Skip to content

Environmental Health Perspectives

Facebook Page EHP Twitter Feed Open Access icon  

Research October 2014 | Volume 122 | Issue 10

Email this to someoneShare on FacebookTweet about this on TwitterShare on LinkedInShare on Google+Share on StumbleUpon
Children's Health
Environ Health Perspect; DOI:10.1289/ehp.1307602

Residential Levels of Polybrominated Diphenyl Ethers and Risk of Childhood Acute Lymphoblastic Leukemia in California

Mary H. Ward,1* Joanne S. Colt,1* Nicole C. Deziel,1 Todd P. Whitehead,2 Peggy Reynolds,3 Robert B. Gunier,2 Marcia Nishioka,4 Gary V. Dahl,5 Stephen M. Rappaport,2 Patricia A. Buffler,2 and Catherine Metayer2

Author Affiliations open
1Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Department of Health and Human Services, Bethesda, Maryland, USA; 2University of California, Berkeley, School of Public Health, Berkeley, California, USA; 3Cancer Prevention Institute of California, Berkeley, California, USA; 4Battelle Memorial Institute, Columbus, Ohio, USA; 5Lucile Packard Children’s Hospital at Stanford University, Palo Alto, California, USA

*These authors contributed equally to this work.

PDF icon PDF Version (167 KB)

  • Background: House dust is a major source of exposure to polybrominated diphenyl ethers (PBDEs), which are found at high levels in U.S. homes.

    Methods: We studied 167 acute lymphoblastic leukemia (ALL) cases 0–7 years of age and 214 birth certificate controls matched on date of birth, sex, and race/ethnicity from the Northern California Childhood Leukemia Study. In 2001–2007, we sampled carpets in the room where the child spent the most time while awake; we used a high-volume small-surface sampler or we took dust from the home vacuum. We measured concentrations of 14 PBDE congeners including penta (28, 47, 99, 100, 153, 154), octa (183, 196, 197, 203), and decaBDEs (206–209). Odds ratios (ORs) were calculated using logistic regression, adjusting for demographics, income, year of dust collection, and sampling method.

    Results: BDE-47, BDE-99, and BDE-209 were found at the highest concentrations (medians, 1,173, 1,579, and 938 ng/g, respectively). Comparing the highest to lowest quartile, we found no association with ALL for summed pentaBDEs (OR = 0.7; 95% CI: 0.4, 1.3), octaBDEs (OR = 1.3; 95% CI: 0.7, 2.3), or decaBDEs (OR = 1.0; 95% CI: 0.6, 1.8). Comparing homes in the highest concentration (nanograms per gram) tertile to those with no detections, we observed significantly increased ALL risk for BDE-196 (OR = 2.1; 95% CI: 1.1, 3.8), BDE-203 (OR = 2.0; 95% CI: 1.1, 3.6), BDE-206 (OR = 2.1; 95% CI: 1.1, 3.9), and BDE-207 (OR = 2.0; 95% CI: 1.03, 3.8).

    Conclusion: We found no association with ALL for common PBDEs, but we observed positive associations for specific octa and nonaBDEs. Additional studies with repeated sampling and biological measures would be informative.

  • Citation: Ward MH, Colt JS, Deziel NC, Whitehead TP, Reynolds P, Gunier RB, Nishioka M, Dahl GV, Rappaport SM, Buffler PA, Metayer C. 2014. Residential levels of polybrominated diphenyl ethers and risk of childhood acute lymphoblastic leukemia in California. Environ Health Perspect 122:1110–1116;

    Address correspondence to M.H. Ward, Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, 9609 Medical Center Dr., Room 6E138, Rockville, MD 20850 USA. Telephone: (240) 276-7172. E-mail:

    Patricia A. Buffler is deceased (October 2013).

    We thank the clinical investigators and their teams at the collaborating hospitals for help in recruiting patients including University of California, Davis, Medical Center (J. Ducore), University of California, San Francisco (M. Loh and K. Matthay), Children’s Hospital of Central California (V. Crouse), Lucile Packard Children’s Hospital, Children’s Hospital Oakland (J. Feusner), Kaiser Permanente Sacramento (V. Kiley), Kaiser Permanente Santa Clara (C. Russo and A. Wong), Kaiser Permanente San Francisco (K. Leung), and Kaiser Permanente Oakland (S. Month). We also acknowledge the Northern California Childhood Leukemia Study staff and the Survey Research Center for their effort and dedication, and S. Merkle at Information Management Services for programming and data management support.

    This research was partially supported by the Intramural Research Program of the National Cancer Institute (NCI), National Institutes of Health and through NCI subcontracts 7590-S-04 (University of California, Berkeley) and 7590-S-01 (Battelle Memorial Institute) under NCI contract N02-CP-11015 (Westat). This research was also financially supported by National Institute of Environmental Health Sciences grants R01ES009137 and P-42-ES-04705-18 (University of California, Berkeley) and NCI grant 5R01CA092683-03 (Colorado State University).

    P.A.B. served on the Board of Directors for FMC Agricultural Products from 1999 to 2011. All other authors declare that they have no actual or potential competing financial interests.

