Companies with a financial interest in a weed-killer sometimes found in drinking water paid for thousands of studies federal regulators are using to assess the herbicide's health risks, records of the U.S. Environmental Protection Agency show. Many of these industry-funded studies, which largely support atrazine's safety, have never been published or subjected to an independent scientific peer review.
Meanwhile, some independent studies documenting potentially harmful effects on animals and humans are not included in the body of research the EPA deems relevant to its safety review, the Huffington Post Investigative Fund has found. These studies include many that have been published in respected scientific journals.
Even so, the EPA says that it would be "very difficult for someone to put a thumb on the scale" to slant the outcome.
Atrazine is one of the most widely used herbicides in the U.S. An estimated 76 million pounds of the chemical are sprayed on corn and other fields in the U.S. each year, sometimes ending up in rivers, streams, and drinking water supplies. It has been the focus of intense scientific debate over its potential to cause cancer, birth defects, and hormonal and reproductive problems. As the Huffington Post Investigative Fund reported in a series of articles last fall, the EPA failed to warn the public that the weed-killer had been found at levels above federal safety limits in drinking water in at least four states. Some water utilities are suing Syngenta to have it pay their costs of filtering the chemical.
Now the EPA is re-evaluating the health risks of atrazine, which was banned in the European Union in 2004 due to a lack of evidence to support its safe use. That ban includes Switzerland, where atrazine's manufacturer, Syngenta, is headquartered. The EPA expects to announce results of its re-examination of the herbicide in September 2010. It could take action ranging from restrictions on its use on crops to an outright ban. Or it could permit continued use without additional restrictions.
The company, one of the world's largest agribusinesses, says the chemical has been used safely for decades and restrictions could prove devastating to farmers who are heavily dependent on the inexpensive herbicide. Atrazine poses "no harm" to the general population or to drinking water supplies, said company spokesman Steven Goldsmith.
EPA records obtained by The Huffington Post Investigative Fund show that at least half of the 6,611 studies the agency is reviewing to help make its decision were conducted by scientists and organizations with a financial stake in atrazine, including Syngenta or its affiliated companies and research contractors.
More than 80 percent of studies on which the EPA are relying have never been published. This means that they have not undergone rigorous "peer review" by independent scientists, a customary method to ensure studies are credible and scientifically sound before they can be published in major journals.
At the same time several prominent studies by independent academic scientists in well-respected scientific journals - showing negative reproductive effects of atrazine in animals and humans - are absent from the EPA's list.
That finding may raise concerns about how the agency is doing its work. Rep. Henry Waxman, chairman of the House Energy and Commerce Committee, which oversees environmental regulators, told the Investigative Fund, "it's critically important that EPA use all of the information at its disposal."
Agency scientists may review studies not on the list, but EPA senior policy analyst William Jordan said that the 6,611 studies are those considered "relevant to the assessment of atrazine."
'Not Just Atrazine'
EPA spokeswoman Betsaida Alcantara said the list was not exhaustive and that some studies may not be on the list because they were not given an eight-digit "master record identification number," which the agency uses to keep track of studies. There is "no uniform practice" for assigning numbers to studies submitted by people other than those working for herbicide, fungicide or pesticide manufacturers, she added.
EPA officials said that with a limited budget the agency must rely heavily on research sponsored by parties with a stake in the outcome. The agency's "test guidelines" governing how experiments are conducted - the types and number of lab animals to be used, for instance. These provide sufficient safeguards against skewed results, officials said.
"Companies have a very strong incentive to follow the guidelines," said EPA senior analyst Jordan. "We hope and think that we have written the guidelines with enough detail that it would be very difficult for someone to put a thumb on the scale, as it were, to slant the outcome, [or] to make something look safer than it is."
Jennifer Sass, a senior scientist specializing in health issues at the Natural Resources Defense Council, argues that relying on a company to test the safety of its own product - an "inherent conflict" of interest - is part of a larger pattern at the EPA. "It's not just happening with atrazine," she said.
Hundreds of herbicides, pesticides, and other chemicals are regulated by the EPA, whose decisions can have significant implications for public health and on the abilities of an array of multinational companies to earn billions of dollars in the U.S.
By law, industry influence often is built into the regulatory process of the federal government. At the Food and Drug Administration, for instance, clinical trials conducted by pharmaceutical companies are used to determine whether pills and devices work and are safe. Makers of pesticides, herbicides, and fungicides also must pay for studies on their products. If they meet agency rules for conducting the testing, the EPA must accept them.
The 'Funding Effect'
But is industry-funded research always reliable? A pair of scientists funded by the National Science Foundation, the U.S. Department of Agriculture, and the EPA scrutinized a Syngenta-funded Canadian study - one that is not on the EPA's list. The scientists said they found numerous inaccuracies and misleading statements.
The scientists who questioned the study, University of South Florida biologists Jason Rohr and Krista McCoy, published their critique in the March 2010 issue of the journal Conservation Letters. In all, they tallied what they said were 122 inaccurate and 22 misleading statements, of which 96.5 percent appeared to support atrazine's safety. The widely cited study focused on the herbicide's effects on fish and other aquatic creatures.
Rohr and McCoy also asserted that the Canadian study, which was done in 2008, misrepresented more than 50 other studies. For example, it incorrectly suggested that only one scientist had demonstrated the chemical's gender-altering effects on frogs. In fact, several other scientists demonstrated such effects.
The study dismissed one of Rohr's papers as invalid, noting wrongly that the researcher had filtered atrazine out of a water tank while trying to assess the chemical's effect on the aquatic organisms in the tank.
The Canadian study also misrepresented results, figures, and conclusions of other studies, according to the University of South Florida biologists.
