Where have we seen this before? Oh yes, frequently: numerous activist groups and academics seeking publicity and/or NIH/NIEHS grants feel the urge, every so often, to take a shot at our most disfavored class of chemicals: phthalates. Here we go again.
A group based at the Columbia/Mailman School of Public Health was fortunate enough to receive one of those into the future NIH/NIEHS mega-grants to follow a cohort of pregnant women and their progeny around for two decades or longer and try to correlate various environmental influences with health-related outcomes. They have published intermittent reports of a similar nature, generally addressing data-dredged, statistically-unsound links between certain chemicals, especially phthalates, and ostensibly-measurable outcomes.
This poor excuse for a study which the authors titled Persistent Associations between Maternal Prenatal Exposure to Phthalates on Child IQ at Age 7 Years, posited an association between 6-7 IQ points reduction among the progeny of moms whose one-time, spot urinary levels of two of the several tested phthalates were in the top quartile, as compared to moms progeny from the lowest urinary phthalate quartile. (Note: the authors title evinced considerable restraint compared to the Mailman press release s breathless Exposure During Pregnancy to Common Household Chemicals Associated with Substantial Drop in Child IQ, not to mention several of the even more excited media headers).
There is but one thing wrong with the media maelstrom, abetted eagerly by the PLOS-One journal and the Mailman School: it is entirely based on fantasy, not science. Or perhaps we should say, intentionally wishful thinking of various folks fevered imaginations. What s wrong with it? Is there any actual substance to their analyses and statistics? It is impossible to elucidate, given all the assumptions and controlled for s in their discussion.
But most importantly, this being an observational study, no one with any semblance of responsible scientific ethics would even hint at a cause-and-effect relationship between any of these chemicals and reduced kids IQs. But that should have gone without saying. Here s the real sleazy undercurrent of statistical chicanery, incomprehensible pseudo-mathematical gobbledegook which confounded even responsible journalists, descending into because I said so:
a The authors allegedly controlled for the home environment ¦and other factors known to influence child intellectual development. Really? Those factors are clearly impossible to control for, especially in a low SES group;
b Maternal IQ controlled for uh, was there any other DNA involved in these kids IQs or did they spring from moms ear? OK, just to be clear: no one even expressed any concern for the complete absence of controlling for paternal factors, including IQ;
c Most of the headlines are warning about cosmetic chemicals, yet, the only phthalate found in significant usage in cosmetics DEP, diethyl phthalate showed zero correlation with the study s outcomes (IQs);
d Phthalates are excreted with a half-life of about 4 hours; thus, measuring their levels at one point in pregnancy hardly reflects exposure in any meaningful way.
ACSH s Dr. Gil Ross had this comment: This group has to bend over backwards and then twist in all directions to dredge up statistically significant findings based on their slovenly, not to say unethical and unscientific, methodologies and phony statistics. Studying so many variables at once is bound to come up with several such quasi-associations. Then, they let nature takes its course, as the press release will generate a horde of media slipshod alarms. But they got the job done: they targeted a disfavored chemical class and justified their vast funding from government agencies with deep pockets to carry on their own anti-chemical agendas. Mixed metaphor alert: it s one hand washes the other, at the Big Government feeding trough.