Sign up to receive email updates, action alerts & health tips from EWG. [Privacy]

Detailed Findings

Fire Retardants in Toddlers and Their Mothers: Detailed Findings

September 4, 2008
Nicolas had 2.2 times more PBDEs than his mom Tess.

EWG examined concentrations of PBDEs in 20 American families. Paired tests taken on the same day from moms and kids found higher concentrations of toxic fire retardants in children from nearly every family. In 19 of 20 families, children had greater total PBDE measurement than their mothers, typically 3.2 times higher. The children studied ranged in age from 1.5 to 4 years old, a time when their playing and eating habits lead to increased ingestion of PBDEs and other contaminants that accumulate in the home. Total PBDE concentrations in children averaged 62 parts per billion in lipid, and ranged from 25 to 114 ppb (or 12.9 to 61.9 pmoles/gram lipid). Concentrations for mothers averaged 25 ppb, and ranged from 10 to 74 ppb (or 5.9 to 40.6 pmol/gram lipid). The study suggests that U.S. children 1 to 4 years of age bear some of the heaviest burdens of flame retardant pollution in the industrialized world. Our study is the first systematic monitoring of Deca PBDE in children, which was identified as a critical data gap by EPA in 2005 (EPA 2005). We found 13 of 20 children and 9 of 20 mothers had measurable amounts of Deca PBDE—a widely used chemical that industry has argued poses no risk to health. Deca is found in plastic for TVs, computers and other electronics. However, the chemical is banned in Europe and some U.S. states, and most manufacturers are voluntarily replacing the chemical due to concerns about its toxicity, accumulation in people and breakdown into other toxic PBDEs.

Olivia had 2.3 times more PBDEs than her mom Mary.

The finding of elevated PBDEs in young children is especially concerning because the most sensitive period for their toxic effects appears to be late pregnancy and early childhood. Single day exposures during a period of rapid brain development are linked to permanent changes in learning, memory and behavior. PBDEs are also known to disrupt thyroid and reproductive hormones, which could lead to permanent impacts if exposures occur during early life.

Children at risk

Our findings are strong evidence that children are not adequately protected from harmful contaminants. Efforts to document concentrations of pollution in people almost exclusively focus on healthy adults whose behaviors lead to lower exposures to many contaminants found in the indoor environment.

Hannelore had 2 times more PBDEs than her mother Erika

However the handful of studies like ours, which focus on chemical exposures for young children, indicate that we can't adequately protect children without understanding how they differ from adults. A recent study examined concentrations of 52 industrial chemicals in 3 to 6 year old children, and found 20 chemicals were markedly elevated relative to older children (ages 12 to 18) and adults (Sexton 2006). These contaminants included volatile chemicals, PCBs, lead, mercury and persistent pesticides. Notably children’s exposures to many of these chemicals continue decades after the chemicals were banned from commerce. Childhood is a period of unique vulnerability to PBDEs and other toxic chemicals. The developing brain and reproductive systems are incredibly vulnerable to toxic chemicals. A classic example is childhood lead poisoning, which illustrates that children’s bodies may be permanently impacted by concentrations of a chemical that would not harm a healthy adult. In the case of PBDEs, a single exposure during a day of rapid brain growth and development can cause permanent changes to behavior and activity (Viberg 2003a, Viberg 2003b, Eriksson 2001, Viberg 2007). In contrast to the single day studies of PBDEs administered to mice in a controlled laboratory setting, American children are exposed to PBDEs thoughout pregnancy and early life, during many periods of significant brain development. They are also exposed to PCBs, lead, Teflon chemicals, and other industrial pollutants that affect brain and reproductive development.

Children in our study have consistently higher levels of PBDEs than adults

Source: EWG's tests of 20 children and mothers

In the case of lead poisoning, economists have quantified the societal impact of slight deficits in intelligence or behavior, which amount to billions of dollars of lost wages, special education, and other social services (Schwartz 1994). The decades of interventions to safely remove lead from children’s homes is being undermined by their exposure to new chemicals intentionally added to products for fire-resistance, or to make non-stick pans and non-stain fabrics, despite evidence they could pose similar risks to children’s wellbeing.

