Secret files suggest chemical giant feared weedkiller’s link to Parkinson’s disease

Documents seen by Guardian detail effort to refute scientific research into paraquat and derail nomination of key EPA adviser

by Carey Gillam and Aliya Uteuova

For decades, the Swiss chemical giant Syngenta has manufactured and marketed a widely used weedkilling chemical called paraquat, and for much of that time the company has been dealing with external concerns that long-term exposure to the chemical may be a cause of the incurable brain ailment known as Parkinson’s disease.

Syngenta has repeatedly told customers and regulators that scientific research does not prove a connection between its weedkiller and the disease, insisting that the chemical does not readily cross the blood-brain barrier, and does not affect brain cells in ways that cause Parkinson’s.

But a cache of internal corporate documents dating back to the 1950s reviewed by the Guardian suggests that the public narrative put forward by Syngenta and the corporate entities that preceded it has at times contradicted the company’s own research and knowledge.

And though the documents reviewed do not show that Syngenta’s scientists and executives accepted and believed that paraquat can cause Parkinson’s, they do show a corporate focus on strategies to protect product sales, refute external scientific research and influence regulators.

In one defensive tactic, the documents indicate that the company worked behind the scenes to try to keep a highly regarded scientist from sitting on an advisory panel for the US Environmental Protection Agency (EPA). The agency is the chief US regulator for paraquat and other pesticides. Company officials wanted to make sure the efforts could not be traced back to Syngenta, the documents show.

And the documents show that insiders feared they could face legal liability for long-term, chronic effects of paraquat as long ago as 1975. One company scientist called the situation “a quite terrible problem” for which “some plan could be made”.

That prediction of legal consequences has come to pass. Thousands of people who allege they developed Parkinson’s because of long-term chronic effects of paraquat exposure are now suing Syngenta. Along with Syngenta, they are also suing Chevron USA, the successor to a company that distributed paraquat in the US until 1986. Both companies deny any liability and maintain that scientific evidence does not support a causal link between paraquat and Parkinson’s disease.

“Recent thorough reviews performed by the most advanced and science-based regulatory authorities, including the United States and Australia, continue to support the view that paraquat is safe,” Syngenta said in a statement to the Guardian.

During the years Chevron USA’s predecessor sold paraquat, “it regularly reviewed and considered scientific studies regarding the safety of its products, including paraquat,” Chevron USA said in a statement to the Guardian, adding that none of the studies reviewed “showed a causal link between paraquat and Parkinson’s disease”.

Chevron USA said the company “does not believe that [its former subsidiary that sold paraquat] had any role in causing the plaintiffs’ illnesses and will vigorously defend against the allegations in the lawsuits”.

As part of a court-ordered disclosure in the litigation, the companies provided plaintiffs’ lawyers with decades of internal records, including hand-written and typed memos, internal presentations and emails to and from scientists, lawyers and company officials around the world. And though the files have not yet been made public through the court system, the Guardian has reviewed hundreds of pages of these documents in a reporting collaboration with the New Lede.

Among the revelations from the documents: scientists with the Syngenta predecessor Imperial Chemical Industries (ICI) and Chevron Chemical were aware in the 1960s and 70s of mounting evidence showing paraquat could accumulate in the human brain.

When Syngenta’s own internal research showed adverse effects of paraquat on brain tissue, the company withheld that information from regulators while downplaying the validity of similar findings being reported by independent scientists.

In addition, the records show company scientists were aware of evidence that exposure to paraquat could impair the central nervous system (CNS), triggering tremors and other symptoms in experimental animals similar to those suffered by people with Parkinson’s. A 1975 Chevron communication speaks of concerns about allegations of “permanent CNS effects from paraquat”.

And as independent researchers continued to find more and more evidence that paraquat may cause Parkinson’s, the documents describe what Syngenta called an “influencing” strategy “that proactively diffuses [sic] the potential threats that we face” and seeks to “maintain and safeguard paraquat registrations”, referring to their regulatory approvals. The strategy “must consider how best to influence academia, and regulatory and NGO environments”.

