Harvard University, Cambridge, MA 02138
The data at high dose levels
Books and major compilations
United Nations Scientific Committee on the Effects of Atomic Radiation
Biological Effects of Ionizing Radiations (BEIR 1972)
NCRPM, Bethesda, MD
moderate to high doses
The Effects of Dose Rate
Low and Very Low Doses
Variation of Cancer Incidence with Background Exposure
Epidemiological studies of Low Dose Behavior.
Supra Linearity or Hormesis
This Resource Letter provides a guide to the literature on the effects of ionizing radiation on people at low doses. Journal articles, books and web pages are provided for the following: data at high dose levels, effects of moderate to high doses, (leukemia, solid cancer, lung cancer, childhood cancer and non-cancer outcomes), effects of dose rate, relationship to background, supra linearity and homesis, and policy implications.
That ionizing radiation can have serious adverse effects on people was obvious to the first experimenters 100 years ago. The physicist Geoffrey Crowther produced the first theory of radiation-induced cancer of which this author is aware. The idea is that when a cell was ionized by radiation it would initiate a cancer. The probability of ionizing a cell in a given time is clearly proportional to the radiation intensity and hence one gets a theory that cancer induction is linear with dose even at low doses. But this theory in its simplest form cannot be true. Cosmic rays and background radiation ionize millions of cells every day and yet lifetime cancer incidence is only about 30% in the US population. Other effects must modify this idea drastically. Cells can be repaired; they can be excreted without leading to a cancer, and so on. Whether cancer incidence is linear with dose depends therefore on whether these important mechanisms are constant with dose or otherwise. A discussion of these becomes very complex, and very rapidly a matter of opinion rather than scientific truth. Accordingly this Resource Letter will mostly address what we can discern from direct epidemiological measurements upon people. There are a large number of references to data on the effects of radiation at high doses. These are mostly in books and compilations (refs. 1-26).
The data at high dose levels
I Journals are listed in the following order. General journals, Cancer and Environment journals, and Radiation specific journals. Many of the most important results are published in the general journals, but detail is usually found in the radiation-specific journals:
American Journal of Epidemiology
Journal of the American Medical Association
British Medical Journal
Cancer and Environment
New England Journal of Medicine
Science of the Total Environment
Journal of the National Cancer Institute
Journal of Radioanalytical and Nuclear Chemistry
Applied Radiation and Isotopes
International Journal of Radiation Biology
Journal of Environmental Radioactivity
Radiation and Risk (Obninsk, Russia)
Radiation and Environmental Biophysics
II Books and major compilations
Of the six books listed, the first is intended for physicians. Nonetheless, there is a lot of physics therein, and all can be read by physicists with great profit. The second is the classic text and the third a more recent text on radiation protection.
1. Medical Effects of Ionizing Radiation,
edited by F.A.Mettler and A. C. Upton, 2nd edition. (W.B. Saunders, Philadelphia,
2. Principles of Radiation Protection, K.Z. Morgan and J.E. Turner, (Wiley, New York, 1967). (E)
3. Radiation Protection: a Guide for Scientists and Physicians, J. Shapiro, 3rd edition (Harvard University Press, Cambridge, MA, 1990). (E)
4. Radiation Carcinogenesis: Epidemiology and Biological Significance, J.D. Boice, Jr. and J.F. Fraumeni, Jr. (Raven Press, New York, NY, 1984). (I)
5. Health Effects of Low-Level Radiation, S. Kondo. (Kinki University Press, Osaka, Japan, 1993). (E)
6. Health Effects of Exposure to Low-Level Radiation, edited by W.R. Hendee and F.M. Edwards. (Institute of Physics Publishing, Bristol, UK, 1996). (E)
The reports of the United Nations Scientific Committee on the Effects of Atomic Radiation (abbreviated and pronounced UNSCEAR) are voluminous. They include reports on exposures from many countries and a summary of much of the scientific literature. Although in general the reader should look at the latest report first, they are not completely repetitive and earlier volumes contain some information not present in the later ones. In addition, a study of the changes helps the reader to follow the changes in scientific understanding.
