Clearly Concerning

Do common plastics and resins carry risks?

It’s hard to avoid bisphenol A. One of the highest-volume chemicals in commercial production, it’s the starting material used to make polycarbonate plastics. Those are the hard, clear plastics used in baby bottles, flatware, watercooler bottles, and the work bowls of food processors. Bisphenol A (BPA) also serves as an essential ingredient of epoxy resins used to line food and beverage cans and even to seal cavity-prone teeth.

PLASTIC? OH NO! Plastic household products that are hard and clear are typically made of polycarbonate plastic, which tends to leach bisphenol A with age and after heating. When the plastic crackles, as often happens, the leaching accelerates. Raloff
TELLTALE BLOND. Agouti-breed mice ordinarily grow into svelte, brown specimens (right). However, when a fetus is exposed to bisphenol A, it can turn into a blond, obese adult (left), showing signs of a gene alteration. Duke Univ. Medical Center

But BPA doesn’t stay put. It inevitably leaches into foods and people’s mouths, such that traces of the chemical now show up in everyone’s body.

The universal presence of BPA has raised concerns because hundreds of animal studies have shown that this largely unregulated pollutant can tinker with the development and function of a wide range of tissues. These studies show, among other effects, that BPA can alter rodents’ and other lab animals’ sex-specific behaviors, perturb developmentally important hormones, boost fat cell numbers and their accumulation of lipids, foster precancerous changes in cells, and induce insulin resistance, a harbinger of diabetes.

If all of this happens in animals, can any of it happen in people too?

That’s what the National Institute of Environmental Health Sciences decided to investigate, explains Jerrold J. Heindel, who works at the institute in Research Triangle Park, N.C. Two years ago, its National Toxicology Program (NTP) recruited two panels of experts to review masses of data on BPA’s reproductive and developmental effects. Last month, these panels issued reports offering different—and in some ways conflicting—assessments. One panel found many areas of concern. The other turned up few.

Ultimately, NTP will issue a single report that integrates conclusions from both panels, along with any new information on BPA that comes to light during the next few months. That report probably won’t emerge for at least a year, says the institute’s Michael D. Shelby, whose office will prepare the final document.

In the meantime, how worried should consumers be?

Previous evaluations “support the conclusion that BPA is not a risk to human health at the extremely low levels to which consumers might be exposed,” according to a statement issued last month by the American Chemistry Council, a chemical-industry group based in Arlington, Va. It interpreted a recent report on BPA by the European Food Safety Authority as indicating that “consumers are not at risk from use of products made from BPA.”

Such reassurances don’t satisfy a number of BPA researchers, however—among them Randy Jirtle of Duke University in Durham, N.C. He recently published a rodent study showing that fetal BPA exposure can reprogram lifelong gene activity in the agouti breed of mice and even change the animals’ coat colors (SN: 8/11/07, p. 84). Those data alone prompt Jirtle to say that “if I was a woman who was pregnant—or thinking about becoming pregnant—I would try hard to avoid exposure to BPA.”

The chasm between such opinions explains why NTP’s judgment is so eagerly awaited. Moreover, Heindel says, it emphasizes why human studies that can confirm or refute BPA effects seen in animals and test tubes must become a research priority.

Glass half empty

Ample evidence exists that BPA can harm lab animals at concentrations below those already occurring in most people. That was the primary conclusion of a consensus statement published in the August-September Reproductive Toxicology by 38 scientists on one of the two NTP panels. Several additional scientists on the panel declined to sign the statement because they work for government agencies that didn’t want them weighing in officially or because they professed inadequate expertise on certain topics, says Heindel.

He commissioned the group to evaluate the strength of data from more than 700 BPA studies. Participants met last fall in Chapel Hill, N.C., to review their findings and identify human-health concerns about which they were “confident” and ones they deemed “likely.”

The panel labeled as “confident” its assessment that BPA at low doses has had negative effects on experimental animals.

For example, the panel concluded that BPA exposure in the womb can permanently alter genes of animals, impair the function of organs in ways that persist into adulthood, and trigger brain, behavioral, and reproductive effects, including diminished sperm production. Effects deemed likely included a heightened sensitivity to carcinogens, impaired immunity, and diminished insulin sensitivity.

The scientists also expressed confidence that many of these effects can be explained by data from test-tube studies of the chemical’s properties.

