STATS ARTICLES 2008

2009 | 2008 | 2007 | 2006 | 2005 | 2004 | 2003


Why Journalism is Failing the Public on the Risk from Plastics
Rebecca Goldin Ph.D and Trevor Butterworth, May 6, 2008
In the media's rush to report a health scare over bisphenol a (BPA), crucial facts and scientific consensus have gone missing.

The question of whether bisphenol A (BPA), a chemical in plastics, is safe has been hotly debated in scientific circles and in the media in recent months, particularly given the recent news that the Canadian government is considering imposing a ban on the chemical. But it’s far from clear that the discussion scientists are having is the one that’s being reported in the media. Given the highly-selective and judgmental way many journalists have covered the scientific evidence on BPA,  the public is not getting a complete picture of the evidence even as the clamor for political action to ban the chemical threatens to overrule the present regulatory consensus in Europe, Japan and the U.S. that BPA is safe.

First, let’s start with the current scientific consensus, to the extent that there is one. The U.S. National Toxicology Program recently compiled a Draft Brief on BPA, in which experts evaluated the current state of the literature on BPA. As they say of their task,

“The Draft NTP Brief on Bisphenol A is not a quantitative risk assessment, nor is it intended to supersede risk assessments conducted by regulatory agencies. The Draft NTP Brief on Bisphenol A is intended to serve as an environmental health resource for the public, as well as regulatory and health agencies.”

The goal is to give a qualitative assessment, which makes it, perhaps, harder to determine what the risk is. The NTP uses five categories of “concern” in non-technical language and in order of worry: negligible concern, minimal concern, some concern, concern, and serious concern. The brief was based on the conclusions of an expert panel, public comments, and additional scientific papers not used by the panel of experts (either because they had been dismissed or because they were published after the panel convened). Here’s the “bottom line”, taken from the NIEHS website:

Can BPA affect human development or reproduction?

“Possibly. Although there is no direct evidence that exposure of people to bisphenol A adversely affects reproduction or development, studies with laboratory rodents show that exposure to high dose levels of bisphenol A during pregnancy and/or lactation can reduce survival, birth weight, and growth of offspring early in life, and delay the onset of puberty in males and females. Recognizing the lack of data on the effects of bisphenol A in humans and despite the limitations in the evidence for "low" dose effects in laboratory animals, the possibility that bisphenol A may impact human development cannot be dismissed. More research is needed.”

Instead of dissecting the words in this summary, let us point to the specifics.

  1. some concern for neural and behavioral effects in fetuses, infants, and children at current human exposures.
  2. some concern for bisphenol A exposure in these populations based on effects in the prostate gland, mammary gland, and an earlier age for puberty in females.
  3. negligible concern that exposure of pregnant women to bisphenol A will result in fetal or neonatal mortality, birth defects, or reduced birth weight and growth in their offspring.
  4. negligible concern that exposure to bisphenol A causes reproductive effects in non-occupationally exposed adults and minimal concern for workers exposed to higher levels in occupational settings.

So for all those guilty-feeling parents out there, worrying about whether they have harmed their children by feeding them from a bottle leaching BPA, let’s just read the letter of the assessment. There is a possibility but not a certainty that BPA is causing problems. Time will tell how this all pans out, but these experts have refrained from “serious concern” or even “concern” on the issue. It may be that avoiding additional exposure is a prudent move, but finding out your baby bottles contain BPA is not like finding out your lead paint is chipping (where the science behind lead contamination is irrefutable). As the FAQ of the NTP brief explains:

“The scientific evidence that supports a conclusion of some concern for exposures in fetuses, infants, and children comes from a number of laboratory animal studies reporting that "low" level exposure to bisphenol A during development can cause changes in behavior and the brain, prostate gland, mammary gland, and the age at which females attain puberty. These studies only provide limited evidence for adverse effects on development and more research is needed to better understand their implications for human health. However, because these effects in animals occur at bisphenol A exposure levels similar to those experienced by humans, the possibility that bisphenol A may alter human development cannot be dismissed.”

There are other scientific bodies that disagree with the NTP Brief about the possibility of negative impact on humans, notably the European Food Safety Authority (EFSA). Europe, which has even more stringent standards for safety on chemicals than the U.S. noted in its 2006 risk assessment that mice respond to BPA in a completely different way to humans. The EFSA concluded that “the low-dose effects in rodents have not been demonstrated in a robust and reproducible way such as they could be used as pivotal studies for risk assessment.” And based on “an extensive database of repeated-dose toxicity, reproductive and developmental toxicity of BPA in rodents and on the comparison of toxokinetics in primates, including humans, and rodents,” the panel then raised the Tolerable Daily Intake of BPA from 0.01 mg per kg to 0.05 mg per kg. This included a “conservative” uncertainty factor of 100.

So the scientific bodies in Europe and the U.S. are not totally in agreement, and even the more conservative NTP Brief notes that no human impact has been measured and points to the need for more evidence. Indeed, the NTP Brief’s conclusions are based in part on the most controversial aspect of the science: are low-dose studies finding real effects, or are they observing an anomaly because their sample size is too small or their methods are not scientifically established?

As Michael D. Shelby, PhD, director of the National Toxicology Program's Center for the Evaluation of Risk to Human Reproduction told WebMD, the evidence from rodent studies  was "very limited,"

"But our conclusion was that we couldn't dismiss the possibility that similar effects might occur in humans.This is a very controversial area with obvious differences in the way different groups have interpreted the data. We have tried to do it in as scientifically sound and thorough means as we could. Even in the areas where we've expressed some concern, the literature is not consistent on the endpoints reported. Some people find these effects; others may not, and their relevance to effects on human health -- there's still some uncertainty about that. That's why we didn't have a lower or higher level of concern in our conclusions."