7. Sources and Effects
of Ionizing Radiation, Report of the United Nations Scientific Committee
on the Effects of Atomic Radiation. United Nations, General Assembly Official
Records: 13th Session, Suppl. 17 (A/3838), (UNSCEAR, 1958). (I)
8. Effects of Atomic Radiation, United Nations Scientific Committee on the Sources and Effects of Ionizing Radiation, Report E. 77. (UNSCEAR, 1977). (I)
9. Atomic Radiation Sources and Biological Effects of Ionizing Radiation, United Nations Scientific Committee on the Effects of Ionizing Radiation, Report to the General Assembly, United Nations, New York. (UNSCEAR, 1982). (I)
10. Genetic and Somatic Effects of Ionizing Radiation, United Nations Scientific Committee on the Effects of Atomic Radiation, Report to the General Assembly, with Annexes. United Nations, New York. (UNSCEAR, 1986). (I)
11. Sources, Effects, and Risks of Ionizing Radiation, , United Nations Scientific Committee on the Effects of Ionizing Radiation, Report to the General Assembly, United Nations, New York. (UNSCEAR, 1988). (I)
12. Sources and Effects of Ionizing Radiation, United Nations Scientific Committee on the Effects of Atomic Radiation, Report to the General Assembly, including annexes, United Nations, New York. (UNSCEAR, 1993). (I)
13. Sources and Effects of Ionizing Radiation, United Nations Scientific Committee on the Effects of Atomic Radiation, Report to the General Assembly, with scientific annexes, United Nations Sales Publication E.94.1X.11, United Nations, New York. (UNSCEAR, 1994). (I)
The US National Academy of Sciences has a Committee on the Biological Effects of Ionizing Radiation (abbreviated and pronounced BEIR) that regularly surveys the literature on the effects of ionizing radiation. In contrast to the UNSCEAR reports, which are mainly a compilation of data, BEIR reports are judgmental.
14. The effects on populations
of exposure to low levels of ionizing radiation, Report of the Advisory
Committee on the Biological Effects of Ionizing Radiations (BEIR 1972),
(National Academy Press, Washington, D.C. 1972). (I)
15. The effects on populations of exposure to low levels of ionizing radiation (BEIR III 1980) (National Academy Press, Washington, D.C. 1980). (I)
16. Health Risks of Radon and Other Internally Deposited Alpha Emitters (BEIR IV) (National Academy Press, Washington, D.C. 1988). (I)
17. Health Effects of Exposure to Low Levels of Ionizing Radiation (BEIR V, 1990) ( National Academy Press, Washington, D.C. 1990). (I)
The National Council of Radiological Protection and Measurements (NCRPM) has produced over 100 reports. Most are too detailed to be of general interest, but I list the following particularly useful ones here.
18. Influence of Dose and
Its Distribution in Time on Dose-Relationships for Low-LET Radiation,
National Council on Radiation Protection and Measurements, Report No. 64
(NCRPM, Bethesda, MD, 1980). (I)
19. Evaluation of Occupational and Environmental Exposures to Radon and Radon Daughters in the United States, National Council on Radiation Protection and Measurements, Report No. 78, (NCRPM, Bethesda, MD, 1984). (I)
20. Induction of Thyroid Cancer by Ionizing Radiation, National Council on Radiation Protection and Measurements, Report No.80, (NCRPM, Bethesda, MD, 1985). (I)
21. Ionizing Radiation Exposure of the Population of the US, National Council on Radiation Protection and Measurements, Report No. 93 (NCRPM, Bethesda, MD, 1987). (E)
22. Exposure of the Population of the US and Canada from Natural Background Radiation, National Council on Radiation Protection and Measurements, Report 94 (NCRPM Bethesda, MD, 1987). (I)
23. Risk Estimates for Radiation Protection, National Council on Radiation Protection and Measurements, Report No. 115 (NCRPM), Bethesda, MD, 1994). (E)
24. Principles and Application of Collective Dose in Radiation Protection, National Council on Radiation Protection and Measurements, Report No. 121 (NCRPM, Bethesda, MD, 1995). (E)
International Commission on Radiological Protection (ICRP). Although no specific reports are referred to here, this commission, started in 1928, issues many reports.
III Conference Proceedings
The International Atomic Energy Agency, a UN agency set up to promote peaceful uses of nuclear energy, and more recently to monitor (and aid in controlling) military uses, regularly holds conferences and issues a number of reports. The conference on low doses of radiation is particularly important since it contains reports from a number of people with divergent views.
25. International Conference: One decade after
Chernobyl: Summing up the consequences of the accident, International
Atomic Energy Agency, Vienna (1996). (E)
26. International Conference: Low Doses of Ionizing Radiation: Biological Effects and Regulatory Control, International Atomic Energy Agency, Vienna, IAEA-TECDOC-976, IAEA-CN-67/63, 223-226 (1997). (I)
One crucial feature of the data is that radiation does not seem to cause ANY medical or biological effect that cannot also occur naturally. It merely increases the probability of the effect. This fact is very important both for understanding possible dose-response curves and for deciding what if anything to do about radiation at low doses. It also leads us to ask a general question. Does radiation ADD an additional probability of developing cancer (an absolute risk model) or does it MULTIPLY the probability that is there from natural or other causes (A RELATIVE RISK model). Although both are discussed in all the BEIR reports, it is noteworthy that the emphasis has changed from the ABSOLUTE risk model in BEIR I (1970) to the RELATIVE risk model in BEIR III and BEIR V.
These sites all have information on important reports and papers on radiation from RERF, IAEA, NRPB, ICRP and NCRPM respectively. Some of the more recent can be downloaded.
V What is the effect at moderate to high doses?
