Sign up to receive email updates, action alerts, health tips, promotions to support our work and more from EWG. You can opt-out at any time. [Privacy]


Study findings

September 23, 2003

Mother's Milk: Study findings

Part 2. Mothers' Milk: Findings

Between November 2002 and June 2003, EWG recruited 20 healthy, pregnant women from across the country, all of whom were expecting their first child, to participate in a study of fire retardants in breast milk. Participants collected a breast milk sample within several months of the birth of their child and completed an exposure assessment questionnaire that provided information about their lifestyle and home and work environments. Samples were analyzed by a certified laboratory.

The lab found PBDEs in every breast milk sample tested — 35 different PBDEs in all. More significantly, our tests found levels higher than those reported previously for U.S. women, including two of the highest levels of PBDEs ever reported in human beings in the world. Levels ranged from 9.5 to 1,078 parts per billion (ppb) in milk fat (lipid), with an average level of 159 ppb, and a median value of 58 ppb. Six of 20 participants had PBDE levels above 100 ppb, with two participants exceeding 700 ppb. The highest PBDE level previously reported in the United States was 580 ppb in maternal blood lipid for a woman in Central Indiana. [45]

EWG study results graph

The most commonly found chemical is known as PBDE-47, which accounted for about half of the total PBDE levels in each participant. It and six other PBDEs accounted for more than 90 percent of each participant's total. Each of these PBDE congeners has three to six bromines and is found in the commercial Penta mixture used predominantly in foam products. There is also evidence suggesting that Deca PBDEs used predominantly in plastics and electronics— which do not readily accumulate in people — are breaking down in the environment to form the congeners that are found in humans.

As observed in other studies of fire retardants, contaminant levels varied widely among participants. [24, 45, 46, 47, 50] All of the participants had higher levels than those commonly detected in European women, and were within the range of body burdens associated with permanent neurological impairments to laboratory animals. Unlike most previous studies of PBDE exposure, we had information about our participants' personal habits, home and work environments. None of the participants reported unusual exposure to PBDE-laden foam or plastics. The variability in PBDE levels we observed in study participants was not easily explained by their diet, occupation, age, body mass or the amount of time they had breastfed their infants.

Of the two most highly-exposed participants, one works in computer-related research and one is a Registered Nurse. One of these participants had replaced the cushions in her sofa and loveseat with cushions purchased at her local variety store. She trimmed the foam cushions for the sofa while inside her home. However, another participant with much lower levels of PBDEs had also recently reupholstered and replaced the foam for a set of dining room chairs, without showing similarly high PBDE levels.

Fish consumption is thought to be a major exposure route for people in Sweden and Finland. [34, 37] One study found that frequent fish-eaters had 5 times higher PBDE levels than non-fish eaters (2.2 vs. 0.4 ppb in lipid.) [34] However, the high PBDE levels in the U.S. might be due to other factors. Seventeen of 20 study participants reported that they typically eat fish at least once per week, and 13 had at least one serving in the seven days before they collected the breast milk sample.

There have been few attempts to characterize fire retardant exposures of the U.S. population. The handful of studies conducted show that PBDE concentrations found in U.S. women and children are the highest reported in the world — and increasing rapidly. Prior to mid-2003 the only information about PBDE levels in U.S. breast milk came from two pooled breast milk samples, collected from an unknown number of women in New York State, Austin, Texas and Denver, Colorado in 1997 and 2000. [48, 49] The samples indicated alarming exposure to PBDEs for U.S. women relative to Europeans and Canadians, but two years passed before a more detailed investigation into PBDE exposures for the U.S. population.

This year, Dr. Arnold Schecter of the University of Texas reported PBDE levels ranging from 6 to 419 ppb lipid for 47 nursing women in Austin and Dallas, the first individual findings for breast milk in the U.S. [47] The study found an average level of 74 ppb and a median (midpoint) value of 34 ppb, 10 to 100 times higher than levels in European women. Total PBDE levels did not vary depending on the mother's age, number of previous births or length of time nursing her infant.

EWG's study found PBDE levels at least 56 percent higher than the Texas and Indiana studies. The average and median (midpoint) values for the most prominent PBDE (PBDE-47) are almost double the levels reported in California women in the late 1990s. [50] This might be in part because we included only first time mothers, who are known to have higher concentrations of many persistent contaminants in their breast milk relative to women who have previously breastfed.

The mean level of PBDEs in our study of women's breast milk was 159 ppb, more than 50 times higher than that reported in a recent Swedish study. [51] In fact, the lowest level of PBDEs in the women we tested (9.5 ppb) was higher than the highest levels measured in Sweden (7.7 ppb), [51] where PBDEs have been voluntarily phased out and will be banned completely next year. [52]

Median values around world graph

Women in other European countries and in Japan appear to have PBDE levels akin to those in Swedish women. Mothers' milk collected in Finland in the mid-1990s had no more than 6 ppb [53]; in that same period German women had a maximum of 11 ppb [54]; and Japanese women studied in 2000 had no more than 1.5 ppb PBDEs. [35]

Asian and European women's exposures to PBDEs are likely much lower than Americans' because the U.S. and Canada use about 95 percent of the form of PBDEs used in foam furniture, which are known to accumulate most widely in people and in the environment. PBDE levels as high as 590 ppb have been measured in the breast milk of Canadian women, though average levels in a recent study were about 40 ppb — still a fraction of those we detected in U.S. women. [29]

