Wade Sikorski
v
Montana Department of Public Health and Human Services
Cause number 2002-5
I am filing this motion seeking a writ of mandamus from the court ordering the Montana Department of Public Health and Human Services to publicly acknowledge the existence of high rates of childhood leukemia, breast cancer, colorectal cancer, and birth abnormalities in Fallon County and across Montana, advise the public on ways to reduce risk of contracting these diseases based on current science, investigate local environmental links to these diseases, and intervene in whatever way necessary to reduce toxic exposures. As the following brief will show, the incidence of childhood leukemia in Fallon County is 10 to 20 times higher than national averages, the incidences of breast cancer in Fallon County and in various other communities across Montana, including Billings, Great Falls, and Missoula, are among the highest rates in the world, the incidence of colorectal cancer in Fallon County is about 7 times higher than it is in neighboring counties to the north and south, and the incidence of birth abnormalities are several times higher throughout the southeastern quadrant of Montana than it is for most of the rest of the state.
In filing this motion, I am claiming my right under Article II, Section 3 of the Constitution of Montana to a clean and healthful environment:
Inalienable rights. All persons are born free and have certain inalienable rights. They include the right to a clean and healthful environment and the rights of pursuing life's basic necessities, enjoying and defending their lives and liberties, acquiring, possessing and protecting property, and seeking their safety, health and happiness in all lawful ways. In enjoying these rights, all persons recognize corresponding responsibilities.
I am also acknowledging my duty under Article IX, Section 1, to maintain and improve a clean and healthful environment, and I am insisting that the state do its duty as well:
Protection and improvement. (1) The state and each person shall maintain and improve a clean and healthful environment in Montana for present and future generations. (2) The legislature shall provide for the administration and enforcement of this duty. (3) The legislature shall provide adequate remedies for the protection of the environmental life support system from degradation and provide adequate remedies to prevent unreasonable depletion and degradation of natural resources.
I am also claiming that the Department of Public Health and Human Services also has a broad and specific duty under statutory law to protect public health, and that it has failed to execute it properly, acting arbitrarily, capriciously, and unlawfully against my petitions to investigate cancer in Fallon County. According to 50-1-202, MCA, under the heading of "General powers and duties," the Department of Public Health and Human Services shall:
(1) study conditions affecting the citizens of the state by making use of birth, death, and sickness records;
(2) make investigations, disseminate information, and make recommendations for control of diseases and improvement of
public health to persons, groups, or the public;
Seeking quick action, I have petitioned DPHHS repeatedly, most recently with a report more than 20 pages long, to investigate the possible environmental causes of these high rates of disease in Fallon County, and in each instance DPHHS has refused, arguing that the rates of cancer are not unusual and that there are no unsafe environmental exposures. I maintain that available evidence directly contradicts the department's position. This brief will argue that thereis a cancer problem in Fallon County, if not all across Montana, particularly for breast cancer, that under Montana's Constitution the people in areas that have these high rates have been denied their right to a clean and healthful environment, and that the department has a duty under both the Constitution and statutory law to admit the truth about the links between environmental pollution and certain diseases, adequately warn the public about them, investigate the causes, and do everything necessary to remove these causes from the environment.
In particular, I want DPHHS to warn the public that there is a problem with childhood leukemia in Fallon County and that
available research suggests that a variety of precautionary measures may lower the incidence of it. A variety of studies, for
example, have found that the exposure of both parents and children to pesticides, solvents, and ionizing radiation increases
the risk of childhood leukemia. (1) By precautionary inference, reducing the use of pesticides, particularly household and
garden pesticides, and materials containing benzene would likely substantially reduce the risk of childhood leukemia. In
addition, one recent study also found that maternal folate supplementation during pregnancy also substantially reduced the
risk of childhood leukemia. (2) Even without knowing the specific causes of the 4 cases of childhood leukemia in Fallon
County, enough is known overall about the environmental causes of childhood leukemia to warrant the department, as a
precautionary measure until a more specific understanding of the causes is available, telling the public how reducing
pesticide and solvent use, and how folate supplementation may prevent future cases of childhood leukemia. While this
would be good advice everywhere, DPHHS has an especially pressing duty to give it when a community has a particularly
high incidence of childhood leukemia, as Fallon County does. Its efforts to convey this potentially life saving information to
the public should be detailed, clear, and frequent-sufficient to adequately warn every resident in the affected area how to
protect their health and life.
I am a longstanding resident of Fallon County, born, raised, and currently living on the family farm 20 miles south of Baker, the county seat. My mother, Lucille Sikorski, died of breast cancer in 1994, my father, Edward Sikorski, died of leukemia in 2002, my cousin, Dalton Beyer, is one of the cases of childhood leukemia, and neighbors that live immediately beside our ranch have developed, among other things, Hodgkins lymphoma, pancreatic cancer, stomach cancer, bone cancer, and prostrate cancer. I am particularly concerned by the number of children and young adults developing cancer throughout Fallon County. Besides the childhood leukemia cases, I know of young adults or children living in Fallon County who have recently developed brain cancer, Hodgkins lymphoma, and testicular cancer. I also know of several cases of cancer in young adults who grew up in Fallon County but moved away, and were residents of other communities when they were diagnosed with cancer. Because they were living outside the county for years, they are unlikely to be included in the Tumor Registry's case count, even though it is possible that childhood exposures in Fallon County caused their cancer. If these people were included, Fallon County's cancer rates could be even worse than they are.
I am also deeply concerned about the high rates of breast cancer in various communities across Montana, including some of our most populated, like Billings, Great Falls, and Missoula-all of which have incidence rates that are among the highest ever recorded in the world. Because so many women in Montana are suffering and dying from this disease, because there is every reason to believe that the incidence of breast cancer could be significantly reduced by reducing the amount of xenoestrogens and other environmental hormones released into the environment, because these chemicals poison not only women but also may be the cause of falling sperm counts in men, the cause of developmental problems in children, and may the cause of significant disruption throughout the whole ecosystem, I feel that the alarmingly high rates of breast cancer in various communities across Montana are an indication of environmental harm that has become everyone's problem.
If these cancers that are appearing all around me are being caused by environmental toxins, it seems likely that my own
health would be threatened. I have been diagnosed with severe allergies, chronic Lyme Disease, and depression. I also have
urinary tract problems. All these health problems, I have found out from reading the scientific literature, can be either
exacerbated or caused by the same kinds of environmental hormones that I suspect may be causing cancer in Fallon County
and across Montana. And so I move that the court accept my standing to sue the department not only in my own behalf but
to protect the environmental life support system from degradation.
After my father was diagnosed with leukemia in the Spring of 2000, I decided that it was too much. I had helplessly watched my mother die of breast cancer years earlier, had watched neighbor after neighbor get cancer, and had watched my cousin recover from childhood leukemia. Finally, with my father's diagnosis, suspicion grew into conviction, and I knew something had to be done. On the advice of Anne Hedges, who works with the Montana Environmental Information Center, I petitioned the Agency for Toxic Substances and Disease Registry (ATSDR) to investigate cancer in Fallon County. As is its practice, the ATSDR asked the Montana Department of Public Health and Human Services (DPHHS) for information on cancer in Fallon County. Thinking I would cover all possibilities, I sent a copy of my petition to Governor Marc Racicot, and he asked various state agencies to investigate. Awhile later, I got a letter back from Governor Racicot saying:
Based on your letter, Montana's Medical Officer Dr. Michael Spence asked persons in the Montana Central Tumor Registry (MCTR), Office of Vital Statistics, DPHHS, to provide information about the incidence of cancer and mortality in Fallon County and other similar counties (enclosed). After studying the results from MCTR's information search, Dr. Spence was unable to identify any unusual incidences of cancer or mortality in the Fallon County data. (3)
Feeling that this simply could not be true, I wrote Governor Racicot back a letter insisting that Dr. Spence was wrong, that we did too, have a childhood leukemia cluster in Fallon County, and I offered my best shot at a statistical analysis. (Which turned out to be wrong--Iunderestimated the ratio between expected and observed rates by half, I later found out.) I concluded my letter by pleading my inadequacies and asking for enlightenment, in the forlorn hope that while explaining things to me and having to explicitly describe its methods and assumptions, DPHHS would have to concede it was wrong:
I am not an expert in epidemiology, and I acknowledge that I may have misunderstood how to interpret the data that I have available to me. If I am in error, and the incidence of cancer in Fallon County is not remarkable and nothing can be done to prevent it, I would very much appreciate someone explaining to me where my error lies. In your letter you said that you were enclosing the information that Dr. Spence used to come to his conclusions. Unfortunately, it was not enclosed. I would very much like to see it, as well as an explanation from him on how he came to the conclusion that he did. (4)
Perhaps persuaded by my argument, Governor Racicot ordered a more through investigation of cancer in Fallon County. The task was apparently assigned to Todd Damrow, an epidemiologist in the Department of Public Health and Human Services. Damrow contacted me several times by phone and email. I told him repeatedly about the cases of childhood leukemia that I knew of in Fallon County and my various concerns about other cases of childhood cancer, the improbable incidence of cancer in my family and our neighbors, and my various guesses about what could be causing it. He told me that he was working hard on it and that it was going to take some time.
Months later, Damrow wrote me a letter back saying in part that:
DPHHS was not able to verify all of the cases of childhood leukemia in Fallon County that you proclaim. The case number on record in the registry is within the expected range for the time period. Your method of case ascertainment was applied selectively to Fallon County, and biases the findings. It is important that study techniques used to capture cases be applied uniformly across all referent populations so that one ends up comparing "apples to apples." . . . In summary, the data do not support the contention that cancer rates are elevated in Fallon County or in southeastern Montana. These findings are consistent with the earlier study by Dr. Michael Spence, State Medical Officer, who reported that cancer incidence rates in the area are not unusual, i.e., did not differ from state and national cancer rates. These findings are also consistent with the review of existing regulatory data by DEQ which showed the current environmental quality to be protective of health. (5)
I was very disappointed with Damrow's response. Though he repeatedly chided me for failing to apply proper scientific method to cancer in Fallon County, (6) Damrow did not, despite my explicit request to Governor Racicot that the state do so, describe the statistical procedures used to analyze the data or the standards used to interpret it. Damrow just said that there was not a cancer problem in Fallon County. Good science means being able to offer evidence and justification for your conclusions. Instead of offering that, the department seems to be asking me to trust them to do the right thing, to take their word that everything is OK. However, after the tragedy in Libby, where various government agencies ignored an obvious threat to public health from asbestos, it is simply not within me to trust the government anymore. It has failed miserably in Libby, leaving it to an out-of-state newspaper to warn the public of a toxic disaster. (7) After very carefully considering the evidence, I feel that the government is failing the public trust again in Fallon County.
Dissatisfied with Damrow's response, trying to make a more persuasive argument for state action, I wrote a lengthy report, more than 20 pages, documenting just why I felt that Fallon County had an excessive cancer rate. I calculated the Poisson distribution test, the standard test used to detect disease clusters, for the cases of childhood leukemia, finding that they were, indeed, statistically significant, did a comparative analysis of the breast and colorectal cancer incidence rates, and rebutted various arguments that Damrow, Spence, and Racicot made. It was a carefully prepared report, and yet, a couple of weeks after I sent it to various members of the department, the reply that I got did not even offer any indication to me that anyone had actually read it. There was no analysis of my arguments, no dispute of the evidence I offered. All I got was a very brief and pointed letter back from Gail Gray, the director of the department, saying:
Health professionals in our department have conducted a through investigation of your report. While the findings may not meet with your approval or satisfaction, our administration remains in support of the findings. Further investigation will not be undertaken at this time. Our department takes very seriously its responsibility to protect the health of the public, and fully accepts responsibility for decisions made. (8)
With this last dismissal, trying to get by court order what I have failed three times to get by petitioning DPHHS, I am filing this suit against the department, and again I insist that DPHHS is wrong when it says that the number of cases of childhood leukemia in Fallon County are "within the expected range for the time period." This brief will establish that, in fact, the number of cases of childhood leukemia in Fallon County is excessive, considerably outside the range of anything that we should tolerate. Furthermore, this brief will establish that DPHHS's own statistics on breast cancer and colorectal cancer in Fallon County show that these other cancers are elevated--indeed, that the incidence of breast cancer, along with a dozen other counties throughout Montana, is one of the highest incidences ever reported in the world and that the incidence of colorectal cancer is about 7 times higher than neighboring counties, even if the overall incidence of all cancers in Fallon County is not remarkable. And this brief will also show that incidence of birth abnormalities in Fallon County, along with the entire quadrant of southeastern Montana, is much higher than in most of the rest of the state, suggesting further, if more subtle, health problems destined to emerge in this area.
In short, this brief will show that Fallon County does have an environmental health problem. And it will argue that it is time
that DPHHS stopped giving the residents of Fallon County false reassurances about the environmental threats to their
health. DPHHS should, instead, exert its efforts toward finding out what is causing these cancers and doing everything it
can to prevent them. At a minimum, it should make a special effort to advise my community that the risk for childhood
leukemia, breast cancer, and colorectal cancer is elevated, that early detection makes these cancers more treatable, and that
reducing exposure to various environmental hazards can reduce the possibility of getting cancer.
Four cases of childhood leukemia have been diagnosed since 1992 in Fallon County. There are 3 cases of ALL strongly linked to Fallon County, and 1 case of AML not so strongly linked, but still, I think, an indication of possible environmental harm through parental exposure in Fallon County. All of the mothers of the children who developed leukemia grew up in Fallon County, and, for the most part, lived in Fallon County when they reached adulthood, as did most of the fathers, breathing the air, drinking the water, and eating the food here. During their pregnancy and for at least a short period afterwards, all of the mothers of children diagnosed with ALL were living in Baker, though one of them had been briefly outside the county for the first part of her pregnancy, and another had moved around inside the county afterwards.
The main point is that all of these children with childhood leukemia, even the case of AML, which lived outside the county until being diagnosed with leukemia, are intimately linked to the environment in Fallon County. The bodies of their parents are made of the land, the air, and the water here, and their own bodies, either directly by living here or indirectly through an inheritance of genetic damage or toxic exposures in the womb, would have harvested the fruit of any toxic releases in Fallon County.
Pregnancy is the period when the link between the body's health and the environment's health is most manifest. Because it is rapidly growing and taking form, the fetus is vulnerable to even the slightest toxic exposures. At Woburn, Massachusetts, where two wells were contaminated by industrial solvents, trichloroethylene, perchloroethylene, chloroform, and other organic agents, and a cluster of childhood leukemia resulted, the greatest risk for the children, it turned out, was exposure to the water while their mother was pregnant with them. A report by the Massachusetts Department of Public Health describes the risk this way:
Strikingly however, the adjusted odds ratio for maternal exposure during pregnancy suggests that the risk of developing childhood leukemia was greater for those children whose mothers consumed water from Wells G and H during pregnancy than for those whose mothers did not. In contrast, odds ratios suggest that the relative risk of developing childhood leukemia among children who consumed water from Wells G and H from birth through childhood was virtually identical for cases and controls. This means that the cases' consumption of contaminated public water as children was not associated with their subsequent development of leukemia. A mother's consumption of contaminated public water while pregnant with the case was the strongest predictor of childhood leukemia. (9)
However, pregnancy is not the only period of vulnerability. Lifetime parental exposures are important too, perhaps even more important. For instance, at Sellafield, England, it was discovered that men exposed to low levels of radiation from a nuclear power plant were 8 times more likely to have children who would develop leukemia, although the children, themselves, were not exposed to higher levels of radiation. (10)
Parental exposure to radiation, and perhaps other toxins, even when the parents show no health damage, has been repeatedly associated with an increased risk of their children developing childhood leukemia. (11) This happens with radiation, some scientists believe, because radiation alters the human gene pool, disrupting genome stability. According to scientists like John Gofman, while high doses of radiation kills cells, low doses disrupt genetics, leaving the cells alive, making low doses of radiation disproportionately mutagenic and carcinogenic. According to scientists, radiation has doubled the rate of genetic mutation in children born up to 300 km from the city of Chernobyl in the former Soviet Union, where a nuclear reactor melted down. Another study found that radiation caused genetic mutation that feed through to phenotypic changes in swallows near Chernobyl. (12) Still another recent study found a dramatic increase in the incidence of ALL among male children exposed to fallout from Chernobyl, with a less dramatic increase for females. (13) This isn't a result isolated to Russia either. A high rate of childhood leukemia was also found near German nuclear power plants, even though the radioactive exposure was much lower than at Chernobyl. (14)
Parental exposure to radiation is not the only parental exposure linked to childhood leukemia. Recently, a study of Agent Orange veterans by the Institute of Medicine of the National Academies found that childhood AML was associated with parental exposure to dioxin in Vietnam. (15) Other studies have found that other parental occupational exposures are linked to various childhood cancers, according to Joanne S. Colt and Aaron Blair, who did an overview of 48 studies linking childhood cancer to occupational exposures of parents:
The evidence for an association between childhood leukemia and paternal exposure to solvents is quite strong. All five of
the studies addressing solvent exposures have reported positive associations. A number of the relative risks were quite
large (i.e., greater than 3.0), and despite the small number of exposed cases in many of the studies, several were
statistically significant solvents in general, (16) chlorinated solvents, (17) and benzene, carbon tetrachloride, and
trichloroethylene. (18) Buckley et al. found a significant trend by duration of exposure for unclassified solvents, but could
not identify with confidence the specific solvents associated with acute nonlymphocytic leukemia (ANLL) risk. The
association between childhood cancer and solvents is an added concern because benzene (19) is a well-established risk
factor for adult leukemia and other solvents are suspected leukemogens. (20)
Interestingly, because of the oilfield in Fallon County, which flares off large amounts of various gases, Colt and Blair go on to observe that parental exposure to hydrocarbon exhaust gases is associated with childhood leukemia:
There have been 12 studies of childhood leukemia and paternal employment in occupations related to motor vehicles or involving exposure to exhaust gases. Elevated risk was found in most of these studies, with statistically significant findings in six. Significant associations were found among diverse occupations such as motor vehicle or lorry drivers, (21) mechanics and gas station attendants, (22) and broader groups of motor vehicle-related occupations. (23) In their review of leukemia Linet and Cartwright suggested that the link between motor vehicle occupations and adult leukemia may be due to benzene and other components in engine exhausts. (24)
It may also be relevant that the Montana Department of Transportation, acting upon a request by the Southeastern Montana Alliance, a former affiliate of the Northern Plains Resource Council, discovered that Ross Management, a transformer salvage company operating in Baker, was tested and was found using transformer oil to fuel its trucks. (25) Leaving behind 2 Superfund sites and banned from operating in Washington State, Ross Management turned up in Baker just a couple of months before the first case of childhood leukemia. (26) There were also rumors that Ross Management was selling or giving people transformer oil to use as heating oil in their shops and homes in Fallon County. Although there has been no proof that Ross was responsible, there also was a cluster of childhood cancer in Onalaska, Washington, one of the towns in Washington where the Ross family lived and operated before moving to Montana.
The Department of Environmental Quality has ruled that Ross needs a resource recovery permit to burn transformer oil in their trucks. Curiously, however, despite the ruling that Ross needs this permit, the DEQ is allowing Ross to operate without it. My feeling is that if the state says a permit is needed, a company should not be operating without it. Besides running the risk of containing PCBs, transformer oil contains a variety of other substances that when burned may be more toxic than ordinary exhaust gases. Since the exhaust of trucks is low to the ground, and people could get large momentary doses from being next to them, say at a traffic light or idling at a truck stop, the effect of short-lived toxic exposures could be high. Although the volume of the releases may be relatively small from Ross Management's trucks, at least in comparison with exhaust gases from the oilfield near Baker, where large amounts of gas are routinely flared off, the coincidence of Ross arriving and the first cases of childhood cancer happening months later seems rather striking.
However, while hydrocarbon exposure has a strong link to childhood leukemia, one of the strongest links to leukemia in children is use of pesticides in the home or garden. Ruth Lowengart, John Peters, and et al. in a case-control study of children 10 years and under in Los Angeles County, besides finding significant risks associated with chlorinated solvents, cutting oil, methyl ethyl ketone, spray paint, father's employment in aircraft manufacturing, and mother's employment in personal services industries, also found a highly significant risk for use of pesticides in the home (OR=3.8, P=0.004) or garden (OR=6.5, P=0.007). (27)
As a practical matter, it is hard to sort out whether the childhood leukemia was caused by parental exposure before conception or direct exposure to the child afterwards. As Sandra Steingraber observes, causes can be very mixed, overdetermined, and redundant. Exposures often start before conception, continue through pregnancy, and persist afterwards:
The children of adults who work in specific occupations also have higher rates of cancer. Childhood brain cancers and
leukemias are consistently associated with parental exposure to paint, petroleum products, solvents, and pesticides. Some
exposures may occur before birth. Children can also be exposed when these materials are carried into the home on their
parents' clothes and shoes, through breast milk (which can be contaminated directly or through maternal contact with the
father's clothing), or even through exhaled air: because solvents are, in part, cleared by the lungs, parents can expose
their children to carcinogens simply by breathing on them. In this way, a father's homecoming kiss and work-clothed
embrace can contaminate his child. (28)
Whatever caused it, the first case of ALL in this cluster was born in 1990, and diagnosed at the age of 2 in 1992. The second case of ALL was born in 1992, and diagnosed at the age of 5 in 1997. The third case of ALL was born in 1993, and diagnosed at the age of 7 in 2000. The single case of AML was born in 1990, and diagnosed at the age of 4 in 1994.
Fallon County covers a large area, 1,620 square miles, but it has a small population, only 3,103 people, as counted by the 1990 census, with 236 children under age 5, and 713 between ages 5 to 17 years. (29) According to recent U.S. Census Bureau estimates, the population of Fallon County has since fallen 7%, to a total population of 2,885, with 26.6% less than 18 years--which would be 767 children. (30) Perhaps this estimate is too optimistic. I've been told that enrollment in the Baker School system, grades K through 8, has dropped dramatically from 441 in the fall of 1990 to only 292 in the fall of 2000.
In 1990, Baker, the county seat of Fallon County, where all the mothers of the children diagnosed with ALL lived during pregnancy, had a population of 1,818 people, with 144 less than 5 years of age, and 406 between the ages of 5 and 17. (31)
That's Fallon County, the place where the childhood leukemia is happening. Now, the general demographics of childhood leukemia:
Childhood cancer should be a rare disease. According to the National Cancer Institute, for children under the age of 15, there are only 14.1 cases of cancer of all kinds per 100,000 children per year. Leukemias of all varieties have an incidence for children under the age of 15 of only 4.3 per 100,000 children per year. (32) For ALL, and children under the age of 15, it is 3.2 cases per 100,000. (33) According to the Centers for Disease Control's SEER data set, from 1973 to 1994, the incidence of childhood ALL rose approximately 1.0% a year for all sexes and races combined. Between 1973 and 1994, the incidence of ALL in female children increased by 32%. (34) (Leukemia is not the only childhood disease that has been increasing, it might be added as well. The incidence of asthma, hypospadias, brain cancer, Wilms tumor, and testicular cancer has been increasing as well since the early 1970s. There has been a 39% aggregate increase in brain cancer in children, and a 68% increase in testicular cancer. (35))
The incidence of leukemia varies dramatically as children age, with the highest incidences in the first years of life, with boys consistently higher than girls, and falling dramatically as children age. In California, for instance, the incidence of ALL, the most common form of childhood leukemia, was 6.5 per 100,000 children per year for ages 0-4. It fell to 3.5 for ages 5-9, and to 0.9 for ages 20-24. (36) According to the American Cancer Society, by the time all of this is averaged together for the ages of 0-19, the incidence of ALL is 2.7 per 100,000 per year. (37) So, the incidence rate depends a lot on which age group is specified.
Unfortunately, at the time that this was written, Montana's Tumor Registry did not appear to have accurate age-specific incidence data for leukemia in Montana, so we are limited to using either national rates or the rates of other states like California.
For the limited purpose of calculating the observed rate, and making the assumptions as conservative as possible to satisfy all doubters, I will exclude the single case of AML, since its connection to Fallon County may be disputed, using only the 3 cases of ALL. However, given the studies linking childhood cancer to parental exposures to radiation, Agent Orange and other pesticides, hydrocarbons, and various solvents, as we saw above, I nevertheless maintain that the case of AML is likely caused by the same things causing the other cases of childhood leukemia in Fallon County.
Specifying the appropriate boundaries of an exposed population for a cancer cluster and matching it with the appropriate incidence rates is fraught with many complexities. Perhaps the most convenient incidence rate would be the National Cancer Institute's rate of 4.3 cases of childhood leukemia per 100,000 children under age 15. Using this rate for Fallon County would overstate the expected rate because it would include AML cases when we are excluding the only childhood AML case.
So, assuming that we had about 800 children under the age of 15 in Fallon County (This is just a round estimate, averaging out the differences over the years between high and low.), with a rate of childhood leukemia of 4.3/100,000 per year, we should expect to have 0.034 cases of childhood leukemia per year in Fallon County, or 0.34 over a 10-year period. However, we have had at least 3 cases of childhood leukemia in Fallon County in less than 10 years--which is a rate just about 10 times higher than expected. If we use an incidence rate for only ALL, which is 3.2 per 100,000 for children under age 15, we get an expected incidence rate of 0.26 in 10 years. This would make the actual rate about 12 times higher than expected.
Note that we are including all the children in Fallon County in these calculations when the relevant population may only be the children in Baker. All of the children who developed ALL were in Baker at the time of their greatest vulnerability, the period in the womb, so, if we assume the causes are similar to the Woburn model, where the exposure was from contaminated wells, the exposure risk may be limited to just the town of Baker. Fallon County is a big place, covering an area the size of a small New England State. Some children are home-schooled, or attend rural schools, or live 30 miles away from Baker. If the environmental risk were limited to Baker, these children living outside Baker would not be exposed, and so we can exclude them from the risk.
If we limit the possible exposure population to children under age 15 living only in Baker, their number would fall to about 500, probably less. And so, using the same leukemia incidence rate of 4.3/100,000 children, we would expect to have only 0.02 cases of childhood leukemia per year in Baker, or 0.2 in 10 years. However, Baker's actual rate in the 90's is at least 15 times higher than that. (Or, calculating it just for ALL, using an expected rate of 0.16 in 10 years, it would be 19 times that.)
To sum this all up, depending on the assumptions made, the incidence of childhood leukemia ranges from being just about10 times higher than expected to 19 times higher than expected. The next step that epidemiologists usually take is to calculate how unlikely this all is.
The standard test used by epidemiologists for determining the statistical significance of cancer clusters is called the Poisson Distribution Test. The Poisson test measures how likely, or unlikely, an event is to happen by comparing expected outcomes with actual outcomes. (38) If we assume all of the children in Fallon County are exposed to whatever is causing the leukemia and use a rate that includes not just ALL but all forms of childhood leukemia, the Poisson Distribution calculation for this outcome is: 0.004663. This means that the chance of having 3 cases of childhood leukemia in Fallon County in a 10-year period is 1 in 216. That is to say, if you were to imagine 216 clusters of childhood leukemia like there are in Fallon County only 1 of them would have happened by chance. The other 215 would presumably be caused by something in the environment.
However, things are probably even worse than this, since this is, in all likelihood, an excessively conservative estimate, using unreasonable assumptions, like including AML when we excluded the single case of it, and including the entire population of Fallon County when the 3 cases of ALL are most tightly tied to Baker. If we limit the rate to only the incidence of ALL in Fallon County, the Poisson Distribution number for this event falls to .002259, or 1 chance in 443. If we limit the population of children to Baker, but use the rate for every kind of childhood leukemia, the Poisson Distribution number falls to 0.001092, or one chance in 916. If we limit the population of children to Baker, and limit the rate to the national incidence of ALL, the Poisson Distribution number falls to 0.000582, or 1 chance in 1,718.
Finally, if you do include the 4th case of childhood leukemia-as I think is entirely justified, given the evidence that parental exposures are a plausible cause-the possibility that this distribution could happen by chance becomes astronomical. When I calculated it, my computer had to resort to exponential notation. Too many zeros. Using an expected rate of 0.2, the possibility that 4 cases of childhood leukemia could happen by chance in Baker in 10 years is: 5.458E-05.
To sum this all up, the possibility that 3 cases of childhood leukemia could happen by chance in Fallon County in the last 10 years ranges from once in 216 times to once in 1,718 times, depending on the assumptions made. When you add a 4th case, the odds become astronomical.
Even the most conservative of these estimations of Poisson improbability, which would be 0.004663, exceeds what epidemiologists accept as a "significant" outcome. The standard for significance using the Poisson Distribution Test is 0.005, or 1 chance in 200. Anything more unlikely than that is considered nonrandom, a cancer cluster caused by something atypical in the environment. (39) It should be noted that the standard for significance here is 10 times higher than is usually the case. Most of the time, epidemiologists consider anything more improbable than 0.05, or 1 in 20, significant. They apply the higher standard here because post hoc evaluations, some statisticians believe, should be more stringent than for new data. The argument is that applying the Poisson Test on a cluster that you know exists is like Texas Sharpshooting, where you shoot a gun at the broad side of a barn and then paint a bull's eye around the hole. The chance of missing has been eliminated. (I think that there are fundamental flaws in this argument, since it assumes that cancer is randomly distributed in a way that is not supported by the evidence, but I will go into that in a later section.)
Nevertheless, that is not the problem in Fallon County. The cluster of childhood leukemia is significant by any professional standard for investigating cancer clusters. Since 1961, for example, the Centers for Disease Control has been investigating cancer clusters, most often leukemia clusters. Between 1961 and 1982 it conducted 108 cancer cluster investigations. The CDC's criteria for proceeding with a cluster study in these cases were:
1) the number of cases exceeded the crude expected number by a factor of 2 or with a probability of 0.05
2)the cases were of one type or one system (i.e., all leukemia or leukemia and lymphomas, but this was not always followed . . . )
3) a specific population could be defined
4) a specific time frame for the occurrence of cases could be set
5) occasionally, when a specific exposure was known or suspected, even if the foregoing criteria were not met (40)
Notice that the CDC got interested when the probability of the cluster exceeded 0.05, a standard ten times easier than what Fallon County exceeds, or when its odds ratio was only 2, a standard that could be 5 times higher and Fallon County would still exceed it. Many childhood leukemia clusters investigated by the CDC never rose to the level of statistical significance Fallon County has. A typical childhood leukemia cluster, one that would very likely be investigated by the CDC or a state health department, is where the children at risk are only 5 times more likely to develop leukemia than children not at risk. (41)
The higher rate in Fallon County also suggests that the relative risk is higher because the exposure is more toxic, as Wartenberg argues:
The relative risk is the typical measure of the severity of the hazard. It tells how much more likely a person is to get the disease if the person is exposed to the hazard. Risks of known environmental hazards usually range from about 1 to 5, with unusually strong hazards being as large as 10.
So, contrary to what the DPHHS has concluded, the cluster of childhood leukemia in Fallon County is not only a valid cluster, it is an unusually strong cluster, suggesting an unusually toxic exposure. Other states, with similar incidence rates of childhood leukemia in their communities, have been investigated. Woburn, Massachusetts, a city of 35,000 people, had accumulated only 21 cases of childhood leukemia by 1986. (42) Yet this was enough to generate a massive lawsuit, a long list of epidemiological investigations by the Massachusetts Public Health Department and Harvard University, an EPA Superfund site, a best-selling book, and a movie staring John Travolta.
More recently, the State of Nevada's public health division promptly launched an investigation when the town of Fallon, Nevada, with a population of 8,300 people, accumulated 8 cases of childhood leukemia. (It has since risen substantially.) (43) The state epidemiologist, Randal Todd, said of the tragedy, "If there is something in the environment, we need to find it so we can prevent further cases." (44)
If Woburn, MA was throughly investigated when it had a childhood leukemia case for every 1,666 people, and Fallon, NV was investigated when it has a childhood leukemia case for every 1,038 people, surely Baker, MT has ample reason to be investigated when it has a childhood leukemia case for every 606 people. Why is the Montana Department of Public Health and Human Services refusing to do for Baker something that health departments across the country have felt obligated to do in their states?
Cancer and Radioactive Fallout in Montana
The causes of cancer are complex, an interplay of environmental, psychoneuro-immmunological, and genetic dynamics. Because of this complexity, a case of cancer is hard to trace back to the cause that started it, the one thing that can be blamed for it. In Fallon County, as everywhere else, we probably have had multiple layers of causes piling on top of each other, confounding and magnifying each other. Sorting them out will be as difficult as it is necessary.
