Unmasking the truth: The science and policy of low-dose ionizing radiation

Abstract

There is scientific consensus on a prevailing hypothesis that, down to near-zero levels, the occurrence of future cancer is proportional to the dose of radiation received. Some experts and professional bodies in the field, however, subscribe to this linear no-threshold (LNT) model in scientific discussions but object to the use of the model for policy-related purposes. Given the large economic interests that are affected by policy decisions, this article recommends that experts and professional bodies avoid the intermingling of scientific and policy debates and acknowledge a logical implication of the LNT hypothesis: Low-dose radiation will sicken and kill a number of people over time.

We are all exposed to low doses of ionizing radiation. ¹ There is scientific debate about the health impacts of this radiation and policy debate about its management and regulation. Regrettably, the two debates are often intermingled.

Although scientific debate continues, there is consensus on a prevailing hypothesis regarding the relationship between low-dose radiation and cancer. For the sake of simplicity, consider low linear energy transfer radiation— especially x-rays and gamma rays—and solid cancers. Given that focus, there is consensus that the dose–effect relationship can be expressed mathematically as a linear no-threshold (LNT) model. In other words, across a range of doses down to near-zero levels, the occurrence of future cancer is proportional to the dose of radiation received.

Despite the scientific consensus, some experts and professional bodies in the field oppose the use of this model for policy-related purposes. Jeffry A. Siegel, president of Nuclear Physics Enterprises, and Michael G. Stabin, associate professor of radiological sciences at Vanderbilt University, are two such experts. They have summarized objections to the LNT hypothesis as follows: “Many have argued that its use has resulted in unjustified levels of public fear concerning low-level radiation exposure, unnecessarily large expenditures of (limited) public funds, and misconceptions with respect to the overall safety of nuclear power and medical uses of nuclear materials” (Siegel and Stabin, 2012).

Their statement is about policy rather than science. It focuses on supposed public fear, misconceptions, and public expenditures. Others have raised additional arguments about the uncertainty of the LNT model, but few have addressed yet another powerful factor in this debate: the potential for private profit.

The LNT hypothesis rests on both epidemiological and biophysical evidence (Brenner, 2009; National Research Council, 2006). Comparison of groups of adults who have and who have not been exposed supports an LNT model down to doses of about 100 millisieverts (mSv), and, for exposure of children in utero or in early life, this proportional relationship between radiation and cancer is seen at doses as low as 10 to 20 mSv. Biophysical evidence is also available at even lower levels, including the simple fact that some human cells are crossed by ionization tracks left behind by charged particles even at very low levels of radiation.

For contemporary policy purposes, the LNT hypothesis can be regarded as well-established science. ² Yet, the consensus around this hypothesis dissipates as one moves from the realm of science to the world of policy making. Experts and professional bodies who support the LNT hypothesis from a scientific perspective may reject some of its policy implications. This change of attitude is especially evident regarding one logical outcome of the LNT model: The model says that radiation-caused health effects will arise across a population even if individual doses are small, and these effects can be estimated through “collective dose.” The concept of collective dose means that individual doses are aggregated across the population, regardless of whether an individual’s exposure is a one-time event, episodic, or continuous.

Through such aggregation, a simple calculation emerges: In the United States, an average individual dose of 10 mSv would cause about 500 excess cancer deaths over time across a population of one million. ³ This statement can be extended to any level of low dose, or any large population, by simple proportion.

To be sure, these excess deaths would occur against a background of a much larger number of cancer deaths of different origin. ⁴ Yet, if the LNT model is valid, the excess deaths from radiation are real events; they are just masked by a greater number of other cancer deaths, so that they are not directly observable.

Some professional groups recoil from public acknowledgment that low-dose radiation causes cancer. For example, the International Commission on Radiological Protection (ICRP) supports the LNT hypothesis, describing it as “a prudent basis for the practical purposes of radiological protection.” Yet, the commission goes on to say that “it is not appropriate, for the purposes of public health planning, to calculate the hypothetical number of cases of cancer or hereditable disease that might be associated with very small radiation doses received by large numbers of people over very long periods of time” (ICRP, 2007: paragraphs 65 and 66). Thus, the ICRP rejects public discussion of real, but masked, health effects from radiation.