Now, let’s take a closer look at the recent Washington Post coverage of the science behind this issue. The title of the WP piece was “Studies on Chemical In Plastics Questioned,” suggesting that the studies themselves were called into question; the article continues to suggest rather blatently that industry influence corrupted the scientific evaluation of BPA. Whether or not this is the case, the National Toxicology Program did not call into question the studies themselves – in fact, the most controversial studies on BPA (according to the NTP) are the low-dose studies, the very studies that the NTP decided to take seriously in increasing the level of concern for infants, while maintaining that no conclusive human evidence has been found. One of the champions of the low-dose studies, ironically enough, is Frederick vom Saal, the one and only scientific source in the Washington Post story. Professor vom Saal is a biologist who has argued for the past decade that BPA poses a grave risk to infants. As the Post reported:

“A decade ago, Frederick vom Saal, a reproductive scientist at University of Missouri at Columbia… theorized that because BPA can mimic estrogen, a female sex hormone, minuscule amounts introduced to fetuses or infants could change cell structure and cause significant health problems later in life. He found that doses 25,000 times below what the government has labeled as safe harmed developing cells in mice…”

What the Post didn’t tell readers was that independent scientists – not only in America, but also in Europe (commissioned by the EFSA to do a risk assessment focused on reproductive effects) and in Japan, specifically, explicitly and repeatedly rejected vom Saal’s research on the grounds of method and statistical power.

Normally, when Europe bans a chemical it creates banner headlines in the U.S. media. Curiously, Europe’s decision not only not to ban BPA but to increase the level permitted has been almost entirely ignored by the U.S. media, of which the Washington Post is the latest example.

The Post didn’t tell readers that the science is not established in part because it’s so difficult to tease out the effects of one chemical versus other influences on people’s lives. The Post didn’t tell readers about the lack of understanding about what happens to BPA in the human body, and why animal studies sometimes do and sometimes do not give us useful information. For example, many of the studies rejected by BPA expert panel were based on what happens when BPA is injected instead of ingested. Some groups have been highly critical of excluding these studies, while others insist that the route of exposure is essential because of how we process BPA.

Instead, the Post has tried to portray the regulation of BPA as something which might have been rigged by industry and the regulatory bodies, and the sourcing follows that line of suggestion. A fair-minded and rigorous reporter would have tested that hypothesis by thoroughly gauging the range of scientific opinion on BPA and considering the objective standards which lead scientists to the weight of evidence for there being a risk or not.

For instance, the Post appears dismissive of studies funded by industry without acknowledging that these studies, precisely because they are well-funded, provide the statistical power to draw valid conclusions that many of the independent studies lack. That’s why many of the independent studies on BPA have repeatedly been rejected by risk assessments around the world: their sample size is too small; if they are low dose experiments, they fail to assess a range of doses; the route of exposure is different from human exposure (and as a result creates different effects).

In fact, if reporters read the European Risk Assessment, they would see a detailed explanation of why many of the studies claiming a risk from BPA were junked. The same studies were rejected by Center for Evaluation for Risks to Human Reproduction in the original BPA panel of experts. Some of those studies were then reconsidered by the government in preparing the current BPA Brief, and played a role in the increased level of concern. The idea that corruption and money were the only influential forces in evaluating these complicated scientific issues is an insult to those scientists who served on the various panels.

A story with such profound implications for public well-being should not report the findings of scientist with a vested personal and economic interest in a particular outcome without balancing them against the perspective of independent toxicologists who understand the fundamental scientific issues in this debate. And given that multiple risk assessments conducted by many independent experts across different countries disagree with vom Saal on objective, verifiable scientific grounds, to report the following in the kicker to the Post’s story is quite simply astonishing:

“Vom Saal said a flood of recent BPA studies have validated his work. "The scientific community basically said, 'This argument is over,' " he said. "It ended a long time ago. There's only been an illusion of a controversy created by a well-financed public relations outfit. The idea that the FDA tells people this is safe is offensive."

If the reporter had gone to any of the research databases to see if that quote was true, they would have found the exact opposite, which on the most basic journalistic grounds should flag vom Saal’s bias as a source.

Just to cite the most glaring example, there was a massive two-year risk assessment of BPA conducted by NSF International, an independent not-for-profit organization devoted to consumer affairs, which was published in February’s edition of the Journal of Toxicology and Environmental Health.  The lead author was Calvin Willhite, a toxicologist with California’s Department of Toxic Substances Control. The risk assessment rejected the research claiming a risk to humans from BPA.
 
The problem with the coverage of BPA in the media is that the scientific community has been systematically ignored by the press in their rush to report a health scare and an industry-brokered cover up. The Washington Post isn’t alone in creating this spin; there are dozens of articles by other publications that all reflect the same point of view, because they all have the exact or similar sourcing.

The point is not about being skeptical of industry or blithely rushing in to support industry: it’s being explicit about the criteria for why one study is included in a risk assessment or excluded. It’s about understanding the strengths and limitations of research. It’s about thinking scientifically when reporting science.

Our reporting this fact has led some to decry us as industry-funded and denounce us as “astroturfing” on BPA. We have not and do not receive industry funding. We have no stake in whether BPA is banned or not. Our stake is on whether people panic and waste their time and energy on a media-hyped health risk rather than a real one. Of course, BPA could be a genuine risk in the end; but when journalism determines the scientific outcome, rather than vice-versa, the public loses.

There is a societal cost to focusing on threats without carefully considering what the data really says, and it is not just in economic terms, as politicians and activists clamor for action to eliminate the perception of risk, it is that the consequences of banning something may well be worse than not banning something. This is why our interest in BPA is simple: report the full story, not just the side of it that makes a better, more gripping, more alarming headline.

Digg!

Technorati icon View the Technorati Link Cosmos for this entry