In the case of BPA, we know that 36 scientists signed a document called the Chapel Hill Consensus expressing their concern about the chemical, but many of these signatories had been expressing concern for years. We also know that USA Today has given prominent coverage to these researchers at the expense of their critics. And on the critical side of the story which says BPA is safe there is a growing number of risk assessments and reviews dismissing the “consensus” concerns as the product of flawed or inadequate methodology and declaring that BPA does not pose a risk at current exposure levels.

Since the European Union’s risk assessment in 2006, there has been a review by Japan’s National Institute of Advanced Industrial Science and Technology (2007), an examination of claims of neurotoxicity by the Norwegian Scientific Committee for Food Safety, (2008), an update to the European Union’s risk assessment  (2008), an evaluation by the French Food Safety Agency (2008), a risk assessment by NSF International, a World Health Organization collaborative center (2008), a review of new data by the German Federal Institute for Risk Assessment (2008), a joint regulatory review for manufacturers by the FDA and Health Canada, a survey by Health Canada (2009), a risk assessment by Food Standards Australia/New Zealand (2009), two more surveys by Health Canada, one on canned powdered infant formula, the second on bottled water products (2009), a hazard assessment by California’s Environmental Protection Agency (2009), and a modeling study of BPA in humans by the German Federal Institute for Risk Assessment (2009).

To reiterate, all of these reviews and assessments found no cause for concern about BPA – and they analyzed the research by the supposedly “growing” band of concerned scientists, subjecting it to statistical and methodological critique.

A survey STATS conducted with George Mason University’s Center for Health and Risk Communication and the Society of Toxicology found that just 9 percent rated BPA as a high risk (by way of comparison, 26 percent rated exposure to sunlight as a high risk).

As for consumers growing concerned about BPA, how could they not be, given the volume of stories telling them they are growing concerned?

McStatistics

USA Today wrote the headline on how not to report percentages in 2004 with “Meth presence surges68% in workplace tests.” In a front-page story, the paper claimed that there was “evidence meth is becoming the workplace's latest drug headache…If use continues to rise at this pace meth will pass cocaine this year as the illegal stimulant of choice.”

Buried at the bottom of the story were numbers that driving this picture of a tripped out workforce. The 68 percent “surge” was a product of a rise from 1.9 meth users per 1000 to 3.2 per 1000; in other words, 99.68 percent of job applicants and workers did not test positive for meth use.

Sample sizes (a crucial factor in determining reliable from unreliable BPA research) have also over-stimulated news coverage at USA Today. For example, a 2008 story “Primary care doctors in short supply” warned that “[p]rimary care doctors are an endangered breed of physician,” and that the survey revealing this trend had a margin of error “slightly less than 1 percentage.” That certainly sounded impressive – especially as the paper reported that “virtually every primary care doctor in the USA as well as 50,000 doctors in subspecialties such as cardiology” were surveyed. What the paper failed to point out was that only 4 percent of doctors completed the survey – a response rate so low it defied reliable prognostication about the future of primary care.

Statistical significance has also been a challenge. The paper reported in 2005 that “Diet study reaffirms red meat as a culprit in colon cancer” But the increased risk for men was not statistically significant, and there was no indication of an association between the disease and meat in women.

The paper has also failed to vet stories it runs from other sources, such as the Associated Press. This produced the memorable claim from a coroner that “Rocking your baby to sleep in a mechanical swing can trigger a deadly attack on the child by the family dog…”

The evidence for running the story? Three cases. A check of dog bite data from the Centers for Disease Control for children in the age group likely to be rocked in a swing found four cases – or 15 percent of all fatal dog attacks on children.

The FDA – as with all the other assessments cited above – does not believe BPA is anywhere close to being a threat, while a brief from the National Toxicology Program said there was “some concern” over some neurotoxicity and developmental endpoints based on a handful of what the director of the National Toxicology Program's Center for the Evaluation of Risk to Human Reproduction the “very limited” studies.

But guess what? The big news last week was that large, multi-generational reproductive toxicity study by the EPA failed to find evidence of neurodevelopmental effects! USA Today did not report this study. In fact, USA Today seems allergic to reporting any of the major studies that have found BPA safe.

According to the NIEHS, animals [sic]studies link BPA with infertility, weight gain, behavioral changes, early onset puberty, prostate and breast cancer and diabetes. New research will focus on low-dose exposures to BPA and effects on behavior, obesity, diabetes, reproductive disorders, asthma, cardiovascular diseases and various cancers. Researchers will also see if the effects of BPA exposure can be passed from parents to their children.”

What USA Today’s Szabo missed was that the NIEHS has done something else that is, arguably, even more important. It has changed its funding requirements for BPA research because much of the work it had funded in the past was deemed to be methodologically inadequate to determine risk. As the NIEHS notes,

Want money for BPA research? Here's the fine print USA Today missed from the NIEHS website...

In order for the data produced to have the greatest impact on our ability to assess the human health effects of BPA, animal studies (developmental exposures or adult chronic exposures) must pay special attention to the following:

• diet (must not interfere with the sensitivity
  of the model to BPA),
• species and strain of animals (must be
  sensitive to estrogenic chemicals at
  low doses),
• sufficient sample size to ensure power to
  detect a statistical difference between
  experimental groups.,
• internal dose of BPA (total and free BPA
  should be measured in blood and if possible
  also in urine throughout the study),
• dose responses (single dose experiments
  are not acceptable),
• phenotype (endpoint must be an actual
  phenotype, disease/dysfunction not
  just toxicity),
• litter must be used as statistical unit for developmental exposures,
• route of exposure should be oral or justified
  to provide similar blood levels as oral route,
• males and females should be used
  when feasible,
• molecular targets and mechanism should be assessed when possible including gene
  expression, receptor binding and epigenetic
  studies. These effects should be linked to
  the exposure and the disease/
  dysfunction endpoints.