32. "Hazards of Ionising Radiation: 100 Years of
Observations on Man", R. Doll, Br. J. Cancer 72, 1339-1349 (1995).
(E) This very important review paper by the leading epidemiologist Sir
Richard Doll discusses the effects that one might expect from general biological
principles and the general status of the field. It is a good start to studying
the subject. He asks several questions:
(1) Does radiation exposure lead to cancer?
(2) Does radiation exposure lead to heart disease?
(3) Does radiation exposure lead to other diseases?
(4) Does radiation exposure lead to genetic anomalies passed to following generations?
(5) Does radiation exposure lead to birth defects?
Most of the studies address only (1) cancer, and the data do indeed suggest that cancer induction is the dominant effect. The data are better than for the other outcomes largely because the observed effects are greater. Several groups of radiation-induced cancer can be distinguished with different characteristics.
33. ."Studies of the Mortality of Atomic Bomb Survivors Report 12, Part I. Cancer:1950-1990," D.A. Pierce, Y. Shimizu, D.L. Preston, M. Vaeth, K. Mabuchi, Radiation Research 146, 1-27 (1996). (I)
Although there have been studies of the effects of radiation on people for 100 years, the most important studies are the studies of the consequences of the Hiroshima and Nagasaki atomic bombs. This study of the survivors has involved many good scientists, and considerable effort and expense. The exposures occurred over 50 years ago, but an increase in cancers over that expected in the general population is still occurring. Therefore the most recent of these papers are the important ones to read. In addition, the radiation dose to which the population was exposed is uncertain. It was derived from measurements at other (similar) explosions, and for the neutron dose by measuring long-lived radioactivity in the region.
Leukemia was the first cancer to be noticed in the atomic bomb survivors. Leukemias began to appear 5 years after the exposure and continued to appear for another 20 years, after which the number of leukemias (almost) ceased. Radiation induces leukemias more readily than other cancers. Therefore leukemias are often considered to be a "marker" for radiation effects. But the variation with age is clearly in great contrast to that of the "solid" cancers, and at old age even an "absolute risk" model would over predict the number.
B. Solid Cancers (other than lung)
Also in reference 33 are data on cancers in tissues of the body (other than those connected with blood) (hereinafter called solid cancers). These do not appear until 20 years after the exposure and have continued to appear for 50 years. The increase of cancer with radiation dose seems to follow a "Relative Risk" pattern whereby the risk, after the latent period of 20 years, is increased by a constant fraction relative to the background at that age.
C. Lung Cancers
Lung cancers have a special place because the lung is also the point of entry for a major exposure route. One of the surprises of the atomic age (i.e. since 1945) was that an excess of lung cancers appeared among uranium miners. In some of the early studies, they seemed to appear only among Caucasian miners and not among native Americans (Indians). There was some speculation of a racial difference in sensitivity. This speculation was rejected and the data somewhat resolved by further follow up, but also by the observation that there are fewer tobacco smokers among the Indian miners. Most analysts have assumed that the lung cancers are due to radon rather than to any other pollution in the mines (dust, etc.) although there is still room to question this. The main concern is that inhaled particles or radionuclides might cause lung cancer. For example radon gas in uranium mines produces lung cancers by alpha particle irradiation to the lung. This is reviewed in references 16 and 19.
34. "Lung Cancer Mortality Among U.S. Uranium Miners: a Reappraisal," A.S. Whittemore and A McMillan, J. Natl. Canc. Inst. 71, 489-499 (1983). (I) The work in this paper summarizes the evidence for the idea that radiation is synergistic with tobacco smoking, in that the effects MULTIPLY and do not merely ADD. This makes some intuitive sense since both smoke and radiation are highly irritant to the lung tissue. This idea can therefore be used to guide the questions which are asked of the data at low doses.
D. Childhood Cancers
35. "Radiation Dose Effects in Relation to Obstetric X rays and Childhood Cancers," A. Stewart and G.W. Kneale, Lancet, June 6, 1185-88 (1970). (E) Children in the age group 0-8 seem to develop leukemia naturally at a greater probability than children and adults a little older. It is usually accepted that these leukemias are caused by something that happened in utero. In the period 1940-1970 it was common to give an X ray to pregnant women to discover any problems with the infant fetus. This gave doses to the fetus approaching 1 Rem. This classic study (the Oxford study) showed that the probability of developing childhood leukemia increased with the number of X rays.
