Factual · Powerful · Original · Iconoclastic
Household radon gas causes seven thousand to 30,000 lung cancer deaths each year in the United States. Or so says the EPA. As such, the invisible gas would be the second-leading cause of lung cancer, which in turn is the leading cause of cancer death.
Identifying and remediating this problem would set the nation back $45 billion. Peanuts, yes, but every cent helps these days. Which is why we need to consider that the EPA’s case against household radon is becoming more translucent than the gas itself.
The EPA’s claim that radon levels in homes are carcinogenic, like so many of their assertions concerning carcinogenicity, are based on what’s called a linear, no-threshold extrapolation. This theory says that because a substance, such as Alar, or a number of pesticides which the EPA is preparing to ban, causes tumors in lab animals at doses hundreds of thousands of times greater than the doses that humans could possibly absorb, that humans are nonetheless at risk of developing tumors from these chemicals. But radon may turn out to provide the best evidence that this assumption, beyond being scientifically unproven, is demonstrably false.
The problem here isn’t reliance on furry little lab animals. Rather, the radon data for which the EPA has relied on for its extrapolations aren’t rodents but miners. Miners who breath in huge amounts of radon, especially those who smoke, have extraordinarily high rates of lung cancer. But is it fair to say their experiences have any bearing on those who breath the gas in at fractious levels?
Happily, there have been a great number of studies on just this, and the overwhelming majority say they do not. Several early studies of housing exposures, mostly done in Sweden, did seem show a household radon risk.
But Dr. William J. Blot and other researchers of the National Cancer Institute writing in the Journal of the National Cancer Institute (JNCI) have observed, "Most of these investigations were based on small numbers of subjects, however, and few incorporated actual measurements of radon in homes."
Another Swedish study, released a year ago from the Karolinska Institute, played to such headlines as "Study Confirms Radon Risks of Lung Cancer." But this may say more about Swedish methodology than radon. From pesticides to power lines to organochlorines, the Swedes have an incredible knack for finding cancer links wherever they happen to look for them. At any rate, even the Karolinska study also showed no significant increase in lung cancer among those persons who never smoked.
On this side of the Atlantic, Dr. Bernard Cohen, a radiation physicist of the University of Pittsburgh, has analyzed in repeated studies the relationship between lung cancer and residential radon levels in over 400 U.S. counties. Far from finding higher rates of disease among those with higher exposures, he found lung cancer rates in high radon areas were much lower, prompting his observation that, "The results are contrary to the predictions of the linear, non-threshold theory..."
Confirming Cohen’s work is that of physicist and radiation consultant Dr. Ralph E. Lapp of Alexandria, Virginia, who compared cancer rates to household radon rates in Morris County. In this area, the EPA shows an average radon screening measurement twice the EPA’s recommended maximum allowable level.
Using the EPA’s own risk calculation, this exposure alone should produce 377 lung cancer deaths per year, over and above smoking- caused lung cancer deaths. Yet, in all of Morris County, there were only 190 lung cancer deaths in the year Lapp studied.
Lapp then looked at New Jersey counties with much lower radon levels, averaging less than a fourth that of the Reading Prong average. Yet, far from having a considerably lower rate of lung cancer deaths, these counties averaged a much higher rate. Looking outside of the state, he compared New Jersey’s Reading Prong to counties in Texas.
He found that, although according to the EPA data, radon levels in New Jersey were seven times greater than they were in Texas, lung cancer makes up an equal fraction of the total cancers in both states.
Yet another research team doing testing in Florida reported, "People in localities with the highest radon levels have a lower percentage of malignant neoplasms [cancerous tumors] compared to people from areas with no radon exposure risks."
Likewise, populous areas with high lung cancer rates where radon levels have been studied have generally been found to have below average radon levels. This includes the New York City area, the San Francisco area, England, and China.
In August, a Canadian study in the American Journal of Epidemiology, according to one of its co-authors, found "that after adjusting for cigarette smoking and education level, there was no increase in the relative risk of any histologic [different] types of lung cancer. There was no increase in the relative risk correlated to cumulated exposure to radon."
Finally, the current issue of the JNCI again found no correlation between radon and lung cancer in non-smoking Missourians.
None of the previous studies have served to move the EPA’s radon policy a fraction of an inch and there’s no reason the latest JNCI one will either. Why? As Bernard Cohen and a co-author, Harvard Medical School Professor of Medicine Graham Colditz stated in a 1991 paper, "A great deal more than radon is at stake here. If the linear no-threshold theory fails for radon, it must surely fail for all other types of radiation, and very probably also for chemical carcinogens."
To quote Consumer Reports, which duly accepts all EPA pronouncements on environmental risks — except to occasionally say it has understated those risks — "Radon poses a far greater, or at least better- documented, risk to life than other atmospheric pollutants such as hydrocarbons, carbon monoxide, or industrial waste gases; or than such contaminants of food and water as Alar, aflatoxin, nitrosamines, or PCBs."
What would happen if our environmental regulators were to admit that this "far . . . better documented risk" is groundless?