    Received: 5 September 2013
    Accepted: 30 May 2014
    Advance Publication: 3 June 2014
    Final Publication: 1 October 2014

  • PDF icon Supplemental Material (386 KB) PDF


Acute lymphoblastic leukemia (ALL) is the most common form of childhood cancer, accounting for about 80% of childhood leukemia in most Western countries (Ross and Spector 2006). Incidence peaks at 2–5 years of age, and in the United States rates are highest among Hispanics (Ries et al. 2004). Childhood leukemia incidence is generally higher in industrialized countries than in developing ones (Ross and Spector 2006). The international variation and the early age peak in incidence suggest that environmental exposures during pregnancy or early life are important to disease development; however, to date, few risk factors for ALL have been identified (Buffler et al. 2005).

Polybrominated diphenyl ethers (PBDEs) are flame-retardant chemicals used in electronics, furniture, and some textiles (Birnbaum and Staskal 2004). They are structurally similar to polychlorinated biphenyls (PCBs), are persistent in the environment, and bioaccumulate through the food chain [U.S. Environmental Protection Agency (EPA) 2012]. Use of the three commercial formulations—pentaBDE, octaBDE, and decaBDE (named for the average number of bromines in the congeners of each formulation)—increased in the United States from the 1980s until 2004, when U.S. manufacturing of penta and octaBDE was voluntarily phased out due to concerns about toxicity (U.S. EPA 2008, 2012). DecaBDE was voluntarily phased out in the United States at the end of 2013. However, because of their persistence and their continued use in some imported products (U.S. EPA 2013), human exposure to PBDEs continues. Further, decaBDE can break down in the environment into more bioavailable PBDE congeners (Söderstrom et al. 2004; Stapleton and Dodder 2008), which may result in continued exposure to lower brominated PBDEs.

The primary sources of PBDEs are consumer products including plastic housings of computers and other electronics, foam upholstery, carpet pads, and mattresses (Birnbaum and Staskal 2004; U.S. EPA 2012). Because PBDEs are not chemically bound to treated products and often constitute several percent by weight, they can migrate into the indoor environment resulting in high indoor concentrations (Birnbaum and Staskal 2004). Approximately 80% of PBDE exposure in the United States is estimated to come from nonfood sources (Lorber 2008), and ingestion of and dermal contact with PBDEs in house dust is a major route of exposure (Coakley et al. 2013; Lorber 2008; Stapleton et al. 2012; Watkins et al. 2012; Wu et al. 2007). Young children tend to have higher PBDE body burdens than adults (Rose et al. 2010; Sjödin et al. 2008) most likely because of their smaller body size and greater intake of dust through frequent hand-to-mouth activity (Harrad et al. 2010; Lorber 2008; Tulve et al. 2002). Further, levels in California children are among the highest in the world because California has had the strictest fire retardant standards for consumer products in the country (Bradman et al. 2012; Rose et al. 2010).

Environmental chemical exposures that are hematotoxic, damage DNA, or interfere with immune system function are of interest as possible leukemia risk factors (Buffler et al. 2005; Ross and Spector 2006). A cross-sectional study of adolescents in the Netherlands found that higher serum concentrations of PBDEs were associated with decreased numbers of lymphocytes (Leijs et al. 2009). Animal studies have shown that PBDEs cause perturbations of cytokine and chemokine levels (Lundgren et al. 2009), increases in lymphocyte proliferation, and decreases in peripheral monocytes and natural killer (NK) cell activity (Fair et al. 2012). Moreover, exposure to PBDEs has been associated with adverse neurological outcomes in children (Eskenazi et al. 2013; Herbstman et al. 2010; Roze et al. 2009), congenital cryptorchidism (Main et al. 2010), and alteration of thyroid hormones in adults (Turyk et al. 2008).

Previously, we reported that PCBs in house dust were positively associated with childhood ALL in California (Ward et al. 2009). Because PBDEs are structurally similar to PCBs and are present at high concentrations in California homes, we also analyzed household dust samples for 14 PBDEs. To our knowledge, no prior epidemiologic studies have evaluated residential PBDEs and childhood leukemia risk.


Study population. We conducted a population-based case–control study of childhood leukemia in Northern and Central California (the Northern California Childhood Leukemia Study [NCCLS]), which included 17 counties in the San Francisco Bay area and 18 counties in the Central Valley (1995–2008) (Bartley et al. 2010). Briefly, cases ≤ 14 years of age were identified from the nine major pediatric clinical centers in the study area. Controls were selected from California birth certificate files and were individually matched (initially 2:1 and then later 1:1 due to funding limitations) to cases on date of birth, sex, maternal race, Hispanic ethnicity, and maternal residence in the study area. Interviews of cases and controls in the main study (tier 1) included questions about child care attendance, residential history, parental occupation, smoking, and other factors. Cases and controls who were < 8 years old at diagnosis/reference date (December 1999–June 2006) and still living in the homes they occupied at the diagnosis/reference date were eligible for a second interview and dust sampling in the home (tier 2).

Eligibility was limited to younger cases and controls so that the dust sample reflected residential exposures over a substantial portion of the child’s life. Because of the requirement that cases and controls were living in the diagnosis/reference home, individual matching of cases and controls was not maintained; instead, analyses were adjusted for the matching factors.

The tier 2 interviews (2001–2007) included questions about household pesticide use, home age and type, carpet age, and parental occupations. A global positioning system measurement of the home location was taken, and we determined whether the residence was located in an urban, suburban, or rural area based on the 2000 U.S. census.

