As mentioned in the Introduction, the final evaluation of the carcinogenicity of extremely low frequency electric and magnetic fields (ELF EMF) was made following the working procedures and evaluation method of the International Agency for Research on Cancer (IARC), as modified in Appendix A. The final evaluations for non-cancer end-points were made by a similar procedure, with consideration of other data relevant to the evaluation of carcinogenicity and its mechanisms (Section 12 (b) in Appendix A). A brief discussion of the levels of evidence as defined by the IARC is warranted.
The predominant evaluations of the various health end-points covered in this report are 'limited' and 'inadequate' evidence. 'Limited evidence' is not an unusual finding for epidemiological data in the IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. This degree of evidence is generally provided by studies for which there is credible evidence of an association and for which a causal linkage cannot be established with a high degree of certainty. This does not mean the effect is weak, nor does it mean there is clearly an effect, although these issues enter into the evaluation. In most cases, this degree of evidence is associated with one or more of the following problems: questionable identification of the exposure factor(s) associated with the disease outcome (either a dose surrogate was used or individuals were missclassified as to their exposure category), bias may have played a small role in the finding, confounders were not ruled out to the satisfaction of the original investigator and/or the Working Group, the observed effect was small, making clear detection of an effect difficult, and/or there is little information on dose-response in the available report. The careful reader is directed to the sections on human epidemiological findings for a clear description of each study. For experimental animal bioassays, 'limited evidence' for an effect is generally driven by clear findings in only a single study for a single end-point or minor problems with a set of data which otherwise would have been positive; confounding, bias, and exposure missclassification generally should not exist in laboratory studies.
'Inadequate evidence' can imply one of four possibilities: (1) there are insufficient data for making a judgment of any kind (e.g. poor study design, making interpretation impossible); (2) the data suggest a positive effect but, due to limitations in design or very weak findings, cannot be interpreted as suggesting a causal linkage; (3) the data suggest a negative effect but, due to limitations in design or very few findings, cannot be interpreted as suggesting no effect; and (4) the data are contradictory and no clear pattern is discernible. For case (1), given a solid hypothesis, it may be beneficial to continue to study an inadequate finding using a better design in the same experimental system. For case (2), if the effect seen is of public health consequence, it should be studied further but with a clear hypothesis and perhaps in conjunction with other studies such as those providing mechanistic interpretation. In case (3), unless there is a clear scientific reason for further study, again involving a defined hypothesis, there is little need to continue to study the observed effect. Finally, for case (4), the effect might be further studied if the scientific issues are compelling or if health concerns are raised, but it is unlikely that another study of similar design would be performed. Additional studies might not be needed. Again, a careful reader searching for scientific hypotheses for further study should read the more detailed descriptions of the findings presented in the three preceding chapters.
The final evaluations and a brief description of the evidence supporting these evaluations are given below.
The Working Group concluded that ELF EMF are possibly carcinogenic to humans (Group 2B, Appendix B). This evaluation was supported by 19 members of the Working Group; 8 members considered that the evidence fell into Group 3 ('ELF EMF are not classifiable as to their carcinogenicity to humans'), 1 member considered that the evidence fell into Group 4 ('ELF EMF are probably not carcinogenic to humans'), 1 member abstained from the vote, and 0 members were absent and did not vote. 2 members of the Working Group who were absent for the final vote left clear instructions as to what their vote would be, and these are recorded in the counts given above.
The majority (20 out of 26) of the Working Group members who voted concluded there is limited evidence that residential exposure to ELF magnetic fields is carcinogenic to children on the basis of the results of studies of childhood leukemia; the remaining 6 voting members concluded that there was inadequate evidence. Three lines of evidence supported the overall finding: the association between exposure to calculated magnetic fields and risk for childhood leukemia, the association between exposure to measured 24-h magnetic fields and risk for childhood leukemia, and continued concern about the association between wire codes and risk for childhood leukemia. There was inadequate evidence from spot measurements of magnetic fields in homes to support this finding.
The majority (14 out of 25) of the Working Group members who voted considered that there is limited evidence that occupational exposure to ELF magnetic fields is carcinogenic to humans on the basis of results of studies of chronic lymphocytic leukemia (CLL); the remaining 11 voting members concluded that there is inadequate evidence. The association between exposure to magnetic fields and risk for CLL was considered in three studies of its incidence (two in Sweden and one in Canada and France) and in one study of mortality from this disease (in the USA). No association was found in the US study, which was based on diagnoses derived from death certificates. In the Canada-France study, a nonsignificantly increased risk was seen overall and in two of the three cohorts. Differences in definition and in follow-up of the three cohorts, however, limit the interpretation of these results. A significant increase in risk was seen in both of the Swedish studies (Feychting et al., 1997; Floderus et al., 1993). Although the study of Feychting et al. provides unique information on the potential importance of combined occupational and residential exposure in adults, it suffers from small numbers and weakness in exposure assessment, particularly for women. In the study of Floderus et al., the risk increased with increasing exposure; it was particularly strong for the highest exposure category and increased somewhat when adjusted for exposure to potential confounders. The refusal rate in that study, however, could have introduced a bias in the results.
