NetNews Usenet Archive 1992 #31

home *** CD-ROM | disk | FTP | other *** search

/ NetNews Usenet Archive 1992 #31 / NN_1992_31.iso / spool / talk / environm / 5273 < prev next >

Wrap

Internet Message Format | 1992-12-29 | 58.7 KB

Path: sparky!uunet!olivea!sgigate!odin!ratmandu.esd.sgi.com!dave From: dave@ratmandu.esd.sgi.com (dave "who can do? ratmandu!" ratcliffe) Newsgroups: talk.environment Subject: "Deadly Deceit," Afterword and Methodological Appendix Summary: an essential democratic tradition -- open access to sensitive info Keywords: measuring statistically significant excess death rate change Message-ID: <1992Dec29.152106.12472@odin.corp.sgi.com> Date: 29 Dec 92 15:21:06 GMT Sender: news@odin.corp.sgi.com (Net News) Organization: Silicon Graphics, Inc. Lines: 913 Nntp-Posting-Host: ratmandu.esd.sgi.com Unless "we the people" conduct our own research and learn for ourselves the actual state of public health in this technocratic, industrialized culture-- what the true risks are and who benefits from taking such risks--we will otherwise be lulled into permanent sleep by official myths serving interests other than those of the children and all life yet unborn. One of the many fission products created in the nuclear fuel cycle and in the detonation of nuclear bombs is strontium-90 which has a half life of almost 30 years. From the Methodological Appendix, below Dr. Gould writes: Strontium-90 is chemically similar to calcium and, therefore, concentrates in the bone of the developing infant, child, and adolescent. Once in the bone, strontium-90 irradiates the marrow where the cells of the immune system originate at a low rate over a period of many years. As first discovered by Dr. Stokke and his co- workers at the Oslo Cancer Hospital in 1968, extremely small doses of only ten to twenty millirads can produce visible damage to the blood forming cells of the bone marrow, probably via the production of free-radical oxygen.[209] This can lead to the development of bone cancer, leukemia and other malignant neoplasms both directly by damaging the genes, and indirectly by lowering the ability of the immune system to detect and destroy cancer cells.[210] Back in 1943 Robert Oppenheimer, Enrico Fermi and Edward Teller all knew that atomic radiation was a biological weapon and, if the bomb failed to work, they could still poison the German food supply with "radioactive fission products bred in a chain-reacting pile. . . . The radioactive isotope the men identified that `appears to offer the highest promise' was strontium, probably strontium 90, which the human body takes up in place of calcium and deposits dangerously and irretrievably in bone." [Richard Rhodes, "The Making of the Atomic Bomb," Simon and Schuster, 1986, p. 511] Excerpts follow from the Afterwood written by Dr. Gould in the first edition of "Deadly Deceit" appearing in the summer of 1990: . . . I believe there is now a great need for the expert testimony of statisticians in evaluating the significance of another ratio, which lies at the heart of toxic tort litigation. The key ratio to study when a population is exposed to an environmental risk is the "observed" number of deaths in a given area and period compared with the "expected" number based on national norms. We then seek to ascertain whether the ratio of the two numbers is too great to be attributed to chance. It is this measure of a statistically significant excess death rate change which lies at the heart of the demonstrations in this book of excess deaths following releases of low-level radiation. . . . After more detailed examination of our databases, I began to believe that most of the excess mortality I had found may have been associated with a few major accidental nuclear releases, especially the accident at Three Mile Island (TMI). Concerned Harrisburg residents urged us to present our TMI findings to the Senate Public Health Committee, with a plea for public debate. At this time I first learned that there were 2,500 lawsuits in progress against the local utility, of which 300 had already been settled on condition that no details of the settlement would ever be revealed. All this despite the claim that "no one died at TMI." . . . Why were birds so affected? Why had AIDS-related deaths in the U.S. doubled in May of 1986? Why were human immune systems so damaged by the Chernobyl fallout? All these were questions we wanted to answer. So we started to investigate the major nuclear releases in the past--from atmospheric nuclear bomb tests in the fifties and the sixties to civilian and military reactor accidents such as those at Savannah River, Millstone, and recurrent releases at TMI and Peach Bottom reactors. Beginning in 1945, the superpowers released massive quantities of fission products into the biosphere from above- and below- ground explosions of nuclear devices with yields of about 600,000 kilotons, according to estimates by the National Resources Defense Council. This was the equivalent of 40,000 Hiroshima-sized bombs. Emissions from nuclear reactors, including major accidents such as Three Mile Island and Chernobyl, added to the total, much of which was composed of long-lived radioactive isotopes that will remain in the stratosphere for millennia. . . . In this book we have drawn heavily on the databases created by Public Data Access, Inc., described in the methodological appendix. I regard these databases as our major achievement, for they represent a wonderful public resource, and a tribute to an important democratic tradition of open access to sensitive information. These databases enable us now to examine, in great detail, clusters of excess deaths from any cause, in any county or groups of counties, anywhere in the U.S., at any time since 1968. While we have focused attention in this book on the long neglected factor of low-level radiation, we fully recognize that all environmental abuses should be analyzed in the same spirit of open public inquiry. Even before the explosion of the first atomic bomb at Alamogordo--long before the 600,000 kilotons of short and long -lived fission products were released into the biosphere of our Earth--the creators of this technology understood its lethal effects as a devastating biological weapon. Yet this knowledge was assiduously covered up and denied to the public at large. It is still being denied today by the public relations arm of the nuclear industry through organizations like the U.S. Council for Energy Awareness which makes statements like "All the medical and scientific evidence gives nuclear power plants a clean bill of health. . . . the radiation from natural and man-made scources presents little or no risk. Scientific studies over several decades show no health effect on people from exposures below about 10,000 millirem." Whose interests are being served by the promulgation