Many people cannot imagine life without a cell phone. Since commercial cell phone networks were first introduced in the 1980s, over half the world’s population now uses them. While becoming a way of life and work, the radio frequency (RF) waves from cell phones have also been depositing energy into users’ brains. With billions of people exposed to more RF energy than ever before, scientists and the public have asked if a lifetime of cell phone use might be a health risk.
Cell phones are carefully designed and tested so that their RF emissions comply with all current health standards. However, the research upon which these safety guidelines are based cannot rule out the possibility of phone RF causing tumors in the head after 10 to 20 years of use. Since many people use cell phones and other RF devices at their work, potential hazards from cell phones are also of interest to the occupational safety and health community.
To answer these questions, an epidemiologic study called INTERPHONE was funded by the European Union and health agencies in 13 countries. From 2000 to 2005, INTERPHONE interviewed 14,000 adults about their cell phone use, other exposures to RF radiation, and other factors conceivably related to brain cancer. With this information, the INTERPHONE epidemiologists have been testing whether indicators of the phone’s RF radiation correlate with increased cancers in the head.
After five years of debate, the INTERPHONE epidemiologists recently published their findings on brain cancer, but their paper raise more questions than it answers. Its conclusions are:
Overall, no increase in risk of [brain cancer] was observed with use of mobile phones. There were suggestions of an increased risk… at the highest exposure levels… However, biases and errors limit the strength of the conclusions we can draw from these analyses and prevent a causal interpretation… The possible effects of long-term heavy use of mobile phones require further investigation.
The journal’s commentary on the INTERPHONE paper likened this conclusion to the Delphic oracle in ancient Greece, whose advice to generals and emperors was famously difficult to decipher. In explaining their oracular conclusions to reporters, INTERPHONE epidemiologists offered several conflicting interpretations of the study results. Emphasizing the conclusion’s first sentence, Swedish researcher Maria Feychting said: “The use of mobile phones for over ten years shows no increased risk of brain tumors.”
Others emphasized the “increased risk” in the conclusion’s second sentence. Dr. Siegal Sadetzki from the Israeli INTERPHONE study said: “What worried me was that, in my study, I saw consistent positive results and they always appeared where there is biological plausibility… [The tumors] appeared in the more than ten years [group], they appeared on the same side where the phone was held, they appeared for the heavy users… So the fact that all of these indications appeared where they should have appeared told me that it was really a red light.”
In rebuttal, the first camp cite the “biases and errors” from the third quoted sentence. “The entire association [seen in INTERPHONE] can be explained by bias,” said Joachim Schz from the German study. “We have identified implausible values of [phone] use in the main data set.”
A middle position was expressed by Jack Siemiatycki of the Canadian INTERPHONE group who called the results “genuinely perplexing, enigmatic and paradoxical.” Siemiatycki said his perception of the phone’s risks have gone “from implausible to something higher.”
When asked to advise cell phone users, the differences of opinion became sharper. Principal investigator Elisabeth Cardis said, “Overall, my opinion is that the results show a real effect… It is too early to make strong recommendations to adults and children concerning the use of phones… But there are ways to limit exposure to the brain from mobile telephones, through the use of [text messaging], speakers or other hands-free devices. This may be a reasonable course of action until stronger conclusions can be drawn about risks from mobile phone use.” But Anthony Swerdlow from the British study disagreed: “This study does not give reason for precautionary measures.”
As the only U.S. citizen who was part of INTERPHONE, I have been reflecting on its contradictory findings. After years of collaboration, why are the INTERPHONE epidemiologists now giving conflicting advice on cell phone use? What can people conclude about the safety of cell phone use?
I was not directly involved in the epidemiologists’ discussions since my role in INTERPHONE was to assess RF exposure in workplaces. However, my research on electromagnetic fields (EMF) and cancer has given me some insight into how environmental health researchers draw conclusions from conflicting epidemiologic findings. From a careful reading of the INTERPHONE paper and the epidemiologists’ public statements, I can offer a few observations:
As a fan of the TV show Law and Order, this INTERPHONE paper reminds me of a missing person case without a body to autopsy. To decide whether this is murder, the detectives can only gather circumstantial evidence. INTERPHONE also lacks a “smoking gun”—measurements of the RF energy absorbed from cell phones in the subjects’ brains. So the epidemiologists are analyzing indirect exposure evidence that can be obtained from interviews and medical records: years of use, time spent on the phone, the tumors’ proximity to the ear, and whether subjects used their phone on the side of the head where the tumors appeared. This approach resulted in calculations of cell phone risks from many different perspectives but little certainty.
When INTERPHONE’s conclusions speak of “biases,” the scientists are not referring to their personal state of mind, but to well-known flaws in epidemiology. When studies collect data on the past behaviors of people in everyday life, these flaws inherent to observational epidemiology can tilt the results one way or the other. With INTERPHONE, this problem was compounded by several factors. First, they recruited subjects without cancer by phone, but their requests were refused by 47% of those called, an indication of the public’s growing resistance to telemarketers. Even worse, controls who refused were less likely to use cell phones. This unrepresentative sample tilted INTERPHONE’s risk estimates downward, often giving the appearance that cell phones protected their users against brain cancer. Some argued that the large number of data points with no cancer risk was reassuring. Others said that many were due to the tilted sample from the phone solicitations. So INTERPHONE re-analyzed the risks without the subjects who never used cell phones and found a smoothly increasing risk with the time spent on the phone. The two camps could not agree on what this re-analysis meant, so it was consigned to an online appendix.