The majority (22 out of 25) of the Working Group members who voted concluded that there is inadequate evidence for an association between occupational exposure to ELF EMF and the risk for other cancers; 2 voting members concluded that there is limited evidence and one that there is evidence for lack of an effect. The cancers considered were acute myelogenous leukemia (4 studies), brain cancer (5 studies), male breast cancer (10 studies), and female breast cancer (4 studies). Overall, the inadequacy of the findings is due to limitations in study design, inconsistency in findings across studies, and/or a lack of association.
The majority (24 out of 25) of the Working Group members who voted concluded that there is inadequate evidence that residential exposure to ELF magnetic fields is carcinogenic to adults; the other voting member concluded there is evidence suggesting lack of effect. The cancers considered in this evaluation included leukemias (2 studies of use of appliances, 3 studies of distance from power lines, 2 studies of measured fields, and one study of calculated fields), breast cancer (4 studies), and cancers of the nervous system (3 studies). Overall, the inadequacy of the findings is due to limitations in study design, inconsistency in findings across studies, and/or a lack of association. In general, none of the associations between residential exposure to magnetic fields and the risk for cancer was by itself convincingly positive; however, the quality of the exposure assessment is a serious limitation in all of these studies.
There is inadequate evidence with respect to childhood nervous system tumors. This conclusion was supported by 25 Working Group members; there were 2 abstentions and 2 absent.
There is inadequate evidence with respect to childhood lymphoma. This conclusion was supported by 25 Working Group members; there were 2 abstentions and 2 absent.
The majority (19 out of 27) of the Working Group members who voted concluded that there is inadequate evidence in experimental animals for the carcinogenicity of exposure to ELF EMF; the other 8 members concluded that there was evidence for a lack of effect. The overall conclusion of the Working Group was that most of the studies suggest a lack of carcinogenicity, and the few that gave results of borderline positivity are inadequate to conclude that exposure to magnetic fields at the magnitude and configurations at which they were investigated increases the incidence of cancer in rodents. Two long-term bioassays showed no carcinogenic response, but one showed an equivocal response at one tumor site in animals of one sex of one species. Within the limits of the experimental model of multistage mammary carcinogenesis, the results of the ensemble of experiments did not provide convincing evidence for a promoting effect of EMF on chemically induced mammary cancer. In another commonly investigated model, skin carcinogenesis, exposure to magnetic fields had no effect. EMF did not induce leukemia or lymphoma in mice or rats in several studies.
When mechanistic data are available to support reported toxicological results, the IARC guidelines allow that support to be categorized as 'weak', 'moderate', or 'strong'. All (27) of the Working Group members who voted concluded that a limited number of well-performed studies provide moderate evidence for mechanistically plausible effects of ELF EMF in vitro at intensities greater than 100 µT on end-points generally regarded as reflecting the action of toxic agents. All (26) of the Working Group members who voted concluded that the in-vitro evidence and information on physical mechanisms provide weak support for an effect of fields of intensities below approximately 100 µT.
A series of recent (1996-98) studies has shown that field exposure induces gene mutations. ELF fields at flux densities below 100 µT have consistently been shown to have no effect on mutation rates, but fields of 200-400,000 µT reproducibly and significantly enhance mutation rates after initiation with either X-rays or gamma-rays. Moreover, exposure to 400,000 µT has been shown to increase the number of mutations in the absence of ionizing radiation in two human cell lines. Thus, multiple, self-consistent reports demonstrate a dose-dependent effect on a process or end-point commonly considered to be associated with carcinogenesis. Importantly, the flux densities used in all of these studies (> 100 µT) are within the range of a single physical transduction mechanism, specifically magneto-chemical transduction. The potential for magneto-chemical effects at flux densities above 100 µT has been firmly established in both theoretical analyses and biochemical investigations.
Numerous well-performed studies have also shown strong effects on other end-points commonly associated with carcinogenesis, including significantly increased cell proliferation, disruption of signal transduction pathways, and inhibition of differentiation, all of these effects occurring at field intensities greater than 100 µT. Like the studies of gene mutations, these investigations were performed at sufficiently high intensities that magneto-chemical transduction is a plausible mechanism of field-cell transduction, although this does not preclude other mechanisms of interaction.