and perpetuation of such "knowledge"? We must all ponder deeply and consider this critical question of our times. Preventing the exposition and understanding of such crucial information reduces the general public to an infantile status and paves the way to oblivion. -- ratitor "It is not enough for a handful of experts to attempt the solution of a problem, to solve it and then to apply it. The restriction of knowledge to an elite group destroys the spirit of society and leads to its intellectual impoverishment." -- Albert Einstein the following is taken from the revised and updated softcover 1991 edition of "Deadly Deceit, Low-Level Radiation, High-Level Coverup" by Dr. Jay Gould and Benjamin A. Goldman with Kate Millpointer, published by Four Walls Eight Windows, New York, and reprinted here with the permission of Dr. Gould. ___________________________________________________________________________ Afterword By Jay M. Gould As I now see it, a curious twist of fate was responsible for my embarking on this controversial voyage of discovery, in such sharp contrast to my professional interests as an economist and statistician. I was retained by Westinghouse Electric Corporation in 1979 as an expert antitrust statistician in a suit involving the price-fixing of uranium by a so-called "international uranium cartel." My job was to ascertain whether forces of supply and demand could justify a six-fold rise in the price of uranium in the wake of the oil embargo of the early seventies. It became clear to me that uranium shortages could not have been a factor in the rise in the price of uranium. Because of frequent shutdowns, the performance of civilian nuclear reactors in the seventies fell far below expectations, and thus so did their demand for uranium. I was also surprised to find that the monthly reports of reactor operations in the nuclear trade press were more realistic than the contradictory statistics available on the consumption of uranium from the Atomic Energy Commission. My interest in nuclear problems began with that assignment, especially as I had been appointed to the Science Advisory Board of the Environmental Protection Agency in 1977 as a result of my discovery of a way to estimate regional concentrations of toxic wastes. This, in turn, led me to reflect on possible linkages of environmental problems with effects on public health. It was in this connection that I first read "Secret Fallout," by Dr. Ernest Sternglass, which I found to be most provocative and disturbing. Here was an eminent professor of radiology at the University of Pittsburgh Medical School, and a former senior physicist at Westinghouse Research Laboratories, who had become convinced, against his will, that the government was lying about the mortality impacts of an early accident at the Shippingport reactor near Pittsburgh. I felt there could be no question about his honesty as he described the process by which he was forced to assume the role of an anti-nuclear whistle blower. But I never dreamed that someday I, too, would be impelled, out of what was only a modest degree of intellectual curiosity, to confirm his findings and play a similar role. My career as an expert statistical witness began, quite by chance, when I was retained by the Department of Justice in 1955 to prepare statistical exhibits for a small antitrust suit. To the surprise of all concerned, the suit went all the way to the Supreme Court and became the famous Brown Shoe Case. In addition to establishing precedents for all postwar antitrust litigations, it established the role of a statistician as an expert witness in evaluating the significance of estimates of a company's "market share." As a result of this case, my career as a statistical expert blossomed. For the next two decades I participated in more than two dozen antitrust cases involving many major American companies, including IBM, Beatrice Foods, Greyhound, Armour, Occidental Petroleum, R. J. Reynolds, Emerson Electric, North American Phillips, and Westinghouse. Many of these companies have disappeared in the "merger mania" of the Reagan years so that antitrust litigation is today merely a historical footnote. While there may no longer be a demand in antitrust litigation for the analysis of market share ratios, I believe there is now a great need for the expert testimony of statisticians in evaluating the significance of another ratio, which lies at the heart of toxic tort litigation. The key ratio to study when a population is exposed to an environmental risk is the "observed" number of deaths in a given area and period compared with the "expected" number based on national norms. We then seek to ascertain whether the ratio of the two numbers is too great to be attributed to chance. It is this measure of a statistically significant excess death rate change which lies at the heart of the demonstrations in this book of excess deaths following releases of low-level radiation. My active antitrust practice led me to establish Economic Information Systems Inc., a company which successfully developed large computer databases for the analysis of the market shares of major companies in all industries. After the sale of my company to Control Data Corporation in 1981, 1 felt free to use my special database expertise to explore environmental problems. In the course of this I came across a remarkable book, "The Next Nuclear Gamble," by Marvin Resnikoff, published by the Council On Economic Priorities. From Dr. Resnikoff, I learned that highly radioactive used nuclear fuel assemblies were piling up in refrigerated pools at each reactor, awaiting the time, presumably early in the next century, when a "Great Nuclear Cemetery" would be built to receive them. Resnikoff estimated that the task of transporting the hundreds of millions of tons of radioactive materials away from the reactors was likely to involve an average of sixteen nuclear accidents each year, each of which could spew as much radioactivity into the environment as had the Three Mile Island catastrophe of 1979. The Resnikoff book was so impressive that I paid a visit to the Council on Economic Priorities (CEP), and accepted an invitation from CEP to continue my environmental research at the Council when my responsibilities to Control Data ended in 1984. My first research project at CEP affirmed the National Cancer Institute's 1977 findings that cancer mortality at the county level was correlated with concentrations of petrochemical activity. Reasoning that environmental pollutants, initially at least, were highly localized, and remembering that the 1980 Census had for the first time included the five-digit ZIP code area as a geographic unit, I made some calls to the Census Bureau and other federal agencies. To my surprise, I discovered that at very little cost (and with help from the Freedom of