Swedish Study Finds Cell Phone Cancer Risk 282
dtjohnson writes "A new Swedish
study has
found that heavy users of cell phones had a 240 percent increase in
brain tumors on the side of their head that the phone was used
on. The study defined 'heavy' use as more than 2,000 total hours,
or approximately one hour of use per workday for 10 years. An
earlier British
study was previously discussed
here that didn't find an increased risk, although that study
covered fewer subjects and only followed one type of brain tumor for a
shorter period of time. Or course, the biggest epidemiological
study of all is the one we are all participating in whenever we use our
cell phone. The results from that study won't be available for a
while."
Re:News? (Score:4, Insightful)
Assumptions (Score:5, Insightful)
"The way to get the risk down is to use hands-free," he told Reuters.
How does he know that? Did his study make that conclusion? The article doesn't say anything about use of hands free kits beyond that statement.
I think Mr Mild is making assumptions about the reason for the apparent 240% increase, and factors which he thinks may be important.
Not really (Score:5, Insightful)
Not really. The metering is lousy. The control group is corrupted. Heck, the technology is changing, so the signals are different. As a study, the world at large makes a lousy experiment for this.
dangerous use of statistics (Score:5, Insightful)
For example, to say something is associate with a 240% increase in risk can be technically accurate, but horribly misleading to most readers. If one in a billion people get a disease, a 240% increase makes your chance of getting it 2.4/1000000000. That is absolutely nothing to worry about.
Also, with this studay, they found out people who had tumors, then asked them if they used cell phones. The subjects probably had no doubt as to why this question was being asked, therefore this was not really a double blind experiment.
Has anyone ever been able to give a rat cancer by blasting it with amplified cellphone-type radiation? That would convince me of the possibility of cell phone risk much more than digging backward through statistical inormation does.
How high is the absolute risk (Score:5, Insightful)
Re:dangerous use of statistics (Score:4, Insightful)
Not to mention I suspect the people who used cell phones extensively twenty years ago are probably a very special group... probably with all kinds of interesting common factors.
Re:suprised? (Score:5, Insightful)
Why? As has been repeated ad nauseam whenever this debate comes about, the frequencies used by cell phones are non-activating. If holding a tiny, low-power transmitter next to your head causes cancer, then people who work at TV and FM stations should be dropping like flies.
All we know at this point (assuming the study's methodolgy holds water) is that there is a correlation between cell phone use and brain tumors. It could mean that cell phones cause brain tumors, it could mean that people prone to brain tumors talk on the phone a lot.
And even if it is eventually shown that cell phones cause brain tumors, that still doesn't necessarily mean it is the radio transciever aspect of the phone that is the culprit. It very well may be exposure to toxic chemicals used in the displays or the batteries, for example, much the same way toxic pesticides used around electrical pylons had people thinking high-voltage lines caused cancer.
Re:dangerous use of statistics (Score:5, Insightful)
Re:News? (Score:2, Insightful)
Re:but it can't be (Score:5, Insightful)
Re:News? (Score:1, Insightful)
Let's just say that these are complex issues and there are a lot of scientific opinions and interests in play.
At least, the surgeon general isn't recommending the use of cell phones for health reasons, like in the old days of cigarettes.
Re:Heh. Stupid study. (Score:2, Insightful)
Link to the former-pdf, now HTML-ized Google cache of the study from the original site, in both Swedish and English: Here. [72.14.203.104]
Even a cursory look at the linked study will show that there have been many, many studies on the effects of RF on animals with conflicting, confusing, and uncertain results. Unfortunately, I'm not a scientist specializing in this field, so I really can't comment one way or the other on the validity of the tests.
It's difficult not to hand-wave this study away without some real, significant, reproduceable results.
An increase from 1 person to 2.4 people getting cancer is serious, if your sample size is 10 people. If your sample size is 10,000 people, or 50,000 people, the difference between 1 and 2.4 is statistical error. To really derive anything further, we'd have to go read the study.
