POWER LINES AND LEUKEMIA

The idea that the field of electric power lines has health effects is as old as the power lines themselves, and predates the present fear-mongering with a coercive social engineering agenda that started in the early 1970s. But almost as long is the history of searching for such effects, and nothing convincing has ever been found, in spite of numerous and thorough studies on both sides of the Atlantic. There was, for example, a Soviet report in the 1950s that linesmen suffered from a decreased sex drive; this led to a flurry of serious studies in the West, none of which found any evidence for the allegation.

But that does not mean that no health effects are possible, for thorough studies must confirm or refute each new allegation. If it were to be alleged that living near a power line shortens one's life span, or causes pneumonia, refutation would be very speedy, because everybody dies, and pneumonia is a common ailment. But if the allegation is a very rare phenomenon such as child leukemia, convincing confirmation or refutation is extremely difficult, and until it is obtained, the fearmongers have a field day, especially when they are vigorously supported by the social engineers, de-industrializers and a sensationalist press..

In 1979 the Amer. J. of Epidemiology published a study by Dr Nancy Wertheimer, a psychologist, claiming to have found a cor-relation between fatal cancers in persons under age 19 and high current configurations in or near their homes in the Denver, Colo., area. Although the paper left much to be desired (small sample, rough guesstimate of exposures, etc.) it deserved to be, and was, taken very seriously.

By far the most thorough study is that by a group of scientists led by Dr. D.A. Savitz of the School of Public Health, U. of N.C., originally commissioned by a special New York State panel ap-pointed to investigate the matter. Since my last report [AtE Sept 87], their study has been published.(1) It raises suspicions that a cor-relation between exposure to magnetic fields and childhood can-cers may exist: the odds ratio (probability of relation to investigated source to probability of randomness), though only half that found by Wertheimer, is 1.5. However, the uncertainties due to insufficient samples and other reasons is still so large that the authors characterize the results as "suggestive, but inconclusive."

The main difficulty is getting a large enough sample for an event as rare as childhood leukemia. In US children under 10 it occurs on an average of 3.1 cases per 100,000 children. If you actually have a sample of 100,000 children, many laymen will guess that is "more than plenty" and that it almost certainly contains the expected number of leukemias, namely 3. In reality the probability of finding exactly 3 random leukemias is only 22.4%. The probabilities of finding n random cases [call them P(n)], are (2): P(0)=4.5%, P(l) = 14.0%, P(2) =21.6%, P(3) =22.4%, P(4) = 17.3%, P(5)=10.7%, P(6)=5.5%, . . . P(10)=0.3%. The probability of getting more than the expected number, 3, is no less than 37.5%. All of this may seem surprisingly high to people unfamiliar with statistics, and using such catch-phrases as "above the national average," the scare mongers go to court against power lines, and ignorant judges issue injunctions.

To confirm or reject the allegation, a much larger numbers of children, each in a well known surrounding of power lines outside and in the home, is needed. It is my understanding that the Savitz group is continuing this work.