Design flaws make INTERPHONE non-informative
On May 18, 2010 was published in the International Journal of Epidemiology the summary result of INTERPHONE project the effect of mobile phone radiation on development of brain cancer (http://ije.oxfordjournals.org/cgi/content/abstract/dyq079). Reading this article has confirmed my earlier concerns about the INTERPHONE, that I have presented, among others, in my earlier blog (How reliable is the epidemiological evidence on mobile phones and cancer?). For already some time I have suggested that INTERPHONE will not be able to give reliable answers whether there is, or is not, any link between use of mobile phones and cancer. This latest publication confirms my earlier concerns.
I could briefly summarize this latest INTERPHONE study as non-informative. This is in spite of INTERPHONE being the largest and the most expensive to date executed epidemiological study on mobile phones and cancer. At its conception, INTERPHONE was expected to provide reliable (even final) answers whether mobile phone radiation can cause brain cancer. However, it seems that flaws in the design of the study, which in fact were known from its beginning, have caused that the outcome of INTERPHONE is scientifically unreliable.
The major flaw in the study is the lack of information on how much people were exposed to mobile phone radiation. In case of INTERPHONE, when it was set up in 2000, network operators were not providing information about the use of mobile phones by the study subjects because the networks considered this information as trade secret. Therefore, researchers had to rely on peoples’ memory. But, honestly, who remembers how much has used mobile phone even year, not mentioning 10 years, ago and on which side of the head? And such information was requested from the study subjects. It means, at least in my opinion, that the half of the INTERPHONE information – the dosimetry – about study subjects’ exposure was extremely poor, to say the least (the second half is the pathological information about the cancers).
Further complication is the long latency period for brain tumors which can be somewhere between 10 – 20 years. At the time when the study was launched, in year 2000, there were very few people who used mobile phone for more that 10 years. That is why the group of the longest users (over 10 years) in the study is so small, what causes that its statistical evaluation is so unreliable.
Some of the other methodological problems with the INTERPHONE are listed in my earlier blog (How reliable is the epidemiological evidence on mobile phones and cancer?).
The INTERPHONE results concerning mobile phone use for less than 10 years are suggesting, according to the authors’ conclusions, that there is no effect of mobile phone radiation on brain cancer. However, such conclusions are premature and unreliable because the results presented in the article suggest some kind of “protective effect“. INTERPHONE scientists dismiss such result as unlikely and suggest that it was caused by the insufficient number of responding study subjects. Well, it might be so. But, this flaw should not be used as a proof that there is no effect. Furthermore, the potential possibility of existence of some kind of “protective effect” might not be so far fetched; especially that similar “protective effect” was seen in some animal studies. What this result means to me is that it is unreliable and it does not provide assurances that the use of mobile phone for less than 10 years will not cause cancer.
The results concerning possible increase of brain cancer in people who used mobile phones for longer than 10 years are also unreliable because this group of study subjects was relatively small and because of it the statistical analysis is mathematically uncertain. So, again, the reliable information is missing.
And there is still one issue that was not addressed by the INTERPHONE. Hypothetically, assuming that over 10 years of use of mobile phones increases risk of developing brain cancer, is it so that when one uses mobile phone for only 10 years and then stops, that then the cancer will not develop? Or, is it so, that already then the cancer development cascade has begun and stopping the use of mobile phone will not affect it? And, of course then comes to mind another question: for how many years one would need to use mobile phone to start such cascade of events leading to the development of cancer – for 10 years or 5 or 1 only? Such questions should be addressed and answered in the future epidemiological study too.
In summary, the cause of such inconclusive result of INTERPHONE is, in my opinion, dosimetry. In INTERPHONE was being compared a very good and reliable pathological information about the cancer with very poor and flawed dosimetrical information about the exposure. Such combination of “something reliable” with “something flawed” can only produce one outcome – the scientifically unreliable and non-informative result.
What it all means is that after 10 years of research and millions of euros used for it we are still in the starting point and do not have the answer whether, or whether not, mobile phone radiation could cause brain cancer.
In my opinion, anyone claiming based on the just published INTERPHONE study that we have the answer concerning brain cancer and mobile phone radiation, should read it again.
However, because the result of the INTERPHONE presented in International Journal of Epidemiology is so inconclusive and non-informative, anyone can “pick-and-chose” suitable fragments and get “suiting conclusions”. In short, put a spin on the INTERPHONE story that is the most confusing to decision makers and to the general public.
Some examples are below:
Reuters has quoted IARC Director Dr Christopher Wild as saying “the results really don’t allow us to conclude that there is any risk associated with mobile phone use, but… it is also premature to say that there is no risk associated with it.”
PR Newswire has quoted the IARC director Dr Christopher Wild, as saying: “An increased risk of brain cancer is not established from the data from INTERPHONE. However, observations at the highest level of cumulative call time and the changing patterns of mobile phone use since the period studied by INTERPHONE, particularly in young people, mean that further investigation of mobile phone use and brain cancer risk is merited.”
It is confusing to think what it means. But at least to me, the Reuters report indicates that IARC Director is uncertain whether any valid conclusions can be drawn from INTERPHONE at all whereas PR Newswire indicates that the same IARC Director is assuring that INTERPHONE has not shown any brain cancer risk.
Also, when reading headlines, “head can spin” and one can agree with the Science News headline: Interphone’s data on cell phones and cancer: The spin begins.
In the examples of media reports, listed below, one can find everything from yes to no to may be so. One can wonder whether the writers of these stories have all read the same article form the International Journal of Epidemiology (just being sarcastic…):
Yes: Half an hour of mobile use a day ‘increases brain cancer risk’ Telegraph.co.uk; WHO study says mobile use ‘can raise brain cancer risk’ The Hindu; Brain cancer link to mobile phones Herald Sun; Ten-year worldwide study links mobile phone use to cancer Daily Mail; Heavy mobile users risk cancer Times Online; Landmark study set to show potential dangers of heavy mobile phone use Telegraph.co.uk; Study links mobile phone use to brain tumours Scotsman;
No: ‘No proof’ of mobile phone cancer link Mirror.co.uk; Study finds no link in cell phone use, brain tumors CNN International; Study finds no brain cancer link to mobile phone use AFP; Study fails to link mobile phones, cancer ABC Science Online; Experts deny mobile phones cause tumours Sydney Morning Herald;
May be so: Cell phone, cancer link inconclusive Detroit Free Press; Cancer from mobile use ‘uncertain’ The Press Association; Massive study can’t say whether cellphone use causes brain cancer Vancouver Sun; Much mobile phone use causes brain cancer? Xinhua; Cellphone cancer study inconclusive; researcher urges more study Washington Post (blog); WHO study on mobile phone cancer risk ‘inconclusive’ BBC News; WHO study has no clear answer on phones and cancer Reuters; Mobile Phone Use and Brain Cancer Risk: Inconclusive Results from … Science Daily;
And my favorite title comes from MSNBC: No answer, just fuzz, from cell phone study. Unfortunately, this seems to be pretty precise description of the outcome of the INTERPHONE.
MY FINAL CONCLUSION
We still do not know whether mobile phone use could cause brain cancer or any other disease. Because the use of mobile phones has become essential part of our life we can not and we should not discourage people from using them. However, it is prudent to advice caution and urge to reduce exposures to mobile phone radiation whenever possible or feasible, especially in children.