Carpet dust samples were collected primarily using a specialized vacuum, the high-volume small-surface sampler (HVS3; Envirometrics, Inc.), in the room where the child spent the most time while awake. We sampled one or more carpets/rugs that were at least 1 m2 in size and were present before the reference date (Colt et al. 2008; Ward et al. 2009). Most samples were from the living or family room. Interviewers also collected existing dust from the household vacuum cleaner. HVS3 dust was analyzed preferentially, with household vacuum dust used only when there was insufficient HVS3 dust. After we determined that HVS3 and household vacuum dust concentrations of pesticides, PCBs, and polycyclic aromatic hydrocarbons were highly correlated (Colt et al. 2008), we sampled dust only from household vacuums from July 2006 through 2007 in lieu of the more time-intensive HVS3 method.

During the tier 2 study period, the participation rate for the first interview among families of cases and controls < 8 years old was 86% (Metayer et al. 2013). A total of 324 cases and 407 controls met the eligibility criteria for tier 2; 296 leukemia cases (91%) including 269 ALLs, and 333 controls (82%) participated. Among those with approximately ≥ 0.5 g of dust, we selected all cases and controls with HVS3 dust and a sample of those with household vacuum dust, giving 167 ALL cases and 214 controls for PBDE analysis. HVS3 dust was analyzed for 101 (61%) cases and 133 (62%) controls. Because of the longer time needed to enroll birth certificate controls in the main study, the time between the reference date and dust collection was shorter for cases [median years, interquartile range (IQR): 0.92 (0.74–1.26] than for controls [1.66 (1.26–2.24)]. The study was reviewed by institutional review boards at the University of California, Berkeley, and the National Cancer Institute. Participants (child’s parent or legal guardian) gave written informed consent before the interviews and dust collection.

Laboratory methods. PBDEs were analyzed by the same laboratory that was used previously for PCBs, pesticides, and other chemicals; details of the sample processing have been described (Colt et al. 2008). Dust samples were sieved and the fine fraction (< 150 μm) retained. A 0.1-g portion of each dust sample was analyzed for 14 PBDEs in batches with 15 participants’ samples, one participant’s sample in duplicate, the same sample spiked with low and high concentrations of all analytes, and a solvent method blank. All batches included at least 4 case and 4 control samples; laboratory personnel were blind to case/control status. Each dust sample was spiked with 250 ng of the following Surrogate Recovery Standards (SRS) that are not found in commercial flame retardant mixtures or are at trace levels (LaGuardia et al. 2006): BDE-126 [5 bromine (Br)], BDE-177 (7 Br), BDE-195 (8 Br), and 13C12 BDE-209 (10 Br). Dust was extracted with 1:1 hexane:DCM, and extracts were analyzed using negative chemical ionization gas chromatography (GC)/mass spectrometry (MS) in multiple ion detection mode (see Supplemental Material, “Analytical method for quantification of the PBDEs”).

The method detection limit (MDL) was based on the concentration giving a peak approximately three times the matrix/instrument noise level. In < 1% of samples, PBDEs were quantified below the usual MDL; these were treated as detections in our analyses. The recoveries of the 4 SRSs were used to evaluate the method performance on a sample-by-sample basis. Recoveries from participants’ samples ranged from 66 ± 14% for SRS 13C12 BDE-209 to 78 ± 12% for BDE-126. In Supplemental Material, Table S1, we show mean sample recoveries for dust spikes and mean percent differences for duplicate samples. Mean low level spike recoveries ranged from 85% for BDE-208 to 122% for BDE-154. Mean percent differences in duplicates (n = 28) ranged from 4% for BDE-99 to 33% for BDE-207. The mean percent difference for BDE-207 included three pairs of samples with a detection in one sample and no detection in the duplicate; removing these gave a mean of 24%.

We used BDE-209 standards to estimate the rate of degradation of BDE-209 in the GC injector. Based on these percentages, we corrected the concentrations of BDE-206 through BDE-209 in participants’ samples. On a molar basis BDE-209 conversion was 3% to BDE-206, 2% to BDE-207, and 1% to BDE-208; we did not observe any conversion of BDE-209 to BDE-203. The correlations between concentrations corrected and uncorrected for BDE-209 degradation ranged from 0.63 for BDE-206 to 1.00 for BDE-209 (see Supplemental Material, Table S2). Analyses were based on the corrected concentrations.

Statistical analysis. Because PBDE concentrations had wide ranges and were right-skewed, natural logarithm transformations of concentrations were used in statistical analyses. For HVS3-sampled dust, we also calculated the chemical loading (nanograms per square meter), by multiplying the chemical concentration (nanograms per gram dust) by the dust loading [total fine dust collected divided by the sampled area (grams per square meter)]. Our results were similar when we used SRS-corrected and uncorrected concentrations; therefore, we report the uncorrected concentrations.

For samples with no detections of specific PBDEs, we used a single imputation method (Lubin et al. 2004) that selects values from the modeled (log-normal) distribution to assign values below the MDL. To create imputation models, we evaluated demographics, home characteristics, sampling year, and other factors in relation to PBDE concentrations by comparing medians and IQRs across levels of these variables. Factors associated with the PBDEs found at highest concentrations in homes (BDE-47, BDE-99, BDE-209) were sample year, time between diagnosis/reference date and dust sampling, residence location (urban, suburban, rural), income, and maternal age, but not season of dust sampling or sampling method. Imputation models included sample year and residence location—factors most strongly associated with PBDE concentrations.

In addition to evaluating the continuous PBDE concentration in relation to ALL risk, we categorized the distributions of most PBDEs into quartiles based on the distribution among controls. For BDE-196, BDE-203, BDE-207, and BDE-208 (detection rates < 80%) the lowest category was homes with no detections; concentrations above the MDL were categorized into tertiles. Odds ratios (OR) and 95% CIs were calculated using logistic regression. Tests for trend were based on the natural-log BDE concentration (or loading) in the regression models. The alpha level for statistical significance was 0.05.