Overarching the various explanations that may emerge, possibly concealing them by disrupting what can be considered normal, (45) is the largely unappreciated fact that the whole state of Montana was heavily contaminated with fallout from nuclear bomb testing in Nevada in the 1950's and 1960's. Although Utah is widely believed to have born the blunt of this fallout, Montana probably received as great, or greater share of it. Of the 25 counties in the United States that received the highest levels of Iodine-131 in the nation, Montana had 15--Meagher with a dose of 12-16 rads, and Broadwater, Beaverhead, Jefferson, Powell, Judith Basin, Madison, Fergus, Gallatin, Petroleum, Lewis and Clark, Blaine, Silver Bow, Choteau, and Deer Lodge all receiving 9-12 rads. (46) Almost all the rest of Montana was in the next highest dose range of 6-9 rads, a level that is still high compared with the rest of the nation. No other state, including Idaho and Utah, was so evenly covered by such high levels of Iodine-131. Apparently, to protect the populations of Denver on one side and Los Angeles on the other, the federal government waited until the wind was blowing toward Montana before setting off any tests.
Although Iodine-131 has been described as the worst risk factor from fallout, Strontium-90 may be more significant for Fallon County and all of southeastern Montana. According to a special report that appeared in The Fargo Forum, by the time a ban on above ground nuclear testing stopped the fallout, an area centered around Belle Fourche, SD, starting in the southeastern quarter of Montana, going through the western third of South Dakota, touching the southwestern corner of North Dakota and reaching down to Nebraska, had accumulated the highest levels of Strontium-90 in the nation--except for a small area in Utah. (47) Unlike Iodine-131, which has a half-life of 8 days, Strontium-90 has a half-life of 28 years. To the body, Strontium-90 resembles calcium, so it concentrates in the bones, where it irradiates the bone marrow. According to The Fargo Forum:
Strontium-90 is not distributed uniformly in the body. Swedish scientists determined there are radioactive hot spots in people's bones, just as there are hot spots of unusual radioactivity around the earth.
The hottest spots in the skeleton may be 6 to 60 times hotter than the skeletal average, the scientists found.
Assuming a person had a 180 strontium-unit average in his skeleton, hot spots as hot as 10,000 strontium units or more could occur. William Caster, formerly a researcher at the University of Minnesota, said concentrations in that range could cause bone cancer--or at least damage the bone, and double or triple the chances of contracting leukemia or anemia. (48)
Other scientists did statistical studies that found that nuclear fallout in the United States increased cancer incidence in the areas most affected. Victor Archer reported that fallout exposure was associated with two forms of childhood leukemia. He also found that a peak in childhood leukemia would follow, reliably and predictably, 5 and 1/2 years after each peak in nuclear testing in 1951, 1953, 1957, and 1962. (49)
Carl Johnson, a former medical officer in the South Dakota, found that nuclear testing in Nevada had the same effect on the people of South Dakota that the bombs dropped on Hiroshima and Nagasaki had on the people of Japan: Leukemia deaths peaked 5 and 9 years after the test bombs were set off, then slowly settled. (50) Other studies have found similar links between fallout and cancer, particularly leukemia. (51)
As is widely known in Fallon County, Terry Cameron, a former sheriff, went around with a Geiger counter after one of the radioactive clouds came through in the late 50's and found that it was pegging his counter. Too high to measure. Given such high levels of exposure to radiation, perhaps it is not entirely a coincidence that the National Cancer Institute would later report that cancer for men in Fallon County from 1950 to 1969 had an overall cancer incidence that put it in the top 10% of counties in the United States. (52)
After all these years, it would seem unlikely that this sudden cluster of childhood cancer, which started in 1992, could be linked to fallout in the '50's and '60's. Overall cancer incidence in Fallon County had been declining. Yet, is it entirely impossible that the Strontium-90 that fell on Fallon County decades before is not linked to these tragedies? Is it not possible that the Strontium-90 was laying in wait, irradiating the genes of everyone living in Fallon County for decades, leaving behind a growing accumulation of chromosomal damage, so that the children who inherited these genes were more vulnerable to disease than they otherwise would be? (53) And could it not be that when these children were exposed to another toxin, perhaps from the oilfield, perhaps from some sort of environmental hormone, perhaps from a virus infection, that they were pushed over the edge, from health to disease, when they otherwise might not have been?
Yet perhaps the radiation exposure that is causing this cluster of childhood leukemia is more recent. In April of 1986, about the same time that the Chernobyl nuclear reactor melted down, covering the entire world with its radioactive core, an underground nuclear weapons test designed to develop the Ronald Reagan's Star Wars technology vented radioactive fallout from Nevada, as Janette Sherman reports:
Compounding fallout from the Chernobyl release was a United States hydrogen bomb test, called Mighty Oaks, that went
badly wrong in April of the same year. The Department of Energy test blast released radioactivity from the test shaft into
the atmosphere, and fallout was measured as far away as Burlington, Ontario. (54)
If the fallout from Mighty Oaks made it to Ontario, it is possible, as happened so frequently in the 50s and 60s when nuclear testing was above ground, that a particularly large amount concentrated and fell on Fallon County, perhaps in a rainstorm. The cases of childhood leukemia in Baker would fit well in the time frame of 5 to 9 years later that Archer and Johnson had earlier discovered.
Another possibility is that the children were exposed to natural radium from disruption of geological structures underneath
Baker. As I argued in one of my first petitions, the Williston Basin Pipeline company, a subsidiary of Montana Dakota
Utilities, stores an enormous amount of natural gas underneath the geological formations under Baker. At one time, I heard
that it was a billion dollars worth. There are naturally occurring deposits of uranium throughout Fallon County. There is
even a site north of Baker that was extensively explored as a potential uranium mine. One concern that I have is that
injecting such large amounts of natural gas underneath Baker could increase the amount of radium, a decay product from
uranium, released into the environment around Baker when overpressured natural gas seeps out of the ground, or perhaps
into the water table. I've been told that so much natural gas leaked out of the ground in one spot that it killed a wheat field.
The possibility that overpressured natural gas is leaching uranium and radon out of the ground or into the water table,
exposing the public to radiation, is not an unreasonable concern, I think. Radium-contaminated drinking water was linked to
a cluster of childhood leukemia in Germany. (55) Childhood leukemia has also been linked to nuclear power plants in northern
England. (56)
Comparisons With Other Counties and Nations
Despite the reassurances the residents of Fallon County have been getting from their government and community leaders, besides the childhood leukemia, Fallon County also has a very high incidence of breast cancer and colorectal cancer. According to the Tumor Registry's annual report for 2000, the incidence of breast cancer in Fallon County is 154.3 per 100,000. This compares with a Montana rate of 108 per 100,000, and a national rate of 111.9 per 100,000. (57) Compared with the incidence of childhood leukemia, this may not seem like a large departure from the norm in surrounding populations, but a 30% higher rate of breast cancer involves a lot more people than a 30% higher rate of childhood leukemia. Breast cancer is a major killer of American women, and the high incidence of it in Fallon County flatly contradicts the reassuring statements that Racicot, Spence, and Damrow, have been making.
According to the statistics in the Tumor Registry's report, Fallon County is tied for second with Sweet Grass County, behind only Wheatland County, for the worst incidence of breast cancer in Montana. 13 counties in Montana have half the incidence of breast cancer that Fallon County does.
Half the incidence!
Something in Fallon County, we might reasonably assume, is causing the women living here to have twice the risk of developing the breast cancer as women living in many other counties in Montana. Surely, that says something about the environment in Fallon County, even if the numbers are small.
One thing that Damrow took me to task for was that I was not using data with the same underlying statistical foundations when I interpreted cancer in Fallon County--I was not comparing apples with apples. So, let's compare apples with apples.
The U.S. Health Resources and Services Administration has a web site where you can compare a county with peer counties across the United States to see how it compares with them on a variety of health indicators. (58) You put in a county, and the HRSA computer, using data from the U.S. Census Bureau's projections for 1997, identifies counties similar to it in "frontier status, population size, poverty, and age structure." It turns out that there are 37 counties "like" Fallon County across the United States, including, for example, Dolores County, CO, Kent County, TX, and Beaver County, UT.
The advantage of comparing a county with its peers is that you eliminate many possible explanations for something like a high incidence of breast cancer. If all the peers of a county are high, it probably means that this shared problem has something to do with their "frontier status, population size, poverty, and age structure," or, perhaps, with something like dietary and exercise habits that may go along with these variables. On the other hand, if a county has a higher incidence of breast cancer than its peers, it probably means that there is something unique about its environment that is causing it, some sort of toxic exposure.
Indeed, it turns out, Fallon County's incidence of breast cancer is quite a bit higher than its peers, according to the HRSA web site. Fallon County's peer range for mortality from breast cancer, age adjusted to the year 2000 standard, ranges from 17.4 to 43.7 deaths per 100,000 women. In comparison, Fallon County's mortality from breast cancer is 54.7 per 100,000. So, Fallon County is not just high compared with its 37 peers, it is 11 points beyond their range, ruining the curve for all of its peers across the nation. More to the point, it is 3 times higher than the lowest of its peers.
The question is: If at least some of Fallon County's peers can have such a low incidence of breast cancer why can't Fallon County? Why should women here have to suffer so much more from this disease than their peers elsewhere?
Still, comparing Fallon County with its peers is not the only frame of reference that is useful. Things get really ugly when we compare Fallon County with international incidences. It turns out that the United States, along with many other industrialized nations, has one of the highest incidences of breast cancer in the world. According to the World Health Organization, the United States ranks 16th in the world in 1992 for mortality from breast cancer, behind such countries as England and Wales, Denmark, Canada, and Israel. With a mortality rate of 22.4 per 100,000 from breast cancer, the United States has a rate that is 22 times higher than Thailand's, which is only 1.0 per 100,000 women per year. The U.S. rate is 8.6 times higher than Korea's, 4.7 times higher than China's, and 3.9 times higher than Ecuador's. (59) Against these numbers, even the counties with the lowest incidence of breast cancer in Montana are doing poorly compared with the lowest incidences in the world.
Although a country like Thailand is a third world country, and has quite different demographics, comparing its cancer incidence with a country like the United States is still possible. The problem is that proportionally speaking, there are more young people in Thailand than there are in the United States, and since cancer is primarily a disease of old age, the two countries cannot be directly compared for cancer incidence. The relatively large number of young people in a Third World country would distort the comparison. However, epidemiologists are well aware of how differences in age distributions affect cancer distributions, and they routinely "correct" for age when they compare populations. What the World Health organization did when it drew up this comparison was make the populations of all countries it was studying have the same age distribution, adjusting all of them to what it calls the "world standard." In effect, through a statistical manipulation called "age correction," it compares the breast cancer incidence of 100,000 women who are, say, age 60 in Thailand with the breast cancer incidence of 100,000 women who are age 60 in the United States. With this age correction, the differences in cancer incidences between industrialized nations and third world nations are put on an equal footing, and any difference that remains is not an artifice of different age distributions but a real difference based on a comparison of identical age groups.
In other words, the incidence of breast cancer in Thailand really is that much lower than it is in the United States.
Things can get very confusing because many U.S. cancer rates, including the Montana Tumor Registry's, are age-corrected not to the world standard, but to what the age distribution of the United States was in 1970. This correction makes it possible to compare the cancer incidence of people living in Florida, where many people go for retirement, with people living in Silicon Valley, where there are many young engineers. Yet these rates cannot be compared with international rates. To do that you have to take the raw data and correct it to the world standard.
I say this to make one further point about the incidence of breast cancer in Fallon County, and across Montana. Although the overall mortality rate from breast cancer puts the U.S. 16th in the world in 1992, the incidence of breast cancer for women living in San Francisco, at least until 1992, was the highest in the world, as the National Cancer Institute Reports:
White women in the San Francisco Bay area experienced the highest incidence (of breast cancer) among 162 areas reporting incidence data to the IARC (The International Agency for Research on Cancer), with an annual rate of 104.2 per 100,000 adjusted to the world standard population. (60)
As the NCI says above, the rate of 104.2 is adjusted to the world standard population, so, if we are going to compare apples with apples, we need to see what it is when the raw data is corrected to the same standard the Montana Tumor registry is using. Fortunately (because Congress ordered them to do so) the Centers for Disease Control and Prevention and the Agency on Toxic Substances and Disease Registry did this when they did a review of a study on breast cancer in San Francisco by the Northern California Cancer Center. They used the standards the National Cancer Institute uses in its SEER program, which are what the Montana Tumor Registry has adopted. This is what the CDC concluded:
CDC has confirmed the trends in breast cancer incidence in the San Francisco Bay Area as presented in the Status Report (done by the Northern California Cancer Center). Since 1973 the incidence rates of invasive breast cancer (all races) in the San Francisco Bay Area have been higher than the rates of invasive breast cancer in all SEER areas combined. However, the rate of newly diagnosed breast cancers in the Bay Area peaked at 123.4 per hundred thousand population in 1987 and decreased 12.0 percent to 108.6 per hundred thousand by 1994. Since 1991, there has been no statistically significant difference between the incidence rate in the San Francisco Bay area when compared to the rate in all SEER areas combined or to the rate in all other SEER areas combined (i.e., excluding the San Francisco Bay Area). Breast cancer incidence rates in the United States are higher than in other parts of the world. (61)
So, let me emphasize this, according to the IARC, San Francisco officially had the highest incidence of breast cancer in the world, and that incidence, according to the CDC, at its peak was 123.4 per hundred thousand per year, corrected to the U.S. standard. And the Montana Tumor registry, using the same SEER standards the CDC was using when it reported this statistic, reports that Fallon County has a breast cancer incidence rate of 153.4 per hundred thousand per year.
This is a number that is a full 30 points higher than what was widely reported to be the highest incidence in the world!
Yet, let's get the record straight, after I make a big fuss about cancer in Fallon County, Governor Marc Racicot, Montana Medical Officer Michael Spence, and State Epidemiologist Todd Damrow, all come out and say, in one way or another, that there is not an unusual incidence of cancer in Fallon County! I am left wondering what qualifies as 'unusual' if it isn't an incidence of cancer that is 30 points higher than the highest incidence in the world? Is something like what happened in Libby the only thing that is going to qualify as "unusual" now? Or are we on the road to a place where even that going to become "typical?"
I really would like to know because Fallon County isn't the only county in Montana that has a higher incidence of breast cancer than the San Francisco Bay Area had, and in several of these counties the possibility that these incidence rates are not statistically significant cannot be raised. By my count, using the Montana Tumor Registry's own statistics, there are 14 counties in Montana that have a higher incidence of breast cancer than the San Francisco Bay Area had at its peak. They include: Cascade (138.5), Daniels (152.6), Deer Lodge (152.6), Fallon (154.3), Fergus (128.3), Garfield (129.9), Jefferson (147.3), Missoula (127.7), Musselshell (136.2), Prairie (144.6), Sweet Grass (154.4), Treasure (144.6), Wheatland (191.3), and Yellowstone (127.6). Several of these counties, including Missoula, Cascade, and Yellowstone, have large numbers of people, at least for Montana. DPHHS simply must not be allowed to dismiss the suffering going on in these places as an aberration common to small numbers. Too many counties in Montana have a truly intolerable rate of breast cancer by both U.S. and World standards, and it is time we accepted this unhappy fact, and did something about it.
Fifty years ago the incidence of breast cancer in women was 1 in 20. Now it is 1 in 8. (62) From 1940 to 1980, breast cancer rates increased by 1.2% each year. However, since 1980 the rates have been increasing by 2% a year, though some recent reports have argued that the upward trend is leveling off. Worse than that, breast cancer rates for women less than 35 have been increasing too, and women in this age group have the poorest survival rate. (63) Because of such large changes over such a short time, and because they happened while corporations were rapidly increasing both the number and amount of toxic chemicals released into the environment, environmentalists have argued that the dramatic increase in breast cancer is the result of increasing environmental pollution. Countries that are the most industrialized tend to have the highest incidences of it, and countries that are less industrialized tend to have less.
According to many scientists, heritable factors play, at best, a limited role in breast cancer, although two gene mutations have been identified, BRCA-1 and BRCA-2, which, when present, dramatically, perhaps by 86%, increase a woman's risk of developing breast cancer. Despite this, only between 5 and 10% of breast cancer cases can be attributed to a heritable factor. Furthermore, studies of women who have moved from low-incidence areas of breast cancer to high-incidence areas show that their rates of breast cancer gradually rise to the level of the new country in which they are living. Epidemiologists generally agree that differences in breast cancer rates between different nations are not due to heritable factors. (64)
The risk of developing breast cancer has been shown to vary with the age at which a woman has her first pregnancy, with delaying pregnancy increasing the risk. An early menarche and late menopause also increase a woman's risk, while breast feeding reduces it. (65) These variables point to the possibility that increased lifetime exposure to estrogen increases the risk of breast cancer. If this is the case, exposure to xenoestrogens, or industrial chemicals like DDT, DDE, and some PCBs (66) that mimic the effect of estrogen, likely increase the risk of breast cancer. (67)
In 1992, Frank Falck, Mary Wolff, and associates found a link between exposure to PCBs and breast cancer. They compared chemical residues in the breast fat tissue of 40 women with palpable breast masses, some of whom would have breast cancer and others would simply have a lump. After these women were screened for breast cancer, those who had breast cancer had 50 to 60% higher levels of PCBs, DDT, DDE, and other hydrocarbon-based pesticides than the women who simply had a lump. The women who developed breast cancer had, on average, 1,965 ppb of PCBs in their breast tissue, about 1,000 times higher than what the FDA considers safe in food. (68) This is important because it also turns out that exposure to organochlorines, like PCBs, not only increases the risk of developing breast cancer, it also increases the likelihood that a woman will die of it. (69)
Some scientists have argued that a high fat diet may be a cause of breast cancer. However, despite the blame placed on behavior, it is unclear whether it is the fat that is causing it or the xenoestrogens and dioxin that invariably bioaccumulate in the fat people are eating. A few studies have shown a link between high fat diets and breast cancer, while others, some of them large prospective studies, have found no link. (70)
Although it is a estrogen blocker, dioxin, an extremely toxic environmental hormone, may also play a significant role in causing breast cancer. In 1998, researchers at the University of Birmingham in England exposed pregnant rats to small amounts of dioxin on the 15th day of pregnancy. The dose was small enough that the female offspring were born normal. However, by the time they were 7 weeks old, their mammary glands developed a high number of "of terminal end buds," which 4 different studies had earlier established are directly correlated with breast cancer. (The more buds the more risk of cancer.) The researchers then exposed these rats to a well-known carcinogen. The dioxin-exposed rats developed more breast cancer than rats not exposed to dioxin but also exposed to the carcinogen. (71)
In addition, high doses of radiation, particularly in infancy, but also from puberty through the childbearing years, is a well-known cause of breast cancer. Low doses of radiation are more controversial as a cause of breast cancer, but a possible factor. (72) According to Janette Sherman, radiation might explain a large number of breast cancer cases:
Comparing breast cancer incidence in 1967-75 with a 1951 cohort, Dr. Carl Johnson found a near doubling of breast cancer in Utah women who lived in the fallout path from the Nevada test site. He found 27 cases versus 14 expected cases. (73) Radiation-induced breast cancer susceptibility is increased when exposure occurs in adolescence and early adulthood, (74) and is promoted by hormonal stimulation. (75) Japanese bomb survivors have increased breast cancer. (76) (77)
According to Sherman, contrary to the assumption that the genetic causes of cancer are a given, and are independent of environmental exposures, radiation from fallout may also explain many genetic mutations leading to cancer. A little appreciated fact is that some of the key studies that identified the breast cancer gene BRCA1 were done in Utah, a state that was heavily exposed to fallout from nuclear testing. (78) She writes:
Considering the cancer clusters and the BRCA1 were identified in Utah, did the genetic alterations result from various radioactive nucleotides, carried from the Nevada test site, that rained down upon their homes and farms? We do not know if the BRCA1 gene has been present for generations or not. Did the Utah "clan of one-breasted women" suffer genetic damage from radiation as believed by Terry Tempest Williams and so lovingly described in her book Refuge? (79) In addition to radiation, what other gene-damaging stimuli were active? Were the insults pesticides, used in homes and farms, or where they hormones, or hormone-like chemicals, contaminating the food, fish and livestock of Utah residents?
Two years after I wrote the above paragraph, I was given a copy of a hearing held in 1959. Before we are led to believe that the breast cancer gene is long-inherited, consider the implications of the following, which was know by our Congress: "The genetic injury from both weapon-produced carbon-14 and fission products occurs at the moment the genetic molecule is affected and is the result of absorbed radiation. The actual effect does not appear until this particular gene is found in either the sperm or ovum at the time of fertilization. This event may occur several generations after the initial injury . . . Most of the injury due to fission products will be initiated within a 30-year period. By contrast, after the initial transient increase in atmospheric carbon-14 is past, half the remaining carbon-14 injuries will be introduced over a period of 5,600 years, 75 percent in 11,200 years, and so on. (80) (81)
If this is true, and radiation can cause genetic harm that will lead to cancer generations later, then isn't it possible that the
high rates of cancer in Fallon County, perhaps in combination with other chemical factors, is a lingering effect of a high
exposure to nuclear fallout? There are no obvious explanations for which of these causes may be prevailing in Fallon
County, but the fact that there is such a high incidence of breast cancer within it, by itself, suggests that something is
causing it.
Comparisons With Other Counties and Nations
Fallon County has the second highest incidence of colorectal cancer in Montana, according to the Montana Tumor Registry's October 2000 report. (82) At 70.4 cases per 100,000 people, it is second only to Meagher County in Montana, which comes in at 76.5 cases per 100,000 people. In Montana, 22 counties have half, or less than half, the incidence of colorectal cancer that Fallon County has. One of its immediate neighbors, Prairie County, has an incidence rate of only 9.7 cases per 100,000, which is 7 times less than Fallon County's. Of course, Fallon County is considerably above Montana's incidence, which is 39.8 per 100,000, and the national incidence, which is 43.9 per 100,000.
Again, colorectal cancer is a cancer that is high in industrialized nations and low in third world countries. Death rates from colorectal cancer are highest in New Zealand and Europe and lowest in Thailand, Kuwait, Mexico, Ecuador, Korea, Venezuela, and Panama. According to the World Health Organization, the incidence of mortality for men in the United States from colorectal cancer was 17.2 per 100,000 in 1992. On the other hand, it was 1.6 in Thailand, 2.4 in Kuwait, and 3.1 in Mexico. (83) So, men in Thailand were about 11 times less likely to die of colorectal cancer than men in the United States.
In other words, Fallon County could do a lot better than it is against this major killer. It is spectacularly out of line with neighboring counties, with Montana as a whole, and with the international community. The high incidence of colorectal cancer in Fallon County, as reported by the Montana Tumor Registry, flatly contradicts any assertion that there is nothing unusual about cancer incidence in Fallon County.
According to the Tumor Registry's 2000 annual report, several hereditary conditions are associated with high risk of colorectal cancer--such as familial adenomatus polyposis, Gardner's Syndrome, or Peutz-Jegher's Syndrome. While it is likely that hereditary conditions play a greater role with colorectal cancer than with most other cancers, it remains true, as with other cancers, that the role of hereditary is limited. A variety of studies of migrants show a convergence for colorectal cancer with the destination country when they move from country with a different incidence. For instance within their own lifetimes, migrants to Australia from Greece and Yugoslavia, where the incidence of colorectal cancer is low, ended up having a similar incidence as Australians, where the incidence is high. (84)
However, the Tumor Registry also reports, as do many other sources, that colorectal cancer is "associated with diets high in fat, especially through red meat intake." Then it adds: "Diets low in fruits, vegetables, high-fiber grains, and folic acid contribute to increased risk." In other words, when it comes to colorectal cancer, the victim is to blame.
Few cancer risks are as well established, at least in the public's mind, as the links between colorectal cancer, a high fat diet, and red meat consumption. Animal experiments have shown that a high fat diet promotes large bowl tumors. (85) In a study of about 90,000 nurses, scientists found that those nurses who where highest in consumption of animal fat had twice the risk of developing colon cancer as the nurses who ranked lowest in consumption of animal fat. (86)
Yet perhaps, despite its wide acceptance, something is confounding this link between a diet high in animal fat and colorectal cancer. If eating red meat is a major risk factor, why is it that Fallon County's neighbor to the south, Carter County, and its neighbor to the north, Prairie County, both of which are surely populated by people as likely to consume beef as people in Fallon County, are both vying to have the lowest incidences of colorectal cancer in the state? With colorectal incidences of 11.6 for Carter County and 9.7 for Prairie County, both counties have a risk from colorectal cancer that is about 7 times lower than Fallon County's. How can there be such dramatic differences in the incidence of colorectal cancer between counties with such similar diets?
Samuel Epstein agrees that diets high in animal fat may be linked to various cancers, but he argues that it isn't necessarily the animal fat, itself, that is the problem. Instead of blaming the victim, we should also consider the effect of environmental exposures, as he writes:
Much has been made of the relationship between modern eating habits--particularly high caloric intake, high consumption of animal fats, cholesterol, dairy products, and meat, and the low consumption of grain and fiber--and the twentieth century cancer epidemic. On the basis of indirect evidence, it has been suggested that a low-fiber, high-fat diet increases the risk of cancer of the colon and possibly of other cancers, including breast, while a high-fiber, low-fat diet protects against these. As far as dietary fat is concerned, there is no question that a very wide range of environmental carcinogens, particularly pesticides and industrial chemicals, are fat-soluble and are likely to accumulate in the food chain. So the more animal, dairy products, and other fats you eat, the greater will be your intake of these fat-soluble carcinogens. (87)
Epstein goes on to argue that the value of a high fiber-diet for preventing cancer may be only be because it reduces "transit time" in the colon, reducing the contact time of dietary carcinogens with sensitive tissues. The faster a toxic chemical goes through the body, the less the body is exposed to it, and the less harm it can do. According to Epstein there is little evidence that animal fat causes cancer and much evidence that the pesticides and other industrial toxins that accumulate in animal fat cause cancer. So, although colon cancer is not ordinarily thought of as an environmental cancer, it may in fact be strongly associated with environmental exposures.
Along with a high fat diet, obesity has been identified is a key risk factor for both breast and colorectal cancer, but is it being overweight that is the cause or is it that both obesity and cancer are caused by exposure to environmental hormones? Janette Sherman thinks we should take very seriously the possibility that obesity is not just a behavioral problem but an effect of industrial pollution as well:
It appears that obesity confers not only an increased risk for breast cancer, but an increased risk for heart disease, hypertension, and degenerative arthritis. Is the obesity link to breast cancer due to toxic chemicals in our diet or to having a bountiful layer of fat on ones' body? Forming a reservoir, obese people have a greater volume of adipose tissue in which to store toxic chemicals than do thin people, and fatty foods are efficient carriers of carcinogenic and hormonal substances.
More basically, do the hormonal chemicals in our diets promote both breast cancer and obesity? After all, hormones are administered to meat animals to promote growth and weight gain. Why should humans expect to not respond similarly to such chemical stimuli? . . . In addition to purposely administered hormonal growth agents, meat and dairy products may contain whatever chemicals the animals ate, were injected with, were applied to their bodies, or were used in their barns. Fish and shellfish may contain whatever toxins are in the water in which they grow, whether in the wild or in fish pond culture. Vegetable oils absorb whatever pesticides were applied to the plant and soil. Additionally, plastic containers and cans lined with plastics can leach contaminants into the food stored therein. (88)
We know from the hormones that we purposely feed animals that they make them fatter than they otherwise would be. If these hormones didn't increase profits, producers wouldn't use them, would they? And so, why isn't it possible that the same hormones that are feed to animals to make them fatter are doing the same thing to the people that eat them? Making them fatter? And if this is possible, why wouldn't it also be possible that other environmental hormones in the environment coming from pesticides and such are doing the same thing to people? Making them fatter? Lectures from public health departments about the risks of being obese may well be missing the point. Being obese may be as much a symptom of exposure to environmental hormones as cancer is.
To a certain extent, the cancer differences between Fallon and Carter Counties can be interpreted as a controlled study for pesticide exposure. Although the consumption of red meat in the two counties is likely identical, the use of pesticides is not. Compared with the rest of southeastern Montana, a relatively large proportion of Fallon County is committed to farm production. On the other hand, with one of the lowest densities of people per square mile (0.5) in the state, Carter County is second only to Garfield County in wide open space. Because of its low rainfall and poor soil, most of Carter County is devoted to beef production on large ranches. Use of agricultural pesticides is high in Fallon County, at least for eastern Montana, while, aside from a few farms, it is low in Carter County. Because of these different agricultural practices, it is likely that beef raised in Fallon County would be exposed to pesticides, while it is less likely in Carter County. (89) Indeed, there are probably huge stretches of Carter County that have never had even a slight exposure to pesticides. And so, most of the cattle raised in Carter County probably have low concentrations of pesticides in their fat.
In both Fallon and Carter Counties, in contrast to what typically happens nationally, a significant portion of the beef people consume within their respective counties was raised on local farms and ranches and processed by local meat cutters. Although some meat produced in Fallon County might be consumed in Carter County, and vice versa, the proportion probably is not too high, given the distance between the two county seats. So, to a larger extent than is the national practice, meat consumed in Fallon County was raised in Fallon County, and meat consumed in Carter County was raised in Carter County.
So, if it is the pesticides that accumulate in animal fat that is causing colon cancer, and not the animal fat itself, then one would expect there to be a difference in colon cancer between the two counties--as there, indeed, is. Although the results have not been consistent, farming and pesticide exposure have been linked to colorectal cancer. For example, a study of farmers in central Italy found a link between farming and colorectal cancer, in particular fruit farming:
An association between fruit farming and colon cancer was detected in this study. Physically active jobs, including farming, have previously been associated with a decreased risk of colon cancer. Siemiatycki et al., (90) however, investigated several site-exposure combinations, reported an increased risk of colon cancer in association with grain dust exposure. Increased risks were also reported among agricultural extension agents (91) and among soil and forest conservationists. (92) The increased risk of rectal cancer among licensed pesticide users in this study was an unusual finding; previous studies reporting a similar association are lacking. For both colon and bladder cancer there was a strong excess among the subjects employed in fruit production, and the risk was higher among the younger farmers. Many chemicals are normally employed in this agricultural practice but, in the absence of specific exposure information, it is difficult to determine the role of any specific agent. (93)
Overall, many studies have found a link between certain kinds of cancer in farmers and exposures to pesticides used in farming. This is despite the fact that farmers really should be healthier than the general population, as the authors of a meta-analysis of some of these studies observed:
The mortality experience of farmers is favorable in terms of all causes, all cancers combined, and ischemic heart disease. The low rates for cancers of the lung, esophagus, and bladder, as well as heart disease may be explained by the low prevalence of smoking among farmers. In addition, the physical demands on farmers may account for their low body fat and high levels of physical fitness, which in turn may contribute to lower risks for heart disease and colon cancer. Dietary factors (such as high intake of fresh fruits and vegetables), residence in areas with little air pollution, and selective migration may influence the deficits of cancer observed, but these factors have not been evaluated among farmers.
In contrast to the deficits for most major disease categories, farmers had significantly elevated risks for leukemia,
multiple myeloma, Hodgkin's disease, melanoma, and cancers of the lip, prostate, and stomach. Mortality from
non-Hodgkin's lymphoma and cancers of connective tissue and brain was increased in most studies, although the MRR
estimates were not significant. These tumors do not fall into an obvious grouping, other than the fact that they are not
strongly associated with smoking. They vary in frequency, history, and prognosis. The excesses among farmers for a few
specific cancers, against a background of low risks for most cancers and nonneoplastic disease, suggests a role for
work-related exposures. These patterns may have broader public health implications, since several of the high-rate tumors
among farmers also appear to be increasing in the general population of many developed countries. (94)
Another study may be directly relevant to the differences in cancer between Carter County and Fallon County. Dina Schreinmachers did a study of cancer incidence in areas with high rates of spring wheat production compared with areas of low rates of spring wheat production in Minnesota, North Dakota, South Dakota, and Montana. She found that in areas where use of chlorophenoxy herbicides, such as 2,4-D, was likely high, the incidence of certain cancers was also high. The more the relative acreage of wheat increased in a county, the more the incidence of cancer increased.
The cancer sites that showed positive trends of increasing cancer mortality with increasing wheat acreage were
esophagus, stomach, rectum, pancreas, larynx, prostate, kidney and ureter, brain, thyroid, bone, and all cancers (men) and
oral cavity tongue, esophagus, stomach, liver and gall bladder and bile ducts, pancreas, cervix, ovary, bladder, and other
urinary organs, and all cancers (women). Rare cancers in men and women and cancers in boys and girls were studied
comparing counties above and below the median of wheat acreage per county. There was increase mortality for cancer of
the nose and eye in both men and women, brain and leukemia in both boys and girls, and all cancers in boys. These
results suggest an association between cancer mortality and wheat acreage in counties of these four states. (95)
There is one thing, though, that may confound the contrast between cancer in Fallon County and cancer in Carter County.
