You Make Pearson Cry #2: Deadly Mobiles? February 8, 2007Posted by Johan in Rants, You Make Pearson Cry.
While international media appears to have been spared so far, Swedish media decided to go creative when interpreting results from studies by IARC (the International Agency for Research on Cancer) on the possible dangers of mobile phone use. National tabloid Expressen goes for the big guns with a headline that roughly translated goes “cell phones damage your child’s brain.” In the article, it is claimed that the risk of “brain tumour” increases by 39 percent if you have used your mobile phone regularly for over 10 years. There’s no mention of children being at risk, so I presume this bit of the story was inspired by the fact that the risk only appears over the long term.
So is there truth to this? Yes, and no. If you go to the source, it turns out that people who have used cell phones for over 10 years do have an increased risk of cancer, and the difference is on the order of 40 percent.
But the increase is not for brain tumours in general. Rather, there is an increased risk of acoustic neurinoma, which IARC describes as “slowly growing benign tumours that generally have a good prognosis, as they only rarely undergo malignant change.” Now that’s interesting enough in itself, but wait, it gets better: the IARC report also states that the incidence of acoustic neurinoma in adults is on the order of 1 per 100 000 per year.
Let’s do some numbers on that. If the incidence for the normal population is 1/100 000, or a risk of 0.00001 percent per year, and regular mobile phone use over 10 years increases the risk by 39 percent, the incidence for these people is 1.39/100 000, and the risk grows to a whooping 0.0000139 percent.
So to summarise, heavy mobile phone use over a long period of time increases the risk that you will catch a benign, slow-growing tumour by a number that is, in real terms, somewhere on the order of being struck by lightning in your pinky.
The real, undiscovered danger in mobiles lies in their use as throwing weapons. I’m willing to bet that the risk of getting a mobile thrown at your head with deadly results is at least 39 percent greater than the risk of the mobile causing you cancer. Of course, such an argument relies on treating proportions as absolutes…. and that would just be silly, wouldn’t it?
By the way, this particular case has nothing to do with misinterpreting correlations, so from now on, I guess You Make Pearson Cry will concern rants on any bad science reporting, correlational or not.