A 2017 draft assessment of formaldehyde that was suppressed by the Trump administration found that the chemical causes myeloid leukemia, according to several sources familiar with the document. The draft assessment concludes that 1 microgram of formaldehyde in a cubic meter of air increases the number of myeloid leukemia cases by roughly 3.5 in 100,000 people, more than three times the cancer risk in the assessment now in use. The EPA currently regulates formaldehyde using an outdated set of calculations, finalized in 1991, based on the risk of nasopharyngeal cancer. If the nasopharyngeal cancer and myeloid leukemia risks are combined, the cancer risk could be 4.5 times higher than the current value.
Although the suppressed assessment, produced by a division of the Environmental Protection Agency known as the Integrated Risk Information System, or IRIS, has grave implications for public health, Trump administration officials, including former EPA Administrator Andrew Wheeler, refused to allow the agency to release the assessment, several sources told The Intercept.
If the Biden EPA permits IRIS to finalize the assessment, which is among the most controversial in the agency’s history, other branches of the EPA will likely set regulatory limits based on the new risk value. IRIS values are used to determine safety thresholds for the amounts of the chemical permitted in air and water, emissions caps from industrial facilities, and guidance for cleanups. Such thresholds also alert exposed people to the dangers they face.
“This is a big, big deal. It could trigger significant interest by the personal injury bar,” said Bob Sussman, who served as the senior policy counsel to the EPA administrator during the Obama administration and deputy EPA administrator under President Bill Clinton. “Even without including the leukemia, the risk is a significant one. With the leukemia, it’s certainly a very significant public health concern.”
Even using the much lower, outdated cancer risk number set in 1991, formaldehyde is already the greatest source of nationwide cancer risk from industrial air pollutants, estimated to cause roughly 18 of the 32 cancers in every 1 million people in the U.S. that are caused by toxic pollutants in the air, according to the EPA’s own data. If the risk values are increased by a factor of four or more, the reported cancer risk from formaldehyde will go up accordingly, revealing previously unrecognized cancer hot spots around the country.
Formaldehyde, mainly used to make plywood, particleboard, and glues, is also a hazard for millions of workers who are exposed to the chemical through the production of resins, wood composite and furniture production, plastics manufacturing, paper production, embalming, foundry work, fiberglass production, building construction, agriculture, firefighting, and the teaching of biology, among other occupations. The general public is exposed to formaldehyde through tobacco smoke, cosmetics, food and water, as well as “off-gassing” from consumer goods, construction materials, and furniture. In 2019, 662 facilities reported releasing 47.7 million pounds of formaldehyde.
In addition to looking at the chemical’s relationships to nasopharyngeal cancer and leukemia, the draft assessment also factored in evidence that formaldehyde causes decreased lung function, allergic conditions, reproductive and developmental toxicity, and sensory irritation.
In part because formaldehyde is so pervasive, a wide range of business interests have attempted to prevent the EPA from finalizing the assessment and publicly acknowledging the connection to leukemia. Until now, the EPA has kept the contents of its assessment under tight wraps. The leaked information provides a window into how the science and regulation around the chemical will finally be updated.
“Just knowing that the IRIS assessment acknowledges that formaldehyde causes myeloid leukemia is huge, regardless of what the value is,” Jennifer McPartland, a senior scientist at the Environmental Defense Fund, said after being told about the main findings of the draft assessment. “But the numbers tell you even more. They set the tone for how the agency will proceed with risk management.”
Assessment Was “Ready to Go”
While the EPA has claimed in court that the formaldehyde assessment did not exist in a form that allowed it to be shared — and used that argument to refuse to release it in response to a Freedom of Information Act request from the environmental whistleblower group Public Employees for Environmental Responsibility, or PEER — former EPA officials told The Intercept that the completed assessment not only existed but that it had also been prepared for release in early 2018.