We adjusted all analyses for age at diagnosis/reference date (< 1, 1 to < 2, 2 to 5, > 5 to < 8 years), sex, and race/ethnicity (non-Hispanic white, Hispanics, non-Hispanic other races), the original matching factors. We also adjusted for household income, year of dust sampling, and sampling method, factors that changed the OR for one or more PBDE by ≥ 10 percent. Adjustment for breastfeeding duration, child care attendance, mother’s or father’s education or age, age of carpets or residence, and geographic location (urban, suburban, rural) did not meet our criterion for confounding. We conducted stratified analyses to evaluate the consistency of associations by sex, child’s age at diagnosis/reference date, race/ethnicity, income, dust sampling method (HVS3 only), breastfeeding duration, and maternal age.

Although our analyses of BDE-206, BDE-207, BDE-208, and BDE-209 were based on the corrected concentrations, most studies of PBDEs and health have not corrected for BDE-209 degradation. Therefore, we show results for quartiles of the uncorrected concentrations of decaBDE in Supplemental Material, Table S3.


The MDLs, detection frequencies, and distributions of the PBDE concentrations among control homes are shown in Table 1. BDE-47 and BDE-99, the main congeners in the pentaBDE formulation, and BDE-209, the main constituent of the decaBDE formulation, were found at the highest concentrations in homes. The median concentration of BDE-183, the main congener in the octaBDE formulation, was > 50-fold lower than BDE-99 and BDE-209. Among controls and cases, the distributions of total PBDEs and individual PBDEs were similar by age, sex, race/ethnicity, and income level (not shown).

Table 1 - See HTML for full tableTable 1 – Distribution of concentrations (ng/g)a of PBDEs in tier 2 NCCLS dust samples among controls (n = 214).

View Table (HTML Version)
View larger image (TIF File)

Congeners in the penta and octaBDE formulations were highly and significantly correlated within formulations [Spearman correlation coefficients (rS): pentaBDEs 0.79 to 0.997, octaBDEs 0.50 to 0.78] (see Supplemental Material, Table S4). PBDEs with 9 bromine atoms (BDE-206, BDE-207, BDE-208) were highly correlated (rS = 0.79–0.81) but correlations with BDE-209 were lower (rS = 0.13–0.16).

As shown in Table 2, most (62%) cases were male, with equal percentages (38%) being Hispanic and non-Hispanic white. Among controls, there were fewer Hispanics than non-Hispanic whites primarily because fewer Hispanic controls met the eligibility criterion of not moving since the reference date. A greater percentage of controls (50%) was in the highest income category (≥ $75,000) than cases (34%). Slightly more controls had dust sampled in the spring or summer (61%) than cases (53%). Cases and controls did not differ by residence age, location (urban, suburban, rural), sampling year, breastfeeding status, or maternal age at birth (not shown). The duration of residence at the sampled home before diagnosis/reference date [median (IQR)]: was 2.5 years (1.5–3.8) for cases and 2.4 years (1.6–4.1) for controls.

Table 2 - See HTML for full tableTable 2 – Characteristics of ALL cases (n = 167) and controls (n = 214) included in analyses of PBDEs in carpet dust samples, 2001–2007 [n (%)].

View Table (HTML Version)
View larger image (TIF File)

Neither concentrations nor loadings of total PBDEs, nor the sum of PBDEs in each formulation, were associated with ALL risk (Table 3). None of the individual pentaBDE congeners was associated with risk of ALL. However, increasing residential concentrations of octaBDE congeners BDE-196 and BDE-203 were each associated with significantly increased risk of ALL (BDE-196: OR = 1.20, 95% CI: 1.01, 1.42; BDE-203: OR = 1.25, 95% CI: 1.05, 1.48). Results were similar for PBDE loadings (HVS3 samples: 101 cases, 133 controls). ORs were elevated for a 1-ln(ng/g) change in concentration and loading of BDE-206, BDE-207, and BDE-208 but did not reach statistical significance. There was no association with BDE-209.

Table 3 - See HTML for full tableTable 3 – ORs (95% CIs) for acute lymphoblastic leukemia associated with the natural log of PBDE concentrations and loadings.a

View Table (HTML Version)
View larger image (TIF File)

Results for categorical analyses of PBDE concentrations are shown in Table 4. We observed elevated ORs for the second quartiles of the summed pentaBDEs and for BDE-47, BDE-99, BDE-100, BDE-153, and BDE-154 but little or no increase for the third or fourth quartiles. Results were similar for pentaBDE loadings (not shown). Risk was not significantly elevated for the summed octaBDEs and decaBDEs but was significantly increased for specific congeners in these formulations. Compared with no detections, risk increased with increasing tertiles of BDE-196, BDE-203, BDE-207, and BDE-208 and was significantly increased in the highest tertiles for BDE-196 (OR = 2.08; 95% CI: 1.14, 3.77), BDE-203 (OR = 1.98; 95% CI: 1.09, 3.62), BDE-206 (OR = 2.08; 95% CI: 1.08, 3.98), and BDE-207 (OR = 1.97; 95% CI: 1.03, 3.76). We observed similar associations and positive trends for PBDE loadings (not shown).

Table 4 - See HTML for full tableTable 4 – ORs (95% CIs) for the association of quartiles or tertiles of PBDE concentrations in the penta, octa, and deca formulations and risk of ALL.

