I am writing to highlight what appears to be a major flaw in the paper by Wen et al.1 on levels of per- and polyfluoroalkyl substances (PFAS) and mortality. The authors drew on several rounds of the National Health and Nutrition Examination Survey (NHANES) during the years 1999–2014 and reported mortality risk during 1999–2015 in relation to serum PFAS measured in the NHANES rounds. They noted significant excess risk from all causes, cardiovascular disease, and all cancers in the highest total PFAS tertile and higher risk still in the highest tertile of perfluorooctane sulfonic acid (PFOS) specifically. They provided very large estimated attributable burdens—for example, 15% of all deaths from all causes each year. Such strong effects for these disease categories are quite at variance with results of cohort studies of PFAS-exposed populations.2,3 In the “Methods” section, Wen et al. described adjustment for a number of potential confounders in the NHANES data, such as sex, race, smoking, and body mass index, but they did not mention adjustment for either date of sample taken or NHANES round.
During the period 1999–2014, the serum levels of the two largest components of serum PFAS fell, most dramatically for PFOS, which dropped by over 80%.4 If the outcome measure also changed over those years, then date of sample taken would obviously be a confounder. That is indeed the case with all-cause mortality, cardiovascular mortality, and all cancers mortality, all of which fell over the years 1999–2015.5,6 The results reported by Wen et al. are therefore at risk of confounding; if, indeed, such confounding took place, the results and conclusions could be substantially exaggerated. Depending on the degree of correction, many of the results and much of the discussion in the paper may need replacing.
Refers to https://doi.org/10.1289/EHP10393
Conclusions and opinions are those of the individual authors and do not necessarily reflect the policies or views of EHP Publishing or the National Institute of Environmental Health Sciences.
References
- 1.Wen X, Wang M, Xu X, Li T. 2022. Exposure to per- and polyfluoroalkyl substances and mortality in U.S. adults: a population-based cohort study. Environ Health Perspect 130(6):067007, PMID: 35731224, 10.1289/EHP10393. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
- 2.Winquist A, Steenland K. 2014. Modeled PFOA exposure and coronary artery disease, hypertension, and high cholesterol in community and worker cohorts. Environ Health Perspect 122(12):1299–1305, PMID: 25260175, 10.1289/ehp.1307943. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Li H, Hammarstrand S, Midberg B, Xu Y, Li Y, Olsson DS, et al. 2022. Cancer incidence in a Swedish cohort with high exposure to perfluoroalkyl substances in drinking water. Environ Res 204(pt C):112217, PMID: 34662573, 10.1016/j.envres.2021.112217. [DOI] [PubMed] [Google Scholar]
- 4.Lin PID, Cardenas A, Hauser R, Gold DR, Kleinman KP, Hivert MF, et al. 2021. Temporal trends of concentrations of per- and polyfluoroalkyl substances among adults with overweight and obesity in the United States: results from the diabetes prevention program and NHANES. Environ Int 157:106789, PMID: 34333293, 10.1016/j.envint.2021.106789. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sidney S, Quesenberry CP, Jaffe MG, Sorel M, Nguyen-Huynh MN, Kushi LH, et al. 2016. Recent trends in cardiovascular mortality in the United States and public health goals. JAMA Cardiol 1(5):594–599, PMID: 27438477, 10.1001/jamacardio.2016.1326. [DOI] [PubMed] [Google Scholar]
- 6.National Center for Health Statistics, US Centers for Disease Control and Prevention. Mortality Trends in the United States, 1900–2018. https://www.cdc.gov/nchs/data-visualization/mortality-trends/index.htm [accessed 10 February 2025].