“We had the document ready to go for review within the agency, but we were not allowed to proceed,” said Jennifer Orme-Zavaleta, former director of the EPA Office of Research and Development, who retired from the agency in July. “Political leadership did not want me to move it forward.”
Wheeler, the former EPA administrator, was among the Trump appointees who were involved in suppressing the calculation of the myeloid leukemia risk, according to one former agency official who worked on the assessment. “The administrator would not allow us to put out that number,” the former official said, going on to explain how members of the administrator’s office, particularly his chief of staff Ryan Jackson, kept an unusually close eye on the formaldehyde assessment.
“Normally when we’re doing an assessment, they [the staff of the administrator’s office] don’t have access to the guts of it, because it’s deliberative. But they would demand that we give them pieces of the assessment, including what literature we were looking at for dose responses,” the former official said, referring to the precise relationship between exposure and cancer risk. “Then they would share it with ACC,” an acronym for the American Chemistry Council. Although the former official provided no direct evidence of the administrator’s office sending out the segments of the assessment, they said the actions of the trade group made it clear that it had reviewed the unpublished material.
“All of a sudden, after we gave the information to Ryan Jackson, there would be all this targeted activity by ACC that would basically dismiss or disprove whatever mathematical methodology that we were going to be using even though it hadn’t been published yet,” said the former official. If the EPA did share the information with the trade group, the report would be subject to release under FOIA and would not be exempt under the “deliberative process privilege,” contrary to the agency’s claim in the PEER lawsuit.
Orme-Zavaleta confirmed that Jackson, who now works for the National Mining Association, played a key role in preventing the assessment from moving forward. “The ACC was in contact with Ryan a lot,” she said.
Asked about his role in the formaldehyde assessment, Jackson stated, “Following enactment of the Lautenberg Act, EPA placed formaldehyde though the risk evaluation required in that new law so formaldehyde could be directly regulated by EPA,” referring to the Trump administration’s decision to add formaldehyde to the list of chemicals to be evaluated under the Toxic Substances Control Act. That assessment has also yet to be completed or released. Jackson also stated that “I’m not contributing to anonymous rumors.”
Andrew Wheeler declined to comment. In an emailed statement, a spokesperson for the ACC wrote that the group “has never had access in any manner to unpublished findings from the formaldehyde draft IRIS assessment.”
Orme-Zavaleta also pointed to other Trump appointees who she said were apparently involved in suppression of the document. Among them was the director of the Office of Air and Radiation, Bill Wehrum, whose input Orme-Zavaleta needed to move the assessment forward. “A couple times we set up briefings for Bill Wehrum. But they canceled those briefings, and I was not given the OK to further the review within the agency.”
David Dunlap, who served as deputy assistant administrator in the EPA’s Office of Research and Development, also helped suppress the assessment, said Orme-Zavaleta. “When we were in the process of coming up with the priority list for IRIS assessments, a lot of those discussions were held during ‘politicals-only’ meetings, so I wasn’t there. But [Dunlap] was, he was our representative. Formaldehyde was on the list of chemicals being considered. And in November, I got back a short list of the chemicals IRIS would assess, and formaldehyde was not part of it,” she said. Dunlap worked for Koch Industries and defended formaldehyde until shortly before he joined the Trump administration in 2018. Dunlap did eventually recuse himself from work on the assessment, though only after formaldehyde was removed from the list of chemicals IRIS would be allowed to assess, according to Orme-Zavaleta. “I only received a copy of his recusal letter after I got the list of chemicals.”
David Dunlap declined to comment on the matter.
“EPA lied to us. And they lied to the court.”
Presented with information from Orme-Zavaleta and others about the completion of the assessment in 2017, PEER attorney Kevin Bell expressed anger and frustration. Bell oversees 3-year-old litigation against the EPA over the group’s Freedom of Information Act request for the assessment. In that case, the agency has insisted that it didn’t have a completed version of the document that it could publicly release.
