Toxic beryllium and genetic testing.

Cancer: Current Research Two prominent medical journals recently published epidemiologic studies [Inskip et al. (1) and Muscat et al. (2)] on cellular telephones and brain cancer. These papers were featured in a front-page story in The New York Times (3). The message of the story was essentially that these were definitive studies that put the question of handheld cell phone hazard to rest. The television networks and other media also ran major stories and conveyed essentially the same message. Most of the patients and most of the data analyses in the Inskip et al. (1) paper as well as in the Muscat et al. (2) paper are irrelevant to the issue of whether handheld cellular telephones cause brain cancer. Most of the patients (86%) in the study by Muscat et al. (2) used car telephones or bag telephones, not handheld telephones; the antennae used with car and bag telephones are well away from the head, so there is little, if any, exposure of the head to the energy. Most of the patients (82%) in the study by Inskip et al. (1) had no or negligible use of a handheld telephone. This crucial information is buried in both the Muscat and Inskip reports. Thus, most of the analyses in these studies (1,2) show that if you have no exposure you have no effect, and are irrelevant to the issue. Further, how is it possible to see an effect with the small group that is left, which is the only group that is relevant, particularly given the long latency for cancer to develop? It is well established that cancers resulting from an environmental insult typically have a long latency. In the study by Muscat et al. (2), for example, the mean time from the date the 66 patients began to use a handheld cellular phone to the date when the authors assessed for cancer was less than 3 years. The situation is similar in the Inskip et al. study (1). A considerable body of cancer literature indicates that a statistically significant increase in cancer is unlikely to appear in such a short time, even if the exposure caused cancer. Thus, a no-effect result would be expected in these small groups, even if handheld cellular phone radiation exposure did cause cancer. Shortly after the papers by Inskip et al. (1) and Muscat et al. (2) were published, another epidemiologic study by Johansen et al. was published in another prominent medical journal (4). The comments that apply to the other studies, in general, also apply to Johansen et al.’s study, but one more comment can be made about it. Johansen et al. (4) lumped together in their analyses car, bag, and handheld telephones as though the use of all these types of telephones gave the same head exposure as handheld telephones. Because most of the people at that time were using car and bag telephones (1,2), most of Johansen et al.’s cases (4) had no exposure. Thus, Johansen et al. (4) had confounded data and could not see an effect even if a large one existed. All Johansen et al.’s analyses show is that if you have no exposure, you have no effect; thus, these analyses are irrelevant to the issue. The carrying out and publication of these studies seems to fall into the pattern that I summarized in a recently published paper concerned with ethical questions in this area of research (5). The public has not been well served with respect to a significant public health issue.