In this issue of EPIDEMIOLOGY, Elliott et al1 report results from a large-scale UK case-control study on power lines and risk of cancer in adults—concluding that their findings “should help to settle a long-standing debate on the safety of residential exposures to extremely-low-frequency magnetic fields from high-voltage overhead power lines and adult cancers.”1 Strengths include the large sample of >50,000 cases and ~80,000 controls as well as the register-based approach with excellent source data for estimating distance of residences to nearby power lines and the extremely-low-frequency magnetic fields around those power lines. Limitations, well discussed by the authors, include omission of exposure sources other than high-voltage power lines and lack of migration information.
The major limitation, however, seems to be the rationale behind this particular study design. Using some types of cancers as cases and other types as controls, in a situation of weak evidence of any risk for any cancer, leads to a situation—expressed here in admittedly slightly provocative terms—where cancers with some concern, but little evidence and weak biologic plausibility of an association, are compared with cancers that have less concern, but with equally little evidence and weak biologic plausibility of an association. The concerns, such as they are, have emerged from general considerations to investigate possible adverse effects of an ubiquitous exposure, isolated findings from small studies, anecdotal reports, or layman concerns about “electrosmog.” For instance, breast cancer has received some attention in the past in relation to the role of extremely-low-frequency magnetic fields in the light of the melatonin hypothesis,2 so this cancer is included in the case group even though recent epidemiologic risk assessments concluded lack of evidence of an association (with breast cancer being the only adult cancer in this category).3 Colon cancer, on the other contrary, has not received attention in this area and so, given the lack of studies and data, this cancer is included in the control group. In fact, all cancers in both case and control groups are in exactly the same category of “not classifiable” in regard to the potential carcinogenicity of extremely-low-frequency magnetic fields (group 3) of the International Agency for Research on Cancer’s Monograph program on the evaluation of carcinogenic risks to humans.4 Thus, the approach adopted may be a risky gamble, because it is unclear what to expect—both under the assumption of a true association for any of the cancers and under the null hypothesis of no association. Indeed, if extremely-low-frequency magnetic fields were a universal carcinogen, the study would not show any association, even if there were a cancer risk.
There is also a substantial risk of false-positive results. First, because of the high number of effect estimates (48 of such shown in Figs. 2 and 3 in the article by Elliot et al),1 deviations from the null could have simply occurred by chance. Second, because it cannot be assumed that living close to power lines identifies a truly random segment of the population, distance to power lines might be a proxy for true risk factors (such as lifestyle or other environmental exposures) that could, for example, effect breast cancer or melanoma risk. Although neither concerns apparently affected the present results, it was impossible to predict this in advance, and an excess of any of the investigated cancers near the power lines would have inevitably created alarm.
The authors are not to blame for responding to political demand. Using available routine data sources, they conducted a well-designed study to address the hypothesis that living close to power lines does not pose a greater risk of getting leukemia, brain cancer, melanoma, or breast cancer compared with any other cancer. If this were indeed the best hypothesis, and perhaps it was, their study should settle the debate. But it needs to be acknowledged that this conclusion involved some luck, with all findings being close to null. The hypothesis highlights the fact that this issue has become a political, rather than scientific, debate. This order needs to be reversed and enforced with words of warning about the possible outcomes of studies conducted because of public, rather than scientific, concern. Indeed, concern-driven research is particularly damaging in times of limited funding, redirecting precious resources away from potentially promising hypotheses in epidemiology and other disciplines. So, although the potential adverse health effects of electromagnetic fields have become a topic of controversy going beyond science itself, they also have had a negative impact on the field of bioelectromagnetic research, causing distraction from the unresolved questions in bioelectromagnetics that merit attention.5
ABOUT THE AUTHOR
JOACHIM SCHÜZ is the head of the Environment and Radiation Section at the International Agency for Research on Cancer (IARC). Additional contacts with the debate on electromagnetic fields come from being past president of the Bioelectromagnetics Society, member of the European Commission’s Scientific Committee on Emerging and Newly Identified Health Risks, and principal investigator of several epidemiologic studies on electromagnetic fields and childhood leukemia, brain tumors, or neurodegenerative disease.
1. Elliot P, Shaddick G, Douglas M, de Hoogh K, Briggs DJ, Toledano MB. Adult cancers near high-voltage overhead power lines. Epidemiology
2. Stevens RG. Electric power use and breast cancer: a hypothesis. Am J Epidemiol. 1987;125:556–561
3. Sienkiewicz Z, Schüz J, Poulsen AH, Cardis EEFHRAN Project. . Risk analysis of human exposure to electromagnetic fields (revised). European Health Risk Assessment Network on Electromagnetic Fields Exposure (EFHRAN). Published October 2012 Available at: http://efhran.polimi.it/docs/D2_Finalversion_oct2012.pdf
. Accessed 26 November 2012
4. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Non-ionizing Radiation, Part 1: Static and Extremely Low-frequency (ELF) Electric and Magnetic Fields. 2002;Vol. 80 Lyon, France IARC Press