“EPA lied to us. And they lied to the court,” said Bell. “They said, ‘We just can’t give you anything like this, because it would be totally full of all sorts of irrelevant things, so that it would totally undermine the deliberative process,’ whereas in reality they had a fully drafted document that was all set to go.”
The EPA declined to comment on Bell’s statement but did provide a broader response, saying that “EPA is committed to ensuring that science is at the backbone of everything the Agency does to deliver on our public health and environmental protection mission, and that we operate in a transparent manner to ensure public trust.” Bill Wehrum did not respond to requests for comment.
Decades of Industry Influence
The suppression of the most recent draft assessment follows many years of interference with government science on formaldehyde. IRIS began updating its original assessment of the chemical in 1997 but almost immediately ran into resistance from industry, particularly around the link to leukemia. A group called the Formaldehyde Council, Inc., which represented companies that used the chemical, began challenging the numbers in the assessment and came up with a cancer risk that was thousands of times lower than the EPA’s. Political appointees in the George W. Bush administration also helped industry postpone the release of the updated document. A 2008 Government Accountability Office report documented those delays and lamented the public health implications of having to rely on an IRIS assessment that was then 18 years old.
Yet that first assessment, completed three decades ago, is still guiding regulation. In the intervening years, both the International Agency on Cancer Research, or IARC, and National Toxicology Program, whose chemical reviews don’t have the regulatory weight of IRIS assessments, classified formaldehyde as a known human carcinogen and recognized it as a cause of leukemia. In 2010, the EPA released a draft of the updated assessment, which also found that formaldehyde caused leukemia, but did not finalize it. The ACC was harshly critical of that draft, and the agency sent the report to the National Academy of Sciences for peer review. The external agency confirmed its finding that formaldehyde causes cancer and encouraged the agency to quickly rewrite and reissue its report. But the EPA didn’t complete that work until 2016.
Vincent Cogliano, who directed the IRIS program from 2010 to 2017, said the delays during those years weren’t due to political or industry pressure. “We were really trying to get a very objective assessment released,” said Cogliano, who attributed the slow progress on the assessment in that period to the heavy workload of the small program. While Cogliano characterized some of the science around formaldehyde as complex, he said that no one on the IRIS staff disagreed about the relationship between myeloid leukemia and formaldehyde. “When I left the IRIS program in January 2017, I thought that it was ready to be released,” he said.
And yet more than four years later, the EPA has yet to release the updated IRIS assessment.
“It’s very frustrating,” said Linda Birnbaum, who directed the National Toxicology Program from 2009 to 2019. “Here there is evidence that IARC said that formaldehyde was a known human carcinogen. And we said it was a known human carcinogen and that the data was clear. And more data has come out since then, especially from the National Cancer Institute, yet no word from EPA.”
Formaldehyde was added back on to the list of chemicals being assessed on the IRIS public calendar in March. According to an EPA spokesperson, the IRIS Program anticipates releasing the draft assessment of formaldehyde for public comment and external peer review by September 30, 2022. The agency is widely expected to ask the National Academy of Sciences to weigh in on how and whether the agency should combine the cancer risk estimates for myeloid leukemia and nasopharyngeal cancer.
Orme-Zavaleta said that no one in the Biden EPA expressed hesitation about resuming work on it when she raised the issue in the first weeks of President Joe Biden’s tenure. “I asked them several times. I said, ‘I’m going to unsuspend formaldehyde, are you good with that?’ And every time, everybody said, ‘Fine, do it,’” she said. “It is moving!”
Despite the assurances, those intimately familiar with the history of formaldehyde remain guarded in their optimism that the EPA is finally about to update the assessment of one of the most significant and widespread carcinogens in the U.S.
“For decades, up to now, the industry has pretty much managed to put out all the fires,” said Sussman. “But perhaps something really significant might be about to happen.”
Update: August 19, 2021
This article has been updated with a comment from the American Chemistry Council.