Information Act) I could purchase computer tapes from the Census Bureau, the EPA, the National Cancer Institute, and the National Center for Health Statistics, containing unpublished but politically sensitive information that had cost the government an estimated $40 billion to collect! This treasure trove of environmental and public health data had never been integrated and analyzed comprehensively. I can only speculate that epidemiologists, mainly employed by state departments of health, rarely investigated the causes of wide variations in local mortality rates, perhaps because of the political consequences of finding a correlation with some local environmental abuse. I believe most environmental epidemiological studies are self-limiting as a result. For someone with a database background like myself, these unused files were a researcher's paradise. It proved to be relatively easy to find statistically significant differences in geographic mortality rates because the official U.S. mortality databases were based on all death certificates filed in a given year. Even small differences might prove to be significant because of the large numbers involved. However, processing such large databases requires a professional staff of computer programmers and analysts. So, in 1985 I helped organize a small company called Public Data Access, Inc. (PDA) in the hope that it could eventually serve as the computer research arm of the environmental movement. It was in this context that I first worked with Ben Goldman, who at the time was a project director at CEP. He had discovered the same need to integrate diverse government data sources for a study of the hazardous waste industry, and had ventured into the world of micro-computers to accomplish the task. In fact, the database he developed on a personal computer for "Hazardous Waste Management: Reducing the Risk" (Island Press, 1986) ended up being PDA's first commercial product, available through a computerized information network called Chemical Information Services, Inc. Ben helped me start PDA, and eventually became its president. This fledgling effort received generous support from environmental foundations and concerned individuals. Even more important was the "sweat equity" contribution of a small group of enthusiastic young analysts, who in time became adept in interfacing large mainframe computers with the increasingly powerful and flexible personal computers, which among their other effects have made possible a decentralizing of data processing and great cost reductions. Since 1986, PDA has helped produce numerous environmental research studies. "Quality of Life in American Neighborhoods: Affluence, Toxic Waste and Cancer Mortality in Residential Zip Code Areas" (Westview Press, 1986), was a "database publication" with data for each of some 35,000 residential five-digit ZIP Code areas taken from EPA and Census Bureau databases. Although the 1980 Census cost about one billion dollars, the Reagan Administration decided that it would be too expensive to publish results for small areas such as ZIP codes, so this book remains the only public source for such localized information. PDA also produced in 1987 "Toxic Wastes and Race in the United States" for the Commission for Racial Justice, which showed that a significantly disproportionate number of toxic waste facilities were located in African-American neighborhoods. "The Philadelphia Toxics Story," for the National Campaign Against Toxic Hazards, and "Toxic Waste and Cancer Mortality in Michigan," for the Public Interest Research Group in Michigan, also were produced by PDA in 1987. In 1988, PDA completed "Mortality and Toxics Along the Mississippi" for Greenpeace USA. In all of these publications, the focus was on geographic areas with significantly high mortality rates and the extent to which these were correlated with exposures to toxic chemicals. Although frequently the correlations were good, in important cases they were poor. This apparent paradox was only resolved for me by discussions with Dr. Ernest Sternglass, who pointed out that a large culprit in the mortality situation might not be toxic chemicals, but low-level nuclear radiation. He argued that at least some part of the positive correlations we were obtaining between toxics and mortality were due to the overlap of nuclear pollution and toxic pollution in industrial areas, and also that where we were not getting a good correlation it might be because our studies did not contain a nuclear pollution variable. Dr. Sternglass also realized that our large mortality databases would overcome earlier criticisms of his research that had used small bodies of data to show that low-level radiation had a major impact on mortality. In response to his challenge, I decided to examine recent mortality trends in areas most exposed to nuclear emissions since 1975. In this work I did find small but statistically significant increases in total mortality, cancer mortality and infant mortality, for the period 1975-82 (compared with 1965-69) for "nuclear" as compared to "non-nuclear" areas. Nuclear areas were defined as the 160 counties which either had at least one commercial nuclear power plant or were downwind of such a county. After months of recalculation and checking, CEP published the nuclear county results in its December 1986 newsletter, entitled "Nuclear Emissions Take Their Toll." In the months that followed, I was disappointed with the complete lack of U.S. news coverage. However, it became a front-page story in Italy in January of 1987. Fabrizio Tonello, who reported on my findings for the Italian news weekly "Il Mondo," told me that his story played a significant role in generating the 80 percent referendum vote to halt work on Italy's two proposed nuclear reactors. Later that year the Italian cabinet accepted the popular decision, and Italy today has banned the construction of new nuclear reactors. European environmentalists were not the only people who were aware of my findings, however. I learned the CEP newsletter was making waves in Italy after I received a telephone request from the Italian Atomic Energy Commission for a copy. Only a few hours later, I received a call from a Philadelphia engineering firm involved in the construction of nuclear reactors. My caller seemed somewhat flustered at reaching me directly, as I had no secretary at the Council. I asked whether he wanted a copy of the report. "No," he replied, the question he wanted answered was, "who authorized this report?" In the U.S., various anti-nuclear groups throughout the country, disturbed by the CEP newsletter, clamored for detailed reports on each reactor. After more detailed examination of our databases, I began to believe that most of the excess mortality I had found may have been associated with a few major accidental nuclear