Is the EPA Changing the Regulatory Paradigm?


Sometimes it’s worth reading the fine print in obscure regulatory proposals. One such example is contained in a “proposed rulemaking” by the EPA on what are called “dose-response models.”

Buried in the Federal Register a few months back (on April 30) is this seemingly innocuous verbiage:

EPA should also incorporate the concept of model uncertainty when needed as a default to optimize dose risk estimation based on major competing models, including linear, threshold, U-shaped, J-shaped and bell-shaped models.

Your eyes glaze over, right?

Instead they should be popping out. EPA is proposing a major change in the way we regulate radiation, carcinogens, and toxic substances in our environment.

Since the 1950s, environmental regulations are largely based upon something called the “linearity-no threshold” (LNT) model, which holds, for example, that the first photon of ionizing radiation has the same probability of causing cancer as the bazillionth one.

The hypothetical mechanism is that each photon has an equal probability of zapping a single base pair in our DNA, and that can result in a breakage which may have the very remote possibility of inducing cancer.

The LNT is in fact fallout from, well, fallout—the radioactive nuclides that were the widely dispersed byproduct of atmospheric testing of nuclear weapons from the 1940s through the early 1960s. University of Massachusetts toxicologist Ed Calabrese has painstakingly built a remarkable literature showing that the LNT was largely the work of one man, a Nobel-prize winning mutation geneticist named Herman Muller. His work was classified by the old Atomic Energy Commission, and, amazingly, as shown by Calabrese, likely not peer reviewed.

His work was also wrong. Since Muller’s early work on x-ray induced mutations in fruit flies, it has since been discovered that DNA breaks all of the time, and that our cells carry their own repair kit. Cancer can occur when they are overwhelmed by large numbers of breakages.

Think of the sun’s radiation, which includes the ionizing ultraviolet wavelengths. The LNT model implies the first photon we experience can cause cancer. In reality there is clearly a threshold above which the cancer probability rises. Everyone is exposed to the sun, but only some populations are prone to basal cell skin cancers and they are in very sunny environments, which explains why it is so prevalent in Australia, and pretty much nonexistent in the planet’s northernmost populated areas.

In fact the LNT model isn’t just wrong—nature actually works opposite to it. Small amounts of exposure to things that are toxic in large amounts can actually be beneficial. Again, consider sunlight. It’s required to catalyze the final synthesis of Vitamin D. Absent Vitamin D, pretty terrible diseases, like Ricketts, with terrible disfigurement and sometimes fatal sequelae.

The alternative model is also largely the handiwork of Dr. Calabrese, which he calls the “biphasic dose-response,” or “hormetic” model. Note this is not homeopathy, which erroneously holds that the smaller the dose of something, the greater the therapeutic effect, which is nonsensical.

The biphasic response is so ubiquitous that it forms much of the basis for modern pharmacology. Small doses of things like certain snake venoms can reduce high blood pressure with all the resultant clinical benefits. Large doses are obviously fatal. In fact, the first Angiotensin Converting Enzyme inhibitor antihypertensives (referred to as ACE inhibitors) were derived from the venom of a Brazilian viper.

The obtuse verbiage quoted from the Federal Register means that EPA is proposing to, where appropriate (and that may be for a large number of instances), substitute other models, including the hormetic one, for the inappropriate LNT.

This rather momentous proposed change owes itself largely to one man—Ed Calabrese, who just happens to be a Cato adjunct scholar in the Center for the Study of Science. Calabrese has hundreds of papers in the scientific literature largely documenting the prevalence of hormesis and the rather shoddy way that the LNT was established and maintained, as well as a tremendous review article on it published last year in one of the Naturefamily of journals.

Used with permission. Cato Institute / CC BY-NC-SA 4.0.

Patrick J. Michaels is the director of the Center for the Study of Science at the Cato Institute. Michaels is a past president of the American Association of State Climatologists and was program chair for the Committee on Applied Climatology of the American Meteorological Society. He was a research professor of Environmental Sciences at University of Virginia for 30 years. Michaels was a contributing author and is a reviewer of the United Nations Intergovernmental Panel on Climate Change, which was awarded the Nobel Peace Prize in 2007. His writing has been published in the major scientific journals, including Climate ResearchClimatic Change,Geophysical Research LettersJournal of ClimateNature, and Science, as well as in popular serials worldwide. He is the author or editor of six books on climate and its impact, and he was an author of the climate “paper of the year” awarded by the Association of American Geographers in 2004. He has appeared on most of the worldwide major media. Michaels holds AB and SM degrees in biological sciences and plant ecology from the University of Chicago, and he received a PhD in ecological climatology from the University of Wisconsin at Madison in 1979.