The commission seeks to justify its position by pointing to uncertainty underlying the LNT hypothesis (ICRP, 2007: paragraph 66). A similar argument is offered by the UN Scientific Committee on the Effects of Atomic Radiation (UNSCEAR), which has declined to estimate numbers of radiation-caused health effects attributable to the Chernobyl reactor accident “because of unacceptable uncertainties in the predictions” (UNSCEAR, 2011: paragraph 98).

UNSCEAR’s claim of uncertainty is somewhat more persuasive than is the ICRP’s argument, because in the Chernobyl case uncertainty encompasses the difficulty of calculating the collective dose from an accident that distributed radioactive material very unevenly over a huge area. Still, the failure to define “unacceptable” leaves considerable room for avoiding unpleasant calculations. The approaches of both the ICRP and UNSCEAR demonstrate how science and policy debates around low-dose radiation have become intermingled.

In practical situations where the LNT model and the concept of collective dose have policy and regulatory relevance, uncertainty-based arguments may be used to support a position that is also influenced by economic considerations. For example, the residual contamination that is permissible after the decommissioning of a nuclear facility, or the permissible release of radioactive material from a nuclear-fuel reprocessing plant, are not only regulatory issues involving scientific uncertainty about low-dose radiation. They are also issues in which large amounts of money are at stake and political pressure can be significant (Fairlie and Sumner, 2000).

A question for professional bodies is whether, in a politically pressurized environment, they will not only speak about the uncertainties of the LNT hypothesis, but will also acknowledge its logical implication: Even very low-dose radiation can be expected to sicken and kill a number of people over time.

Background to the low-dose radiation debate

An average resident of the United States receives an annual dose of about 3 mSv from natural sources and about 3 mSv from human-made sources. The natural dose is attributable to cosmic rays, radiation from the ground, ingestion of radioactive material, and—accounting for about two-thirds of the total—inhalation of radon. The average natural dose has held steady during the past three decades. By contrast, the average human-made dose has risen from about 0.5 mSv annually in the early 1980s to the present annual level of about 3 mSv. Most of this dose is attributable to medical radiation, of which about half is from computed tomography (CT) scans. A chest CT scan, for example, exposes the patient to about 15 mSv (Einstein, 2009).

Industrial events that release radioactive material into the environment can create additional exposure. For example, it has been estimated that the 1986 Chernobyl reactor accident created an average individual dose of about 6 mSv to residents of the European part of the Soviet Union and about 1 mSv to residents of Europe outside the Soviet Union (Energy Department, 1987). ⁵ These averages reflect wide variation of individual dose, in part because the deposition of radioactive material varied greatly across this region. The 530,000 workers in the Chernobyl recovery operation received much higher doses, an average of about 120 mSv each between 1986 and 1990 (UNSCEAR, 2011).

Public fear, economic considerations, and the role of experts

Those who intermingle scientific and policy arguments about low-dose radiation typically raise one or more of four issues: uncertainty about basic science, uncertainty about calculation of dose, economic considerations, and public fear about radiation. In depicting the public-fear argument, Siegel and Stabin, the two authors mentioned above who oppose the use of an LNT model for policy purposes, have stated, “[T]he idea that no radiation exposure is ‘safe’ generates significant public fear and in the event of a significant radiological event in the US, it may be all but impossible to effectively manage this fear by public education and risk communication (another use of LNT) unless started now” (Siegel and Stabin, 2012).

Public fear about natural and human-made hazards does not always accord with expert knowledge. At the same time, experts can appear oblivious to real and significant factors that influence public attitudes. This phenomenon is evident in the policy debate about low-dose radiation. Some experts seem to be unaware that economic considerations in this area can include private profit.

Exposure to human-made sources of radiation is closely associated with industrial activities that touch directly upon people’s sense of safety and agency. These industries are also profitable, politically powerful, and can appear indifferent to public concerns. It is, therefore, unsurprising that some members of the public distrust experts who appear to be associated with these industries. To illustrate, consider two industrial activities—medical imaging and nuclear power.

As cited above, medical radiation is responsible for a six-fold increase in the average, annual, human-made radiation dose to residents of the United States since the early 1980s. Medical imaging is the major driver of this increase. It is, therefore, significant that the US Food and Drug Administration (FDA) sees fit to warn the public about the marketing of CT scanning as a disease-screening procedure. The FDA’s website advises against such screening, saying that, although the scans are invaluable in diagnosing and guiding treatment of injured patients or those with symptoms of disease, they should not generally be used “as a preventive or proactive health care measure to healthy individuals who have no symptoms of disease” (FDA, 2012).

