener categories: triBDE, tetraBDE, pentaBDE, heptaBDE, hexaBDE, octaBDE, nonaBDE, and decaBDE. Whereas most of these categories contained multiple congeners, the decaBDE category was limited to a single congener, BDE 209, which has come into widespread use as a result of the ban on products that contain lower molecular weight congeners. The exposure parameters needed to calculate the body burden from the three exposure routes were extracted from published and unpublished sources. Literature searches were performed to ensure that the most recent exposure and hazard information from all international studies, including New Zealand specific data was incorporated into the assessment. The primary factors affecting the uptake of PBDEs during vehicle occupancy were determined to be the air and dust concentrations inside the car and the amount of time a child is restrained in a car seat. All available information for each of these three parameters was compiled and analysed to assess either the upper and lower limits or the 5th and 95th percentiles for the confidence interval. A total of 15 publications were identified that quantified the car dust levels of individual congeners. Another five papers documented the airborne level within this microenvironment, whereas seven literature sources contained time activity data for infants/toddlers that were relevant to the assessment. Following summation and categorisation of the results for individual congeners, the data was statistically analysed to establish the values to be used to compute whole body uptake under low, medium, and high exposure conditions. Although the PBDE measurements appeared to be log-normally distributed, confidence intervals could not be reliably established for dust and air levels due to the high variance for the reported values. The time spent in a car was sufficiently robust that 5th, 50th, and 95th percentile values were calculated. These data along with physiological and pharmacokinetic parameters applicable to infants/toddlers in the 0 to 5 year age range were used to calculate uptake by all three exposure routes and under all three exposure circumstances (high, medium, and low). The total body burden for all three exposure routes was then compared with effect estimates judged to provide the best margin of safety. The reference dose took into consideration observed adverse health effects of PBDEs on the endocrine system and neurological development from animal studies, and reported associations with adverse neurological development from human epidemiological investigations. The U.S. Environmental Protection Agency’s Reference Dose (RfD) values of 0.1 µg/kg/day for tetraBDE, and pentaBDE were determined to be the most applicable for the lower molecular categories and were rationally applied to triBDE through hexaBDE body burden

estimates. Likewise, the RfD of 3 µg/kg/day for the octaBDE was deemed to be also relevant for heptaBDE. The final two categories, nona- and decaBDE were assessed using the RfD value of 7 µg/kg/day for decaBDE. These reference values are based on the most sensitive toxicological endpoints and are supported by recent toxicology data and epidemiological reports of neurobehavioural effects in young children from pre- and post-natal exposures to PBDEs. While a cancer potency slope factor is available from the U.S. EPA for BDE 209, this was not used in the current assessment, due to questions about the nature and relevance of the rodent tumours on which it based, and the lack of corroborating data from other PBDEs. Hazard indices (HI) were calculated for each PBDE congener category using the body burden and effect estimates described above to determine if the ratio was above unity, which would indicate that the exposures were excessive and in need of control. The hazard ratios were calculated assuming that infant/toddler car seats account for 20% of the concentration measurements in car air and dust. This adjustment factor was based on empirical evidence that suggests that the relative surface area of an infant car seat is no more than one-fifth of the total upholstered surface within a small compact automobile. Under these circumstances, the adjusted HI for the eight congener categories ranged from 10-8 to 10-6 for the low exposure scenario to 10-6 to 10-3 for the worst-case scenario. The highest HI’s were observed for the tetraBDE and pentaBDE categories where values of 2.13E-03 and 4.85E-03 were determined. Since these HI values were all less than unity, the overall assessment indicates that PBDE exposures from car seats or the car micro-environment in general, and in isolation of other sources of PBDE exposure, likely do not pose a risk for adverse health effects in infants and toddlers. These conclusions, however, should be interpreted cautiously since they exclude consideration of the background exposures from other indoor and outdoor sources. Likewise, the risk determinations are constrained by the fact that two potential routes of exposure could not be completely quantified (i.e. direct dermal absorption from the seat itself (i.e. not from dust), and indirect pulmonary absorption from inhalation of the dust). These additional routes are not expected to be major contributors to the overall exposure assessment, but could cause slight movement in the estimated uptakes. It has also been assumed, for the purposes of this report, that the high end of the range of published values reported internationally, provides a conservative representation of what could be found in New Zealand car dust PBDE concentrations and uptakes. The validity of this assumption can only be explored through the collection of additional New Zealandspecific data. Any future assessments may need to consider the fact that infant car seats are also used as child carriers outside of the car environment and that the resulting increase in occupancy duration could appreciably impact the contribution of car seats to PBDE body burdens.


