On August 23, 2023, the EPA Human Studies Review Board (HSRB) approved a report which recommended major revisions to EPA’s approach for assessing formaldehyde risks. The HSRB was provided a detailed charge to inform EPA’s assessment of formaldehyde under both the Toxic Substances Control Act (TSCA) and the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA).1
The charge to the HSRB focused on assessing the acute endpoints related to sensory irritation, with two distinct sets of charge questions. One set of charge questions focused on ethics, asking the HSRB to evaluate whether the studies were conducted ethically and whether they met prevailing ethical standards at the time studies were conducted. The HSRB found that all the studies evaluated were conducted ethically. A second set of questions asked the HSRB to evaluate four specific studies and determine whether they were each “scientifically sound, providing reliable data for use in a weight-of-evidence to determine a point of departure for acute inhalation exposures to formaldehyde.” In addition to reviewing four studies individually, the HSRB was asked to comment on the use of the four studies in EPA’s weight of evidence evaluation for acute inhalation endpoints, as well as on the proposed points of departure (PODs) presented by EPA as the basis for calculating safe exposure limits.
These charge questions required the HSRB to do a deep dive into the scientific issues and conduct its own evaluation of the studies. This is in stark contrast to the recent review of formaldehyde by the National Academies of Sciences, Engineering, and Medicine Committee (NASEM committee) on the Review of EPA’s 2022 Draft Formaldehyde Assessment, where EPA explicitly told the NASEM committee that that NASEM “shall not conduct an independent assessment separately from the IRIS document.”2 Unlike the constraints that EPA put on the NASEM committee, the HSRB review “considered materials presented at both meetings, both from the EPA and from the public comments, research articles, and related materials, the Agency’s review, the Agency’s statistical analysis of the research data, and oral comments from Agency staff during the HSRB meeting discussions.”3 In fact, the HSRB review was sufficiently broad that the HSRB took it upon itself to review not just the four studies it was asked about, but also two additional studies that were part of EPA’s weight of evidence evaluation for the sensory irritation endpoint.
The HSRB Deep Dive into the Methods of the Studies Led to Findings of Significant Concern
The HSRB review provided many important recommendations to EPA that have a broad impact not just on the TSCA and FIFRA risk evaluations of formaldehyde, but also on EPA’s Draft IRIS assessment for formaldehyde, which the agency is seeking to finalize. A few of these important recommendations are described below.
- When commenting on study designs and which studies should be used to develop a point of departure (POD) for effects, the HSRB favored the use of controlled chamber studies (like Kulle et al. (1987) and Lang et al. (2008)) over the use of observational studies (like Hanrahan et al. (1984) and Liu et al. (1991)). HSRB found that the controlled chamber studies had a “preferred study design and greater scientific rigor” than the observational epidemiology studies and recommended that EPA rely on the controlled chamber studies rather than observational epidemiology studies.
- When reviewing EPA’s finding that Haber’s Rule applies to formaldehyde, the HSRB, after considering the scientific evidence (including responses from EPA staff, consideration of previous EPA applications of Haber’s Rule, and public comments), determined that formaldehyde does not follow Haber’s Rule, and no exposure adjustment is necessary when calculating a POD.4
- When evaluating the sensory irritation endpoint, the HSRB questioned whether it met EPA’s definition of an “adverse” health effect and instead suggested that these endpoints could be used as a lower bound for potential adverse effects, rather than a POD for adversity. Consistent with the determination that sensory irritation is not an adverse health effect, the HSRB recommended that uncertainty factors do not need to be applied to the POD.
- When evaluating which populations are most sensitive to sensory irritation, the HSRB concluded that younger individuals are more sensitive to sensory irritation than older individuals. Consistent with this finding, the HSRB recommended that an uncertainty factor need not be applied when the study population included younger individuals. The HSRB stated that this approach is consistent with the approach EPA used in its assessment of chloropicrin.
- The HSRB recommended that EPA have a “more coordinated approach” as its different offices conduct their evaluations for formaldehyde. In particular, the HSRB cited NASEM and the TSCA Science Advisory Committee on Chemicals (SACC), as well as other state and federal agencies, as groups which EPA should bring together, rather than isolating them, when establishing PODs.
The HSRB In-Depth Science Review Led to Findings that Contradict the Limited NASEM Review
The NASEM committee recently released its process review of EPA’s draft assessment of formaldehyde, and, not surprisingly, the HSRB review findings contradict some of the NASEM findings. The NASEM committee’s task was to assess “whether EPA’s draft document adequately and transparently evaluated the scientific literature, used appropriate methods to synthesize the current state-of-the science, and presented conclusions regarding the hazard identification analysis and dose-response analysis of formaldehyde that are supported by the scientific evidence.”5 Importantly, as noted by the NASEM committee, “[t]he committee did not conduct an independent hazard evaluation or dose-response assessment, and therefore does not recommend alternative hazard identification conclusions or toxicity values. The committee also was not charged with commenting on other interpretations of scientific information relevant to the hazards and risks of formaldehyde, nor did its statement of task call for a review of alternative opinions on EPA’s formaldehyde assessment.” Unlike the HSRB, the NASEM committee did not consider alternative opinions, did not conduct its own independent review, and did not conduct an evaluation that looked at the weight of the scientific evidence, due to the narrow scope with which it was charged.
