Recently came across an especially irritating editorial in the Washington Times and decided I really could not let the contentions stand.
Here it is, by Dariusz Leszczynski:
Helsinki/Finland, January 11, 2012-Epidemiological studies are given the most weight in evaluation of human health effects. Therefore, when researchers started their effort to find out whether cell phone radiation causes brain cancer, epidemiology was given the most of attention – and the most funding.
Well… yes, since Epidemology is the study of health events, disease patterns, health statistics and disease rates and their relation to factors like environment, lifestyle and other causes, it would seem to be the field of study that would apply to such a question.
It’s as straight forward as determining that geology is the appropriate field of science to look to when trying to determine the characteristics of a rock.
However, and please let me play “devils advocate”,
Only if I can play with science advocate.
is the epidemiology overrated?
There, are we done?
Will epidemiology ever give us reliable answers concerning cell phone radiation and brain cancer?
Yes, and they have. Or is it simply that you don’t like the answer and want it to be something else, therefore you consider it flawed?
In 2010 and in 2011, two of the largest epidemiological studies on brain cancer were published. It appears that the time and money were used generously,
There’s a lot of interest in the topic, so a lot went into it. I’m not certain which studies you mean, but there have been some enormous ones recently.
but the studies failed to provide reliable answers concerning cell phones radiation and brain cancer. Flaws in the design of both studies prevented delivering conclusive answers.
Really? Well, if you say so. But thankfully, we don’t have to rely on any two studies. Two studies don’t mean much in the world of epidemiology anyway. To actually get a conclusive answer, you need to have confirming data coming from many studies. In this case we’re lucky enough to have literally thousands. So, you could actually discard two of them if you so choose and it won’t change the balance of the evidence much, because there’s such a huge amount from other sources.
It was 1999 when the largest case-control epidemiological study, INTERPHONE, was planned. At that time, optimists hoped that by the end of this project in 2004 we would know whether cell phone radiation causes brain cancer.
Actually, I think we had a pretty good idea even back in 1999, so it doesn’t seem very optimistic to think we would by 2004. That would be like me predicting that in the year 2017 we’ll know that the earth revolves around the sun. Unless there’s some kind of complete collapse of civilization that leaves behind only a handful of completely uneducated people, I am pretty sure we will know that in 2017, since we do already know it now.
I think I see where this is going though. The Interphone study was supposed to be one of the largest studies of this type and would dispel the doubt forever. It pretty much did.
After several delays, INTERPHONE published the results of the glioma brain cancer study in 2010.
The results were confusing, to say the least. Use of the cell phone for less than 10 years seemed to have a “protective” effect, whereas the use of the cell phone for more than 10 years showed a small increase in glioma occurrence.
Well I agree on one thing: The study abstract didn’t do a very good job of putting this all in context. It might simply be that research scientists are very apprehensive about using absolutes and tend to talk in degree of confidence. The tiny increase in giloma, but only in certain subsets was almost certainly statistical noise. It was miniscule. The “protective” effect can be attributed to a combination of statistical noise and possibly some slight confounding factors.
The balance of the data provides pretty good confirmation of no overall risk increase. Again, this should have been made more clear. The problem largely stems from having non scientifically literate persons get involved in the reporting. Reports and public officials have a tendency to focus on very narrow portions of a study like this and take them out of context. They will generally then demand to know whether the researchers can be 100% confident that this is not in fact a risk effect. The answer to that question is always no, statistical analysis never regards anything as being 100% certain. Then the study gets reported as if it raised doubts, when it actually does not.
Several problems with the design of INTERPHONE were debated. By design, the INTERPHONE study was unable to detect brain cancer induced by cell phone radiation because of its long (over 10 years) latency period.
Okay, that might be the case, but plenty of other studies did look at longer latency periods. A few went so far as to track down some of the early adopters of cell phones who started using them frequently in the early 1980′s and they also found no increase in brain cancer.