Fallon County has a large, aging, oilfield and Carter County does not. It might be that the high incidence of colorectal
cancer in Fallon County is confined to the area immediately within the oilfield, and there is no link to pesticide use. It may
only look like there is. Nevertheless, to rule out one possible cause of cancer in Fallon County, it would be a good idea to
test the animal fat of animals raised in Fallon County and compare it with animals raised in Carter County. It may be that an
environmental hormone, such as dioxin, is present in quantities sufficient to explain the high incidence of leukemia, breast
cancer, and colorectal cancer in Fallon County, but not in Carter County.
Dioxin is likely to be found in animal fat in Fallon County, at least to some extent. Although Agent Orange is generally thought to have been used only in Vietnam, it was also used here, in Montana. As a child, I saw my family, our neighbors, and the county use a lot of Agent Orange throughout the 1960's. It came in orange drums marked "2,4-D and 2,4,5-T." People didn't call it "Agent Orange"then, but that is what it was, the same chemical used in Vietnam. On one of the drums I even remember seeing the word "dioxin" listed in the ingredients. I have not found any remaining empty drums on our place disclosing dioxin on its list of ingredients, but I distinctly remember seeing the word "dioxin" once on a barrel when I was in High School, maybe college. It was surrounded by a lot of other words, like "tetra" and "chlor," but there it was. I remember it because I saw it after I had just read an article in Mother Jones about what Agent Orange had done in Vietnam. I remember wondering if the drum that had dioxin listed on it was same stuff in the Mother Jones article. Young and innocent, I told myself that our "dioxin" couldn't be like that "dioxin" because it had all these strange words around it that made it different.
Now I know better.
And now I'm wondering if the widespread use of chemicals like Agent Orange in Montana isn't coming back to haunt our health.
The incidence of different cancers can be used to interpret different environmental exposures. When a county has a low incidence of lung cancer and a high incidence of breast cancer it means something about the specific environment of that county, what people there are breathing, drinking, and eating. Whether or not that county has an environmental problem does not necessarily hinge on whether the overall incidence of cancer in that county is high but on the incidences of particular cancers. This point needs to be made because Fallon County, consistent with its rural status, does have a relatively low overall incidence of cancer. Contrary to what the department seems to be assuming and what some people in the community are saying, this does not, however, mean that Fallon County could not have an environmental problem. As many scientists agree, cancer is not one disease but many different diseases, each with different causes. Although there are toxins, such as dioxin, that contribute to many different kinds of cancer, different cancers are usually caused by different toxins.
After Clair Johnson published several articles in the Billings Gazette about my petition to the ATSDR, the Fallon County Times published an article by David Espeland, the manager of the Fallon Medical Complex. Under the headline, "Fallon Medical Complex Responds to Cancer Scare," Espeland reassured the residents of Fallon County that we didn't have a cancer problem. To prove his point he referred to an annual report that the Montana Central Tumor Registry had released in October of 2000. (96) Following the example of Governor Racicot, Dr. Spence, and Todd Damrow, Espeland told the residents of Fallon County that:
Last month the (Tumor) Registry published their 2000 annual report, covering the five years spanning 1993 to 1997. In
this report, Fallon County does not display remarkably higher numbers than those of Montana and the rest of the nation.
For instance, the overall rate of cancer in Fallon County 3.5 per 1,000 people. This rate is 3.6 in all of Montana and 4.0
in the nation. (97)
He went on to list the incidences of various cancers in Fallon County. Then, ignoring all the emphasis I had been placing on
childhood leukemia, he simply reported the incidence of leukemia for all ages, as if we couldn't have a childhood leukemia
problem because it wasn't reflected in the overall rate:
Leukemia, a focus of the Billings Gazette articles, has an incidence rate of 0.14 per 1,000 people in Fallon County (versus 0.10 in both Montana and the nation). There are also many cancers that had a 0% rate in Fallon County during the years 1993-1997, including cervical cancer, skin melanoma and stomach cancer.
All in all, cancer is a very emotional subject. Everyone has known one or more people who have been plagued by cancer
and at times we feel surrounded by it. However, the empirical data indicates that overall Fallon County residents are not
realizing a significantly higher incidence of cancer. Only after comparing our rates with those of Montana and the rest of
the nation does this become apparent.
I did not have the Tumor Registry's annual report when I first filed my petition. When I did finally get chance to look at it,
I was not nearly as reassured by it as Espeland was. In fact, as I read it, the statistics in it directly contradict what Espeland,
Racicot, Spence and Damrow have been saying about cancer in Fallon County. While Fallon County does, indeed, have an
unremarkable overall incidence of cancer, it has a very high incidence of breast cancer and colorectal cancer. The relatively
"normal" overall incidence of cancer in Fallon County is only possible because Fallon County has a low incidence of all the
cancers associated with smoking.
According to the Tumor Registry's report, the incidence of lung cancer in Fallon County is 36.5 per 100,000. This compares with 54.1 per 100,000 for Montana and 56.2 per 100,000 for the nation. Fallon County has about half the incidence of lung cancer that the 4 worst counties in Montana do. (98)
It doesn't stop there. There are 7 cancers associated with smoking--oral, esophagus, pancreas, larynx, lung, bladder, and kidney. For men, according to the National Cancer Institute, 90% of the cases of oral cancer are attributable to smoking, as are 77% of the cases of esophagus cancer, 26% of pancreas cancers, 79% of larynx cancers, 89% of lung cancers, 43% of bladder cancers, and 45% of kidney cancers. (99) As the National Cancer Institute vehemently insists, smoking is by far and away the largest cancer risk that the American public faces. Not smoking substantially reduces the risk of getting cancer.
Apparently, if one can conclude anything from the incidences of these cancers, many residents of Fallon County are heeding the National Cancer Institute's advice on smoking. According to the Tumor Registry's annual report, the incidence of bladder cancer in Fallon County is half the national rate, kidney and renal cancer are a third, oral cavity and pharynx are a half, and pancreas is a couple of points lower. (100) Taken all together, the low incidence in Fallon County of the cancers associated with tobacco smoking substantially lowers its overall incidence of cancer.
Saying that we can't have a cancer problem in Fallon because the overall incidence is low is like saying that the ground
water can't be contaminated because people aren't smoking. The logic is twisted, distorted, and fallacious. Although Fallon
County has an overall incidence of cancer that puts it in the midrange of counties in Montana, incidence rates for breast
cancer, leukemia, and colorectal cancer are quite high, and these high rates, I think, strongly imply that Fallon County does
not have a clean and healthful environment. If the residents of Fallon County have made the effort to protect their health by
not smoking as much as people in other counties, polluters should not be entitled to make up the difference, taking lives that
otherwise would have been spared by healthful behavior.
While I was trying to figure out what was happening in Fallon County, I went surfing through the Montana Department of Public Health's web page. (101) On the web page that deals with health planning you can bring up the "health" profiles of various counties one at a time. When I brought up Fallon County's, I discovered that 26% of the births in Fallon County had some sort of birth abnormality. I was caught between being horrified and convinced that it couldn't possibly be true. All the babies that I knew seemed normal.
Trying to get some sort of perspective on how large Fallon County's problem with birth defects was, I started bringing up the health profiles of all the other counties in Montana. I was even more amazed because there were 10 counties with high incidences of birth defects, and all but 1 of them, Liberty County, which is up by the Canadian border, were clustered around each other in southeastern Montana. Some of these counties even had worse incidences of birth abnormalities than Fallon County did. These are the counties in southeastern Montana that have elevated birth abnormality rates: Bighorn 16%, Carter 21%, Custer 34%, Dawson 17%, Fallon 26%, Garfield, 31%, Powder River 25%, Prairie 34%, and Rosebud 20%. Except for Liberty, which had 15%, all of the other counties in Montana were either 10% or lower, with most of them coming in around 7%. So, counties in southeastern Montana had about 3 times as many abnormal births as counties in the other 3 quadrants of Montana.
This seems very strange to me. While there are significant ecological, climatic, economic, and cultural differences between western and eastern Montana, there are almost no such differences between northeastern and southeastern Montana. Basically all of eastern Montana is pretty much the same--an endless prairie only lightly populated, a few farms and ranches here and there, and every 50 or 60 miles a small town that supports them with schools, hospitals, county government, a Main Street shopping district, and equipment dealers. Toss in a couple of coal mines and an oilfield or two, and you have eastern Montana. There are no industrial, cultural, or ecological differences between southeastern and northeastern Montana that I can think of that would explain such a large differences in birth abnormalities.
It is almost like DPHHS misplaced a decimal point when it counted birth abnormalities in southeastern Montana. In fact, moving the decimal point one step to the left would bring southeastern Montana in line with national reports of birth defects, which occur at a rate of 1 in 33 births. The problem is that DPHHS insists that it put the decimal point in the right spot.
When I discovered the high incidence of birth defects, I contacted DPHHS and asked them for an explanation. This is the response I got from Bruce Schwartz, via email:
Dear Dr. Sikorski;
I apologize for not responding sooner. I took some time off last week and neglected to change my outgoing phone message to reflect this. I am responding to your inquiry via e-mail rather than phone so my response can be as complete as possible.
Your inquiry about the nature of reported birth outcomes in certain south-eastern Montana counties is timely; the matter came to my attention six or seven months ago and I have been looking for an explanation since then. The statistics to which you refer are reported in the Montana County Health Profiles and are labeled "Percent of Newborns with any Anomaly or Abnormality." The observations are for a five-year period (1993-1997) and we cannot yet say whether this observation would be similar for a longer or different period of time.
This category is very broad, including eight defined abnormalities and fifteen defined congenital anomalies. In addition to the defined abnormalities and anomalies, birth attendants are allowed to list one "other" abnormality and up to seven "other" congenital anomalies. These "other" categories are open-ended (i.e., the attendants can write anything they choose in the space provided) and do not easily lend themselves to analysis. Frankly, I personally find such a broad category to be of little use, since the reader is not told what it contains. At best, it can be thought of as an index number, calling attention to a possible problem.
As you can imagine, this analysis is slow-going, since analysis requires visual inspection of the records. We began by listing the number and percent of all anomalies and abnormalities, by type, for a multi-county region in southeastern Montana (including Fallon county) and the remainder of the state. Preliminary examination of these tabulations showed no difference between the southeast and the other counties except in the "other abnormalities" category. Observations in this category of abnormalities showed that most were remarks about the umbilical cord. I am not sure that these cases represent a serious matter. Remarks about the nuchal cord are difficult to interpret.
One of my colleagues (who has a doctorate in Endocrinology and Reproductive Physiology, and is thus much more informed on such matters than I), is unable to determine whether these comments indicate a health problem or just represent observations of relatively harmless phenomena. This phenomenon may be--at least in part--a reporting phenomenon, a matter of what different reporting facilities feel is worth reporting.
I have been consulting with the state medical officer on this matter and intend to continue the investigation until we have a satisfactory answer (assuming one can ever be found). We are continuing to examine a number of alternative theories, involving medical and non-medical risk factors, but have found no clues so far. I am told by the State Medical Officer that abnormalities of the newborn are generally not preventable and are linked to conditions of the delivery, not to environmental exposures. So far we have found little evidence to support any plausible biological theory explaining these observations.
I appreciate your interest in this matter. As I said, I have been aware of it for several months, but we also rely on citizen
reports to help focus our investigations.
Sincerely,
Although DPHHS has at least recognized the existence of this problem of birth abnormalities, it does not seem to be taking it as seriously as I would like. Furthermore, the claim by State Medical Officer Mike Spence that "abnormalities of the newborn are generally not preventable and are linked to conditions of delivery, not to environmental exposures" seems to me highly contestable. I doubt very much the implied charge that doctors in southeastern Montana are more incompetent than in the rest of the state, and it seems likely to me, given all the developmental problems associated with environmental hormones, various toxic chemicals, and radiation, that birth abnormalities would be strongly associated with environmental exposures.
For instance, scientists were recently alarmed to discover that not only has the human male sperm count declined dramatically over the last couple of decades, it is increasingly genetically damaged as well, as the Montreal Gazette recently reported from a conference on male-mediated developmental toxicity:
Scientists from around the world are alarmed by a dramatic increase in genetically damaged human sperm - a trend that
is not only causing infertility in men, but also childhood cancers in the offspring of those who can reproduce.
It's now estimated that up to 85 per cent of the sperm produced by a healthy male is DNA-damaged, a leading authority
on the subject revealed yesterday at an international conference being held in Montreal.
"That's very unusual," said John Aitken, head of biological sciences at the University of Newcastle in Australia.
"If you were to take a rat or a mouse or a rabbit, usually more than 80 per cent of their sperm would be normal."
For the last 20 years, scientists have known about declining sperm counts. But researchers are now learning that the
quality of human sperm is steadily eroding, and might be causing birth defects as well as brain cancer and leukemia in
children. (102)
Nevertheless, despite wrongly dismissing environmental exposures as a possible cause, at least DPHHS seems to be willing
to acknowledge that birth abnormalities in southeastern Montana appear to be a real phenomenon, and is not just the effect
of a misplaced decimal. Against DPHHS's apparent complacency, I would argue that even if the birth abnormalities that are
appearing disproportionately in southeastern Montana are only on the umbilical cord, and the baby seems unaffected, it may
be only be because the effects of the abnormal umbilical cord had not appeared yet. Umbilical cords are important. They
are what supply the fetus with nutrients, oxygen, and waste disposal. They also are a barrier separating the mother's immune
system from the baby's, protecting the baby, who has a cellular identity that is distinct from its mother's, from the mother's
immune system. Without the mediation of the umbilical cord, the mother's immune system would identify the baby as
"other," a foreign element, and attack it. If that crucial supply line or barrier were reduced or distorted somehow, especially
during a crucial stage of development, it could have a subtle, though possibly significant, effect on the lifetime health of the
baby. Perhaps the harm done would not show up until years later in reduced intelligence, behavioral problems, poor motor
coordination, cancer, or a chronic disease such as asthma.
Despite the huge area involved, a common environmental cause might explain the abnormalities. Southeastern Montana is downwind of Coal Strip, with all of its potential for toxic emissions, such as mercury. It is also downwind of the refineries in Billings. Most chilling of all, and perhaps most likely, this whole area where the birth abnormalities are concentrated is also where, according to the map I saw in The Fargo Forum, the highest levels of Strontium-90 were recorded after open air nuclear testing stopped. Strontium-90 has a half life of 28 years. It has been a long time since the nuclear testing stopped, but not long enough for all the Strontium-90 that fell on southeastern Montana to have decayed to negligible levels. Perhaps there is still enough left throughout the region to leave subtle traces of its presence on the umbilical cords of newborn babies.
I have no idea if that is the right explanation, or whether the birth abnormalities are linked to the high incidence of childhood leukemia, breast cancer, and colorectal cancer in Fallon County, but I do think that the DPHHS should exert every effort it can toward finding out why there is such a high incidence of birth abnormalities, however minor, in southeastern Montana.
If Fallon County has a problem with childhood leukemia because of toxic exposures, I expect that it would have other problems in children too. Many people have remarked to me about the number of children in Fallon County on Ritalin, a drug used to treat what is called Attention Deficit Hyperactivity Disorder, or ADHD, for short. One neighbor blames it on the schools. The teachers, I am told, are pushing the drug because of financial incentives. While financial motives may contribute, I am nevertheless struck by the number of children who actually seem to need the drug. Looking at the children who I know are on the drug, I don't think that it is being unnecessarily prescribed. The eyes of these children are wild, sometimes hauntingly inhuman, and their behavior is uncontrolled, sometimes cruel-at least in my experience. So why, suddenly, when we could get by without Ritalin for centuries, are children in Fallon County in such need of it? I think that DPHHS, while it is considering the high incidence of childhood leukemia and birth abnormalities in Fallon County should also consider the possibility that these problems may be far from being the only risks to children in our community.
Subtle health effects on children should be especially considered because dioxin, and all the chemicals--like PCBs and furans--that resemble it, are amazing toxins, capable of harming children in many different ways. They are all what the EPA calls environmental hormones. This means that they are toxic not just because they damage or kill cells, or because they attack DNA, causing mutations that can lead to cancer or birth defects. Instead, they are toxic because they disrupt the information systems that regulate the body. Unlike conventional toxins, environmental hormones act like pieces of information once they get inside the body. Because the body responds to them as information, it doesn't really matter how much of them are there, only that they are there. They tell the body to do something. Letting dioxin or PCBs get into someone's body is like altering the code of a computer program. Changing one word in a line of code, like putting an "and" where an "or" is, messes up everything. Think of all the problems the Y2K bug might have done, how that just in the United States hundreds of billions of dollars were spent just changing the way computers read dates. Because one bit of information can do so much, you cannot assume that just because you can't measure the amount of PCBs or dioxin present in something it is safe.
Perhaps modern science has placed too much responsibility upon our genes for what we are. Genes, it would seem, are responsible for everything we are or suffer from--inherited diseases like cystic fibrosis, predispositions to cancer, appearance, mental and physical abilities, and about everything else. In fact, genes are only partially responsible for all of this. Genes may be the blueprint, the design of what we are, but they do not exclusively determine how we are built. The building process, and what we become because of it, are shaped by many things that have nothing to do with genes. The hormonal environment the fetus, and then later the child, develops in is crucial.
As Colborn, Dumanoski, and Myers point out, human intelligence is affected as much by the amount of thyroid hormone reaching the brain during a crucial phase of development as it is by genes. A young man may develop testicular cancer not because of his genes but because of abnormal levels of hormones in his mother's womb. When hormonal messages are disrupted early in life, they disrupt everything afterwards. Colborn, Dumanoski, and Myers use this kind of metaphor:
Imagine what would happen if somebody disrupted communications during the construction of a large building, so the
plumbers did not get the message to install the pipes in half the bathrooms before the carpenters closed the walls. Imagine
that the wrong instructions arrived when the program for the climate control system was being set up, and the building
thermostat was fixed at eight-five degrees rather than sixty-eight. Imagine what it would mean if, through a
communications mix-up, the high-rise ended up with only one elevator instead of eight. (103)
In other words, the construction of anything is as important as the blueprint, probably more so. Genes are not destiny,
hormones are. As the body's messengers, hormones are the instructions that activate genes, turning merely potential
instructions into something that has real consequence. If the fetus is bathed in counterfeit hormones during development,
and it is falsely instructed to do things it would not normally do, it is going to turn out much different than its genetic
blueprint says it ideally would.
While the body can repair damaged DNA, which will lead to cancer, it cannot tell the difference between environmental hormones, which are dysfunctional, and natural hormones, which are necessary. It reads them both the same way. Evolution has made cells very responsive to hormones. Because there were no artificial hormones in the environment, at least not until the industrial age, cells have no way of telling if the hormone is a real hormone, naturally and appropriately arising in the body, or not. If something fits into a cell's hormone receptors, the cell reads it as information and responds accordingly. If it is a false hormone, the body has no way of knowing it is being damaged. Because of this vulnerability of the cell, the toxicity of environmental hormones, and the threat they pose to children, is completely unlike that of most other carcinogens.
Because hormones are information, and will affect body processes in very small amounts, environmental hormones can cause damage in vanishingly small amounts. Conventional toxins, like arsenic and cyanide, are measured and regulated in parts per million. Dioxin, on the other hand, is typically measured and regulated in parts per trillion, sometimes in parts per quadrillion. The EPA has had to go to such extremes because it has found no threshold below which dioxin does not have an effect.
Indeed, quite unlike conventional toxins, environmental hormones can actually be more toxic in smaller quantities than they are in larger ones. Contrary to all expectations, according to Fred vom Saal, the toxic effects of exposure to environmental hormones will, within a certain range, increase with smaller doses and decrease with larger ones. (104) The result, once plotted on a graph, looks like an U, with increased toxic effects on the top and decreased effects on the bottom. Unlike conventional toxins, it is not a smooth linear line rising upward, the larger the dose the more toxic the result.
Because a small dose can be more toxic than a larger dose, dilution, while preferable, is not the solution to pollution, at least not the kind of pollution environmental hormones pose. Once environmental hormones are in the environment, no matter in how small a quantity, they can bioaccumulate until they become a problem. Both dioxin and PCBs, accumulate in body fat, remaining in the body for many years without being broken down or excreted. As a result, they can increase in concentration many thousands of times as they go up the food chain. Because of this, no amount released into the environment is safe.
Concern about the toxic effects of dioxin and PCBs would be moot if in fact the average person was not accumulating enough of them to have toxic effects. Not long ago, the EPA released a draft of its reassessment of dioxin. According to the studies that it cited, the typical American eats 119 pg/day of dioxin (TEQ) (105). (Of that 37 pg comes from eating beef, 24 pg from dairy, 17 pg from milk, 13 pg from chicken, and 12 pg from pork.) (106) That amount, the EPA has found, puts the typical American close to the threshold where dioxin will have what it considers a toxic effect. The average level of dioxin in a middle-aged person is 9 ng/kg, or 9 parts per trillion. However, according to the EPA, the immune system begins to be suppressed at 7 ng/kg, a mere 2 units less than what most of us have. At 14 ng/kg, which is only 5 units more than average, human glucose tolerance is altered and testes size is decreased. At 19 ng/kg, monkeys, which are good laboratory models for humans, show learning disabilities. (107)
According to the EPA's most recent reassessment, the dioxin present in the average American's body is enough, in the worse case scenario, to cause as many as 1 in 100 people to develop cancer--a much higher incidence than any other carcinogen the EPA has ever studied. (108) In other words, according to the EPA, the average American is not only close to the threshold where environmental hormones will begin to affect their health, they are actually over it.
However, as is usually the case, things may be even worse than the EPA is saying they are. According to a group of 40 scientists who meet under the auspices of the World Health Organization, dioxin is 2 to 10 times as toxic as it had seemed in 1990, when the EPA was doing its first reassessment of dioxin. (109) Even worse, according to a group of German scientists, the cancer hazard from dioxin for people living in industrialized countries may even go as high as 1 in 10. (110) Of course these findings are the product of a relatively small group of scientists, and remain tentative, while the EPA's reassessment reflects the consensus of the entire scientific community, at least at the time it was written. (111)
Whatever scientists eventually agree it is, the level of threat that dioxin poses to public health, and in particular to children, is astonishingly high. Even more astonishing is the small amount of material that is causing all this harm. According to the EPA, all of this harm is being caused by about 3000 grams (or 6.6 pounds) of dioxin TEQ per year. (112) Imagine that! The equivalent of a little more than six pounds of material spread over the entire United States, a concentration so diluted as to be inconceivable, and yet scientists believe it is causing possibly as many as 120,000 cancers a year. Every year that's more than twice the total number of Americans who died altogether in the Vietnam war. (113) As we saw above, since environmental hormones act in entirely different ways than conventional toxins, they can be toxic in even very small concentrations. More than that, because they concentrate and persist in fatty tissues, PCBs biomagnify dramatically as they go up the food chain. By the time PCBs move up the food chain in the Great Lakes, for example, starting with phytoplankton and zooplankton, which larger species like mysids eat, which are eaten by smelt, which are eaten by trout, which are eaten by herring gulls, they can be concentrated 25 million times. (114) PCBs might not be detected in the lake water, but by the time they reach the top of the food chain they become a deadly threat to species like the bald eagle.
In other words, because PCBs can be biomagnified so many millions of times, the differences between concentrations of only an order of magnitude or so, like the differences between 2 ppm, 50 ppm, and 500 ppm, are almost irrelevant. Released into the environment, PCBs end up becoming concentrated in human fat tissue, very possibly in concentrations large enough to cause cancer, whatever their original concentration.
Hormones, Disabilities, and Behavior
However, as Theo Colborn insists, cancer is not the most threatening aspect of environmental hormones. The other effects are potentially much more of a threat to our civilization because they affect the development of children, permanently altering their bodies. These effects include immune system dysfunction, endocrine disruption, declining sperm counts, rising infertility, diminished mental capabilities, birth defects, and behavioral problems. (115)
According to a study published in the New England Journal of Medicine by Joseph and Sandra Jacobson, children exposed to low levels of PCBs in the womb have lower IQ's, reduced reading comprehension, attention problems, and reduced memory. (116) The children who had these problems were exposed when their mothers ate 2 to 3 meals a month of fish from Lake Michigan for at least 6 years before becoming pregnant. The greatest deficits were in the children whose mothers ate the most fish. They were twice as likely to be two years behind their peers in word comprehension. They also had more coordination and behavioral problems.
Significantly, how much fish the mothers ate while the children were in the womb was not as important as the total amount they ate in the years before becoming pregnant. This means that mothers cannot protect their children by controlling what they eat during pregnancy. Cumulative lifetime exposure to mothers is the deciding variable. A study of North Carolina children also showed that those with the higher levels of PCBs in their bodies did worse on tests requiring fine motor coordination.
Reproductive dysfunction also is associated with exposure to PCBs and dioxin. Boys in Taiwan whose mothers consumed PCB-contaminated rice oil while they were in the womb during 1979 developed smaller penises than other boys who were not exposed. (117) Some species of wildlife are having similar problems. Alligators exposed to environmental hormones have penises one-third to one-half normal size, and are having problems reproducing because of it.
According to more than 60 studies, the sperm count in men world wide has declined dramatically over the last 50 years. (118) In most of the studies, the decline approaches 50%. Other, more recent, studies say that it is more than that. A long series of studies have shown that exposure to environmental hormones decreases testes size, decreases the size of accessory sex organs (penis, prostate, seminal vesicle), lowers sperm count, lowers testosterone level, decreases sexual behavior, and causes abnormal testes. According to one study, these kinds of effects can come from a single dose while the fetus is developing. (119) In this study on rats the dose was small enough that the males showed no effects at birth. However, at puberty, the sex organs of the exposed rats were smaller, testosterone levels were lower, and sperm counts were low. Furthermore, when the dioxin-exposed males were placed with females, they took much longer to mate, they had to try several times to achieve ejaculation, and they took a lot longer than the controls before attempting to mate again. When these same males were placed with other males, they assumed a female mating position.
The effects on men exposed to environmental hormones are similar to laboratory animals. Men exposed to dioxin in accidents or from occupational exposure consistently complain of reduced sex drive and difficulty with erection and ejaculation. (120)
Disruptions in the reproductive system may even be more serious for women. In 1820, according to Stephanie Coontz, the average age girls reached physical maturity was 16. In 1900 it was 14, and in 1940, when the chemical industry really took off, it was 13. (121) Now, according to a recent study, breast or pubic hair development begins at 10 to 10½ for whites and 9 for African Americans. (122) That's the average age. The same study also found that 1% of white children and 3% of African American children had either breast or pubic hair development at the age of 3.
The age of 3!
At least one study has linked early onset of puberty in girls to exposure to PCBs. This study monitored PCB levels in the blood and breast milk of hundreds of pregnant women. After birth, the development of the children was monitored. Girls with the highest prenatal exposure to PCBs entered puberty 11 months earlier than girls with the lowest exposure. It is important to note that this study was of women and children exposed to only normal levels of PCBs. (123) This means that there is no safety margin left. Any additional amount of PCBs released to the environment, no matter how small, could push the average age of puberty downward.
Laboratory experiments have yielded similar results. Young female rats treated with Arochlor 1221 (a PCB) on the second or third day of life achieved sexual maturity in 28 days, but untreated rats did it in 42 days. (124) In other words, exposure to a PCB made female rats achieve sexual maturity in almost half the time they normally would.
Many politicians have been elected blaming the epidemic of children having children on working mothers, absent fathers, feminists, sex education in schools, and the decline of family values. However, in doing so, they have overlooked their own moral responsibility to provide every child with a clean and healthful environment. The truth is that much of this epidemic of children having children could not be happening if the age girls reached puberty had not declined so much.
Another thing that many conservative politicians have somehow overlooked in their zeal to bash liberals and feminists and get the 10 Commandments hung on the wall of every classroom in America is the extent to which violence and social dysfunction may be caused by exposure to toxins. One telling study was done in Mexico by Elizabeth A. Guillette and her colleagues. They compared two groups of children living in the Yaqui Valley in northern Sonora, Mexico. Although identical in every other way, one group of children lived in the lowlands, which were dominated by pesticide intensive agriculture, and the other lived in the highlands, where the children's parents made a living without using pesticides. The children from the lowlands, who were exposed to pesticides, had far less endurance, inferior hand-eye coordination, and poor drawing skills in comparison to the children from the highlands, who were not exposed to pesticides. The researchers also noticed that the children exposed to pesticides were more likely to fight with their siblings and were more likely to become easily upset or angry when their parents mildly corrected them. These aggressive behaviors were not noted in the children not exposed to pesticides. (125)
Exposure to environmental hormones has been found to affect not only the health of laboratory animals but change their behavior too. Frederick vom Saal and his team, for example, found that mice that were exposed to low levels of the pesticides DDT and metho-xychlor in the womb were more territorial and aggressive than mice that were not exposed in the womb. (126) Other researchers found the same thing when they fed laboratory mice water from wells in rural Wisconsin that were contaminated with farm chemicals. The exposed rats showed unpredictable outbursts of aggression. (127)
According to Colborn, Dumanoski, and Myers, although the dysfunction of animals may not necessarily translate into human social dysfunction, it is a very real possibility that endocrine disruptors at least partially explain the unacceptable incidence of child abuse, crime, and violence in human society:
What about the breakdown of the family and frequent reports of child abuse and neglect? If scientists have found
evidence of careless parenting in contaminated bird colonies, do these chemicals have any role in similar phenomena
among human parents? Reacting to reports of growing neglect and violence against children by their parents, some
commentators have ventured that there must be something wrong with these people; some basic instincts seem to be
missing. Hormones do not determine our behavior, but it is likely that they influence mating and parenting behavior in
humans just as they do in other mammals. . . Hormone disruption can increase the tendency toward a certain kind of
behavior, such as territoriality, or attenuate normal social behaviors, such as parental vigilance and protectiveness. Given
this provocative evidence, we should consider chemical contamination as a factor contributing to the increasing
prevalence of dysfunctional behavior in human society as well. (128)
If children in Fallon County are suffering from an incidence rate of childhood leukemia and birth abnormalities that is much
higher than expected, it is entirely possible that whatever is causing these problems is also causing many other
developmental problems. Yet there would be a difference: These problems, though subtle, like a slightly lower IQ, early
sexual development, and behavioral problems, would not be happening to just a few children in the county, they could be
happening to all of them.
Whether or not there is a cancer problem in Fallon County, or across the state, is not just a scientific question, it is a legal and moral question as well. (129) Science is needed to inform a decision on whether a community has too much cancer. It can tell us how much there is, the likelihood or not it is caused by environmental exposures, the future risk to the general population, and so on, but what it cannot do is tell us is how much cancer is too much, or when the heath department must take action to prevent more. How much is too much is a legal, moral, and political question, involving, in the end, a decision that weighs the right to a clean and healthful environment against the greed of corporate America.
In other words, the court has jurisdiction over this matter. The Montana Constitution says that we have a right to a clean and healthful environment, and also a duty to maintain and improve it. This right and this duty give the court the authority needed to interpret the adequacy of DPHHS's decision to describe the incidence of cancer in Fallon County as "not unusual" and then to do nothing about it. I make this point anticipating that DPHHS will attempt to justify its decision to do nothing as a purely technical one, obscuring the legal, moral, and political issues with a claim that, to a substantial extent, cancer happens randomly, and that establishing the realness of a cancer cluster, or rather its link to environmental exposures, depends upon proving that it is "statistically significant." (130) Although statistics can be very helpful for identifying cancer clusters, tests of statistical significance are an unconstitutional way of deciding whether a community has too much cancer.
This brief is going to argue that testing for the statistical significance of a cancer cluster is a frame-up, a use of bad science to conceal bad public policy. By presuming randomness when it needs to be proven, the appeal to statistical significance as the criterion of proof to decide the existence of a cancer cluster conceals the harm that massive amounts of toxic exposures are causing.
Various times when I have argued that there is a childhood leukemia cluster in Fallon County people have responded by invoking chance, as if cancer were a random distribution of bad luck and life were a game of chance. "A cancer cluster is just a streak of bad luck. It happens for no other reason than it happens." I've been told. It is like cancer is a slot on a roulette wheel, and the victims of it are selected for no better reason than the marble lands on their number. Although he did not say it in nearly such unequivocal terms, Todd Damrow, the state's epidemiologist, seems to agree with this metaphor. In a phone conversation, he told me that reaching any conclusion about cancer in Fallon County was hard because the "numbers were so darn small." Meaning that they weren't large enough to establish the statistical power needed to disprove the hypothesis that any excess was merely random.
