how much data fiddling was required to find a result..
Washington, D.C. (PRWEB) July 25, 2007
A great deal of media and activist attention has been directed recently at so- called "human studies" regarding phthalates. They are a major part of the information cited by phthalate critics to support their calls for bans or restrictions on phthalates. Marian K. Stanley, manager for Phthalate Esters Panel, a group of the American Chemistry Council, takes a look at the leading studies to clarify the matter.
"Scientists do not dose humans with chemicals as if they were rodents in a toxicology lab. Instead, they typically look in the urine for the presence of a particular chemical, and attempt to relate that presence to a health effect," explains Stanley. "Even the best of this kind of research cannot prove that the chemical caused the effect; the most these studies can do is to suggest a link between the effect and the chemical exposure that should be pursued in follow-up studies," Stanley adds.
Stanley concludes that it would help if these studies looked to see if people without the effect - a control group - have the same levels of exposure, indicating that the effect is from some other cause. Regrettably, most do not examine control groups.
Here are the leading studies that claim to show a link between phthalate exposure and some biological effect, listed by key author:
SHANNA SWAN - a statistician, Dr. Swan claimed to show a link between maternal exposure to some phthalates and subtle changes in the reproductive systems of their male infants. She was careful to state that no actual damage - no health effect -- was detected, just changes in penis size and ano-genital distance (called AGD). Key problems: the report has been heavily criticized. The attempt to convert AGD into a kind of index for phthalate effects has no standing in human biology; Swan had to combine the levels of four different phthalates to show a statistically significant effect. One critique by a biostatistician and mathematician wondered "how much data fiddling was required to find a result.." An EPA panel reviewing one of the cited phthalates to set safe exposure levels looked at the Swan study but ruled that it should not influence the panel's decision to reduce the estimated risk level by 300 percent.
SUSAN DUTY, RUSS HAUSER - when Dr. Duty was a post-doctoral fellow at Harvard in Dr. Hauser's lab, she and her mentor published papers claiming to show a correlation between phthalate exposure and sperm damage. The authors conceded that studies were preliminary, and that the study sample was small. Key problems: Critics pointed out three serious failings: One, her male subjects were subfertile, drawn from a fertility clinic. Would current phthalate levels in their bodies have any bearing on fertility problems that already existed? Two, there was no control group of normal men, which is standard procedure in all such tests. Three, the primary phthalate fingered by one study was DEP, which is biologically implausible because DEP does not show the same toxicological effects in rodents that some other phthalates do. See Response to Study of Phthalates and Sperm Quality on Phthalates.org
GIUSEPPE LATINI - his team published a paper claiming a link between phthalate exposure and the birth of newborns one week early (which is pre-term but not premature). Key problems: why would the presence of DEHP right before delivery reveal anything about the duration and nature of the pregnancy? Second, the researchers tested for the intact molecule. But intact DEHP breaks down within 30 minutes of entering the body. That means the exposure was very recent, and most likely from medical procedures involving vinyl tubing that actually helped guarantee the safe delivery of a healthy baby.
I. COLON - reported higher levels of DEHP in blood from 41 girls in Puerto Rico experiencing premature breast development. Key problems: first, ample toxicological evidence shows phthalates do not act like estrogens that stimulate breast development. Second, the study reported high levels of the intact DEHP molecule and low levels of its first metabolite. That makes no sense, because DEHP breaks down very rapidly in the body. The most likely explanation: contamination from lab equipment containing DEHP - a very common problem in early research on phthalates. Find comments on this study on Phthalates.org.
KATHARINA MAIN - searched for evidence of a correlation between cryptorchidism in infants (undescended testicles) and levels of phthalates in their mothers' breast milk. Key problems: the study failed to find the effect it was looking for. The paper was nevertheless published because it did claim a correlation between some levels of hormone production in the infants and phthalate levels in the breast milk, but the changes are of no known significance. (Any substance encountered by a mother will find its way into all body fluids at some level, including breast milk.)