View Table (HTML Version)
View larger image (TIF File)

Risk estimates for PBDEs were not changed by adjustment for residential concentrations of PCBs (not shown), which were associated with increased ALL risk in this population (Ward et al. 2009). Our results were similar when we restricted our analyses to those who lived at their home for at least 1 year before diagnosis/reference date, to subjects whose homes were sampled within 2 years of diagnosis/reference date, or to HVS3 dust (not shown). We observed consistent results by age, sex, race/ethnicity, maternal age, and breastfeeding status (not shown).

Our primary analyses of decaBDEs were based on concentrations corrected for BDE-209 degradation. Correction resulted in a small increase (5%) in the median BDE-209 concentration. Compared with uncorrected medians, corrected concentrations of BDE-206, BDE-207, and BDE-208 decreased about 49%, 75%, and 70%, respectively. Rank correlations between corrected and uncorrected concentrations were 0.63, 0.66, and 0.69 for BDE-206, BDE-207, and BDE-208, respectively; whereas corrected and uncorrected BDE-209 concentrations had a correlation of 1.00 (see Supplemental Material, Table S2). The associations we observed between the corrected BDE-207 and BDE-208 concentrations and ALL risk were monotonic and demonstrated stronger positive trends than the associations based on the uncorrected concentrations (tertiles above the detection limit vs. nondetections; see Supplemental Material, Table S3).


We found no association with ALL risk for the total summed PBDEs, for the sum of congeners in the penta, octa, and decaBDE formulations, or for the dominant constituents of these formulations (BDEs 99, 183, and 209, respectively). However, children whose homes had the highest concentrations of specific, less common PBDEs with 8 or 9 bromine atoms (BDE-196, BDE-203, BDE-206, BDE-207) had about twice the risk of ALL compared with children in homes with low or nondetected levels of these PBDEs. Results were similar, although not all associations were statistically significant, for PBDE loadings in the subgroup with HVS3 dust. Adjustment for BDE-209 degradation in the GC injector resulted in stronger positive trends for BDE-206, BDE-207, and BDE-208 compared with analyses of unadjusted concentrations of these congeners, suggesting that adjustment may have improved exposure classification.

To our knowledge, no prior epidemiologic studies have evaluated residential PBDE exposure and childhood cancer risk. In animals, only the carcinogenicity of BDE-209 has been evaluated; increases in thyroid gland follicular cell adenomas and carcinomas (combined) were observed for male and female mice, and dose-related increases in liver adenomas were found in male and female rats [National Toxicology Program (NTP) 1986]. These findings were described as equivocal by the NTP (1986); however, the U.S. EPA considered them to be suggestive of carcinogenic potential (U.S. EPA 2008). Chronic long-term carcinogenicity studies of the pentaBDEs are currently being conducted by the NTP (2014).

BDE-209 is less easily absorbed and more rapidly excreted than lower-brominated PBDEs (Birnbaum and Cohen Hubal 2006); therefore, toxicities may be different from other PBDEs (Birnbaum and Cohen Hubal 2006; Dunnick and Nyska 2009). It is likely that PBDE congeners differ in toxicities and carcinogenic potential because differing toxicities are well-documented for the structurally similar PCBs, furans, and dioxins (Lorber 2008). However, to date, there is limited information on the toxicities of the octa and nonaBDEs (Birnbaum and Cohen Hubal 2006; Lorber 2008). Most short-term toxicity studies have focused on pentaBDE formulation congeners, and have demonstrated toxic effects on the liver and thyroid gland (Dunnick and Nyska 2009; Dunnick et al. 2012). Lower-brominated PBDEs also have endocrine-disrupting properties similar to dioxins (Fowles et al. 1994). PBDEs can be contaminated with dioxins and furans (Ren et al. 2011); however, tetrachlorodibenzo-p-dioxin (TCDD) at contaminant levels has not shown toxicity to the liver and thyroid gland to the same extent as the PBDEs (Suzuki et al. 2010). Nevertheless, the dioxin-like activity of the PBDEs is a concern and a possible mechanism for carcinogenicity. Further mechanistic studies of specific PBDE congeners would be informative.

Studies of the immunotoxicity of PBDEs are quite limited. A mouse study showed adverse effects of BDE-99 and Bromkal 70-5 DE, a commercial pentaBDE mixture, on cytokine and chemokine levels (Lundgren et al. 2009). A recent animal study (Fair et al. 2012) of DE-71, a commercial pentaBDE mixture, found increased lymphocyte proliferation, decreased NK cell activity, and decreased numbers of peripheral monocytes. A cross-sectional study of Dutch adolescents found decreased lymphocytes associated with higher serum concentrations of summed congeners in the penta and octaBDE formulations (Leijs et al. 2009). The clinical implications of these effects are unknown. Chemicals that perturb the immune system may increase chances of an inappropriate response to infection, which is a hypothesized cause of childhood leukemia (Greaves 1988).

Blood levels of PBDEs in California children (2003–2005) (Rose et al. 2010), first-time California mothers (similar time frame) (Park et al. 2011), and Mexican-American children in California (2007–2008) (Bradman et al. 2012) are among the highest in the world. Levels of BDE-47 in breast milk (Park et al. 2011) exceeded the U.S. EPA reference dose for neurodevelopmental and liver toxicity for 60% and 7% of mothers, respectively. Given the potential for high exposure to the developing child, further research on the health effects of PBDEs is warranted.