The trouble with doing scientific studies on real, moving people is that it's exceptionally difficult to control external variables. For instance: GSM cell phones (Cingular, T-Mobile, a few minor regional carriers) have a total of four bands they operate on, 850, 900, 1800 and another band that escapes me. CDMA phones (Verizon, Sprint, etc) operate on others, and iDen (Nextel, Southern LINC, etc) phones operate on yet another. Each type varies in wavelength and power output, so it's a vast generalization to say "Cell phones are bad for your brains", because of the vast differences between the services, the cell phones, and the effects of different frequencies on different parts of your brain.
Random appeal to authority: I'm a ham radio operator, and they make us learn interesting things about what too much RF does to you. But at the frequencies we operate, site surveys start being required when you're pushing more than 50 watts at 146 MHz (for instance). 50 watts is something like 50-100 times the amount of power that cell phones push, but, again, at different frequencies, so I'm not really sure I said anything relevant there. It's just hard to tell.
By the time the studies start showing reproduceable evidence, I'll be out of college and far away from the wireless industry, hopefully reducing my chances of being sued
Re:suprised? (Score:2, Insightful)
That's how.
Re:Heh. Stupid study. (Score:2, Insightful)
Re:Studies like this are always a problem (Score:2, Insightful)
Isn't that a bit like saying
The American study is automatically suspect. I don't think the country that gave us Intelligent Design can be trusted with anything scientific. Ever.
I'm not even going to touch the German/Human Rights issue. Europe doesn't have its own PATRIOT act or Guantanamo yet, you know.
Re:suprised? (Score:3, Insightful)
Re:Not really (Score:3, Insightful)
Re:Assumptions (Score:5, Insightful)
Riiight... If you can make the GHz RF radiation travel into the wire of your earpiece, then you should patent it quickly, because then you've managed to do something that no radio engineer deemed possible... There's something called matching impedance you might want to investigate.
Re:Assumptions (Score:2, Insightful)
While we are at it we can argue about how a few watts of photons with less energy than infrared can cause cancer while kilowatts from a nice comfy open fire do not.
Re:public health (Score:2, Insightful)
Flawed (Score:5, Insightful)
First off, despite multiple studies done that prove no correlation between brain tumours and mobile phones this claims to have found something. Now I guess other factors may have come in to those other studies some bias etc. However, this article details an initial study that also showed _no_ connection. It was only after they altered the questionaires and retested people that they found something. Whats more, they then did no further alteration to the questions and simply ran with the same test only on a bigger scale.
There may be a detailed explanation of why that occured but with currently released information weve no idea how many times they were willing to alter the questions to get the answers they wanted, and no explanation of which questions were altered or why. What adds to the suspicion is the fact that the only reason the first test was thrown out was 'short latency' and 'low numbers' of people. Neither of which affect the questionaire.
So what we have is a group of people who rely on getting a result for there funding. (No differently to the previous studies.) After they got no real results from a first test, altered it in a way that appeared to have no bearing on that initial test. They then found they got results... Doesnt really inspire any confidence in there impartial testing.
Secondly, something others have pointed out already, asking a bunch of people with tumours when they started using mobile phones and then roughly getting rid of other factors that could have caused them based on a questionaire... Not a great method of working this out.
Whatever you thought of the study seen on the BBC site it raised a very good point about something that would cause a bias. 'reporting from brain tumour sufferers who knew what side of the head their tumours were on.' etc. This test doesnt even begin to try clamp down on these kinds of bias. Even if this test was entirely fair, the results are far from dramatic. With excessive use it shows only a relatively small increase in cases. With a potential for people to be increasingly suspicious of there mobiles the more they use them this could easily be put down to false assumptions.
As far as im concerned this study is severly flawed. The other studies are also flawed, to a degree, but until someone actually has decent evidence that these things are causing damage then its not going to stop the millions of people who use them. I certainly wouldnt say mobile phones are safe but there is still little to no evidence suggesting they harm us. (and arguably more evidence to suggest that they dont.) The presure is definately on those who have to prove a link.
