Mobiles don't kill
The results of the Interphone study into the relationship between cellphones and cancer, 10 years in the making, are finally out.
The study, conducted by the World Health Organisation's International Agency for Research on Cancer involved 13 countries including New Zealand and questioned thousands of people.
Published in May in the International Journal of Epidemiology, the conclusion is that cellphones don't cause brain cancer.
The results are pretty convincing. In fact a naive interpretation would be that cellphones actually prevent cancer - though that is not what the study's authors conclude.
The design of the study was quite simple. In each of the participating centres, an attempt was made to interview every person who received a firm diagnosis of meningioma or glioma - respectively cancer of the brain lining or the brain itself.
For each person with a case of cancer recruited into the study, at least one similar but healthy control subject was recruited.
Once selected, all subjects were interviewed about their cellphone use. They were asked if they used a cellphone regularly, and if so how long they had been using one, how many calls they made and how much time in total they spent on the phone.
The answers were totted up, combined, and used to determine an odds ratio for different degrees of cellphone use.
Odds ratios are related to relative risks, which are a familiar sight. Whenever you read about one thing increasing or decreasing risk of some other thing (usually a dread disease) by some percentage you're looking at a relative risk.
Odds ratios are different but convey a similar idea. Here they were used to measure whether a person was more likely to get cancer if they used a cellphone regularly than if they did not.
Boiled down to simple terms, an odds ratio greater than one would suggest cellphones are bad for you.
The interesting thing about the Interphone study is that almost all the odds ratios came out less than one. That means that people who used cellphones regularly were in general less likely to get cancer. In only one category, those who reported using their phones more than 1640 hours, did the odds ratio rise above one.
The authors concluded that cellphones do not cause cancer in the brain. But what about the exceptional ratios for the high users? For a start, these exceptional results don't fit with the others. There is a principle in medicine that as the dose goes up so should the response. But in this study there was no gradual trend upwards in cancer rates with increasing use, just a single spike.
Another hint that something is amiss in the high-use data is that many of the figures don't make any sense. Some of the subjects would have to have talked on their phone for 12 hours a day, every day to get the use they reported. Don't forget that subjects in this study had brain tumours, which could well affect their recollection and reporting. It's probable that tumours caused some to report impossibly high use of their phone, rather than high use leading to tumours.
We're in the hands of epidemiologists here, so we should exercise caution. While epidemiology is fantastic at finding associations between things, it can't actually establish that one thing causes another. That doesn't stop imagination filling in the gap, inferring causes and effects that are not really there. This particular study also depends on interviews and brings the frailty of human memory into the mix.
Nevertheless, the Interphone study didn't find an association between cellphone use and brain tumours. This fits with other kinds of research that have looked at cellphone radiation in animals. It also fits with the absence of any known physical means by which the radiation emitted by cellphones could cause cancer.
In the end, this huge study has determined that there is a slight possibility that very high cellphone use increases risk of brain cancer. But it probably doesn't.
You may prefer to be cautious, but I am encouraged by the results of this study, and many others, to use my cellphone without worrying that it will kill me.
* Hayden Walles holds a PhD in computer science from the University of Otago and writes the True/False column for Your Weekend magazine. See truefalse.co.nz.