About PBDEs

PBDEs are a class of toxic fire retardants added to household furniture and electronic items. As persistent and bioaccumulative chemicals, their use has lead to accumulation in wildlife, food products, and the general environment. Two of 3 types of PBDEs were withdrawn from commerce in the U.S. in 2005, due to concerns of harmful effects to people and the environment. The third form, known as Deca, is still in widespread use. Studies indicate that daily exposures to PBDE can exceed the government’s safe daily exposure levels for some children. Duke University researchers calculated the amount of PBDEs sticking to adult and children’s hands and determined that young children were most at risk, due to normal play and exploration which leads to putting their hands and other non-food items in their mouths (Stapleton 2008). They estimate children's exposure to PBDEs in dust to be 10 times higher than that of adults.

Maija's daughter Eva had 30% more PBDEs than she did

This is especially concerning because childhood is the period of greatest sensitivity to PBDEs' toxic effects. The most sensitive laboratory tests find that exposure to PBDEs during early life permanently affects learning, memory and behavior in juvenile animals (EPA 2008 a,b,c,d). EPA's recent safety value for Deca PBDE exposure is 500 times lower that the previous estimate by the National Academy of Sciences and promoted by the American Chemistry Council in 2003 (NAS 2000, ACC 2003). The reason is because EPA considered toxicity of Deca exposures during early childhood, while the NAS used a 1986 study that only considered toxicity in adult animals (EPA 2008d). Studies by EWG and others of umbilical cord blood, breast milk, childrens' blood and homes find that American children are exposed to these chemicals throughout early life, including during pregnancy, infancy and childhood (EWG 2003, 2004, 2005, Herbstman 2007, Wu 2007, Athanasiadou 2008).

Zade had 3.4 times more PBDEs than mom Tracy

Tracy says: "My son's concentrations of PBDEs were 3.4 times higher than my own. While this study shows that is common for kids his age, it is certainly worrisome."

In the recent analysis of the health effects of the four most common PBDEs in laboratory studies, EPA determined that the most sensitive impacts are to mice brain and behavior after receiving a single dose of PBDEs (EPA 2008 a,b,c,d). The studies test a period of rapid brain development that is similar to brain growth during the 3rd trimester of a human pregnancy. But these tests do not account for the additional impact of exposures during other life stages, the cumulative impact of exposure to other PBDEs, and other toxic chemicals like mercury, lead and PCBs which also affect intelligence and behavior. As a result, the EPA admitted "low confidence" that the current “daily safe exposure level” for oral exposure assigned by the Agency fully protected health (EPA 2008 a,b,c,d). Scientists measuring PBDEs that accumulate on people's hands calculate that children ingest roughly 10 times more PBDEs than adults from hand-to-mouth contact (Stapleton 2008). Based on these estimates children (weighing 22 to 44 pounds) get about half their daily safe exposure to Penta type PBDEs from the chemical accumulating on their hands alone. The children with higher than average exposures would exceed the safety level based on hand-to-mouth exposures alone. In addition to dust ingestion children face PBDE exposure from foods, inhalation and directly mouthing items with PBDEs in them. Despite this worrisome evidence about PBDEs and the availability of safer replacements, U.S. health and environmental agencies have done little to address children’s on-going exposures to these chemicals. Deca continues to be widely used despite clear evidence that suitable replacements exist and a voluntary shift by major retailers away from the compound. Penta, the form used in foam furniture, is no longer produced in the U.S. However, gaps in EPA’s regulatory safety net allow Penta to legally enter the U.S. via imported foam products. Biomonitoring studies by CDC and others continue to focus on healthy adults and older children, overlooking the population most at risk.

Are current exposures safe?

Susan's son Louis had 3.3 times more PBDEs than she did. Daughter Ella was not tested.

