How to Interpret Scientific Findings In the Cell Phone Radiation Controversy
By Rong Wang, PhD | November 13, 2013 | Categories: Cell Phone Radiation Protection | Wireless Energy
As cell phones and other wireless devices become increasingly popular, there is a growing concern over the possible health impact of wireless technology. Science is still inconclusive on whether cell phone radiation is safe or harmful to humans. Consumers are often confused by conflicting study results and mixed media messages. This article is intended to explain the primary causes of the scientific dilemma and offer some suggestions on how to interpret scientific findings in the field of cell phone radiation and human health.
Cell phone radiation is a form of electromagnetic radiation (EMR) in the radio-frequency (RF) range. Since the inception of commercial cell phones in the 1970s, thousands of studies have been conducted on the biological and health effects of RF radiation, including molecular and cellular studies, animal studies and human studies. Human studies can be performed in various ways. Laboratory human experiments are used to investigate the acute or short-term effects of RF radiation on the human body and population-based studies (epidemiology) are used to examine the relationship between cell phone use over a period of time and certain health outcomes.
In the past decade, several large epidemiological studies have been carried out around the world to examine possible links between cell phone use and brain cancer. Results from those studies have been mixed and sometimes contradictory. However, following a comprehensive review of the existing scientific evidence, the World Health Organization (WHO) classified cell phone radiation as “possibly carcinogenic to humans” in 2011. The WHO’s classification illustrates our current state of knowledge on this issue and calls for more research. The controversy is likely to continue for many years or even decades to come.
In the midst of uncertainty, there are a few things that consumers should be aware of when interpreting the existing and upcoming scientific findings and media messages related to this topic.
Study Biases and Methodology Limitations Impair Study Quality
There are numerous sources of biases  involved in epidemiological studies. In one type of study, called a case-control study, prior exposure to cell phone radiation is compared between people with and without brain tumors. Participants are asked to report their cell phone use in the past for an exposure assessment, which can lead to recall bias. In addition, cell phone use habits seem to affect people’s likelihood to participate in a study, which can lead to participation or selection bias.
In another type of study, called a cohort study, a group of healthy people with different exposure status (e.g. cell phone users vs. non-users) was followed over time to compare their brain tumor incidence. The prospective cohort study minimizes recall bias but comes with other problems. For example, the Danish Cohort Study, the only cohort study targeting cell phone use and brain cancer risk to date, has a serious misclassification problem (information bias). Among the total of 420,000 cell phone subscribers under study, about 200,000 corporate cell phone users were excluded from the “user” group and were classified into the “non-user” group while in reality they could have been among the heaviest users. Besides, as cell phones and other wireless technologies become more and more universal, it is increasingly difficult to find a truly unexposed control/reference group for risk comparison.
Another common methodological limitation of epidemiological studies is the inaccurate assessment of exposure to cell phone radiation. In the retrospective case-control study, the self-reported cell phone use information can often be inaccurate because it is very difficult for the participants to remember their cell phone use from a long time ago, especially if they were suffering from a brain tumor. Even in the prospective cohort study, an accurate assessment of exposure can still be difficult. For example, the Danish Cohort Study mentioned above used the number of years of cellular subscription instead of actual mobile phone use for exposure assessment. This meant that a person who used a cell phone for five minutes a week was considered to have the same exposure level as a person who spent five hours per day on a cell phone only because they had the same subscription period. Other factors that can further complicate an exposure assessment include different cell phone models (different amount of RF emission), user environment (rural users typically experience greater exposure from their cell phones than urban users), use scenario (e.g. calls made with or without a headset) and other sources of RF exposures (such as cordless phones).
Furthermore, for a rare disease like brain cancer that affects about 20 in every 100,000 people, a large sample size is necessary to produce meaningful and reliable statistical analysis. Studies involving a relatively small number of people are limited in their ability to detect small increases in risk and the results are less reliable.
Funding Source and Author Affiliation Influence Study Outcome
In an ideal scientific world, one would expect all studies to be performed with perfect objectivity. However, recent systematic reviews of the influence of financial interests in medical research concluded that there is a strong association between industry sponsorship and pro-industry conclusions. Unfortunately, the same phenomenon has also been shown to be true in the field of cell phone radiation and human health.
In 2006, Dr. Henry Lai, a research professor in the bioengineering department at the University of Washington did an analysis of 326 existing studies on possible biological effects of RF radiation published between 1990 and 2006, and where their funding came from. He found that about 50% of the studies showed a biological effect and 50% did not. But when he filtered the studies into two groups – those funded by the wireless industry and those funded independently – Lai discovered that industry-funded studies were 30% likely to find an effect, as opposed to 70% of the independent studies. A 2007 systematic review of 59 experimental human studies found a similar phenomenon – studies funded exclusively by industry reported the largest number of outcomes, but were 90% less likely (odds ratio 0.11) to report an effect or a link than studies funded by public agencies or charities. A 2010 analysis concerning the same topic showed that the funding source and author affiliation significantly affect whether or not a study shows a correlation between cell phone use and cancer.