36. "Cancer Mortality Among Atomic Bomb Survivors
Exposed in utero or as Young Children (1950-1992)," R.R. DeLongchamps,
K. Mabuchi, Y. Yamamoto, D.L. Preston, Radiation Research 147, 385-395
37. "Risk of Childhood Cancer from Fetal Irradiation," R. Doll and R. Wakeford, Br. J. Radiol., 70: 130-139 (1997). (I)
38. "Childhood Leukemia in Gomel, Mogilev, Vitebsk, and Grodno Oblast (Regions) of Belarus Prior and After Chernobyl Disaster," E. P. Ivanov, G.V. Tolochko, L.P. Shuvaeva, V. S. Lazarev, M. A. Bogdasorova, A. V. Planko. (E)
The Chernobyl accident is sufficiently recent that only childhood cancers (leukemias, reference 38 and thyroid, reference 39) should have been seen so far. The lack of leukemias is consistent with a linear extrapolation from higher doses. But although the idea that X rays increase leukemia is generally accepted, it is still possible that causality works in the other direction, and that the reason for the X rays was a medical problem associated in some way with a latent leukemia. For that reason it is especially interesting to look at the pregnant women who were exposed at Hiroshima or at a radiation accident such as at Chernobyl. An effect is seen. It is not larger than and could be smaller than the effect suggested by Stewart and Kneale.
39. "Chernobyl-Related Thyroid Cancer in Children of Belarus: A Case-Control Study," L.N. Astakhova, L.R. Anspaugh, G.W. Beebe, A. Bouville, V.V. Drozdovitch, V. Garber, Y.I. Gavrilin, V.T. Khrouch, A.V. Kuvshinnikov, Y.N. Kuzmenkov, V.P. Minenko, K.V. Moschik, A.S. Nalivko, J. Robbins, E.V. Shemiakina, S. Shinkarev, S.I. Tochitskaya, M.A. Waclawiw, Rad. Research 150, 349-356 (1998). (I)
The thyroid cancers, while mostly non-fatal are numerous and were a surprise. This observation has been confirmed by others.
E. Medical Exposures
40. "Mortality from Breast Cancer After Irradiation
During Fluoroscopic Examination in Patients Being Treated for Tuberculosis",
A.B. Miller, G.R. Howe, G.J. Sherman, J.P. Lindsay, M.J. Yaffe, P.J. Dinner,
H.A. Risch, and D.L. Preston, N. Engl. J. Med. 321, 1285-1289 (1989).
41. "Radiation Induced Cancer and Leukemia Risk in Patients Treated for Cancer of the Cervix," J.D. Boice, M. Blettener, R.A. Kleneman et al. J. Nat. Cancer Inst. 79,1295-1311 (1987). (I)
It might be expected that people already being treated with radiation would be examined more carefully than others. This makes a study of such patients particularly interesting. The above studies are typical. They are often interpreted as consistent with a linear dose response relationship with a usual slope. But they can also be interpreted as showing evidence for a threshold.
F. The Other Outcomes
42. "Studies of the mortality of A bomb survivors: non cancer mortality based upon revised doses DS86," Y. Shimuzu, H.Kato, W.J. Schull and D.G. Hoel, Radiation Research 130, 249-266 (1992). (I)
These authors found an increase in several other medical outcomes as a result of the lower levels of exposure. In particular, heart disease appears with a frequency about one third of the frequency of cancer.
43. "The Children of Parents Exposed to Atomic Bombs: Estimates of the Genetic Doubling Doses of Radiation for Humans," J.V.Neel, W.J. Schull, A.A. Awa, C. Satoh, H. Kato, M. Otake, and Y. Yoshimoto, Amer. J. Human. Genet. 46, 1053-1072 (1990). (A)
Careful search has been made in the RERF data for genetic effects, which are so often portrayed in science fiction as dominant effects of radiation. But these are very small. Statistically significant effects are not observed.
44. The Growth and Development of Children Exposed in utero to the Atomic Bombs in Hiroshima and Nagasaki," J.W. Wood, R.J. Hoehn, S. Kawamoto, and K.G. Johnson". Amer. J. Public Health 57, 1374-1380 (1967). (A)
One statistically significant effect has been found in the RERF data on children whose parents were irradiated while they were in utero . Those whose parents were exposed within 1500 m of the hypocenter of the Hiroshima bomb were on average 2.25 cm shorter, 3 kg lighter and 1.1 cm smaller in head circumference than those exposed further away. This has not been seen in other data sets although there seems to have been no careful look.
45. "Threshold for Radiation-Related Severe Mental Retardation in Prenatally Exposed A-bomb Survivors: a Reanalysis," M. Otake, M.J. Schull, and Lees Brit. J. Radiation Biology 70(6), 755-763 (1996). (A)
Another effect that is perhaps associated with or a consequence of the reduction in head size is a statistically significant reduction in Intelligence Quotient (IQ) among children exposed in utero. In other cases there is severe mental retardation. Although these effects are generally believed to have a threshold, the data on reduction in IQ are consistent with a linear relationship between reduction and dose.
V. The Effects of Dose Rate
From general principles one might guess that a high radiation dose given at a low rate over a period of years might have a different (probably smaller) effect from the same dose given in a short time, although the very use of a total DOSE summed over a long time period, of the order of a lifetime implies that the difference is unlikely to be large. Data from animal exposures to shows that there is a reduction in cancers (for the same total dose) at low dose rates. A Dose Rate Reduction Factor (DRRF) is usually introduced to describe this. The following papers can be used to address this directly.