PBDEs have been proven to be especially toxic to the developing brain, leading to permanent changes in learning and behavior (EPA 2008 a,b,c,d). Laboratory studies administered single doses of each type of PBDEs to newborn mice and found permanent effects which worsen with age (Eriksson 2001, Viberg 2002, 2003a, 2003b, 2007). Similar findings have been confirmed by other scientists (Branchi 2003, Kuriyama 2005, Talsness 2005, Rice 2007). It is difficult to compare the doses administered in these 1-day studies to human exposures. PBDEs are widely detected in people, including nearly all of the 2,000 participants in a recent CDC study (Sjödin 2008). Body burden measurements indicate that Americans are exposed consistently from development in the womb through adulthood to a mix of PBDEs and other chemicals with similar impacts to neurodevelopment. A senior scientist at EPA estimates that PBDE concentrations in the most highly exposed Americans are within a factor of 10 from the doses found toxic in the laboratory (Birnbaum 2006). Many Americans are likely to exceed EPA's safety guideline, which calls for human exposures to be 300 to 3000 times lower than those found to be toxic in laboratory studies. Despite widespread exposure during pregnancy and early childhood, only a few studies have examined toxic effects of PBDE exposures in people. A case-control study in Denmark and Finland found statistically significant differences between PBDE concentrations in the breast milk of mothers of boys with cryptorchidism (in which one or both testicles is undescended at birth) as opposed to mothers with unaffected sons (Main 2007). The results from this study suggest that the mothers with higher body burdens of PBDEs were more likely to have sons born with cryptorchidism. However, placental levels of PBDEs did not correlate in the same manner as did the breast milk levels and it is not clear why. These findings are also difficult to interpret given that the concentrations in both groups studied are typically less than 5 parts per billion, dramatically lower than levels in Americans.

Hillary's daughter Brynn had 3.4 times more PBDEs than she did.

A recent study from the U.S. looked at the effects of PBDE exposures on the hormone systems of a group of men who consume sport fish (Turyk 2008). The researchers found that in this group of men, exposure to PBDEs at levels similar to those found in the U.S. population were associated with significant effects on thyroid hormone levels. In addition, those men with the highest 5 percent of PBDE levels also had increased thyroid antibodies, which is a risk factor for thyroid disease later in life. Although this study did not include pregnant women, infants, or children, these results are especially worrisome because even minor fluctuations in thyroid hormone levels can have permanent negative effects on brain development and growth in the developing fetus, infants, and children.

PBDEs in young children

Ours is the first U.S. study to look carefully at PBDE levels in young children. The only previous U.S. study of children was of a single family with 2 young children living in Oakland, California (Fischer 2007). The study reported consistently high measurements of PBDEs and other chemicals in the toddler and 5 year-old child relative to their parents. The significant differences between children and their parents call into question the previous idea that concentrations in mothers is a good predictor of exposures to their children. While maternal monitoring is a proxy for exposure to the developing fetus or nursing infant, the concentrations we measured in young children were higher than levels reported in a large umbilical cord blood study for U.S. infants (Herbstman 2007). They are also higher than concentrations measured in teens and adults in CDC's recent study of 2,000 Americans. Interestingly the moms in our study had lower concentrations than the CDC group, which could be due to the limited number in our study. However, these findings point to the need to evaluate behavioral factors that lead to increased exposures during early life.

EWG findings indicate that young children have higher levels of PBDEs than other age groups

Graphic shows median values reported for 6 common PBDEs primarily used in the Penta mixture. Data are drawn from Herbstman's (2007) tests of 297 umbilical cord blood samples, EWG's test of 20 children ages 1 to 4 years old, and CDC's tests of 2,062 Americans age 12 and older (Sjödin 2008). Deca PBDE was not reported in newborn and adults, but estimated to be around 2 ppb in both groups.

