Saturday, October 18, 2014

BPA Research Funding Linked to Researcher Bias?

From Angela Logomasini of the Competitive Enterprise Institute.
"The number of studies that have appeared in the news during recent years on the chemical bisphenol A (BPA) is staggering. Few substances undergo such scrutiny. So why BPA? Mattie Duppler of American’s for Tax Reform’s Cost of Government project answers that question in an article for The Hill’s Congress Blog: Congress has poured millions of dollars ($170 million since 2000) into BPA research for what amounts to little more than a witch hunt.
Follow the money and you may find a strong statistical association between government funding and the increased number of research studies that link BPA to various health ailments.

Money goes out to researchers motivated to produce studies that report positive associations that easily get published and that gain more funding. And the more money politicians spend for research studies, the more likely some portion of studies will come up with positive associations between BPA and various health aliments, which is likely to happen by mere accident. In addition, many positive findings appear to be attributable to activist agendas among some researchers who make creative interpretations of largely meaningless data. And the studies that come up negative usually don’t get published or end up in the news either because negative findings as simply not interesting.

Thus far, the allegedly most damning studies on BPA are extremely weak. Most don’t really find what the researchers claim they do, and they are often poorly designed. Consider the latest BPA study in the news. Published in the Journal of the American Medical Association Pediatrics (JAMA Pediatrics), it claims that BPA is associated with wheezing and reduced lung function in children.

Specifically, the authors conclude:

These results provide evidence suggesting that prenatal but not postnatal exposure to BPA is associated with diminished lung function and the development of persistent wheeze in children.

That claim has produced alarming headlines like: “BPA Exposure During Pregnancy Linked To Lung Problems In Children (STUDY),” “Kids exposed to BPA before birth at risk of wheeze study says,” and “BPA Tied to Breathing Problems in Kids: Study.”

Yet, despite what the authors or the headlines say, a study that is simply “suggesting” a finding, offers no real “evidence” of anything. That’s simply doublespeak. And apparently, the entire study is doublespeak. The authors themselves point out many weaknesses of their study, which arguably should have led them to a different conclusion.

The study measured BPA in women’s urine during pregnancy and took measurements of BPA in children’s urine sampled once a year for five years. Then the authors ran statistical models to correlate “BPA exposure” with semi-annual reports from the mothers on how much their children wheezed over a five-year period as well as with medical measurements of lung function. The researchers said they found that some of the children with higher BPA exposures reportedly wheezed more or had more lung function issues than did the rest, although no association was found after children reached five years old.
There are numerous problems here. First, the authors did not have enough information to accurately measure the long-term BPA exposures of any of the subjects. BPA levels can vary substantially over time—even over just a few hours—and a few “spot measurements” are not good proxies for actual exposure over time. The authors allege to have at least partially addressed this problem by doing a series of measurements. But their series amounted to just two measurements of BPA during each woman’s pregnancy! And the children were measured just once a year for five years, but many subjects did not complete all the visits for BPA measurements and measurements of lung function. So, at best, the study authors report, “The BPA and spirometry data [medical measurement of lung performance] were available for at least 1 time point (age 4 or 5 years) for 208 children (155 at age 4 years and 193 at age 5 years).” These few annual spot measurements among a portion of the sample are simply not sufficient to establish BPA exposure levels.

More importantly, the study measured BPA levels in urine, simply showing humans metabolize the BPA and pass it out of the body. They provide no evidence that the babies were exposed in the womb nor do they show that BPA has any impact on the human body at any point in time. In fact, most evidence indicates that BPA passes out of the body without having the opportunity to have any health effects, which is an issue the authors never addressed.

The researchers also admit that they only found associations within subsets of their data and that such associations were “marginal” as well as “limited and inconsistent.” Pair that with the fact that the data on lung function were incomplete and data on wheezing were reported by mothers every six months, making recall bias a potentially big factor. They further admit that their sample was not particularly random and that there may have been other confounding factors not addressed in the study.

A blog post published by the European industry group, the BPA Coalition, sums up some of these obvious problems and also rightly notes that the study is also too small for drawing conclusions and the limited statistical associations it found do not build the case for a plausible cause-and-effect relationship. While some may question the BPA Coalition’s motives because it represents industry, you can’t argue with the points they make; the facts speak for themselves.

This study is a classic example of hype and exaggeration using largely meaningless data. It does not begin to challenge the body of solid research and weight of the evidence showing that consumer exposures to BPA and the resulting risk is exceedingly low, outweighed by the substantial benefits of this chemical."

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.