PCBs, which are structurally similar to PBDEs, are immunotoxic (Heilmann et al. 2010) and considered to be probable human carcinogens by the International Agency for Research on Cancer (Lauby-Secretan et al. 2013) and the U.S. EPA (1996). We observed a positive association between residential levels of PCBs and ALL risk in this study population (Ward et al. 2009); however, concentrations of PBDEs and PCBs were not significantly correlated in our study, and adjustment for PCBs levels did not change the associations with PBDE congeners. Similarly, studies of Hispanic women with young children in Salinas County, California (Bradman et al. 2007), and first-time mothers in California (Park et al. 2011), found no significant correlation between concentrations of PCBs and PBDEs in maternal blood or breast milk, suggesting different routes and sources of exposure. Differences in the time periods of peak production and differences in uses may account for the low correlations between PCB and PBDEs in homes (Harrad et al. 2008).

Strengths of our study were rapid case ascertainment, population-based selection of controls, and high participation rates. An additional strength was the measurement of numerous chemicals that allowed us to adjust for co-occurrence with other chemicals in the homes. We were able to evaluate PBDE loadings among cases and controls with HVS3 dust, and we observed results that were similar to PBDE concentrations. Loadings are postulated to be a more accurate indicator of exposure for small children (Bradman et al. 1997; Lanphear et al. 1998), for whom dust ingestion is a major route of exposure. A previous study (Allen et al. 2008) found low correlations between PBDE concentrations in household vacuum bags and interviewer-collected dust. In our study, concentration ranges and associations with ALL were similar for HVS3- and vacuum-collected dust samples combined and for HVS3-collected dust samples only, providing some evidence that concentrations of PBDEs in household vacuum dust may be a useful exposure indicator.

A limitation of our study was that we measured PBDEs in carpet dust only. However, a few studies have measured PBDEs in both dust and serum and found dust to be an important source of exposure for first-time mothers and young children (Rose et al. 2010; Stapleton et al. 2012; Wu et al. 2007). Further, a recent study in New Zealand (Coakley et al. 2013) found significant correlations between PBDE concentrations in dust from mattresses and floors and mothers’ breast milk. Another limitation was our use of one dust sample to characterize exposure, which may result in exposure misclassification if PBDE levels vary over time. In a subset of homes in our study (Whitehead et al. 2013), two dust samples from household vacuums taken 3–8 years apart and analyzed for PBDEs by a different laboratory in 2010–2012 had moderate Spearman rank correlations (range, 0.16–0.56). Correlations were generally higher for samples 3–6 years apart compared with 7–8 years and were higher for the lower-brominated PBDEs compared with the higher-brominated PBDEs (e.g., pentaBDE rS > octaBDE rS > decaBDE rS). A prior study (Allen et al. 2008) demonstrated strong correlations between PBDEs measured in two dust samples taken about 8 months apart. Both studies provide some support for the usefulness of one dust sample to characterize residential exposure over a period of a few years. Further limitations included the small sample size, especially for analyses of HVS3 dust, and relatively small numbers of Hispanics and lower-income participants in our sampled population. However, we found little evidence that PBDE levels varied substantially by race/ethnicity or income levels. Results among Hispanics and non-Hispanic whites and stratified by income level were similar to results overall, providing some assurance that our findings were not biased by these factors.

In summary, we found no association with ALL for most common PBDEs in residential dust, but observed positive associations for specific octa and nonaBDEs. Among those with residential concentrations in the highest tertile of specific octaBDE congeners (BDE-196 and BDE-203) and decaBDE congeners (BDE-206 and BDE-207), we observed a significantly increased risk of ALL. Positive trends for BDE-206, BDE-207, and BDE-208 were stronger after adjusting for BDE-209 degradation in the GC injector. Additional animal carcinogenicity and toxicity studies and epidemiologic studies with repeated sampling and biological measures of exposure would be informative.


Allen JG, McClean MD, Stapleton HM, Webster TF. 2008. Critical factors in assessing exposure to PBDEs via house dust. Environ Int 34:1085–1091.

Bartley K, Metayer C, Selvin S, Ducore J, Buffler P. 2010. Diagnostic X-rays and risk of childhood leukaemia. Int J Epidemiol 39:1628–1637.

Birnbaum LS, Cohen Hubal EA. 2006. Polybrominated diphenyl ethers: a case study for using biomonitoring data to address risk assessment questions. Environ Health Perspect 114:1770–1775; doi: 10.1289/ehp.9061.

Birnbaum LS, Staskal DF. 2004. Brominated flame retardants: cause for concern? Environ Health Perspect 112:9–17; doi: 10.1289/ehp.6559.

Bradman A, Castorina R, Sjödin A, Fenster L, Jones RS, Harley KG, et al. 2012. Factors associated with serum polybrominated diphenyl ether (PBDE) levels among school-age children in the CHAMACOS cohort. Environ Sci Technol 46:7373–7381.

Bradman A, Fenster L, Sjödin A, Jones RS, Patterson DG Jr, Eskenazi B. 2007. Polybrominated diphenyl ether levels in the blood of pregnant women living in an agricultural community in California. Environ Health Perspect 115:71–74; doi: 10.1289/ehp.8899.

Bradman MA, Harnly ME, Draper W, Seidel S, Teran S, Wakeham D, et al. 1997. Pesticide exposures to children from California’s Central Valley: results of a pilot study. J Expo Anal Environ Epidemiol 7:217–234.