A Syngenta “regulatory strategy” document from 2003 refers to paraquat as a “‘blockbuster’ product” that must be “vigorously” defended to protect more than $400m in projected annual global sales. Ensuring what Syngenta called its “freedom to sell” paraquat was a top priority, the internal records show.

Syngenta also created a website the company used to publicly dismiss concerns about links between paraquat and Parkinson’s disease and provide positive product messaging. On that website, the company asserted that paraquat did not readily cross the blood-brain barrier, even when the company had evidence from animal and human data that paraquat accumulated in brain tissue. The company no longer uses that language on its website.

“It is highly unethical for a company not to reveal data they have that could indicate that their product is more toxic than had been believed,” said Bruce Blumberg, professor of developmental and cell biology at the University of California, Irvine, speaking generally about corporate conduct. “[These companies are] trying to maximize profits and they jeopardize public health, and it shouldn’t be allowed. That is the scandal.”

‘A unique herbicide’

Paraquat is one of the most widely used weedkilling chemicals in the world, competing with herbicides such as glyphosate, the active ingredient in Monsanto’s Roundup brand for use in agriculture. Farmers use it to control weeds before planting their crops and to dry out crops for harvest. In the United States, the chemical is used in orchards, wheat fields, pastures where livestock graze, cotton fields and elsewhere. As weeds have become more resistant to glyphosate, paraquat popularity has surged.

It is used on approximately 15m acres of US farmland. US government data shows that the amount of paraquat used in the United States has more than tripled between 1992 and 2018.

Estimated US agricultural use of paraquat

On the Syngenta-run Paraquat Information Center website, the chemical is described as “a unique herbicide” that “can deliver safe, effective weed control, generating social and economic benefits, while protecting the land for future generations”.

Paraquat has been the subject of more than 1,200 safety studies submitted to, and reviewed by, regulatory authorities around the world, according to Syngenta.

Though it is widely used, paraquat has long been known to be dangerous to ingest – a tiny swallow of the chemical can kill a person within days. Scores of people around the world have died from ingesting paraquat either intentionally or accidentally. The EPA restricts use only to people certified to apply it. It is not sold to consumers, and paraquat warning labels carry the symbol of death – a skull and crossbones.

Syngenta maintains on its website that if users follow directions and wear proper protective clothing, including gloves and boots, “there is no risk to human safety”. Paraquat is “not a neurotoxicity hazard” and “does not cause Parkinson’s disease”, the company states.

Despite the company’s claims, dozens of countries have banned paraquat, both because of the acute dangers and mounting evidence of links to health risks such as Parkinson’s from chronic, long-term exposure. Syngenta currently sells paraquat products in more than two dozen countries, from Australia to Uruguay.

Paraquat was banned in the European Union in 2007 after a court found that regulators did not thoroughly assess safety concerns, including scientific evidence connecting Parkinson’s to paraquat. It is also banned in the UK, although it is manufactured there. The chemical was banned in Switzerland, Syngenta’s home country, in 1989. And it is banned in China, the home base for ChemChina, which purchased Syngenta five years ago.

In the US, the EPA has largely agreed with Syngenta and other chemical companies that say paraquat can be safely used. Last year, the EPA said it would continue to allow farmers to use paraquat, including spraying it across fields from small airplanes.

A ‘Parkinson’s pandemic’

Concerns about possible ties between paraquat and Parkinson’s disease have grown as the spread of Parkinson’s has accelerated; the disease is now considered one of the world’s fastest-growing neurological disorders. Prevalence of Parkinson’s more than doubled from 1990 to 2015 and is expected to continue to expand rapidly, impacting millions of people around the world. Along with paraquat, toxins in air pollution and other pesticides, and to a smaller extent genetic factors, also are considered by many researchers as risk factors for the disease.

Roughly 60,000 Americans are diagnosed each year with Parkinson’s, and in recent years it has been ranked among the top 15 causes of death in the United States, according to the Centers for Disease Control and Prevention. Moreover, the death rate from Parkinson’s has climbed more than 60% in the United States over the past two decades, according to research published last year. It is considered the fastest-growing neurological disease in the world.

