I recently wrote a New Republic column about cell phones and brain tumors. I was gratified to see my column get some attention. Yet when I started to see it pop up on websites such as verizoncustomer.net, and when I started getting emails from other public health researchers, I realized that some clarification and correction were in order.
Fortunately, I am blogging for TNR this week, helping to tend CitizenCohn while Jon is overseas. That gave me a chance to address the issue once again:
In my view, anyone considering these questions should keep in mind two apparently discordant insights: First, the absolute cancer risk to a given cell phone user—though imperfectly known—is very small. Cell phone radiation would not make my top-20 list of population health problems. Second, there remains a real specific issue regarding potential links between cell phone use—particularly long-term use—and brain cancers that should be addressed.
My original piece for TNR effectively communicated the first insight. The second, not so much….
More here.
Given the reliance of the WHO warning on purely epidemiological studies - that is, the total absence of a plausible mechanism - isn’t it a bit worrisome for the quality of the study that rates for some of the cancers were lower among the heavier phone users?
Arguing for more funding for research on cell phones & cancer is arguing for less funding for research on other things. Given finite resources to commit, wouldn’t it be better to focus on other things where there’s a higher likelihood of finding a real problem with real solutions?
And since epidemiological research can never prove the complete absence of a link (it can only provide likely upper bounds on rates), no amount of money and effort will ever satisfy those who equate “radiation” with “health hazard” without concerning themselves with the underlying physics.
I agree with most of Foster’s comment - but I think he may go too far towards disparaging epidemiological studies. The utter lack of a plausible mechanism is of course a concern for the (literally!) tinfoil-hat crowd warning us of cell phones (Bob Park is great on this issue), but while this problem with their narrative would demand greater scrutiny and more skepticism of purely epidemiological claims, an attempt to ask whether their studies were properly constructed and the like, if those studies held up methodologically and demonstrated a great increase in risk then the lack of an apparent mechanism would hardly be the fault of those ringing the warning bells. The whole point hear is that it’s a two-fer: the physics can’t explain why there would be a risk, and the epidemiology does not in fact convincingly demonstrate a meaningful risk. But epidemiology, as such, and properly conducted, remains valuable.
Ben Goldacre of the (London) Guardian’s Bad Science column, blog, and eponymous book argues, convincingly to me, that science reporting on risks should always use absolute risks, not relative ones. For (wholly invented) example, not “New study shows that microwave popcorn doubles the risk of tongue cancer!” but “Microwave popcorn increases risk of tongue cancer from 1 in 100,000 to 2 in 100 000″.
The two messures converge when the absolute risk is high, as with Paleaolithic dangers of being killed by predators or enemies. Our brains are perhaps not well adapted to deal with the much lower risks of our safer lives.