To the Editor:
Elliott et al1 conclude that their results “do not support an epidemiologic association of adult cancers” with high-voltage overhead power lines. Important shortcomings in their methods have hidden an underlying association.
Omission of 132 kV lines (typical magnetic fields >150 nT within 50 m) from the study causes substantial exposure misclassification, even with a study population within 1000 m of National Grid lines. We were members of the UK government advisory group Stakeholder Advisory Group on ELF EMFs (SAGE) that found2 there are four times more homes close to 132 kV lines than to 275 and 400 kV lines combined. Even if mainly nondifferential, the misclassification will, under reasonable assumptions, have a small bias that slightly understates the strength of association. The reduction in numbers of exposed cases and controls will materially diminish statistical power and understate the significance of any underlying associations.
Choice of a 50 m cut point further diminishes statistical power and obscures the association. Kroll et al3 found that 91% of addresses with estimated fields >100 nT are located within 99 m of a high-voltage power line. Typical 50 Hz magnetic field exposures from other sources lie between 10 and 100 nT. This makes 100 m a more suitable cut point, with better statistical power.
The authors’ adjustments to the data have a material effect but are unconvincing with respect to their unpublished area-level socioeconomic assumptions. The selection of four cancers is inappropriate in a comparison with other cancers. Leukemia and brain cancer stand out as a priori candidates4,5 with magnetic field associations with (express or implied) aggregated P values around 10−7. This is comparable with the 5-sigma level accepted as a “discovery” in the field of particle physics.
Using the same reference populations1 (600–1000 m), we grouped people living 0–99 m from a National Grid power line at the time of their cancer diagnosis. The adjusted data were grouped by inverse-variance weighting. Results are summarized in the Table.
Despite their conclusions to the contrary, the proximity data reported by Elliott et al1 (as well as the persistent prior magnetic field associations) in fact add some substantial support to the hypothesis that high-voltage power lines may increase adult cancer risks.
Sutton, Ely, United Kingdom
Professor Emeritus, University of Sunderland
Sunderland, United Kingdom
Professor Emeritus, Atmospheric Chemistry Group
University of Bristol Bristol, United Kingdom
Sutton, Ely, United Kingdom
1. Elliott P, Shaddick G, Douglass M, de Hoogh K, Briggs DJ, Toledano MB. Adult cancers near high-voltage overhead power lines. Epidemiology. 2013;24:184–190
2. UK Department of Health. . SAGE First Interim Assessment Supporting Papers 2007:page 63 Available at: http://www.emfs.info/Related+Issues/SAGE/SAGE+downloads.htm
. Accessed March 16, 2013.
3. Kroll ME, Swanson J, Vincent TJ, Draper GJ. Childhood cancer and magnetic fields from high-voltage power lines in England and Wales: a case-control study. Br J Cancer. 2010;103:1122–1127
4. O’Carroll MJ, Henshaw DL. Aggregating epidemiological evidence: comparing two seminal EMF reviews. Risk Analysis. 2008;28:225–234
5. Kheifets L, Monroe J, Vergara X, Mezei G, Afifi AA. Occupational electromagnetic fields and leukemia and brain cancer: an update to two meta-analyses. J Occup Environ Med. 2008;50:677–688