That said, even if the AVERAGE latency period were something like twenty or thirty years, it’s hard for me to imagine that there could be a bell curve so narrow as to have zero detectable risk increase after a much shorter period of time.
At the time of execution of INTERPHONE (2000-2004), cell phones were in common use for only a few years. There would be not enough time for the development and diagnosis of brain cancer if it was caused by cell phone radiation.
It does not matter how common they were by the early 2000′s. The fact of the matter is that they have existed since the late 1970′s and they have been used by many people since then. Sure, the actual proportion of the population that began using cell phones a lot in the early 1980′s is small, but it’s still more than large enough to produce good study results.
It’s not even really a cell phone issue. Wireless phones are just UHF/Microwave transmitters and those have been around for ages. There are studies that have been done on others exposed much longer. Police officers started using radar guns in the late 1950′s to measure the speed of motorists and some cops spent thirty years working highway patrol with a radar gun in their car. Others spent their careers as microwave technicians for AT&T or television networks. Military personnel worked on the deck of ships with radar antennas energized nearby.
Studies have been done on these individuals. Many of them, in fact. The results are consistent and compelling: The only health effects ever detected are acute thermal injuries and no chronic effect of exposure to RF fields has ever been documented.
However, there was an even more important design flaw. The information about the extent of exposures to cell phone radiation was based on individual recollection of the subjects in the study. The study subjects were asked about their history of using cell phone, including how long and how many phone calls they made in the past.
Perhaps in this study, but not in all. While it may introduce a potential source of error, I’m hard pressed to see how this could possibly skew the studies that badly. Even if you rely on spotty recollection, the fact that people who reported being heavy phone users show no greater cancer risks than those who never owned a cell phone at all would seem to be pretty hard to mess up.
By the way: Studies on cigarette smoking and cancer have largely been based on the subject’s recollection of how many packs they usually smoked a day. Despite this, they had no problem picking up on the fact that tobacco causes lung cancer.
It is a very unreliable method. Who of us remembers how many and how long calls made a few days ago? The study subjects were asked to recall cell phone use up to ten years before the study.
Okay, lets see if I can do this…
Got my first cell phone in the summer of 2001. Before that I had used cell phones a bit, but only occasionally when on that belonged to someone else. I worked for a company that sold cell phones so I had a good plan with a discount. Consequently, I used it a good few minutes a day or more. I would say my use has generally been on the increase since then, although not always. I’ve generally made or received three or four calls per day, usually each one only being a few minutes. Occasionally I have longer calls. In 2004 and 2005 I had a job that had me on the road a lot and my usage went up to about a dozen calls a day, but mostly short. As it stands now I use about 180 minutes of talk time in a month, but occasionally one or two long calls can push that way up. That’s how it’s been for the past few years.
Therefore, by design, INTERPHONE compared reliable information concerning diagnosed cancers with entirely unreliable information about exposures. Such kind of comparison can not produce reliable result, as was seen in the confusing results of the study published by INTERPHONE in 2010.
Again, you’re presuming that this error is so great that it would make someone who has never owned a cell phone indistinguishable in risk from someone who says they’ve been a heavy cell phone user for the past ten years. That just does not make sense. Even if recollection skewed the data, it shouldn’t so enough to cause that kind of discrepancy.
In 2011, the Danish Cohort published another largest study, evaluated in this column in December 2011.
Similarly to INTERPHONE, the Danish Cohort compared reliable information on diagnosed brain cancers with the absolutely unreliable information about exposures based not on the use of cell phone but on the length of subscription with the network operator.
No. That’s actually perfectly reasonable. It stands to reason that a person who has a cell phone contract and owns a cell phone will be more prone to using a cell phone than one who does not. This is even more true in the early years. In 1983, a handheld cell phone cost about four thousand US dollars. Anyone who pays that much for something obviously has reason to do so. For example, real estate agents were some of the first to embrace the technology, because even given the high cost, they needed to make appointments while traveling between properties.
It might be imperfect in that some cell phone owners will use it more than others, but a cell phone owner will always use it more than one who does not own a cell phone.
