May 04 2018

Be Wary of Dubious Brain Cancer Study

We have yet another example of a scientifically complex study being mangled by the mainstream media, who simply do not have the chops to provide an adequate analysis. The Telegraph reports:

Fresh fears have been raised over the role of mobile phones in brain cancer after new evidence revealed rates of a malignant type of tumour have doubled in the last two decades.

They further report about the study:

They analysed 79,241 malignant brain tumours over 21 years, finding that cases of GBM in England have increased from around 1,250 a year in 1995 to just under 3,000.

This sounds alarming. The mainstream reporting was not horrible (unlike some of the fearmongering by advocacy groups) but was completely inadequate to really put this study into context. The Science Media Center put together analysis from various experts, and the entire page is worth a read. It is a good demonstration of how to really analyze a scientific paper.

The Telegraph article does point out that this new study is looking at brain cancer incidence only, and did not present any data that correlates the risk of brain cancer with any specific risk factor. The paper only speculates about possible causes, including the rise in cell phone use. But that is just scratching the surface.

The paper itself is of dubious quality. The journal is an open-access journal with a history of poor peer review. If these findings were really robust, one would expect to see the paper published in a high quality journal. So we essentially cannot count on the quality of the peer review for this study. This is important because none of the authors have an active academic appointment. The lead author, Alasdair Philips, is an electrical engineer with a history of anti-EMF activism. He is an advocate with an ax to grind, not a disinterested academic.

Perhaps this explains some of the curious aspects of the study itself. Prof David Coggon commented:

Their report is unconventional in its format and terminology, making it difficult to follow in places, but the finding on brain tumour incidence is in line with an earlier study in the United States.

How about the data itself. There is some indication of cherry picking going on. First, the choice to look at the time period of 1995 to 2015 does not appear random. That is the exact time period over which cell phone use increased dramatically, from 5% to 95% of the UK public. But, as several of the expert reviewers pointed out, the trend that the authors describe actually started in 1971 when registries of brain tumors were first kept. So if you have a trend that is continuing unchanged for 40 years, it does not make sense to ascribe as a cause for that trend something that was introduced half-way through. But if you want to point a finger at cell phones, just present the data for the second half of the trend.

The same data also shows that some other types of brain cancers were decreasing over the same period of time, partially offsetting the rise of GBM. This suggests the possibility of changing diagnostic patterns, not a real biological change. It is possible that better diagnosis, the introduction of CT scan and MRI scans, and changing diagnostic patterns is entirely responsible for this trend over the last 40 years. The authors conclusions that their data points to an environmental cause is not warranted.

It is also worth noting that the rise in GBM was mostly in the oldest age group, which is not the group with the most cell phone use.

The Telegraph does point out another limitation of the data, but does not really go into sufficient detail – the study did not even try to correlate brain cancer with any environmental factor. That is fine in itself – the study is what it is, and no study needs to do everything. But then why even speculate about cause when the data did not address cause?

In order to address possible causes much more analysis would need to be done. Does the rise in GBM correlate with actual cell phone use, or show an age pattern consistent with use? Does it occur more on the side of cell phone use than the other side? Is there a dose response effect with amount of cell phone use? And is there a plausible mechanism.

None of this data is presented. Other studies which have looked at this type of data have generally shown that there is no correlation between cell phone use and risk of brain tumor.

So in the end this study does not really add anything new to our existing data. We already knew there was this trend of a slow increas of certain types of rare brain tumors over the last 40 years, probably due to changing diagnostic patterns. But the phenomenon does deserve more study. There may be an actual increase happening, due to radiation exposure from CT scans and X-rays, increased air travel, prior atomic weapons testing, or another cause.

And while the data on cell phone use is so far reassuring, because cell phones are so common it remains important to keep monitoring for any possible health risk. The plausibility of a cancer risk from cell phones is very small, however, because they only produce non-ionizing radiation, which is not strong enough to break chemical bonds and cause mutations. But there may be a subtle effect, and we should continue to monitor for any warning signs.

I don’t think this new study really changes anything, but it does seem to have served what is perhaps its real purpose – to provide press releases implying a risk from cell phones.

No responses yet