Issue Date: June 6, 2011
Debating BPA’s Toxicity
Industry livelihood and modern conveniences versus uncompromised human and environmental health advocacy have framed what remains an open-ended debate over the safety of the widely used chemical bisphenol A. Given mounting knowledge of BPA’s estrogen-mimicking properties and possible toxic effects, decisions have to be made on whether it should continue to be freely used, restricted for some uses, or banned completely. There are no simple answers.
On one side of this debate, consumer and environmental advocacy groups who emphasize potential health problems stemming from steady, low-level exposure to BPA support their position by invoking the precautionary principle, which holds that even in the absence of scientific consensus it’s prudent to act as though a potentially harmful substance is, in fact, harmful. They argue that there is already enough scientific evidence of BPA’s toxicity to warrant a ban and that the general population shouldn’t have to be human guinea pigs while waiting for more compelling evidence.
On the other side of the debate, manufacturers and chemical industry groups out to protect a multi-billion-dollar market from regulation are defending BPA as safe. They say the problem with applying the precautionary principle is that it could lead to a ban on a useful chemical that poses no actual danger.
Scientists have played leading roles throughout the debate and tend to align with one side or the other. At times, they have let their emotions drive arguments over experimental design and interpretation of research results. These scientific disagreements may arise because the researchers hail from different disciplines such as endocrinology, toxicology, epidemiology, and chemistry, and because they work in different cultures such as academia, industry, government, and environmental advocacy.
Meanwhile, a consumer revolt is in full swing to remove BPA from food and drink packaging. Since 2008, manufacturers have voluntarily been replacing BPA-containing polycarbonate as the plastic of choice for essentially all baby bottles and toddler sippy cups. Most reusable water bottles are also now BPA-free (see cenm.ag/4).
BPA-based epoxy resin food- and beverage-can liners, however, are still widely used. Most manufacturers maintain that there are few viable alternatives for can-liner applications that match the cost and shelf-life performance of current epoxy resins.
Chemical companies and trade groups aren’t willing to divulge numbers, but market analyst reports estimate global annual BPA demand at up to 12 billion lb and growing at about 5% per year. Polycarbonate makes up 74% of BPA use, and epoxy resins consume about 20% of production. But the baby bottle and food-can applications combined accounted for less than 5% of BPA consumption. That means the phaseouts haven’t put a big dent in overall BPA use.
The BPA debate’s ultimate outcome is still to be decided by government regulatory agencies. These authorities have treaded carefully in assessing the available information. In the U.S. they officially advocate staying on the current course of exercising caution, recommending ways to reduce exposure, and doing more research.
According to the Environmental Protection Agency, people can safely be exposed to up to 50 µg of BPA per kg of body weight per day. It is that number, set in 1988, that keeps the debate alive: The chemical industry is satisfied with it, but many endocrinologists and toxicologists, and by extension environmental health and safety advocates, think it is too high.
One set of 38 scientists came to the latter conclusion while attending a National Institute of Environmental Health Sciences (NIEHS) workshop on BPA in November 2006. Studies dating from the 1990s had reported that amounts of BPA at or below the 50 µg/kg/day safety threshold could cause developmental disorders in lab animals. The findings suggested that low doses of BPA and other endocrine disrupters can lead to adverse health effects, whereas higher doses short of acutely toxic levels might result in no observed effects.
In reviewing hundreds of animal and cell-culture studies, the workshop scientists concluded that human exposure to BPA is likely in the range that had been observed to cause the low-dose effects in animals.
At about the same time, a panel established by the National Toxicology Program (NTP), an interagency program administered by NIEHS, was also evaluating many of the same studies. The panel estimated that formula-fed infants up to a year old ingest up to 13 µg/kg/day of BPA; the levels were lower for breast-fed babies. For adults, NTP estimated BPA intake was up to 1.5 µg/kg/day.
In its final report, issued in September 2008 and taking the NIEHS workshop findings into consideration, the NTP panel stated that it had “some concern” regarding BPA effects on the brain, behavior, and prostate gland in fetuses, infants, and children. “Some concern” represents the middle level of NTP’s five-level concern scale, which ranges from “serious” to “negligible.”