releases, especially the accident at Three Mile Island (TMI). Concerned Harrisburg residents urged us to present our TMI findings to the Senate Public Health Committee, with a plea for public debate. At this time I first learned that there were 2,500 lawsuits in progress against the local utility, of which 300 had already been settled on condition that no details of the settlement would ever be revealed. All this despite the claim that "no one died at TMI." I presented our TMI findings to staff members of the Committee twice in the Spring of 1987. Early in 1988, Senator Edward Kennedy, Chair of the Public Health Committee, requested that the National Institutes of Health conduct a study of mortality near nuclear reactors. Senator Kennedy cited reports of high leukemia rates found by epidemiologists of the Harvard School of Public Health near the Pilgrim reactor in 1982-84, which were published in a letter to the British medical journal "The Lancet" on December 5,1987. "The New York Times" on July 7,1988, reported that the National Cancer Institute (NCI) had agreed to study cancer deaths among people living near nuclear plants. The "Times" quoted Dr. John Boice, chief of radiation epidemiology at NCI, who said "the study was prompted by a British survey completed last year . . . [that] found a higher incidence of leukemia among children and teenagers living near nuclear plants." We expect that this book may contradict the NCI findings, promised for 1990. Our hypothesis suggests that many more counties need to be studied than the number proposed by NCI. Dangerous fission products, particularly radioactive iodine and strontium, can be borne by winds and waterways for hundreds of miles and then come down in the rain, contaminating sources of fresh water and milk far from the reactor site itself. These ingested fission products can then wreak serious damage on immune systems. All of this was suggested by our Chernobyl findings. We began our Chernobyl study after an invitation to present our TMI findings at the European Conference on Chernobyl Radiation in Amsterdam at the end of May 1987. While there, I not only gained much anecdotal knowledge about the impact of Chernobyl radiation in Europe, but also learned that no European nation publishes monthly mortality reports similar to the "Monthly Vital Statistics Report" of the U.S. National Center for Health Statistics. I wondered whether mortality in the U.S. could have been affected by the small percentage of Chernobyl radiation that drifted over in the stratosphere and came down in the May 1986 rains. We began our research on this question when I returned to New York in June 1987, and were intrigued to find that abnormally high levels of radioactive iodine had been detected by Environmental Protection Agency milk-monitoring stations in almost every state beginning on about May 9th, 1986. We also found that significant mortality increases had occurred in May for which the Chernobyl fallout seemed to be the only plausible explanation. We revealed our Chernobyl findings in two papers delivered at the First Global Radiation Victims Conference held in New York City in September 1987. As we had come to expect, the U.S. press did not cover the story, but it was front-page news in the Japanese and Canadian press. This was followed by major stories in leading English papers such as the "Independent" and the "Economist." Finally, almost six months after we presented our findings, "The Wall Street Journal" broke the silence in the U.S. by reporting on our results on February 8, 1988. A few weeks after the publication of "The Wall Street Journal" story, I received a fascinating letter from Dr. David DeSante, a researcher at the Point Reyes Bird Observatory in California, enclosing an article he had published in an ornithological journal early in 1987. He had recorded a 62 percent drop in the number of newly hatched landbirds during the period from mid-May to mid-August, 1986. The only explanation he could find for the landbird reproductive failure was the radiation fallout from Chernobyl. Why were birds so affected? Why had AIDS-related deaths in the U.S. doubled in May of 1986? Why were human immune systems so damaged by the Chernobyl fallout? All these were questions we wanted to answer. So we started to investigate the major nuclear releases in the past--from atmospheric nuclear bomb tests in the fifties and the sixties to civilian and military reactor accidents such as those at Savannah River, Millstone, and recurrent releases at TMI and Peach Bottom reactors. Beginning in 1945, the superpowers released massive quantities of fission products into the biosphere from above- and below- ground explosions of nuclear devices with yields of about 600,000 kilotons, according to estimates by the National Resources Defense Council. This was the equivalent of 40,000 Hiroshima-sized bombs. Emissions from nuclear reactors, including major accidents such as Three Mile Island and Chernobyl, added to the total, much of which was composed of long-lived radioactive isotopes that will remain in the stratosphere for millennia. In the Fall of 1988, the Senate Government Operations Committee, under the leadership of Senator John Glenn, held hearings about a series of accidents and safety problems at military nuclear facilities operated by the Department of Energy. It turned out that crucial information about some of these accidents had been withheld from the public and the Congress for as long as 25 years. We immediately realized that our mortality database, with its ability to yield calculations of excess mortality at any place and at any particular time, could be brought to bear to illuminate the health consequences of these incidents. In late 1988, we joined efforts with the Commission for Racial Justice of the United Church of Christ (CRJ) in creating the Radiation and Public Health Project (RPHP). The project grew out of a long-standing relationship between PDA and CRJ. Since 1982, CRJ has investigated the presence of toxic substances in residential communities across the country, and has challenged the disproportionate impact on racial and ethnic neighborhoods. This pursuit lead CRJ to engage PDA in 1986 to prepare the ground-breaking study "Toxic Wastes and Race in the United States," which was the first comprehensive empirical study of race and toxics in the U.S., and which has been instrumental in influencing the Centers for Disease Control to undertake epidemiologic studies in this area. The function of RPHP is to extend our studies of the health effects of nuclear releases, and to stimulate public debate on these controversial issues. In this book we have drawn heavily on the databases created by Public Data Access, Inc., described in the methodological appendix. I regard these databases as our major achievement, for they represent a wonderful public resource, and a tribute to an important democratic tradition of open access to sensitive information. These databases enable us now to examine, in great detail, clusters of excess deaths from any cause, in any county or groups of counties, anywhere in the U.S., at any time since 1968. While we have focused attention in this book on the long neglected factor of low-level radiation, we fully recognize that all environmental abuses should be analyzed in the same spirit of open public inquiry. ____________________________________________________________________________ Methodological Appendix Statistical epidemiology, the study of the distribution and determinants of disease among human populations, goes back many years. Indeed, as John Allen Paulos notes in his book "Innumeracy," probability theory began in the seventeenth century with gambling problems and "statistics began in the same century with the compilation of mortuary tables, and something of its origins stick to it as well."