Researchers who examined Dow Chemical Company-sponsored animal tests performed two decades ago on the insecticide chlorpyrifos found inaccuracies in what the company reported to the U.S. Environmental Protection Agency compared to what the data showed.
And, according to internal EPA communication, agency scientists also had issues with the study interpretations, yet the agency approved the compound for continued use anyway.
“EPA staff scientists and staff were telling management there were problems,” said Jennifer Sass, senior scientist with the Natural Resources Defense Council, who was not involved in the current study but has worked on issues related to toxics, including chlorpyrifos, for decades.
“And management disregarded it.”
Those 20-year-old industry studies are still used by regulatory agencies such as the EPA and the European Food Safety Authority in approving continued use of the controversial insecticide, which is used on beans, citrus, corn, cotton, wheat and soybeans.
“Exaggerated trust in the reporting” by regulators led to a “failure” of both U.S. and EU authorities to act on red flags, the authors wrote.
The results, published today in the journal Environmental Health, are timely: The EPA is appealing a court decision that would mandate a ban on chlorpyrifos residue on food (which would effectively mean a ban on farm-use); and the European Union is considering a ban as well.
The Obama Administration’s EPA in 2015 proposed a ban of the chemical on food (that would have likely taken effect in early 2017), but President Trump’s former EPA Administrator Scott Pruitt reversed the decision.
In August, however, a three-judge panel on the Ninth Circuit Court of Appeals ordered the EPA to ban the chemical. The EPA has appealed the decision.
Multiple studies since the industry-funded research have shown toxic impacts, especially to children, from chlorpyrifos exposure. Those studies have linked fetal exposure to lower IQ’s and reduced gray matter in the brain later in life. Health researchers have increasingly sounded the alarm that the chemical should be banned due to its potential for impacts on young nervous systems.
The chemical was developed as a nerve gas during World War II.
There’s no surprise it’s toxic “because it was designed from chemical warfare agents,” Sass said.
The new study is a peek behind the curtain at the stark discrepancy between industry and independent science on the chemical.
“If all of this raw data had been scrutinized properly, it should have at least required further testing to see if these findings were abnormal,” Philippe Grandjean, senior author on the study and a researcher and adjunct professor of environmental health at Harvard T.H. Chan School of Public Health, told EHN. “In our minds, their [Dow] data are not appropriate to prove that [chlorpyrifos] is not a neurotoxicant.”
Inaccuracies in the reporting
The researchers, led by Axel Mie, an assistant professor in the department of clinical science and education at the Karolinska Institute, requested the data for two industry lab animal studies – one from 1998, and one in 2015.
One study tested chlorpyrifos exposure on rats, while the other was a rat study of chlorpyrifos-methyl, a breakdown chemical from chlorpyrifos.
- The lab, Argus Research Laboratories in Pennsylvania, used a 2 percent cut off for what constitutes “statistically significant” findings throughout most of the study, instead of the scientific standard of 5 percent. This is important because it is a stricter interpretation of data and would make it more likely that they wouldn’t find impacts from exposure.
- When the lab looked at dimensions of the brain after exposure, they didn’t look at individuals but put them all together and took an average. “When we looked at least one dimension in the rats, cerebellum height was decreased and linked to exposure to chlorpyrifos in newborn pups,” Grandjean said. “In the other test study where they examined chlorpyrifos-methyl those data were in part missing, so we were unable to see if the same thing happened with the sister compound. And there was no explanation for the data being unavailable.”
- The rat studies failed to model human exposure and potential brain impacts. “The brain growth spurt occurs mainly postnatally in rats but prenatally in humans,” Mie and colleague wrote. However, the newborn pups in the industry studies had decreased levels of exposure once born because only a fraction of chlorpyrifos is transferred via milk.
- The test facility for the studies was “unable to detect neurobehavioral effects of elevated developmental exposure to lead nitrate, although lead is a confirmed developmental neurotoxicant at very low doses,” the authors wrote.
“We believe there were some inaccuracies in the reporting and in the summary provided by Dow to the EPA and EFSA,” Grandjean said. “And this goes back something like 20 years, when all of this testing was being done, and this is what current approval of chlorpyrifos relies on.”
“Federal agencies need to stop doing negotiations with registrants”
Grandjean said there were several hundred pages of data.
In communication between EPA toxicologists and those responsible for registering pesticides, it’s clear agency scientists were well aware of study interpretation problems.
“The study was graded unacceptable due to an inadequate presentation of the statistical data analysis,” wrote Susan Makris, formerly with the toxicology branch of the EPA, in a 2000 note to the agency’s reregistration branch.
An EPA spokesperson said the agency is reviewing the new study.
“What happened in the end was EPA management overriding their own science and technical experts,” Sass said.
Sass added that EPA scientists are now on the “right track” – looking at low dose exposures and specific impacts to developing children.
And now it’s up to management and administration officials to follow the science.
“This [study] just shows that industry can’t be trusted on how it reports data, and federal agencies need to stop doing negotiations with registrants,” Sass said.
EHN has reached out to Dow Chemical Company and will update the story when they respond.