PFAS (per- and polyfluoroalkyl substances) are in the news again, following major settlements reached by dominant U.S. chemical manufacturers (Dupont, Chemours, and Corteva) and 3M. 

The settlements are in the billions of dollars and include pledges to test and eliminate PFAS in drinking water, and in some cases, to pay damages. Notably, the settlements do not require a company to admit liability, they just require it to pay.

The Center for Truth in Science has been working to address PFAS issues since 2021, including funding two important reviews that assess the quality of the principal studies used to develop PFAS regulations and inform policy. These reviews are completely independent of the Center.

The first was a critical review completed by academic toxicologists at the University of Rovira i Virgili in Barcelona, published in Environmental Research. It used a systematic process to evaluate selected studies, identify important gaps in research, and suggest more up-to-date and rigorous methodologies for future PFAS research.

The second was an expert panel review organized and implemented by SciPinion, a firm started by two highly experienced scientists who bring together subject matter experts to examine and provide professional opinions on timely issues facing society. 

SciPinion invited 15 senior scientists to participate in this review, which is described below. A full report on their methods and findings is available here. 

An expert panel review of 11 key studies cited in the risk assessment, public policymaking, and judicial decisions on the toxicity and carcinogenicity of PFAS

Eleven studies were divided into five groups: 

  1. Toxicokinetics/biomonitoring in human participants
  2. Toxicokinetic analyses
  3. Key effects on immune/vaccine response
  4. Key effects on reproduction and development
  5. Key effects on other health issues

Each group was scored by a team of three topic expert reviewers and a lead scientist in the first round. Results were reviewed again by expert reviewers from the first round who expressed an interest. The overall mean and minimum/maximum scores were calculated across reviewers.

Most of the studies had a large amount of variability in scores among the reviewers, with some exceptions, which can be seen in the full report here. 

The panel included a diverse mix of qualified reviewers, all with Ph.D.’s. Eleven were from the U.S., two were from Canada, and one each from Portugal, India, and Germany. Six have academic faculty appointments or recently retired from academia, three consult for private industry, one is in the chemical manufacturing industry, two were former EPA staff, one was a former staff at the National Cancer Institute, one is with the Institute of Health in Portugal, and one works with a respected scientific institute located in India.

General conclusions

Two studies were selected for review on toxicokinetics and biomonitoring in humans: Emmett et. al. (2006) and Olsen et. al. (2007).

  • While it had some limitations, Emmett et. al. (2006) was consistently rated higher for use in risk assessment decision making. The reviewers supported its conclusion that serum levels are strongly and positively correlated with drinking concentrations of PFOA.
  • The reviewers noted that Olsen et. al. (2007) was an important early contribution to the overall literature on variability in serum elimination half-lives of PFOA, PFOS, and PFHxS in humans. However, several limitations in the methodology contributed to lower ratings.
  • Until there are new data, Emmett et. al. (2007) is the preferable half-life estimate to use.

Two papers were selected on the toxicokinetics analysis of data in humans, which revisit earlier estimates of elimination half-life values for select PFAS using more advanced methods: Russell et. al. (2015) and Dourson et. al. (2019).

  • Russell et. al. (2015) ranked highest in the review and showed how relevant background exposures are to analyzing kinetics, and the risk for biased results if not done properly. 
  • Dourson et. al. (2019) found that the published literature on serum elimination half-life of PFOA is highly variable and proposed a biological mechanism behind this finding. 
  • Interestingly, both studies do not appear to be factored into agency assessments of PFAS toxicokinetics, which suggests current estimates of elimination half-lives may be too high.

Two papers were selected on effects associated with immune function and vaccine response in humans: Abraham et. al. (2020) and Grandjean et. al. (2012).

  • The European Food Safety Authority used human data from the Abraham et. al. (2020) study to determine a total weekly intake threshold. But this study scored poorly in the review and scores varied widely due to potential confounders, lack of an unvaccinated control group, and lack of additional study biomarkers.
  • The Grandjean (2012) study also scored poorly. While it has not been directly incorporated in deriving toxicity values and thresholds by agencies, the concern generated by the study has been a rationale for applying a database uncertainty factor of 3 or 10. 
  • Risk assessment practitioners should understand the endpoint of immune/vaccine response for PFAS is still highly uncertain.

Two papers were selected on effects on reproduction and development in mice: Koskela et. al. (2016) and Lau et. al. (2006). Three studies in rats examined other effect endpoints: Butenhoff et. al. (2012), Perkins et. al. (2004), and Luebker et. al. (2005).

  • The five studies looked at a range of effects from exposure to PFOA and/or PFOS. Each has been used in one or more federal and state regulatory agencies as one of the main studies to support their current health benchmark for PFOA/PFOS.
  • The reviewers found no significant risk of bias for any of these studies.
  • Lau et. al. (2006) received the most consistent and highest scores from the reviewers, and it found a dose/response relationship between the administered dose of PFOA and increased maternal liver weight in mice. 
  • It also demonstrated a similar rate of kidney elimination between males and females. This is strong evidence that extrapolating from one species to another has a high uncertainty level. For example, in rats there is a much lower elimination rate in males compared to females.
  • Koskela et. al. (2016) had too many limitations to show evidence of effects of PFOA on bone morphology and density.
  • The three rat studies had high variability in scores from reviewers, but received consistently high average scores. Two of the studies provided relatively strong and consistent evidence for adverse effects on tissue.
  • The reviewers concluded the PFOS study by Luebker (2005) adequately demonstrated the potential for in utero transfer of PFOS to the fetus and to pups through lactation. 
  • These studies show that reproduction effects are critical in rodents.

The reviewers also identified more than a dozen other studies that provide additional insights

regarding the exposure, toxicity including immunotoxicity and carcinogenicity, and kinetics of

PFOA, PFOS, and other PFAS compounds. These should be included in future analyses. 

The Center has made the U.S. Environmental Protection Agency aware of this research through the public comment process, in response to the agency’s request for public input to help in the consideration of potential future PFAS regulations. 

The findings from both reviews strongly suggest that more science is needed to understand the issue of PFAS. We hope that continued research will be conducted, and decisions will be made, with the understanding that we are far from scientific consensus when it comes to the hazards of different types of PFAS.