46. " Cancer Mortality Among Techa River Residents
and Their Offspring," M. M. Kossenko, Health Physics, 71, 77-82,
47. "Issues in the Comparison of Risk Estimates for the Population in the Techa River Region and Atomic Bomb Survivors," M. M. Kossenko, M. O. Degteva, O. V. Vyushkova, D.L. Preston, K. Mabuchi, and V.P. Kozheurov, Radiation Research, 148, 54-63 (1997). (I)
In 1955-56 radionuclides spilled from the reservoir in Lake Karachay into the Techa River. Villagers drank the water and ingested many radionuclides. For 40 years the health of 30,000 villagers around the Techa river has been studied. The exposures were mostly internal exposure from the bone seeker strontium 90. The doses can be moderately well determined by subsequent examination of radioactivity of teeth and other bones. This then enables a bone marrow dose to be determined, which is the appropriate organ dose for describing leukemia incidence. In contrast, the "solid" cancers depend upon external doses which are far less well determined. From the Techa River data it seems that the dose rate reduction factor for leukemia is about 3, but with a large error band (perhaps from 2 to 6). The dose rate reduction factor for solid cancers is about 1 with a much larger error band.
48. "Radiation Doses and Cancer," A. Shlyakhter and R Wilson, Nature 350, 25 (1991). (E)
For many years it had been rumored that the workers at the MAYAK atomic bomb plant in the Urals received large radiation doses. Attempts from the western countries to discover what they were fruitless until 1991 when a description was published in Russian in the journal "Priroda." In this paper, these data are discussed and show that the cancer rate was less than suggested by the Japanese atomic bomb data by a Dose Rate Reduction Factor of about 3.
49. "Verification of Occupational Doses at the
First Nuclear Plant in the Former Soviet Union," A. A. Romanyuka, D. Regulla,
E. K. Vasilenko, A. Wierser, E.G. Drozhko, A. F. Lyzlov, N. A. Koshurnikova,
N. S. Shilnikova, and A. P. Panfilov Appl. Radiat. Isot., 47, 11-12,
1277-1280 (1996). (A)
50. "Mortality Among Workers with Chronic Radiation Sickness," N. S. Shilnikova, N. A. Koshurnikova, M. G. Bolotnikova, N. R. Kabirova, V. V. Kreslov, A F. Lyzlov, and P. O. Okatenko., Health Physics 71(1), 86-89 (1996). (A)
The doses to the MAYAK workers are in principle well determined by personal monitors. Personal monitors even at that early date were reliable in the USA. In principle it should be possible to determine a dose rate reduction factor both for leukemia and for the principal solid cancers.
51. Lung Cancer in Radiochemical Industry Workers,"V.Hohryakov
and S. Romanov, The Science of the Total Environment, 142, 25-28,
Elsevier Science B.V. (1994). (A)
52. "Lung Cancer in Nuclear Workers of MAYAK," V.F. Khokhryakov, A.M. Kellerer, M. Kreisheimer and S.A. Romanov. Radiat. Environ. Biophys. 37, 11-17 (1998). (A)
An interesting subsidiary set of data comes from the MAYAK workers. This is because the workers were exposed to plutonium by inhalation and might be expected to develop lung cancer in the same way that uranium miners develop lung cancer from uranium. These are the ONLY workers exposed to plutonium (239 mostly) at doses high enough to have an appreciable incidence of lung cancer. While the first studies suggested that the dose-response relationship is quadratic with dose (in qualitative agreement with animal data) and therefore a low dose effect approaching zero seemed to make sense, a more careful look at the data suggests that a linear dose response relationship fits the data better. However the doses are NOT low and a threshold or reduced effect at low doses is possible.
53. "Estimated Long Term Health Effects of the Chernobyl Accident," E. Cardis, G. Anspaugh, V.K. Ivanov, et al., presented to the IAEA Conference: One Decade after Chernobyl: Summing up the Consequences of the Accident, International Atomic Energy Agency, Vienna (1996). (E)
VI Low and Very Low Doses
A. The Definition of Low Dose
The concept of what constitutes a low dose has been modified considerably over the last 50 years. In 1945 a typical chest X ray gave a dose of 1 Rem (0.01 Sv) and at least one jurisdiction (UK) went as far as to propose mandating such an X ray every year. (The bill died in the House of Lords because of the objections of the physicist Lord Cherwell.) In contrast, in 1987 a proposal of the US Nuclear Regulatory Commission to call a radiation exposure that gave no more than 1 milliRem (0.00001 Sv) to any person "Below Regulatory Concern" was withdrawn after some vocal public opposition. Yet natural background exposures are a few hundred milliRems or 100 times this amount. Thus "low dose" now means doses as low as, and usually well below background.
