Legal Implications of Hormesis:
Kathleen E. Kunzer*
American Chemistry Council
1300 Wilson Blvd.
Arlington, VA 22209
American Chemistry Council
1300 Wilson Blvd.
Arlington, VA 22209
I first became aware of the concept of hormesis, though not the term, in late 1993 - early 1994 when the Environmental Protection Agency (EPA) was proposing to add over 300 new chemicals to the list of chemicals subject to annual release reporting under the Toxic Release Inventory Program.1 Chemicals may be added to the list because they are known or reasonably anticipated to cause a variety of adverse health effects, including cancer, reproductive dysfunction, neurological disorders, etc. One of the substances being considered for addition was folic acid. At the same time, news stories appeared highlighting the importance of folic acid in the diets of pregnant women for fetal development. 2,3 How could both these things be true! In the end, EPA chose not to include folic acid on its list of toxic substances. Its reasoning was never a matter of public record, but the low-dose benefits of folic acid had to be a factor. It was the right outcome, but, unfortunately, not one that was dictated by the law.
It is difficult to argue with the public health benefit of taking into account all the public health or environmental benefits, as well as risks, posed by a substance when considering whether and to what degree to regulate that substance. It is compelled by the first principle of medicine: "first, do no harm." The consideration of hormesis is subsumed in the larger question of how agencies account for the potential adverse consequences of a proposed regulation often referred to as risk-risk analysis or countervailing risk. Passenger side air bags are the classic example, but others include the risk posed by asbestos removal or the risk posed by removing a product from the market either from loss of the benefits of the product or from risks posed by substitutes. There is no requirement now that an agency consider the risks posed by a certain regulation and no legal imperative to first do no harm in accounting for hormetic effects when setting standards.
Professor Cross correctly notes that hormesis does not fit easily under current law and regulatory practice. In technology based standards, this is clearly the case. Where standards are being set based on health effects outcomes, the opportunity is there to consider hormetic effects in setting the standard, but it is rarely done. The good news is that developing case law may mean agencies will no longer be able to ignore evidence of hormesis.
Professor Cross proposes two different and interesting approaches for incorporating hormesis into standard setting. Both would require a restructuring of our environmental regulatory system. The first is to consider hormesis in determining the net-benefits of a regulation to set a standard that is determined by a benefit cost analysis.
In an ideal world, regulators would always try to maximize the net benefits achieved by a regulation. To do so, an agency would weigh the costs against the benefits and attempt to optimize the benefits for the costs expended. The agency would evaluate the costs and benefits (both quantifiable and non-quantifiable) and take into account the potential adverse consequences of a regulation. Under such a scheme, hormesis would be considered as either an offset in establishing the benefits to be achieved or as a cost if the regulation could impair health by reducing levels of a chemical so low that they were below the optimal level to provide protection.
In the real world, even where agencies are now allowed to take costs into account, no determination of relative costs and benefits is required. True, the current Regulatory Review Executive Order, E.O. 12866, requires for major rules ($100,000,000 or more in annual costs) that all agencies evaluate benefits and costs and determine whether the benefits justify the costs.4 That is not a mandatory requirement, however, because it is not one that can be compelled by a court.
Attempts to change regulatory standard setting practices to require this type of cost benefit balancing have been unsuccessful. Legislation proposed during the 104th Congress in 1995 would have made it mandatory for an agency to determine the net benefits and net costs and set a standard that either was the most cost effective or maximized the net benefits.5 This legislation was roundly rejected and is still described with horror as the attempt to "roll back environmental protection." In subsequent Congresses, this balancing requirement was softened to require on that the agency be able to state that the benefits of a regulation justified its costs.6,7 This legislation, too, has been unacceptable. Thus, the likelihood of achieving a regulatory system that tries to incorporate hormesis as part of a rational decisional criteria, weighing benefits and costs, is very slim.
The second of Professor Cross' approaches for more systematic consideration of hormesis in regulation is using a "risk cup." That approach offers promise, particularly with respect to its reliance on market-based mechanisms achieving risk reduction goals. Market mechanisms can provide cost-effective environmental protection. There are, however, potential problems with a risk cup approach. It is complex. For example, it needs to be based on realistic estimates of risk instead of a series of conservative assumptions, and it must account for back ground risk and risk from non-industrial human activity. Also further explanation is needed on how hormesis would be considered more effectively in a risk cup than under current health based standards.