Page 8

1. Background Polybrominated diphenyl ethers (PBDEs) are commercial mixtures of persistent, halogenated chemicals added to consumer products, because of their excellent fire suppressing qualities, to reduce flammability, and thereby improve fire safety. The predominant applications of these chemicals are as additives to plastics, foams and other polymers that make up a wide range of consumer products, such as electronics, textiles, and insulating materials that can be found in our homes, offices, schools, and cars. Uses of these products include addition to polyurethane foams (e.g. Penta and TetraBDEs), and hard plastics (Octa and DecaBDEs). The automobile interior is made up of various plastics and foams, and will therefore have potentially different PBDEs available for exposure from various sources within the vehicle. While there are 209 theoretical PBDE congeners, in practice, only a sub-set are produced commercially and found in the environment. The general structure is shown in Figure 1.

Figure 1

General Structure of PBDEs.

Three commercial preparations are produced: PentaBDE, OctaBDE, and DecaBDE. These products consist of various congeners, the most notable and environmentally prevalent being tetrabromodiphenyl ether (BDE 47), pentabromodiphenyl ether (BDE 99 and BDE 100), hexabromodiphenyl ether (BDE 153 and BDE154), heptabromodiphenyl ether (BDE 183), octabromodiphenyl ether (BDE 203), and decabromodiphenyl ether (BDE 209) congeners (Bakker et al., 2008). Although, the demand for brominated flame retardants is generally declining in many regions of the world due to environmental and health concerns, some markets have noted a sizable increase. Manufacturers in the United States and Europe have voluntarily ceased, or agreed to cease, the production and use of commercial grade PBDEs (USEPA, 2010). Since 2004, commercial Penta- and OctaBDE mixtures are no longer used in regulated products. These restrictions resulted in a temporary increase in commercial DecaBDE production in the United States; whereas Europe went on to ban the use of DecaBDE in 2008. The US has now followed suit with manufacturers and importers announcing plans to phase out the production and sales of technical DecaBDE by the end of 2013 (USEPA, 2009). In 2006, PentaBDE was banned in California, which has some of the highest recorded PBDE blood levels in the world, due to the pervasive use of PBDEs to meet stringent fire safety standards of polyurethane foam products sold there under what is called Technical Bulletin 117 ((DCA, 2013)). For furniture makers to comply with TB117, large quantities of flame retardant materials were added to foams and plastics, amounting to sometimes over 5% by weight of the final product. The California standard also carried implications for the U.S. generally, and also other parts of the world, where compliance with California TB117 was placed on labels of items as far away as Taiwan and other Asian countries (Stapleton et al., 2013).


Page 9

None of these restrictions have taken place in Asia where PBDEs still enjoy widespread use. The Asia-Pacific region accounted for 41% of the overall demand for all flame retardants in 2010 and the market is anticipated to grow even larger (Ceresana, 2011). For instance, the use of flame retardants in China is anticipated to increase by more 7% per year through 2018; however the contribution of PBDEs to this overall increase is currently unavailable. In the past, commercial grade DecaBDE has been the dominant form used within China with production increasing from 10,000 tons in 2001 to 30,000 tons in 2005 (Ni et al., 2013). Currently, alternative substances are being produced and applied as substitute flame retardants for PBDEs worldwide (Ali et al., 2012). These substances are not reviewed or assessed in this report. PBDEs produced in Asia are sold worldwide via the many manufactured products that contain them. According to Environment Canada, the worldwide market for PBDEs was approximately 67,390 tons in 2001, including 56,100 tons of DecaBDE, 7,500 tons of PentaBDE and about 3,790 tons of OctaBDE (EC, 2006). For the historical commercial and regulatory reasons mentioned, there are significant differences in the usage of PBDEs by continent (see Table 1). For example, PentaBDE was used almost exclusively in the Americas, while in the Asia Pacific region, DecaBDE is and remains the predominant congener produced and used in commerce. Regional differences in human exposures to PBDEs are therefore expected. Table 1

Market demand for PBDEs in 2001.

AmericasA Commercial Product

EuropeB

AsiaC

Market demand

Estimated consumpti on (tons)

Market demand

Estimated consumption (tons)

Market demand

Estimated consumption (tons)

PentaBDE

95%

7,100

2%

150

3%

250

OctaBDE

40%

1,500

16%

610

44%

1,680

DecaBDE

44%

24,500

13%

7,600

43%

24,050

A

All countries in North, South and Central America were included. All countries in Eastern and Western Europe were included. C Australia, New Zealand and the Indian subcontinent were included. Source: Environment Canada, 2006 B