B. Variation of Cancer Incidence with Background Exposure
54. "Altitude, Radiation, and Mortality from Cancer and Heart Disease," C.R. Weinberg, K.G. Brown, and D.G. Hoel, Radiat. Res. 112, 381-390 (1987). (E)
One way of attempting to understand the effect of radiation on people at low doses is to understand the variation of cancer mortality with natural radiation exposure. In many studies cancer mortality seems to be lower in areas with high radiation dose.
55. "Natural Background Radiation and Cancer Death in Rocky Mountain States and Gulf Coast States", J. Jagger, Health Physics, 75(4), 428-430 (1998). (E)
The radiation levels in the Rocky Mountain states are higher than in the Gulf states, yet the cancer rate is lower. This effect may be seen throughout the U.S. and Canada (see reference 22), but confounding factors may exist. Many Mormons live in Utah and mountain states who do not smoke or drink alcohol or coffee and seem to have half the cancer rate of their non-Mormon neighbors. In New Jersey also there is much (presumably polluting) industry. Thus many analysts conclude that the only fact of importance from these studies of geographical variation is that radiation at these levels (a few hundred millRems per year of 10-20 Rems per lifetime) is NOT an important factor in developing human cancers compared to other factors.
56. "High Background Radiation Research in China", L. Wei, et al., Atomic Energy Press, Beijing, China (1996). (I) This reference discusses how high background radiation in China does not seem to lead to high cancer rates. It is unclear how much this is due to life style factors or other pollutant effects.
C. The Relationship of Low Dose Effects and Background
When doses were called LOW even when they were more than background dose, it was possible to discuss logically the effects of radiation independently of whatever causes the background. Now that LOW means radiation doses 100 times smaller than background ,it is necessary to consider them together. However, very few scientists and scientific papers do this logically.
57. "Fundamental Carcinogenic Processes and their
Implications for Low Dose Risk Assessment," K.S Crump, D.G. Hoel, C.H.
Langley and R.Peto, Cancer Research 36, 2973-2979 (1976). (E)
58. "Low-Dose Linearity: the Rule or the Exception?," M. Crawford and R. Wilson, Human and Ecological Risk Assessment, 2(2), 305-330 (1996). (E)
In the first of these papers, Crump et al. point out that whatever the basic biological process relating a dose to cancer, a differential linearity results provided that the radiation dose and the background act on the biological system in the same way. Since the cancers produced by radiation and those produced by background are indistinguishable, this is an assumption that has not been refuted - although of course it is an assumption whose validity must constantly be questioned. Crawford and Wilson went further and pointed out that the argument is a general one and can apply to other outcomes than cancer, such as respiratory problems caused by air pollution or cigarette smoking. This in our view makes it mandatory for any discussion of low dose behavior (meaning as is usual these days doses lower than background) to include a discussion of what causes the natural background of cancers. Unfortunately this is rarely done.
59. "Cytogenetic and Molecular Analysis of Therapy-Related Leukemia," J.D. Rowley and M.M. Le Beau, Ann. N.Y. Acad. Sci. 567, 130-140 (1989). (A)
There is a possibility that the cancers from radiation and background can be in principle distinguished, in which case the above argument would not apply. In this very fundamental set of measurements, Rowley and Le Beau have shown that the chromosome structure of an Acute Myelogenous Leukemia (AML) occurring subsequent to and presumably caused by radiotherapy was appreciably different from those that occur naturally. If this turns out to be a general result, the low dose extrapolation arguments of references 57 and 58 must be drastically reconsidered.
D. Epidemiological studies of Low Dose Behavior.
60. "Threshold Models in Radiation Carcinogenesis", D. G. Hoel and P. Li, Health Physics, 75(2):241-250, (1998). (I)
This paper addresses the few epidemiological studies with a large enough data sample, and with systematic errors well enough controlled, that can be used to discuss directly the shape of the dose response curve below a total dose of 50 rems (0.5 Sv). The paper focuses on the data on atomic bomb survivors.
61. "The Cancer Risk from Low-Level Radiation,"
B.L Cohen, Radiat. Res. 149, 525-526 (1998). (E)
62. "Response to the letter of Bernard L. Cohen,"D.A. Pierce, Y. Shimuzu, D.L. Preston, M. Vaeth and K. Mabuchi Radiat. Res. 149, 526-528 (1998). (I)
The earliest studies of the survivors of the atomic bombs in Hiroshima and Nagasaki suggested (with weak statistical significance since they are on the edge of the natural background) a threshold at a dose of about 20 Rems or even hormesis. The data of the most recent follow up in reference 33 are consistent with linearity down to a dose of 5 Rems. In reference 61 this analysis was questioned. In reference 62 it is pointed out that the recent analysis takes all the data together in a manner similar to a maximum likelihood analysis. It is well known that a maximum likelihood approach can pull out of the "noise" signals that are not clear when binned data are taken, and a Gaussian approximation used. It is, however, admitted that there may well be non-statistical errors that invalidate this conclusion.