The panel also had “some” or lesser concern for other outcomes, including an earlier age for puberty in girls; fetal or neonatal mortality, birth defects, or reduced birth weight; reproductive effects in nonoccupationally exposed adults; and ill effects in workers exposed to higher levels in occupational settings. The panel based its decision on the fact that although there was credible evidence for the low-dose BPA developmental effects, the experimental results weren’t consistent.
The NTP panel’s analysis was interpreted as running counter to the NIEHS workshop’s conclusions, causing what had been a simmering debate to boil over. Members of the NIEHS workshop charged that the NTP panel was biased by industry influences and had wrongly discounted the importance of the low-dose data in its analysis.
The response that followed was as expected.
“The NTP report did not say BPA is bad; it said there is some concern. You can make that statement about anything,” said Jack N. Gerard, then-chief executive officer of the American Chemistry Council (ACC), the chemical industry’s trade organization and lobby group. “That gives us confidence in the safety of BPA in all its multiple uses.”
BPA critics took the finding the opposite way, emphasizing the panel’s concern about prenatal and early childhood exposures. The Environmental Working Group, a nonprofit advocacy organization, said in a statement that the NTP report “raised concerns that exposure to BPA during pregnancy and childhood could impact the developing breast and prostate, hasten puberty, and affect behavior in American children.”
Some of the endocrinologists and toxicologists who were part of the NIEHS workshop, including Frederick S. vom Saal and John Peterson Myers, criticized the NTP assessment and EPA’s safety threshold in journal commentaries. They faulted the panel for several reasons, but overall for not giving sufficient weight to their low-dose results. Vom Saal is an award-winning endocrine-disruption specialist at the University of Missouri, Columbia. Myers is founder of Environmental Health Sciences, a nonprofit environmental advocacy organization.
Vom Saal and Myers point out that current U.S. health standards for exposure to toxic chemicals rest upon the assumption that high-dose testing procedures used in regulatory toxicology adequately predict potential low-dose effects. But that prediction doesn’t hold for BPA and other endocrine disrupters, they say.
BPA binds weakly to the primary estrogen receptor in people, which is bound more strongly by estrogen hormones as well as natural estrogen-like compounds in foods—so-called phytoestrogens (see page 20). BPA binds more strongly to alternative estrogen receptors, and its physiological effects are thought to be mediated by these alternative pathways.
At high BPA doses—some thousands to millions of micrograms per kilogram of body weight—the chemical displays acute toxicity. But at low doses—from about 0.025 µg/kg/day up to 500 µg/kg/day—things aren’t so cut and dried.
As the dose increases, it changes the suite of genes that are expressed by BPA’s action on estrogen receptors, vom Saal says, which means the impact at low doses, or at doses in between, can’t be predicted from high-dose experiments.
This disrupted biochemistry triggers changes in hormone concentration, enzyme function, protein synthesis, and more, which are of greatest concern for developing fetuses and newborns. Scientists propose that such changes stay with a person over a lifetime and play a key role in the development of diseases.
Overlooking low-dose effects creates the strong likelihood that some health standards currently used to set exposure limits for the American public are too weak, vom Saal and Myers assert. They suggest that based on low-dose research findings, no level of BPA might be safe.
In extrapolating the current safety threshold to the level where low-dose effects are observed, they say, the safety threshold should be some 5,000 times lower—0.01 µg/kg/day instead of 50 µg/kg/day. That level would likely require eliminating BPA from all types of products, including CDs, eyeglasses, and football helmets.
In response to the low-dose findings, the chemical industry commissioned its own large-scale low-dose BPA studies. Among the most influential of these were two studies led by Rochelle W. Tyl and coworkers of the North Carolina-based Research Triangle Institute (now RTI International), an international independent research institute.
In 2002, with funding from the Society of the Plastics Industry, Tyl and coworkers monitored three generations of rats dosed with different levels of BPA and found no adverse effects at low doses. But those results were called into question by vom Saal, Myers, and 34 other leading BPA scientists. Among other reasons, they noted the study used a strain of rats that some scientists believe is insensitive to low doses of estrogen-like compounds such as BPA. Using those animals makes it impossible to adequately assess the consequences of low-dose BPA exposure, they say.
