This case study is about environmental and occupational exposure to a chemical called perfluorooctanoic acid (PFOA); the sequestering and the revelation of information about potentially severe health problems associated with exposure. Used to manufacture Teflon and other non-stick and stain-resistant products, PFOA has been detected in a high percentage of samples of human blood and house dust taken from homes in Massachusetts, Maine, New York, Oregon and California, and it has contaminated drinking water in some communities in West Virginia and Minnesota. DuPont and 3M, which manufactured the chemical and used it in manufacturing processes, were aware of the persistence and potentially harmful health effects for many years, without notifying the US Environmental Protection Agency (EPA) of the adverse effects information, as required under the Toxic Substances Control Act (TSCA). EPA will fine DuPont a minimum of $13 million and a potential maximum of $300 million  for its violation of TSCA, and plaintiffs have filed three class action lawsuits against DuPont and 3M, alleging property damage and current and future adverse health effects. One group of plaintiffs in Parkersburg, West Virginia has settled with DuPont; the settlement included provisions for an on-going screening and surveillance program in the community, with the potential for further epidemiologic investigation if recommended by an independent panel of expert epidemiologists. Another lawsuit in Minnesota is pending. The case of PFOA illustrates the importance of litigation as a tool for protecting public health in the context of a regulatory system that requires minimal testing of chemicals before they are marketed and relies on manufacturers to report evidence of risks to health.
When PFOA was first manufactured and used in manufacturing processes, in 1951, no statute required manufacturers of chemicals to submit information prior to marketing. When the Toxics Substances Control Act (TSCA) was enacted by Congress in 1976, more than 63,000 chemicals, including PFOA, were granted blanket approval for use in consumer and industrial products. Under TSCA, EPA could have requested safety testing if the agency believed that the chemical might pose risks to workers or consumers; EPA did not prioritize PFOA for such review. Within ten to twenty years of its initial introduction into manufacturing processes (by the year TSCA was enacted), manufacturers had data suggesting that PFOA accumulates in human blood, does not break down in the environment, and may cause serious health problems, including liver damage, reproductive and developmental defects, and cancer.
Although TSCA required the companies to report adverse health effects, no current statute requires companies to monitor PFOA emissions from production. Since 1961, when the companies such as DuPont and 3M—the two primary producers of PFOA—first learned of the potential threat to health, millions of people have been exposed to PFOA in the workplace, in food, and in water. Widespread exposure to a substance suspected of causing birth defects, cancer, and liver damage might have continued indefinitely, but for litigation initiated by people concerned about contaminated water in their community.
The case of PFOA raises troubling questions about the effectiveness of current mechanisms for assessing and regulating chemicals in the United States. The legislation controlling the manufacturing of industrial chemicals—which make up the vast majority of chemicals produced in the United States—empowers but does not require EPA to request data on potential health problems for people who may be exposed. Each year, thousands of new chemicals (not all of them toxic) are manufactured in the U.S. with only limited or non-comprehensive information about potential acute or chronic effects made available to the government. Pre-market testing of agricultural chemicals is more rigorous because data are required before EPA will grant a license for use. Yet for the most part, knowledge about the full range of potential harmful effects of chemicals in commerce is available for only a small percentage. Industrial chemicals are judged innocent until proven guilty, but the inherent limitations of environmental health sciences—coupled with the impotence of the regulatory framework in the US—make it very difficult to prove guilt, until people who have been affected cry foul.
The Toxic Substances Control Act
The Toxic Substances Control Act (TSCA) was enacted by Congress in 1976 to provide the Environmental Protection Agency the ability to track the nearly 100,000 industrial chemicals currently produced or imported into the United States. TSCA mandated the creation of an inventory of chemicals considered safe unless the EPA can provide evidence of an unreasonable risk to human health or the environment. Several key sections of the Act influence the regulation of new or existing chemicals:
- Under Section 4 of TSCA, an EPA Administrator may require the testing of new or existing chemical substances if the chemical will be produced in substantial quantities and may enter the environment, resulting in significant or substantial human exposure, or if there are data to suggest that the chemical presents an unreasonable risk to health or the environment.
Section 5 of TSCA prohibits the manufacture, processing, or import of a “new chemical substance” or “significant new use” of an existing substance unless a premanufacture notification (PMN) is submitted to EPA at least 90 days before the start of manufacture or processing. The PMN contains information on the chemical identity, physical characteristics, processing and use, and available toxicity data (information on the chemical's potential adverse effects on human health or the environment). During the 90-day period, EPA reviews this and any other available information. It can then request more data, prohibit or limit manufacture, or terminate the review process.