63. "Job Factors, Radiation and Cancer Mortality
at Oak Ridge National Laboratory: Follow-Up Through 1984," S. Wing,
C. Shy, J. Wood, S. Wolf, D. Cragle, W. Tankersley, and E. L. Frome, Amer.
J. Indust. Med. 23, 265-279 (1993). (I)
64. "A Mortality Study of Employees of the Nuclear Industry in Oak Ridge, Tennessee," E.L. Frome, et al., Rad. Res. 148, 64-80 (1997). (I)
65. "Effects of Low Doses and Low Dose Rates of External Ionizing Radiation: Cancer Mortality Among Nuclear Industry Workers in Three Countries," E. Cardis, E.S. Gilbert, L. Carpenter, G. Howe, I. Kato, B.K. Armstrong, V. Beral, G. Cowper, A. Douglas, J. Fix, et al., Radiat. Res., 142, 117-132 (1995). (A)
There have been various studies of the cancer rates among workers in nuclear power plants and other nuclear facilities. The authors of reference 63 find an apparent increase with radiation about 10 times what was seen in the other studies. However an update in reference 64 continues the follow up for a longer period and the effect tends to disappear. Reference 65 is the result of an analysis where several occupational cohorts are combined. This is not merely a "meta-analysis" of several papers but a combined study of those groups where the data are deemed reliable. The additional number of leukemias seen is consistent with an extrapolation from the number at higher doses (but the lower 95th percentile of the number is close to zero). The number of additional "solid " cancers is close to zero but the upper 95th percentile is close to the linear extrapolation from higher doses. This suggests that larger numbers (such as those suggested by reference 63 or reference 75 (5 times that derived from the usual figures) are definitively excluded.
66. "Radon Levels in United States Homes by States
and Counties," B.L. Cohen, R.S. Shah, Health Physics 60, 243-259
67. "Relationship Between Exposure to Radon and Various Types of Cancer," B.L. Cohen, Health Phys. 65(5), 529-531 (1993). (E)
68. "Dose-response Relationship for Radiation Carcinogenesis in the Low-dose Region," B.L. Cohen, Int. Arch. Occup. Environ. Health 66, 71-75 (1994). (I)
69. "Test of the Linear-no Threshold Theory of Radiation Carcinogenesis for Inhaled Radon Decay Products," B.L.Cohen, Health Phys. 68, 157-174 (1995). (I)
70. "Problems in the Radon vs. Lung Cancer Test of the Linear No-Threshold Theory and a Procedure for Resolving Them. B.L. Cohen, Health Physics. 72, 623-628 (1997). (I)
Only recently have there been studies of the effects of radon on people in residential situations. There are two types of study. One, an "ecological" study, compares the AVERAGE lung cancer rate in a community with the AVERAGE radon concentration in the houses of that community. There are several early studies but the most important and most careful are in references 66-70. The average lung cancer rate falls with increasing radon concentration. It would be a logical non sequitur to derive directly from such a study the relationship of the probability of an INDIVIDUAL person succumbing to cancer with the radon concentration to which that individual is exposed (the dose - response relationship). To make such a conclusion is sometimes called "the ecological fallacy." However, Cohen argues that it is legitimate to compare ANY set of data with a theory and if the data do not fit, the theory must be wrong. In particular, he claims that the particular linear dose response relationship espoused by US EPA cannot be correct.
71. "Indoor Radon and Lung Cancer: Risky or Not?," J.M. Samet, J. Nat. Cancer Inst. 86, 1813-1814 (1994).
A distinguished epidemiologist challenges Cohen's studies and implicitly all other "ecological" studies. (E)
72. "Residential Radon Exposure and Lung Cancer
among Nonsmoking Women," M.C.R. Alavanja, R.C. Brownson, J.H. Lubin, J.
Chang, C. Berger, and J.D. Boice, Jr., J. Natl. Cancer Inst. 86,
1829-1837 (1994). (A)
73. "Lung Cancer Risk from Residential Radon: Meta-analysis of Eight Epidemiologic Studies," J.H. Lubin and J.D. Boice, Jr., J. Natl. Cancer Inst. 89, 49-57 (1997). (I)
These are "retrospective cohort" studies in which a group of people are followed and the individual doses estimated. These are free from the ecological fallacy but there are no data in the low dose region where 90% of Americans are exposed. It is important to realize that any conclusion about the risk at low doses (that is doses below natural background) derived from these studies is dependent upon an extrapolation, which may not be in direct disagreement with the ecological study of Cohen.