In comments to Clair Johnson, a reporter for the Billings Gazette, both Todd Damrow and Mike Spence referred to a lack of statistical power to justify not investigating cancer in Fallon County:
Damrow and Spence both said normal variations occur when looking at statistical significance and that Montana's small population relative to other areas makes comparisons difficult. When looking at small numbers, one or two cancer cases can blow up the rates, Damrow said. (131)
This needs some explanation. Epidemiologists typically measure cancer incidence in cases per hundred thousand people. Aside from Yellowstone County, every other county in Montana has less than a hundred thousand people. As a result, to get the incidence rate for any kind of cancer almost anywhere in Montana, you must multiply up instead of dividing down. Multiplying up, particularly when you only have a population of only a couple thousand, the way Fallon County does, can dramatically amplify any sort of error or random occurrence. A couple of cases more or less of any cancer (except the most rare) would not affect the incidence rate much for Yellowstone County, but they would for Fallon County. One or two cases more in a community of a couple thousand people could translate into a hundred cases more once the rate was translated into units of a hundred thousand. Also, because cancer happens to discreet individuals, and not continuously to, say, 0.33 people, cancer rates in small communities are going to inevitably fluctuate from year to year. Some years there will be a run of cancer cases, other years there may not be any-or so it might seem.
But do random fluctuations really blow up the rates? They do, I think, only if you make a lot of contestable presumptions, in particular buying into the cancer as mere chance metaphor. If you presume, as Damrow and Spence seem to have, that cancer is randomly distributed, and thus subject to testing for statistical significance, you are likely to conclude that random clusters are more likely in small populations than in large ones. (132) Large numbers do indeed smooth things out, reducing fluctuations. But is this because the possibility of randomness has been reduced, or is it because very specific toxic exposures have been averaged out over a larger area, diluting the contrasting differences that would come from different toxic exposures? A high rate of cancer in a small number of people may not be random nearly so much as it is testament to the toxic exposures they have suffered. Insisting on large enough numbers to get statistical significance will expand the area and dilute the high rate.
Depending on the circumstances, large numbers will conceal the effects of local and specific toxic exposures as much as they are needed to produce statistical significance. Despite the insistence by epidemiologists that large numbers are needed to identify cancer risks, large numbers conceal at least as much as they reveal. If cancer is not randomly distributed, insisting on statistical significance before a high rate of cancer is acknowledged as high would conceal enormous amounts of harm. Much preventable suffering would be dismissed as statistically insignificant, mere chance when it was not at all chance.
If DPHHS did apply tests of statistical significance to decide if the cluster was real in Fallon County, or merely a random fluctuation, it would have been following the Centers for Disease Control's guidelines, which do emphasize the importance of establishing statistical power, as the quotation below from the CDC instructs:
Determine an appropriate reference population. Occurrence rates (or other statistics) calculated for the cluster should be compared with those for a reference population in order to identify an excess number of cases . . . If the number of cases is sufficient, and if a denominator is available (e.g., population of a community, number of children in school, or number of employees in a workplace), calculate occurrence rates, standardized morbidity/mortality ratios, or proportional mortality ratios . . . Compare the calculated statistic with that for the reference population to assess significance. Chi-square tests and Poisson regression are also commonly used techniques for comparing proportions. (133)
Because of guidelines like this, which frame the decision as a matter of showing statistical significance, health departments have become increasingly reluctant to become involved in cancer cluster investigations. (134) The Centers for Disease Control has taken a position that sets the standard:
The reported experience of health agencies confirms, however, that major associations between exposures and outcomes
are rare. Minnesota, for example, has reported results from over 500 investigations of clusters, six of which were
full-scale investigations. In one instance, in an occupational setting, an important public health outcome concerning
cancer was documented. Missouri and Wisconsin have reported similar experiences: large numbers of requests for
investigations have been received, but only an occasional in-depth evaluation is warranted. CDC has been consulted in
over 100 such investigations, and again, major associations between exposures and outcomes have been rare. . . .The
unofficial consensus among workers in public health is that most reports of clusters do not lead to a meaningful outcome.
Often, a "case" is not clearly defined, and the "cluster" is, in fact, a mixture of different syndromes. Frequently, no
exposure or potential cause is obvious, and-to make the investigation even more difficult-there are many possible causes.
For example, an inactive toxic waste site may contain hundreds of chemicals. An investigation at the site may indicate no
immediate or obvious connection between exposure and disease, and considerable manipulation may be required to
demonstrate a statistically significant excess. Finally, the biologic consequences and public health impact often are not
clear. (135)
Perhaps the result of opinions like this, especially when they come from the CDC, is that public health officials may feel that they already know everything they need to know about a cancer cluster when a citizen reports it. Since health departments are framing the issue in terms of statistical significance, they are assuming cancer is largely random from the start, and because of this, they can be sure that an investigation of a cluster will go nowhere unless there is sufficient statistical power-- large enough numbers--to rule out random fluctuations, which there rarely is. By framing the decision this way, requiring large numbers when they are seldom going to get them, health departments are able to avoid investigating cancer clusters. They can dismiss any but the worst clusters as merely random. As Steingraber observes:
Before the possibility that the cluster has occurred by chance can be ruled out, cancer rates in some small communities must reach extraordinarily high levels-sometimes as high as eight to twenty times higher than levels for the surrounding areas. Because of the small sample size, lesser increases will not attain sufficient power for the study to be conclusive. (136)
Requiring a cancer cluster to pass a test of statistical significance before anything is done, particularly if it is a small
community, as most communities are in Montana, means that the link between cancer and the environment will never even
be considered for most of Montana. Montana's typically small communities will not have the statistical power to reject the
null hypothesis, the possibility the result was generated by random, and so they will never be investigated.
Besides insisting on standards of proof that disqualify small communities from the very start, health departments, lead by the CDC, just plain have an attitude problem when it comes to investigating cancer clusters, as Sandra Steingraber has observed:
In spite of public concern, many public health officials become dismissive-if not downright apoplectic-when the subject of community-level cancer clusters is raised. Some consider the investigation of alleged clusters a disparaged practice and lament the inability of common people to grasp the statistical concept of randomness. Too often, the message relayed back to those vigilant citizens seeking explanations is that their questions are misguided. Too rarely are they told that the tools of epidemiology are just too blunt to provide answers. (137)
Framing the decision as a technical one, as a matter of mere statistical significance and not as a grave moral, legal, and political issue, has a definite political pay off: It insures that the CDC and the health departments won't have to go out into a world run by powerful corporations and face the political risks of stopping a polluter from polluting. They can say the science isn't there to do it.
Statistical Significance and Real Significance
For all the appeals to science they make, the CDC and the health departments are not using good science when they dismiss a high rate of cancer in a small community as statistically insignificant. They are just manufacturing excuses. While statistical tests are powerful and helpful tools in investigating disease, the way the CDC and health departments have been using them to investigate cancer is inappropriate, bad science. They have been using a mathematical test to decide an issue that is not mathematical. Tests of statistical significance simply cannot determine whether a disease cluster really is real. While they can tell us whether a sample of a population is large enough to represent differences in a population faithfully, they cannot tell us whether those differences are something that really matters. It is bad science, poor public policy, and, at least in Montana, I contend, unconstitutional for DPHHS to make what is unavoidably a moral and legal decision using a technical procedure that is incapable of balancing constitutional rights or considering moral obligation.
Strange as it may sound, tests of statistical significance cannot tell us whether a cancer cluster is significant. Let me go back and explain more thoroughly: Although the Poisson Distribution test is somewhat different from a significance test like Chi-squared, mostly because it does not directly include the sample size in its variables, tests of statistical significance are measures of whether a random sample adequately represents the population from which it was taken. Here is how an old textbook of mine describes what a test of statistical significance is:
Precisely applied, tests of statistical significance tell us how likely it is that a tendency we find in a sample is sufficiently
strong for us to conclude that it also occurs in the population from which the sample is drawn. Some assumptions are
involved. One is that we know what population we are generalizing about. Another is that our sample is truly
representative of this population in the sense that only chance fluctuations can throw it off. (138) (Emphasis in original.)
This textbook emphasizes that the technical meaning of "significance" departs from the ordinary meaning:
"Statistical significance" is a technical term with a specialized meaning. Therefore you must be careful to avoid using
"significant" in its ordinary connotation of "important" or "noteworthy" in a report in which you also use "significant" in
the technical sense. (139)
Statistical significance tests only tell you whether your sample is large enough to accurately reflect the population from
which it is drawn. They cannot tell you whether some difference in the variables matters or not. For instance, with a large
enough sample of the U.S. population, you might find, let us suppose, a statistically significant result that Americans of Irish
ancestry are 0.1% more likely than people of all other ethnic origins to be alcoholics. This test of statistical significance
would seem to confirm the stereotype the Irish drink to excess--but does it really? A 0.1% difference from other ethnic
groups? In fact, the most significant thing about such a result would be that it disproves the stereotype. There would be no
real or practical difference between the Irish and other ethnic groups. And so, even if a difference is statistically significant,
it may not matter.
Statistical significance tests are not magic. They do not grasp the deep essence of being. They cannot decide fundamental questions, like whether something is real or not. They cannot make moral, legal, and political decisions for you. Regrettably, although this is a very basic warning, taught in statistical textbooks throughout the social sciences at every level, I know that political scientists have ignored it, I know that economists ignore it even more frequently, and regrettably, I am finding that epidemiologists and public health departments are ignoring it as well, with the result that the links between cancer and the environment, between corporate pollution and public health, are being obscured and concealed.
Applying statistical significance tests, and in particular tests like Chi-squared and Poisson, to decide whether cancer clusters are real is the wrong thing to do for a variety of reasons. First, it is increasingly becoming clear that cancer is not randomly distributed. Although there are many different random aspects to cancer, such as exposure and vulnerability, at a deep level, there is no evidence to conclude, as epidemiologists implicitly must be doing when they use statistical significance tests, that many, if not most, cases of cancer are randomly distributed-that the cancer roulette metaphor actually explains why cancer happens. Although the mechanisms that cause cancer remain as mysterious as they are complex, and are doubtlessly distributed in ways we do not fully appreciate, we do know enough about the gene/environment interactions that cause cancer to know that randomness cannot be assumed in any but a relatively small proportion of cancer cases. Random distributions are not identical with poorly understood distributions, and just because we don't understand what caused any given case of cancer doesn't mean that it happened randomly, that is to say, without any cause other than chance.
Bad Luck does not Explain it, and Neither do Genes
Although it is no doubt true that cancer, to some extent, is an inevitable tragedy of the human condition, happening for no better reason than that life is happening, it is far from established that most cancer is to any meaningful extent randomly distributed. For example, we know that the incidence of cancer has increased 49.3% between 1950 and 1991 in the United States. Not so long ago in 1950, about 25% of Americans could expect to get cancer in their lifetimes. Now, 38.3% of women and 48.2% of men can expect to, almost double what it was a half century earlier. (140) Such a dramatic increase for such a large number of people cannot possibly be random.
Furthermore, as Steingraber argues:
The spatial features of cancer's occurrence around the globe clearly belie the notion that cancer is a random misfortune.
Industrialized countries have disproportionately more cancers than countries with little or no industry (after adjusting for
age and population size). One-half of all the world's cancers occur among people living in industrialized countries, even
though we are only one-fifth of the world's population. Closely tracking industrialization are breast cancer rates, which
are highest in North America and northern Europe, intermediate in southern Europe and Latin America, and lowest in
Asia and Africa. Breast cancer rates are thirty times higher in the United States than in parts of Africa, for example.
Breast cancer incidence in the United States is five times higher than it is in Japan, but this gap is rapidly narrowing. Of
all the world's nations, Japan has the most rapidly rising rate of breast cancer. (141)
The dismissal of cancer as random fate, something a health department would have no authority or power to prevent,
seems to depend mostly upon the theory that genes are destiny. If cancer is usually caused by something that is randomly
distributed, as genes presumably are, DPHHS would be off the hook, wouldn't have to do anything, except in the most
glaring instances, like perhaps Libby. This theory is dangerous because it does have the trappings scientific validity. Of
course, genes are involved in every case of cancer, if only because every cancer case involves a gene/environment
interaction. Of course, people vary in their genetic ability to resist cancer, just as they vary in their ability to win a 100-yard
dash. And of course, there are some versions of cancer that are strictly determined by genes. If you get the gene, you are
very likely to get the cancer. Still, the hope that a random distribution of genes explains most cancer, and in particular most
cancer clusters, is not good science. Not anymore.
Steingraber tells a personal story to disrupt the notion that cancer can be dismissed as mere bad luck driven by a genetic lottery:
I had bladder cancer as a young adult. If I tell people this fact, they usually shake their heads. If I go on to mention that cancer runs in my family, they usually start to nod. She is from one of those cancer families, I can almost hear them thinking. Sometimes, I just leave it at that. But if I am up for blank stares, I add that I am adopted and go on to describe a study of cancer among adoptees that found correlations within their adoptive families but not with their biological ones . . . At this point most people become very quiet.
These silences remind me how unfamiliar many of us are with the notion that families share environments as well as
chromosomes or with the concept that our genes work in communion with substances streaming in from the larger,
ecological world. What runs in families does not necessarily run in blood. And our genes are less an inherited set of
teacups enclosed in a cellular china cabinet than they are plates used in a busy dinner. Cracks, chips, and scrapes
accumulate. Accidents happen. (142)
If random genes were a good explanation for cancer how come is it that when one identical twin gets cancer the other identical twin is only somewhat more likely than the general population to get the same kind cancer? That is what a recent study of twins in Sweden found:
We conclude that the overwhelming contributor to the causation of cancer in the populations of twins that we studied was the environment. For some forms of cancer, in which a shared environment is important, it may be possible to find clues in studies of childhood environment or long-lasting family habits. The relatively large heritability proportions for cancer at some sites, despite the wide confidence intervals, suggest major gaps in our understanding of heritable cancer. Even for cancers for which there is statistically significant evidence of a heritable component, most pairs of twins were discordant for the cancer--indicating that, on the population level, the increase in risk of cancer even among close relatives of affected persons is generally moderate. (143)
The study found that in the twins studied about 28% of stomach cancers could be explained by some sort of inherited trait, while shared environmental factors, such as life in the womb together, explained 10%, and nonshared environmental factors explained 62%. Heritable factors explained only 27% of breast cancers, while an environment the twins shared explained 12%, and a nonshared environment explained 67%. Colorectal cancer had one of the highest heritable risk factors, but even for it an inherited trait could explain only 35% of the cancer cases, with 5% of the risk from a shared environment, and 60% from a nonshared environment. Most relevant to this discussion, leukemia has one of the lowest heritable risks. Something the twins inherited could explain only 21% of the cases of leukemia, while 12% of the cases were explained by a shared environment, and 66% by a nonshared environment. (144)
Leukemia in children is even less likely to be the result of a long line of genetic inheritance. Until recently, leukemia in childhood was almost invariably fatal. The children would die before they could pass their genes on to the next generation, and so any but the weakest tendency toward childhood leukemia would have faded out of the human gene pool a long time ago--at least if it didn't come from more recent mutations in their parent's genes caused by radiation or other toxic exposure.
The Swedish twins study is consistent with many other studies on the extent that genes cause cancer. Although much has been made about the discovery that mutations in the BRCA1 and BRCA2 genes will increase the lifetime risk for getting breast cancer of a woman who has them 60% to 80%, these genes do not explain really all that many breast cancer cases, as Mary Alma Welch reports:
Mutations in genes such as BRCA1 and BRCA2 account for about 5% to 10% of breast cancer cases. (145) The mutations
may account for as many as 25% of cases in women younger than 30 years who have breast cancer. (146) These mutations
occur in less than 1% of the population at large, (147) but they may occur in almost half of the families with a very high
incidence of breast and ovarian cancer . . . On the other hand, a woman who does not carry a BRCA1 or BRCA2 mutation
of known penetrance still has a 1 in 8 chance of developing the disease in her lifetime. (148)
Using a different path, the International Agency for Research on Cancer, an agency within the World Health Organization,
came to a similar conclusion about the environment as the primary cause of cancer. When it compared the incidence of
cancer in countries with the highest rates with countries with the lowest rates, it found that at least 80% of all cancer is
caused by environmental exposures-that's 4 out of 5 cancers world wide that are not random but caused by something
humans have done to themselves, be it cigarette smoking, a toxic diet, industrial pollution, or radiation exposure from
nuclear testing or nuclear power. The lowest rates of cancer in the world define what is humanly possible, the amount of
cancer that is the inevitable tragedy of the human condition, coming from spontaneous mutations, heredity, cosmic radiation,
or other things impossible to avoid. The difference, on the other hand, between the higher rates and the lower rates define
what we ourselves have added to the inevitable tragedies of the human condition. (149) This difference would be the domain
of responsibility for public health departments, the portion of human suffering that they could do something to prevent. No
one is going to fault a public health department for not preventing cancer that cannot be prevented. What is at issue here is
the amount that isn't random, that isn't an effect of the inevitable tragedy of being human, and can be prevented.
To a limited extent, cancer is a product of genetic heritage. To this extent, it may be distributed as randomly as genes are, and clusters of cancer will randomly occur from time to time for no reason better than bad luck. Nevertheless, as science is clearly showing, most cancers are the result of exposure to something in the environment--natural and industrial toxins, radiation (natural and otherwise), infection by viruses and bacteria, and, as I have argued, various psychoneuroimmunological dynamics. (150) Because of this, when cancer happens at an unusually high incidence, or in clusters, health departments must not presume it to be mere bad luck. Balancing the relatively small portion of cancer that can be explained by genetics against the large portion that can be explained by the environment, the presumption should be that something toxic in the environment most likely explains it.
Unfortunately, connecting a case of cancer to something specific in the environment is, as I am sure DPHHS will point out, hard to do. Most cancers have a decades-long period of latency, the time between when exposure to the toxin begins and the time when the cancer becomes manifest. Within this latency period, there are invariably many ways in which the link between cause and effect might be confounded, entangled, or confused. Most cancer victims can look back at their lives before they were diagnosed and point to many different possible causes for their cancer.
So, while the causes of cancer are mostly not random, they can appear to be so because they are so hard to explain with a specific environmental exposure. But just because the environmental causes are hard to isolate does not mean they are random, or should be interpreted using the mathematics of random distributions. Contrasting cancer cluster investigations to a classic case where the health department in New York City was able to quickly solve a mystery of eleven men who had all turned sky blue and become very sick, Sandra Steingraber explains how difficult it can be to isolate the specific environmental causes of a cancer cluster. According to her, in the case of the eleven blue men, the New York health department interviewed the victims, and found out that they all used a salt shaker in a particular neighborhood diner. The health department tested the shaker and found out it contained sodium nitrate, not sodium chloride, as it should. Exposure to sodium nitrate is known to cause methomoglobinemia, whose most striking symptom is stark blue skin. In this disease cluster case, linking cause and effect was easy because everything was so immediate. The men got sick immediately after exposure. However, things are not so easy with cancer clusters, as Steingraber points out:
Now imagine that cancer made people turn blue. And further imagine a skid row saltshaker containing a powerful
chemical carcinogens that eleven customers unwittingly sprinkle over their food and that eventually causes them to develop
cancer. In spite of their telltale hue, the reason for their disease would probably never be uncovered. Because of the delay
between exposure and onset of disease, at least ten years would pass before any of the eleven turned blue, and some of
them would undoubtably move away during this time. Because cancer is a disease with multiple causes, other drifters with
blue complexions, who contracted cancer for unrelated reasons, would move into the area. The saltshaker itself would be
long gone. Thus, although a cluster of people did indeed contract cancer from a single, identifiable source, a study of all
blue-faced people in the neighborhood would not likely be able to establish the fact. (151)
So, while cancer is not a random disease, striking without identifiable cause, it can appearto be such because the latency
period for it is typically decades long, concealing the link between environmental cause and health effect. While the long
latency period does open up the possibility for random clusters coming from people randomly moving in and out of the
community where the exposure occurred, the cause of the cancer itself is not random. It is the environment. So, applying
tests of statistical significance to cancer clusters to decide whether the environment is the cause is justified only when there
is a nonenvironmental possibility for a random generation of the result. A test of statistical significance, for example, would
be justified (though it would be moot) if it were known that most of the cancer victims had recently migrated into the
community. In western Montana, for example, where lots of people are moving into the state, it is conceivable that a cancer
cluster could come from people who had recently moved in. The toxic exposure causing their cancer would have happened
while they were living out of the state, but, by random chance--this victim coming from California, that victim coming from
Washington--they might end up living next to each other and developing cancer at the same time. (Fallon County, it should
be pointed out, has been losing people, and so the cases of cancer that are occurring here, and in particular the cases of
childhood leukemia, are almost inevitably linked to exposures in the county.)
Despite difficulties like this, a cluster of childhood leukemia is different. DPHHS does not have the excuse of being able to point to nearly as many complicating and confounding variables linking the cancer to the environment. The latency period for it is much shorter than adult cancers, perhaps mere months, instead of decades. (152) And so, childhood leukemia can be used as a reliable indicator of toxins in the environment. As Daniel Sinnett and his colleagues put it:
A peak in the incidence of ALL (2 to 5 years of age) suggests that this disease is initiated in utero or very soon in early
childhood. Because of such a short period between the initiating mutation(s) and the appearance and the detection of
tumor cells, childhood ALL offers a conducive model to examine the effect of carcinogens metabolizing genes in cancer
susceptibility. The identification of risk factors related to gene-environment interaction should thus be facilitated as
comparison to adult cancers where different effects are confounded due to much longer latency periods. Although the
results obtained in children will not necessarily apply to the adult cancer, in many respects the pediatric model may turn
out to be more useful to understand the process of oncogenesis. (153)
In other words, children are like the canaries that miners used to take into the mines to detect accumulations of deadly gases--they get sick before anyone else does. Because of the immaturity of their immune systems, because their cells are replicating faster than adults, because they interact more, pound for pound, than adults do with the environment, children are more likely to be affected by toxins in the environment than adults. As a result, their diseases must be interpreted as a warning of accumulating toxins in the environment. By quickly identifying childhood clusters of disease caused by environmental toxins, we put ourselves in a position to stop the exposure before it causes further harm. A cluster of childhood leukemia is not bad luck, a spin on the wheel of cancer roulette. It is a warning. If isolating the causes of many kinds of cancer is difficult because of the long latency periods, it is less difficult with childhood cancers, and so a health department cannot simply brush off a childhood leukemia cluster because it isn't statistically significant.
Statistical Significance and Cancer Clusters
Without exception, testing for statistical significance presumes the possibility of some sort of random generation of the results. There has to be a chance model, a justification for thinking the results are random. Indeed, the validity of a finding of statistical significance depends upon rejecting what is called the "null hypothesis," which means we first propose that the sampled distribution happened only by chance, and then reject that hypothesis if it becomes too improbable. Because a chance model is logically required, it is a misuse of statistical significance tests to apply them to a sample that is not randomly drawn, not even when it appears to be random.
Statisticians do insist on this point. For instance, in their textbook Statistics, David Freedman, Robert Pisani, and Roger Purves consistently maintain that without a "box," without a way of reaching into a population (imagine an urn full of black and white marbles) and randomly selecting a sample, the tests for significance are meaningless, as they write:
Statistical inference can be justified only in terms of an explicit chance model for the data. No box, no inference. (p. 407)
Every legitimate test of significance involves a box model. (p. 440)
If an investigator makes a test of significance when he has data for the whole population, watch out. (p. 490)
Unless there is a clearly defined chance model, a test of significance makes no sense. (p. 497) (154)
We know that around 80% of cancer cases in industrialized countries are from environmental causes. We know that no more than about 20% of cancer cases can come from random genetic distributions. We know, in short, that cancer, for the most part, does not happen randomly, without connection to the environment it occurs in. It is caused by the environment. For cancer there is no viable chance model, no game of cancer roulette. Around four out of five times cancer is linked to the environment the victim has lived in. And so, when a health department is confronted with a cluster of cancer it is obligated not to test the cluster for significance, maybe dismissing a high rate of cancer as mere chance, but to go to that community and ask the victims, or their families: Where have you been living? What have you been doing? What have you been eating? What is your occupation? What have you been exposed to? An explanation is warranted, not a dismissal based on a presumption that it is just chance.
As I argued earlier, doing a significance test on a cancer cluster to test whether it is random is a frameup-it frames cancer, without any scientific justification, as something that is randomly distributed, something that happens by chance, without meaningful cause, and then offers a politically convenient excuse not to investigate it. (155) As long as statistical significance tests are used to identify cancer clusters, and link them to the environment, cancer can never become anything but what statistical significance tests reveal, a randomly distributed object that is only occasionally linked to the environment. (156) From the outset, as soon as a method of random mathematical analysis is chosen, and the metaphor of chance is imposed on the data, cancer becomes a production of what that kind of mathematical analysis can reveal. (157) Once it is presumed to be randomly distributed, cancer becomes available for testing of statistical significance, which, in turn, constitutes it as an object meaningfully understood by statistical significance testing. (158) And so, the decision that decides cancer's essence here is not something that is inescapably existential, an unmediated reality of cancer itself, but an effect of the statistical technology used for researching it-the metaphor used to interpret it. Cancer is "randomly" distributed not because science has established that it is, but because epidemiologists and health departments have insisted on researching it that way. In short, as a "randomly" distributed object available for statistical significance testing, cancer is a production, an artifice of the technology used to research it, not an "objective" reality, whatever that might be. (159)
In other words, we don't have to take epidemiologists or public health departments too seriously when they tell us that cancer isn't linked to the environment it is happening in until the rate is statistically significant. The "randomness" of cancer is purely an effect of the method used to investigate it. A production of the metaphor used to interpret it.
The key issue before this court is how much cancer can exist in a community before it becomes unacceptable and the public health department must act to reduce it. DPHHS says that there isn't a cancer problem in Fallon County, and I say that there is. DPHHS appears to be relying on appeals to statistical significance to get out of doing anything (even though the number of cases of childhood leukemia is highly significant), and I am insisting that tests of statistical significance cannot decide the issue. Although it is tempting to make the question into a technical issue by using tests of statistical significance to settle it, avoiding the tumult of legal, moral, and political debate, it simply isn't possible. There are no short cuts around a debate over the extent to which we, as a society, should sacrifice our health and the health of future generations for convenience and corporate profit.
Contrary to what one might expect and hope, the misuse of statistical significance tests is common among scientists of all disciplines, perhaps because it is a short cut out of difficult political and philosophical arguments. Since the basic statistical mathematics are the same for epidemiology as they are for political science and economics, the criticism of an economic theorist against the misuses of statistical significance in economics should be as valid for epidemiology. Here is how Donald McCloskey, an economic theorist, criticizes the misuse of statistical significance tests in economics:
How large is large? The usual criterion for answering it in economics, which is the statistical test of significance, does not answer it. The criterion is a poor one. It would be good to abandon it. Because it appears to provide a way of processing "how large is large" question on a scale limited only by the computer budget of the investigator, it has crowded out the sounder rhetoric of quantification. In the dear dead days of mechanical calculators, when the inversion of a 5-by-5 matrix was a feat and the inversion of a 10-by-10 not worth risking one's marriage to achieve, economists and other social scientists with quantitative tastes had to know exactly why they were calculating each statistic. They had to think. Giants walked the earth then, chained to a Freiden mechanical calculator and their copies of statistical handbooks by Fox and Snedecor. Or so they say. In any event they were unable to pollute the air with trivial significance tests.
The inadequacies of significance tests as a rhetoric for quantification can be summarized either by the narrowness of what they do or by the breadth of what do not do. What they tell the intrepid investigator what the probability is that because of the small sample he faces he will make a mistake of excessive gullibility in accepting a false statistical proposition, conventionally taken to be the proposition that some number is zero (or some other single interesting alternative). That is all. The procedure keeps one from being made to look a certain, narrow kind of fool. There is no protection against other kinds of foolishness, such as using entirely the wrong variables in one's regression equation; or even using a single wrong variable, which is sometimes enough to make one look very foolish indeed. One is protected from the narrow foolishness in talking about a narrowly defined hypothesis for a narrow sort of error, namely, the error that comes from having too small a random sample--not a nonrandom sample, understand (for that there is no protection), just a too-small but perfectly random one.
Economists use the mathematics of sampling daily, yet rarely in fact take samples. The mathematics applies most naturally to situations in which the statistician literally draws a sample, such as a sample of glebe terriers from English villages to be used to determine the geography of scattered holdings of land. Such "active" sampling entails no metaphysics about unobservable error terms. The statistician knows the error term to be independently and identically distributed because he has used a table of random numbers having such qualities to choose the sample. He could select the entire universe if he wanted, but chooses instead to select the 11th, the 365th, the 7,864th, the 5,6457th, and so forth out of the 10,000 villages in order to make his task manageable. There is still an element of luck about it, but the provenance of the luck is known: remarks about its provenance is what statistical theory is about. (160)
McCloskey continues this argument, insisting that tests of statistical significance simply are not up to the job of answering the questions we most want to know:
It is clear, then, that the breadth of what tests of significance do not do is great. They do not answer the question, How large is large? How large the number of whippings of American slaves had to be to be considered important is a difficult question, on which various sorts of comparative evidence will bear. (Were free workers subject to corporal punishment? What did the slaves consider to be an excessive number of beatings? How much were slaves beaten in other slave societies?) The merely statistical significance of one or another estimate of the number, however, should interest no one very much.
Yet the test of significance is routinely used to decide all manner of economic questions, and economists are drunk with it.
(161)
Just as tests of statistical significance cannot tell us whether slaves in America where whipped too much, they cannot tell us
if a community has too much cancer. Transforming that question into a question about random distributions is a copout. It
avoids the real issue. One of the clearest indications that an economist, a political scientist, or an epidemiologist is copping
out, using a technical gimmick when they should be arguing laws, morals, or politics, is when they start applying tests of
statistical significance to entire populations. Tests of significance, it must be emphasized again, are designed for deciding
whether a random sample is large enough to reflect the characteristics of an entire population. When they are applied not to
random samples but to entire populations to decide whether anything random is happening in the population as a whole
something is very wrong. The metaphor of chance has run amuck, imposing itself on something that is not distributed by
chance.
For instance, if DPHHS dismissed my concern about Fallon County having an incidence of breast cancer that was one of the highest in the world on the grounds that it probably was merely a random fluctuation, perhaps using a test like Chi-squared to justify its conclusion, it would not be using good statistical method, that is to say, good science. The incidence statistics that DPHHS has on Fallon County include virtually everyone living here. There is no sample being drawn. There is no valid chance model. And there is no possibility that the sample could randomly misrepresent the whole population because DPHHS is counting virtually the whole population. If DPHHS dismissed my concerns about the high incidence of breast cancer here (or colorectal cancer) on the grounds that it could have happened randomly, the department, despite its supposed grounding in empirical method, is venturing into metaphysical speculation that is very contestable.
Let's look at the metaphysics implied here: If the whole population of Fallon County could have randomly departed from its "true" rate of cancer, and randomly generated a "false" rate, that means that there is a "real" rate of cancer for Fallon County that transcends anything that actually happened. This would be presuming that there is something more "real" about Fallon County's incidence rates than what actually happened--a really "real" rate of cancer. --Kind of reminds you of Plato, doesn't it, where the "reality"we see in ordinary life is nothing more than shadows flickering on the wall of a cave, while true reality is outside, casting shadows on the wall? If we dismissed the high rate of breast cancer in Fallon County as random, we would be dismissing the truth of what we can see and count for the sake of a theory of randomness that we can neither see nor count. We would be saying that we know something about cancer in Fallon County that only statistical significance tests can reveal, and that this metaphysical standard, with only an appeal to its own internal logic, can override and discount what we see. The actual count of cancer may be high, as it is with breast cancer in Fallon County, but not to worry, it is only a random fluctuation. Just how do we know it is a random fluctuation? Because when we treat the whole population of Fallon County as a sample of itself, and run a test for statistical significance on the whole population, it turns out that the actual count, due to the small numbers of the whole population, is not statistically significant. And so the "sample" of Fallon County is telling us that it may not be adequately representing the whole population of Fallon County-this, even though the "sample" is identical to the whole population.
