November/December 1997 Vol. 53, No. 7 No dose too low By Ian Fairlie & Marvin Resnikoff Is there a safe level of radiation? Is there some standard below which exposure is harmless? Maybe even "good for you?" Years ago, many radiation scientists believed just that. Very low doses of radiation were harmless to human health, they said, and some argued that very small exposures actually had a positive effect. Over time, however, those views have lost favor, and they are now aired mainly by those who appear to have a pro-nuclear bias. Today, most of the world's radiation scientists believe that even very small doses of radiation- those that are well below background levels and whose effects are difficult to detect-increase the risk of developing cancer, however slightly. That view was officially endorsed in 1991, when the International Commission on Radiation Protection (icrp) adopted the "linearno-threshold" model, a theory that had been around, in various forms, for decades. A wide variety of studies-of Japanese A-bomb survivors, nuclear workers, and children who had been exposed with medical equipment-are often cited in support of the no-threshold theory, as well as recent studies that implicate "natural" or background radiation as the cause of many cancers. Although the no-threshold theory is now widely accepted by every major international organization concerned with radiation risks, it continues to be a matter of controversy because a significant number of radiation scientists, particularly in the United States and France, still hold to the earlier view that low doses are harmless. And the views of American scientists in particular remain influential in setting radiation standards in the U.S. nuclear industry. In part, the growing acceptance of the no-threshold theory reflects the latest understanding of biologists of how cancers originate. Their recent findings of radiation's effects on a cellular level raises new and serious questions about what level of radiation exposures should be permitted, both for workers in the nuclear industry and for the public at large. But tighter standards requiring lower emissions are likely to have serious cost implications for the nuclear industry, and they could have implications in lawsuits over radiation overdoses. The current controversy When the icrp, a voluntary international commission that makes recommendations on radiation dosimetry and limits, issued Publication 60, 1990 Recommendations of the icrp, calling for lower limits, it set off considerable controversy.1 The icrp proposed that radiation doses to the general population and to workers in the nuclear industry should be kept to very low levels below background. The proposal was significant, because most countries, excluding the United States and Russia, follow the commission's recommendations in setting national standards. Publication 60 recommended reducing radiation limits for workers from 50 milliSieverts (5 rem) to an average of 20 per year, and limiting permissible exposures to the general public from 5 milliSieverts to one. These standards have now been adopted by a variety of other organizations, including the U.N. Scientific Committee on the Effects of Atomic Radiation, the International Atomic Energy Agency, the World Health Organization, Britain's National Radiological Protection Board, and the European Union. In the United States, the National Radiation Protection Committee, the National Research Council's Committee on Biological Effects of Ionizing Radiation, and the Environmental Protection Agency have all endorsed the no-threshold theory, but the Nuclear Regulatory Commission has yet to adopt the icrp's recommended exposure limit. The current U.S. standard for workers remains at 50 milliSieverts per year. However, the Nuclear Regulatory Commission's standard for the public is already 1 milliSievert. Meanwhile, the lower limits-and the no-threshold theory behind them- have provoked a heated debate in some quarters-in professional associations, radiation and nuclear industry conferences, radiation journals, and on various e-mail listservers. The controversy, which has continued for more than six years, shows few signs of abating. French complaints Initial complaints about the recommendations in Publication 60 came mainly from radiation authorities in countries whose governments were strongly pro-nuclear-France, Canada, and Japan. France is committed to nuclear power, with about 75 percent of its electricity capacity derived from nuclear plants. The French government was particularly concerned about the effect that tighter limits would have on domestic uranium mining; however, a slump in world uranium prices has since shut that industry down. Then, in 1995, the French Academie des Sciences issued a report criticizing the new standards, consistent with the views of the powerful French nuclear industry and research establishment. The Academie's own report was heavily criticized, however, and the French government's Institute of Nuclear Safety and Protection sided instead