So tell me, Doc – I was reading about some recent Executive Orders about nuclear energy and saw something in there about the NRC using radiation exposure models that aren’t based on science. Really? I thought that this stuff was well-known – is it actually a mystery? And if so, why don’t we know more? And how did we end up with the model we’re using, anyhow?
Well, for starters, you’ve asked about what might well be the single most contentious topic in radiation biology and radiation safety. People have been wondering about the exact effects of low doses of radiation for…well…over a century, the current model we use (more on this in a bit) was adopted about 70 years ago, and the arguments about the model have continued more or less unabated ever since. There are passionate adherents to two primary models – one holds that there is no level of radiation exposure below which there is no risk; the other postulates that there is a threshold below which the risk vanishes. There’s also a variant of the second model that postulates that, at doses below this threshold, exposure to radiation might actually be beneficial. But before getting into the details of any of these, let’s go over a little history.
Hermann Muller and DNA Mutagenesis
In the 19-teens Hermann Muller was curious about evolution – he accepted that genes were the basic unit of heredity (this was before we even knew what DNA actually did) and that it was minor changes in our genes (whatever those turned out to be) that drove evolution – what he didn’t know was how genes changed from generation to generation to drive evolutionary change. Speculating that natural radiation might be one of the drivers of spontaneous genetic changes (mutations), as he started graduate studies he began working on genetic experimentation with fruit flies, eventually demonstrating in the early 1920s that genetic changes could be induced using x-rays and radium. In the mid- and late-1920s Muller was able to show that radiation could induce lethal mutations and he almost immediately began warning the public about the possible dangers of working with and around radiation.
Muller was awarded the Nobel Prize in Physiology or Medicine in 1946 for discovering that radiation can induce mutations; in his Nobel lecture he discussed his concern that there might be no threshold below with radiation posed no risk, growing increasingly concerned about this as time progressed. In the aftermath of the nuclear weapons used against Japan and the first several years of atmospheric nuclear weapons testing, which included worries about the health effects of radioactive fallout. When the Soviet Union tested its first nuclear weapon in 1949 and the Cold War began to heat up, fears of global nuclear war gained momentum; learning about the health effects of radiation seemed an ever-more pressing concern.
In the early 1950s a study on low-dose radiation health effects concluded that, while there was no conclusive data on the health effects of very low doses of radiation, it made sense to assume that the roughly linear relationship seen at high doses and dose rates continued all the way to zero dose and to regulate radiation exposure accordingly until later studies provided more accurate information. Seventy years later we are still awaiting definitive studies. In the absence of such a study the overall consensus appears to be that it makes sense to use the LNT hypothesis because, if true, it produces the greatest calculated risk and, if incorrect, the actual risk will be lower than what was calculated – that it will be wrong in the “right” direction.
Other hypotheses
I’m not going to go into a great deal of detail about other hypotheses on our response to radiation exposure. One hypothesis is that, below a dose of about 10 rem or so, there are no effects, good or bad – this is (not surprisingly) called the “Threshold” hypothesis. And a variant of that called “Hormesis” suggests that low-level radiation exposure might actually reduce our risk of getting cancer by inducing the body’s DNA repair mechanisms to repair more damage than the radiation exposure is causing. The problem is that the epidemiological studies aren’t powerful enough to tell us which of these hypotheses is correct, although each of them is plausible.
Why the controversy might never end
Here’s the problem. It’s easy to see if a high dose of radiation will cause harm because the harm shows up in everybody and it shows up fairly quickly. If I’m trying to see if radiation exposure causes nausea and vomiting, I can find out by studying a relatively low number of people exposed – either in accidents or during, say, radiation therapy – and the effects will show up in a few weeks to a few months. But if I want to see if, say, a single whole-body CT scan will increase the risk of cancer then I’ll have to study 10 million people for a half-century or longer. That’s an experiment that cannot be performed, if only because of the ethics and the cost of exposing 10 million people to an unnecessary CT scan and the cost of studying such a large population for a lifetime. To find out what (if anything) happens at lower exposures requires more people and more money, as you can see in this graphic. There have been large-population studies performed, but they’ve all had their problems.
It’s interesting, for example, to note that locations in the US with higher levels of natural radiation exposure tend to have lower cancer incidence than areas with lower natural radiation exposures. The problem is that there are complicating factors from place to place – different levels of smoking, different exposures to other potential carcinogens, different diets, different income levels, different access to medical care, and so forth.
The recent Executive Orders
Why this is all relevant is that all of our radiation regulations – not just in the US, but globally – assume that LNT accurately describes the health effects of exposure to radiation at all levels of exposure. The thing is, it’s relatively easy and inexpensive to protect against high levels of radiation because they don’t happen very often – it’s a lot harder and much more expensive to protect against lower levels of radiation exposure because they’re much more common. Consider, for example, the number of x-ray machines compared to the number of high-activity sources at blood banks. Tens of millions of people are exposed to or work around x-ray machines and trying to reduce their radiation exposure even slightly would be difficult and expensive; the number of people working around irradiators is in the tens of thousands and protecting them is much easier.
These orders, however, are aimed at nuclear energy and what one of them notes is that “A myopic policy of minimizing even trivial risks ignores the reality that substitute forms of energy production also carry risk, such as pollution with potentially deleterious health effects,“ calling for the NRC to “reconsider reliance on the linear no-threshold (LNT) model for radiation exposure and the “as low as reasonably achievable” standard, which is predicated on LNT. Those models are flawed, as discussed in section 1 of this order. In reconsidering those limits, the NRC shall specifically consider adopting determinate radiation limits, and in doing so shall consult with the Department of Defense (DOD), the Department of Energy (DOE), and the Environmental Protection Agency.” Writing as a scientist with several decades of experience working in radiation safety, this makes sense.
There’s a lot more in these Executive Orders than what dose-response model will be used to regulate radiation safety in the US. But the model – our underlying assumptions about the hazards of exposure to low doses of radiation and when we start protecting ourselves from radiation – is one of the most fundamental decisions we need to make. Leaving out the politics and other controversies we’ve had of late, there is much in these Executive Orders that seems reasonable.