Perhaps DPHHS did that, perhaps they will deny it after I put it like this. Yet what arethey doing if they deny the inappropriate use of statistical tests, using a whole population as a sample of itself, and yet insist that high rate of breast and colorectal cancer in Fallon County is but a random fluctuation? How do they know it is random if they cannot prove it with a significance test? And if it is random, allowing them their argument for a moment, how do they know that the really "real" rate isn't actually worse than what they are counting? Surely, if cancer is as random as DPHHS seems inclined to say it is, the department will concede that sometimes an actual cancer rate can be worse than it appears to be, and that, despite the actual count, there really is a link to an environmental problem. Still, the only direction they seem to be inclined to go on this randomness thing is the direction that will get them off the hook.
Surely, the department, being officially staffed by empiricists, will insist that nothing it concludes is based on a metaphysical
system, however worthy Plato's thought may be. Surely it will tell us that it bases its conclusions solely on the data
presented to it. Fine, I say. If that is the case, if the department is truly going to base its decisions on the empirical
evidence, it has no business using tests of statistical significance to decide whether a cancer cluster is real or not. It must, as
I will contend later, base its decision on whether an actual enumeration of the cases of cancer is higher than they would be in
a clean and healthful environment.
State Wiggles out of it Using Risk Assessment
Apparently, another reason that the government used to justify not investigating the links between cancer and the environment in Fallon County was that it decided that there were no toxic exposures in Fallon County that were "unsafe." In a letter to me, Robert Williams, on behalf of the Agency for Toxic Substance and Disease Registry, wrote:
ATSDR staff checked with federal and state agencies regarding the presence of a hazardous waste site or a specific,
significant release of hazardous substances in Fallon County, but could not identify a specific hazardous substance release
that could be a concern. (162)
It appears that the State of Montana used information from risk assessments to dismiss my petition too. After consulting
with Montana's Chief Medical Officer, Mike Spence, and Montana's head of the Department of Environmental Quality,
Mark Simonich, Governor Marc Racicot wrote me a letter saying that a review of environmental exposures did not offer any
explanation for unusual incidences of cancer in Fallon County. According to Governor Racicot:
. . . The environmental investigation included uranium exploration, hazardous materials related to oil and gas production, solid waste disposal, public drinking water, and pesticide use.
Uranium exploration occurred in Fallon County in the late 1950's and again, in the 1970's. The DEQ (then the Department of State Lands) began regulating uranium drilling in the early 1970's. Five permits for uranium were issued. Two were never drilled because of Montana's passage of a bill banning disposal of uranium milling tailings around 1980. Less than 100 holes were drilled in the 1970's. Fallon County never became as significant an area for uranium exploration as Carter County, its southern neighboring county.
Due to Fallon County's geological setting, it is possible that some levels in the regional ground water system could show high levels of radioactivity from soluble forms of uranium. These natural levels could, over time, cause cancer. There is, however, no scientific evidence this has occurred.
According to the DEQ, it would be possible to determine if radionucleide content in ground water could be a contributing factor. It would be a matter of testing the supply wells or sources which have been used by the families for uranium, radon, and probably, a suite of related "daughter" elements or isotopes. While this would not necessarily confirm whether there had been some past, transient episode of high radionucleides, it could indicate whether there is a current, on going problem.
DEQ was not aware of any history of in-situ uranium leach mining in Montana.
The references to alleged mishandling or improper disposal of hazardous materials associated with oil and gas production appeared to be in general rather than relating to specific operations. Drilling fluids, production water, and other wastes associated with the exploration, development, or production of crude oil, natural gas, or geothermal energy are excluded from hazardous waste regulations; however, other hazardous materials that are ancillary to such operations are not excluded and would be subject to state regulations when disposed of it they meet the definition of hazardous waste. DEQ is not aware of any specific oil field operation in Fallon County that is, or has been, unlawfully disposing of hazardous waste.
Regarding the issues of the old landfills as potential problem sources, the City of Baker has two closed landfills west of town. According to DEQ records, the older of these two landfills was operated from the period beginning sometime prior to 1966 to 1986. From 1986 until 1992, the city operated another landfill in the same general vicinity as the first. Both existed during a time when siting and operational criteria were minimal and ground water monitoring was not required.
The suggestion that the health problems could be caused from the improper disposal of hazardous materials is possible. Disposal of materials that we now know to be hazardous could have occurred at either of the two landfills near Baker, or at the sites of any of the gas and oil drilling locations throughout the county, or in an agricultural dumpsite that may have existed at an area farm or ranch. Existing information does not reveal if hazardous materials were disposed of at either of the city landfills. As for the current Fallon County landfill, the siting and operating criteria under which it was developed and operated (full Subtitle D compliance) gives DEQ much more confidence that it would not be the source of any problems in the area.
Although landfill gas can be hazardous, in most situations where landfills have caused environmental or public health damage, it has been from the contamination of water. While DEQ does not have any water monitoring data specific to the old landfills, it does have data for the City of Baker water system, which is a public water supply. As a public water supply, the quality of the water is monitored very closely and must meet specific state and federal standards. DEQ's records for the City of Baker public water system indicates that the water meets all required standards for water quality and shows no signs of any contamination that would be expected to cause the health problems.
You mentioned that Agent Orange was used extensively on your family's ranch as a herbicide in the late 1950's,
throughout the 1960's, and into the 1970's. Some pesticides, when disposed of, are subject to hazardous materials
regulations; however, the use and storage of pesticides are not subject to hazardous waste regulations, but rather the
provisions of the Montana Pesticide Act. It is not clear if the drums are empty or if they contain pesticide. The Montana
Department of Agriculture can assist in analyzing the contents of the drums and provide advice on proper management of
the drums. DEQ is not aware of any specific current concerns regarding the proper disposal of waste pesticides in Fallon
County. (163)
The main emphasis in Governor Racicot's letter was on how there were no "unsafe" environmental exposures in Fallon County, nothing that, as far as the state knew, exceeded state or federal standards. The logic seems to be that there wasn't any environmental cause, as defined by state and federal standards, for a high rate of cancer, and so there was no reason to assume that any high rates were anything more than random fluctuations. To come to this conclusion, means endorsing the risk assessments that set the federal and state standards for what is considered "safe." However, this seems to me backwards. If there is a high rate of cancer, and no toxic environmental exposures exceed state or federal standards, there is probably something wrong with state and federal standards. High cancer rates should be used to decide whether the standards are sufficient to protect public health. The standards should not be used to dismiss high rates of cancer as mere random fluctuations. If cancer rates are too high, and none of the standards have been exceeded, that should mean that the standards are too weak. Nevertheless, implicitly, if not explicitly, the Governor invoked the standards set by risk assessments many times when he concluded that there could not be an environmental cancer problem in Fallon County-risk assessments involving innumerable hazardous materials, uranium mining, drinking water quality, oil field operations, and so on.
In a subsequent letter to me, Todd Damrow, the epidemiologist who reviewed my petition again after I insisted to Governor Racicot that the state was wrong, again makes it clear that the risk assessments used to set federal and state standards were used to inform his decision that there was no need for an investigation in Fallon County. Referring to his own finding that there wasn't a cancer problem in Fallon County, Damrow wrote:
These findings are consistent with the earlier study by Dr. Michael Spence, State Medical Officer, who reported that
cancer incidence rates in the area are not unusual, i.e., did not differ from state and national cancer rates. These findings
are also consistent with the review of existing regulatory data by DEQ which showed the current environmental quality to
be protective of health. (164)
Governor Racicot explicitly referred to the standards regulating the safety of Baker's drinking water in his letter to me, as we saw above. He wrote, "As a public water supply, the quality of the water is monitored very closely and must meet specific state and federal standards. DEQ's records for the City of Baker public water system indicates that the water meets all required standards for water quality and shows no signs of any contamination that would be expected to cause the health problems." In other words, according to the risk assessments used to set the standards for acceptable toxic exposures in drinking water supplies, Baker shouldn't have a cancer problem because of anything in the water. Everything is below what the state calls the "Maximum Contaminate Levels," the levels found by a formal process of risk assessment to be safe.
On Friday, June 1, 2001, the Fallon County Times ran on its last page the annual drinking water quality report for Baker. When you might expect the neutral language common to government efforts to inform the public, the City of Baker seemed strangely enthusiastic about the quality of its water, almost like it was a public relations move, designed to limit public concern:
We're very pleased to provide you with this year's Annual Quality Water Report. We want to keep you informed about the
excellent water and services we have delivered to you over the past year. Our goal is and always has been, to provide to
you a safe and dependable supply of drinking water. Our water source is ground water. We have five wells-all of them in
the Fox Hills Sands Formation, and are located on the west edge of town . . . We're pleased to report that our drinking
water is safe and meets federal and state requirements. (165)
Like you would expect of a public relations maneuver, however, the bad news came toward the last, when people might
have been expected to become bored with the good news and quit reading. Toward the bottom of the ad, it was reported
that the test results did show a trace amount of 1,4-Dichlorbenzene-0.32 parts per billion. This amount, we were told, was
considerably below both the Maximum Contaminate Level and the Maximum Contaminate Level Goal, both of which were
75 parts per billion. This would make a mere 0.32 parts per billion seem safe, but (as we were not told in the ad)
1,4-Dichlorbenzene is a chemical specifically used to kill things. It is an insecticidal fumigant, commonly used against
clothes moths, and as a germicide. It is sometimes used as a deodorant for garbage and in restrooms, as well as an
insecticide for control of fruit borers and ants. It may also be applied to tobacco seed beds for blue mold control, used for
the control of peach tree borers; or used for control of mildew and mold on leather and fabrics.
It also moves around in the environment after it is used, according to a fact sheet I found on the Internet:
Chemical waste dump leachates and direct manufacturing effluents are reported to be the major source of pollution of the
chlorobenzenes (including the dichlorbenzenes) to Lake Ontario. The major source of 1,4-dichlorobenzene emission to the
atmosphere is volatilization from use in toilet bowl deodorants, garbage deodorants and moth flakes. If released to soil,
1,4-dichlorobenzene can be moderately to tightly adsorbed. Leaching from hazardous waste disposal areas has occurred
and the detection of 1,4-dichlorobenzene in various groundwaters indicates that leaching can occur. (166)
So, 1,4-dichlorobenzene can leach out of landfills and into groundwater supplies. Perhaps it is also relevant to note that
the old town landfills are directly over the aquifer where Baker gets its water. However, despite Governor Racicot's and the
City of Baker's reassurances that the water is safe, if one contaminate leached into the ground water supply for Baker, are
we not justified in wondering if others didn't too--but aren't being tested for? Despite the reassurances from Governor
Racicot and the City of Baker about the safety of the water supply, and despite the fact that everything tested for is below
the "Maximum Contaminate Levels," some skepticism should be justified. Something, after all, is causing a high rate of
childhood leukemia, breast cancer, and colorectal cancer in Fallon County. Until that something is identified, anything that
seems suspicious deserves to be held in suspicion. Despite all the risk assessments and all the standards saying everything is
safe, chemicals are not human beings. When they are present in a place they shouldn't be, and when harm is happening that
shouldn't be happening, they should be held guilty until proven innocent.
We are told in the City of Baker's ad that the water is safe, that the Maximum Contaminate Levels "are set at very stringent levels." So high that a person would have to drink 2 liters of water every day at the MCL level for a lifetime to have one-in-a-million chance of becoming sick. (167) Still, is this really the case? Is water contaminated with an insecticide, a chemical produced to kill things in small quantities, really safe, even if the levels are very low? And who is to decide what is "safe" or not? As Mary O'Brien, a scientist whose main interest is risk assessment, points out, too often government officials and industry advocates presume entirely too much when they pronounce something safe:
What is acceptable to any person is a matter of personal judgement, but the word is used by risk assessment's promoters
as if it were something concrete that could be measured by others, or as if it were something about which everyone must
surely agree. This is not accurate. For instance, while a state's Department of Environmental Quality may call some
amount of toxic pollution of well water acceptable, a person who actually drinks the water may not find any unnecessary
pollution acceptable. (168)
Maximum Contaminate Levels, the standards by which the government decides something is "safe" for us, are set by the
formal process of "risk assessment." To the uninitiated, risk assessment may seem the essence of rationality, science, and
democracy. The risks of toxic exposure are researched, assessed, reviewed, commented on, and then, if they are
"acceptable," officially endorsed. But, in fact, despite all the efforts to make them seem scientific and democratic, there are
lots of ways in which risk assessments can seriously underestimate the real risk to the public, be biased in favor of large
corporations, and undermine democracy and our civil liberties.
Risk assessments have long been criticized by environmentalists for being the basis of horrible injustice, and a full accounting of all the ways they fail to account for the harm done to the environment and the public could fill volumes. However, what risk assessments ultimately amount to is a decision by the government to allow private corporations to expose the public to deadly toxins that will kill people without their knowledge or their consent. Ordinarily, a government action like this would be a violation of the 5th amendment to the Constitution of the United States, which says, "No person shall be . . . deprived of life, liberty, or property without due process of law." However, these are not ordinary times. For the economy to operate, we are told by corporate America and its apologists, it has become necessary for all of us to bet our lives and spin the cancer roulette wheel, accepting small amounts of what is called "risk" for what we are told is the "greater good."
Risk, it seems, is something that happens to statistical abstractions, one in a million people, maybe one in a hundred thousand. Because the odds are so remote to the individual invited to spin the cancer roulette wheel, they are not really real for them. That is the seduction. The odds are too unlikely to mean that their life is actually at risk. Someone, in theory, may die, but, with such remote odds, it won't be them. This deployment of the word "risk" transforms what would be murder, a violation of our civil rights, into something that is "insignificant," if not "legal." With this mobilization of risk metaphors, we all are conscripted into taking risks on the behalf of corporate America. The risk is acceptable, we are told, because it is a "negligible" risk, too small to matter. The misfortune is going to be borne by someone else, someone who for all practical purposes won't matter to them. And in case someone doesn't buy the metaphor, and decides the risk isn't negligible, and they become reluctant to spin the cancer roulette wheel, statistical significance testing is deployed to reassure them that the bad luck is just random bad luck. If there is a cancer cluster, it probably happened just by chance. It is nothing to stop them from playing the game that pays off so well for corporate America because, chances are, it won't happen to them.
Peter Montague, the editor of Rachel's Environment and Health News, traces the origins
of risk assessment back to a 1954 decision by the Food and Drug Administration to allow certain amounts of poisons into our food:
Initially these amounts were called "safe" but as knowledge grew, scientists came to realize that, at least in the case of
cancer-causing poisons, someone, somewhere, would be harmed if any amount appeared in the nation's food supply. After
that realization, "acceptable" amounts were set on the basis of predictions of how many people would be killed. Usually
the official goal was to kill no more than one-in-a-million people, though sometimes one-in-a-hundred-thousand is deemed
acceptable. This technique is now called "risk assessment," and today it is so widely practiced that some people consider it
the only possible way to think about such matters, which of course isn't true. (169)
Make no mistake about it, although justified as a economic necessity, this decision was not a morally neutral one. It
amounted to nothing less than sacrificial murder on behalf of corporate America. The plain fact is, even under the most
conservative (i.e., false) assumptions being put forward by the risk assessors, people are dying from pollution, as Steingraber
points out:
Suppose we assume for a moment that the most conservative estimate concerning the proportion of cancer deaths due to environmental causes is absolutely accurate. This estimate put forward by those who dismiss environmental carcinogens as negligible, is 2 percent. Though others have placed this number far higher, let's assume for the sake of argument that this lowest value is absolutely correct. Two percent means that 10,940 people in the United States die each year from environmentally caused cancers. This is more than the number of women who die each year from hereditary breast cancer-an issue that has launched multimillion dollar research initiatives. This is more than the number of children and teenagers killed each year by firearms-an issue that is considered a matter of national shame. It is more than three times the number of nonsmokers estimated to die each year of lung cancer caused by exposure to secondhand smoke-a problem so serious it warranted sweeping changes in laws governing air quality in public spaces. It is the annual equivalent of wiping out a small city. It is thirty funerals every day.
None of these 10,940 Americans will die quick, painless deaths. They will be amputated, irradiated, and dosed with
chemotherapy. They will expire privately in hospitals and hospices and be buried quietly. Photographs of their bodies will
not appear in newspapers. We will not know who most of them are. Their anonymity, however, does not moderate this
violence. These deaths are a form of homicide. (170)
To compare it another way, the10,000 people every year who are deemed an acceptable sacrifice by risk assessors in industry and our government are three times the number of people the terrorists killed in the attack on the World Trade Center and the Pentagon. If we objected so mightily when terrorists kill perhaps 3,000 of our citizens, immediately mobilizing for war, generously donating funds to the victims, and allowing the erosion of our civil liberties, how can we offer no protest when the companies kill 10,000 people each year from toxic environmental exposures? How can this sacrifice be justified? What in our Constitution would allow our own corporations to kill American citizens this way?
Since risk assessment is a government action, and since the standards which allow these deaths are government standards, the government is effectively taking these lives. The 5thamendment to the Constitution of the United States says, "No person shall be . . . deprived of life, liberty, or property without due process of law." To hear corporations such as the mining companies and the logging companies cite the 5th amendment, claiming that anything that increases the cost of their operations is a regulatory takings, one would think that the only takings possible is the takings of private property. But in fact, the government is also forbidden to take life and liberty without due process. If our civil rights mean anything, the government should be bound by the Constitution to go through a lengthy process to justify the taking of life-an investigation, a trial, a sentencing-and it should offer evidence that the person's whose life is being taken did something to merit this punishment.
What is the due process that allows the pesticide industry, with government sanction, to take the lives of the 10,940 people Steingraber describes above? Was there a trial? Were any of these people found guilty of anything? Did they deserve to die? Let us be honest and admit that risk assessment is not due process, and that the lives taken by it are taken illegally, without the victim's knowledge, consent, or representation by legal consul to protect their rights.
There is not even a hint in the Constitution that the government may take a life without just cause, so long as it is a benefit to the economy. Surely the commerce clause cannot possibly reach so deep as to allow the taking of life without due process. Given all the controversy surrounding due process and the death penalty, all the appeals that are still required by even the most aggressive judicial advocates of the death penalty, it isn't even possible to imagine that aneconomic taking of life could be so light and easy under a plain reading of the Constitution. And the anonymity of the victims of risk assessment does nothing to change any of this.
Anonymity, as Paul Merrell and Carol Van Strum point out, while convenient for corporate America, is not a meaningful moral or legal distinction:
Why should it matter that contemporary "negligible risk" victims are known only in number and not by name? We do not
excuse the killer who shoots into a large crowd of strangers because he doesn't know his victims' names and he kills only a
few people or even just one. Why, then, do we tolerate those who spray the crowd with poisons rather than bullets? Is it
not still murder? The corpses lie just as dead. (171)
Let us suppose, as Merrell and Van Strum do in the same article above, that the government did a really, really good risk
assessment, one where a permit to release a carcinogenic pesticide into the food supply was so accurate that it included the
names of the people who would eventually get cancer from it. This one would die, that one would get sick. Surely, the
marked people would be entitled to an injunction, no matter how few they were in number, against using that pesticide based
on their right to life in the 5th Amendment. (172)
Despite the anonymity of the victims of risk assessment, it still is murder to kill people with toxins, as Peter Montague argues:
A few years ago, nationwide fear and outrage erupted when a small number of people died after a few Tylenol bottles were spiked with cyanide.
Courts have declared that it is murder for a wife to kill her husband by lacing his chili with parathion (a pesticide), so how can we excuse those who authorize the poisoning of the entire nation's food supply? Is it sufficient justification for murder that only a few will die? If so, how can we justify punishing the Tylenol killer or the wife who murders only one husband?
We need to recognize risk assessment (for pesticides and other hazardous chemicals) for what it is: evidence of
premeditated murder. It documents the intent of regulators and polluters to sacrifice individual lives on the altar of profit.
The person who writes a risk assessment is an accessory to a felony. The concept of "negligible risk" is tolerated only
because of the anonymity of its intended victims. (173)
When a community has an incidence of childhood leukemia that is 10, maybe 20, times higher than the national average, an incidence of breast cancer that is one of the highest in the world, and an incidence of colorectal cancer that is 7 times higher than neighboring counties, as Baker does, the presence of even a very small amount of a pesticide in its drinking water should be a cause for concern, not an occasion for reassuring statements. With the kind of incidence of cancer Baker has, the City of Baker may be presuming a bit much when it decided for everyone, without their consent or full appreciation of the facts, that the water was "safe." The evidence available would surely indicate that something isn't nearly as safe as it should be. Something is wrong, and the presence of a pesticide in the drinking water, however small its concentration, may be an indication of what it is, whatever the risk assessment says is "safe."
When the City of Baker, the State of Montana, or the federal government uses the result of a risk assessment, an MCL level for 1,4-dichlorbenzene, to reassure the public that the water is "safe," they are doing something legally questionable. They are validating a process that we know, even by the people who advocate it, is killing people, taking life without just cause.
Lies, Damn Lies, and Risk Assessments
Using risk assessments that say that an exposure is safe and that the state is justified in not investigating a cancer cluster would be a violation our right to a clean and healthful environment for several different reasons. Not only, as we saw above, does risk assessment presume consent to risk when none is given, conscripting people into the game of cancer roulette against their will, there is every reason to believe that risk assessment consistently underestimates the real risk that people are facing. The odds of the game aren't nearly as favorable as we've been told.
We've been doing risk assessment, as Peter Montague observes above, since 1954, and yet from then until now the evidence indicates that it has been a miserable failure. Risk assessments are always telling us that the risk from a pesticide, or some other toxic chemical, is always so very small-something like one-in-a-million-and yet the reality for Americans has been a huge increase in risk, as Sandra Steingraber points out:
All types combined, the incidence of cancer in the United States rose 49.3 percent between 1950 and 1991. This is the longest reliable view we have available. If lung cancer is excluded, overall incidence still rose by 35 percent. Or, to express the figures in another way: at midcentury a cancer diagnosis was the expected fate of about 25 percent of Americans-a ratio (Rachel) Carson found so shocking that it inspired the title of one of her chapters (in Silent Spring)-while today, about 40 percent of us (38.3 percent of women and 48.2 percent of men) will contract the disease sometime in our lifespans. Cancer is now the second leading cause of death overall, and the leading cause of death among Americans aged thirty-five to sixty-four.
More of the overall upsurge has occurred in the past two decades than in the previous two, and increases in cancer
incidence are seen in all age groups-from infants to the elderly. If we exclude cancer of the lung and restrict our view to
the period covered by SEER, overall incidence rose 20.6 percent between 1973 and 1991, while mortality declined 2.8
percent. (174)
If we go back further in time than Steingraber did, the increase in cancer is even worse than that. The American Cancer
Society in 1919 published posters reporting that only one out of every ten persons more than forty was then dying of cancer.
(175) Although these last statistics surely are not as methodologically sound as the ones sited by Steingraber above, perhaps
in error for reasons ranging from poor diagnosis to inaccurate classification, and subject, as well, to much different economic
and health care situations, they do suggest that the incidence of cancer has increased dramatically over the last century.
Obviously, there is nothing in the way the universe is put together that means that cancer has to be as common as it now is. Almost certainly, the high incidence of cancer is linked, at least in part, to the truly staggering amount of toxins now released into the environment annually. In the most recent Toxic Release Inventory by the EPA in 1999, the government reports that 7.77 billion pounds of toxic substances were released from 22,639 facilities across the U.S. (176) And even this staggering number underestimates the amount of toxic substances released into the environment because it didn't cover all facilities or all toxins. Despite its small population, according to this TRI report, Montana ranked ranked 19th in the nation for total on- and off-site releases, releasing 48,659,575 pounds in 1999. In pounds of toxics per capita, Montana ranks as one of the worst states in the nation.
Yet, despite the increase in cancer, despite the staggering amounts of toxins and carcinogens released into the environment, we are so often told by the government and industry apologists that everything is safe. The risks have been carefully assessed, and they are, for all practical purposes, negligible. However, being told that everything is safe when various disease rates have increased so much is like being the woman who returned home early from work and found her husband in bed naked with another woman, and he protested his innocence by saying, "Honey, who are you going to believe, me or your lying eyes?"
Despite the claim by industry and government that everything is safe, looking at the dramatic increase in cancer that has happened along with the dramatic increase in toxic releases, the inescapable truth is that risk assessments radically underestimate the real amount of harm being done by industrial toxic releases. They cover up what is really happening to public health. Risk assessments are subject to distortion in many different ways, as William Ruckelshaus, a former head of the EPA has admitted: "We should remember that risk assessment data can be like the captured spy: If you torture it long enough, it will tell you anything you want to know." (177) Torture is not the wrong metaphor to use to describe what is happening to the truth, either.Scientists who advise the U.S. EPA on regulatory decisions are not neutral advocates for public health but often financially and professionally linked to the very industries that would be affected most by the regulations being assessed, according to a study by the General Accounting Office, a congressional watchdog agency. In one case, seven of 17 members of a Science Advisory Board panel studying the cancer risks of a toxic chemical either worked for chemical companies or for industry-affiliated research organizations. In addition, five of the other members had received consulting or other fees from chemical manufacturers. Only a small minority had no direct conflict of interest. (178) With so much conflict of interest, the integrity of risk assessment processes cannot be assumed.
Risk assessments underestimate the risk in a variety of ways. Risk assessments claim to be a complete assessment of all the risks, as they would have to be to be valid, but this is impossible for several different reasons. First, for a risk assessment to be complete, all of the toxic effects from a toxic release would have to be known. In fact, science has just begun to explore the toxic effects of all the pollutants we are releasing into the environment, as Steingraber points out:
The rapid birthrate of new synthetic products that began in 1945 far surpassed the ability of the government to regulate their use and disposal. Between 45,000 and 100,000 chemicals are now in common commercial use; 75,000 is the most frequently cited estimate. Of these, only about 1.5 to 3 percent (1,200 to 1,500 chemicals) have been tested for carcinogenicity. The vast majority of commercially used chemicals were brought to market before 1979, when the federal Toxics Substances Control Act (TSCA) mandated the review of new chemicals. Thus, many carcinogenic environmental contaminates likely remain unidentified, unmonitored, and unregulated. Too often, this lack of basic information is paraphrased as "there is lack of evidence of harm," which in turn is translated as "the chemical is harmless." (179)
O'Brien is even more blunt about the lack of knowledge of the toxic effects of many chemicals, and how risk assessment implicitly uses this lack of knowledge to conceal the harmful effects of pollution. Ignorance is not bliss, according to O'Brien:
Most risk assessments are unscientific because they cannot estimate how much of a damaging activity poses no risks or "insignificant" risks.
Imagine someone assessing how many Bloody Marys it is "safe" to drink in one day, but measuring only the tomato juice and not the vodka. The assessor might conclude that drinking a gallon of Bloody Marys (i.e. tomato juice) would pose no danger at all. If someone enthusiastically relied on that risk assessment, the effects could be fatal.
Industry and government enthusiastically rely on the conclusions of many such partial risk assessments. The results are
generally destructive, and often fatal. (180)
Besides being radically incomplete, the information used in risk assessments is highly abstract, disconnected from the actual effect toxins would have in the environment. Typically, estimates of exposure come from laboratory studies on animals. One group of animals is exposed to the toxin and another group, which is like the first in every other way possible, is not. After a period, the two groups of animals are compared. This kind of comparison, the controlled study,does produce vital information about a toxic effects of chemicals. However, it is not going to tell us everything we need to know about what effect the toxin is going to have in the real world. In the real world, people are simultaneously exposed to many different toxins, each of them having a different effect on the body, this one suppressing the immune system, that one causing DNA damage, another one disrupting liver function, and all of them making the body more vulnerable to the effects of the others. As we have found in the treatment of AIDS, combinations of drugs can be much more effective than any are alone. Similarly, the combined effects of a toxic soup of environmental exposures would likely be greater than the sum of the effects would be separately.
For example, when farmers are spraying Roundup, a herbicide that kills all plants, on their fields, they often add a small amount of ammonium sulfate to the mix. By itself, ammonium sulfate is only a fertilizer, not even very toxic, except in very large doses. However, when combined with Roundup, it significantly increases the toxic effect on plants. Acting as a fertilizer, the ammonium sulfate stimulates the plants to take in more Roundup, making the same amount of herbicide more effective at killing. The differences between spraying on straight Roundup and adding a little bit of ammonium sulfate are considerable, as I have seen myself. With just Roundup, the weeds take weeks to die, and sometimes they don't. With just a little fertilizer added, they fall over, turn brown, and die much more quickly.
Every competent farmer knows this, many people are living with AIDS now because of combinations of drugs, and yet this notion that combinations of things can have a synergistic greater than the parts is "unproven" to the technicians of risk assessment. Or at least ignored. Risk assessments invariably do not assume the synergistic effects of different chemicals. They simply add the toxic effects of various emissions together from, say, an hazardous waste incinerator-which would be a long list of different kinds of dioxins, furans, undestroyed PCBs and heavy metals-and then decide whether the risk is acceptable or not. They do not consider how these chemicals will interact or how they will react when people are also exposed to chemicals in the environment from other sources like pesticides, fertilizer, and solvents. They can't do this because, in addition to the incomplete knowledge about the toxic effects of many chemicals, there is almost no science out there to quantify the synergistic effects of combinations of toxic chemicals. Indeed, there probably is never going to be that kind of science because the effect of exposure to multiple combinations of toxins is simply too complex to research-at least by conventional quantitative means. Try to imagine all the chemicals that need to be studied-but largely haven't been yet-for their toxic effect. There's tens of thousands of them. Now imagine studying these in combinations-say three or four at a time. To cover everything, you would drop one chemical out at a time, replacing it with another, because, remember, we are looking for the effect of a combination of chemicals, not a single chemical. Then, when we finished with that, we would have to try varying the doses of chemicals relative to each other. By the time all the bases were covered for a truly complete risk assessment, the number of studies needed would grow exponentially, far beyond the ability of the government to finance. In short, there really is no way of calculating a realistic risk assessment of the combined effects because there really is no way to know what the specific effect is going to be when different chemicals are combined with other chemicals in the environment or in someone's body. The world is just too complex to realistically assess risk.
Although the rule on incineration in Montana does take the greater vulnerabilities of children and vulnerable adults into consideration, risk assessments also understate the toxic effect because they are typically calculated only for health male adults. Effects on children are often ignored, as a National Academy of Sciences committee comprised of pediatricians, chemists, developmental biologists, and food scientists reported to U.S. Senate Committee on Agriculture:
The committee noted that children are not little adults. They are more heavily exposed to toxicants in the environment because, pound for pound, they eat more food, drink more water, and breathe more air than adults. In addition during childhood, children's brains, lungs, reproductive organs, endocrine systems, and immune systems are all growing and developing, with synapses being established and immune tolerances being built. All of these developmental processes are extraordinarily complex and can be disrupted by toxic chemicals.
The NAS committee also evaluated risk assessment, as practiced by the EPA. There are underlying assumptions in the
risk-assessment methods that underestimate children's exposure. For example, the prototypical person used in risk
assessment is a 70-kg young adult male-not a child, an elderly person, or a person with a chronic disease. A second
problem is that traditional risk assessment calls for examining only one chemical at a time, whereas people are often
exposed to multiple agents from the same foods and from multiple sources. For instance, a pesticide analysis was
conducted by the Environmental Working Group, an environmental advocacy group based in Washington, D.C., on
strawberries, apples, and pears taken directly from grocery shelves. They found as many as seven different pesticide
residues on some of the fruits and vegetables. Yet, when risk assessments are done, each chemical is considered in
isolation, and the assumption is made that no other chemicals exist on the fruits and vegetables. In addition, the wide
range of residue levels make it difficult for investigators to conduct valid assessments. The concept that synergism or
interaction of chemicals is possible is not considered in the risk-assessment models. (181)
Besides the problems with synergy, vulnerable populations, and incomplete knowledge, another problem that risk
assessment typically ignores is baseline risk, the risk from toxins already present, accumulating in the background from other
sources. With 7.77 billion pounds of toxins being released into the environment annually, it should be a mere truism that we
live in a polluted environment, and, because of it, we all already have significant body burdens of various toxins. If these
toxins have a threshold, a point below which there is little effect but above which a little bit more suddenly becomes toxic,
then adding a little bit more to a body burden that is already high is an entirely different kind of risk than if the body burden
is nonexistent or low. As O'Brien argues:
(E)stimates of "safe" exposure to a particular hazardous substance or activity are indefensible because they pretend that
we are not already damaged, or contaminated . . . or otherwise under stress. Risk-assessment scientists assume in their
calculations that the activity they are analyzing is being experienced by a healthy, uncontaminated, unstressed organism
living in a healthy, unpolluted world. These scientists or risk assessors are pretending the real world doesn't exist. (182)
Not taking into account background accumulations results in risk estimates that are entirely divorced from reality. For instance, as I found out when I participated in state rule-making on incinerators, the Department of Environmental Quality's rules governing the permitting of incinerators in Montana allow the department to give a permit to an incinerator without in any way taking into account background exposures in the risk assessment. An incinerator may be entirely "safe" by state standards, releasing no toxins at a level that will cause cancer in more than 1 in 100,000 people, but no account is taken of the possibility that another identical incinerator may be nearby, or even a hundred such incinerators may be nearby. Without any of these incinerators releasing an "unsafe" amount of toxins, enough of them together could theoretically release enough toxins to kill virtually everyone for miles around, and yet, by the standards the DEQ, nothing about these emissions would be classified "unsafe." DEQ and DPHHS could walk among the bodies, strewn here and there, shrug their shoulders, and tell us no state emission standards were ever violated.
