Scientist defends study linking phthalates to feminising boys
Professor Shanna H. Swan, of the University of Rochester, has rebutted charges that included she had used “unproven methods to compile questionable data” and “selectively excluded data” to support her conclusions.
The allegations were made by Steve Risotto of the American Chemistry Council (ACC) after publication of a paper in the International Journal of Andrology linking prenatal phthalate exposure to reduced masculine play in boys.
Phthalate exposure via
The study evaluated the playing behaviour of 145 children aged 3 ½ to 6 ½ years who were exposed to higher concentrations of metabolites of two common phthalates, di(2-ethylhexyl) phthalate (DEHP), and dibutyl phthalate (DBP) during pregnancy. The research concluded that boys exposed to these chemicals exhibited less-masculine behaviour – although girls were unaffected. The findings were reached after mothers completed a questionnaire on their children’s play habits.
Phthlates are endocrine disrupting chemicals and have been banned in toys in the European Union for some time. Recent studies have shown that the major source of human exposure to the two phthalates of most concern (DEHP and DBP) is through food, said a statement from the researchers. “These phthalates are used primarily in polyvinyl chloride (PVC), so any steps in the processing, packaging, storage, or heating of food that use PVC-containing products can introduce them into the food chain,” it added.
“Our results need to be confirmed, but are intriguing on several fronts,” Swan said. “Not only are they consistent with our prior findings that link phthalates to altered male genital development, but they also are compatible with current knowledge about how hormones mould sex differences in the brain, and thus behavior. We have more work to do, but the implications are potentially profound.”
Research row
Risotto challenged the research after quoting the scientists as admitting their results were “not straightforward”. But Swan rejected the charge saying the phrase used in the ‘Discussion’ part of the paper related only to the “technical issue” of how to communicate statistical findings relating to regression co-efficients to a non-scientific audience - and not the results themselves. She stressed the research methods they used had been straightforward.
The senior director of the ACC phthalate esters division also said: “Dr. Swan uses unproven methods to compile questionable data to reach conclusions that are consistent with her well-publicised opinion which is not based on the weight of the scientific evidence surrounding the safety of phthalates.”
Swan again dismissed the allegation, saying: “There is nothing we did that is unproven. The Preschool activities inventory has been used in close to 5,000 children, since 1993 and is well standardised. The measurement of phthalate metabolites is routine for the CDC, and the statistical methods we used are taught in any statistical methods class.”
Risotto also charged the researchers of biasing the results by using mothers from a previous phthalate study who had a higher level of concern about the chemical. However, Swan said that as none of the mothers knew their own phthalate levels, there was no way their earlier participation could have biased the results.
Responding to the final claim that it appeared the researchers had selectively excluded data, eliminating certain subjects from the analysis, in order to strengthen their conclusion, Swan said: “The three children, who were excluded because of extremely high phthalate metabolite values, were excluded prior to the play behaviour analysis. This is also included in the paper. Excluding such outliers is standard statistical practice and is necessary to avoid violating assumptions of the model used.”
“Prenatal phthalate exposure and reduced masculine play in boys” by Shanna H Swan et al is published in the International Journal of Andrology 32, 1-9
