Update on Childhood Leukemia Research
The preceding review of expert scientific opinion since the NIEHS Q&A booklet was published in 2002 condensed the panels’ and agencies’ conclusions regarding the many health outcomes that have been the subject of EMF health research. It was evident that, repeatedly, mainstream expert opinion has found no evidence that everyday exposure levels of magnetic fields cause effects on such varied health endpoints as pregnancy outcome (e.g., miscarriage and birth defects), neurodegenerative illnesses (e.g., Alzheimer’s disease), cardiovascular disease, electromagnetic hypersensitivity (EHS, see sidebar titled “Related Topics”), and others. The concerns about the association between childhood leukemia and magnetic fields remains, but a causal role for magnetic fields is cast in significant doubt because of the persistent absence of effects on leukemia development in whole animals, the absence of an explanatory mechanism, and the uncertainties surrounding the epidemiology studies.
As context, the IARC classification of magnetic fields as a Group 2B (possibly carcinogenic to humans) was based to a major degree on two ‘pooled’ analyses of the epidemiology literature published in 2000 that addressed the association of magnetic fields with childhood leukemia. The term, pooled, means that the raw data from a collection of studies were combined as if constituting a single study. One analysis was conducted in the U.S. and the other in Europe using an overlapping but not identical set of studies, with the two arriving at similar conclusions. These studies reported statistically significant relative risks (RRs) of between 1.7 and 2.0 associated with average residential magnetic fields above 3 to 4 mG (see sidebar on relative risk). In 2010, an international group of investigators published a pooled analysis of the studies available since the IARC report. The updated pooled analysis reported a comparatively weaker association, relative risk of 1.44, that was not statistically significant. Although consistent with the earlier pooled studies the investigators concluded that, “[o]verall, the association is weaker in the most recently conducted studies, but these studies are small and lack methodological improvements needed to resolve scientific uncertainties regarding the apparent association. We conclude that recent studies on magnetic fields and childhood leukaemia do not alter the previous assessment that magnetic fields are possibly carcinogenic.”
During this period, several studies reported the association of childhood leukemia with distance from overhead high voltage transmission lines. A study conducted in the UK of childhood cancer from 1962 to 1995 published in 2005 reported that although childhood leukemia was associated with close proximity to the transmission lines (within about 650 feet), the associations remained with a weaker though statistically significant relative risk at distances at which the magnetic fields from the lines are negligible (about 650 to 2,000 feet). Other cancers, including brain cancer, bore no relationship to distance from overhead transmission lines.
A follow-up study in the UK published in 2014 extended the period of observation to 2008, reporting that the childhood leukemia risk associated with proximity to overhead lines, though evident in the 1960s and 1970s, disappeared in subsequent decades. The fact that magnetic fields from the lines were a constant presence in residences located near the lines’ corridors throughout the five-decade period, but the occurrence of leukemia in those residences diminished to background levels over the five-decade period, provided strong evidence that some other unknown factor aside from magnetic fields had played a role in the association with elevated risks of childhood leukemia in the earlier periods. Two other studies of the risk of childhood leukemia versus distance to transmission lines were conducted in France (2013) and in Denmark (2014) with inconclusive results. A study of childhood leukemia (nearly 6,000 cases) and distance to overhead transmission lines across California found associations for distance and magnetic fields (greater than 4 mG) consistent with, but weaker than previous studies, despite its large size and rigorous methodologies. More recently, a pooled analysis of proximity to overhead lines published found increased risk of childhood leukemia within a subset of those living closest to the highest voltage lines (within 164 feet).
Based on findings from several pooled analyses and recent studies, such as the California study, a 2019 meta- analysis of magnetic fields and childhood leukemia studies assessed the hypothesis of a decline in the association over time. The findings suggest that the associations were stronger during early time periods (late 1980s), and that risk of childhood leukemia in relation to magnetic fields has weakened in more recent studies (2000s).
The childhood leukemia studies summarized thus far addressed the question: Is the risk of an initial diagnosis of childhood leukemia associated with exposure to residential magnetic fields? In 2006 and 2007 two studies looked at a different question: After the initial diagnosis and treatment is the magnetic field in a child’s residence associated with that child remaining disease-free? A U.S. study published in 2006, and a German study published in 2007 each suggested that survival was poorer in children living in residences with higher magnetic fields, but both studies had small sample sizes limiting one’s ability to draw firm conclusions. To overcome this problem, investigators from eight countries pooled all of the available data from over 3,000 children to assess whether either the risk of relapse or overall survival was associated with residential magnetic fields. The results of the pooled analysis were published in 2012, concluding: “In this large pooled analysis of more than 3000 children diagnosed with ALL in eight countries, no statistically significant associations were observed between exposure to ELF–MF and event-free survival or overall survival of ALL. These results provide no evidence that ELF–MF has a role in predicting outcome of childhood ALL.” This case serves to emphasize a point made earlier that it is premature to draw conclusions that rely on a small set of early studies with inadequate numbers of subjects.