with the icrp. Even more remarkable, the institute asked Britain's National Radiological Protection Board to review the Academie's report for the French parliament.2 Although the Academie opposed the conclusions of that review, the Academie itself refused to make recommendations one way or the other. As a result, the French government has now agreed to implement the recommendations in Publication 60. The United States In January 1996, the Health Physics Society, a professional association of radiation protection personnel and scientists working in the U.S. nuclear power industry, weapons laboratories, and nuclear research establishments, issued a statement opposing the no-threshold theory.3 The society asserted that no adverse effects had been observed in humans exposed to less than 100 milliSieverts, and further, that below such a dose, radiation risks were either nonexistent or too small to be observed. Estimating the risk of doses below 50 milliSieverts, the society said, was too speculative. The society also asserted that applying the no-threshold theory without accounting for biological mechanisms like "cellular repair" resulted in overstatements of risk. It was soon clear that many of the society's members disagreed. The society's May 1996 newsletter contained many dissents from that January statement. One member pointed out that the society's statement that low doses had no health effects was in conflict with the society's simultaneous recommendation that risk estimates not be used for low doses. Another member pointed out that the society's position sounded "more political than scientific." And a third correspondent predicted that the statement "will harm the credibility of the [Society] as a radiation protection organization." Meanwhile, the American nuclear waste industry also became concerned. In July 1996, the Nuclear Regulatory Commission was petitioned by its own advisory committee on nuclear waste, most of whose members are representatives of industry, warning that there would be significant "societal" costs if the no-threshold position was adopted.4 In addition, the petition argued strongly against the idea of "collective" or population-wide dose estimates when calculating risks. The scientific debate Epidemiological studies. Both sides agree that studies of Japanese bomb survivors show strong evidence of adverse effects at a dose of 200 milliSieverts for adult survivors and 100 milliSieverts in the case of children. Below that, the Health Physics Society says, there is an "inability to detect any increased health detriment." But most others disagree. They cite Alice Stewart's pioneering Oxford studies, which revealed that children whose in utero exposures were as little as 10 to 20 milliSieverts had 40 percent more childhood leukemias than those who were not exposed-a statistically significant increase in risk at low doses.5 Although Stewart's work, published in 1970, was long attacked by nuclear industry proponents, subsequent studies have supported its main findings and government radiation authorities now quote the Oxford studies without caveat.6 In addition to Stewart's work, a 1995 study that pooled the results of seven epidemiological investigations showed a statistically significant increase in the incidence of human thyroid cancer in groups that received doses of between 10 and 100 milliSieverts.7 Carcinogenesis. Perhaps more important than epidemiological studies, where there is always room for argument, are recent advances in understanding how cancers are triggered. Radiobiologists now agree that cancer can be initiated as a result of a single radiation track passing through a single cell nucleus.8 Most damage to cellular genetic material is corrected by repair enzymes. When cellular dna is misrepaired, however, it can result in a mutation that, years later, may develop into cancer. And if a single track through a single nucleus-the lowest possible radiation dose-can cause cancer, then any exposure to radiation is hazardous. Of course, the odds that a particular individual will develop cancer from a particular exposure are extremely low. Still, it means that one can no longer speak of a "safe" dose level. A more troubling problem arises if one calculates the chance that some members of a large population-not a particular individual-will develop cancer (calculated by adding all the individual risks). Background radiation. Everyone in the United States is exposed to 4 to 5 milliSieverts per year of natural, or background, radiation from a variety of sources-radon, cosmic rays, and the nuclides found in soil, diet, and in the body. Individuals may also be exposed to additional radiation through medical diagnostic and treatment programs. No-threshold opponents argue that because emissions from nuclear installations are already well below background levels, there is no reason to worry. But this argument has always assumed that natural radiation is safe, an assumption that is no longer tenable, given what we now know about carcinogenesis at the cellular