(Jun. 19, 2010) The massive Interphone cellphone study released earlier this year dismissed its own work because it failed to prove what researchers were looking for, namely, evidence of adverse effects of cellphones. Also rejected was evidence in their research that showed cellphones may reduce brain-tumour risk. The Interphone researchers may well have missed the real story in their work because they didn’t want to see it.
Cellphones can cause cancer and other medical problems, one whack of studies says. Cellphones are safe, says another batch.
Which is it? Both sets of studies can’t be right.
The definitive study that many expected would settle the controversy, a multi-year, 13-country effort organized by the World Health Organization’s International Agency for Research on Cancer, came out last month. To its researchers’ surprise, the data indicated that cellphones can be both safe and dangerous, and that cellphone use can even reduce the risk of cancer. It all depends on how much cellphones are used.
The researchers’ reaction to the results they obtained? Don’t believe the data, they effectively warn. We must have botched this up somehow, especially since most of the data shows cellphones actually reduce cancers.
The researchers don’t disown their work in its entirety — they do find believable the small subset in their data that indicates cellphones can be dangerous. This bad-news subset, they recommend, warrants further research. But not the overall good-news thrust of the data.
The study that these researchers disavow, called Interphone, is the largest-ever case-control study of cell-phone use and brain tumours, with the largest-ever numbers of users of 10 years or more. To the researchers’ dismay, those who didn’t use cellphones were likelier to get meningloma and giloma, two common cancers of the brain, than those who did. Worse from the perspective of the researchers, veteran cellphone users tended to fare even better than those who had logged less time on cellphones. For example, those who had been using cellphones for more than 10 years, and had cumulative call times of 115 to 360 hours, had a 50% lower chance of contracting meningloma than a non-cellphone user. With five to nine years of cellphone use, the odds dropped to 33% and those who used it for only one to two years saw a 20% reduction.
Where could the researchers have gone so wrong? In lots of places, they report, going to great pains to list the different types of biases in their data that they may have inadvertently introduced. The potential deficiencies in their data, in fact, dominated their discussion of the study. The researchers’ ultimate conclusion: “We have no certain explanation for the overall reduced risk of brain cancer among mobile phone users in this study, although selection bias is almost certainly a contributor.”
But there is one possibility that the researchers — so focused on bias in their data — did not explore. The researchers dismissed the possibility that the data could be perfectly valid, even though they noted that other studies had also found that cellphone use can promote human health. In other words, the only certain bias in this study lies among the personal biases of the researchers.
The notion that small amounts of radiation can be beneficial to human health has a large and growing following, in the form of a scientific discipline known as hormesis. Under hormesis, what can be harmful in large doses — whether arsenic, iron, vitamins or radiation — is necessary to health in small quantities. This theory of hormesis, until the last decade on the fringes of science, has been entering the mainstream. Theses on hormesis are winning prizes for their scholarship; hormesis now increasingly appears in academic texts and in the curriculum of schools of medicine and toxicology.
But the bias against hormesis persists, particularly among official agencies associated with governments and the United Nations, the backers of the Interphone study. This hostility is so intense that, even though the results of the Interphone study are perfectly consistent with those that hormesis might predict — low amounts of radiation provide benefits, while at high levels the radiation becomes deadly — the Interphone researchers couldn’t bring themselves to recommend research into this possibility.
Why not? I asked one of the Interphone authors, a scientist at the Sydney Medical School in Australia. It would be too complicated to set up such a study, he suggested, and too expensive, with too many confounding variables. And besides, he explained, some hormesis proponents are apologists for polluters who seek “to ‘sanitize’ low levels of exposure to harmful agents,” adding “it certainly has been used by some to try and minimize concerns about health in the peaceful use of nuclear energy.”
The biggest confounding variable of all, it seems, may be the injection of politics into science.