As a disease of the central nervous system, common Parkinson’s symptoms include tremors, or a rhythmic shaking in arms and legs, stiffness and rigidity of the muscles, a loss of balance and coordination, and difficulty speaking. Parkinson’s symptoms develop when dopamine-producing neurons in a specific area of the brain called the substantia nigra are lost or otherwise degenerate. Without sufficient dopamine production, the brain is not capable of transmitting signals between cells to control movement and balance.

“The Parkinson’s pandemic has exacted an enormous toll on tens of millions of individuals who bear the brunt of the disease,” Ray Dorsey, a neurologist at the University of Rochester Center for Health + Technology in New York, wrote in a 2020 book about the rise of the disease.

Dorsey is one of a number of leading scientists from around the world who say research clearly shows paraquat exposure can cause Parkinson’s disease.

“Paraquat is considered the most toxic herbicide ever created,” Dorsey said in an interview.

Syngenta said the weight of evidence actually shows that paraquat does not cause Parkinson’s and said a 2021 study co-authored by its chief medical office backs that position. The company also pointed to a 2020 update to the US Agricultural Health Study (AHS) as supporting its position. (The 2020 AHS looked at a much larger group of people than prior AHS research has linked paraquat to Parkinson’s, however.)

“There is no properly designed epidemiological study that shows a link between paraquat and Parkinson’s disease,” the company said in a statement.

“To this day, and despite hundreds of studies being conducted in the past 20 or so years, a causal link between Paraquat and Parkinson’s disease has not been established,” Chevron USA said in a statement to the Guardian.

Toxic timeline

Syngenta predecessor ICI first recognized paraquat’s value as a herbicide in 1955, launching its paraquat brand Gramoxone in the UK in 1962 and then in the United States shortly after.

Even as the company was bringing paraquat to the market, its scientists were starting to see early signs of possible problems with the product. Internal records show that in 1958, an ICI researcher reported to a colleague that company tests on laboratory animals found exposure to a chemical compound related to paraquat appeared to affect the central nervous system.

A 1964 ICI study on rabbits noted dermal exposure to paraquat caused symptoms such as “weakness and incoordination” in some of the animals receiving very high doses. In 1966, ICI scientists studying paraquat exposure effects on a variety of animals noted that large doses given to rats and mice showed effects on the central nervous system, with various impacts, including some animals displaying “hyper-excitability”, a stiff gait or tremors.

In 1968, paraquat poisoning deaths were starting to mount around the world, as many people intentionally used the herbicide as a tool for suicide. With the deaths, according to the documents, came multiple autopsies and analyses revealing that paraquat was accumulating in brain tissue in people who had ingested small amounts of paraquat.

In the early 1970s, animal studies by ICI researchers found more evidence of the chemical’s ability to move into the brain, as well as the lungs, and spinal cord. Field workers exposed to the chemical were complaining of health problems, and the documents indicate that by 1974 some state regulators were expressing concerns about the potential long-term, chronic effects on workers who might inadvertently lick small quantities of paraquat residue off lips, or inhale paraquat mist. Company officials were also warned of rumors that some people inside the EPA were in favor of banning paraquat.

In response, Chevron executives decided the labeling on Gramoxone needed stronger warning language, including advising users to wear goggles and a respirator when spraying. Notes from a February 1974 meeting referred to the “paraquat toxicological problems in the USA” and “increasing numbers of reports of toxicological effects of paraquat to applicators in the field”.

ICI expressed concern about market “repercussions” outside the US from added warnings, but agreed to the changes, according to the meeting notes.

Notes from a follow-up meeting a month later quoted a Chevron lawyer as saying “to a lawyer there is evidence now that paraquat could cause industrial injury and it should be recognized that Chevron could face suits totalling millions of dollars”.

A year later, Chevron fears were growing. In a July 1975 letter to ICI, a Chevron toxicologist noted “problems of nosebleed and sore throat in our own plant workers”, as well as studies indicating the potential for central nervous system effects from paraquat. The Chevron scientist asked ICI for information, saying “anything you have on the question of permanent injury from paraquat, or any follow-up evaluations several years after spraying would be of benefit to us.”