PBDE levels rising rapidly in people

Swedish researchers first reported an exponential increase in PBDE levels in 1999. After detecting a new contaminant in a breast milk sample, researchers examined archived breast milk samples collected over a 25-year span and found an amazing 60-fold increase in the concentrations of PBDEs in breast milk between 1972 and 1997 — equivalent to a doubling every five years. They noted that the increase was startling, given that levels of many persistent chemicals (including PCBs) declined sharply in the same period. [30, 55] The Swedish findings garnered worldwide attention, revealing a global pollution problem, spurring additional studies and a phase-out and eventual ban of most PBDEs in the European Union, beginning in 2004. [52]

swedish study results graph

Inspired by the work of Swedish researchers, Canadian experts later discovered that PBDE levels in the breast milk of women in Vancouver, B.C. were 15 times higher in 2002 than they had been just 10 years earlier. Not only were PBDE levels rising more quickly in Vancouver than in Sweden, doubling in 2.6 years as opposed to five years, but also average PBDE levels were at least 10 times higher in Canadian women than their Swedish counterparts. [29]

Recent evidence of a time-related increase in PBDE levels in the United States comes from the San Francisco Bay Area. It indicates that PBDE levels in Bay Area women's blood and breast tissue have more than tripled since PBDEs were first introduced in commercial products about 30 years ago. [50] Also, a recent study of archived blood bank samples from the Southeast and Washington state shows a steady rise in PBDE levels in adult blood serum since 1985, and a steadily decline in PCB and PBB levels over the same period. [28]

Exposures to fetuses and babies

Measuring PBDE levels in breast milk is a useful and accurate way to investigate exposures for the developing fetus and newborn baby. PBDE levels are highly similar when measured in a woman's blood serum or in her breast milk or in her infant's umbilical cord blood. [45, 51, 53] Our findings of high levels of PBDEs in breast milk, and recent studies of PBDE levels in infant cord blood, demonstrate that developing babies' exposure to PBDEs is ubiquitous in the United States.

For many years scientists considered the developing fetus to be shielded by the placenta from many chemicals. However there is a growing awareness that virtually any substance present in the mother's body is transported to some extent into the womb. [56] New evidence suggests that most chemicals that accumulate in a mother's body fat can cross through the placenta and be incorporated into the developing infant's body at high levels. Studies of PBDE's cousin, PCBs, indicate that the developing fetus is particularly sensitive to toxic insult. [4] In-utero exposures are dangerous because they occur during a period of dramatic mobilization of maternal fat stores to nurture the rapidly developing fetus especially during the third trimester of pregnancy a time where there is rapid development of the brain and nervous system.

This year, researchers in Central Indiana studied 12 infant-mother pairs to see whether PBDEs were reaching the bodies of developing babies. They found ample evidence that exposures were occurring before birth. Almost identical contaminant levels were measured in maternal and fetal blood. [20] Another study found a similarly strong concordance between levels of PBDE in maternal and cord blood and breast milk samples. The researchers also note that the smaller, more accumulative, and most toxic PBDEs, those with fewer bromines, moved from maternal blood to cord blood most easily. [51]

The high levels of PBDEs in the maternal body and ease of transfer to the developing fetus result in highly exposed newborns. The Indiana-born babies begin their first day of life with PBDE levels ranging from 14 to 460 ppb PBDEs, some 30 to 100 times higher than Swedish newborns [51], and 35 times higher than blood samples collected from adults in Indiana just 15 years ago. [57] Data from these studies show that US babies are born with PBDE levels much higher than European babies.

Exposures through breastfeeding

One of the most meticulous studies to explore adverse health effects resulting from breastfeeding indicates that infant exposure to PCBs via breast milk is much less of a threat than in-utero exposures. This study tracked children over eleven years born to Michigan mothers who ate PCB-contaminated fish during pregnancy. The researchers reported that babies who were breastfed at least six weeks did better than non-breastfed babies on tests designed to measure subtle neurological differences in young children. [4]

The researchers concluded, "Virtually no adverse effects were found in relation to postnatal exposure to PCBs from breast-feeding, indicating that the fetus is particularly vulnerable to this exposure." [4] This may be because the newborn baby has better mechanisms to detoxify chemicals; because the newborn is not as vulnerable to the impacts of PCBs; or because beneficial compounds in breast milk like omega-3 fatty acids protect the developing brain or repair subtle deficits incurred during the prenatal period. [4, 58]

Despite the numerous benefits associated with breastfeeding the detection of PBDEs in breast milk signals dangerous exposures to the developing fetus and baby. A recent study of human serum in Norway found that infants had higher levels of PBDEs and other brominated fire retardants than any other group. [59] Average PBDE levels for infants aged zero to four years were 2.8 times higher than all other age groups. In-utero transfer and breastfeeding are suspected as major sources of PBDE levels in infant serum. Dust is another possible source, because infants ingest more dust than older children or adults, and these fire retardants are detected in high levels in household dust. [39] The evidence that young children are especially exposed and uniquely vulnerable to these toxic chemicals heightens the need for action to limit exposures.