Honest, I am not making this up. When I blasted the rule before the Board of Environmental Review, the DEQ did not object to the logic of my characterization. (183) That's simply the way things are in Montana. Background accumulations of toxins are not taken into account in risk assessments. The presumption starting every risk assessment is that everyone is always living in a perfectly unpolluted environment. (Although I was never offered an explanation for not including background accumulations in a risk assessment, my guess is that the reason they aren't included is that doing so would deny polluters equal access to the risk they impose on the public. Limiting background accumulations would mean that some companies would be denied permits for no better reason than that risk was accumulating and they got there too late to get their fair share.) Without including background accumulations, however, risk assessments are no more honest than the entries people get in the mail from lotteries like the Publisher's Clearing house, where they tell you that you are on a short list to win millions. The real odds are entirely different from the ones they tell you about. When people see a risk like 1 in a million in a risk assessment, they will think that is the risk they face. However, in the fine print, things read a little differently . . .
O'Brien sums up her assessment of risk assessment this way:
Risk assessors can't account for all the conditions in the real world. They can'taccount for all the toxic substances
already contaminating and affecting people and other species out in the world. They can't account for individual
sensitivities. They therefore simply put on blinders and proceed full steam ahead, estimating the safety of some substance
or activity as if the real world doesn't exist. (184)
And if risk assessors are confronted with this:
The most they answer (if they answer at all) is that all other assumptions in their risk assessments are "conservative." By
this the assessors imply that the numbers they have put into their formulas assume "unrealistically" large exposures and
"unrealistically" high toxicity. In this way they claim even if they haven't considered all types of stresses . . . they have
surely overconsidered some of the stresses. At that point, they have left science, and you are left arguing with numbers that
might as well have been pulled out of thin air. (185)
For these and other reasons, risk assessment is simply an inadequate tool to protect public health. While I would like to
argue that risk assessment is unconstitutional whenever it is used in Montana, I am simply going to argue that it is
unconstitutional as applied here in this case. Because they are so abstract, DPHHS must not use the inferences drawn from
any sort of risk assessment to justify not investigating or limiting the investigation of a cancer cluster. Ignoring the
synergistic effect of combinations of chemicals, ignoring baseline accumulations of chemicals, and frequently presuming that
the person at risk is not more vulnerable than a "typical" person, risk assessments are disconnected from the real world, and
cannot reflect what is actually happening to real people, who inevitably are being exposed to wide combinations of toxins,
who may have had a long history of toxic exposures, and who may have a compromised immune system, be a child, or be
elderly.
Using risk assessments to say that a high rate of cancer can't be real because all of the toxic exposures are "safe" would be
using a hypothetical conclusion to dismiss actual results; it would be ignoring the hard truth of numbers to prop up a
politically convenient methodology that only benefits the bottom line of polluting corporations; and it would have the effect
of denying the people living in the cancer cluster their right to a clean and healthful environment. If it is going to protect
public health, DPHHS must deal with reality when it confronts a cancer cluster, what is actually happening to people, not
what a risk assessment says should happen.
One night, the joke goes, someone was walking home and found a drunk outside a bar, on his hands and knees underneath a street lamp. The drunk told the passerby that he was looking for his keys. Wanting to be helpful, the passerby got down on his hands and knees, offering to help. They went on like this for awhile, staring intently at bare pavement, and when it was clear no keys were to be found, the passerby asked, "Are you sure you lost your keys here?"
"Oh no," the drunk said, "I lost them out in the parking lot. The light is just so much better here."
This is an old joke, and I apologize for offering it. Still, it does describe, I think, the way health departments have been looking for the causes of cancer clusters. They have been looking where the light has been the brightest, not where the key is. Many epidemiologists who have investigated cancer clusters insist that they almost always fail to find statistically significant explanations for the cause. Investigating cancer clusters is a vain and futile effort, too many public health officials and epidemiologists have told us, unlikely to yield scientifically meaningful results. As a result, investigating cancer clusters has become little more than a public relations effort for health departments, done mostly to allay public anxiety and take the heat off public officials. (186)
One of the purposes of this lawsuit is to contest the methodological framework that the CDC, state health departments, and even too many university epidemiologists have imposed on the study of cancer and chronic diseases. (187) By limiting the possibilities for revealing the truth about cancer to quantitative method and risk assessment, which is to say to a formal, highly mathematical, analytic reductionism, they have framed the exploration of cancer with presumptions, such as the randomness of cancer and the effectiveness of reducing a whole to its parts, that cannot reveal much of its true nature (188) While quantitative method has been helpful in revealing some aspects of disease clusters, and while risk assessment does identify part of the harm being done by environmental pollution, they are far from what is needed to adequately protect public health. These methods have failed to help us much, and the best evidence for concluding this is that the CDC, state health departments, and university epidemiologists have gotten nowhere using them. They are as baffled now about why cancer clusters happen as they were decades ago. (189) And if they continue to use these methods, bending knee faithfully in the church of positivism and random distributions, they will probably know as little about the toxic causes of cancer a century from now as they do now.
Health departments and university epidemiologists could do much better at protecting public health than they are doing, and the reason they have been failing is that they just aren't looking in the right place.
As Sandra Steingraber observed, statistical testing is a very blunt tool when it comes to investigating cancer clusters. To get any kind of sampling validity, you have to have large numbers. (190) However, cancer clusters are typically limited to small numbers-the people living near a contaminated well, a hazardous waste dump, or an incinerator plume. Although the numbers of people exposed the worst are going to be limited, often to only a couple of hundred or thousand, the first thing health departments do upon getting a complaint is apply statistical tools that start by presuming that it is a random result and then require much larger numbers than are available to come to any conclusion. When proof of an environmental link inevitably fails to appear, failing to confirm that the cluster is not random or that any of the toxic exposures are unsafe, they lecture the public about the need to be more scientific and to not make inflammatory statements about cancer, getting everyone all excited about nothing. (191)
To put it simply, this policy of shiftless denial is blind ideology, not good science, as unjust as it is inadequate. Applying tests for randomness when there is no justification for it, using risk assessments to say everything is safe when the number of people suffering indicates it is not, this policy has the effect of ignoring the links between people's suffering and environmental pollution, letting corporations get away with literal murder.
Whatever it is, cancer is clearly a very complex disease with multiple causes, involving exposures to tens of thousands of industrial chemicals, interactions with a wide range of natural viruses, bacteria, and complex environmental chemicals, as well as being situated within a unique personal, social, political, and economic circumstance. Because cancer develops out of the interplay of all these complex circumstances, the methodology of statistical testing is simply not subtle enough to elaborate the true complexity of the links between cancer and environmental pollution. Needing to mathematically operationalize the terms of analysis to meet their criterion of "truth," significance testing and risk assessment have to oversimplify the complexity of the real world, chopping it up into categorical blocks. In doing this, they conceal with simplification the complexity that needs to be revealed if public health is going to be protected.
Before any real progress can be made in preventing it, health departments are going to have to accept that cancer is uniquely complex and deeply situated in people's lives, and that this means that there are simply not going to be any universal causes to it-that is to say, statistically plausible explanations that would apply everywhere, every time. When cancer appears, it is most likely the result of a unique interplay of events and exposures within complex physiological and ecological systems, a very specific interplay of pollutants and situation--this chemical causing a genetic mutation, that chemical suppressing the immune system, another chemical damaging the liver. None of these things by themselves may be sufficient to be the trigger, but only have that effect when they combine together in a uniquely complex situation. Given this complexity, and the likelihood that every cancer cluster will always be unique, there is little hope that any full explanation of any one cancer cluster will be much more than a suggestion about what is causing another similar cancer cluster. Each will invariably involve different chemicals, different concentrations of chemicals, different environmental interactions, and different vulnerabilities in the people exposed to them.
The CDC, state public health departments, and epidemiologists have been getting nowhere investigating cancer clusters because, seeking a general and mathematically formal explanation, they have presumed that this cancer cluster is like that cluster, thinking that if a chemical didn't cause the problem in one circumstance, it can't be causing it in another. Propelled by the methodological needs of statistical analysis and not by the nature of the problem, this assumption is entirely unwarranted. Just because a chemical may not be an efficient cause of cancer in the laboratory says nothing about what it may do amid the complexity of the real world.
In other words, epidemiologists are too tied up in the methodology of the formal experiment, which requires analytic reduction, to really deal with the complex synergies of cancer. They need to explore the causes of cancer more like Aristotle would, less like Galileo. (192) They need to look at cancer in its natural state, in the community where it happens, instead of trying to abstract it from its context, trying to derive a universal law that will explain it.
Below is a quotation from my book, Modernity and Technology: Harnessing the Earth to the Slavery of Man. It describes how that the modern experimental method, for all its strengths, can conceal a much more complex reality:
Despite seeming similarities that have been interpreted as the origin of the modern experiment in ancient knowledge, a decisive difference remains, separating their techniques of knowledge from ours. "Twisting the lion's tail," as Francis Bacon would say, the modern experiment disturbs things, separates them from their context and subjects them to rigorously planned control to reveal the mechanical causes determining them. Invoking this utopian perspective, it begins by laying down a law as a basis for the experiment, abstracting the thing from the forces, variables, or complications that make impossible a formal observation of the facts or any possibility of the determining causes that will either prove or disprove the law. Where Aristotle sought to observe things in their natural condition, surrounded by all the four modes of occasioning in all their complexity and interplay so that he could know the thing as it is, the modern experiment, invoking its utopian world, establishes artificial, controlled, and planned circumstances, eliminating the complexity and interplay of "irrelevant" variables based on an already known law so that it can know the simple mechanical causes of the thing. It is this controlling based on an already known, this reduction to mechanical procedure, that makes the truth of the modern experiment mathematical, and distinguishes it from ancient and medieval observation.
Contrary to the experiment's mythology, its self-proclaimed progress is not a progress that all ages could acknowledge and revere. By abstracting itself from the natural situation of the thing, by reducing its truth to a mechanical cause or a mathematically formalizable entity, the experiment conceals many things. For instance, the use modern Agribusiness makes of chemicals to kill pests directly that damage crops could only have arisen and been acknowledged as an improvement of technique over the older methods because the experiment concealed the true complexity of nature from itself. Aristotle, for instance, would not have been convinced that the development of a chemical that kills a pest simply and directly would be a useful tool, an advance beyond old techniques, because, by abstracting the insect from its environment, the experiment ignores the place of the insect in the nature of things, making it vulnerable to external consequences that it could not anticipate. What other insects would the chemical kill? What effects would that have on other life forms? What would it do to the soil after many years of use? What would it do to the farmer, her state, her gods? In order for a chemical pesticide to be judged an improvement for Aristotelian science it would have to open itself up to many more questions than it does in modern experimental science. An Aristotelian science might well reveal that organic farming is a much better technique than chemical farming because it looks at a technique from the vantage point of its place in the whole of things. Chemical farming can be revealed as an improvement or as progress only by invoking the utopianism, the unsituated perspective, of modern science. (193)
Donna Haraway agrees with me. (194) In contrast to the universal and timeless aspirations of experimental knowledge, she argues that only partial vision, limited voice, and situated knowledge is going to reveal the complexity of the world. Disembodied knowledge, unlived truth, and genderless science reveals nothing because it is connected to nothing. It isn't alive. To act in the world and make it better you have to be someone, be somewhere, tied to institutions, related to people-and be limited by that body, place, and time. You have to have a place, a home.
This is why so many cancer cluster investigations have failed, I contend. Seeking formal mathematical validity at the expense of all else, the CDC, state health departments, and too many university epidemiologists have never realized that cancer has a home, that it is always uniquely bound up in the life situations of the people affected by it.
To put it in more practical terms, instead of processing statistics in their offices in Helena, DPHHS needs to drive all the 500 miles out to Baker and pay attention to what is going on here. This would include talking a lot with the victims of cancer, testing their bodies for chemicals, and then testing the environment that they live and work in. Perhaps, in a variation on the experiment's method, they would test the combination of chemical exposures they found prevalent in the bodies of the victims on laboratory animals to see what effect the combinationwould have. Of course, such an experiment would be scientifically meaningless because, failing to isolate the various chemicals, it would not be relevant to any other situation than Fallon County's, but this sacrifice of general relevance might greatly enhance the specific relevance for Fallon County. It might tell us if in fact the complex combination of chemicals that people have accumulated in their bodies in Fallon County is causing the cancer.
Public health investigators should go back to their roots and investigate cancer clusters the way that John Snow investigated a cholera cluster in 1845. Snow was the famous physician who was able to stop the cholera epidemic in London by removing a pump handle from a contaminated well--discovering the waterborn nature of the disease 32 years before the discovery of Vibrio cholerae, the bacteria that caused it, and a decade before Pasteur had proved that bacteria caused infectious diseases. (195) Unlike modern epidemiological studies, Snow did not make cholera cases into a short list of variables, locating them on a grid where they could be statistically tested. Instead, he made an opened history of each case, looking for similarities and differences in diet, occupation, exposure, habit, location, and so on as the cases accumulated. He worked with specific people, their lives, their habits, their specific activities, not with statistical abstractions. For instance, to explain why a woman developed cholera outside the cluster, away from the area where the contaminated well was located, Snow discovered that, far from contradicting his thesis, the apparent aberration actually supported his thesis that a single well was the source of the problem:
I was informed by this lady's son that she had not been in the neighborhood of Broad Street for many months. A cart went
from Broad Street to West End every day, and it was the custom to take out a large bottle of water from the pump in Broad
Street, as she preferred it. The water was taken on Thursday, 31st August, and she drank of it in the evening, and also on
Friday. (196)
If Snow had investigated cholera the way health departments investigate cancer clusters, using tests of statistical
significance to measure the strength of associations, this "aberration" who lived quite a distance from the well might well
have been taken as proof that cholera was, at least to some extent, randomly distributed, perhaps linked to bad air currents,
which might briefly and randomly flow anywhere. (Remember, Snow was before the germ theory of infection. He could
have gone off in many different directions.) And this might have led him astray, leaving the pump handle on the
contaminated well. But instead, Snow paid close attention to the details, going out and looking for an explanation for
aberrations, instead of chalking them up random nature.
If health departments investigated cancer the way that Snow investigated cholera, we might just find that cancer is not nearly as randomly distributed as is presumed when the framework of statistical significance tests are imposed on the data. The CDC, the health departments that follow its bad example, and the university epidemiologists that lend their support to both have become prisoners of statistical method. The "random" character of cancer is more an effect of the method used to reveal it than it is a reality of the disease itself. Drunk with the power of computer-supported statistical analysis, the CDC, state health departments, and university epidemiologists have been looking where the light is the brightest, not where the key is. Using a method that presumes randomness, they can only "discover" that cancer is randomly distributed, when in fact every "aberration" may have an entirely adequate explanation, if they would only adopt a method that would allow them to look for it. As in the Steingraber's example of the eleven blue men that I used earlier, cancer may appear to be randomly distributed only because of long latency periods and multiple causes, but if you could trace the causes back to their origin, you might find that cancer is anything but a random disease, striking without environmental exposure. I do not require that the CDC, state health departments, or university epidemiologists throw away their computers. Quantitative method can shed a lot of light on disease distributions, increasing our understanding of them. Nevertheless, the key to cancer may be hid in a place where this light cannot shine into. Health departments need to recognize the limits of any one method. They need to understand that some conceal what others would reveal.
They also need to recognize that their goal, protecting public health, is not identical with the goal of research scientists who are studying public health and disease. They each have different goals, and must take different precautions pursuing those goals. The goal of the research scientist is knowledge, and taking precaution to avoid falsely reporting as knowledge something that may only be the result of random distributions is appropriate for them. In studies of drug trials, toxins, surgical procedures, and so on, research scientists are always dealing with many unknown variables, some of which may be randomly distributed. They don't really know, and because they don't, they have to be on guard against random distributions. In a cancer treatment, for example, it may just happen that a researcher winds up with a string of patients who respond better than most would, or worse than most would to a treatment. To make sure that they don't publish false information, doing everything they can to rule out randomness as an explanation might be a reasonable precaution for research scientists, even when the justification for random distributions is lacking or weak. Depending on the nature of the study, statistical significance testing may be an appropriate tool to help do this.
However, the goal of public health departments is not primarily knowledge; it is public health. Where research scientists
are trying to reduce uncertain knowledge, increasing our understanding of disease, public health departments must cope with
uncertain knowledge, protecting public health as best as is possible when knowledge is inadequate and limited. A research
scientist may shrug their shoulders in face of a cancer cluster, saying they really can offer no explanation for it, concluding
that perhaps it is merely random. A public health department does not have that luxury, and cannot dismiss a high rate of
disease as merely random. When a disease rate is higher than it would be in a clean and healthful environment, it must take
precautions to protect public health, using the best information available, however inadequate, to reduce the disease rate. Of
course it may error, removing chemicals from the environment that were not a part of the problem or working on a cancer
cluster that truly is a random result from time to time, but when a disease rate is higher than it would be in a clean and
healthful environment, error in favor of public health is always an acceptable risk. People have a right to a clean and
healthful environment, toxic chemicals don't.
It is significant, I think, that the preamble of Montana's constitution begins as a celebration of Montana's natural environment--its beauty, the grandeur of its mountains, the vastness of its rolling plains. After describing what Montanans are most grateful to God for, it goes on to list a variety of goals for our life together-improving the quality of life, promoting equality of opportunity, and securing liberty for present and future generations:
We the people of Montana grateful to God for the quiet beauty of our state, the grandeur of our mountains, the vastness
of our rolling plains, and desiring to improve the quality of life, equality of opportunity and to secure the blessings of
liberty for this and future generations do ordain and establish this constitution.
The preamble unmistakably establishes the environment as a primary and essential good affecting the quality of our life, something that the citizens of the state have a responsibility to out of gratitude toward God and for the benefit of present and future generations. In case the point is missed, Article II, Section 3 explicitly recognizes an inalienable right to a "clean and healthful environment," and then, in a move that is perhaps the first in the modern constitutional tradition, insists upon personal responsibility for it:
Inalienable rights. All persons are born free and have certain inalienable rights. They include the right to a clean and
healthful environment and the rights of pursuing life's basic necessities, enjoying and defending their lives and liberties,
acquiring, possessing and protecting property, and seeking their safety, health and happiness in all lawful ways. In
enjoying these rights, all persons recognize corresponding responsibilities.
Later, in Article IX, Environment And Natural Resources, in Section 1 the novel notion of a responsibility for environmental
quality returns, this time not only as a personal responsibility but as a collective one as well, one also held by the state and
the legislature:
Protection and improvement. (1) The state and each person shall maintain and improve a clean and healthful environment in Montana for present and future generations. (2) The legislature shall provide for the administration and enforcement of this duty. (3) The legislature shall provide adequate remedies for the protection of the environmental life support system from degradation and provide adequate remedies to prevent unreasonable depletion and degradation of natural resources.
Although the judiciary is not specifically charged with the duty to maintain and improve a clean and healthful environment, it clearly is included in Section 1, when the duty is given to "the state and each person." The word "adequate," used 2 times in Section 3, would, on the plain language of it, give specific responsibility to the judiciary to review the adequacy of the remedies provided by the legislature. With this language, the legislature or a state agency whose policy is set by the legislature, like DPHHS or DEQ, cannot absolve themselves from an independent review of their efforts. They cannot just say that whatever they do is good enough, and then leave it at that. If there is a scientific or aesthetic reason to doubt the adequacy of either the efforts of the legislature or the efforts of a state agency, or even a private party to protect the environmental life support system from degradation or prevent unreasonable depletion and degradation of natural resources, the judiciary must review the adequacy of these efforts and order more appropriate efforts.
In other words, the word "adequacy" in Section 3 means that truth matters, that evidence counts, that review is necessary. Section 3 offers up a way to test the adequacy of the efforts of the state and the legislature with evidence. If the evidence shows that the environmental life support system is being degraded, as it would be if, say, an incidence of a certain form of cancer was higher than it would be in a clean and healthful environment, the judiciary, the branch of government where an appeal to evidence is formally heard, is accountable for it. The judiciary cannot ignore evidence of inadequate environmental stewardship by state agencies and the legislature, letting them neglect their duties. If evidence is presented of state or individual neglect of the environment, the judiciary must hear it and resolve it in favor of a clean and healthful environment. That is the policy set by the Constitution. Like everyone else in Montana, the judiciary has a duty to act to maintain and improve a clean and healthful environment.
In Montana Environmental Information Center and et al. v. Montana Department of Environmental Quality, the Montana Supreme Court ruled that the right to a clean and healthful environment is a fundamental right, requiring strict judicial review:
53.Applying the preceding rules to the facts in this case, we conclude that the right to a clean and healthful environment is a fundamental right because it is guaranteed by the Declaration of Rights found at Article II, Section 3 of Montana's Constitution, and that any statute or rule which implicates that right must be strictly scrutinized and can only survive scrutiny if the State establishes a compelling state interest and that its action is closely tailored to effectuate that interest and is the least onerous path that can be taken to achieve the State's objective.
54.State action which implicates those rights provided for in Article IX, Section 1 would normally not be subject to strict scrutiny because those rights are not found in Montana's Declaration of Rights. Those rights would normally be subject to a middle-tier of scrutiny because lodged elsewhere in our state constitution. However, we conclude that the right to a clean and healthful environment guaranteed by Article II, Section 3, and those rights provided for in Article IX, Section 1 were intended by the constitution's framers to be interrelated and interdependent and that state or private action which implicates either, must be scrutinized consistently. Therefore, we will apply strict scrutiny to state or private action which implicates either constitutional provision. (197)
The language in these sections the court was referring to cannot be taken lightly, ignored, or dismissed as mere poetry, then. Put simply, the residents of Fallon County, as well as every citizen of Montana, has a right to an environment that will not cause them to get cancer. From the transcripts of the constitutional convention, where the founders were discussing these sections, they clearly meant what the language plainly says. They wanted an environment that not only did not cause disease but was clean and not degraded, as Delegate McNeil explained why the word "clean" was also used to modify the word "environment":
The majority felt that the use of the word "healthful" would permit those who would pollute our environment to parade in
some doctors who could say that if a person can walk around with four pounds of arsenic in his lungs or SO2 gas in his
lungs and wasn't dead, that that would be a healthful environment. We strongly believe-the majority does-that our
provision-or proposal is stronger than using the word "healthful." (198)
The right to a clean and healthful environment is a very high standard, leaving very little room for misinterpretation, the
delegates believed. In adopting this language, the delegates were intending that the environment in Montana not merely not
be a cause of disease, but also not be allowed to be degraded, as Delegate McNeil explained:
We did not want the Supreme Court of this state or the Legislature to be able to say that the environment in Montana, as
we know right now, can be degraded to a healthful environment. . . . I submit if you will read that majority proposal again
and again, you will find that it is the strongest of any constitution. . . . (199)
Arguing against an amendment that would have weakened the language, Delegate Foster also gave the following defense of
the language finally adopted:
I feel that if we, as a Constitutional Convention of Montana, use our line of defense on the environment on the basis of
healthful, then we, in fact, might as well forget it, because what I'm concerned about in Montana is not a healthful
environment. This country is going to have to address itself to the question of a healthful environment. What I'm concerned
about is an environment that is better than healthful. If all we have is a survivable environment, then we've lost the battle.
We have nothing left of importance. The federal government will see to it one way or another, if it's in its power, that we
have an environment in which we can manage to crawl around or to survive or to in some way stay "alive." But the
environment that I'm concerned about is that stage of quality of the environment which is above healthful; and if we put in
the Constitution that the only line of defense is a healthful environment and that I have to show, in fact, that my health is
being damaged in order to find some relief, then we've lost the battle; so I oppose this amendment. (200)
The notion that a merely healthful environment is not good enough, that it should be better than healthful, came as
something of a surprise to me. The common usage of "health" today already contains within it the notion of an abundance
of well-being, an unqualified zest for life and not just the absence of disease. The World Health Organization, for example,
defines health in the preamble to its 1948 constitution as "a state of complete physical, mental, and social well-being and not
merely the absence of disease or infirmity." (201) This notion of health as being an abundance of well-being, not merely as the
absence of dysfunction, is also hardly a rare one in the American tradition. Walt Whitman defined health this way:
(T)he condition (in which) the whole body is elevated to a state by other unknown-inwardly and outwardly illuminated,
purified, made solid, strong, yet buoyant. A singular charm, more than beauty, flickers out of, and over, the face-a curious
transparency beams in the eyes, both in the iris and the white-the temper partakes also. The play of the body in motion
takes a previously unknown grace. Merely to move is then a happiness, a pleasure-to breathe, to see, is also. All the
beforehand gratifications, drink, spirits, coffee, grease, stimulants, mixtures, late hours, luxuries, deeds of the night seem
as vexatious dreams, and now the awakening; many fall into their natural places, wholesome, conveying diviner joys. (202)
But perhaps Delegates Foster and McNeil were concerned not so much with the commonly accepted notion of what health
is as they were with the ways it might be corrupted by the likes of Anaconda Copper, W.R Grace, and Montana Power.
Under corporate wordsmithing, "health" would likely become merely the ability to show up to work for them, last the day,
and return for a couple more. That perversion could not be allowed, and so the delegates to the convention left a clear
record that the environmental standards must be high enough not only to prevent disease but also prevent environmental
degradation. Not just any possibility of disease being caused by environmental pollution was forbidden but also any harm to
the environmental life support system was forbidden as well. And so, the standard was not just physiological harm to
humans but harm to the environmental life support system. Either was cause for invoking the right to a clean and healthful
environment, and insisting upon adequate measures to maintain and improve the environmental life support system. When
the convention delegates referred to "the quiet beauty of our state," "the grandeur of our mountains," and "the vastness of
our rolling plains," they wanted it all to stay that way, if not even become better.
In Montana Environmental Information Center v. Montana Department of Environmental Quality, the Montana Supreme Court, referring to both the Constitution and the transcripts of the Constitutional Convention, ruled that the state does not even need to wait until proof that public health is being damaged before it acts to protect the environmental life support system from degradation. Writing for the majority, Justice Trieweiler concluded:
67.We conclude, based on the eloquent record of the Montana Constitutional Convention that to give effect to the rights guaranteed by Article II, Section 3 and Article IX, Section 1 of the Montana Constitution they must be read together and consideration given to all of the provisions of Article II, Section 1 as well as the preamble to the Montana Constitution. In doing so, we conclude that the delegates' intention was to provide language and protections which are both anticipatory and preventative. The delegates did not intend to merely prohibit that degree of environmental degradation which can be conclusively linked to ill health or physical endangerment. Our constitution does not require that dead fish float on the surface of our state's rivers and streams before its farsighted environmental protections can be invoked. The delegates repeatedly emphasized that the rights provided for in subparagraph (1) of Article IX, Section 1 was linked to the legislature's obligation in subparagraph (3) to provide adequate remedies for degradation of the environmental life support system and to prevent unreasonable degradation of natural resources.
68.We conclude, therefore, that the District Court erred when it held that Montana's constitutional right to a clean and healthy environment was not implicated, absent a demonstration that public health is threatened or that current water quality standards are affected to such an extent that a significant impact has been had on either the Landers Fork or Blackfoot River.
69.We conclude that the constitutional right to a clean and healthy environment and to be free from unreasonable
degradation of that environment is implicated based on the Plaintiffs' demonstration that the pumping tests proposed by
SPJV would have added a known carcinogen such as arsenic to the environment in concentrations greater than the
concentrations present in the receiving water and that the DEQ or its predecessor after studying the issue and conducting
hearings has concluded that discharges containing carcinogenic parameters greater than the concentrations of those
parameters in the receiving water has a significant impact which requires review pursuant to Montana's policy of
nondegradation set forth at § 75-5-303, MCA. (203)
There are several different reasons why MEIC v. DEQ is relevant to the state's failure to address the high cancer rates in
Fallon County. First, as we saw above, it found that the right to a clean and healthful environment holds the same strict
standard for judicial review as other fundamental rights, such as voting, speech, press, assembly, and so on. Secondly, it
begins to set a standard for how much environmentally caused disease is too much. If, as Justice Trieweiler wrote, "Our
constitution does not require that dead fish float on the surface of our state's rivers and streams before its farsighted
environmental protections can be invoked," then it also seems likely that state action most certainly would be required when
a community had an incidence--as Fallon County does--of childhood leukemia that is, at the very minimum, 10 times higher
than the national average, an incidence of breast cancer that is among the highest in the world, and an incidence of colorectal
cancer that is about 7 times higher than neighboring counties.
Besides protecting the environmental life support system from degradation, one of the goals of the court's decision was to prevent the merest possibility of cancer in humans. The release of arsenic was an issue in MEIC v. DEQ because . . . "tests proposed by SPJV would have added a known carcinogen such as arsenic to the environment in concentrations greater than the concentrations present in the receiving water . . . " Trieweiler's opinion notes that although the plaintiffs had not proved that theirs, or anyone else's, health had yet been harmed by the release of arsenic, a known carcinogen, into state waters, the mere possibility that it could harm health was sufficient to justify review. With this move, the court was clearly showing an interest inprecaution, in preventing not just actual harm but possible harm. The court was takingprecaution to prevent cancer.
In Fallon County, it must be emphasized, we do not have just the possibility of harm, we have actual evidence of harm-high incidences of cancer, high incidences of birth abnormalities. So, if the state was obligated to take every precaution in MEIC v. DEQ, reviewing arsenic releases even when there was no proof of human harm, it certainly is obligated to do so here. A high cancer rate is evident, and the state must, at a minimum, take every precaution to insure that it is not caused by an environmental exposure.
It also means that, until it is proven otherwise, the state must presume that a high rate of a disease that has been scientifically linked to toxic exposures is caused by a toxic exposure. This might not always be the case, but the state must presume it until it proves otherwise, and, with this presumption, adequately warn the public of the toxic exposures that have been scientifically linked to the disease. For instance, various studies have linked childhood leukemia to household, garden, and agricultural use of pesticides, various solvents, benzene, and radiation. As soon as it is apparent that a high rate of a disease like leukemia is occurring in a community, the state must take the precaution of adequately warning the affected public of all the toxic links that science has identified for this disease, without even knowing which one is appropriate. Then, it must proceed with an through investigation to try and isolate the cause.
How Much Cancer is Too Much Cancer?
DPHHS will surely argue that a certain amount of cancer is the inevitable tragedy of the human condition, as I will agree is indeed the case. Some people are going to get cancer no matter what we do to prevent it, no matter how clean and healthful our environment. To a certain extent cancer is an inevitable result of the aging process, to some extent it is also the result of natural mutations in genes, and to some extent it is simply an inheritance. However, given the dramatic increase in the incidence of cancer since the start of the industrial age, given the wide variation of cancer that exists between industrialized and non industrialized countries, and given the massive amount of scientific evidence mechanically linking carcinogens, such as dioxin, in the environment to cancer, a very troubling and entirely unacceptable number of cancer cases have to come from toxic environmental exposures. To the extent that cancer and other chronic diseases come from environmental exposures, the state is obligated to take adequate action to warn the public and prevent these diseases from happening. That is clearly what our right to a clean and healthful environment is all about, actually having the quality of life that comes from living in a clean and healthful environment.
So, the question becomes: how do we decide in any given community whether the cancer we find is a part of the inevitable tragedy of the human condition or comes from environmental exposures? How much is too much? I submit that the plain language of the Montana Constitution sets a self-executing standard for how much cancer is too much. When it gives us a right to a clean and healthful environment, it requires that no community have an incidence of cancer higher than it would if the people had a truly clean and healthful environment. That is to say, to find out if the cancer rate is too much we look for a truly clean and healthful environment, one where people are not exposed to any toxins from environmental pollution, (perhaps a nonindustrialized Third World country, perhaps a rural community in Montana with little exposure to pesticides, little air and water pollution, and so on) and we compare the incidence rates of cancer in these unpolluted places, where that cancer rates would be taken as an indication of the inevitable tragedies of the human condition, with the incidence rates of the community under question. If there is a higher incidence rate in the community under question, it is too high, requiring state intervention to take adequate measures to clean up the environment. If it is not, it is adequate.