Page 10

The first applications of PBDEs were in the late 1970’s and early 1980’s, at which time virtually nothing was known about their environmental persistence or health effects. Up until 2005, there were approximately 300 published articles relating to the health and environmental effects of PBDEs. Now there are well over 2,000 publications, largely focused on experimental toxicology and environmental persistence. Comparatively few articles have been available on human health outcomes, although this is now changing rapidly. Among the still emerging reported health effects of PBDEs are interference with brain and neurological development, interference with or mimicking critical hormones, including testosterone, estrogen, and thyroid hormone, and associations with impairment of children’s learning, behaviour and memory, as well as effects on fertility and reproduction. Although population-wide surveys of PBDEs in blood and urine of US residents have been included in the NHANES (National Health and Nutrition Examination Survey) data collected by the US Centers for Disease Control and Prevention (CDC), critical questions remain about the magnitude of human exposures, and from where these exposures originate (CDC, 2009). The biomonitoring studies reveal widespread exposure throughout the population and indicate significant variations in the distributions of population exposures to PBDEs. Analysis of the NHANES 2003-2004 sub-sample showed detection of BDE 47, in nearly all participants and detection of BDE 28, BDE 99, BDE 100, and BDE 153 in greater than 60 percent of participants (Sjodin et al., 2008). It is not yet understood what health effects could be occurring at the doses experienced through normal contact with these substances within households, offices, schools, or cars. Exposures to PBDEs have been reported to be highest for young children aged 1 to 5 years (EC, 2012). While few studies have measured levels of PBDEs in young children, one large study of 2420 individuals, conducted in Australia, found that levels of PBDEs in blood are greatest for children aged 2.3 to 6 years, compared with older children and adults. This implies that exposure routes aside from breast-feeding must contribute significantly to early exposures (Toms et al., 2008, Toms et al., 2009a). Similarly, a study in the United States of 20 young children (ages 1.5 to 4 years) found that the sum of their PBDE blood levels was consistently higher than those of their mothers (Lunder et al., 2010). In this study, increased hand-to-mouth behaviour in children was hypothesised to account for the significantly higher blood levels. Despite the bans placed on Penta, Octa, and DecaBDE products, their long-lived persistent nature in the environment, and continued presence in older furniture and consumer products maintains a significant exposure source and therefore a level of human exposure, as demonstrated by recent national biomonitoring programmes, such as the Health Canada Survey and U.S. NHANES survey of environmental chemicals in the blood the general population (CDC, 2009, EC, 2012). 1.1. PBDEs in New Zealand The use of PBDE flame retardants in New Zealand is not expected to be particularly high since there are no regulations requiring their use in manufactured goods (Keet et al., 2010). A comparison of PBDE levels in house dust revealed similar findings across countries (Ibarra et al., 2013). Figure 2 shows that average total PBDE (ΣPBDE) levels in PDBE Risk for Infants/Toddlers

Page 11

New Zealand house dust of 164 ng/g are comparable to those found in the United Kingdom and Europe. These levels, however, may not reflect current usage patterns since authors now report average PBDE levels in New Zealand homes of 2756 and 3079 ng/g in floor and mattress dust, respectively (Coakley et al., 2013). BDE 209 levels dominated the speciation profile accounting for approximately 90% of the total concentration. A recent investigation into the prevalence of PBDE flame retardants in New Zealand evaluated products used by children in the 0-5 year age group and found limited presence in most footwear and toys with levels not exceeding 0.04% (measured as total bromine (Keet et al., 2010). A notable exception was automobile booster seats which contained PBDE levels ranging from 0.10 - 0.27% (n=4). By comparison, the ΣPBDE content in passenger car seats ranged from 0.16 - 0.70% (n=6). The PBDEs in these samples were associated with the expanded polystyrene (EPS) foam used as a lining or cushioning material under the fabric.

Figure 2

Total PBDE levels (ng/g) in house dust from different countries.

Although, alternative flame retardants (e.g. chlorinated Tris, and alkyl and aryl phosphates) have largely supplanted PBDEs in many products, it is unclear if this also applies strictly to infant and toddler car seats. Many PBDE-containing car seats remain on the market and are sold secondhand or passed on to other family members. Consequently, the potential still exists for infants and toddlers to be overexposed as a result of their contact with the PBDEs in car seat foam (Keet et al., 2010). Although the quantities of specific PBDE congeners were not evaluated in this study, the authors noted that levels of total PBDEs in car seats were appreciably higher than the levels found in other products. In fact, a US study of bromine and brominated flame retardants in various car seat components showed that only 29% were completely free of brominated substances and that 85% of the seats tested contained bromine-related materials in the EPS foam (see Table 2).


Page 12

Table 2 Percentage of car seats free from bromine and brominated flame retardants. Car Seat Component

Bromine/BFR