[202] Epidemiologists emphasize the fact that statistical correlation cannot prove causality. Indeed, they have coined the phrase "ecological fallacy" to indicate cases where people have mistakenly drawn the conclusion that A caused B from parallel movements of factors A and B. Every epidemiology textbook is full of examples of how one can make this erroneous conclusion. Common sense tells us the same thing. Any baseball fan knows that if a team won 15 out of 18 games (factor B) in August, this probably has less to do with the fact that the temperature on the winning days was over 90 degrees (factor A), than with a star pitcher or hitter returning to the lineup (factor C). We would reach this conclusion even if 90-degree days correlate better with winning, because the star player may have been in the lineup on losing days too. Thus, if there is a plausible theoretical mechanism by which factor A could cause factor B, then their correlation is much more plausible than one with some factor C for which there is no obvious causal connection to B. Statistical correlation has always been used to help identify potential causal factors. The search for statistical correlation is not meant to replace causality studies, but rather to provide clues where to look. For example, if one finds an outbreak of sickness in a local community, the epidemiologist would look at many factors, trying to identify which correlates with the illness. If it turns out that most of the people in the community who became sick had eaten dinner at Tom's Restaurant within the last week, it would be a reasonable first step to examine the restaurant or something in it for a causal factor. Follow-up studies scrutinizing the restaurant's cleanliness, food packaging, etc., could ultimately prove that something at Tom's caused the outbreak. However, an examination of the people who became ill in that community may also show a statistical correlation between the illness and wearing the color red. The sensible epidemiologist, faced with two factors to investigate, factor A (eating at Tom's restaurant), or factor C (wearing red), would undoubtedly put his/her resources into investigating factor A. Common sense suggests that eating at a restaurant is a much more plausible candidate for causing the illness than is wearing red; though, a curious epidemiologist might also check out the possibility that people were wearing clothes dyed red with a harmful chemical. In this book, the causal hypothesis linking factor A (low-level radiation) with factor B (excess deaths) is the "Petkau effect," discussed in detail below. The Petkau effect offers a plausible explanation that suggests low-level radiation may in fact cause excess mortality, and this hypothesis is supported by the statistically significant correlations. This brings us to the important notion of "excess death" used throughout the book.[203] Epidemiologists use the concept of excess death to show that certain geographic areas and demographic groups suffer from unexpectedly high mortality rates. Excess deaths may be roughly defined as the difference between the number of deaths observed in a given population and the expected number. It is relatively easy to measure the observed number of deaths using government tabulations of death certificates. The more difficult question is: how do we know how many deaths to expect? The most common method epidemiologists use to estimate expected deaths is to compare the population of concern, for example, residents in counties surrounding the Savannah River nuclear plant, with a much larger population, such as all residents of the United States. The basic idea is that the much larger U.S. population experiences a "normal" or average rate of mortality that can be used as a yardstick. Thus, the observed mortality among the smaller population, be it a locality, a particular age cohort, or other grouping, is tested to determine if it is significantly different from the national norm. The smaller group is often called a "sample" taken from the "universe" of the U.S. as whole. To make this comparison, epidemiologists first "standardize" the population of concern to rule out the influence of peculiarities in age, gender, and racial composition. These three characteristics are most commonly standardized, partly because such data are systematically collected on death certificates. For example, if a county has a much higher proportion of older people, then it is natural to expect a higher mortality rate. Similarly, since women tend to live longer than men, if a county has an unusually high proportion of men, this too might account for a higher mortality rates. In addition to classifying deaths according to age and sex, the government differentiates between "whites" and "nonwhites," with the latter including a wide mix of African Americans, Spanish Americans Asian Americans, Native Americans, and so on. Although "non-whites" is thus a very imprecise category, it generally has been observed to have significantly higher rates of mortality than the category "whites." Epidemiologists deal with these variations by first dividing the population of concern into the different age, gender, and race groups, and then calculating the expected mortality rates of each age-sex-race group based on the corresponding national rate. The calculation of theoretically expected mortality can then be compared with the mortality rate that is actually observed in each group to yield the difference. Thus excess death is defined as the number of observed deaths that are significantly higher than expected for each race-sex- age group based on their corresponding national average; although, sometimes, only adjustments for age are used. It is important to note one flaw in the conventional standardization technique used by government epidemiologists, and