How Much Radiation Is Too Much?

An Interview with Edward Calabrese

Everyone knows that the dose is critical when you are taking a prescription medication: a small amount can provide significant benefit, but a large dose can kill you. This “non-linear” effect is taken for granted in pharmaceuticals, but is not generally adopted for regulating the risks of radiation. Dr. Edward Calabrese is a professor and toxicologist at the University of Massachusetts Amherst's Department of Environmental Health Sciences. He has spent his career studying non-linear effects in different carcinogens. From hundreds of studies, he has concluded that radiation should be treated more like pharmaceuticals, and regulators needs to change how they think about radiation risks and harm.

What’s the history of the linear-no-threshold, or LNT, framework, and how did it come to be the standard?

The rise of LNT theory was really the result of a political motivation by a group of radiation geneticists. I’m sure they believed that dose response was linear, but they also wanted to scare the hell out of society to increase their stature and grant funding for their research, and they wanted people to think they were the only ones who could save the world from the harms of atomic weapons testings, etc. It was a paternalistic behavior.

We know it’s the dose that makes the poison, but is it also the dose that makes the benefit? Or are some substances simply benign or have no effect at low doses?

It depends on how you define “low” and how you define “benign.” Essentially everything has a dose response relationship, meaning everything is harmful at high enough doses, but then we get into some complexity. A lot of things are essential, like minerals and vitamins, and it may well be that radiation is essential at low doses. Most agents are going to be toxic at high doses, but we’ve found that most agents have a biphasic relationship.

How do regulatory bodies like the Nuclear Regulatory Commission or International Atomic Energy Agency set dose limits for radiation?

They have a long history, and they make use of their long history. When you regulate carcinogens, you assume that the dose response in linear. If it’s a non-cancer endpoint, you assume there’s a threshold. If you’re dealing with radiation, you have much more stringent controls. The buy-in to linear dose response has a huge effect on industry. And this decision to go linear, which was made in the 1950s, was the most impactful environmental regulatory change in the modern world.

What’s the evidence that LNT is not the right model, can you give a concrete example?

From 1985 on, I began to study this and tried to take a look at alternative models, including the hormetic dose response model, which was marginalized and neglected. I found that this phenomenon occurs quite widely. The EPA’s default models, for example, are arbitrary. Numerous studies are cited in my publications that support the hormesis model. The reader simply has to read the literature and get educated. While this is not an easy exercise, it is necessary in order to understand the breadth and significance of the hormesis database and its impact on toxicology and risk assessment. Such an assessment will show that the hormesis phenomenon is far more common than previously recognized and outcompetes the LNT and threshold models in head-to-head competition. This is documented in the peer-reviewed literature.

How do you determine the threshold between beneficial and harmful?

You have to do this through study, and that’s what we’ve done. There’s a lot of historical clinical research that does exist which we have now summarized and evaluated in the biomedical literature. If people wanted to use this, the dosages are pretty well worked out. This country needs to overcome its often irrational fear of radiation to take a more informed approach to exposure and use the newly assessed information for clinical benefit.

What’s been the experience of others who had tried to replicate your results?

I have a long experience conducting my own lab experiments. But I have also done a large amount of research assimilating the work of others. My laboratory work is very replicable. In the general world of hormesis, these studies are more challenging than most. In the traditional work of toxicology, you’re looking for the dose that produces harm. In hormesis, the beneficial effects are very constrained. You have to have enough doses and replications of your findings to ensure that your results are reproducible and real.

If regulators shifted from relying on an LNT model to a hormetic dose model, how much would limits actually change?

Getting rid of regulations based on the LNT theory would result in an increase in the acceptable dosage by at least several hundred fold. And that would have a huge impact. This would have a positive effect on human health as well as save billions and billions and billions of dollars. The regulatory agencies are kind of a cult, but they don’t know they’re part of a cult.

One challenge with low dose radiation, is that certain radioisotopes accumulate in the food supply or in certain organs. If we had higher limits for radio-isotope releases, would we need to pair that with better monitoring of food or personal doses, making sure that individuals don’t exceed the threshold and go into high dose range?

Well, I think that that would be true. They’d have to look at every situation. The optimal dose is very close to the level where it crosses over into being harmful. And there are personal differences as well, what’s optimal for you might be different for another person. But it’s an important challenge, you have to solve the problem differently now. We can’t just keep saying “lower is always better.”

For vaccines we know the mechanism for how they work, what’s the mechanism for the benefits of low-dose radiation.

Our understandings of hormetic mechanisms in the last decade have really exploded. We’ve learned a lot about signaling pathways and so many different components for the mechanisms have been identified. I can’t say that the mechanisms are fully understood, but we can tell which receptor and cell signaling pathways are involved in mitigating or promoting a certain effect.

Finally, it seems like there has been some movement at NRC to revisit and maybe revise LNT, have you been involved in this effort and what’s your prognosis?

I submitted comments during the review process. I never thought I’d live to see the day where a regulatory agency would take our efforts seriously. I hoped they would, but I never imagined it would happen. They opened it up for four months of comments. I read all of them, most comments weren’t serious. About 60% of those that were serious were in support of the change. There’s little question that the science is on the hormesis side. You know, we use hormesis in the whole pharmaceutical industry, and if the NRC is taking any of these medical drugs they are already relying on hormetic science.