Rohr, who served on an EPA advisory panel examining atrazine last year, told the Investigative Fund that he felt compelled "to set the record straight given the potential policy and environmental implications of these misconceptions and inaccuracies."
The author of the Canadian study, University of Guelph (Ontario) biologist Keith Solomon, declined to respond to questions from the Investigative Fund about his financial ties to Syngenta, the company's influence, or the inaccuracies and mischaracterizations the South Florida biologists said they had uncovered. Solomon noted that other scientists had come to similar conclusions, and that governments in the U.S. and Australia had not found any significant risk to creatures living in water.
While the critiqued study is not on the EPA's list, several other studies by Solomon are.
Wendy Wagner, an expert in environmental policy at the University of Texas law school, said that the criticism of the Canadian study demonstrates a phenomenon sometimes referred to as "the funding effect."
"It is next to impossible to squeeze all of the discretion out of a researcher, and when he has a strong incentive to find a particular result, the result can be unreliable and badly biased research," said Wagner, an authority on the influence of politics and special interests on science. "There is compelling evidence that bias still pervades sponsored pesticide research - research that presumably is done in accord with EPA's guidelines."
Meanwhile, some independently-funded academic research published in major scientific journals is missing from the list of papers the EPA is using to make its decisions on atrazine. Absent are studies published in Proceedings of the National Academy of Sciences, Environmental Health Perspectives, and Nature. Many works by independent academic scientists such as Tyrone Hayes and Rohr - who have demonstrated a range of potential reproductive and hormonal effects of the chemical - are not on the list.
Some peer-reviewed studies from prestigious journals fail to meet the agency's standards, said EPA analyst Jordan, citing as an example work by scientists such as Hayes, who recently found that low doses of atrazine could turn male frogs into female frogs.
Jordan explained that the agency couldn't rely on Hayes' and the other scientists' research in part because the government lacked protocols for testing chemicals on frogs. So the EPA developed those guidelines and asked Syngenta to study the issue. The company's researchers reported that they were unable to replicate Hayes' findings. Jordan said the Syngenta study "superceded" Hayes' and the other scientists' studies. The EPA, on its Website, currently states that atrazine causes no such adverse effects on frogs and that "no additional testing is warranted" to address the issue.
Environmental groups have in the past criticized the EPA for allowing chemical companies to wield disproportionate influence over regulatory decisions. While evaluating the safety of atrazine in 2003, the EPA allowed representatives from Syngenta to participate in closed-door negotiations with the agency, according to documents obtained by the NRDC in 2004.
Dale Kemery, an EPA spokesman, defended the practice of omitting some studies. The agency's safety "review may not include every study that has been conducted, since some may not meet the standards that are appropriate for a regulatory setting or they may not be on target for the issues to be assessed."
The EPA considers industry-sponsored studies "scientifically more robust than are the studies generated by people in academia," said Jordan, the agency's senior policy analyst. "That's generally because companies spend more money on their studies and can attend to details that are potentially important that people in academia just can't afford to do."
Jordan added that agency oversight of the thousands of unpublished studies on the list is just as rigorous as a peer-review by scientists prior to publication in a scientific journal. "I know that people might not agree with this proposition, but I believe that the scientists at EPA constitute a peer-review," he said. "Our scientists go over the studies with a fine tooth comb."
EPA officials said they were not able to provide a list of all omitted research.
A spokeswoman for CropLife America, the Washington D.C.-based trade association that represents pesticide and herbicide manufacturers, said EPA oversight is thorough, regardless of whether studies have appeared in peer-reviewed journals.
"Whether or not they have been published, the studies submitted to EPA for registration support of pesticide products are subject to scientific review by EPA scientists that is equally, if not more, rigorous and demanding than the pre-publication peer review conducted by any scientific journal," said spokeswoman Mary Emma Young.
Some people are skeptical about the rigor of the EPA's scrutiny. "What worries me," said the University of Texas' Wagner, "is the possibility that there isn't time or energy within EPA to give a lot of oversight to this unpublished, industry-funded research, especially when the number of unpublished studies for a chemical like atrazine are in the thousands."
A former EPA official, epidemiologist Lynn Goldman, said it is normal and necessary for the agency to accept unpublished and industry-funded studies, most of which would not be interesting enough to publish in scientific journals.
"This is the way that the system was built by Congress. It could be changed but the EPA does not have the authority to turn the system upside down," said Goldman, a former assistant administrator for toxic substances during the Clinton administration.
The existence of a list of relevant research for EPA review has played a prominent role in public arguments for the herbicide's safety. Journalists, scientists, and advocates for atrazine have frequently cited the "6,000" studies.
In 2005, Anne Lindsay, then a top official in the EPA's Office of Pesticide Programs, brought up the number of studies during congressional testimony. "Atrazine is one of the most well-examined pesticides in the marketplace," she said, noting that "there are nearly 6,000 studies in EPA files on the human health and environmental effect of atrazine."
Syngenta now cites the number in its press materials and on its website - not merely as a tally of studies but as proof of its safety. "Atrazine passes the most stringent, up-to-date safety requirements in the world," said spokesman Paul Minehart. "In 2006, the U.S. Environmental Protection Agency (EPA) re-registered atrazine in 2006 based on the overwhelming evidence of safety from nearly 6,000 studies."
- Follow the Huffington Post Investigative Fund on Twitter or fan us on Facebook.
- Do you have information about this story? Send us a tip or submit a correction.
- REPUBLISH THIS STORY FOR FREE: The Huffington Post Investigative Fund licenses its content through Creative Commons. We encourage you to republish our stories in full with proper attribution.