employed here. By standardizing for the vague racial categories of white and nonwhite, this technique underestimates the excess deaths suffered by people of color. Rather than considering the cause of higher mortality among nonwhites, this technique simply defines it as expected. There are clear biological and behavioral explanations for expecting higher mortality among certain sex and age groups; the same cannot be said for the multi-racial grouping called nonwhite. A person over eighty is more likely to die than a young adult. Females are more likely to get breast cancer than males, and males are more likely to have a heart attack than females. The majority of differences in mortality among racial and ethnic groups, on the other hand, are caused by environmental factors, including living conditions, diet, pollution, etc. Only a very few diseases have been genetically linked to certain racial and ethnic groups, Tay-Sachs among Jews, for example, and sickle-cell anemia among African Americans. Defining higher mortality among nonwhites as expected is thus similar to saying that society expects nonwhites to be exposed to unhealthy environmental factors. Future research should adjust for this distortion; however, this was not done here. What elevates an excess death to the level of being "statistically significant?" In rough terms, an event is statistically significant when it is "improbable" that it would be observed in the real world if merely the laws of chance were operating. Epidemiologists seek fluctuations in mortality that exceed the limits of chance variation. They do this by determining the "improbability" of the difference between observed and expected mortality. A difference so great that it is improbable that it results from chance means the observed mortality increase or excess is "significant." Throughout this book, an increase in mortality is characterized as significant if the probability that it could be due to chance is less than one out of 100 (a "P value" of less than 0.01). This judgement can be made precisely, because variations in mortality conform to the bell-shaped "normal" curve. Because the statistical demonstrations are intended to develop hypotheses rather than to prove their validity definitively, significant divergences have been identified occasionally with P values of less than 0.05 (less than five percent probability of a chance result). Both confidence levels are commonly used by statisticians. Any individual case that passes a significance test may still reflect a random variation. But the cumulative significance of the five sets of correlations between low-level radiation and increased mortality, considered in Chapters Two, Four, Five, Eight and Nine, means that the likelihood that they are all chance occurrences is remote. Imagine repeatedly tossing one hundred coins and recording the percentage of heads that turns up in each repetition. The following sequence could result: 51 percent heads, 48 percent heads, 50 percent heads, 47 percent heads, 52 percent heads, 50 percent heads, etc. If we keep tossing the one hundred coins one thousand times and then plot the number of times each percentage appears, we would generate the bell-shaped normal curve, and 50 percent heads would be the most frequently recorded, or "mean," result. Statistical theory enables us to determine that the "standard deviation" of this distribution is plus or minus five percent heads. This means that about two-thirds of all results are expected to fall within one standard deviation on either side of the mean result (from 45 or 55 percent heads). About 95 percent of all possible results are expected to fall within the interval between 40 and 60 percent heads, or two standard deviations on either side of the mean. Most statisticians would regard the remaining possible outcomes (i.e., less than 40 or more than 60 percent heads) as highly improbable and thus statistically significant. The following table indicates the P value, or degree of statistical improbability associated with three increasingly improbable outcomes: PERCENT STANDARD P HEADS DEVIATIONS VALUE 65% 3 0.001 70% 4 0.0001 75% 5 0.000001 When an increase in mortality has a P value of less than 0.001, that is, less than one out of one thousand, this is equivalent to the highly improbable act of tossing one hundred coins and getting 65 heads and 35 tails. The formula used in this book for computing the significance of mortality phenomena is the standard one described by the National Center for Health Statistics in the annual volumes of the "Vital Statistics of the United States:" (O - E) / SQRT ( (O^2 + E^2) / N) where O = observed mortality rate; E = expected mortality rate; SQRT = the square route of; and N = observed number of deaths. Expected rates are calculated as a function of the original observed rate multiplied by the change in the U.S. rate. This formula is based on a Poisson distribution, which is appropriate for statistically rare events such as mortality. The formula yields that number of standard deviations by which the observed rate differs from the expected rate. This value can be converted to a probability estimate with a table of the area under the normal curve, which can be found in the back of any statistics textbook. Since we calculated that it is highly unlikely that the excess deaths found in the case studies were due to chance, what could have caused them? The hypothesis proposed here is that they were caused by a biochemical mechanism whereby ingested fission products promote the formation of "free radicals" that damage the immune system. This mechanism was discovered in 1972 by Abram Petkau.[204] The statistical tests in this book demonstrate that there were highly significant events among large human populations, each of which requires a reasonable explanation. The Petkau effect is a plausible biochemical mechanism (though significance tests cannot prove it was a cause), and thus must be considered. Dr. Abram Petkau is a Canadian physician and biophysicist who until recently managed the Medical Biophysics Branch of the Whiteshell Nuclear Research Establishment, located in Pinawa, Manitoba. While studying the action of radiation on cell membranes in 1971, Dr. Petkau conducted an experiment never done before. He added a small amount of radioactive sodium-22 to water containing model lipid membranes extracted from fresh beef brain. To his surprise, the membranes burst from exposure to just one "rad" (a measure of the amount of radiation absorbed) over a long period of time. Conversely, Dr. Petkau had previously found that 3,500 rads were required to break the cell membrane when X-rays were applied for only a few minutes. He concluded that the longer the exposure, the smaller the dose needed to damage cells. After several more experiments, he discovered the cause of this surprising effect from low-level radiation. The irradiation process was liberating electrons, which were then captured by the dissolved oxygen in the water, forming a toxic negative ion known as a free- radical molecule. The negatively charged free-radical molecule is attracted to the electrically polarized cell membrane. This causes a chemical chain reaction that dissolves the lipid molecules, which are the principal structural components of all membranes in cells. The wounded and leaking cell, if unable to repair the damage, soon dies. If the free radicals are formed near the genetic material of the cell nucleus, the damaged cell may survive, but in mutated form. Subsequent research by Dr. Petkau and other scientists ultimately demonstrated that this process occurs even at background radiation levels.