- Section 6 of TSCA authorizes the EPA to issue regulations to address the risks of existing substances; these regulations range from labeling requirements to an outright ban of the substance.
- Section 8 of TSCA authorizes the EPA to require industry to maintain records on and/or report environmental and health effects. Specifically, manufacturers, processors, and distributors of chemicals must maintain records of significant adverse effects on health or the environment caused by the substance or mixture. They must submit lists and copies of health and safety studies to EPA; studies showing adverse effects previously unknown must be submitted to EPA as soon as they are completed or discovered.
TSCA has several limitations. To restrict a new or existing chemical under the Act, the EPA must prove that the chemical “will present an unreasonable risk;” that the proposed regulation is the least burdensome possible to reduce risks to a reasonable level; and that the benefits of regulation outweigh the costs to industry. This proof must be developed on a chemical-by-chemical basis, creating an enormous workload for EPA. The short time frame for action, the minimal information available, the potential inaccuracy of short-cut measures to assess hazard and exposure , and the allocation of the burden of proof to the government, rather than to industry, has resulted in minimal action by EPA: as of 1994, EPA had restricted only four new and five existing chemicals. TSCA’s adverse reporting requirement is one of the few sections of the Act that has been used to successfully to make data available for government consideration. Nonetheless, it creates a disincentive for voluntary testing.
Health Risks of PFOA
Perfluorooctanoic acid is used by DuPont in the manufacturing of several stain resistant products including Teflon. Also known as C8 (as it has eight carbons), PFOA is used in the manufacturing of these products to hold raw materials in suspension while they are they are spread on surfaces and dried. As the waxy non-stick material is drying, the PFOA is removed. While most of the solids are recycled or incinerated, some are emitted into the environment through the air and water.
Early Industry Studies
As early as 1961, DuPont scientists issued internal warnings about the health risks of PFOA. Following these warnings, DuPont and 3M, the chief supplier of the chemical, began studying the health outcomes associated with PFOA. These studies demonstrated that PFOA accumulates in human blood and does not break down in the environment and therefore may cause serious health problems, including liver damage. In 1978, DuPont began a testing program to understand the health effects of PFOA and by 1980 had found elevated levels of the chemical in workers at the DuPont Washington Works plant in Wood County, West Virginia.
Based on its findings, the company recommended that workers with potential exposure to PFOA wear special protective clothing and equipment, including breathing devices. A company memo, dated July 1980, concluded that, "continued exposure (to C8) is not tolerable." Between 1979 and 1981, two of seven children born to female plant workers at the Washington Works plant had birth defects, which led DuPont to reassign 50 women who worked in the plant's Teflon production division to other areas of the plant. 3-M sent Dupont results from a study demonstrating that PFOA causes birth defects in rats prompting Dupont to conduct additional occupational studies. In addition to these occupational studies (which are still ongoing), DuPont conducted tests, between 1981 and 1984, for PFOA contamination in tap water in Little Hocking, Ohio, which lies on the Ohio River opposite the Washington Works plant in West Virginia. A DuPont memo stamped "personal and confidential" stated that these tests found evidence of C8 contamination in drinking water on both the Ohio and West Virginia sides of the river. None of this information was reported to EPA.
Industry and Academic Research: 1980’s and 1990’s
Throughout the 1980’s and 1990’s several animal toxicological studies were conducted on PFOA, both by industry and academics. These studies demonstrated that PFOA had a direct effect on the liver of rats and in particular increased the size of the liver. Increased liver size is non-specific, but is evidence of some adverse effect on the liver. PFOA was also shown to be associated with an increase in Leydig cell adenomas (tumors) (also here), tumors in the pancreatic cells, decrease in thyroid gland weight, and loss of body weight and decreased resting heart rates. Studies on plant workers who were exposed to PFOA soon followed. In 1993, researchers in Minnesota found levels of PFOA in exposed workers to be higher than the general population. In this retrospective cohort mortality study, the investigators examined the relationship between PFOA exposure and mortality due to various causes. In a cohort study of 3M workers employed between 1947 and 1983 in their Minnesota plant (2,788 male and 749 female employees), ten years of employment in jobs in which workers were exposed to PFOA was associated with a 3.3-fold increase (95% CI, 1.02 to 10.6) in prostate cancer mortality compared to those with no exposure. Both DuPont and 3M dismissed these results on the basis of small sample size.