It should be really very simple-only we all know that it isn't going to be that simple. DPHHS, and its industry friends, can be expected to argue that this would be a departure from accepted practice. The accepted practice is to compare expected rates of cancer, as determined by either state averages or by national averages, with the rate of the community in question. This is what I did when I found that the incidence of childhood leukemia in Fallon County was, at the very least, 10 times higher than the national average. However, while this is the accepted procedure and, in this case, indicates a statistically significant excess of cancer, I contend that using state or national averages for the expected rate is a violation of the right to a clean and healthful environment because both state averages and national averages inevitably include within their number people living in very polluted environments-next to hazardous waste dumps, downwind of incinerators, downstream of chemical plants, or beside fields frequently sprayed with pesticides. As we saw from the most recent Toxic Release Inventory, around 7.77 billion pounds of toxic materials are released into the national environment annually. (204) The average national incidence of cancer includes virtually everyone exposed to this historically unprecedented toxic exposure.
Many Montanans do not live in an unpolluted environment, either. The average rates in Montana would include Libby, where we know people are dying at an unprecedented rate from toxic environmental exposures, Butte, where the largest Superfund complex in the nation is located, and Billings, which has some of the worst sulphur dioxide levels in the nation. How can a national or a state average of cancer possibly be what should be expected, that is to say, what would happen but for exposure to industrial pollution? Billings is not an unpolluted environment, what with all its refineries, and it is wrong to use it as a benchmark, even as only part of the average, to decide whether the incidence of breast cancer is too high in Fallon County, particularly since Billings, too, has one of the highest rates of breast cancer in the world.. Not having a breast cancer rate that is high in comparison to Billings, says nothing about how low breast cancer rates could be in Fallon County, but for toxic exposure from the environment. State or national averages include too many people who have suffered the consequences of toxic exposures, and their health simply cannot be used as a reflection of what disease rates would be if everyone were living in a clean and healthful environment.
If we focus on the plain language of Montana's Constitution, the right to a clean and healthful environment simply does not
allow DPHHS to use either state or national averages of cancer incidence as a standard of comparison to decide whether
community's cancer rate is caused by environmental exposures. The only valid rate to use for comparison is a rate from a
truly clean and healthful environment, one that is wholly unpolluted. Unless that is the standard used, we just won't know
what the right to a clean and healthful environment means.
I argued earlier that using statistical significance tests to decide whether a cancer cluster was real was bad science. Here, I want to argue that it is unconstitutional as well because evaluating whether a cancer cluster is real is like counting people to apportion voting districts. For both voting apportionment and cluster detection how a single person or how a single case of cancer is going to be counted determines whether they are going to represent anything or not, whether fundamental human rights are going to be respected or not. So, the requirements for one should be appropriate for the other. If using statistical sampling to apportion votes is unconstitutional, it is equally unconstitutional to use statistical sampling to identify cancer clusters.
In the Department of Commerce v. the U.S. House of Representatives, the U.S. Supreme Court upheld a lower court decision that statistical sampling was not an acceptable means for apportioning voting districts. The court held that the Census Bureau must count actual people, not statistically contrived ones, when it comes to apportioning voting districts, and I similarly argue that DPHHS must count actual cancer cases, not statistical contrivances, in deciding whether a cluster of cancer cases are environmentally caused, and thus subject to government remedy. Just as statistical sampling allowed the Census Bureau to ignore people who had been actually counted in the census in favor of people who were mathematically fabricated, testing for statistical significance in small communities has the effect of allowing departments of health to ignore high rates of cancer because they may be random. In other words, even though a case by case count of cancer may be high, it isn't counted because of a mathematical theory that the count may not really be as high as it actually appears. As with the Census Bureau's proposal, an actual count is ignored in favor of a theory of what the count really is. (Note: Age correction would not share this problem because with it every case is counted, they are just weighed differently so that populations with different age distributions can be compared. Age correction makes it possible to compare rates in different populations; inappropriate statistical significance testing makes it difficult if not impossible to compare rates in small populations with large ones.)
As Chief Judge Claude M. Hilton acknowledges in his opinion in Matthew Galvin, et al. v. William J. Clinton, et al., which was consolidated into Department of Commerce v. U.S. House of Representatives when the issue made it to the U.S. Supreme Court, there has always been a problem with undercounting in the census--not counting people who were low income, homeless, recent immigrants, part of a minority group typically discriminated against, or living in a rural area. (205) This undercount is a constitutional problem because, according to the Constitution, Article I, Section 2, Clause 3, the census is supposed to count everyone every 10 years and use the results to apportion voting districts:
Representatives and direct Taxes shall be apportioned among the several States which may be included within this Union,
according to their respective Numbers, which shall be determined by adding to the whole Number of free Persons,
including those bound to Service for a Term of Years, and excluding Indians not taxed, three fifths of all other Persons.
The actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and
within every subsequent Term of ten Years, in such Manner as they shall by Law direct.
If an area has many people who were likely to be undercounted, the district would be too large to represent them
adequately. The undercount favors Republicans, who are more likely to counted, and hurts Democrats, who are less likely
to be counted.
Under the Clinton Administration, the Census Bureau proposed to use a method called "statistical sampling" to help deal with the inevitable undercount in the census, correcting what it saw as a bias in favor of Republican representation. In the first stage of this proposal, the Census Bureau would do everything pretty much as it had always done, trying to get a direct count of as many people as it could. However, then, as Judge Hilton describes it in Glavin v. Clinton:
The second phase of the enumeration is the Integrated Coverage Measurement Survey, in which Census Enumerators will conduct interviews in a random population sample, separate from each state, to determine what proportion of people living in the sample blocks were included in the initial enumeration. The Census Bureau's plan will classify each of the country's seven million blocks into groups known as sampling strata based on the characteristics of the block's residents according to the 1990 Census results, such as racial and ethnic composition, proportion of homeowners to renters, etc. The Bureau will select a controlled scientific sample of these blocks and enumerators will then conduct an independent second roster and ICM interview.
Each person and each enumeration is then assigned to a unique poststratum, a group of persons having similar
probability of having been enumerated in the initial phase. The Bureau will then estimate the number of persons in each
poststratum who were correctly counted, missed, or over counted in the initial data phase. The Bureau will use that
estimate to create an adjustment factor for each poststratum, and then multiply the number of people counted in each in
each poststratum in the initial data collection phase by the appropriate adjustment factor to adjust the census
synthetically. Once the adjustment factors have been applied to each poststratum in a block, the statistically adjusted
population figures for each block will be aggregated at the tract, county, state, and national levels. This will be the
reported population number used for Congressional apportionment and other purposes. (206)
In other words, instead of perfectly counting everyone, the Census Bureau would try to get a very accurate sample of the nation, and then, using the conventional count that it would do everywhere, it would develop a statistical profile of different population strata, and then compare them with the sample as to such things as race, home ownership, response to the census, region, and rural/urban factors. From these comparisons, it would project what the "true" count was, and then adjust their actual count to reflect their "true" count.
Naturally, Republicans in the House of Representatives were not wild about the idea of a Democratic administration adjusting the actual count of the census to reflect what it was going to call a "true" count, however much it looked like science, and however justified the concerns about an undercount were. So Republicans in Congress took the Census Bureau to court, citing examples of actual people who would be counted by the census but who would be discounted in the adjusted census in favor of "virtual" people who did not exist. For example, according to the U.S. Census Monitoring Board, which is run by Congress:
The preliminary results of the A.C.E. (Accuracy and Coverage Evaluation) for 2000-consistent with the results in 1990-show that approximately one million people who filled out a census questionnaire would have had their records nullified in order for the adjustment methodology to work properly. It is not just Whites who would be eliminated or nullified through adjustment. Many Asian children would also be eliminated or nullified in the process. Many Black, Hispanic, Asian, and Native Hawaiian homeowners would also be likely eliminated. These "nullified" persons represent the elimination of real people who did their civic duty by returning their census forms.
For example, to determine a total net under count of the African-American population of 2.17 percent, the Bureau would
need to add and nullify hundreds of thousands of records within the census count. The Bureau, using the A.C.E.
methodology to adjust the census, would add approximately 845,000 records to the census for African-Americans to adjust
the undercount. In contrast, the Bureau would negate or nullify approximately 100,000 records for those post-strata in the
African-American community where they estimated overcounts. These post-strata included approximately 35,000
African-American men and women living in large metropolitan areas, over 50 years of age, who own their own homes. The
Bureau identified an overcount for these individuals, while at the same time identifying their younger children and
neighbors as undercounted. (207)
The main problem with using statistical sampling to adjust the census to correct for the undercount is that hypothetical
people are being counted while real people are being ignored or actually discounted, according to the Congressional Census
Monitoring Board:
It's important to first understand what the A.C.E. does and does not do. The A.C.E. does not determine how many real
people are missed by the census. The A.C.E. produces an estimate of the population of the United States for every level of
geography by comparing, or matching, the results of an independent survey to the results of the census. Because statistical
adjustment does not involve real people, it cannot eliminate the undercount or "correct" all of the error in the census. (208)
In Galvin v. Clinton, which was upheld by the U.S. Supreme Court, Judge Hilton agreed with the plaintiffs that using
statistical sampling to adjust apportionment resulted in violation of their voting rights, as he observed:
Individual citizens have standing to allege vote dilution resulting from allegedly unlawful legislative apportionment. See Baker v. Carr, 369 U.S. 186 (1962) (holding that plaintiffs, Tennessee voters, had a "plain, direct, and adequate" interest in maintaining the effectiveness of their votes and therefore had standing to maintain the action). Plaintiffs are individual taxpayers in Connecticut, Massachusetts, Minnesota, Missouri, Pennsylvania, and Wisconsin, all which are substantially likely to lose a seat in the House of Representatives solely because of the Department's plan.
Plaintiffs allege that the plan will dilute the voting strength of Plaintiffs at the intrastate level. Specifically, several
plaintiffs reside in counties whose relative population will be diminished by operation of the Department's plan. This
"elimination" of population constitutes vote dilution and a tangible injury resulting from the use of sampling. But for this
statistical "adjustment" Plaintiffs' counties would have a larger population. When the population of neighboring counties
is being increased by the addition of computer generated persons, this injury is compounded. Plaintiffs who reside in
counties which will have their population increased by less than the average of the other counties in their state will
necessarily suffer a loss in relative political representation. (209)
In other words, under the Census Bureau's proposal to use statistical sampling to try to correct the undercount, statistical artifacts are counted, not real people. Trying to correct inadequacies in counting, we venture into the presumption that there is something more real than what has actually happened. This is always problematic. We can be quite sure that there is an undercount-there are suggestions of it everywhere-but when it comes right down to it, how do we really know the extent? What if the theory that produces the abstract people is wrong, and they aren't real after all? What if there is an error in some of the assumptions? What if the assumptions were manipulated by political interest? Finally, how can you be sure none of these errors has happened? In the end, you just don't have the actual people there to prove the truth. You only have your faith that they exist, and in the numbers you think they do. You do not have actual proof, an actual body.
At any rate, the U.S. Supreme Court, in an opinion written by Sandra O'Conner, ruled that the Census Bureau cannot use sampling to apportion voting districts:
For the reasons stated, we conclude that the Census Act prohibits the proposed uses of statistical sampling in calculating
the population for purposes of apportionment. Because we so conclude, we find it unnecessary to reach the constitutional
question presented . . . Accordingly, we affirm the judgment of the District Court for the Eastern District of Virginia in
Clinton v. Glavin, No. 98--564. As this decision also resolves the substantive issues presented by Department of Commerce
v. United States House of Representatives, No. 98--404, that case no longer presents a substantial federal question. The
appeal in that case is therefore dismissed. (210)
Although the Supreme Court did not rule on the constitutionality of the Census Bureau's proposal, the intent of the Supreme Court is clear: The Constitution requires that actual people should be counted, not statistical abstractions, at least as far as voting apportionment goes. But why should it stop there? What is the difference between actual people being counted when their right to vote is at stake and actual people being counted when their right to a clean and healthful environment, if not their right to life, is at stake?
As we saw, a high count of people getting cancer is not taken seriously in small populations by health departments until it becomes statistically significant. The actual count of people getting cancer is ignored, even when it is several times what would be expected, until a sampling procedure (testing for statistical significance) shows that the number is large enough not to be random. In other words, the actual count is ignored in favor of a statistical sampling theory that the distribution may be random. Thispossibility that the count may be a random fluctuation is taken to be more real than the actual count, at least until the numbers become large enough to show statistical significance. So, as with the Census Bureau's proposal, a count of actual people in a cancer cluster is ignored in favor of a statistical sampling theory that purports to be more accurate than an actual count. Although Fallon County has one of the highest incidences of breast cancer in the world, this rate is apparently not considered "real" by DPHHS because the numbers are not large enough to establish statistical significance. The high rate might be random. But, in fact, the "randomness" of cancer distributions is largely an effect of the method used to investigate them, not of any scientific proof that cancer is, for the most part, randomly distributed.
In as far as doing an actual enumeration instead of executing a statistical abstraction would make any difference in government action, making the government more aggressive in protecting public health when it otherwise would do nothing, the Montana Constitution, following the precedent set by the U.S. Supreme Court, would surely require using an actual enumeration to identify an environmentally caused excess of cancer, not a statistical abstraction which would ignore the suffering of real people. When a count of cancer cases is higher than it should be, it is higher than it should be, and nothing should be allowed to dilute that fact. The government is deciding here the distribution of a fundamental right--determining how much cancer is too much-and it must base its decision on what is known, concretely counted, not on an statistical extrapolation, which may be altered and twisted into its opposite by merely changing assumptions. If a person's life is at risk from exposure to environmental pollution, and a high incidence of cancer has been counted in their community, the government cannot be allowed to error by dismissing it as a statistical anomaly when an actual enumeration of the incidence rate compared with what it would be in a clean and healthful environment shows that actual people are becoming sick and dying. Just as they are entitled to just one vote, people should be entitled to live their life to a natural end, unaffected by involuntary exposures to toxic pollution.
Tests of statistical significance for cancer clusters are unavoidably biased in favor of corporate profit and against
environmental protection and public health. By assuming that cancer is randomly distributed, then requiring mathematical
testing to make sure that the cluster has enough statistical power to overcome such an assumption, testing for statistical
significance too frequently places an insurmountable barrier in the way of pursuing environmental justice. It is wrong to do
this, it allows life to be taken when the state should be protecting it, and the court must not allow DPHHS to ignore such
things as a high rate of breast cancer in Fallon County because it is, as DPHHS seems to have concluded, not statistically significant.
In his letter to me, Governor Racicot conceded several times that environmental exposures in Fallon County could have caused health problems, even as he concluded that there wasn't an unusually high incidence of cancer in Fallon County. Referring to the uranium deposits found in Fallon he conceded that it is "possible that some levels in regional groundwater system could show high levels of radioactivity from soluble forms of uranium." He also conceded that "the suggestion that the health problems could be caused from the improper disposal of hazardous materials is possible." (211) Indeed, throughout his letter, the governor was very careful to hedge his language, conceding the possibility that environmental exposures had caused the cancer in Fallon County, while denying that there was a problem. The governor admits that harm may be happening, but he takes no precaution against it.
But he should have.
In fact, there is a problem: The incidence of childhood leukemia in Fallon County for the last 10 years is at least 10 times higher than the national average, the incidence rate for breast cancer is one of the highest in the world, the incidence rate for colorectal cancer is about 7 times the rate in neighboring counties, and the incidence of birth abnormalities is about 3 times higher than it is for most of Montana. Since cancer, by and large, is not an inevitable tragedy of the human condition but is largely caused by environmental exposures, the state has a constitutional obligation to prevent it. As we saw Article IX, Section 1, Clause 1 of the Montana Constitution says: "The state and each person shall maintain and improve a clean and healthful environment in Montana for present and future generations." Unless DPHHS aggressively investigates high cancer rates and other environmental diseases, warns the public about them, and, along with the Department of Environmental Quality and the EPA, removes the toxins causing them from the environment, this section of the Constitution is not being faithfully executed.
The plain language of the Constitution clearly sets a policy of environmental protection and public health precaution. No qualifying words are limiting the duty to maintain and improve a clean and healthful environment, such as "as far as is reasonable," "to the extent it is technologically possible," or "insofar as it is economically feasible." The duty toward public health and the environment is unqualified by any other duty or right, such as an public obligation--as many polluters might have desired-to undertake "risk" for their benefit. While possessing and protecting property is an inalienable right, the right of property ownership is absolute only to the extent that enjoyment of this right does not harm others or the environment. Nothing in the Constitution obligates anyone to risk their health, or the environment at large, so that someone else can profit from property ownership. I owe no obligation to Monsanto to risk cancer for its benefit.
Indeed, any "legal" requirement that we do would, itself, be a violation of our property rights. Surely the right to own property, if it means anything, means that we have exclusive control over what others may do to what we own. If we do not consent to a trespass--in this case a toxic trespass--they cannot legally do it. Pollution violates the autonomy of personal ownership, exposing our bodies (and surely we have a property in our own bodies) and our other property to toxins that will harm and devalue them, and so, pollution is inevitably and unavoidably a transgression of property rights. The right to a clean and healthful environment and the right to possess and protect property are not in conflict with each other, however much large corporations, seeking to maximize market externalities, may say they are, and there is no need to "balance" one right against the other. The right to a clean and healthful environment is but an explicit declaration of what is already implicit within the right to possess and protect property. The two rights share an underlying philosophy, a unity of purpose, which is to maximize the individual's freedom, the self's autonomy from infringement by others.
Since there is nothing in the Constitution obligating us to undertake any sort of health or environmental risk for the benefit of corporate profit, and since, on the other hand, we are obligated to maintain and improve a clean and healthful environment for present and future generations, DPHHS, as a public health department, is constitutionally bound to act only to maintain and improve a clean and healthful environment. When there is scientific uncertainty about whether this carcinogen or that one is causing the cancer, or whether any toxin is present in sufficient quantities to cause harm, or whether the numbers of people suffering harm are large enough to justify concluding that there is environmental harm, the state, as a precaution against violating our right to a clean and healthful environment, must act to protect public health and the environmental life support system. If mistakes are to be made, they are to be made on the side of protecting the environment and public health, not on the side of protecting corporate profits. People have rights, not industrial toxins, and if an industrial toxin is possibly causing harm to the environment, threatening life and body, any corporation profiting from its release in Montana has the same duty that the state and every individual in Montana does.
Harm done to corporate profit from getting the wrong toxin, or from acting on a cancer cluster when it happened by chance, is insignificant against the harm that comes from not acting when a toxic release is harming public health and the environment. Corporations can look for alternative ways of making a profit if they are prevented from releasing a given toxin into the environment, recovering their loss by manufacturing substitutes, but recovering their health is not nearly so easy for people exposed to a toxic release, and undoing the harm of a lost life is impossible. Preventing the loss of health or the loss of life has to be a greater priority than the loss of profit.
Since science has found that most cancer cases are caused by environmental exposures, an incidence of cancer that is higher than it would be in a clean and healthful environment is, all by itself, sufficient cause for DPHHS to take the precaution of informing the public that an environmental threat to their health may exist. There is no need for statistical significance testing. The evidence is already presumptively there in the high cancer rate. Whatever risk assessments may say about the safety of the toxins prevalent in the local environment, whatever the statistical significance of a high incidence of cancer, DPHHS must act on what it knows, the actual counts of disease. A risk assessment that says that an exposure to a toxin is safe must be regarded with suspicion and, as a precaution to ensure our right to a clean and healthful environment is not wrongly sacrificed, discounted when the actual counts of people exposed to a toxin are higher than they would be in a truly clean and healthful environment.
Tests of statistical significance are fundamentally the wrong tools to use when deciding whether a toxic exposure has placed the right to a clean and healthful environment in jeopardy because they take exactly the wrong precaution. Statistical significance testing protects industry and the state from having to assume responsibility for high rates of disease if it is possible that they might be merely random. With statistical significance testing, high rates of disease can be dismissed as merely random in small communities. Instead of giving the benefit of doubt to corporate interests, the precaution must go to public health, and the "randomness" of cancer discounted until proven. When a rate of cancer is high, it should, as a precaution, be presumed to be high, not dismissed as random.
This is especially important in Montana. Statistical significance tests depend on large numbers to draw conclusions, but most of Montana, except for Yellowstone County, has only small numbers. In the typical Montana community, incidence rates must be many times higher than expected to become "statistically significant." The practical effect of using tests of statistical significance to decide whether an environmental toxic exposure is causing a "real" rise in disease rates in small communities is to deny the people in these communities their right to a clean and healthful environment, allowing corporate polluters to do almost anything they want to them, as W.R. Grace did in Libby. (212)
Let me repeat this again one last time: Fallon County has an incidence of childhood leukemia that, at the minimum, is10 times higher than the national average (an average, which, itself, is a growing outrage). It has an incidence of breast cancer that is one of the highest in the world, and an incidence of colorectal cancer that is about 7 times higher than neighboring counties. It also has an incidence of birth abnormalities that is about 3 times higher than most of the rest of the state. Although the numbers of people afflicted are small, the actual number of cases of these diseases compared either with national and state averages or, more appropriately, what they would be in a truly clean and healthful environment, indicate that there is an environmental problem in Fallon County. Nevertheless, DPHHS has repeatedly issued an opinion that these rates are not unusual and justify no intervention.
I move the court order DPHHS to adequately warn the public in Fallon County of a high rate of childhood leukemia, breast cancer, colorectal cancer, birth abnormalities, and whatever other diseases are similarly high here, advise the public of ways science has suggested might reduce the risk of contracting them, investigate possible causes of these diseases, and intervene in whatever way appropriate to remove these causes from the environment so that they no longer pose a threat to anyone's right to a clean and healthful environment. Since the standard applied in Fallon County must be the same for the whole state, I also move that DPHHS be ordered to do the same for communities across the state that have similarly high rates of diseases that may be caused by environmental exposures. In particular, this order would include the communities that share a high breast cancer rate with Fallon county, as the cities of Billings, Great Falls, and Missoula do.
In general, whenever it is dealing with a disease cluster that may be linked to environmental exposures I move the court order DPHHS to:
Written By: Wade Sikorski, Ph.D.
Box 202
Willard, MT 59354
(406) 775-6378
1. Lowengart, R.A., Peters, J.M., Cicioni, C., et al., "Childhood leukemia and parents' occupational and home exposures," Journal of the National Cancer Institute Vol. 79 (1987), pp. 856-864.
See also: Arundel, S.E., Kinner-Wilson, L.M., "Parental occupations and cancer: a review of the literature," Journal of Epidemiology and Community Health Vol. 40 (1986), pp. 30-36.
2. Thompson, J.R., Fitz Gerald, P. Willoughby, M., Armstrong, B., "Maternal supplementation in pregnancy and protection against acute lymphoblastic leukemia in childhood: a case-control study,"The Lancet Vol. 358 (December 8, 2001), pp. 1935-1940.
3. Governor Marc Racicot, letter to Wade Sikorski (August 7, 2000).
4. Wade Sikorski, letter to Governor Marc Racicot (August 13, 2000).
5. Todd Damrow, Montana State Epidemiologist, Department of Public Health and Human Services, letter to Wade Sikorski (December 19, 2000), p. 3.
6. Clair Johnson, "Cancer Rates Not Higher in Fallon County, State Says," Billings Gazette, January 4, 2000, p. 1.
7. See the following articles in the Seattle-Post Intelligencer by Andrew Schneider: "While people are dying, government agencies pass buck," (Friday, November 19, 1999), "Deadly ore was shipped around U.S., Canada, (Wednesday, December 22, 1999), "A town left to die," (Thursday, November 18, 1999). All articles are available on the Internet: http://seattlep-i.nwsource.com/uncivilaction/
8. Gail Gray, Director of DPHHS, letter to Wade Sikorski (March 2, 2001).
9. Massachusetts Department of Public Health, Bureau of Environmental Health Assessment,Woburn Childhood Leukemia Follow-Up Study Vol. I, Analyses (July 1997).
10. F.E. Alexander and P. Boyle, Methods for Investigating Localized Clustering of Disease(World Health Organization: International Agency for Research on Cancer, 1996), p. 15.
11. V. Beral, E. Roman, M. Bobrow, "Childhood Cancer and Nuclear Installations," British Medical Journal (1993).
12. "Low Level Radiation and Health: How Dangerous is it Really?"
http://www.llrc.org/health.htm.
13. A.G. Noshchenko, K.B. Moysich, A. Bondar, P.V. Zamostyan, V.D. Drosdova, A.M. Michalek, "Patterns of acute leukemia occurrence among children in the Chernobyl region,"International Journal of Epidemiology Vol. 30, No. 1 (2001), pp. 125-129.
14. Alfred Korblein and Wolfgang Hoffman, "Childhood Cancer in the Vicinity of German Nuclear Power Plants," Medicine and Global Survival Vol.6 (1999), pp.18-23.
15. Institute of Medicine, National Academy of Sciences, Veterans and Agent Orange: Update 2000, http://www4.nationalacademies.org.../7a99ce2d3b3cc1c985256a330069e92e?OpenDocumen
16. Buckley J.D., Robison L.L., Swotinsky R., Garabrant D.H., Lebeau M., Manchester P., Nesbit M.E., Odom L., Beters J.M., Woods W.G., and et al., "Occupational exposures of parents of children with acute nonlymphocytic leukemia: a report from the Children's Cancer Study Group,"Cancer Research Vol. 49 (1989), pp. 4030-4037.
17. Lowengart, R.A., Peters J.M.., Ciconi, C., Buckley, J., Bernstein, L., Preston-Martin, S., Rappaport, E., "Childhood leukemia and parents' occupational and home exposures," Journal of the National Cancer Institute Vol. 79 (1987), pp. 39-46.
18. McKinney, P.A., Alexander, F.E., Cartwright, R.A., Parker L., "Parental occupations of children with leukemia in west Cumbria, north Humberside, and Gateshead," British Medical JournalVol. 302 (1991), pp. 681-687.
19. Linet, M.S., Cartwright, R.A., "The leukemias," Cancer Epidemiology and Prevention(Schottenfeld, D., Fraumeni, J.F. Jr. eds), (New York: Oxford University Press, 1996), pp. 841-892.
20. Colt, J.S., and Blair, A., "Parental Occupational Exposures and Risk of Childhood Cancer,"Environmental Health Perspectives Vol. 106, Supplement 3 (June 1998).
21. K. Hemminki, S. Saloniemi, T. Salonene, T. Partanen, H. Vinio, "Childhood cancer and parental occupation in Finland,"
Journal of Epidemiology and Community Health Vol. 35 (1981), pp. 11-15.
See also: C. Magnani, G. Pastore, L. Luzzato, B. Terracini, "Parental occupation and other environmental factors in the etiology of leukemias and non-Hodgkin's lymphomas in childhood: a case-control study," Tumori Vol. 76 (1990), pp. 413-419.
22. J. Fabia, T.D. Thuy, "Occupation of father at time of birth of children dying of malignant diseases," British Journal of
Preventive Social Medicine Vol. 28 (1974), pp. 98-100.
See also: J.D. Buckley, L.L. Robison, et al., Occupational exposures of Children with acute nonlymphocytic leukemia: a
report from the Children's Cancer Study Group," Cancer Research Vol. 49 (1989), pp. 39-46.
See also: N.J. Vianna, B Kovasznay, A. Polan, C. Ju, "Infant leukemia and paternal exposure to motor vehicle exhaust
fumes," Journal of Occupational Medicine Vol. 26 (1984), pp. 679-681.
23. Hemminki K, Saloniemi S, Salonen T, Vainio H, "Childhood cancer and parental occupation in Finland," Journal of Epidemiology and Community Health Vol. 35 (1981), pp. 11-15.
24. Joanne S. Colt and Aaron Blair, "Parental Occupational Exposures and Risk of Childhood Cancer," Environmental
Health Perspectives Vol. 106, Supplement 3 (June 1998).
See also: Rolf Norlinder and Bengt Jarvholm, "Environmental exposure to gasoline and leukemia in children and young adults-an ecology study," International Archives of Environmental Health Vol. 70 (1997), pp. 57-60.
25. Eric Rydbom, compliance specialist, Fuel Tax Management & Analysis Unit, Montana Department of Transportation,
"Investigative Report, Subject: Ross Management," (January 15, 1998).
See also, Memo to SEMA officers, Aaron Browning, NPRC organizer (January 22, 1998).
26. Wade Sikorski, Sacrificial Rituals: A Murder Mystery in Montana, (2000), http://www.midrivers.com/~wds.
27. Ruth A. Lowengart, John M. Peters, Carla Cicioni, and et al., "Childhood leukemia and parents' occupational and home exposures," Journal of the National Cancer Institute Vol. 79 (1987), pp. 39-46.
28. Sandra Steingraber, Living Downstream: An Ecologist Looks at Cancer and the Environment (Reading, MA: Addison-Wesley, 1997), p. 65.
29. U.S. Census Bureau, 1990 Census,
http://factfinder.census.gov/servlet/Basi...me=DEC_1990_DP1&_geo_id=05000US30025
30. U.S. Census Bureau, http://quickfacts.census.gov/cgi-bin/county?cnty=30025
31. U.S. Census Bureau,
http://factfinder.census.gov/servlet/Basi...e=DEC_1990_STF1_DP1&_geo__id=160000US300020
32. National Cancer Institute, Cancer: Rates and Risks, 4th Edition 1996 (NIH Publication No. 96-691, May 1996), p. 31.
See also, Malcolm A. Smith, Lynn A. Gloeckler Ries, James G. Gurney, Julie A. Ross, "Leukemia,"Cancer Incidence and
Survival Among Children and Adolescents: United States SEER Program 1975-1995 (National Cancer Institute, Pub. No.
99-4649)
http://seer.cancer.gov/Publications/PedMon/leukemia
33. S.A. Kaye, L.L. Robinson, W.A. Smithson, P. Gunderson, F.L. King, J.P. Neglia, "Maternal Reproductive History and Birth Characteristics in Childhood Acute Lymphoblastic Leukemia,"Cancer, Vol. 68 (1991), pp. 1351-5.
34. National Institute of Health, "Childhood Cancer: A Growing Problem," Environmental Health Perspectives Vol. 106, No. 1 (January 1998).
35. Dona Schneider and Natalie Freeman, "Children's Environment Health Risks: A State-of-the-Art Conference," Archives of Environmental Health Vol. 56, No. 2 (March/April 2001), pp. 103-110.
36. California Department of Health Services, California Cancer Registry (11/99), "Leukemia: Acute Lymphocytic," http://www.ccrcal.org/Cancer--00/index
37. American Cancer Society, Cancer Facts and Figures 2000 (Atlanta, Georgia: American Cancer Society, 2000), p. 20.
38. B. Bert Gerstman, Epidemiology Kept Simple: An Introduction to Classic and Modern Epidemiology (New York: John Wiley and Sons, 1998), p. 229.
39. B. Bert Gerstman, Epidemiology Kept Simple: An Introduction to Classic and Modern Epidemiology (New York: John Wiley and Sons, 1998), p. 233.
40. Glyn G. Caldwell, "Twenty-two Years of Cancer Cluster Investigations at the Centers for Disease Control," American Journal of Epidemiology Vol. 132, Supplement No. 1 (1990), pp. S43-S47.
41. F.E. Alexander and P. Boyle, Methods for Investigating Localized Clustering of Disease(Geneva: International Agency for Research on Cancer, 1996), p 4.
42. Massachusetts Department of Public Health Bureau of Environmental Health Assessment,Woburn Childhood Leukemia Follow-up Study Vol. 1, Analysis (July 1997).
43. Michael Weissenstein, "Children's Illness Confounds, Frustrates Small Town," Las Vegas Review-Journal, Sunday, February 4, 2001, p 25A.
44. Billings Gazette, "New Childhood Leukemia Case Surfaces in Fallon,"
http://www.billingsgazette.com/archive.php?section=health&display=rednews/2001/.../fallon.in
45. For a discussion of the discursive politics of nuclear pollution, see: Valerie L. Kuletz, The Tainted Desert: Environmental and Social Ruin in the American West (New York: Routledge, 1998).
46. - -
47. 47Patrick Springer, "Forgotten Fallout: What is the Legacy of the Radioactive Rains?" The Fargo Forum (May 1,1988), p. 7.
48. 48Forgotten Fallout," p. 7.
49. Victor Archer, "Association of Nuclear Fallout with Leukemia in the United States,"Archives of Environmental Health Vol. 42, No. 5 (September/October 1987), pp. 263-271.