Re:Actually, it's quite correct (Score:2, Insightful)
You seem to think that this very small risk is something that can be overlooked. However, I could point a few million people who believe otherwise.
I think you see where I'm heading - while it's true that there are higher risks than the one mentioned here (or in the article), the whole point is that these are risks that we can easily avoid. And the increase in relative risk has nothing to do with the actual incidence of the disease in the studied population.
Incidentally, this "there are worse things to die of" argument was -and still is- used by a lot of smokers. Not that I would compare using cell phones to smoking (there is still too little data concerning the risks -or lack of them- associated with cellphones), but it is interesting to see how the human mind tends to follow certain patterns...
Re:suprised? (Score:3, Insightful)
Well, if it is unrelated, that would be easy to prove. What side of the brain are most tumors found? Or is it equal on both? What side of the head do most people hold the cell phone on? I know I use phones mainly with my right-hand/ear, so preferences do exist. If there is an unrelated propensity for right-brain tumors and right-hand cell usage, then the causation implied with the correlation is flawed. There is rarely a "cause" found for things in these types of scientific studies, just correlations explored.
And, of course, the study wasn't double-blind. I didn't RTFA, but if they found people that were diagnosed with cancer, then asked them questions, it would be human nature to assume the answer and change their answer to match the expected answer. No, this isn't lying, it's an actual subconscious alteration of perceptions, like the placebo effect. Tell someone they have cancer in their left hemisphere and ask them if they usually use the cell phone on that side, and they will be more likely to answer that they do. So, the only way to fix that problem is to lie to the patient about their disease (usually unethical) in order to remove that bias. Since that's impractical, these studies will forever be flawed. Or, we could study 1,000,000 people for 20 years and get a reasonably accurate answer to the question.
Read the Study (Score:4, Insightful)
Re:public health (Score:4, Insightful)
For example, should we introduce new regulations on cell phones that force cell companies to build twice as many towers which will statistically save ten lives over the next twenty years or use the same amount of money to introduce subsidized prostate and breast cancer screening programs that will statistically save a thousand lives per year?
Public health is all about the economics. You put your money where it will do the most good. Not that any of these studies are actually conclusive enough to justify anything.
Re:link to the actual study (PDF) (Score:3, Insightful)
First off, in response to another poster in this thread, the choice of controls is correct. In case control studies you look at groups with and without an outcome, in this study various brain tumors, and then examine whether the rate of exposure, cell or cordless usage here, differs between the two groups. Having other cancers is an entirely different outcome, and case control studies do not allow one to examine multiple outcomes by definition. A cohort study would allow for examination of multiple outcomes, but is inappropriate here since the incidence of brain tumors is so low as to make a cohort study prohibitively large and expensive.
Second, from reading the actual study as opposed to the news summaries I believe the results to be valid. Why? The results meet many criteria for causality and are strong statistically. Read on for what I mean by this.
The case for causality: First off there is biologic plausibility. Read the second full paragraph on page 9 of the pdf for discussion of this issue by the authors. Incidentally the assertion by other posters that these results are invalid because they show roughly similar odds ratios for analog, digital, and cordless phones is addressed and shown to be untrue in the first full paragraph on page 9 (as well as in the discussion of frequencies used in the introduction).
Next there is a clear dose-response relationship, as the odds ratio increases with greater cumulative wireless phone usage. This also partly addresses the issue of temporal relationship, as long term cell phone usage would necessarily predate the onset of recently diagnosed tumors.
Finally, the results seem statistically sound. By this I mean that the 95% confidence intervals do not cross 1.0, and that the relationship between exposure and outcome persists after correction for age and socioeconomic class. (Sex wasn't corrected for since the controls were already matched by the study design.)
Does this mean that I'm going to immediately stop using my cell phone? No. However, I'm going to keep on using it because I value its convenience more than the possibility of developing brain cancer at some multiple of a low rate.