In order to evaluate this complex array of evidence, some epidemiologists based their interpretation solely on INTERPHONE, but others took earlier research on RFs health effects into account. For example, Microwave News reported:
“If there were a stronger effect, we would have seen it.” … [Joachim] Schz said that his outlook on long-term risk is largely based, not on INTERPHONE, but on his … Danish cohort studies and … a more recent analysis of the incidence of brain tumors in the Nordic countries…. Neither points to an increase in brain tumors among the general population.
Dr. Swerdlow cast his net even wider to include animal cancer assays and previous epidemiologic studies. “The balance of evidence from this study, and in the previously existing scientific literature, does not suggest a causal link between mobile phone use and risk of brain tumours,” he concluded.
How do epidemiologists add up all this statistical data with their pluses and minuses to decide whether cell phones cause cancer? Unfortunately, the INTERPHONE epidemiologists and scientists in general have not agreed on how to keep score.
Should risks tainted by possible “biases and errors” be rejected or thoughtfully interpreted? Should the pattern of elevated risks from several “biologically plausible” conditions suggesting RF exposures be seen as a “red light,” as Dr. Sadetzki argued? Or do the study’s complexities make any patterns illusory?
Is the re-analysis in the online appendix a valid method for adjusting for the study’s control selection problems? Should its pattern of a growing risk with increasing phone use be taken as evidence of a cancer risk? Or is this an invalid change of the rules after the cards were dealt?
Should past studies that found no cell phone risks raise the bar for INTERPHONE’s evidence? Or should each study be analyzed on its own merits before comparing its findings with other studies done under different conditions?
The INTERPHONE experience shows that epidemiologists do not have universally-accepted answers to these questions, especially for this complex study with multiple lines of indirect evidence and imperfect data.
Cell Phones and Public Health
After five years of debate over the conclusions, the epidemiologists could only agree that the “possible effects of long-term heavy use of mobile phones require further investigation.” While waiting for these new studies, should people change how they use cell phones? There again, the INTERPHONE investigators disagree.
Prof. Swerdlow concluded no changes are needed. On the other hand, Drs. Cardis and Sadetzki suggest spending less time speaking on the phone and moving the phone away from the head by using text messaging, ear pieces and the speaker phone option. This advice is based on the elevated cancer rates associated with the total time spent talking on phones and with the older “analog” cell phones that direct more RF energy to the brain.
These contradictory interpretations of the INTERPHONE evidence are rooted in two philosophies held by epidemiologists and other public health scientists. One philosophy holds that an agent like RF radiation has no effect unless data clearly shows otherwise. This cautious approach to drawing conclusions is a cornerstone for sciences of all kinds and has been a key to its many successes over the past 300 years.
In opposition is the precautionary principle that endorses action when research findings are suggestive but not conclusive. From a scientific viewpoint, the precautionary approach can be justified for questions of public health because the errors in a study can also obscure risks.
Dr. Sadetzki suggests how these two philosophies can be balanced when she said, “As a scientist, this is not enough [evidence] … for causality, but an indication that, according to my judgment, it is enough in order to advise the precautionary principle.”
Judge for Yourself
In all these reflections on the INTERPHONE paper, the common thread is that its interpretation depends as much on judgment as on the data. Since the experts disagree, each person must now judge whether to reduce his or her RF exposure from cell phones based on these controversial reports of brain cancer.
Here are a few facts that can be helpful in making these judgments:
- The contested INTERPHONE finding is a doubling of brain cancers in people who used cell phones for over a half-hour a day over ten or more years.
- The type of brain cancer that reportedly increased is called “glioma.” It is a relatively rare cancer which strikes 6 to 8 people per 100,000 every year. This incidence is far lower than the 68 cases of lung cancer per 100,000. Thirty-three percent of brain cancer patients are alive five years after diagnosis compared to 15% of the lung cancer patients.
- The radiation received from a cell phone can be greatly reduced by using it away from the head. Texting, the speaker phone option, ear buds and Bluetooth® all lower the radiation received from the phone.
- People absorb far more RF radiation from a cell phone held to the ear than from Wi-Fi, portable phones, wireless computer networks, cell phone towers, and all other wireless devices (outside of a few workplace sources).
- Cell phones and all other wireless consumer devices comply with all safety standards which protect against RF’s proven health hazards.
- The greatest risks from cell phones are their distractions while people are driving cars or doing other potentially hazardous activities.
Other data collected by INTERPHONE may eventually provide more clarity on RF cancer risks. Papers on tumors of the acoustic nerves and salivary glands are scheduled. INTERPHONE investigators have been using the detailed interviews of cell phone usage to estimate the RF energy absorbed at the tumor’s location. By combining all of the collected information into a single measure of RF exposure, this approach may provide more direct evidence on the phone’s brain cancer risk.
In addition, NIOSH has been collaborating with some INTERPHONE investigators to analyze its data on occupational RF exposure in a new study called INTEROCC. Funded by a research grant from the National Institutes of Health, INTEROCC is using the INTERPHONE interviews to test whether brain cancer risks are associated with occupational exposures to RF radiation and chemicals. NIOSH heads the evaluation of the subjects’ exposures to occupational RF and other electromagnetic fields.
Although INTERPHONE and other health research have not yet given a clear answer on the cancer risks from cell phones, the data collected and the supporting studies by engineers, physicists, and biologists have greatly increased our understanding of people’s exposures to RF radiation from these new wireless technologies. If this additional research firmly establishes any health risks from cell phones, NIOSH and other public health agencies will have much of the information needed to make science-based recommendations to protect people’s health.
Dr. Bowman is a certified industrial hygienist in the NIOSH Division of Applied Research and Technology. He thanks Bob Park, Louis Slesin, and Greg Lotz for helpful discussions.
All quotes from the INTERPHONE investigators are from Interphone Resources and U.S. Senate Hearing on Cell Phones and Health: A Recap by Microwave News.