Studies including young children in Norway and Australia both detected elevated concentrations in 0 to 4 year olds, relative to older children and adults (Thomsen 2002, Toms 2008). Notably, many studies of young children outside the U.S. find significantly lower PBDE levels. This is due to the fact that almost all of the Penta PBDE manufactured was used in the U.S. and thus resulted in the addition of thousands of pounds of toxic chemicals to home products. The average PBDE level for 0 to 4 year old Norwegian children was less than 10 ppb in lipid (Thomsen 2007). Each of our child participants exceeded this amount. Four-year-old Spanish children had an average of 1.3 to 3.6 ppb PBDEs in their serum, with higher concentrations noted in breastfed kids (Carrizo 2007). Another study examined paired samples from moms and kids in the Faroe Islands. In this study, children's concentrations at age 7 were roughly equal to maternal concentrations during pregnancy of 4 to 5 ppb in lipid (Fängstrom 2005). The Faroe Islander exposures to Penta-like PBDEs are likely limited to dietary sources. Deca concentrations are much more similar between Americans and Faroe Islanders. A study of 4 pooled blood samples collected from 0 to 4 year old Australian children found slightly higher concentrations than those measured in children in our study (Toms 2008a). Concentrations of PBDEs ranged from 63 to 80 ppb in lipid. The average age of children was 2 years. Each sample contained blood from 100 children. A follow up study was designed to take a closer look at the differences over the first 4 years (Toms 2008b). This study pooled samples from children 0 to 6 months in age, 6 to 12 months in age and 1, 2, and 3 year olds. This study found lower levels in the 0 to 6 month-old children, and similar concentrations for kids 6 months to 3 years. The authors concluded that increasing concentrations in children post-weaning suggest a role for interior sources of PBDEs. Notably the concentrations for Australian children were much lower in the second study, only 35 to 43 ppb (Toms 2008b).

Why do American children have such intense exposures?

Laurie's son Conner had 60% more PBDEs than his mom.

Childhood exposures to household items containing PBDEs have been implicated as the predominant route of daily exposure for U.S. children (Jones-Otazo 2005, Lorber 2008, Allen 2007, Stapleton 2005). This is because the US fire safety standards have been passed without in-depth consideration or any public discussion of likely chemical hazards, and thus resulted in thousands of pounds of toxic chemicals added to home products. Other countries don't require chemical fire retardants, and the U.S. accounted for roughly 95 percent of the world's annual Penta consumption before the mixture was phased-out. Moreover, as a result of widespread environmental contamination, PBDEs are now also detected in common foods, including meat, fish and dairy products (Schecter 2006), which are secondary sources of exposure. Researchers at Duke and Boston Universities have taken a close look at the role of household sources of PBDEs in the measurements in Americans. One recent study found a positive association between the concentrations of PBDEs in household dust and concentrations of PBDE in the bodies of nursing mothers (Wu 2007). Another took a detailed look at brominated chemicals in the home, and found a strong link between Deca measurements and televisions (Allen 2007) and found that PBDE levels in house dust did not vary significantly over an 8 month period (Allen 2008). Finally, researchers measured PBDE concentrations on the hands of adults and older children and used these numbers to estimate the daily intake for young children who have high rates of hand-to-mouth contact (Stapleton 2008). These findings indicate that PBDE particles stuck to hands contribute to about 50 percent of EPA's safe daily exposure level for 2 types of PBDEs for a typical small child, and highly-exposed children will exceed the safety level based on hand-to-mouth contact alone. Nursing can be a significant source of PBDEs for infants, but there is no indication from this or other studies that the harmful effects of PBDEs or other persistent contaminants outweigh the benefits of breastfeeding (Jorissen 2007, Dorea 2006). PBDE transfer from mother to child during pregnancy and nursing could skew the maternal-child comparison. However, studies of American mothers find that PBDE concentrations in mothers’ milk dropped by just 1 to 3 percent per month over the first 6 months of nursing when milk production is greatest (Hooper 2007). This is different than the observation that PCB levels slowly decline over the duration of nuring (LaKind 2001, Bloom 2007). The relatively steady concentrations of PBDEs is likely due to frequent and continuous contact with PBDE-containing consumer products, and possibly shorter half-lives than other persistent chemicals (Hooper 2007).