Tyl led a follow-up ACC-funded study in 2008 that used mice instead of the rats. It also found no adverse effects from low-dose BPA. In this case, the study was criticized for, among other reasons, problems with the animal chow fed to the mice—phytoestrogens in the feed may have masked BPA’s effects.
Tyl, who has moved on to other areas of research, is unconvinced that BPA is harmful. The argument against BPA “now transcends data, facts, and results and is driven by an unjustified fear,” she says.
For their part, regulatory agencies have given more weight to the Tyl experiments and others like them because they are large-scale studies and adhered to so-called Good Laboratory Practice (GLP), whereas most of the other hundreds of studies showing negative effects, which were smaller and in some cases considered only screening tests, did not.
GLP is a set of federal rules for conducting research on health effects or safety testing of drugs and chemicals. The rules—which were established in 1978 after a spate of fraudulent toxicology studies on pesticides—spell out how such studies must be executed and reported. EPA’s current safety threshold is based on GLP-compliant animal studies.
Critics of BPA’s continued use state that the standardized assays used in GLP studies are outdated because they are incapable of detecting the types of low-dose effects that state-of-the-art techniques and experimental designs show (Environ. Health Perspect., DOI: 10.1289/ehp.0800173).
“The tools we have been dependent on to assess risk are largely blind to these endocrine-disrupter effects,” Myers says. “We haven’t done a good job advancing standardized assays at a pace that can keep up with new scientific observations.”
Nevertheless, low-dose methods and data aren’t yet being considered concrete enough to be used for BPA risk assessments and regulatory decision making, to the continued frustration of the scientists designing and carrying out the studies (C&EN, Nov. 8, 2010, page 40).
One sticking point has been proving that BPA exposure is causing adverse health effects in people during adulthood like those seen in short-lived lab animals, when there could be many causes. These may include other man-made chemicals or combinations of chemicals coupled with high-calorie diets and lack of exercise.
The first major epidemiologic study to make this connection came out in 2008, at about the same time as the NTP report. A team led by David Melzer of the University of Exeter, in England, analyzed data from the 2003–04 U.S. National Health & Nutrition Examination Survey (NHANES), which was conducted by the Centers for Disease Control & Prevention. The survey found microgram-per-liter levels of BPA in the urine of 93% of the hundreds of people tested.
The Melzer study revealed that people with higher urine concentrations of BPA are at greater risk for heart disease, type 2 diabetes, and increases in liver enzyme activity (J. Am. Med. Assoc., DOI: 10.1001/jama.300.11.1303). The researchers concluded that “these findings add to the evidence suggesting adverse effects of low-dose BPA in animals.”
Last year, when NHANES data were released for 2005–06, Melzer and coworkers repeated their analysis. The newer data showed lower levels of BPA in U.S. adults compared with the earlier data, but the heart disease and liver enzyme associations were replicated; the diabetes association was not apparent in the 2005–06 data (PLoS One, DOI: 10.1371/journal.pone.0008673).
To be fair, the researchers show an association between BPA and the human-health conditions, not proof of causation. Nevertheless, Melzer tells C&EN that he believes these associated effects shown by analyzing independent data sets “should be taken seriously.”
Because testing the effects of BPA directly in people is unethical, scientists looking to clarify the BPA debate are designing studies that can ascertain toxicity without dosing people.
In one such example, a team led by Ruthann A. Rudel of the Silent Spring Institute, a nonprofit environmental advocacy organization, reported in March that BPA levels in the urine of five families in San Francisco dropped on average 66% when they stopped eating canned and plastic-packaged foods for three days and ate only freshly prepared food at home. When the families went back to their normal diets, their BPA levels returned to the original levels (Environ. Health Perspect., DOI: 10.1289/ehp.1003170). The researchers’ conclusion: Removing BPA from food packaging will significantly decrease exposure for adults and especially children.
But these types of studies still don’t connect BPA exposure with ill health effects. In other experiments, scientists are trying to connect adverse effects observed in lab mice and rats to people. Two such studies reported last year further fueled the BPA debate.
In one study, a team led by Daniel R. Doerge of the Food & Drug Administration’s National Center for Toxicological Research dosed adult and baby rhesus monkeys with BPA (Toxicol. Appl. Pharmacol., DOI: 10.1016/j.taap.2010.07.009). The researchers reported that the monkeys metabolized BPA at rates similar to those reported for humans and that their results were “clearly different” from those reported in studies using rats. The team concluded that “any toxicological effect observed in rats in early postnatal exposures to BPA could overpredict those possible in primates.”
