What follows is the continuation, in serial form, of a central chapter from my book A Primer in the Art of Deception: The Cult of Nuclearists, Uranium Weapons and Fraudulent Science.
SCAM NUMBER THIRTY-EIGHT: Design epidemiological studies in such a way as to guarantee that the results will underestimate the risk to health from radiation in the environment.
What knowledge exists about the medical effects of radiation on populations has been garnered from epidemiological studies. This research also has been used to validate the models of radiation effects upheld by the radiation protection agencies. Since nuclear/radiological weapons, commercial nuclear power and the biological impact of low-level radiation are such highly politicized subjects, it is not surprising that epidemiological studies are sometimes structured and implemented, either by intention or accident, to reflect the prejudices of the researchers conducting them. Biased studies pollute the knowledge base and are a propaganda device. Of interest here are the distortions of fact that can be insinuated into population studies in order to “prove” the correctness of ICRP models and thereby “verify” the minimal hazard predicted by the risk factors.
The ECRR has identified a number of common errors that have appeared in published epidemiological studies of radiation risk:
1) Wrong Doses: Many of the dosage scams mentioned previously have become incorporated into studies of contaminated populations. When dosages are assessed inaccurately, no meaningful conclusions can be drawn about risks. In most studies of contaminated populations, dosages are not actually measured in each member of the study group but estimated or, in the vernacular of radiation epidemiologists, “reconstructed”. This practice, based on numerous assumptions about the migration of radionuclides through the environment, is a simple means by which radiation studies can be subtly manipulated to deliver predetermined or politically acceptable results.
A common tactic used by the AEC during the period of aboveground weapon testing was to formulate dosages to the population from fallout in terms of external radiation. Conveniently, this served to downplay the level of exposure to those people living downwind by ignoring the additional dosages caused by internal emitters. This error of ignoring the cumulative effects of all the radioisotopes involved has compromised the Hiroshima Life Span Study, studies of populations living downwind of nuclear weapon detonations and studies of Chernobyl. As a variation on this theme, population studies are invariably based on the currently accepted models of external radiation. In instances where internal emitters are taken into account, the contribution made from external radiation and internal radiation are usually combined to derive a single dose estimate which is treated as if it were completely delivered externally. If internal emitters pose an enhanced hazard, as this book argues it does, no reliable conclusions about the hazard posed by radionuclides in the environment can be produced by this methodology.
2) Wrong Controls: To study radiation effects, the incidence of illness in an exposed population must be compared to that of a similar population that did not receive the exposure. If an inappropriate control group is selected for study, the health risks from radiation can be rendered woefully inaccurate. One way where this can occur is when the population chosen as the control group has been likewise exposed to radiation. When this occurs, the radiation-induced cancer rate in the study population will be made to appear lower than it actually is, perhaps even “nonsignificant,” due to the heightened incidence of radiation-induced cancer in the control group. This error is a central shortcoming of the Hiroshima Life Span study, where members of the control group received undetermined dosages from internal emitters. It has also crept into studies of the inhabitants of the Marshall Islands, of people living downwind of the Nevada Test Site and of populations contaminated by the fallout from Chernobyl.
It is important to note that the entire human population has been exposed, both internally and externally, to radiation from weapon-test fallout, accidents, and routine ventings from nuclear installations. There is no uncontaminated subpopulation that can serve as a control group to test the impact of this pollution on the health of the human species. Internal contamination by novel fission products befouling the environment may be more hazardous than Natural Background Radiation but its effect is masked by this universal contamination. The presence of these radionuclides in the environment might well explain the rise in the incidence of cancer across the population since the middle of the last century. This lack of a suitable control group has important implications for interpreting of epidemiological studies of radiation and cancer causation. When the incidence of cancer in a study group exposed to the effluent of a radiation accident is compared to that of members of the general population, the frequency of radiation-induced cancer will end up appearing less than it in fact is, and the risk of cancer from radiation will be underestimated.
The ECRR notes that it may be inappropriate to select members of the general population as a control group if the study group does not itself reflect the general population. One example is the “war survivor effect” prejudicing the Life Span Study. Those in Hiroshima who survived a long war, the atomic bombing and the hardships that followed this holocaust became members of the study group in a project designed to determine radiation effects in humans. However, the survivors of this horrific ordeal might not have been representative of either the Japanese population or of the entire human race. The hardships of living through the war followed by the extreme trauma of the bombing, and survival through the subsequent five years before the onset of the study may have preferentially selected individuals with stronger immune systems or genetic resistance to certain types of radiation effects. As a consequence, their incidence of cancer may have been atypically low compared to members of the population at large. Similarly, employees in the nuclear industry may manifest a “healthy worker effect” which lowers their rates of cancer in comparison with equivalent age groups within the general population. Fit, employable individuals undergoing regular medical exams who selected themselves to work in the nuclear industry and were then selected again for employment may not represent a true cross-section of the general population. Such comparisons may generate spurious results about the risks to health from radiation exposure.