level. Background radiation has already been calculated to cause some 45 percent of Britain's cancer deaths.9 Similar rates occur in the United States and other countries. The International Committee on Radiation Protection does not use background radiation as a criterion for acceptable radiological practices. Dna damage. The integrity of a cell's dna is constantly under assault, mostly by thermodynamic instabilities or attacks by chemical radicals. Because almost all of the resulting dna breaks are repaired with great fidelity, those who adhere to the idea of a safety threshold claim that a few more breaks pose an insignificant risk. But this claim ignores the nature of repair. Damage caused by heat instability or chemical radicals affects only one of the two strands that make up the dna; in repairing the break in a single strand, the other strand is used as a template. Radiation, however, is more likely to cause double-strand dna breaks, which have a greater risk of misrepair. "Adaptive response." Many nuclear proponents point to research findings that appear to show that cells given small doses of radiation suffer less damage when subsequently given larger doses than cells that received no preliminary dose.10 In other words, they say that cells benefit from low doses in some circumstances. They postulate that a low priming dose of radiation may increase the number of enzymes available to repair subsequent radiation damage. But this argument ignores the enormous capacity already available for repair, and the fact that it is fidelity of repair, not the number of repairs, that is critical. The U.N. Scientific Committee on the Effects of Atomic Radiation examined the notion of "adaptive response" in 1993, and concluded that while it was an interesting phenomenon that occurs in some cell systems at various stages of development, it had little relevance in radiation protection.11 The latest word In 1996, the latest mortality data from the continuing study of Japanese A-bomb survivors was published, revealing a statistically significant upward trend of risk with doses in the region of 50 milliSieverts.12 Warren Sinclair, president emeritus of the U.S. National Committee for Radiation Protection, concluded that these new results vindicated the position the icrp took in 1991.13 The results are another blow to the Health Physics Society's claim that there is no scientific proof of negative effects at exposures below 100 milliSieverts. The Academie des Sciences has concluded that these new results require it to reconsider its position on the no-threshold theory. Three important findings come out of the latest study of survivors: on effects at low exposures, on the shape of the dose-response curve, and on gender and age differences in response to radiation. The increased statistical power of the study, with five added years (to 1990) and 10,000 additional subjects, allowed the authors to divide the population that received lower exposures into three groups that received estimated doses of no more than 20, 50, or 100 milliSieverts. Excess cancers occurred in each group, and the authors conclude that the "data do not suggest the existence of a threshold below which there is no excess risk." Their results are also consistent both with Stewart's study on fetal exposures, and Thomas Mancuso's study of workers exposed at Hanford, both of which involved low exposures. Indeed, Stewart's study of pregnant women showed a linear relationship down to 10 milliSieverts.14 (It should be noted that all the subjects in these studies were exposed to similar rates of background radiation.) Although the latest survivor study suggests a straight dose-response curve, the authors also raise the possibility of a "superlinear" relationship-that is, that the incidence of some solid cancers increases more steeply at lower doses. The latest survivor study also found important cancer mortality differences by gender and age at the time of the bombing: Generally, risk dropped as age at exposure increased. Women were at twice the risk as men when exposed at the same age, and children under 10 were at greatest risk. Collective dose The icrp's adoption of a no-threshold model for radiation effects has stimulated renewed interest in calculating collective dose. Until recently, this concept was used mainly for collating occupational doses or doses to small populations living near nuclear facilities. Two reports in the mid-1980s provide theoretical underpinning to the concept of collective dose. Both reports recommended that, since radiation protection practices are concerned with the protection of the population as whole, potential collective as well as individual harm should be calculated.15 The icrp has also endorsed the use of collective dose: "Radiation detriment should be explicitly included," and "collective effective dose is an adequate representation of the collective detriment." The icrp is also concerned that collective dose be kept "as low as reasonably achievable," as the regulations of most European countries