- Jan Beyea Guest editor
Every time a release of radioactivity occurs, questions arise—not only about the true exposures, but also about the health risk at low doses. Predictably, debates unfold in the news media and galvanize social media networks. Sometimes these conversations enlighten the public, but often times they only exacerbate the confusion and fear about the significance and reality of exposure. Fukushima is the latest example of this warped communications strategy.
This special issue of the Bulletin examines what is new about the debate over radiation risk, specifically focusing on areas of agreement and disagreement, including quantitative estimates of cancer risk as a function of dose. In this issue, we don’t pretend to put the questions about the scientific jigsaw puzzle to rest, but we do hope to provide a sophisticated update for you, presented by people whose work has increased understanding within the field. For example, social scientist Paul Slovic updates his classic work on perception of radiation risk. Roger Kasperson, another social scientist, writes on the intriguing framework that he and colleagues developed about the social amplification of risk, which helps to explain public reactions to events like Fukushima and Chernobyl. By implication, Kasperson’s analysis raises a challenge for those who communicate risk information, whether professionally or informally. To provide information needed in a democracy, these communicators may amplify risks to the point where needless fear is generated, or they may attenuate the risks to a degree that desirable responses are avoided.
Today, the scientific and medical establishment of most countries (with the exception of France, where the public strongly supports nuclear power) accepts a default hypothesis on the effects of radiation at doses below the range where epidemiologic data are conclusive. This is the so-called linear non-threshold theory (LNT), which the review committee of the US Institute of Medicine and the National Academy of Scientists refers to in these words:
A comprehensive review of the biology data led the committee to conclude that the risk would continue in a linear fashion at lower doses without a threshold and that the smallest dose has the potential to cause a small increase in risk to humans. (National Research Council, 2006)
Radiation protection organizations, such as the International Commission on Radiation Protection, also use the LNT to justify minimizing future exposures; however, they have a tendency to focus on the uncertainties of the hypothesis and oppose its use to estimate consequences from releases such as Fukushima and Chernobyl—no doubt out of concern that such estimates may amplify the perception of risk. Whether or not avoiding predictions of low-dose consequences really attenuates risk perception or, in fact, amplifies it by increasing public suspicion about a cover-up is an interesting question. Technical and policy analyst Gordon Thompson, in his contribution to this issue of the Bulletin, discusses some aspects of this dilemma from the perspective of a scientist who often works with community groups.
Whatever the use of the LNT, the data from the one-time exposures of the Japanese atomic-bomb survivors provide most of the quantitative data on the linear slope, meaning the magnitude of the dose response. Epidemiologist David Richardson, whose work with these data has provided much new information about risks of low-dose radiation, writes on the history of this most important studied population, discussing its strengths and limitations. Radiobiologist Colin Hill examines some of the new biological research, particularly, on genomic instability, bystander effects, and adaptive response—effects that may lead to a better understanding of responses at very low doses and may help quantify any deviations from the LNT. An important question is whether or not any of the epidemiologic evidence has been interpreted properly. Answering no to that question is biostatistician Sander Greenland, who writes that misleading interpretations of low-dose epidemiologic data result in an underestimate of the full health impacts, because of failure to account for diseases with accelerated onsets.
Quantitative perspectives on risk at low doses have changed dramatically over the last 40 years, back when I first engaged in public debates on the subject. Has it made any difference outside political campaigns and in the culture wars? In particular, how do quantitative risk estimates affect rules and regulations? Terry Brock and Sami Sherbini from the US Nuclear Regulatory Commission examine the role that risk estimates of health effects play in regulating nuclear power in the United States.
In my own contribution to this special issue, I survey data, arguments, and debates surrounding low-level radiation risks. Historically—in the absence of human epidemiologic data—biologic arguments and cell data, fiercely debated, were used to convert risk estimates derived from the atomic-bomb data to protracted exposures. My article explores the new, large-scale epidemiologic studies that are directly relevant—not to one-time exposures received at Hiroshima and Nagasaki, but to the protracted exposures that are received from continuous decay of radioactive isotopes associated with releases from Fukushima or from the Soviet and US weapons complexes.
I also analyze contrasting data that suggest that dose responses might be higher or lower than predicted by the LNT. Some researchers believe that the dose response is higher than the LNT at low doses (supralinear response), while others maintain the dose response drops rapidly below the range covered by epidemiologic data (quasi-threshold); both groups can find some support in recent epidemiologic studies demonstrating the complexity of the scientific jigsaw puzzle that researchers face. There are other researchers who believe that the dose response turns around at some point as dose is decreased, actually reducing the risk of cancer (hormesis theory); this evidence can be found in data collected from home radon measurements correlated to county lung-cancer rates—albeit in contradiction to more standard epidemiologic studies of the same association, which do show the expected dose response.
If our efforts in this issue of the Bulletin are successful, the reader will be ready to join the debate armed with a broad-based view of the epidemiologic evidence and its differing interpretations, along with an awareness of the stakeholder and researcher landscape.