Though the animal and occupational studies did not consider drinking water as a route of exposure, DuPont established an internal "community exposure standard" of 1 part per billion in drinking water in 1991. With these standards in place, DuPont continued the use of PFOA. By May, 2000, 3M was sufficiently concerned about the health and safety risks of PFOA to stop manufacturing the chemical.
Class Action Lawsuit
In August of 2001, residents living near Dupont’s Teflon plant in Wood County, West Virginia filed a class action lawsuit alleging that DuPont had knowingly discharged PFOA into water supplies in amounts exceeding the company’s 1 part per billion guidelines, that the PFOA releases adversely impacted and impact property values, and that PFOA exposure resulted in increased risk of illness. Following the filing of relevant documents, including a symptom and disease prevalence study by Dr. James Dahlgren, et al , in April of 2003, Judge George W. Hill, the presiding judge in the class action lawsuit, determined that PFOA is toxic and hazardous to humans. Judge Hill ordered DuPont to obtain, test and analyze samples of blood from members of the class action suit. The 25,000-50,000 members of the suit were defined as individuals who consumed water contaminated with PFOA from any water supply in the vicinity of the plant, not just those whose household drinking water was supplied by the two specific water companies.
In May 2003 DuPont appealed the court order stating that the ruling was not based on, as stated in a DuPont press release, “sound science.” They requested that Judge Hill step down from the case because he lived in a neighborhood near Parkersburg affected by the PFOA releases and would therefore be considered a member of the class. Judge Hill rejected the motion to disqualify himself from the class action suit because he did not live in the area contaminated with PFOA. DuPont’s lawyers appealed the case to the West Virginia Supreme Court of Appeals on the basis of these two key issues: the blood testing ordered for members of the class (DuPont stated that the blood tests were unnecessary) and the move to disqualify Judge Hill from the case. The Supreme Court ruled in favor of the defendants, agreeing that the blood tests were not necessary. However, the Court denied the defendants move to disqualify Judge Hill and thus the Judge continued to preside over the case.
In May 2004, the West Virginia State Supreme Court ruled that three previously sealed DuPont documents be unsealed and entered into the evidence of the class action lawsuit. One of the documents was a November 2000 communication from DuPont attorney John R. Bowman, prepared after the water district responsible for supplying water to one of the affected areas (Lubeck, in Mason County, WV) sent a letter to its customers notifying them C8 had been found in their drinking water (the letter stated that DuPont found “low concentrations” of C8 in drinking water in two wells). The memo from attorney Bowman states: "In view of the interest the letter is getting I think we need to make more of an effort to get the business to look into what we can do to get the Lubeck community a clean source of water or filter the C8 out of the water." In the memo, Bowman also states, "Our story is not a good one, we continued to increase our emissions into the river in spite of internal commitments to reduce or eliminate the release of this chemical into the community and the environment because of our concern about the biopersistence of this chemical."
In September 2004 DuPont agreed to a settlement of the class action lawsuit with terms for property claims, medical monitoring, and potentially, further health studies and compensation for medical conditions identified as being linked to PFOA exposure worth up to $343 million. In February 2005, Judge Hill approved a $107.6 million initial settlement between the residents and DuPont. The settlement agreement requires the company to clean water systems contaminated with C8 and to establish an unbiased scientific panel to explore potential links between C8 exposure and human disease. The settlement further stipulates that if a link is found, residents with a disease or health problem related to C8 may seek additional compensation beyond the $343 million settlement. If no link is established, the company will be free from any future lawsuits from class members related to C8 health claims. A panel of three well-respected epidemiologists, Anthony Fletcher of the London School of Hygiene, David Savitz of the University of North Carolina at Chapel Hill and Kyle Steenland of Emory University, was established to oversee studies into associations between PFOA and health effects.