To add to the confusion, when a study generates multi-fold findings, its general conclusion or press release may not objectively reflect all aspects of its findings. When citing the findings of a study, different media may use different headlines and emphasize different aspects of the findings. This can be misleading for general public who often rely on media messages to gain understanding.
Therefore, consumers should be aware of the impact of financial interests on science and public issues, and take the funding source and authorship into account when interpreting scientific findings and media messages related to cell phone radiation and human health.
Long Latency Impedes Scientific Conclusion
Further complicating the epidemiological evidence is the long latency period between the exposure to carcinogens and the clinical diagnosis of cancer. The recent WHO official classification of outdoor air pollution as a leading cause of cancer (carcinogenic) to humans helps illustrate the issue of long latency. When asked why it had taken so long to reach the conclusion, IARC director Dr. Christopher Wild said that one problem was the time lag between exposure to polluted air and the onset of cancer and “often we’re looking at two, three or four decades once exposure is introduced before there is sufficient impact on the burden of cancer in the population to be able to study this type of question.” In additional to the long latency between exposure and the diagnosis of cancer, it also takes time for science to gather information, perform analysis, resolve controversy, and finally reach consensus. In the case of cigarettes, it took more than 100 years to definitively link cigarette smoking to lung cancer. Before that, numerous studies had actually concluded that there was no link between cigarettes and cancer.
Cell phones have only become prevalent in the past 15 years. Most of the existing studies have covered only a few years, with very limited cases covering more than 10 years. Therefore, one should not expect to have sufficient evidence to support a link between cell phone radiation and brain cancer, even if it does exist. The controversy is likely to continue for many years or even decades to come. While waiting for the final answer, consumers should not ignore the early evidence of risk  and the lessons that history teaches us when trying to understand the health impact of cell phone radiation.
Not all scientific studies should be treated equally in terms of quality and reliability. They can be influenced by multiple factors such as study biases, methodology limitations, funding sources, author affiliations, and latency periods. Consumers should consider all of these factors when evaluating the current body of scientific findings related to cell phone radiation.
 Epidemiology is the study (or the science of the study) of the patterns, causes, and effects of health and disease conditions in defined populations. It is the cornerstone of public health, and informs policy decisions and evidence-based medicine by identifying risk factors for disease and targets for preventive medicine.
 Biases in research is any factors that produce systematic variations or errors in research findings.
 For example, the 13-country case-control Interphone study has a high refusal rate of the control (41%) – it is assumed that healthy (non-cancerous) people who did not use a cell phone were less likely to participate in the study compared to healthy people who used a cell phone and the end result of the selection bias is an underestimate of risk (Interphone study, 2010)
 Lönn 2004 Output power levels from mobile phones in different geographical areas; implications for exposure assessment
 For example, the authors of the children-focused CEFALO study acknowledged that “our study also has limitations…because childhood brain tumors are rare, we could eventually include only 352 case patients and about two control subjects for each patient. Thus, the statistical power of the study to detect small risk increases was limited.” (The CEFALO report, 2011)
 Bekelman et al. 2003; Yaphe et al. 2001
 For example, in the CEFALO study, both the results and conclusion of the abstract are contradicted by the reported results. The general conclusion in the press release that “children and adolescents who use mobile phones are not at a statistically significant increased risk of brain cancer compared to their peers,” is a misrepresentation of the study’s actual finding of increased risks. The study was partly funded by the mobile phone industry funded Swiss Research Foundation on Mobile communication and the study coordinator Dr. Martin Röösli is a member of its board.
 Cigarettes had been around in the United States in crude form since the early 1600s and became widely popular after the Civil War (1861-1865). By 1944, the American Cancer Society began to warn about possible ill effects of smoking, although it admitted that “no definite evidence exists” linking smoking and lung cancer. In 1964, a report by the Surgeon General’s Advisory Committee on Smoking and Health concluded: “cigarette smoking is causally related to lung cancer in men.” In 1965, Congress passed the Federal Cigarette Labeling and Advertising Act requiring the surgeon general’s warnings on all cigarette packages.
 For example, the large 13-country Interphone study found no overall increased risk of the brain cancer glioma from cell phone use. However, it observed a 40% increase in risk for people with highest exposures. A 2013 study of 790,000 women in the UK found a possible increased risk of acoustic neuroma in women who had used a cell phone for more than 5 years compared to women who never used a cell phone, and the risk of acoustic neuroma increased with increasing duration of cell phone use.