Buffler PA, Kwan ML, Reynolds P, Urayama KY. 2005. Environmental and genetic risk factors for childhood leukemia: appraising the evidence. Cancer Invest 23:60–75.

Coakley JD, Harrad SJ, 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.

Colt JS, Gunier RB, Metayer C, Nishioka MG, Bell EM, Reynolds P, et al. 2008. Household vacuum cleaners vs. the high-volume surface sampler for collection of carpet dust samples in epidemiologic studies of children. Environ Health 7:6; doi: 10.1186/1476-069X-7-6.

Dunnick JK, Brix A, Cunny H, Vallant M, Shockley KR. 2012. Characterization of polybrominated diphenyl ether toxicity in Wistar Han rats and use of liver microarray data for predicting disease susceptibilities. Toxicol Pathol 40:93–106.

Dunnick JK, Nyska A. 2009. Characterization of liver toxicity in F344/N rats and B6C3F1 mice after exposure to a flame retardant containing lower molecular weight polybrominated diphenyl ethers. Exp Toxicol Pathol 61:1–12.

Eskenazi B, Chevrier J, Rauch SA, Kogut K, Harley KG, Johnson C, et al. 2013. In utero and childhood polybrominated diphenyl ether (PBDE) exposures and neurodevelopment in the CHAMACOS study. Environ Health Perspect 121:257–262; doi: 10.1289/ehp.1205597.

Fair PA, Stavros HC, Mollenhauer MA, DeWitt JC, Henry N, Kannan K, et al. 2012. Immune function in female B6C3F1 mice is modulated by DE-71, a commercial polybrominated diphenyl ether mixture. J Immunotoxicol 9:96–107.

Fowles JR, Fairbrother A, Baecher-Steppan L, Kerkvliet NI. 1994. Immunologic and endocrine effects of the flame-retardant pentabromodiphenyl ether (DE-71) in C57BL/6J mice. Toxicology 86:49–61.

Greaves MF. 1988. Speculations on the cause of childhood acute lymphoblastic leukemia. Leukemia 2:120–125.

Harrad S, Goosey E, Desborough J, Abdallah MA, Roosens L, Covaci A. 2010. Dust from U.K. primary school classrooms and daycare centers: the significance of dust as a pathway of exposure of young U.K. children to brominated flame retardants and polychlorinated biphenyls. Environ Sci Technol 44:4198–4202.

Harrad S, Ibarra C, Diamond M, Melymuk L, Robson M, Douwes J, et al. 2008. Polybrominated diphenyl ethers in domestic indoor dust from Canada, New Zealand, United Kingdom and United States. Environ Int 34:232–238.

Heilmann C, Budtz-Jørgensen E, Nielsen F, Heinzow B, Weihe P, Grandjean P. 2010. Serum concentrations of antibodies against vaccine toxoids in children exposed perinatally to immunotoxicants. Environ Health Perspect 118:1434–1448; doi: 10.1289/ehp.1001975.

Herbstman JB, Sjödin A, Kurzon M, Lederman SA, Jones RS, Rauh V, et al. 2010. Prenatal exposure to PBDEs and neurodevelopment. Environ Health Perspect 118:712–719; doi: 10.1289/ehp.0901340.

LaGuardia MJ, Hale RC. 2006. Detailed polybrominated diphenyl ether (PBDE) congener composition of the widely used penta, octa, and deca-PBDE technical flame retardant mixtures. Environ Sci Technol 40:6247–6254.

Lanphear BP, Matte TD, Rogers J, Clickner RP, Dietz B, Bornschein RL, et al. 1998. The contribution of lead-contaminated house dust and residential soil to children’s blood lead levels. A pooled analysis of 12 epidemiologic studies. Environ Res 79:51–68.

Lauby-Secretan B, Loomis D, Grosse Y, El Ghissassi F, Bouvard V, Benbrahim-Tallaa L, et al. 2013. Carcinogenicity of polychlorinated biphenyls and polybrominated biphenyls. Lancet Oncol 14:287–288.

Leijs MM, Koppe JG, Olie K, van Aalderen WMC, de Voogt P, ten Tusscher GW. 2009. Effects of dioxins, PCBs, and PBDEs on immunology and hematology in adolescents. Environ Sci Technol 43:7946–7951.

Lorber M. 2008. Exposure of Americans to polybrominated diphenyl ethers. J Expo Sci Environ Epidemiol 18:2–19.

Lubin JH, Colt JS, Camann D, Davis S, Cerhan JR, Severson RK, et al. 2004. Epidemiologic evaluation of measurement data in the presence of detection limits. Environ Health Perspect 112:1691–1696; doi: 10.1289/ehp.7199.

Lundgren M, Darnerud PO, Blomberg J, Friman G, Ilback NG. 2009. Polybrominated diphenyl ether exposure suppresses cytokines important in the defence to coxsackievirus B3 infection in mice. Toxicol Lett 184:107–113.

Main KM, Skakkebaek NE, Virtanen HE, Toppari J. 2010. Genital anomalies in boys and the environment. Best Pract Res Clin Endocrinol Metab 24:279–289.

Metayer C, Colt JS, Buffler PA, Reed HD, Selvin S, Crouse V, et al. 2013. Exposure to herbicides in house dust and risk of childhood acute lymphoblastic leukemia. J Expo Sci Environ Epidemiol 23:363–370.