In June 2000, EPA announced it was expanding its investigation of perfluorooctyl sulfonate (the investigation began in 1999) to include other fluorochemicals, including PFOA. Informed by documents unearthed during the class action lawsuit, in 2001, EPA received copies of documents from one of the plaintiffs’ lawyers, Robert Bilott, indicating that DuPont had: 1) studied PFOA in pregnant workers and their offspring as early as 1981 and 2) performed substantial drinking water testing in homes and businesses showing that Dupont understood in 1987 and confirmed repeatedly in later years that PFOA levels in drinking water near its Washington Works facility exceeded its community exposure guidelines (these documents later formed the basis of EPA’s case against Dupont showing violations under TSCA). The EPA Office of Pollution Prevention and Toxics released a preliminary hazard assessment of PFOA (“Draft Hazard Assessment of Perfluorooctanoic Acid and Its Salts”) in April 2002. By the end of September 2002, EPA had initiated a "priority review" (an analysis used to determine whether expedited regulatory action is warranted) of C8. The priority review was prompted by articles in the scientific literature documenting associations between C8 exposure and reproductive and developmental effects; these results and information revealed in the class action litigation raised the possibility that PFOA might meet criteria for consideration under TSCA. As part of the hazard review, EPA published an interim revised hazard assessment (“Revised Draft Hazard Assessment of Perfluorooctanoic Acid and Its Salts”) in November 2002.
In the spring of 2003, EPA decided to defer regulation of C8 and instead pursued a consent decree with DuPont industry officials requiring new research into routes of exposure. This decision had implications for plaintiffs' attorneys in the West Virginia class action lawsuit that was pending at that time: attorneys could not cite EPA regulation as evidence of the chemical's toxicity. In April 2003, EPA released a preliminary risk assessment of the developmental toxicity associated with PFOA, which demonstrated widespread exposure in the U.S. population and linked exposure to the chemical with developmental and other health effects. The risk assessment also stated that there were considerable gaps in data about the health effects of the chemical which had to be addressed before the EPA could decide in favor of or against regulatory action.
In June 2003 the EPA held a public hearing on an enforceable consent agreement (ECA) development process for PFOA (ECAs are publicly negotiated agreements among industry, EPA, and others which require industry to generate data and submit them to EPA by stipulated deadlines). During the five hour hearing, officials from DuPont and other major chemical companies, environmentalists and others discussed the merits of requiring testing of PFOA under TSCA.
Having decided against regulatory action under TSCA, EPA filed a suit against DuPont in July, 2004, alleging that the company caused “widespread contamination” of PFOA in drinking-water supplies near its Parkersburg, West Virginia plant at levels exceeding the company’s community exposure guidelines. The suit also alleged that DuPont had - over a 20 year period - repeatedly failed to submit information on adverse effects (in particular, information on liver enzyme alterations and birth defects in offspring of female Parkersburg workers) that it had obtained regarding the chemical as required under TSCA. Last, the suit alleged that DuPont had violated Resource Conservation Recovery Act (RCRA) hazardous waste management regulations. The suit seeks millions of dollars in fines from DuPont for violations under these two statutes. On December 6th, the EPA issued a second complaint, expanding its case against DuPont in order to include evidence of elevated blood levels of PFOA in residents near the West Virginia plant, discovered by the company in July 2004.
In January 2005, the EPA issued a new draft risk assessment of the health risks of PFOA. The report states that people nationwide could face "a potential risk of developmental and other adverse effects" from exposure to low levels of C8. The report, which relies primarily on animal studies, cites evidence that C8 is carcinogenic in rats, but considers the human evidence inadequate. The assessment suggests the chemical targets the liver and is present in the breast milk of rats. It concludes that the chemical could raise cholesterol and triglyceride levels in humans. The risk assessment does not include information from studies of workers exposed to C8, as EPA did not believe such information was adequate for the report. In releasing the risk assessment, EPA officials stated they have yet to reach a conclusion on how C8 affects human health and will consult with their Science Advisory Board prior to issuing a more definitive statement. Thus, there have been several drafts of the risk assessment, but a final risk assessment document has not yet been published. A recent draft Science Advisory Board review of EPA’s draft risk assessment published in June, 2005, reaches very different conclusions than EPA: the SAB panel suggests that PFOA should be classified as a likely human carcinogen. In February 2006, the peer review panel voted to approve the recommendation that PFOA be considered a likely carcinogen, as outlined in its January 2006 draft report. The panel's final report is available here.