In the other study, vom Saal’s group also dosed rhesus monkeys with BPA (Environ. Health Perspect., DOI: 10.1289/ehp.1002514). But the researchers came to a completely opposite conclusion—mice and monkeys metabolize BPA at similar rates as people. As a result, they argue that it is reasonable to directly extrapolate health effect studies of BPA in mice to humans.
Vom Saal and coworkers also concluded that people likely have significant BPA exposure from nondietary sources. They further confirmed earlier studies that show people might be exposed to BPA at up to 500 μg/kg/day to achieve the levels observed in biomonitoring studies. That figure is 10 times higher than EPA’s safety threshold.
In their paper, vom Saal and coworkers reanalyzed the data in the Doerge study, suggesting that those researchers made erroneous assumptions and that their data did not support their conclusions. Vom Saal suggests that the results of the two studies actually are the same. FDA declined to make Doerge available for comment on the vom Saal study.
FDA, taking into account the plethora of scientific studies, reports, and policy statements from environmental groups, industry trade groups, independent labs, and federal agencies—including its own BPA task force—issued a statement in January 2010 in which it agreed with NTP’s analysis of “some concern” regarding BPA. But FDA did not restrict BPA.
Then-FDA deputy commissioner Joshua M. Sharfstein, a physician, commenting on FDA’s position on the safety of BPA for pregnant women and children, said: “If we thought it was unsafe, we would be taking strong regulatory action.”
In addition, in March 2010 EPA added BPA to its list of chemicals targeted for possible regulation under the Toxic Substances Control Act (TSCA) and announced it would be gathering more safety data. As part of the plan, EPA is using its Design for the Environment program to encourage manufacturers to implement green chemistry to redesign products to reduce the chemical’s presence in the environment. But EPA has not restricted BPA use.
Last fall, FDA’s counterpart in Europe announced that it too had concerns about BPA safety. But in its review of some 800 research studies on BPA published since 2007, the European Food Safety Authority found no new evidence to justify lowering its tolerable daily intake value of 50 μg/kg/day, which is the same as the U.S. level.
The European Commission, which had charged the agency with evaluating BPA safety, subsequently invoked the precautionary principle and banned BPA in baby bottles. China recently did the same.
In the absence of U.S. regulatory action, Congress has toyed with the idea of superseding federal agencies by imposing legislation banning BPA. Congress has also considered revising TSCA, which both sides of the debate agree is in need of an overhaul. Indeed, some BPA critics suggest regulatory agencies haven’t acted on BPA because they do not have appropriate regulatory authority under TSCA.
In effect, many people would like to see a reversal of U.S. policy: Abandoning the current approach in which chemicals can be used until proven toxic in favor of a policy more like Europe’s Registration, Evaluation, Authorization & Restriction of Chemical substances (REACH) program, in which the industry must provide data demonstrating a chemical is safe before it can be used in commerce.
For the moment, U.S. federal regulators continue to consider BPA safe by a bare margin. But the debate is not over. New studies are addressing the earlier research design criticisms in order to test the low-dose hypothesis. NIEHS has invested some $30 million on additional BPA research by 10 leading science groups, including vom Saal’s, that will be reaching journals during the coming year. These studies might show more convincingly that the low-dose data can reliably be used to inform regulatory decisions. Or they might show that BPA is safe for continued use.
“There are some experts who are passionately convinced of the counterintuitive fact that small amounts of BPA are more problematic than large amounts, and that a lot of bad things can come from it,” observes John C. Warner of the Warner Babcock Institute for Green Chemistry, in Wilmington, Mass., who has investigated BPA in cash-register receipts. “But then there are equally competent and passionate scientists who believe BPA is perfectly safe. It’s difficult to know who is wrong and who is right. Our society is so frustrated.”
Warner suggests the role of scientists “should not be to come down on the side of being right or wrong, but to quantify the magnitude of the problem and let consumers make their choices, which will ultimately decide what happens in the marketplace.”
- Chemical & Engineering News
- ISSN 0009-2347
- Copyright © American Chemical Society