While a risk-cup approach may have appeal as another objective and rational approach to environmental protection, it is probably both a practical and political impossibility at this time. We do not have the technical and scientific sophistication necessary to allocate risk fairly among industry sectors or companies. Moreover, politically, unless we can figure out how to address potentially disparate impacts of risk in a market-based risk cup approach, the of opposition would be substantial.
In the end, we may not have to wait for new regulatory schemes to begin regularly factoring in hormesis to standard setting. It may be compelled under a recent decision by the U.S. Court of Appeals of the District of Columbia Circuit, American Trucking Association, et al. v. EPA.8 In its 1999 decision remanding the National Ambient Air Quality Standards on particulate matter and ozone, the Court of Appeals held that in setting health based standards, the agency had to consider both the beneficial and the adverse health effects of substances. It was not strictly hormetic effects, as characterized by Professor Cross, that were at issue in that case, but the situation is plainly analogous.
When the Environmental Protection Agency was in the process of promulgating the standard for ambient ozone, it was presented with information about benefits of tropospheric ozone in preventing harmful effects of the sun's ultraviolet rays cataracts, melanomas, and other skin cancers. The Agency disregarded these benefits in setting its standard. Without commenting on the scientific validity of those purported benefits, the court stated that "(l)egally, then, EPA must consider positive identifiable effects of a pollutant's presence ."7 The court went on to state that the Agency cannot impose a higher information threshold for beneficial effects than for adverse effects.7 This aspect of the court's decision was specifically upheld on rehearing.8,9
Of course, the court's decision turned in substantial part on EPA's interpretation of the Clean Air Act. Pointing to references to the term "adverse effects" in the Act, EPA , argued that it did not need to consider beneficial effects. The court refused to defer to the agency's interpretation of the statute, citing other language to consider "all identifiable effects on public health and welfare."10 The court concluded the statute unambiguously contemplated consideration of all effects. Reaching beyond this conclusion, however, the court noted that, even if the language were ambiguous, it would reject the agency's interpretation as unreasonable: "it seems bizarre that a statute intended to improve human health would, as EPA claimed at argument, lock the agency into looking at only one half of a substance's health effects in determining the maximum level for that substance."8 Whether this decision has precedence beyond the specific provisions of Clean Air Act at issue remains to be seen. What is important, though, the court opens the door for future challenges to require an agency to consider evidence of hormesis when setting a standard.
The true obstacle to fairly considering
hormetic effects in setting regulatory standards will be the
difficulty of proving their existence and the dose response
relationship for hormesis. This difficulty is illustrated by a
case cited by Professor Cross, the decision by EPA in setting
the MCGL for chloroform. As pointed out,
notwithstanding considerable evidence to the contrary and the advice
of the Scientific Advisory Board, EPA refused to depart
from it linear, no threshold default assumption for cancer.
The Court of Appeals for the D.C. Circuit vacated the
rule, finding setting of standard zero to be arbitrary
and capricious.11 Notwithstanding this decision, the
institutional barriers remain to accepting evidence
regarding effects at very low levels. Hormesis often occurs at
those low levels where effects are difficult or impossible
to measure through traditional testing. So, it too will
likely suffer a similar fate.
1. Emergency Planning and Community Right-To-Know Act, 42 U.S.C. §11023. Chemicals may be added to the list because the are known or reasonably anticipated to cause a variety of adverse health effects, including cancer, reproductive dysfunction, neurological disorders, etc.
2. Carole Sugarman, Folic Acid Crucial for Fetal Development, Washington Post, Nov. 9, 1993, at Z 20.
3. John Schwartz, To Reduce Birth Defects, FDA Requires Folic Acid in "Enriched" Breads and Grains, Washington Post, Oct. 8, 1993, at A 2.
4. Executive Order 12866, Regulatory Planning and Review, September 30, 1993.
5. Comprehensive Regulatory Reform Act of 1995, S. 343, 104th Cong., 1st Sess. (1995).
6. Regulatory Improvement Act of 1998, S. 981, 105th Cong., 1st Sess. (1997).
7. Regulatory Improvement Act of 1999, S.746, 106th Cong., 1st Sess. (1999).
8. American Trucking Association v. EPA, 175 F.3d 1027 (D.C. Cir. 1999).
9. American Trucking Association v. EPA, 195 F.3d 4, 17 (D.C. Cir. 1999).
10. Clean Air Act, 42 U.S.C. § 7408(a)(2).
11. Chlorine Chemistry Council v. EPA, 206 F.3d 1286 (D.C. Cir. 2000).
* Counsel, American Chemistry Council. The opinions
expressed in the comment are solely those of the author and should
not be taken as reflecting the views of the American