Bias affects cell phone cancer risk findings

October 15, 2009 by Lin Edwards, Medical Xpress weblog
Bias affects cell phone cancer risk findings

( -- A group of South Korean and American researchers has found studies of possible links between cell phones and brain tumors and other cancers vary in quality, and those suggesting there is little or no risk may have some bias.

The scientists were from the National Cancer Center in Goyang, South Korea, the Seoul National University Hospital, Ewha Womans University and the University of California in Berkeley. The team carried out a meta-analysis of 23 published epidemiological studies, covering a total of 37,916 people who had been investigated to determine if there was a link between cell phones and cancer.

The findings of the analysis, published in the Journal of Clinical Oncology, were that the results of the studies varied widely, depending on who carried out and funded the research and what controls they had in place for bias and errors.

The researchers, led by Dr. Seung-Kwon Myung, divided the studies into "high quality" if the research blinded subjects of the study to bias (because the researchers did not know which subjects had tumors when they did their interviews), and "low quality" if the research did not counter bias, or if it was funded by the .

They discovered that eight high quality studies found a 10-30% increase in tumors in people who regularly used cell phones over those who rarely or never used them. The same studies found an increased risk of benign (non-malignant) tumors in people who had been using the phones for ten years or more.

These high quality studies were funded by the Orebro University Hospital Cancer Fund in Sweden, the Orebro Cancer Fund, and the Swedish Work Environment Fund, but seven of the eight studies were carried out by Dr. Lennart Hardell, a Swedish . This leads to a supposition that the results might be related to the Swedish environment: many Swedes live in rural areas, and the radiation emitted by phones is greater in rural areas where the signal is weaker.

The "low quality" studies, which did not correct for bias, found no link, or found that people who used mobile phones faced lower risks of tumors than people who rarely or never used the phones. Myung's team said some of these studies were funded by the Global System for Mobile Communication Association and the Mobile Manufacturer's Forum.

The analysis also concluded the studies were not broad enough to prove a link exists between cancers and phone use, and larger cohort studies would be needed to finally settle the question. Until now, the only published cohort study showed no link between tumors and phone use, but this study, from Denmark, used telephone subscription data rather than actual phone use.

The studies being analyzed were all case studies. This means the researchers divided their subjects into two groups: those who had or other cancers and those who did not, and then interviewed the subjects about their use of mobile phones. The researchers concluded that cohort studies are needed, in which groups of phone and non-phone users are followed over a long time period to see which group develops cancer more often.

The use of cell phones and other cordless phones has increased dramatically over the last decade, with an estimated 4.6 billion users today, according to the United Nations International Telecommunication Union. There have been concerns the radiation emitted by the phones may cause health problems, but so far there is no proof, and the debate continues.

More information: Mobile Phone Use and Risk of Tumors: A Meta-Analysis, Seung-Kwon Myung et al., , doi:10.1200/JCO.2008.21.6366

© 2009

Related Stories

Recommended for you

Magnetized wire could be used to detect cancer in people, scientists report

July 16, 2018
A magnetic wire used to snag scarce and hard-to-capture tumor cells could prove to be a swift and effective tactic for early cancer detection, according to a study by researchers at the Stanford University School of Medicine.

Researchers suggest new treatment for rare inherited cancers

July 16, 2018
Studying two rare inherited cancer syndromes, Yale Cancer Center (YCC) scientists have found the cancers are driven by a breakdown in how cells repair their DNA. The discovery, published today in Nature Genetics, suggests ...

Researchers map 'family trees' of acute myeloid leukemia

July 16, 2018
For the first time, a team of international researchers has mapped the family trees of cancer cells in acute myeloid leukaemia (AML) to understand how this blood cancer responds to a new drug, enasidenib. The work also explains ...

Scientists sharpen the edges of cancer chemotherapy with CRISPR

July 13, 2018
Tackling unsolved problems is a cornerstone of scientific research, propelled by the power and promise of new technologies. Indeed, one of the shiniest tools in the biomedical toolkit these days is the genome editing system ...

Products of omega-3 fatty acid metabolism may have anticancer effects, study shows

July 13, 2018
A class of molecules formed when the body metabolizes omega-3 fatty acids could inhibit cancer's growth and spread, University of Illinois researchers report in a new study in mice. The molecules, called endocannabinoids, ...

Looking at the urine and blood may be best in diagnosing myeloma

July 13, 2018
When it comes to diagnosing a condition in which the plasma cells that normally make antibodies to protect us instead become cancerous, it may be better to look at the urine as well as the serum of our blood for answers, ...


Adjust slider to filter visible comments by rank

Display comments: newest first

1 / 5 (4) Oct 15, 2009
Meta-analysis is like metaphysics - complete garbage. Two different researchers can use the same studies and apply "accepted" rules for meta-analysis and get completely different conclusions. Whenever something stupid is being proposed there is always a meta-analysis behind it. Enough already!
1 / 5 (3) Oct 15, 2009
This is absolute garbage. It seems they classify any study that disagrees with their hypothesis as "low quality" If you look at most meta-analysis the odds ratio fit into a region of roughly .7 to 1.3. If you ignore all of the studies that show a below unity value you are are inherrently shifting the results above unity. This behvior seems to be consistent with other people's work in this field. Dr. Hardell's own meta-analysis openly comments on this perception. He claims that there is no reason to believe that low level microwave radiation should inhibit cancerous growths, while ignoring that there is no reason to believe they would? Compared to the odds ratios for established cancer causing behaviors, which are often well into the double digits, all of the cell phone studies seem to settle around unity with some statistical noise.
1 / 5 (2) Oct 16, 2009
This is garbage squared, or cubed. The industry studies imperiously treated as low quality are the only ones that take the technical issues into account. First, cell phones operate in different frequencies bands depending on where they are used, and whether they use analog or digital signals.

Next, in most of the studies that find a potential (but not statistically significant) correlation, long term cell phone users are the most likely to have tumors--but they are also the most likely to have used high-powered analog phones that are no longer manufactured or sold.

I remember when Motorola came out with the first cell phone to weigh 10 ounces. No one would buy--or use--such a big bulky phone today. Well, actually they might. Remember Iridium? International newspaper correspondents and others still use these phones, which need a much higher powered signal to reach a satellite not a nearby cell tower. But no high brain cancer risk has been associated with those phones.
1 / 5 (2) Oct 16, 2009
eachus - I don't believe the sat phones operate on the same frequencies as cellular phones.

How are they going to do cohort studies since everyone uses cell phones? How about looking at population data pre & post cell phones?

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.