The study also contaminated the control group with the cell phone users.
The study looked at the habits of long term user as compared to the general population and to groups of similar demographic profiles. Some of those included those who had used a cell phone as well, but didn’t you just assert that it would not matter since the latency period is very long? In any case, it’s all but impossible to find a large group these days which has never owned a cell phone. So the study compared long term cell phone users to those who either had recently acquired a cell phone, never owned a cell phone or had been very light user. The study actually looked at the groups using more than one method. It examined it based on the length of the phone ownership, the average usage of the phone, the reported habits etc.
In all cases, no coloration to increases in brain cancer was ever detected.
Again, as with the INTERPHONE, the Danish Cohort made comparison of reliable data on cancer with the unreliable information about exposures cannot produce reliable final result.
And what the hell would you consider to be reliable data?
Brain cancer is a rare disease, somewhat in the range of around 10 cases per 100,000 people. It means that in order to reliably detect the change, which seems to be less than 50% according to flawed INTERPHONE, tens of thousands of the study subjects should be analyzed. This is very expensive but not necessarily productive.
It’s actually not quite that rare. In fact, it’s about twice as common as cited.
But regardless, the fact is that if the probability of brain cancer were increased by using a cell phone, it would be easy to detect if that probability increase were large. In other words, if it increased the risk from, 22 per 100,000 people to 23 per 100,000 people, that would be very hard to find and a massive sample would be needed. On the other hand, if it increased it from 22 per 100,000 people to 100 per 100,000 people, that would be easy to detect and would stand out from the statistical noise in even a modest study.
Therefore, what we can say from these studies, without doubt, is that while it is impossible to rule out the possibility that there is an increased risk, it must be vanishingly small, if it does exist, because otherwise it would have been easily detected.
As shown by the experiences with INTERPHONE and Danish Cohort, large amounts of money (tens of millions of Euros) and ample amounts of time (over 10 years) were used and no reliable answers received.
No, we have reliable answers. They’re just not the ones you want.
In the current situation, with the above presented experience, should the epidemiology be the first kind of studies to use our scarce research resources? Epidemiology is very expensive and takes a very long time to get results. Any flaw in the study design sets us back by ten or more years.
Well I agree in so much as there’s no point in throwing more money at this. We have plenty of data. The jury is not out. The questions have been answered. It’s time to consider spending money on things we don’t know.
Would we be we better off using the available funding for the human studies examining acute effects of cell phone radiation on physiology? This would, of course, include studies of the known molecular events leading to initiation and development of cancer. We still do not know if cell phone radiation triggers any such events in living humans.
We’ve actually done that too.
And as far as molecular events that lead to initiation and development of cancer, those are not observed with microwaves. No mechanism by which that could happen has ever been discovered, despite more than a century of study of RF fields and electromagnetic radiation.
Performing physiological studies on volunteer will provide information whether any known carcinogenic events are triggered by cell phone radiation. Depending on the result, we could act immediately by imposing preventive measures based on scientific evidence.
Yes, we have done that. We’ve done it on humans. We’ve done it on animals. We’ve done it on live tissue cultures. We’ve done it on chemical systems that mimic what goes on in cells.
To provide such information, epidemiology will still need tens of years before it is able to perform effective studies, assuming that studies will be designed without any major flaws. Volunteer studies examining physiology and pro-carcinogenetic events would provide information much faster.
It’s been done. At some point it becomes time to give up on the existence of something which has been studied for so long and has not been determined to exist.
In this time of scarce resources, we need to make choices how to obtain, most reliably and expeditiously, information about the possible effect of cell phone radiation on brain cancer.
Based on the experience of the last 10-15 years, epidemiology does not seem to be the method of choice.
Well, compared to an assclown with an ax to grind and a desire to be in the newspaper, it actually does pretty well.
This entry was posted on Friday, January 20th, 2012 at 9:15 pm and is filed under Bad Science, History, media, Misc, Quackery. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
View blog reactions