74. "Radon-exposed Underground Miners and Inverse Dose-rate (protraction enhancement) Effects," J.H. Lubin, J.D. Boice, Jr., C. Edling, R.W. Hornung, G. Howe, E. Kunz, R.A. Kusiak, H.I. Morrison, E.P. Radford, J.M. Samet, et al., Health Physics 69(4), 494-500 (1995). (A) This paper is a summary of the evidence that suggests that the mining cancer data underestimate the risk of uranium miners. Such an underestimate would be even harder to reconcile with the data of Cohen.
F. Supra Linearity or Hormesis
75. Radiation and Human Health, J.D. Gofman, (Sierra
Club Books, San Francisco, 1981). (I)
76. "Radiation Induced Cancer from Low-Dose Exposure: an Independent Analysis," J.D. Gofman (Committee on Nuclear Responsibility, San Francisco, 1990). (I)
This is the foremost and most logical of a set of claims that the effect of a low dose is 5 or more times the "establishment" wisdom. Although primarily concerned with radiation exposures from peaceful nuclear energy, Gofman is consistent in also pointing out high medical exposures although he is clearly less eager to oppose them.
77. "Is there a Large Risk of Radiation? A Critical Review of Pessimistic Claims," A. Shihab-Eldin. A.S. Shlyakhter and R. Wilson, Environmental International 18, 117-151 (1992). (E)
This paper discusses several claims that at radiation doses at or below the background cancers are produced. These reports often select data or otherwise fall into statistical "traps" or errors. Rarely (Gofman is an exception) is there a discussion of the effect of the background and why many more people are not dying naturally from cancer in high radiation areas, which would be expected if their claims were true.
78. "Cancer in Populations Living Near Nuclear Facilities: A Survey of Mortality Nationwide and Incidence in Two States," S. Jablon, Z. Hrubec, J. D. Boice, Jr., J. Amer. Med. Assoc. 265(11), 1403-1408 (1991). (E)
This paper directly responds to claims that living near nuclear power plants in ordinary operation gives high radiation doses.
79. "Health Effects of Low-dose Radiation: Molecular, Cellular, and Biosystem Response," M. Pollycove, and C.J. Paperiello, in: Low Doses of Ionizing Radiation: Biological Effects and Regulatory Control, International Atomic Energy Agency, Vienna, IAEA-TECDOC-976, IAEA-CN-67/63, 223-226 (1997). (A)
There is also a strong movement in the opposite direction, suggesting that radiation at low doses and low dose rates is good for you. This paper is typical of several papers in this conference report that address this proposition. In addition, this view is strongly supported in reference 5.
80. "Report of the National Institutes of Health Ad Hoc Working Group to Develop Radioepidemiological Tables", NIH Publication No. 85-2748. (A)
In 1985 Congress requested a set of tables to determine the probability that a person's cancer was due to his radiation exposure. These tables assume a linear dose response relationship. The dose for which the probability of causation is greater than 50% (and therefore compensable by ordinary legal rules) is very high and very few people will receive it.
81. Policy Statement of the Health Physics Society, 1313 Dolley Madison Boulevard, Suite 402, McLean, VA 22101. (1996). (E)
In accordance with current knowledge of radiation health risks, the Health Physics Society recommends against quantitative estimation of health risks below an individual dose of 5 Rem (the Rem is the unit of effective dose; in international units, 1 Rem = 0.01 sievert (Sv)) in one year or a lifetime dose of 10 rem in addition to background radiation. Risk estimation in this dose range should be strictly qualitative, accentuating a range of hypothetical health outcomes with an emphasis on the likely possibility of zero adverse health effects. The current philosophy of radiation protection is based on the assumption that any radiation dose, no matter how small, may result in human health effects, such as cancer and hereditary genetic damage. There is substantial and convincing scientific evidence for health risks at high dose. Below 10 Rem (which includes occupational and environmental exposures), risks of health effects are either too small to be observed or are non-existent.
82. "Implementation of the Principle of As Low As Reasonably Achievable (ALARA) for Medical and Dental Personnel, National Council on Radiation Protection and Measurements, Report No 107 (NCRPM, Bethesda MD, 1990). (E)
The policy recommendation of the International Commission
on Radiological protection (ICRP) dating back to 1928 is that it is prudent
to assume that a risk remains at low doses and that no exposure should
be accepted without expectation of some benefit. This has led to the principle
of ALARA - that doses should be reduced to As Low As Reasonably Achievable
(economic and other factors taken into account). This report outlines suggested
procedures and on page 25 suggests that if doses can be reduced at a cost
of $10 - $1,000 per man Rem or less ($1000 - $100,000 per person Sv) that
should be done. The Nuclear Regulatory Commission in 1975 had already suggested
a number at the high end - $1,000 per Man Rem - for nuclear activities
under their purview. Reference 82 also recommend a de minimis level
for an individual dose of 1 milliRem (100 m
Sv). Many scientists believe that their time and that of the public is
better spent in insisting that these guidelines be followed (and not exceeded)
with a coherent risk/cost/benefit analysis rather than addressing the possibility
of a threshold at low doses that may be impossible to prove rigorously.