Synergistic endocrine disrupters are scheduled to return from the dead in February 1999. That's when the journal Environmental Health Perspectives plans to publish a new study "suggesting" (again) manmade chemicals may act synergistically to disrupt hormone systems.
The new study is very weak -- and that's only if it hasn't been cooked. But first, a little background...
When last we left synergism in July 1997, Tulane University researchers John McLachlan and Steven Arnold had just retracted their June 1996 study from Science.
McLachan and Arnold had claimed that combinations of PCBs and pesticides were potent endocrine disrupters -- up to 1,500 times more potent than the individual chemicals.
Media-, EPA- and enviro-hype surrounding the report intimidated the environmentally-gunshy 104th Congress to require the EPA to develop a regulatory program for screening and testing chemicals for their potential to disrupt hormone systems.
But soon after, independent laboratories from around the world began to report that the Tulane-PCB study could not be replicated -- such replication being required for the study results to be taken seriously by scientists, if not Congress!
By July 1997, the study had become such an embarrassment that McLachlan and Arnold were forced to retract it from publication. Of course, the new law remained, though without scientific underpinning.
And so now we have the new study. If anyone in Congress was hoping that its legislative faux pas would eventually be vindicated, this study won't do the trick.
First, even accepting the study results at face value -- which I don't -- the study did not involve PCBs, pesticides or other alleged endocrine disrupters. I must wonder why, since it would have been just as easy to run this laboratory experiment with manmade chemicals as natural ones. While the study authors only "suggest" that environmental estrogens may synergize with stronger natural estrogens, you can bet the media, EPA and enviros will gloss over this distinction.
The study involves turtle embryos exposed to combinations of weak and potent natural estrogens. Apparently, sex determination in this species of turtles depends on temperature. Incubation below 28.6 degrees Centigrade produces males. Incubation above that temperature produces increasingly more females, until at 29.6 degrees Centigrade, only females are produced. Reportedly, prior experiments have shown that combinations of natural estrogens and temperature synergize to produces more females than expected if the effects of the estrogens and temperature were additive.
Reportedly in the study at hand, turtle embryos were exposed to the natural estrogens estrone, estradiol and estriol individually and in combination at male-producing temperatures. Table 1 from the study shows the 30 trials conducted.
Of the 15 trials involving combinations of these estrogens, only two trials (nos. 22 and 26 in Table 1) reportedly produced significantly more females than expected if the effects of the estrogens were additive. While 2 of 15 trials is hardly a strong result, it appears that some hanky-panky occurred with even these two trials.
Table 2 from the study shows that, if the effects of the natural estrogens were additive, the number of expected females in Trial 22 would be 11.5. This expectation is compared against the observed 18 females to get a statistically significant difference.
But from Table 1, their own experiment shows that the expected number of females and males should be 14 (4 from estriol 0.01 and 10 from estradiol 0.20), not 11.5. I'd bet that the significance of the reported result evaportates if the correct comparison is made.
A similar error is made for Trial 26. According to Table 1, the expected number of females should be 18 (14 from estrone 0.25 and 4 from estriol 0.01). But the Table 2 comparison is made with only 17.
While the Trial 26 error may be less significant, it could be this study really only reported a statistically significant difference in one trial -- hardly something to write home about.
I'm not surprised that some fudging is involved. In the study, the authors thank McLachlan and Arnold "for discussions on the research design and interpretation of the results." But McLachlan and Arnold were the ones compelled to withdraw their own bogus claim of endocrine disrupter synergy. Do they get the credit for this questionable claim, too?
The study authors state in the Introduction of their paper "Although the pesticides endosulfan and dieldrin were reported to produce similar synergy using a yeast gene expression system (footnote 5, omitted), subsequent studies failed to replicate this result (footnotes 6&7, omitted)."
Footnote 5 is the McLachlan/Arnold study! The study was retracted. For all intents and purpose, it no longer exists. It should not be cited. There is no "result." And while that may be a bulletin to Congress, it should not be for the editors of Environmental Health Perspectives.
One final question: When may expect this study to be retracted?
Comments on this posting?
Click here to post a public comment on the Trash Talk Bulletin Board.
Click here to send a private comment to the Junkman.
Copyright © 1998 Steven J. Milloy. All rights reserved on original material. Material copyrighted by others is used either with permission or under a claim of "fair use." Site developed and hosted by WestLake Solutions, Inc.
Material presented on this home page constitutes opinion of Steven J. Milloy.