Benefits of breastfeeding outweigh contaminant concerns

EWG strongly recommends that all mothers breastfeed their children if possible. The health benefits of breastfeeding are numerous and not overshadowed by concerns about contaminants in mothers' milk. For starters, breast milk contains essential fatty acids that are not present in infant formula. These components of milk may bolster brain and neurodevelopment to offset some of the impacts of chemicals like PBDEs (Jacobson 2002). Several studies find that in-utero, not breast milk, exposures to brominated fire retardants are more correlated with adverse health effects in children. These include findings of worse neurological outcomes and reduced growth based on in-utero not lactational PCB exposure (Jacobson, 2002, Koopman-Esseboom 1996), and lower body weight for height (Blanck 2002). Conversely studies of Michigan girls exposure to polybrominated biphenyls (PBBs) found that those with high pre- and post-natal PBB exposure underwent earlier menarche than girls with lesser exposures in-utero or who were not breastfed (Blanck 2000). Notably both PCBs and PBBs are structurally similar to PBDEs, and exhibit many of the same toxic effects. PBDEs have not been extensively tested in people. An additional group of studies find that breastfed children show lesser effects than formula-fed members of the cohort, despite higher body concentrations of toxic chemicals. These studies include those examining DDT, PCBs and dioxins as they affect infant neurodevelopment (Ribas-Fito 2007, Eskenazi 2006, Jorissen 2007, Jacobson 2002, Boersma 2000), mercury's effects on neurodevelopment (Jensen 2005) and DDT and asthma (Sunyer 2005, Sunyer 2006).

More information about contaminants and breastfeeding


Liz's son Evan has 3.3 times more PBDEs in him than she does.

One Spanish study found higher concentrations of PBDEs in 4 year olds who had been breastfed (typically for about 4.5 months) compared to those who had not (Carrizo 2007). However, the absolute difference was small: an average of 3.6 ppb in the lipid of breastfed 4 year-olds compared to 1.3 ppb in formula-fed children. It is unclear how much of a role breastfeeding plays in our measurements. But U.S. children have both greater exposures to PBDEs in mothers' milk and in household products than children in other countries. The study of Australian children looked separately at 0 to 6 months and by 6 month to year increments until 4 years old. This study concluded that PBDE concentrations rise after weaning, so exposures in 2- and 3-year-olds were likely due to contact with household items (Toms 2008b). All children in our study were breastfed for at least 4 months, and many of them for much longer. We found no relationship between the duration of nursing and children's PBDE levels. Additionally 8 mothers and children from the group were participants in a 2003 study of PBDEs in breast milk. There was no clear relationship between a mothers' breast milk samples, her blood samples 3 to 4 years later, and her child's PBDE measurements at that age. In 5 cases children's blood concentrations were greater than PBDEs in mothers' milk. In the remaining 3 cases children's concentrations were lower. These findings underscore the importance of reducing PBDE exposures for women of childbearing age and children.

Types of PBDEs

Jennifer's son Elijah had 3.6 times more PBDEs than his mom.

There are 209 types, or congeners, of PBDEs, and their name indicates the number and arrangement of bromine atoms in a molecule. Penta is the common name for a mixture of PBDEs used primarily in foam furniture. Penta mixtures vary, but typically include primarily PBDEs with 4 or 5 bromines, and a smaller amount of PBDEs with 6-bromines. The Octa mixture includes a mix of PBDEs with 6 to 10 bromines. The product known as Deca is primarily made up of PBDE-209 with 10 bromines. Deca formulations can include PBDEs with eight or nine bromines, higher purity formulations were recently introduced as a result of the European restrictions against PBDEs with 1 to 9 bromines. Overall, a small number of PBDE congeners predominate in commercial fire retardant mixtures, humans and in the environment. These include PBDE-47 (4 bromines), PBDE-99 (5 bromines), PBDE-153 (6 bromines), and PBDE-209 (10 bromines). Our tests revealed high concentrations of a trace PBDE in at least 4 participants. This is due to the fact that another brominated fire retardant, a polybrominated biphenyl (PBB-153) coeluted with PBDE-154 in our analysis.



These sum values of PBDE do not include PBDE-154 which coeluted with a different brominated fire retardant PBB-153 in our analysis. This chemical was excluded in the total because at least 2 moms and their 2 children had abnormal amounts of this trace congener, suggesting relatively high levels of PBB-153. Hence our reported levels are slightly underestimating the total PBDE levels in the children and their mothers.