[205] At high levels of radiation, Petkau found less cellular damage from free-radical production per unit of energy absorbed than at low levels of radiation. Free radicals are so dangerous to living systems because they form in water, and water comprises eighty percent of a cell. Free radicals not only destroy healthy cells, but also affect normal cell function in a way believed to speed the aging process. Nature has provided some protection from free radicals, probably because they are normally produced by the oxygen metabolism within the cell. The protector, superoxide dismutase, quenches the chain reaction.[206] It is now believed that superoxide dismutase is found in all cells which use oxygen in their life processes. For example, human tissues that contain naturally high levels of superoxide dismutase, such as the brain, liver, thyroid, and pituitary, are more resistant to the effects of radiation than tissues low in superoxide dismutase content, such as the spleen and bone marrow. Apparently this enzyme evolved to protect biological systems from superoxide, or free-radical, damage caused by ultraviolet light, background radiation, and the result of normal energy production in the cell. However, radiation which is produced by fission products and ingested through the food chain, or applied externally, can produce more free radicals than the body can deactivate (or "dismutate"), resulting in gross damage that may be irreparable. Furthermore, Dr. Petkau and others have found that only ten to twenty millirads will destroy a cell membrane, in the absence of the protective superoxide dismutase. The free-radical reaction can be quenched in another way. At higher intensities of radiation, the free-radicals become so concentrated that they tend to deactivate each other. If this were not so, medical X-rays would cause far greater biological damage than they do. A simple analogy, first used by Dr. Sternglass, can explain this phenomenon. Think of the free radicals as individuals in a crowded room. A fire starts and everyone tries to get out at the same time. As a result, everyone bumps into each other and very few escape. If only a few people are in the room when the fire occurs, however, everyone leaves easily through the door. The rate of escape is very high, and therefore, efficient. Chronic exposure to low-level radiation produces only a few free radicals at a time. These can reach and penetrate the membranes of blood cells with great efficiency, thus damaging the integrity of the entire immune system although very little radiation has been absorbed. In contrast, short, intense exposures to radiation, as with medical X-rays, form so many free radicals that they bump into each other and become harmless ordinary oxygen molecules. Short exposures thus produce much less membrane damage than the same dose given slowly over a period of days, months, or years. More recently, Charles Waldren and co-researchers have found that when a single human chromosome is placed in a hybrid cell and irradiated, the ionizing radiation produces mutations much more efficiently at low than at high doses, as in the case of cell membrane damage.[207] They found that very low levels of ionizing radiation produce mutations two hundred times more efficiently than the conventional method of using high dose rates, or brief bursts from X- ray machines. They found that the dose-response curve exhibits a downward concavity (logarithmic or supra-linear relationship) in mammalian cells, so that the mutational efficiency of X-radiation is maximal at low doses, exactly as was found by Petkau for free-radical mediated biological damage. Thus, their findings contradict the conventional scientific dogma that the dose-response curve is linear, and that a straight line can be used to estimate low-dose effects from studies of high doses. A protracted exposure to ingested beta emitters can be one thousand times more harmful to cell membranes than a brief external exposure to X-rays, because DNA repairs itself relatively efficiently after an X- ray hit compared to the damage caused by oxygen free-radicals at very low doses.[208] This type of exposure may thus account for the jump observed in mortality immediately after nuclear plant accidents, or after fallout from atmospheric bomb tests. Strontium-90 is chemically similar to calcium and, therefore, concentrates in the bone of the developing infant, child, and adolescent. Once in the bone, strontium-90 irradiates the marrow where the cells of the immune system originate at a low rate over a period of many years. As first discovered by Dr. Stokke and his co- workers at the Oslo Cancer Hospital in 1968, extremely small doses of only ten to twenty millirads can produce visible damage to the blood forming cells of the bone marrow, probably via the production of free-radical oxygen.[209] This can lead to the development of bone cancer, leukemia and other malignant neoplasms both directly by damaging the genes, and indirectly by lowering the ability of the immune system to detect and destroy cancer cells.[210] A peak accumulation of strontium-90 in the body after two or three years could explain the delayed peaks in total mortality as observed after the Savannah River Plant accidents described in Chapter Four. This accumulation results from the combination of growing uptake and slow excretion, and the consequent mortality primarily involves deaths from heart diseases, as well as from cancers and other causes. Free- radical oxygen, produced most efficiently by internal beta emitters such as strontium-90, may be a factor in coronary heart disease as well as cancer. The theory is that the free radicals oxidize the low-density cholesterol and cause it to become more readily deposited in arteries, thus blocking the flow of blood and inducing heart attacks.[211] Recent medical research from across the country has provided new evidence linking cancer to impaired immune systems.