NTP (National Toxicology Program). 1986. Toxicology and Carcinogenesis Studies of Decabromodiphenyl Oxide (CAS No. 1163-19-5) in F344/N Rats and B6C3F1 Mice (Feed Studies). NTP Technical Report 309. Available:​pts/tr309.pdf [accessed 18 August 2014].

NTP (National Toxicology Program). 2014. 2,2′4,4′-Tetrabromodiphenyl Ether (DE-47) – M980090. Available:​agents/ts-m980090.html [accessed 18 August 2014].

Park JS, 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.

Ren M, Peng P, Cai Y, Chen D, Zhou L, Chen P, et al. 2011. PBDD/F impurities in some commercial deca-BDE. Environ Pollut 159:1375–1380.

Ries LAG, Melbert D, Krapcho M, Mariotto A, Miller BA, Freuer EJ, et al. 2004. SEER Cancer Statistics Review: 1975–2004. Bethesda, MD:National Cancer Institute. Available:​2004/ [accessed 18 August 2014].

Rose M, Bennett DH, Bergman A, Fangstrom B, Pessah IN, Hertz-Picciotto I. 2010. PBDEs in 2–5 year-old children from California and associations with diet and indoor environment. Environ Sci Technol 44:2648–2653.

Ross JA, Spector LG. 2006. Cancers in children. In: Cancer Epidemiology and Prevention (Schottenfeld D, Fraumeni JF Jr, eds). New York:Oxford University Press, 1251–1268.

Roze E, Meijer L, Bakker A, Van Braeckel KN, Sauer PJ, Bos AF. 2009. Prenatal exposure to organohalogens, including brominated flame retardants, influences motor, cognitive, and behavioral performance at school age. Environ Health Perspect 117:1953–1958; doi: 10.1289/ehp.0901015.

Sjödin A, Wong LY, Jones RS, Park A, Zhang Y, Hodge C, et al. 2008. Serum concentrations of polybrominated diphenyl ethers (PBDEs) and polybrominated biphenyl (PBB) in the United States population: 2003–2004. Environ Sci Technol 42:1377–1384.

Söderstrom G, Sellström U, de Wit CA, Tysklind M. 2004. Photolytic debromination of decabromodiphenyl ether (BDE 209). Environ Sci Technol 38:127–132.

Stapleton HM, Dodder NG. 2008. Photodegradation of decabromodiphenyl ether in house dust by natural sunlight. Environ Toxicol Chem 27:306–312.

Stapleton HM, Eagle S, Sjödin A, Webster TF. 2012. Serum PBDEs in a North Carolina toddler cohort: associations with handwipes, house dust, and socioeconomic variables. Environ Health Perspect 120:1049–1054; doi: 10.1289/ehp.1104802.

Suzuki G, Someya M, Takahashi S, Tanabe S, Sakai S, Takigami H. 2010. Dioxin-like activity in Japanese indoor dusts evaluated by means of in vitro bioassay and instrumental analysis: brominated dibenzofurans are an important contributor. Environ Sci Technol 44:8330–8336.

Tulve NS, Suggs JC, McCurdy T, Cohen Hubal EA, Moya J. 2002. Frequency of mouthing behavior in young children. J Expo Anal Environ Epidemiol 12:259–264.

Turyk ME, Persky VW, Imm P, Knobeloch L, Chatterton R, Anderson HA. 2008. Hormone disruption by PBDEs in adult male sport fish consumers. Environ Health Perspect 116:1635–1641; doi: 10.1289/ehp.11707.

U.S. EPA (U.S. Environmental Protection Agency). 1996. Polychlorinated Biphenyls (PCBs) (CASRN 1336-36-3). Available: [accessed 7 May 2014].

U.S. EPA (U.S. Environmental Protection Agency). 2008. Toxicological Review of Decabromodiphenyl Ether (BDE-209). EPA 635-R-07-008F. Available:​pdf [accessed 13 December 2013].

U.S. EPA (U.S. Environmental Protection Agency). 2012. Technical Fact Sheet—Polybrominated Diphenyl Ethers (PBDEs) and Polybrominated Biphenyls (PBBs). Available:​fact_sheet_pbde_pbb.pdf [accessed 4 September 2013].

U.S. EPA (U.S. Environmental Protection Agency). 2013. Polybrominated Diphenyl Ethers (PBDEs) Action Plan Summary. Available:​s/pubs/actionplans/pbde.html#action [accessed 13 December 2013].

Ward MH, Colt JS, Metayer C, Gunier RB, Lubin J, Crouse V, et al. 2009. Residential exposure to polychlorinated biphenyls and organochlorine pesticides and risk of childhood leukemia. Environ Health Perspect 117:1007–1013; doi: 10.1289/ehp.0900583.

Watkins DJ, McClean MD, Fraser AJ, Weinberg J, Stapleton HM, Sjödin A, et al. 2012. Impact of dust from multiple microenvironments and diet on PentaBDE body burden. Environ Sci Technol 46:1192–1200.

Whitehead TP, Brown FR, Metayer C, Park JS, Does M, Petreas MX, et al. 2013. Polybrominated diphenyl ethers in residential dust: sources of variability. Environ Int 57–58:11–24; doi: 10.1016/j.envint.2013.03.003.

Wu N, Herrmann T, Paepke O, Tickner J, Hale R, Harvey LE, et al. 2007. Human exposure to PBDEs: associations of PBDE body burdens with food consumption and house dust concentrations. Environ Sci Technol 41:1584–1589.

WP-Backgrounds Lite by InoPlugs Web Design and Juwelier Schönmann 1010 Wien