STATS ARTICLES 2005
In any discussion of phthalates in the media over the past months, readers and viewers have been warned to some degree that this family of chemicals that make plastic flexible has been linked, in some way, to reproductive problems in baby boys.
As with any new and complex scientific research, the media’s attempt to explain the actual study that suggested a link left a lot to be desired: for example, the Wall Street Journal recently summarized the June article by Dr. Shanna Swan in the journal Environmental Health Perspectives as “a small study found that baby boys whose mothers had the greatest phthalate exposures while pregnant were much more likely than other baby boys to have demasculinized traits.”
The only part of that sentence that accurately describes Swan’s article is “a small study.”
What Swan claimed to have found was a correlation between the anogenital index (AGI) of a baby boy, and the level of residual chemicals from phthalates (also called metabolites) in the mother’s urine before his birth. The anogenital index is a measurement of the distance from the anus to the base of the penis, divided by the weight at the time of measurement.
The boys in the Swan study had a smaller than average anogenital index. In rats, under high doses of phthalates, this anatomical change is also evident, as is damage to the reproductive systems. This lead to worries that human reproductive health may be at risk from the chemicals.
As STATS quickly pointed out there were problems with the methodology – and contrary to media accounts such as USA Today’s “common chemical may cause defects in baby boys” – the study did not record any reproductive harm to the babies. (At the time, the New York Times was the only newspaper to note our criticisms.)
And then last week, an independent expert panel convened by the Center for the Evaluation of Risks to Human Reproduction (CERHR), of the NIEHS and National Toxicology Program, to examine the reproductive health hazards and risks of the phthalate DEHP was unable to validate Swan’s key finding.
Swan found a correlation between AGI and four phthalates and argued that there is a statistically significant relationship between a short AGI and abnormal testicular descent (P<.001) – a relationship that is important, because it suggests that a short AGI is an important biological marker for testicular descent. (The “p-value” is a measure of how likely it is that we would see the data from the study purely by chance and not because there’s an actual correlation, or a causal relationship, between the two.)
However, when the review panel did the calculation, they did not find a correlation. The review panel did not have the original data (typically, such data is not released with research papers), but the reported percentages in Swan's paper fall short of her thesis. In other words, Swan's study did not correlate any phthalate with any biological markers for fertility or normal genital development.
The panel contacted Swan to ask her how she got her p-value, and it would appear that Swan was unable to provide a satisfactory explanation. Accounts diverge over whether this constituted an acknowledgement of error or merely a need for clarification. All Dr. Swan would tell STATS is that she “is in the process of clarifying the issue of our calculations regarding AGD and testicular descent in a letter-to-the-editor of EHP.”
Such expert review keeps this important field of research in check, and it will eventually determine whether there are actual associations between phthalates and fertility in this study or other human studies. However, the media and numerous activist groups have cited Swan's research without any regard for literal accuracy or the kind of balanced reporting that would have alerted readers to the underlying problems in the study.
Ironically, on the same day as the NIH panel made public their findings on Swan, the U.S. Public Interest Research Group (PIRG) released a report warning of toxic chemicals in many baby products. The report cited Swan’s study, claiming it “reveals that normal exposure to phthalates can harm genital development of unborn baby boys.” On the basis of this (and other claims about other chemicals), PIRG is calling for the Consumer Product Safety Commission to adopt the precautionary principle and force manufactures to remove such chemicals from their products.
The PIRG report was covered by dozens of news organizations; the NIH panel on phthalates and reproductive health was completely ignored.