50. 50Forgotten Fallout," p. 8.
See also:Carl Johnson, "Cancer incidence in an area of radioactive fallout downwind from the Nevada test site," Journal of
the American Medical Association Vol. 251 (1984), pp. 230-236.
Carl Johnson, "Leukemia death rates of residents in areas contaminated by plutonium," Abstract, American Public Health Association, Washington, D.C. (November 1, 1977).
51. J.L. Lyon, M.R. Klauber, J.W. Gardner, and K.S. Udall, "Childhood luekemias associated with fallout from nuclear
testing," New England Journal of Medicine Vol. 300 (1979), pp. 394-402.
See also: M. J. Gardner, and P.D. Winter, "Mortality in Cumberland during 1959-1978 with reference to cancer in young
people around Windscale," Lancet Vol. 1 (1984), pp. 216-217.
See also: H.L. Beck, and P. W. Krey, "Radiation exposure in Utah from Nevada nuclear tests,"Science Vol. 223 (1984), pp. 139-144.
52. 52National Cancer Institute, Atlas of Cancer Mortality in the United States: 1950-94 (NIH Publication No. 99-4564, September 1999), p. 4.
53. One way of finding out would be to test for Strontium-90 in teeth. See J.M. Gould, E.J. Sternglass, J.D. Sherman, J. Brown, W. McDonnell, J.J. Mangano, "Strontium-90 in deciduous teeth as a factor in early childhood cancer," International Journal of Health Services Vol. 30, No. 3 (2000), pp. 515-539.
54. Janette D. Sherman, M.D., Life's Delicate Balance: Causes and Prevention of Breast Cancer (New York: Taylor and Francis, 2000), p. 67.
55. W. Hoffman et al, "Radium-226-Contaminated Drinking Water: Hypothesis on an Exposure Pathway in a Population with Elevated Childhood Leukemias,"Environmental Health PerspectivesVol. 101, Supplement 3, (1993), pp. 113-115.
56. Valerie Beral, "Childhood Leukemia near Nuclear Power Plants in the United Kingdom: The Evolution of a Systematic Approach to Studying Rare Disease in Small Geographic Areas," American Journal of Epidemiology Vol. 132, Supplement No. 1 (1990), pp. S63-S68.
57. Montana Tumor Registry, Annual Report (October 2000) p.18.
58. U.S. Health Resources and Services Administration,
http://www.communityhealth.hrsa.gov/countyInfo.asp
59. National Cancer Institute, Cancer: Rates and Risks, 4th Edition, 1996, National Institutes of Health (NIH Publication No. 96-691), p. 43.
60. National Cancer Institute, Cancer: Rates and Risks, 4th Edition, 1996, National Institutes of Health (NIH Publication No.
96-691), p. 120.
The actual report identifying San Francisco with the highest breast cancer incidence in the world is: Parkin D.M., Muir C.S.,
Whelan S., and et al., eds.: Cancer Incidence in Five Continents Vol. VI, IARC Scientific Publication No. 120 (Lyon: World
Health Organization, International Agency for Research on Cancer, 1992).
61. Center for Disease Control and Prevention, Report to Congress: CDC Review of the Northern California Cancer Center Report: Status of Breast Cancer Research in the San Francisco Bay Area, Executive Summary, May 1998, p. 5.
62. Eric J. Feuer, Lap-Ming Wun, Catherine C. Boring, W. Dana Flanders, Marilytl J. Timmel, Tolly Tong, "The Lifetime Risk of Developing Breast Cancer," SEER Cancer Statistics Review 1973-1997 (National Cancer Institute, Pub. No. 00-2789).
http://seer.cancer.gov/Publications/CSR1973_1997/breast
63. Liane Clorfene-Casten, Breast Cancer: Poisons, Profits, and Prevention (Monroe, ME: Common Courage Press, 1996), p. 2.
64. National Cancer Institute, Cancer Rates and Risks, 4th Edition, 1996 (NIH Publication No. 69-691, May 1996), p. 120.
65. J.L. Kelsey and M.D. Gammon, "Epidemiology of Breast Cancer," Epidemiology Review,Vol. 12 (1993), pp. 228-240.
66. Eva Cecilie Bonefeld-Jorgensen, Helle Raun Anderson, Thomas Hoj Rasmussen, Anne Marie Vinggaard, "Effect of highly bioaccumulate polychlorinated biphenyl congeners on estrogen and androgen receptor activity," Toxicology Vol. 158 (2001), pp. 141-153.
67. D.L. Davis, H.L. Bradlow, and M. Wolff, et al., "Medical Hypothesis: Xenoestrogens as Preventable Causes of Cancer,"
Environmental Health Perspectives Vol. 101 (1993), pp 372-377.
See also, T. Key, G. Reeves, "Organochlorines in the Environment and Breast Cancer," British Medical Journal Vol. 308
(1994), pp. 1520-1.
See also, Vessela Nedelcheva Kristense, Anne Lise Borresen-Dale, "Molecular epidemiology of breast cancer: genetic
variation in steroid hormone metabolism," Mutation Research Vol. 462 (2000), pp. 323-333.
68. F. Falck, A. Ricci, M. Wolff, J. Godbold, and P. Deckers, "Pesticides and Polychlorinated Biphenyl Residues in Human Breast Lipids and Their Relation to Breast Cancer," Archives of Environmental Health Vol. 47 (1992), pp. 143-146.
69. Annette Pernille Hoyer, Torben Jorgensen, John Brock, Philippe Grandjean, "Organochlorine exposure and breast cancer survival," Journal of Clinical Epidemiology Vol. 53 (2000), pp. 323-330.
70. D.J. Hunter and W.C. Willett, "Diet, Body Size, and Breast Cancer," Epidemiology ReviewVol. 15 (1993), pp. 110-132.
71. Nadine M. Brown and others, "Prenatal TCDD and Predisposition to Mammary Cancer in Rats," Carcinogenesis Vol. 19, No. 9 (1998), pp. 1623-1629.
72. E.M. John, J.L Kelsey, "Radiation and other Environmental Exposures and Breast Cancer,"Epidemiology Review Vol. 15 (1993), pp. 157-62.
73. C. J. Johnson, "Cancer incidence in an area of radioactive fallout downwind from the Nevada test site," Journal of the American Medical Association Vol. 251, No. 2 (1984), pp. 230-236
74. R. Bertell, "Handbook for Estimating Health Effects from Exposure to Ionizing Radiation," (Toronto, Ontario: Institute of Concern for Public Health, 1984).
75. A.C. Upton, "Biological basis for assessing carcinogenic risks of low-level radiation,"Carcinogenisis Vol. 10 (1985), pp. 381-401.
76. T. Wakabyashi, H. Kato, T Ikeda, W. Schull, "Incidence of cancer in 1959-1978, Based on Tumor Registry," Radiation
Research Vol. 93 (1983), pp. 112-146.
See also: M. Tohounaga, C.E. Land, T Yamamotor, M. Asano, S. Tokuoka, H. Ezaki, I Nishimori, "Breast cancer in
Japanese A-Bomb Survivors," Lancet (October 23, 1982).
77. Janette Sherman, Life's Delicate Balance: Causes and Prevention of Breast Cancer (New York: Taylor and Francis, 2000), p. 65.
78. Y. Miki, and 44 other authors, "A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1," Science Vol. 262 (1994), pp. 66-71.
79. Terry Tempest Williams, Refuge: An Unnatural History of Family and Place (New York: Vantage Books, 1991).
80. U.S. 86th Congress Joint Committee on Atomic Energy, Summary-Annalysis of Hearings: Fallout from Nuclear Weapons Tests, May 5-8, 1958 (Washington: U.S. Printing Office, 1959), pp. 28.
81. Janette Sherman, Life's Delicate Balance: Causes and Prevention of Breast Cancer (New York: Taylor and Francis, 2000), p. 167.
82. Tumor Registry, Annual Report (October 2000), p. 22.
83. National Cancer Institute, Cancer: Rates and Risks, 4th Edition, 1996, NIH Publication No. 96-691 (May 1996), p. 47.
84. A.J. McMichael and G.G. Giles, "Cancer in Migrants to Australia: Extending the Descriptive Epidemiological Data," Cancer Research Vol. 48 (1988), pp. 751-6.
85. B.S. Reddy, "Dietary Fat and Colon Cancer," Experimental Colon Carcinogenesis, H. Autrup and G.M. Williams, editors (Boca Raton, Florida: CRC Press, 1983), pp. 225-239.
86. W.C. Willett, M.J. Stampfer, G.A. Colditz, et al, "Relation of Meat, Fat, and Fiber Intake to the Risk of Colon Cancer in a Prospective Study Among Women," New England Journal of Medicine Vol. 323 (1990), pp. 1664-7.
87. Samuel Epstein, The Politics of Cancer Revisited (Freemont Center, New York: East Ridge Press, 1998), p. 298-9.
88. Janette Sherman, Life's Delicate Balance: Causes and Prevention of Breast Cancer(New York: Taylor and Francis, 2000), pp. 16-17.
89. Aaron Blair, Shelia Hoar Zahm, Neil E. Pearce, Ellen F. Heineman, Joseph F. Fraumeni Jr, "Clues to Cancer Etiology from Studies of Farmers," Scandinavian Journal on Work and Environmental Health Vol. 18, No. 4 (1992), pp.209-215.
90. J. Siemiatycki, L. Richardson, M. Gerin, et al, "Associations between several sites of cancer and nine organic dusts: results from a hypothesis-generating case-control study in Montreal, 1979-1983," American Journal of Epidemiology Vol.123 (1986), pp. 235-249.
91. M.C.R Alavanja, A. Blair, et al., "Mortality among agricultural extension agents," American Journal of Industrial Medicine Vol. 14 (1988), pp. 167-176.
92. M.C.R Alavanja, A. Blair, et al., "Mortality among forest and soil conservationists,"Archives of Environmental Health Vol. 44 (1989), pp. 94-101.
93. Francesco Forastiere, Augusto Quercia, Maria Miceli, et al, "Cancer Among Farmers in Central Italy," Scandinavian Journal of Work and Environmental Health Vol. 19 (1993) pp. 382-389.
94. Aaron Blair, Shelia Hoar Zahm, Neil Pearce, Ellen F. Heineman, Joseph F. Fraumeni Jr., "Clues to cancer etiology from studies of farmers," Scandinavian Journal of Work and Environmental Health Vol. 18 (1992), pp. 209-215.
95. Dina M. Schreinemachers, "Cancer Mortality in Four Northern Wheat-Producing States,"Environmental Health Perspectives Vol. 108, No. 9 (September 2000), pp. 873-881.
96. Pete Feigley, Debbi Lemons, and Fred Ramsey, Cancer in Montana, 1993-1997: An Annual Report of the Montana Tumor Registry, Montana Central Tumor Registry, Department of Public Health Services (October 2000).
http://www.dphhs.state.mt.us/divisions/otd/vital/statistical_tables.htm
97. David Espeland, "Fallon Medical Complex Responds to Cancer Scare, Fallon County Times, (December 1, 2000), p. 1.
98. Montana Tumor Registry, Annual Report, p. 28.
99. National Cancer Institute, Cancer: Rates and Risks, 4th Edition, 1996, National Institutes of Health (NIH Publication No. 96-691), p. 69.
100. Montana Tumor Registry, Annual Report, pp. 14, 24, 34, and 38.
101. Department of Public Health and Human Services, http://www.dphhs.state.mt.us/divisions.
102. Aaron Derfel, "Sperm on the wane: scientists express alarm," The Montreal Gazette(Friday, June 22, 2001). http://www.montrealgazette.com/news/pages/010622/5081326.html
103. Theo Colborn, Dianne Dumanoski, and John Peterson Myers, Our Stolen Future (New York: Dutton, 1996), p.203-205.
104. Fred vom Saal, S. Nagel, P. Palanza, M. Boechler, S. Parmigiani, W. Welshons, "Estrogenic Pesticides: Binding Relative to Estradiol in MCF-7 Cells and Effects of Exposure During Fetal Life on Subsequent Territorial Behavior in Mice," Toxicology Letter (in press, 1995).
105. TEQ stands for toxic equivalents. There are 75 different dioxin congeners, 135 different furan congeners, and 209 different PCB congeners, all of which have different toxicities. Some dioxin congeners are much more toxic than others with 2,3,7,8-TCDD being the worst. When dioxin is found in the environment, it is typically a mixture not only of different congeners of dioxins, but also different congeners of furans, PCBs, and other environmental hormones which have a similar effect as dioxin. To simplify things, scientists take each congener, multiply it by a toxic equivalency factor (TEF), and convert everything into one toxic unit, the TEQ number. Through this conversion, a whole mixture of different congeners are reduced to the equivalent toxic effect that the worst congener of dioxin would have. Thus, 100 pounds of a couple of relatively nontoxic congeners of dioxin and furans becomes a pound of dioxin TEQ. Of course this all is very confusing, but it is less confusing than trying to talk about all the different toxic effects of different congeners. Generally, when scientists talk about amounts of dioxin, they are not talking about the actual quantities of the different congeners added together, but the equivalent toxic effect that 2,3,7,8-TCDD would have.
106. EPA, Estimating Exposure to Dioxin-Like Compounds, Volume I: Executive Summary(Washington D.C.: Office of Health And Assessment, June 1994), EPA/600/6-88/005Ca, p. 37.
107. Lois Marie Gibbs, Dying from Dioxin (Boston, MA: South End Press, 1995), pp. 74-5.
108. EPA, Health Assessment Document for 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds: Volume III of III (Washington, D.C.: Office of Health and Environmental Assessment, EPA, August 1994), EPA/600/BP-92/001c, p. 9-86.
109. "Executive Summary: Assessment of the Health Risk of Dioxins: Re-evaluation of the Tolerable Daily Intake (TDI); WHO Consultation, May 25-29, 1998, Geneva, Switzerland," World Health Organization, WHO European Centre for Environment and Health, International Programme on Chemical Safety (December 8,1998).
110. Heiko Becher, Karen Steindorf, and Dieter Flesch-Janys, "Quantitative Cancer Risk Assessment for Dioxins Using an Occupational Cohort," Environmental Health Perspectives Vol. 106, Supplement 2 (April 1998), pp. 663-670.
111. EPA, Estimating Exposure to Dioxin-Like Compounds, Volume I: Executive Summary(Washington D.C.: Office of Health and Assessment, June 1994), EPA/600/6-88/005Ca.
113. See also, Peter Montague, "Dioxins--The View from Europe," Rachel's Environment and Health Weekly #636 (February 4, 1998).
114. Theo Colborn, Dianne Dumanoski, and John Peterson Myers, Our Stolen Future (New York: Dutton, 1996), pp. 26-28.
115. Theo Colborn, Frederick S. vom Saal, and Ana M. Soto, "Developmental Effects of Endocrine-Disrupting Chemicals in Wildlife and Humans," Environmental Health Perspectives Vol. 101, No. 5 (October 1993), pp. 378-384.
116. Joseph Jacobson and Sandra Jacobson, "Intellectual Impairment in Children Exposed to Polychlorinated Biphenyls in
Utero," New England Journal of Medicine Vol. 335 No. 11(September 12, 1996), pp. 783-789.
See also Joseph Jacobson and Sandra Jacobson, "Dose-response in Perinatal Exposure to Polychlorinated Biphenyls
(PCBs): The Michigan and North Carolina Cohort Studies," Toxicology and Industrial Health Vol. 12, Nos.3/4 (1996),
pp.435-445.
117. Marguerite Holloway, "Dioxin Indictment," Scientific American Vol. 270 (January 1994), p. 25.
118. Elisabeth Carlsen and others, "Evidence for the Decreasing Quality of Sperm During the Past 50 Years," British Medical Journal Vol. 305 (1992), pp. 609-613.
119. T. Mably and others, "In Utero and Lactational Exposure of Male Rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin Exposure," Journal of Toxicology and Applied Pharmacology Vol. 114 (1991), pp. 118-126.
120. G.M. Egeland and others, "Total Serum Testosterone and Gonadotropin in Workers Exposed to Dioxin," American Journal of Epidemiology Vol. 139 (1994), pp. 272-281.
121. Stephanie Coontz, The Way We Really Are: Coming to Terms with America's Changing Families (New York: Basic Books, 1997), p 13.
122. Marcia E. Herman-Giddens and others, "Secondary Sexual Characteristics and Menses in Young Girls Seen in Office Practice," Pediatrics Vol. 99, No. 4 (April 1997), pp. 505-512.
123. Nell Boyce, "Growing up too Soon," New Scientist (August 2, 1997), p. 5.
See also: Ronald J. Gellert, "Uterotrophic Activity of Polychlorinated Biphenyls (PCB) and Induction of Precocious
Reproductive Aging in Neonatally Treated Rats," Environmental Research Vol. 16 (1978), pp. 123-130.
125. Elizabeth A. Guillette and others, "An Anthropological Approach to the Evaluation of Preschool Children Exposed to Pesticides in Mexico," Environmental Health Perspectives Vol. 106, No. 6 (June 1998), pp. 347-353.
126. Fred vom Saal, S. Nagel, P. Palanza, M. Boechler, S. Parmigiani, W. Welshons, "Estrogenic Pesticides: Binding Relative to Estradiol in MCF-7 Cells and Effects of Exposure During Fetal Life on Subsequent Territorial Behavior in Mice," Toxicology Letter (in press, 1995).
127. W. Porter, S. Green, N. Debbink, and I. Carlson, "Groundwater Pesticides: Interactive Effects of Low Concentrations of Carbamates, Aldicarb, and Methomyl, and the Triazine Metribuzin on Thyroxine and Somatotropin Levels in White Rats," Journal of Toxicology and Environmental Health Vol. 40 (1993), pp. 15-34.
128. Theo Colborn, Dianne Dumanoski, and John Peterson Myers, Our Stolen Future (New York: Dutton, 1996), pp. 237-8.
129. For a discussion of the discursive politics of risk assessment see: Sylvia Noble Tesh,Uncertain Hazards: Environmental Activists and Scientific Proof (Ithaca: Cornell University Press, 2000).
130. For a discussion of how politics underlays scientific discourse in the social sciences, see: William Connolly, Appearance and Reality in Politics (Cambridge: Cambridge University Press, 1981)
131. Clair Johnson, "Cancer Rates Not Higher in Fallon County, State Says," Billings Gazette, Thursday, January 4, 2000, p. 1.
132. Kenneth Rothman, "Keynote Presentation: A Sobering Start for the Cluster Busters' Conference," American Journal of Epidemiology Vol. 132 (July 1990) No. 1, p. 1.
133. Centers for Disease Control, "Guidelines for Investigating Clusters of Health Events," MMWR 39(RR-11), (July 27, 1990), pp. 1-16.
134. Raymond Richard Neutra, "Reviews and Commentary: Counterpoint from a Cluster Buster," American Journal of
Epidemiology Vol. 132 (July 1990) No. 1, p. 1.
Richard A. Goodman, James W. Buehler, and Jeffrey P. Koplan, "The Epidmiologic Field Investigation: Science and
Judgement in Public Health Practice," American Journal of EpidemiologyVol. 132 (July 1990) No. 1, p. 9.
Beth J. Fiore, Lawrence P. Hanrahan, and Henry A. Anderson, "State Health Department Response to Disease Cluster
Reports: A Protocol for Investigation," American Journal of Epidemiology Vol. 132 (July 1990) No. 1, p.14.
135. Centers for Disease Control, "Guidelines for Investigating Clusters of Health Events," MMWR 39 (RR-11), (July 27, 1990), pp. 1-16.
136. Steingraber, Living Downstream, p. 76.
137. Sandra Steingraber, Living Downstream: An Ecologist Looks at Cancer and The Environment (New York: Addison Wesley, 1997), p. 75.
138. William Buchanan, Understanding Political Variables: Second Edition (New York: Charles Scribner's Sons, 1974), p. 92.
139. Understanding Political Variables, p. 92.
140. E.J. Feuer, "Lifetime Probability of Cancer," Journal of the National Cancer Institute
Vol. 89, (1997), p. 279.
141. Steingraber, Living Downstream, p. 59.
142. Sandra Steingraber, Living Downstream: An Ecologist Looks at Cancer and the Environment (Reading, MA: Addison-Wesley Publishing Company, 1997), p 251.
143. Paul Lichtenstein, et al, "Environmental and Heritable Factors in The Causation of Cancer: Analyses of Cohorts of Twins from Sweden, Denmark, and Finland, The New England Journal of Medicine Vol. 343, No. 2 (July 13, 2000), p. 84.
144. "Environmental and Heritable Factors," p. 82.
145. W. Burke, M. Daly, J. Garber, et al., "Recommendations for follow-up care of individuals with an inherited predisposition to cancer," Journal of the American Medical Association Vol. 277, No. 12 (1997), pp.997-1003.
146. E.B Claus, N. Risch, W.D. Thompson, "Autosomal dominant inheritance of early-onset breast cancer: implications for
risk prediction," Cancer Vol. 73, No. 3, (1994), pp. 643-651.
See also D.A. Berry, G. Parmigiani, J. Sanchez, et al., "Probability of carrying a mutation of breast-ovarian cancer gene
BRCA1 based on family history," Journal of the National Cancer InstituteVol.89, No. 3 (1997), pp. 227-238.
See also: D. Ford, D.F. Easton, M. Stratton, et al. Genetic hetrogeneity and penetrance analysis of the BRCA1 and BRCA2 genes in breast cancer families," American Journal of Human GeneticsVol.62, No. 3 (1998), pp. 227-238.
147. A.S. Whittemore, G. Gong, J. Itnyre, "Prevalence and contribution of BRCA1 mutations in breast cancer and ovarian cancer: Results from three U.S.population-based case-control studies of ovarian cancer," American Journal of Human Genetics Vol. 60, No. 3 (1997), pp. 496-504.
148. Mary Alma Welch, "What Genetic Testing Can Tell You About Breast Cancer Risk,"Journal of the American Academy of Physician Assistants Vol.14, No. 3 (March 2001), pp. 50-61.
149. L. Tomatis, Cancer: Causes, Occurrence and Control, IARC Scientific Pub. 100 (Lyon, France: International Agency for Research on Cancer, 1990), p. 21.
150. Wade Sikorski, Infected with Difference: Healing Dis/ease in the Body Politic,unpublished, 1999, http://www.midrivers.com/~wds.
151. Steingraber, Living Downstream, p.78.
152. C.E. Land and J.E. Norman, "Latent Periods of Radiogenic Cancers Occurring Among Japanese A-Bomb Survivors,"
Late Biological Effects of Ionizing Radiation, Vol. I, (Vienna: IAEA, 1978) Report No. IAEA-sm-224/602, pp. 29-47.
P.G. Smith and R. Doll, "Age and Time-Dependent Changes in the Rates of Radiation-Induced Cancers in Patients with
Ankylosing Spondylitis Following a Single Course of X-Ray Treatment, "Late Biological Effects of Ionizing Radiation, Vol.
I, (Vienna: IAEA, 1978) Report No. IAEA-sm-224/602, pp. 205-18.
153. Daniel Sinnett, Maja Krankinovic, and Damian Labuda, "Genetic Susceptibility to Childhood Acute Lymphoblastic Leukemia, Leukemia and Lymphoma Vol. 38, (2000), pp. 447-462.
154. Quoted from: Donald N. McCloskey, The Rhetoric of Economics (Madison: The University of Wisconsin Press, 1985), p.167.
155. For a discussion of how technology enframes the world, see: Martin Heidegger, "The Question Concerning Technology," in The Question Concerning Technology, (New York: Harper and Row, 1977).
156. For a discussion of how "reality" shifts, see: Michel Foucault, The Order of Things (New York: Vintage Books, 1973).
For a discussion of how "population" became a target of power see also: Michel Foucault, The History of Sexuality, Volume
I (New York: Random House, 1973).
157. For a discussion of how mathematics views the world, see: Martin Heidegger, "Science and Reflection," and "The Age of the World Picture," The Question Concerning Technology (New York: Harper and Row, 1977).
158. For a discussion of how scientific theory mediates the existence of data, see: Robert Ackermann, Data, Instruments, and Theory: A Dialectical Approach to Understanding Science(Princeton: Princeton University Press, 1985).
159. For a discussion of the pitfalls of scientific methodism, see: Paul Feyerabend, Against Method (London: Redwood Burn Limited Trowbridge and Esher, 1975).
160. Donald N. McCloskey, The Rhetoric of Economics, (Madison, Wisconsin: The University of Wisconsin Press, 1985), p. 159.
161. McCloskey, The Rhetoric of Economics, p. 162.
162. Letter to me from Robert C. Williams, P.E., DEE, Assistant Surgeon General, Director, Division of Health Assessment and Consultation, ATSDR, Dec. 18, 2000.
163. Letter to me from Governor Racicot, August 7, 2000.
164. Todd Damrow, Montana State Epidemiologist, Department of Public Health and Human Services, letter to Wade Sikorski (December 19, 2000), p. 3.
165. City of Baker advertisement, Fallon County Times, (June 1, 2001), p. 8.
166. Regrettably, I lost the address to this page on the internet.
167. City of Baker, "Annual Drinking Water Quality Report," Fallon County Times (June 1, 2001), p. 8.
168. Mary O'Brien, Making Better Environmental Decisions: An Alternative to Risk Assessment(Cambridge, MA: The Massachusetts of Technology Press, 2000), p 10.
169. Peter Montague, "Risk Assessment-Part 2: No Person Shall be Deprived of Life . . . "Rachel's Environment and Health News No. 195 (Environmental Research Foundation, August 22, 1990).
170. Steingraber, Living Downstream, p. 269.
171. Paul Merrell and Carol Van Strum, "Negligible Risk or Premeditated Murder?" Journal of Pesticide Reform Vol. 10 (Spring 1990), pp. 20-22.
172. Lots of other examples could be used. For instance, the people that died of the cross contamination of anthrax from the letter sent to Senate Majority Leader Tom Daschle. I use the example of the Unabomber, who selected his victims randomly, was tried for murder, but in contrast the executives of W.R. Grace were not tried for what they did in Libby in: Wade Sikorski, "On Mad Bombers: A Review of the Unabomber's Manifesto," Theory & Event, February 1997.
http://muse.jhu.edu/journals/theory_&_event/v001/1r.sikorski.html
173. Peter Montague, "Risk Assessment-Part 2: No Person Shall be Deprived of Life . . . "Rachel's Environment and Health News No. 195 (Environmental Research Foundation, August 22, 1990).
174. Sandra Steingraber, Living Downstream, p. 40.
175. James T. Patterson, The Dread Disease: Cancer and Modern American Culture(Cambridge, MA, Harvard University Press, 1987), p. 85.
176. EPA, "Toxics Release Inventory 1999: Executive Summary," http://www.epa.gov/tri, p. E-3.
177. O'Brien, Making Better Environmental Decisions, p. 26.
178. Eric Pianin, "The Ties that Blind," Washington Post (July 16, 2001).
http://www.washingtonpost.com/wp-dyn/articles/A59494-2001Jul13.html
179. Steingraber, Living Downstream, p. 100.
180. O'Brien, Making Better Environmental Decisions, p.59.
181. Dona Schneider and Natalie Freeman, "Children's Environmental Health Risks: A State-of-the-Art Conference," Archives of Environmental Health Vol. 56, No. 2 (March/April 2001), p.104.
182. O'Brien, Making Better Environmental Decisions, p. 70.
183. Wade Sikorski, Sacrificial Rituals: A Murder Mystery in Montana (2000), http://www.midrivers.com/~wds
184. O'Brien, Making Better Environmental Decisions, p. 66.
185. O"Brien, Making Better Environmental Decisions, pp. 66-67.
186. Tee L. Guidotti, Philip Jacobs, "Implications of an Epidemiological Mistake: A Community's Response to a Perceived
Excess Cancer Risk," American Journal of Public Health Vol. 83 (1993), pp. 233-239.
Beth J. Fiore, Lawrence P. Hanrahan, and Henry A. Anderson, "State Health Department Response to Disease Cluster
Reports: A Protocol for Investigation," American Journal of Epidemiology Vol. 132, Supplement No. 1 (1990), pp. S14-S21.
Alan P. Bender, Allan N. Williams, Rebecca A. Johnson, and Helen G Jagger, "Appropriate Public Health Responses to Clusters: The Art of Being Responsibly Responsive," American Journal of Epidemiology Vol. 132, Supplement No. 1 (1990), pp. S48-S52.
187. Kenneth J. Rothman, "Keynote Presentation: A Sobering Start for the Cluster Busters' Conference," American Journal
of Epidemiology Vol. 132, Supplement No. 1 (1990), pp. S6-S13.
Raymond Richard Neutra, "Reviews and Commentary: Counterpoint from a Cluster Buster,"American Journal of
Epidemiology Vol. 132, Supplement No. 1 (1990), pp. 1-8.
188. Richard A. Goodman, James W. Buehler, and Jeffrey P. Koplan, "The Epidemiologic Field Investigation: Science and
Judgement in Public Health Practice," American Journal of EpidemiologyVol. 132, Supplement No. 1 (1990), pp. 9-16.
Robert J. Hardy, Gene D. Schroder, Sharon P. Cooper, Patricia A. Buffler, Howard M. Prichard, and Martin Crane, "A Surveillance System for Assessing Health Effects From Hazardous Exposures,"American Journal of Epidemiology Vol. 132, Supplement No. 1 (1990), pp. S32-S42.
189. Glyn G. Caldwell, "Twenty-two Years of Cancer Cluster Investigations at the Centers for Disease Control," American Journal of Epidemiology Vol. 132, Supplement No. 1, (1990), pp. S43-S47.
190. Daniel Wartenberg, Michael Greenberg, "Detecting Disease Clusters: The Importance of Statistical Power," American Journal of Epidemiology Vol. Vol. 32, Supplement No. 1 (1990), pp. S156-S166.
191. David Savitz, "Environmental Exposures and Childhood Cancer: Our Best May Not Be Good Enough," American Journal of Public Health Vol. 91, No. 4, (April 2001), pp. 562-563.
192. For a vigorous defense of Aristotle's science see: "Aristotle not a Dead Dog," in Paul Feyerabend's Science in a Free Society (London: Verso Editions, 1978).
193. Wade Sikorski, Modernity and Technology: Harnessing the Earth to the Slavery of Man(Tuscaloosa, AL: The University of Alabama Press, 1993).
194. Donna Haraway, Simians, Cyborgs, and Women: The Reinvention of Nature (New York: Routledge, 1991), p. 196.
195. B. Burt Gerstman, Epidemiology Kept Simple (New York: John Wiley and Sons, 1998), p. 8.
196. John Snow, quoted by Gerstman, Epidemiology Kept Simple, p. 9.
197. Terry Trieweiler, Montana Supreme Court (1999) No. 97-455, 988 p.2d 1236
198. Delegate McNeil, Montana Constitutional Convention, Vol. IV, March 1, 1972, p.1201.
199. Delegate McNeil, Montana Constitutional Convention, Vol. V, March 1, 1972, p. 1243.
200. Delegate Foster, Montana Constitutional Convention, Vol. V March 1, 1972, p.1243-44.
201. Quoted in Gerstman, Epidemiology Kept Simple, p. 2.
202. Walt Whitman, in J. Diamond, Your Body Doesn't Lie (New York: Warner Books, 1979), pp. 55-56.
203. Terry Trieweiler, Montana Supreme Court (1999) No. 97-455, 988 p.2d 1236.
204. EPA, "Toxics Release Inventory 1999: Executive Summary," http://www.epa.gov/tri, p. E-3.
205. Chief Judge Claude M. Hilton, United States District Court, Alexandria, Virginia, Matthew Glaving, et al. v. William J. Clinton, et al., No. 98-207-A (Filed: Sept. 24, 1998)
206. Chief Judge Claude M. Hilton, Alexandria, Virginia, "Matthew Glavin, et al. v. William J. Clinton, et al., No. 98-207-A, pp. 4a-5a.
207. U.S. Census Monitoring Board, "Report to Congress" (April 1, 2001), CMBC 72-424, p. 11.
208. U.S. Census Monitoring Board, "Report to Congress" (April 1, 2001), CMBC 72-424, p. 4.
209. Chief Judge Claude M. Hilton, Alexandria, Virginia, "Matthew Glavin, et al. v. William J. Clinton, et al., No. 98-207-A, pp. 10a-11a.
210. Sandra Day O'Conner, Department of Commerce v. U.S. House of Representatives, Nos. 98--404 and 98--564 [January 25, 1999]
211. Letter to me from Governor Racicot, August 7, 2000.
212. John Ritter, "Town clenched in suffocating grip of asbestos," USA Today (Tuesday, February 1, 1999), p. 8A.