The FDA’s warning exemplifies a growing concern within the health professions that disease screening has been overemphasized, as well as a recognition that profit has influenced this trend. As a Dartmouth professor of medicine put it: “Think about it this way: In the past, you went to the doctor because you had a problem and you wanted to learn what to do about it. Now you go to the doctor because you want to stay well, and you learn instead that you have a problem. How did we get here? Or perhaps, more to the point: Who is to blame? One answer is the health care industry: By turning people into patients, screening makes a lot of money for pharmaceutical companies, hospitals and doctors” (Welch, 2012: A25).

In the case of nuclear power, public concern centers on the potential for an unplanned release of radioactive material, as a result of an accident or malevolent act. The March 2011 accident at the Fukushima Daiichi Nuclear Power Station in Japan demonstrated that this concern has a reasonable basis. At the same time, operation of nuclear power plants in the United States has become a highly profitable business. Given these facts, a US citizen could reasonably demand that the risk posed by nuclear power should be stringently regulated.

Yet, many observers believe that present regulation is deficient. Victor Gilinsky, a former commissioner of the US Nuclear Regulatory Commission (NRC), has summarized the policy situation in language suggesting that public fear about nuclear power is appropriate: “Unfortunately, US regulators have been overly accommodating to the industry they supervise. The Nuclear Regulatory Commission has been handing out 20-year extensions to plants, whose original licenses were for 40 years. This includes the country’s oldest operating plant, New Jersey’s Oyster Creek, which went into operation in 1969 and now holds a license to operate until 2029. These extensions tend to be granted after NRC reviews that are heavily weighted toward accepting the validity of past technical conclusions” (Gilinsky, 2011).

In light of such controversies, experts on the effects of low-dose radiation would be well-advised to make a clear distinction between scientific debate and policy debate. They should acknowledge the implications of science without regard for any conflict with political or economic considerations. Otherwise, they risk a loss of public faith in their profession, which could contribute to declining support for science in general.

Continued scientific debate about the LNT hypothesis is entirely appropriate, but participants in that debate should avoid introducing policy considerations. Participants should also be alert for evidence that the LNT model may underestimate risk, perhaps by neglecting indirect effects of radiation. ⁶

Within the policy realm, experts should not support the LNT hypothesis and then distance themselves from its logical implications. They should, therefore, recognize the existence of real, but masked, health effects at low radiation doses, albeit with some quantitative uncertainty. If they argue that a dose is difficult to calculate, they should carefully explain the sources of that difficulty. If they believe that uncertainty renders use of the LNT model unacceptable for policy purposes, they should set forth a standard of acceptability.

Public fear does not provide a reason to hide the logical implications of the LNT hypothesis. An attempt by experts to hide these implications is likely to be counterproductive. The truth would probably be revealed eventually, leading to diminished public faith in the relevant experts and in science in general. Ultimately, public fear could be exacerbated. Also, when experts consider public fear, they should account for contemporary views on individual agency. In past years, well-meaning doctors would often withhold a diagnosis of cancer to avoid alarming a patient. Now, such behavior is generally regarded as patronizing and obsolete.

When regulatory bodies develop limits on low-dose radiation, they will inevitably be obliged to balance the costs and benefits of regulation. As long as the LNT model remains the prevailing hypothesis, this balance should explicitly recognize the existence of real, but masked, health effects. Also, the balance should take a politically mature view of the financial incentives related to human-made radiation exposure. As shown above, exposure is often linked to industrial activities that are, in part, profit driven.

A contemporary policy approach to low-dose radiation should involve four major actions by professional bodies and scientific experts in the field. First, until the LNT hypothesis is unseated, accept its logical implications, which include the premature death of a number of citizens. Second, be explicit about uncertainty, recognizing that further research could show low-dose radiation to be either less or more dangerous than the LNT hypothesis indicates. Third, be transparent in dealing with the public. Fourth, ensure that cost–benefit analyses reflect nothing but science and the public interest. These actions could build a policy framework that enjoys broad public support.