Notes from an October 1975 meeting between Chevron and ICI recorded that “Chevron are concerned on the chronic effects of paraquat sprays … The syndrome is reported as injury to the CNS …”

The notes state that there may be a need for long-term toxicity studies or an epidemiology study because “Chevron would like more positive data to use in litigation cases”. In the same meeting, it was noted that an autopsy of a recent paraquat poisoning victim had found lesions on the motor neurons “sufficient to cause debilitation” but the notes said it was not clear what might have induced this effect. (Motor neurons are cells in the brain and spinal cord that send commands from the brain to the rest of the body.)

In a December 1975 letter to the Chevron toxicologist, an ICI scientist wrote: “We discussed last week the point you raised about possible chronic effects, which you see causing legal problems. This is a quite terrible problem and, frankly, I do not believe a satisfactory investigation can be made. However, I think some plan could be made, and to be as definitive as possible, any study must be as free from doubt as possible.”

Bad news builds

As the companies fretted, the bad news continued to build: a 1976 autopsy of a farm worker reviewed by ICI showed “degenerative changes” in the “cells of the substantia nigra” of the brain. Such changes are a hallmark of Parkinson’s, but the autopsy said they were probably because of lung damage. A Chevron memo that year noted “gaps in our knowledge of the chronic effects of paraquat exposure”.

By 1985 the science on paraquat health effects had become the subject of vigorous research by independent scientists, and the findings were ringing alarm bells within Chevron’s highest ranks.

In October 1985, an internal memorandum circulated to Chevron officials noted that a study by a Canadian researcher had found “an extraordinarily high correlation” between Parkinson’s and the use of pesticides, including paraquat. The memo also noted that paraquat was “chemically very similar” to the byproduct of synthetic heroin called MTPT, “which produces almost instant Parkinson’s, by killing dopaminergic neurons in the brain”.

The author of the Canadian study had warned that an increase in Parkinson’s disease would be seen as a consequence of the relatively recent introduction of paraquat-like pesticides.

The memo then warned that paraquat could turn out to be a huge legal liability, similar to the fate that befell an asbestos company when the common building material was found to cause cancer.

The asbestos situation “highlighted the especially severe financial risks involved in selling a product which contributes to a chronic disease”, the memo states. “Parkinson’s can go on for decades.”

R Gwin Follis, the retired chairman of Standard Oil – which became known as Chevron in 1984 – wrote to GM Keller, the chairman of Chevron: “I cannot think of anything more horrible for us to bequeath to our successors than an asbestos problem.” Chevron stopped selling paraquat a year later, in 1986.

The “decision to exit the paraquat distribution business was made solely for commercial reasons due to increased competition and did not relate to any health concerns regarding paraquat,” Chevron USA said in a statement to the Guardian.

The company added that during the years a former Chevron subsidiary sold paraquat, it “met or exceeded all federal and state requirements for product-safety testing before and after release on the market”.

A ‘defensive position’

Through the 1990s and into the 2000s, the research on paraquat and Parkinson’s expanded, inside and outside Syngenta. Several US researchers did studies that found unsettling impacts of paraquat on mice, adding more evidence the chemical could cause Parkinson’s.

Syngenta noted these “external pressures on paraquat” and decided its own scientists should repeat studies done by the outside scientists to see if they came up with the same results. There was a caveat: the Syngenta science team “avoided measuring PQ [paraquat] levels in the brain, since the detection of any PQ in the brain (no matter how small) will not be perceived externally in a positive light”, according to an internal Syngenta presentation.

“Data generated will be used to build a scientifically robust, defensive position for paraquat in response to the issues already in the scientific literature, and to questions raised by the media, customers and regulatory authorities,” another Syngenta document stated.

“The issue around the claims that paraquat exposure and Parkinson’s disease are linked needs to be addressed if the future Syngenta aspirations for the product are to be realised.”