Although the panel didn’t officially put the “confident” or the “likely” label on any human effects of BPA, “the consensus was that there is no reason to think that effects that occur in animals in response to low doses of BPA would not also occur in humans,” says participant Frederick S. vom Saal of the University of Missouri in Columbia. Although his and others’ studies have identified an animal’s most critical windows of susceptibility to harm as its time in the womb and shortly after birth, he notes that the scientists agreed that BPA exposures even in adult animals can trigger adverse effects.

Peer-reviewed summaries of the panel’s conclusions in the areas of human exposures, molecular mechanisms, rodent data, carcinogenicity, and wildlife effects appear in the August-September Reproductive Toxicology.

Glass half full

At a meeting in Alexandria, Va., less than a week after Heindel’s panel unveiled its conclusions, a second expert panel—organized by NTP’s Center for the Evaluation of Risks to Human Reproduction—finished roughing out its review of the data in 450 to 500 published scientific papers. These experts concluded that current BPA exposures appear to pose little risk to people.

The only substantial exception to that conclusion, this panel reported, was that exposure to the chemical might perturb neural development in the womb or shortly afterward.

Panelist Jane Adams, a neuroscientist at the University of Massachusetts, Boston, observes that only “a handful of papers—fewer than 10” raised “red flags” suggesting neural damage from BPA. In some studies, various brain parts of animals exposed to BPA during development later exhibited altered numbers of cells or of cellular receptors that respond to hormones. One paper described altered behavior—such as a diminished propensity for male rodents to explore a novel environment, a trait more characteristic of females.

Although no human study has ever suggested comparable impacts, Adams says that the panel expressed concern on these points because BPA blood concentrations associated with the animal effects are comparable to concentrations known to exist in people. That suggests, she says, that BPA residues in pregnant and lactating women might pose risks to their babies.

Ingestion versus injection

Why did the two NTP panels come to such different conclusions about the potential risks of BPA to people? Expert-recruiting strategies suggest an answer, vom Saal says.

Most participants in the Chapel Hill meeting, Heindel says, were selected on the basis of their experience in conducting studies on BPA. Other panel members had substantial experience with other pollutants that can mimic estrogen. All these scientists knew the good qualities and shortcomings of past experiments in the field.

Members of the panel that met in Alexandria, by contrast, were selected precisely because they had no direct BPA experience and, therefore, no obvious vested interest in judging the quality of data on the chemical. The team members’ experience spanned a range of disciplines, including toxicology, neuroscience, statistics, and reproductive health.

This second panel rejected many of the studies that had raised concerns for the first one. For instance, the Alexandria group largely discounted findings from animal studies in which BPA had been administered by injection rather than by mouth.

The reason for that decision, the Alexandria panelists explained, was their concern that anything but oral administration of BPA wouldn’t represent the normal route of the chemical into people’s bodies. Ingested compounds enter the blood and then circulate to the liver, which can filter out some BPA. It is then shed in urine. Injected agents can bypass the liver and potentially build up unrealistically high BPA concentrations in the body, this panel worried.

That’s a valid concern, especially for studies evaluating adult exposures, agrees Patricia A. Hunt of Washington State University in Pullman, a member of the Chapel Hill panel. However, her own research group has evidence that oral and nonoral administrations of BPA can have comparable impacts.

In a study of genetic effects on fetal mice, her team administered the chemical orally or via slow-release pellets implanted under pregnant animals’ skins. In the January PLoS Genetics, Hunt and her colleagues reported no difference in the genetic effects from either dose.

Via both routes, low doses of BPA to mother mice affected their female pups. The daughters’ chromosomes were less stable than normal when the pups grew up and mated. Upon fertilization, their eggs’ genes exhibited error-prone separations and copying, leading to chromosome abnormalities in some 40 percent of fetuses developing in female mice whose only exposure to BPA had been in the womb. That’s at least 20 times the incidence of such abnormalities in mice unexposed to the chemical.

“We were stunned to see this effect of an estrogenic substance,” Hunt told Science News.

The BPA dose producing the effect was small: 20 micrograms per kilogram (µg/kg) of body weight per day, which is 40 percent of what the Environmental Protection Agency has judged, extrapolating from other animal data, to be the likely “lowest observable-adverse-effects level,” or LOAEL, in people.