In May, 2005 DuPont and EPA tentatively agreed to a multimillion dollar settlement of EPA’s charges that the company failed to report health and pollution risks of PFOA exposure. On December 14, 2005, EPA forwarded to the Environmental Appeals Board a settlement with DuPont. A DuPont Vice President said their interpretation of the reporting requirements differed from the Agency’s, but the settlement allowed the company to put the matter to rest. The settlement requires DuPont to pay $10.25 million in civil penalties and perform Supplemental Environmental Projects (SEP)—environmentally beneficial projects that violators agree to undertake in exchange for mitigation of the penalty to be paid— worth $6.25 million. The settlement is pending approval by the Environmental Appeals Board. The first SEP to be completed is a project designed to evaluate the potential of nine of DuPont’s fluorotelomer-based products to breakdown into PFOA in order to improve the current understanding of potential sources of PFOA in the environment and routes of exposure.
In January 2006, EPA called upon eight manufacturers to voluntarily reduce PFOA and related chemicals. All eight fluoropolymer and telomer manufacturers have agreed to participate in EPA's global stewardship program for PFOA.
This case study describes a recent example of the interplay between the sequestering of science, differing interpretations of bodies of evidence, and a regulatory system that places the burden for proving risks from exposure to chemicals on the government, the result of which is widespread exposure to a hazardous chemical over many years. The case of PFOA further illustrates the importance of litigation as a tool for exposing failures by manufacturers to comply with legal requirements, and for prompting action by government agencies whose resources are woefully insufficient for the oversight and regulatory functions for which they are responsible by law.
Update (August 2011)\
On August 26, 2011 after 10 years of litigation, a settlement was approved between a groups of plaintiffs living near DuPont's Chamber Works plant in Salem County, New Jersey, and DuPont. The plaintiffs included about 4,000 households, and DuPont agreed to pay $8.3 million for drinking water filtering systems or an equivalent amount for households wanting an alternative (e.g., purchasing bottled water.) The settlement was approved by the U.S. District Court for the District of New Jersey Camden Vicinage. The 61-page decision (Richard Rowe, et al, and Misty Scott et al v. DuPont) is here.
Update (November 2012)
The C8 Science Panel issued their final newsletter noting "[we] have now completed our primary task, which is to provide to the Wood County Court a series of evaluations on whether or not there is a "Probable Link" between exposure to PFOA/C8 and a range of diseases among people living in the Mid-Ohio Valley." The C8 Science Panel determined that there is a probable association between the following health conditions and exposure to PFOA:
- High Cholesterol
- Testicular Cancer
- Thyroid Disease
- Pregnancy-Induced Hypertension/Preeclampsia
- Ulcerative Colitis
The C8 Science Panel's study reports are available here.
Links and Resources
Dupont's PFOA Information page
DuPont's news release (1/11/05)
Press release on PFOA in drinking water from New Jersey's Department of Environmental Protection
Guidance on PFOA levels in drinking water from New Jersey's Department of Environmental Protection, Division of Science, Research and Technology
Ohio legislation requiring regulating PFOA as a toxic air pollutant
Minnesota's addition of PFOA and PFOS to the Health Risk Limits for Groundwater Rules
Mortality of Employees of an Ammonium Perfluorooctanoate Production Facility, 2007 (here)
WV School of Medicine, Stephanie J. Frisbee. C8 Health Project: How a Class Action Lawsuit Can Interact with Public Health--History of Events. (PowerPoint slides here)
Robert A. Bilott letter to EPA Administrator Stephen Johnson and ATSDR Director Howard Frumkin, May 29, 2008 (here)
ATSDR Director Howard Frumkin's letter to Barbara Taylor, Director of West Virginia's Office of Environmental Health Services, February 10, 2009 (here)
EPA has an 8(e) TSCA docket on PFOA (docket #: AR-226); electronic versions of documents are not posted there in a timely manner, but EPA docket staff can retrieve documents for you if you use the AR-226 docket number.
About the Authors
This case study was written by Richard Clapp DSc and Polly Hoppin ScD, along with Jyotsna Jagai (PhD candidate at Tufts University School of Public Health) and Sara Donahue (former SKAPP project staff member). Dr. Clapp served as an expert witness for plaintiffs in the class action lawsuit between DuPont and citizens of Parkersburg, WV.
 EPA has said they will not impose this maximum.
 When toxicological or epidemiological data do not exist, EPA uses methods such as analysis of structure activity relationships -- in which new chemicals are compared to existing chemicals with similar molecular structures for which testing data are available -- to predict health and environmental effects. These predictions are not always accurate. Information used to assess potential exposures to new chemicals, such as production volume and anticipated uses, are estimates provided by industry, and can change substantially once EPA completes its review and manufacture begins.
 Publication forthcoming, report was filed with the court.