Along with making a plan to generate data for its defense, Syngenta began honing a broader “influencing” strategy and “freedom to sell” strategy. A 2003 eight-page document made the objectives clear: the goal was not just to protect paraquat, but to expand its use.

At the time, the chemical was under regulatory review in Australia and the European Union. The company worried about evolving regulatory policies posing “a threat”, including that regulators may start to replace “higher hazard products with lower hazard products”, and apply a “precautionary principle”.

Under that type of regulatory approach, companies seeking to sell a chemical have a burden of proving product safety. In contrast, the US regulatory system largely takes the opposite approach – a chemical must be proven unsafe to be kept off the market.

In response to the growing regulatory threats, Syngenta said it would take several steps, including leading “national, regional and global industry initiatives to influence regulatory policy”.

The company also set as an objective “targeted collaborations with key influencers to improve product image …”

Internal communications show the company discussed consultations with several senior European scientists, and plans to “contribute substantively [sic] to the literature”, including for studies being done for submission to the UK’s Department for Environment, Food and Rural Affairs, and the Agricultural Health Study in the US, a decades-long collaborative research project involving multiple US government agencies.

As Syngenta honed its defenses, the data from its internal studies started to come in. The first internal study done in 2003 was designed to dose mice with paraquat as outside scientists had done, and then measure any loss of dopamine neurons in the substantia nigra of the animals’ brains. Syngenta’s tests did find losses but used a manual counting technique for analyzing those losses that was different from the automated technique used by independent scientists. Under the Syngenta analysis, the impacts of paraquat on the animal’s brains were deemed not statistically significant, a finding Syngenta made public.

What the company did not publicize at the time, however, was the fact that Syngenta scientist Louise Marks, who led the animal studies in question, repeated those studies using the more accurate, automated technique used by independent scientists.

She found that when using an automated analysis technique, paraquat actually did result in a statistically significant loss of the relevant brain cells – just as the outside scientists had found. Marks did another study, and the results were the same. Marks could not be reached for comment.

Deposition testimony given in the current litigation by longtime Syngenta scientist Phillip Botham, which has not previously been made public and during which a judge was not present, indicates that company officials would not tell the EPA of Marks’ research findings until roughly 15 years later, in 2019. The company only told the EPA about the Marks’ data after lawyer Steve Tillery, who in 2019 was suing Syngenta on behalf of people with Parkinson’s, threatened to send the evidence to the EPA himself, according to a transcript of Botham’s testimony.

When asked about the Marks tests, Sygnenta said: “The Marks studies involved a model in which a particular breed of mouse was injected with near-lethal doses of paraquat. Such models are of limited relevance to evaluating the safety of those using paraquat occupationally.”

The deposition also revealed that when Syngenta said on its website that paraquat did not readily cross the blood-brain barrier, and did not reach the specific area of the brain necessary to produce Parkinson’s symptoms, the company knew those statements were not accurate.

When asked in the deposition if that information was true at the time it was posted on the website, Botham admitted it “certainly had some inaccuracies”. “It appears that this communication had not had a chance, for reasons which I can’t fully explain, to catch up with the science that was still emerging,” he said. Part of the reason the company never reported Marks’ findings on its website, he said, was because subsequent research produced different results.

A secret plan

Part of the strategy to influence regulators involved trying to lobby for and against who the EPA looked to for independent scientific advice. In 2005, the EPA was considering appointing Dr Deborah Cory-Slechta to an open position on an important agency scientific advisory panel (SAP) on pesticides. Cory-Slechta was an influential US scientist whose work at the time was establishing ever stronger evidence that paraquat could cause Parkinson’s disease.

“This is important. We do not want to have Cory-Slechta on the SAP core panel,” Syngenta senior research scientist Charles Breckenridge wrote to colleagues in a June 2005 email.

Company emails show Syngenta decided to ask Ray McAllister, a regulatory policy expert at the industry lobbying group CropLife America (CLA), to disparage Cory-Slechta’s work in communications to the EPA. Syngenta officials wrote what they wanted McAllister to tell the EPA, and delivered it to McAllister.