The researchers used BPA-laced implants in the experiment because getting daily measured doses of a substance by mouth stresses pregnant mice. “It’s a good way to lose [fetal] pups,” says Hunt.

Vom Saal adds that the liver in fetal mice isn’t very effective at removing toxic agents anyway, so slow-release implants and even injections probably deliver chemical exposures to the animals comparable to an oral dose.

Indeed, physiologist Angel Nadal of Universidad Miguel Hernández de Elche in Alicante, Spain, and his colleagues obtained comparable effects with oral and implanted BPA in their study using 100 µg/kg doses. Even a single dose by either method triggered insulin resistance in mice.

Nadal says that such doses produce blood concentrations comparable to values that have been recorded in pregnant women. However, in the January 2006 Environmental Health Perspectives, his team reported that it took doses of only 10 µg/kg BPA to impair insulin and blood sugar regulation in mice.

This month, Nadal and his team are launching a study to evaluate whether long-term oral exposure to BPA triggers type 2 diabetes or obesity.

New developments

Some BPA studies were too new to be evaluated by the NTP panels. For example, researchers led by Ana M. Soto of Tufts University in Boston and her colleagues recently reported that a pregnant rat’s exposure to low doses of BPA “resulted in early puberty in female offspring.” Moreover, those daughters’ mammary tissues exhibited changes suggesting elevated susceptibility to cancer. Indeed, when the scientists subsequently exposed these daughters to a carcinogen, the rats were more likely to develop abnormal tissues or outright cancer than were rats with no exposure to BPA, pre- or post-natal. The team reported these findings in the January Environmental Health Perspectives.

Vom Saal’s team reported in the same journal in June that genital cells from fetal-male mice develop additional cellular receptors for estrogens and androgens—female- and male-sex hormones, respectively—when exposed to low doses of BPA. These extra receptors magnified the cells’ sensitivity to the sex hormones, which in other studies have been shown to fuel prostate cancer growth. The authors note that the cell-altering effects of BPA “occurred within the range of concentrations currently measured in human serum.”

These findings may help explain a finding in rats by a group led by Gail S. Prins of the University of Illinois at Chicago. The researchers found that exposure to low doses of BPA in the womb increased the tendency of prostate glands in adult-male offspring to become precancerous upon exposure to extra estrogen.

“Men develop elevated estrogen levels, relatively speaking, as they age,” Prins notes. Although BPA didn’t appear to cause cancer directly, she notes that naturally produced estrogen can. In the June 1, 2006 Cancer Research, her team showed that BPA-exposed animals exhibit a heightened cancer vulnerability to estrogen concentrations typical of advanced age.

Her team also showed that BPA can reprogram genes in the fetal prostate in ways that affect a cellular process that’s been linked to cancer development.

Retha R. Newbold of the National Institute of Environmental Health Sciences led a study that found that female mice exposed to BPA as fetuses had a dramatically increased risk of developing uterine cysts, precancerous changes, and additional types of reproductive-tract disease in middle age (SN: 8/11/07, p. 84). The team’s findings appear in the August-September Reproductive Toxicology.

Finally, Hiroshi Masuno of Ehime Prefectural University of Health Sciences in Japan says that his group will publish data in the next few months on the obesity-fostering potential of BPA in mice. In the new study, the exposure of pregnant mice to BPA increased their offspring’s adipose tissue mass, serum-cholesterol concentrations, and blood-triglyceride readings, the biochemist says.

Help wanted

The biggest shortcoming of data on BPA’s effects is that human studies are all but nonexistent, Heindel says. However, he notes, concerns articulated by the two NTP panels will guide what kind of human studies his agency should fund.

Where do all these hints of danger from a ubiquitous chemical leave consumers? “I don’t want to cause undue worry,” Newbold says, “but I don’t think we can ignore the possibility BPA can pose risks. There’s just too much science [suggesting it can].”

As a precaution, she recommends that pregnant and nursing women limit their BPA exposures “in whatever way they can.” One of her tips: If a new mom must use plastic kitchenware, keep it out of the microwave and dishwasher. That’s one answer that BPA research has already provided: Heating polycarbonate plastics frees more BPA, which then leaches into foods.


Janet Raloff is the Editor, Digital of Science News Explores, a daily online magazine for middle school students. She started at Science News in 1977 as the environment and policy writer, specializing in toxicology. To her never-ending surprise, her daughter became a toxicologist.