require. Death and money If the practice of calculating collective dose is adopted, there will be great controversies over attempts to translate from dose to detriment-that is, to calculate damages in terms of deaths or their money equivalents. For instance, one can predict that the collective doses from Britain's Sellafield fuel reprocessing plant will cause 200 excess cancer deaths worldwide each year (over all future time) for every year that Sellafield's emissions remain at the present level. Cost-benefit studies may use collective dose to calculate whether it is worthwhile to carry out remedial work or to end certain processes. For instance, if the icrp's per-Sievert risk factor and an estimated value of $3 million for each loss of life were used, the global cost of Sellafield's annual estimated emissions would be $615 million.16 Despite their uncertainty, the global effects of collective doses and their monetary costs should be calculated for discharges from military and civilian nuclear facilities, and they should be included in environmental impact statements for proposed nuclear facilities, nuclear waste dumps, and final repositories. Such calculations will undoubtedly raise questions about the environmental costs of proposed nuclear facilities. This concern is reflected in the opposition to collective dose measurements expressed by the Health Physics Society and the Nuclear Regulatory Commission's nuclear waste advisory committee. The inescapable problem for the Health Physics Society and the nuclear industry is that, as knowledge of radiation's effects continues to increase, it becomes more and more apparent that those effects must be taken more seriously than once thought. Over the past 100 years, exposure limits steadily have been tightened as more knowledge has been acquired about radiation effects. Whether they like it or not, global collective dose estimates will increasingly become a part of the language of both environmental agencies and pressure groups, and they will make their way into environmental impact statements and other assessments of nuclear operations. n 1. 1990 Recommendations of the icrp, icrp Publication No. 60, International Commission on Radiation Protection, 1991. 2. "Risk of Radiation Induced Cancer at Low Doses and Low Dose Rates for Radiation Protection Purposes," Documents of the nrpb, vol. 6, no. 1. 3. "Radiation Risk in Perspective, Position Statement of the Health Physics Society," Hps Newsletter, January 1996. 4. "The Health Effects of Low Levels of Ionizing Radiation," press release, Office of Public Affairs, Nuclear Regulatory Commission, 1996. 5. Alice M. Stewart and George W. Kneale, "Radiation Dose Effects in Relation to Obstetric X-rays and Childhood Cancers," Lancet, 1970, vol. 42, pp. 118588. 6. National Radiological Protection Board, "Risk of Radiation Induced Cancer at Low Doses and Low Dose Rates for Radiation Protection Purposes," Documents of the nrpb, vol. 6, no. 1. 7. E. Ron et al., "Thyroid Cancer after Exposure to External Radiation, a Pooled Analysis of Seven Studies," Radiation Research, 1995, vol. 141, pp. 25977. 8. J. Stather, C. Muirhead, and R. Cox, "Radiation Induced Cancer at Low Doses and Low Dose Rates," Radiation Protection Bulletin, July 1995, no. 167, pp. 812. 9. R. Edwards, "Natural Radiation May Kill Thousands," New Scientist, May 4, 1996, p. 4. 10. J. D. Shadley, V. Afzal, and S. Woolf, "Very Low Doses of X-rays Can Cause Human Lymphocytes to Become Less Susceptible to Ionizing Radiation," Mutagenesis, 1987, vol. 2, pp. 9597. 11. "Sources, Risks, and Effects of Ionizing Radiation," U.N. Scientific Committee on the Effects of Atomic Radiation (Vienna: United Nations, 1993). 12. D. A. Pierce et al., "Studies of the Mortality of Atomic Bomb Survivors: Report 12, Part I, Cancer: 19501990," Radiation Research, 1996, vol. 146, pp. 127. 13. "Lnt Model Said Vindicated by New Data," Nucleonics Week, November 14, 1996. 14. Thomas P. Mancuso, et al., "Radiation Exposures of Hanford Workers Dying from Cancer and Other Causes," Health Physics, 1977, vol. 13, pp. 36984; Stewart, "Radiation Dose Effects." 15. B. Lindell, "Concepts of Collective Dose in Radiological Protection," (Paris: Organization for Economic Cooperation and Development/Nuclear Energy Agency, 1984); Sfrahlen Sicherheit Kommission, "Safety Codes and Guides, Possibilities and Limits of the Application of Collective Dose, A Recommendation of the German Radiation Safety Commission, Edition 8," Gesellschaft fur Reaktorsicherheit, 1985. 16. For example, see "Estimating Externalities of the Nuclear Fuel Cycle," Oak Ridge National Laboratories and Resources for the Future (Washington, D.C.: McGraw-Hill, 1994); C.F. Guenther and C. Thein, "Estimated Cost of Person-Sv Exposure," Health Physics, 1997, vol. 72, pp. 204221. Ian Fairlie is a researcher at the Centre for Environmental Technology, Imperial College of Science, Technology, and Medicine, in London. Marvin Resnikoff is a senior associate at Radioactive Waste Management Associates in New York City.