California may soon become the first US state to adopt legislation banning the manufacture and sale of children's products containing certain chemicals designed to soften plastics. The initiative follows mounting scientific evidence in recent years about the potential harm of phthalates and so-called endocrine disruptors such as Bisphenol A (BPA). If bill AB 319, introduced by Assembly-member Wilma Chan (D-Oakland), passes, California would be the first state to follow the lead set by a European Union directive six years ago that put an emergency ban on baby toys made from plastics containing phthalates. At the time, European ministers said their decision was based on the "precautionary principle" and they would review it "in the light of new scientific data."
Data indicate that dangers do in fact exist, and bans set to limit childhood exposure might not go far enough. Yet many argue over the merit of the model.
Chan says she believes there is enough evidence of potential harm to ban these chemicals. Nevertheless, the chemical industry insists that phthalates and BPA are harmless. Gilbert Ross, medical director of the American Council on Science and Health, describes the California bill as a "scare campaign." Steven Hedges of the American Plastics Council says, "The potential human exposure to BPA... is extremely low and poses no known risk to human health," citing safety assessments from European and US regulatory agencies. But Frederick vom Saal, from the University of Missouri-Columbia, argues that these evaluations are outdated. "Most of the important articles have been published since the risk assessment exercises," vom Saal says. He adds that European and US assessments are based on scientific literature published in the 1980s and 1990s. "There are now over a hundred [recent] studies showing adverse affects of BPA in different animal models." But controversy surrounds the choice of animal models as well as the relevant dose ranges for testing.
Ross says that tests performed on rodents cannot be extrapolated to indicate potential human health effects. "Adverse effects of chemicals on rats cannot even predict similar effects on mice, much less on people."
Thomas Zoeller, a thyroid endocrinologist at the University of Massachusetts, Amherst, contends that rodent models are reasonable both for practical and theoretical reasons. But he notes the difficulty in extrapolating between species when it comes to endocrine issues. Other experts echo these concerns. "The mouse is not a good model for the human in terms of fetal effects of estrogens," says Richard Sharpe at the Centre for Reproductive Biology in Edinburgh, Scotland. Murine estrogen levels are only modestly elevated in pregnancy compared with humans, Sharpe says. "So, BPA exposure may add more to the estrogen burden in the mouse," he states. However, vom Saal rejects this argument, explaining that it is the amount of activated estrogens that are important and that small changes in endogenous estrogen levels have been linked to human birth defects.
Appropriate dosing poses another problem. "The dose issue is critical in any study of environmental estrogen effects," says Neil MacLusky of the Helen Hayes Hospital in New York, "because it is possible, with high enough doses, to see responses to just about anything." Regulatory agencies attempt to apply a substantial safety factor when extrapolating from animal models to the human situation. For example, the US Environmental Protection Agency (EPA) calculated the acceptable daily intake level (known as the Reference Dose, or RfD) for BPA by dividing the rodent "lowest effect" level of 50,000 μg/kg/day by 1000 to obtain an RfD of 50 μg/kg/day. The European Commission adopted a more conservative acceptable daily intake level of 10 μg/kg/day.
Nevertheless, many recent studies have highlighted potential harmful effects of very low doses of phthalates and estrogenic compounds. These studies raise questions as to the actual human exposure levels. "Exposures to BPA are estimated in developed countries to be well below the safety limits," says MacLusky. "So, for the average person in the street there is not an immediate cause for concern." The situation changes, however, in the case of pregnant women, as BPA may be concentrated in the human placenta and amniotic fluid. MacLusky notes that fetal development represents the most sensitive period of life in terms of exposure to chemicals like phthalates and BPA.
To explore these issues, vom Saal teamed up with reproductive biologist Barry Timms at the University of South Dakota to investigate the effects of estrogenic compounds on pregnant mice. A decade ago they had found that fluctuations in the levels of the natural endogenous hormone estradiol induced changes in prostate development in utero. "The recent awareness that many environmental chemicals act as hormone mimics led us to study the consequences further, using mouse prostate development as an animal model," says Timms.
They fed low levels of BPA (10 μg/kg/day) to pregnant mice for four days.1 They then removed fetuses by cesarean section and removed the prostates from male embryos. Using sophisticated 3-D image reconstruction, they made precise measurements of the developing prostate. As a positive control, Timms' group used diethylstilbestrol (DES), an estrogenic compound similar in structure to BPA and associated with reproductive organ defects and cancer in humans.
Timms and vom Saal found that low-level exposure to the compounds caused an increase (up to 40%) in the size and the number of the prostate ducts, and increases in proliferation of the basal epithelial cells of the primary ducts. They also observed a narrowing of the bladder neck and urethra. The researchers suggest that the observed deformities could predispose animals to prostate cancer and bladder disease in later life, and vom Saal says that the virtually identical effects of BPA and DES make a strong case for the relevance of the rodent model and the potential human hazards.
Researchers in the field agree that the Timms study has carefully examined the issue of low doses. "The doses used in [the Timms] study are well within the range of what is currently stated by both European and US government agencies to be 'safe' for human consumption," notes MacLusky. Zoeller says that the study is particularly rigorous with respect to the sampling and the high anatomical resolution.
Nevertheless, researchers are divided over the study's relevance to human health. "I have doubts as to what it means," says Sharpe. "Nobody, as far as I am aware, including the authors, has shown that the effects that they report (i.e. an increase in prostatic budding, etc.) actually lead to any adverse change in the prostate in adulthood." Michael Joffe, an epidemiologist at Imperial College in London, agrees. "Whether their endpoint is relevant to future prostate cancer, as they claim, is difficult to judge," Joffe notes.
PASSING ON PROBLEMS
This is just one of many recently published works cautioning about the developmental effects of plastic components. Ana Soto and colleagues at Tufts University Medical School in Boston showed that environmentally relevant doses of BPA affect mammary gland development in mice.2 Shanna Swan's group at the University of Rochester School of Medicine and Dentistry, New York, reported an association between male genital defects and phthalate exposure in pregnant women.3 Also, in collaboration with researchers at Yale University, MacLusky found evidence for BPA effects on brain development in rodents.4
Joffe points out that "the interest in male reproduction was sparked by interest in male fertility – the falling-sperm-counts story – and the widespread rise in testicular cancer throughout the developed world. [But] none of these findings can explain the descriptive epidemiological findings that started the concern." The rise in testicular cancer, for example, began about a hundred years ago, before the introduction of BPA or phthalates.
Michael Skinner from the Center for Reproductive Biology at Washington State University, says, "The take-home message from these recent studies is that the fetal basis of disease is becoming much more critical than we previously thought." His group recently found that the effects of high-dose antiandrogenic compounds could be passed on epigenetically for up to four generations in rats.5 Skinner has begun to study the effects of environmentally relevant doses and is investigating other phenotypes, such as cancer and prostate disease. Skinner says, "Potential biohazards of environmental toxins now need to be reevaluated in terms of transgenerational effects." Moreover, he reiterates the susceptibility during fetal development.
These conclusions will have consequences for future legislative initiatives, which are currently based on exposure to infants. "That's inappropriate," says Skinner, who argues that fetal, prenatal, and adult exposure should all be rethought. The EPA is currently considering this issue. "The Californian bill may not go far enough," says MacLusky, "since the exposure of women during pregnancy may be an even greater cause for concern than the exposure of children. In the case of children and babies, it is essential to err on the side of caution."