The study in question began in September 2020. The research was conducted by Nadia Gaber, UC San Francisco; Lisa Bero, University of Colorado Anschutz Medical Campus; and, Tracey J Woodruff, UC San Francisco, United States.
They first organised and analysed industry documents regarding chemicals of the PFAS family, archived in special collections of the UCSF Chemical Industry Documents Library (CIDL). The initial aim was to compare reports and comments on health research from the internal documents to the public record, establishing a comparative timeline. The second objective was to evaluate the strategies involved in delaying or obscuring the public health knowledge of the harms of PFAS.
They started by applying deductive codes from published studies in the literature detailing industry strategies of manipulating science. Because the body of documents available was relatively small (only 39), it was possible to read each document and assess whether it matched the codes of interest.
What was known to the public could be traced from PubMed search results. Only 66 studies were published before 1980. In comparison, from 2002 to 2020, more than 2,500 papers were published.
One of the earliest studies in the research, in 1959, examined potential respiratory and dermal toxicity from the manufacture of new plastics, including PTFE (polytetrafluoroethylene, commonly known by its trade name, Teflon). Over the years there were many reports, but the manufacturers were never indicted.
The UCSF researchers, however, identified documents supporting industry influence of science and policy in four of the six strategy areas.
One criterion was influencing the research question itself. Herein, industry decides what to study, or not, in order to produce evidence detracting from harms of their product. In 1978, DuPont’s occupational physician noted “unusually high” liver enzyme elevations but dismissed findings as clinically insignificant, despite inadequate statistical power, neglecting to pursue research.
Another tool employed was to fund and publish favourable research. In 1996, 3M funded a study of occupationally exposed men and found no clinical hepatic toxicity.
Then, of course, was the ploy of suppressing unfavourable research. The UCSF researchers found several such instances. As early as 1961, it was found that PFAS compounds C6, C9, and ART increased the liver size of rats even at low doses, should be handled “with extreme care." Four years later, an Industry Lab report found C8 “highly toxic when inhaled and moderately toxic when injected.” Much later, in 1994, 3M knew of “possible” prostate cancer and shared the details with DuPont.
The chemical manufacturers also distorted public discourse. 3M internal communications from 1980 described C8 as “about as toxic as table salt." A year later, joint employee communications denied that workers had been exposed at levels that could cause adverse health effects, and also denied adverse pregnancy outcomes.
Worse, they even tried to change or set scientific standards. In 1991, DuPont insisted that no Environmental Protection Agency (EPA) notification was warranted, years after determining that PFAS were a chronic hazard. In 2000, DuPont insisted to public water utility officials that its own exposure guidelines were enough.
And, it is not that the PFAS manufacturers did not know of the effects. By 1981, DuPont had removed women of childbearing age from any potential exposure to C8, that was deemed by the researchers as a covert admission of harm. By 1994, DuPont’s internal reports state the company should “evaluate replacement of C-8 with other, less toxic materials.”
Gaber, Bero and Woodruff concluded: Our review of industry documents shows that companies knew PFAS was “highly toxic when inhaled and moderately toxic when ingested” by 1970, forty years before the public health community. Further, the industry used several strategies that have been shown common to tobacco, pharmaceutical and other industries to influence science and regulation—most notably, suppressing unfavorable research and distorting public discourse.