Tobacco Documents Online

-26- t APRIL 1?, 1993 /NATIONAL REVIEW _ Junk Science L AST WEEK'S scare from the Environmental Protection Agency (EPA) was radon in schools. It grabbed headlines with the claim that there are 73,000 classrooms in 15,000 schools where this radioactive gas is over the agency's "ac- tion level" of 4 pCi/L. This led Congressman Henry Waxman to say breathlessly that it is "more danger- ous to attend school than work in a nuclear-power plant." (He did not add that nuclear-power plants in the U.S. have proved among the safest places any- one could choose to work. Indeed, in decades'opera- tion of up to two hundred nuclear-power plants not a single worker has died of radiation.) Some months ago we asked the EPA for the scien- tific articles and reports justifying their radon action level, and after a month's delay, during which our interest ebbed, we received an intimidatingly thick package. Last week we took that EPA package off the shelf and spent some hours going through the studies. We were amazed to fmd.that they don't sup- the EPA position at all. lwey fail to find any statistically proven associa- tion between residential or school radon levels and lung cancer. They constantly emphasize the "uncer- tainty" surrounding the arithmetical extrapolation to residential radon levels of lung disease suffered by workers in mines with high radon concentrations- As one cancer scientist, Gio Gori, wrote recently, the official cancer risk assumptions are "poignantly out of step with the scientific evidence." (Regulatory Tox- icology and Pharmacology, 16, 10-20, 1992.) And the EPA omitted from its package the most da,,,.,ing set of radon/lung-cancer studies, from Ber- nard Cohen, professor of physics and radiation health at the University of Pittsburgh.. Cohen's group has measured radon levels in 350,000 homes across the U.S. and subjected the data to every con- ceivable statistical check. He finds no basis for con- cern about low-level radon-indeed, the reverse: "Me [EPA's] linear theory predicts that lung-cancer rates should increase by 7.3 per cent for each pCi/L of radon concentration in homes, whereas our stud- . ies indicate that lung cancer rates actually decrease by about 6 per cent pCi/L." How so? An eminent biochemist, T. D. Luckey, has experimentally shown the health benefits of low- •level radiation and called the process "hormesis." Cohen's statistics suggest that not only is the EPA radon scare phony, but it could deprive millions of people of the benefits of hormesis. After all, rich peo- MAR 31 1993, ~ ple have been seeking better health for centuries by going to spas whose sole distinguishing physical characteristic is that they have higher levels of radon and other sources of ionizing radiation. Another piece of junk science from the EPA is the notion that thousands of non-smokers die of lung cancer from the smoke of smokers-a/kla environ- mental tobacco smoke (ETS). Now, everyone accepts that smokers assume a major risk for themselves. They increase their risk of lung cancer at least ten- fold. But ETS is cigarette smoke diluted thousands of times compared to the smoke smokers inhale di- rectly into their lungs. And it is hard to distinguish chemically from cooking smokes and from boiler- flue, tailpipe, and industrial emissions. The closest thing to science in the debate over ETS is a slew of statistical studies of the incidence of dis- ease among couples where one partner smokes and the other doesn't. Some of the studies show a mild statistical association (risk ratios like 1-2, compared to ratios of 2.0 and more that are normally required to establish association and a ratio of over 10.0 for direct smoking). Most fail to meet the 95 per cent confidence level usually adopted by statisticians to exclude chance clustering. The EPA's recent declaration that ETS is a "Class A carcinogen" was achieved by a quite shameless abandonment of regular scientific procedures. Since the American studies don't prove the case, the EPA dragged in a large collection of studies from Asia and Europe. Though it claimed to have "proved" the asso- ciation by a "meta analysis" or combining of the ex- isting studies, the EPA simply abandoned the con- ventional 95 per cent confidence level and applied a 90 per cent test in order to claim the result was sta- tistically significant. Alvan Feinstein, professor of medicine and epi- demiology at Yale medical school, wrote recently in Toxicologic Pathology that the EPA study on envi- ronmental smoke "simply ignored the inconvenient results and emphasized those that are (in a memora- ble phrase) `helpful-"' He said he had been told by a colleague that the EPA report on ETS was "rotten science" in the worthy cause of getting a smoke-free society. Professor Feinstein observed that govern- ment agencies funding scientific research often be- come "mechanisms of advocacy." That used to be called "lying," and it still should be. CqAPAY7(5)