How America became obsessed with BPA
In the words of Frederick vom Saal, a scientist at the University of Missouri who has led the charge against BPA for over a decade, the chemical was, as he told Discover Magazine in May 2008, the biological equivalent of global warming. And if BPA was just like global warming, it meant that those who begged to differ were – the just had to be – chemical industry stooges. An article by a science writer, Linda Gross, in PLoS Biology got the ball rolling by purporting to explain how the chemical industry conspired to undermine vom Saal’s evidence. It was hailed by Salon’s Andrew Leonard as “a superb exposé,” without any mention that the primary source and focus of the article was vom Saal and that it quotes him and his collaborators extensively and without any serious corresponding attention to the voluminous criticism of his and their experimental methods.
In fact, what has gone missing in the coverage of BPA across the media was that the body of research hailed by vom Saal as demonstrating that BPA was dangerous at low doses had been studied and rejected as scientifically invalid or irrelevant by the European Union’s Food Safety Authority (EFSA), individual European countries, and Japan – along with various evaluations in the United States. As EFSA’s current BPA panel told STATS, “many of the studies indicating low dose effects of BPA were contradictory and not well conducted.” These flaws were first pointed out by the National Toxicology Program in 2001, when it convened a panel of statisticians who then identified the statistical limitations in the way the experiments were conducted and, crucially, how they should be corrected. (These criteria and their significance are discussed below).
Every risk assessment that followed, whether in the U.S. or abroad used these criteria to determine which studies should carry weight and which shouldn’t. And the net result was that NTP dismissed most of the evidence against BPA in a 2007 evaluation by an expert panel, with one caveat, namely, that there was “some concern” over neurodevelopmental risks, a statement which meant that the available data only provided limited evidence of risk, and that more research was needed. Parsed by the media, this was translated into an admission by the NTP that vom Saal was right all along and that the public was in danger. It didn’t seem to register with journalists that the reproductive risks vom Saal and others had been touting for years were rejected.
Even though “some concern” is the lowest level of concern in the NTP’s nomenclature of risk, the European Union in 2008 called into question whether the NTP was now under so much pressure to respond to the campaign against BPA that it was being much too concerned. EFSA looked at the studies that the NTP relied on to determine “some concern” and said they all had “major shortcomings” and couldn’t be taken as evidence of risk. Denmark’s Environment Protection Agency also looked at these studies and agreed that they weren’t worth worrying about, and France’s Food Safety Agency declared them unreliable. This was particularly noteworthy, given that all these countries use the precautionary principle to regulate chemicals – and are, in theory, more risk averse than the U.S. These developments were largely ignored by the U.S. media.
This overall body of evidence claiming BPA was harmful was also dismissed by a risk assessment carried out by NSF International in 2008, a World Health Organization collaborative center, and one of the leading independent, scientific, consumer affairs organizations in the world. The research was carried out by a prominent American toxicologist, Calvin Willhite, who co-authored the chapter of development toxicity for one of the standard textbooks on toxicology. It received no attention from the mainstream media.
And in a recent and revealing development, the National Institutes of Health, which funded some of this BPA research (including vom Saal’s original work on BPA) under the auspices of the National Institute of Environmental Health Sciences, has tightened its guidelines for research on BPA because too many of the studies it funded in the past have been dismissed by the scientists charged with conducting risk assessments –
“due to a variety of experimental limitations including the use of a single dose, small numbers of animals per group, non-oral route of administration, lack of proper statistics and lack of data on specific phenotypic endpoints.”
The NIEHS will now only fund two-year studies on BPA, of sufficient sample size “to ensure power to detect a statistical difference between experimental groups,” and that the “route of exposure should be oral or justified to provide similar blood levels as oral route,” among with other requirements. In effect, the government body that funded much of the so-called “independent” science supporting the idea that BPA is dangerous now demands the kind of study protocols used by the very industry-funded studies that vom Saal, environmental activists and the media have insisted can’t be trusted. This irony hasn’t received any attention in the media, even though it suggests that had these protocols been required over a decade ago, BPA would not have become the issue it is today.
Journalists in the U.S. were so staggeringly resistant to the idea that BPA was safe that the NSF International research was virtually ignored by the American press – as were the risk assessments in Europe, along with any study that showed that BPA didn’t present a threat to health, even if those studies were independently funded and well conducted from an experimental and statistical perspectiveIt is difficult not to see the cumulative effect of these changes as a tacit admission by the NIH and NIEHS that it has funded a lot of useless studies on BPA – particularly as many of the requirements it now demands of research proposals were deemed essential as far back as 2001, when the National Toxicology Program commissioned a subpanel of statisticians to evaluate the quality of emerging research on BPA.
In other words, for almost a decade, the flaws in research claiming that BPA has been a threat to health have been in plain sight; they were discussed openly and sometimes heatedly by scientists, but always with the same conclusion: the research used to claim BPA was a risk was poorly done, or couldn’t be replicated, or was just experimentally irrelevant. But the dogged failure by most reporters and news organizations to examine and understand why certain methodological choices produced irrelevant or flawed studies has kept the threat of BPA a constant source of news and editorial outrage while suppressing the very science which the public ought to be told about.
In fact, journalists in the U.S. were so staggeringly resistant to the idea that BPA was safe that the NSF International research was virtually ignored by the American press – as were the risk assessments in Europe, along with any study that showed that BPA didn’t present a threat to health, even if those studies were independently funded and well conducted from an experimental and statistical perspective. Instead, newspapers like the Journal Sentinel turned to environmental activists and the scientists whose work had been dismissed to press the case that the public was in danger.
When, for instance, the Milwaukee Journal Sentinel reported (August 16, 2008) on the the National Toxicology Program’s evaluation, it focused only on the “some concern” for neurodevelopment risks and ignored all the “neglible” risk conclusions. The context – what “some concern” meant - was not explained; the criticism of the research by other countries was never reported. Instead, the Food and Drug Administration was lambasted for ignoring the alarm raised by the NTP’s evaluation. Environmental activist groups were quoted chiding the FDA for going down the path of the drug Vioxx, and relying on “industry-funded” studies.
This all provided the kind of narrative urgency that journalism thrives on – the sense of being on a mission to hold truth up to power, to make the newspaper a vital public watchdog . But according to the authors of the European Risk Assessment and the NSF International paper, the Journal Sentinel’s mania to indict BPA (and the other news reports that have mirrored its approach) have made a mockery of science. In the following analysis, they explain in detail why this reporting is misleading and inaccurate, and why the low dose research was rejected as irrelevant or scientifically invalid by their assessments. They explain why the Milwaukee Journal Sentinel didn’t just fail the public in basic toxicology, it failed to follow basic scientific principles in evaluating evidence – the principles that govern whether research is reliable and relevant or not.
The analysis also shows how the paper’s sourcing was dramatically skewed towards environmental activists and a handful of scientists whose common bonds, collaborative interests, and, in one case, mutual financial interests were never disclosed to readers. At the same time, the paper doggedly ignored a vast array of longstanding and breaking evidence that BPA was safe.
The newspaper was also explicitly warned by the head of the National Toxicology Program expert panel that it was relying too much on one scientist whose work and perspective had been repeatedly rejected by international risk assessments. Despite this warning, the paper chose to conduct its own scientific study of BPA at that scientist’s lab, and then gave the results to the same scientist and several of his collaborators to validate and interpret. But when STATS asked NSF’s Willhite to analyze these results, he declared them “erroneous.”
The analysis also looks at the Oakes Award jury convened by Columbia University’s Graduate School of Journalism, and whether the decision to give the award to the paper was affected by a jury sensitized to only one side of the debate.