3 ms·
By Christopher J Portier, the same Christopher Portier that served as a special advisor for IARC while at the same time secretly accepting money from law firms
by coin 4y ago
By Christopher J Portier, the same Christopher Portier that served as a special advisor for IARC while at the same time secretly accepting money from law firms involved in glyphosate litigation. The IARC study was edited to remove evidence to the contrary to its findings.
https://www.judicialhellholes.org/2019/05/14/california-courts-drop-science-at-the-door/ https://www.judicialhellholes.org/2019/05/14/california-cour...
"Closer scrutiny of the IARC process reveals that it was advised by an “invited specialist,” Christopher Portier, in its work on glyphosate. At the same time Mr. Portier was working for the agency, he was being paid by the Environmental Defense Fund, an anti-pesticide group. Moreover, Mr. Portier received $160,000 from law firms suing over glyphosate. When asked about this potential conflict of interest, Mr. Portier initially claimed to be advising firms on other IARC-related lawsuits and not glyphosate litigation. He later acknowledged that his statement was wrong. It is also worth noting that Mr. Portier had no experience with glyphosate prior to his work on it for IARC.
Following Mr. Portier’s arrival at IARC, the final glyphosate study was altered in at least 10 ways to remove or reverse conclusions finding no evidence of carcinogenicity. The agency removed multiple scientists’ conclusions that studies found no link between glyphosate and cancer in lab animals and statistical analyses of studies with negative findings were turned into positive ones. The determination that glyphosate was “probably carcinogenic” was based on “limited evidence” of carcinogenicity in humans and “sufficient evidence” in experimental animals."
- Retric 4y agoNegative results on cancer studies don’t mean there isn’t a cancer link. They at best suggest a limitation on how dangerous a substance may be. Aka plenty of 2 pay a day smokers never get lung cancer. The usual approach is to have a reasonably large number of animals and increase their exposure, but that only works if the risks are linear and the animal model maps well to humans.