There are two new studies out this week that raise more questions about the dangers of DuPont Co.’s chemical, C8.
First, there’s this paper published in the journal Environmental Health Perspectives, concluding:
This study suggests an association between PFC exposure and juvenile asthma. Due to widespread exposure to these chemicals, these findings may be of potential public health concern.
Then, there’s this paper, also published yesterday in that same journal:
This study suggests that higher PFOA serum levels may be associated with testicular, kidney, prostate, and ovarian cancers and non-Hodgkin’s lymphoma. Strengths of this study include near-complete case ascertainment for state residents, and well characterized contrasts in predicted PFOA serum levels from 6 contaminated water supplies.
That one is especially interesting, given that it was co-authored by Tony Fletcher of the London School of Hygiene and Tropical Medicine. Dr. Fletcher was a member of the C8 Science Panel, which previously ruled that the only “probable links” between C8 exposure and cancers were for testicular and kidney cancers.
UPDATED: I asked Dr. Fletcher to explain the contradictions between this new study and the Science Panel’s probable link findings, and this is the email he sent me:
This is a general issue of assessing the evidence which has been faced for all the probable link assessments. As described in the Science Panel Probable Links reports, for each evaluation we considered evidence together from published studies done by others, and from published and unpublished work done by the Science Panel. We considered the weight of the evidence, including consistency across studies. We took into account the capability of different study designs to detect a disease risk if such a risk exists: this capability of each study depends on both study size (a large study is better able to show the presence or absence of risk) and study quality (how good was the exposure assessment, how complete was the study population, how well controlled were the potential confounding factors etc). In many cases we found that some results were quite different , where one study suggested a risk and another did not, or within one study, different ways of analysing the results lead to different patterns of results. In each case we then judged whether the balance of evidence convinced us that it is more likely than not that an observed excess risk was explained by the connection between PFOA exposure and the particular human disease. Or alternatively we judged that an excess was more likely than not due to chance or study artefact. Where an individual positive association was based on stronger evidence (for example with a convincing pattern of increasing risk with exposure category and strong statistical significance), this carries more weight than results with weaker significance tests and without supportive results from other studies/analyses. Such considerations guide us in determining what is a “probable link” as defined in the Settlement Agreement.
The specific questions you raise are examples of this. Some of the particular set of results you can see in this paper (prostate cancer, ovarian cancer and non-Hodgkin’s lymphoma) were based on small numbers or showed unimpressive p-values or were not replicated by other studies, and when we put the evidence together we concluded for these that there was overall insufficient evidence to support a probable link with C8.
Finally, there was a third paper, this article, published in late December in the journal Environmental Science and Technology:
The present study investigated variation in perfluorocarbon levels of 9,952 women of childbearing age who had been exposed to perfluorooctanoic acid (PFOA) in drinking water contaminated by industrial waste. An analysis of variance with contrast was performed to compare the levels of PFOA and perfluorooctanesulfonic acid (PFOS) in pregnant and nonpregnant women overall and during each trimester of pregnancy. We found that pregnant women had lower circulating PFOA and PFOS concentrations in peripheral blood than nonpregnant women and that PFOA levels were consistently lower throughout all trimesters for pregnancy, suggesting transfer to the fetus at an early stage of gestation. These results are discussed in the context of the endocrine-disrupting properties of perfluoroalkyl substances that have been characterized in animal and human studies. Our conclusion is that further, systematic study of the potential implications of intrauterine perfluorocarbon exposure during critical periods of fetal development is urgently needed.