“Ray has a tough job to do in providing comments that don’t come back to haunt CLA and be used against us,” one Syngenta executive wrote to colleagues.

Another Syngenta executive wrote to colleagues that it was “going to be very difficult to pin something really specific on D C-S …”

The company decided secrecy would be key. The company did not want the public or the EPA to know Syngenta was behind the effort.

“I would ask that you handle our comments with care and in such a way that they cannot be attributed to Syngenta,” Greg Watson, a Syngenta regulatory affairs executive, wrote to McAllister. He then suggested that the communications to the EPA about Cory-Slechta “should be submitted informally & NOT placed in the public docket”.

In a separate email, Watson wrote that “for many, many of our projects it would be a real disaster to have her on the SAP!”

Watson suggested, among other things, that McAllister tell the EPA that Cory-Slechta used an “over-interpretation of data” to present scientific conclusions that were “in reality, speculation”, and was someone who made “overly dogmatic” statements.

McAllister communicated the concerns about Cory-Slechta to the EPA without mentioning they came from Syngenta. The agency chose someone else for the advisory panel.

The documents show similar efforts to influence the roster of scientists selected by the EPA for a 2010-11 pesticide advisory panel. At that time, Syngenta advised CropLife to tell the EPA that Cory-Slechta was using her research program for “anti-pesticide advocacy” and was identifying effects “without quality data”.

Cory-Slechta was not selected for the panel in question, while a scientist supported by CropLife was.

When asked to comment about the company’s actions against her, Cory-Slechta said she was not surprised. She said Syngenta representatives had tried various tactics over the years to intimidate her, and also at least once to woo her with an invitation to help fund and collaborate on research.

“They would follow me around,” she said in an interview. “It was clear they were not happy with me. Consistently our research showed that when you administer paraquat in rodent models you would see a loss of dopamine cells … in the substantia nigra. That is the hallmark, or the gold standard, of Parkinson’s disease.”

She said: “They didn’t like the data. They saw a threat to a huge market.”

Cory-Slechta said she is not anti-pesticide, nor pro-pesticide. “I want to stay in the middle,” she said. “I pride myself and I go overboard to stay in the middle. I let myself be led by the data.”

When asked about the Cory-Slechta correspondence, Syngenta said: “We disagree and take exception to this mischaracterization.”

No ‘clear link’

The plaintiffs’ lawyer Steve Tillery was poised to present many of these internal documents and other evidence at a June 2021 trial in Illinois that would have been the first major court challenge to Syngenta and Chevron over the Parkinson’s connection to paraquat.

Just as the trial was set to begin, however, Syngenta agreed to pay $187.5m to settle with the plaintiffs in that case and several others, according to a disclosure in the company’s 2021 financial statement. The company did not admit liability as part of the settlement. It is not clear how much, if any, Chevron might have paid.

Other lawyers are now pressing claims for more than 2,000 other plaintiffs with Parkinson’s disease, including filing lawsuits on behalf of people with Parkinson’s in Canada.

The EPA’s agreement to reconsider its assessment of paraquat was welcomed by the farmworker groups, Parkinson’s scientists and others who brought the court challenge. The agency has said it will take another look at the health risks and costs that come with the widespread use of paraquat, and will have a revised report out in a year.

“Our research partners have studied the ample and compelling evidence showing paraquat’s association with neurological degradation and symptoms related to PD,” Ted Thompson, senior vice-president of public policy at the Michael J Fox Foundation for Parkinson’s Research, said in an email.

“We believe the federal government and the EPA should use every tool at their disposal to eliminate its risk.”

It is not clear however, if the EPA’s extended review of paraquat will change the agency’s position. EPA scientists said in its 2019 draft human health risk assessment that its review of research about the potential association between paraquat and Parkinson’s had found only 71 studies out of 489 to be relevant to the agency’s analysis.

The agency “has not found a clear link between paraquat exposure from labeled uses and adverse health outcomes such as Parkinson’s disease”, the agency states on its website.

While the agency conducts its reassessment, paraquat use continues.