Regarding the process EPA followed, the NASEM Committee provided a Tier 1 recommendation that “EPA should revise its assessment to ensure that users can find and follow the methods used in each step of the assessment for each health outcome. EPA should eliminate redundancies by providing a single presentation of the methods used in the hazard identification and dose-response processes. A central roadmap and cross-references are also needed to facilitate access to related sections across the different elements of the assessment (e.g., appendixes, main document) for the different outcomes analyzed. Related Tier 2 recommendations would amplify the impact of this Tier 1 recommendation in improving the assessment.”
A Case Study Shows the Differences Between the HSRB and NASEM Review Approaches, But Both Reviews Find Flaws in a Critical Study
As described above, the HSRB charge required that it conduct an in-depth review of study methods and findings. The NASEM committee did not conduct an in-depth review but did however provide a case study to determine if the committee could replicate the steps described in the 2022 Draft Formaldehyde Assessment to translate findings from Hanrahan et al. (1984) to develop a sensory irritation POD, as replicability is a critical indicator of regulatory transparency. HSRB also evaluated Hanrahan et al. (1984), but as noted above, the HSRB review did look closely at EPA’s methods.
Neither the HSRB nor the NASEM committee supported the use of the Hanrahan et al. (1984) study. The NASEM committee determined that “[t]he committee’s review of the Hanrahan et al. (1984) study generated concerns not captured by EPA’s review.”6 The committee noted, among other findings, that the data collection instrument was not well described in the study and that this was not consistent with rating of “medium” quality that EPA gave to the study. It also noted deficiencies in the analysis domain of the study that EPA did not recognize. Additionally, the committee stated that “[i]nconsistencies are also apparent in how study quality deficiencies are represented in the four-domain summary confidence figure for human sensory irritation studies.” The case study review found inconsistencies between EPA’s evaluation criteria for sensory irritation and how they were applied to the Hanrahan et al. (1984) study, as well as how study limitations were presented for other sensory irritation studies. The case study findings support the NASEM committee’s overall conclusion that it could not replicate the EPA process with “complete fidelity.”
The HSRB found that observational studies, like Hanrahan, cannot and should not be used to support a quantitative POD. The HSRB in-depth review of the methods also determined that the study did not meet its original intent of being representative of Wisconsin mobile homes, as noted by the authors. The HSRB justifiably asked EPA to provide a rationale for the value of the study as the HSRB did not find it representative or generalizable.
HSRB and NASEM Reviews Show That Significant Revisions to Both the Systematic Review and the Scientific Conclusions Reached in the Draft IRIS Assessment Are Needed
A review of general methods and approaches by the NASEM committee found EPA’s methods, as described above in the case study, to be lacking in transparency and consistency. This was reflected in NASEM’s most important Tier 1 recommendation to EPA. Not surprisingly, when HSRB dove deeply into the individual studies and conducted an independent evaluation of EPA’s conclusions, it disagreed with EPA’s choice of preferred studies and dose-response conclusions, including the application of uncertainty factors for sensory irritation endpoints, as is summarized above.
While the EPA’s HSRB evaluated only acute sensory irritation endpoints, the same studies it evaluated were relied upon by EPA for chronic endpoints (e.g., Hanrahan et al. (1984)) in the draft IRIS assessment. Thus, the HSRB findings regarding the strengths and weaknesses of these studies are relevant to the draft IRIS assessment. Importantly, the HSRB review shows that an in-depth science review, according to EPA’s own guidelines, is necessary for all non-cancer and cancer endpoints in the IRIS assessment. While the NASEM committee conducted a review for transparency and clarity, their review did not include an independent evaluation of the science; nevertheless, as is shown by the case study evaluated by the NASEM committee, and the NASEM Tier 1 recommendation to EPA, significant process concerns were identified in the 2022 Draft Formaldehyde Assessment. The HSRB and NASEM reviews support the need for an independent in-depth review of all science underlying the critical endpoints and conclusory revisions to the 2022 Draft IRIS Formaldehyde Assessment.
1According to the HSRB Charter, “[t]he Committee was created to review human research submitted to the Office of Pesticide Programs to be used for regulatory purposes under the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA, 7 U.S.C. 136-136y) or section 408 of the Federal Food, Drug and Cosmetic Act (FFDCA, 21 U.S.C. 346a).”
2For additional details see our previous blog post available here.
3See the HSRB report discussion of its review process available here.
4Haber’s Rule states that the severity of a toxic effect is a function of the time of exposure and the concentration of exposure, and that as these increase, the toxic effect increases. The impact of not applying Haber’s Rule to formaldehyde means that PODs from short-term studies are not adjusted downward when calculating a POD for a longer-term exposure.
5The NASEM 2022 Report can be found here.
6Id. at page 142.