3) Wrong Sample: It is not uncommon for groups that have been differentially exposed to radiation to be pooled together in a common study group. For instance, people living within a defined radius of a nuclear installation may be grouped together in an effort to detect the effect of living near the facility on rates of cancer. This method can mask the fact, for instance, that those people living downwind of the plant will have received higher doses than those living upwind. When the number of cancers recorded is compared to the size of the population under investigation, those living upwind will dilute the findings, lowering the cancer rate and lowering the apparent risk.
4) Wrong Assumptions: This entire chapter has been devoted to proving that the current model of radiation effects is biased toward underestimating the risk to health from exposure to radiation. The assumption built into this model can profoundly influence the interpretation of data collected in epidemiological research. The ECRR provides a clear example. In studies of nuclear workers and the effects of Chernobyl in Europe, “the assumption of a linear no-threshold dose response has resulted in many clear observations of effect being discounted because high-dose groups may have lower cancer rates than intermediate-dose groups.” According to the LNT hypothesis, those in a population who received the highest dosages should manifest the greatest degree of radiation-induced illness. If the greatest number of casualties do not exist in the high-dose group, then radiation is discounted as the cause of any detected cancer increase across the population. This line of reasoning is based on models originally designed to understand the effects of external radiation. It may not be valid for low-level internal exposure. For instance, a study might reveal that the greatest incidence of birth defects or childhood leukemia occurred in the intermediate-dose group rather than the high dose group. Does this finding justify the conclusion that radiation was not the cause? Certainly not! Perhaps a greater number of fetuses were spontaneously aborted in the group suffering the greater exposure, thus lowering the incidence of disease in the children of members of this cohort. The expectation of observing a linear dose response can thus blind researchers into discovering that low-level internal exposure may carry with it an enhanced risk for radiation injury. In line with this observation is another which deserves mentioning. When speaking of the effects of Chernobyl on the rates of cancer in Europe, the ECRR makes the following point:
“In addition, epidemiological studies have been influenced by or countered with the predictions of the ICRP risk models for populations exposed to the doses resulting from the discharges. These predict very modest effects which would generally be difficult to establish against the large background cancer rates experienced by the study populations and therefore, when increases in cancer are seen in such populations they are ignored or at least not ascribed to exposures from Chernobyl” .
Another incorrect assumption mentioned by the ECRR has been incorporated in studies of the rates of cancer in geographical areas of high Natural Background Radiation. To “prove” the harmlessness of low-level radiation, studies have been conducted comparing cancer rates between populations living amidst different levels of naturally occurring radiation. When the areas of high NBR are not found to demonstrate higher rates of cancer, the conclusion jumped to is that low-level radiation is not a hazard to health. One factor that is not taken into account is the selection over time of radiation resistance among members of a population exposed for generations to the elevated radiation in the environment. As noted by the ECRR: “Inducible radiation resistance has been demonstrated in animal studies yet no allowance has been made for this when comparing populations in Natural Background studies.”
As a third example, the ECRR notes that current models of radiation effects are based on the assumption that cancer is initiated directly from a single exposure event which induces genetic alterations. The genetic theory of cancer on which this is based fails to take into account other factors that may come into play in influencing the progression of a cancer. For instance, immune system stress, diet, or other environmental toxins may either aggravate or mitigate the effects of the initial aberration thereby affecting radiation-induced cancer rates within populations.
5) Wrong Conclusions: The ECRR notes that it is quite common for the conclusions drawn in an epidemiological study to be out of sync with the data collected during the course of the study. Journal abstracts or the conclusions appearing at the end of research papers claiming no observed effect between radiation exposure and cancer incidence have been observed to contradict the information included in tables or text within the body of the papers.
For a long and detailed examination of the types of errors that have corrupted important studies that purport to show that radiation exposure has little or no effect on public health, the reader is advised to consult chapter five, “Paradigm Deconstructed,” in Wings of Death.
 European Committee on Radiation Risk (ECRR). Recommendations of the European Committee on Radiation Risk: the Health Effects of Ionising Radiation Exposure at Low Doses for Radiation Protection Purposes. Regulators' Edition. Brussels; 2003. www.euradcom.org.