[212] The studies have focused on transplant patients, who as a group suffer from extremely high rates of a variety of cancers. Their cancers diminished rapidly when the doses of immunosuppressive drugs were reduced. (Such drugs are given to stop normal immune systems from rejecting the transplanted organs.) The researchers suspect two types of cells in the immune system of playing major roles in this phenomenon: natural killer cells and cytotoxic T-cells. They found evidence that during immunosuppression, these cells were more depleted among the transplant patients who developed skin cancers than among those who did not.[213] Earlier research published in 1977 demonstrated that bone-seeking isotopes such as strontium-89 and strontium-90 deactivated precisely such natural killer cells in laboratory mice.[214] These new findings linking cancer to immunodeficiencies, combined with the earlier findings of Petkau and others of higher-than-expected cell damage from low radiation doses, point to a possible explanation for the rapid increases in mortality rates after low-level radiation releases. The correlations of health effects with low-level radiation that are discussed throughout this book may thus be caused indirectly by chronic low-level exposures to ingested radiation through hormonal and immune system damage from free radicals. Low levels of strontium-90 and iodine-131 ingested in food, milk, and water, and breathed in air, may damage the ability of the body to detect and destroy infected or malignant cells. Such damage may occur even if radiation is present at concentrations far below existing standards. These standards were set on the basis of a quite different biological mechanism: cancer cell production caused by the direct impact on genes of high doses of external radiation. [203] For a more technical description of the methodology for calculating excess deaths, see Public Data Access, Inc., "Mortality and Toxics Along the Mississippi River," Washington, DC: Greenpeace USA, 1988. The method used here is fully documented in the Greenpeace report with the one addition that race has been figured into the adjustment process. [204] A. Petkau, "Effect of 22 Na+ on a phospholipid membrane," "Health Physics," Vol. 22, 1972, p. 239. See also A. Petkau, "A Radiation carcinogenesis from a membrane perspective," "Acta Physiologica Scandinavia," Suppl. Vol. 492, 1980, pp. 81-90. [205] A. Petkau and W. S. Chelack, "Radioprotective effect of superoxide dismutase on model phospholipid membranes," "Biochemica et Biophysica Acta," Vol. 433, 1976, pp. 445-456. See also A. Petkau, W. Kelly, W. S. Chelack, S. D. Pleskach, C. Barefoot, and B. E. Meeker, "Radioprotection of bone marrow stem cells by superoxide dismutase," "Biochemical and Biophysical Research Communications," Vol. 67, No. 3, 1975, pp. 1167-1174; A. Petkau, W. S. Chelack and S. D. Pleskach, "Protection of post-irradiated mice by superoxide dismutase," "International Journal of Radiation Biology," Vol. 29, No. 2, 1976, pp. 297-299; A. Petkau, "Radiation protection by superoxide dismutase," "Photochemistry and Photobiology," Vol. 28, 1978, pp. 765-774; A. Petkau, "Protection and repair of irradiated membranes," in "Free Radicals, Aging, and Degenerative Diseases," Alan R. Liss, Inc., 1986, pp. 481-508; and A. Petkau, "Role of superoxide dismutase in modification of radiation in jury," "British Journal of Cancer," Vol. 55, Suppl. VIII, 1987, pp. 87-95. [206] Irwin Fridovich, "The biology of oxygen radicals: the superoxide radical is an agent of oxygen toxicity; superoxide dismutases provide an important defense," "Science," Vol. 201, 1978, pp. 875-880. [207] Charles Waldren, Laura Correll, Marguerite A. Sognier and Theodore T. Puck, "Measurement of low levels of X-ray mutagenesis in relation to human disease," "The Proceedings of the National Academy of Sciences," Vol. 83, 1986, pp. 4839-4843. [208] T. Stokke, P. Offedal, and A. Pappas, "Effects of small doses of strontium-90 on the ratbone marrow," "Acta Radiologica," Vol. 7, 1968, pp. 321-329. [209] Ibid. [210] Peter A. Cerutti, "Prooxidant states and tumor production," "Science," Vol. 227, 1985, pp. 375-381. [211] See New York Academy of Science, "Antioxidants may prevent or slow down heart disease," "Science Focus," Vol. 3, No. 4, Spring 1989, p. 8, and Jane E. Brody, "Natural chemicals now called major cause of disease," "The New York Times," April 26, 1988 and Jean L. Marx, "Oxygen free radicals linked to many diseases," "Science," Vol. 235, 1987, pp. 529-531. [212] Elizabeth Rosenthal, "Transplant patients illuminate link between cancer and immunity," "The New York Times," December 5, 1989. [213] Ibid. [214] O. Heller and H. Wigzell, "Supression of natural killer cell activity with radioactive strontium: effector cells are marrow dependent," "Journal of Immunology," Vol.110, 1977, pp. 1503-1506. Also, E. Sternglass found in 1973 that significant changes in cervical cancer incidence and mortality in Baltimore women were directly correlated with changes in concentrations of the short- lived strontium-89 found in milk. See "Epidemiological studies of fallout and patterns of cancer" in "Radionuclides and Carcinogenesis, U.S. AEC Symposium Series 29, Conference-720505, edited by C. L. Sanders, et al., Washington, DC: U.S. Atomic Energy Commission, June 1973, pp. 254-277. -- "We are able to inform you that ancient grandfathers, the great stands of cedar and redwoods, are in danger of extinction by chainsaws. The maple, chief of trees, is dying from the top down, as was prophesied by Ganiodaiio, Handsome Lake, in 1799. Great rivers and streams are filled with chemicals and filth, and these great veins of life are being used as sewers. "We were told the female is sacred and carries the gift of life as our Mother Earth, the family is the center of our life and that we must build our communities with life and respect for one another. "We were told the Creator loves children the most, and we can tell the state of affairs of the nation by how the children are being treated. "When we return to Onondaga, we will begin our Great Midwinter ceremonies. We will tie the past year in a bundle and give thanks once again for another year on this earth. "This was given to us, and we have despoiled and polluted it. If we are to survive, dear friends and colleagues, we must clean it up now or suffer its consequences. . . . But Lyons also remembered turning to Leon Shenandoah, chief of the Grand Council of the Six Nations Confederacy. "My chief, he doesn't say much, but I asked and he said, `They're not taking it serious enough. I don't think they realize what's going to happen to them. What's coming.' He would have liked to see less posturing. We have our prophecies. We know what is coming down the road.'" -- Onondaga Chief Oren Lyons, on the Global Forum he helped organize on Environment and Development for Survival held in Moscow, January 15 to 19, 1990.