There is little doubt that the evidence in support of the decision to classify ELF EMF into Group 2B is driven by the results of studies on childhood leukemia in residential environments and on CLL in adults in occupational settings. The fact that limited evidence was seen for CLL in adults should not be construed as providing support for the finding with regard to leukemia in children, however: childhood leukemia and adult CLL are very different diseases with different etiologies. Also, the inadequacy of the evidence for an effect on the risk for CLL in adults in the studies of residential exposure neither supports nor refutes the findings in the studies of occupational exposure. The in-vitro and mechanistic data provide, at best, marginal support for the conclusion that ELF EMF are possibly carcinogenic to humans. While ELF magnetic fields at intensities greater than 100 µT provide moderate support for effects in vitro, there was little evidence of effects at intensities below this limit, which cover most of the range of exposure in the studies of residential childhood exposure and adult occupational exposure. Relatively few of the studies of occupational exposure addressed exposure to electric fields. Finally, the inadequate evidence from long-term bioassays for carcinogenicity in rodents is driven more by lingering concerns about single findings in two separate studies than by an overall concern that something has been missed in these studies or that there is a trend toward a positive effect in poorly conducted studies.
None of the evidence for adverse health effects seen after exposure to ELF EMF achieved a degree of evidence exceeding 'inadequate' (for humans) or 'weak' (for experimental animals). The end-points evaluated in humans were adverse birth outcomes after maternal exposure, adverse reproductive effects after paternal exposure, Alzheimer disease, amyotrophic lateral sclerosis and other motor neuron diseases, suicide and depression, and cardiovascular disease.
There is inadequate evidence that maternal occupational exposure to ELF EMF causes adverse birth outcomes. This conclusion was supported by 22 Working Group members; there were 2 votes for 'lack' of evidence, 1 abstention, and 4 absent.
There is inadequate evidence that paternal occupational exposure to ELF EMF causes reproductive effects. This conclusion was supported by 20 Working Group members; there were 3 votes for 'lack' of evidence, 2 abstentions, and 4 absent.
There is inadequate evidence that occupational exposure to ELF EMF causes Alzheimer disease. This conclusion was supported by 23 Working Group members; there was 1 vote for 'lack' of evidence, 1 abstention, and 4 absent.
There is inadequate evidence that occupational exposure to ELF EMF causes amyotrophic lateral sclerosis. This conclusion was supported by 24 Working Group members; there was 1 abstention and 4 absent.
There is inadequate evidence that occupational exposure to ELF EMF causes suicide or depression. This conclusion was supported by 17 Working Group members; there were 6 votes for 'lack' of evidence, 2 abstentions, and 4 absent.
There is inadequate evidence that occupational exposure to ELF EMF causes cardiovascular disease. This conclusion was supported by 24 Working Group members; there was 1 abstention and 4 absent.
There is inadequate evidence that environmental exposure to ELF EMF has adverse effects on pregnancy outcome or is associated with depression. This conclusion was support by 23 Working Group members; there was one vote for 'no' evidence, 1 abstention, and 4 absent.
There is no evidence in experimental animals for effects of ELF EMF on the immune system. This conclusion was supported by 13 members of the Working Group; there were six votes for 'weak' evidence, one abstention, and nine absent.
There is no evidence that exposure to power-line frequency EMF affects the hemotological parameters of rodents. This conclusion was supported by 17 members of the Working Group; there was 1 abstention and 11 absent.
There is weak evidence for the neurobehavioral, neuropharmacological, neurophysiological, and neurochemical effects of electromagnetic fields in experimental animals. This conclusion was supported by 9 members of the Working Group; there were 8 votes for 'moderate' evidence, 3 abstentions, and 9 absent.
There is no evidence for the reproductive or developmental effects of exposure to sinusoidal magnetic fields in experimental animals. This conclusion was supported by 17 Working Group members; there were 3 votes for 'weak' evidence, 8 abstentions, 1 absent.
Because of the complexity of the electromagnetic environment, the review of epidemiological and other biological studies did not allow precise determination of the specific, critical conditions of exposure to ELF EMF associated with the disease end-points studied.
There was one biological effect which the majority (14 out of 19) of the Working Group found to have strong evidence; exposure to electric and magnetic fields affects bone repair and adaptation. The remaining 5 votes were for moderate evidence. There appears to be substantial, accumulating evidence that complex clinical exposures to PEMF have a significant effect on the primary bone healing processes. The studies of both osteotomy and spinal fusion show a robust effect. While quantification and analysis were weak in these two studies, they are prospective, randomized, double-blind trials, a rarity in the field of orthopedics. Perhaps the most convincing trial is that of the response of bone tissue during limb lengthening. While no effect on secondary bone healing was observed, there was significant inhibition of bone resorption and evidence of new bone formation. Studies in animals in vivo indicate only limited efficacy. Magnetic therapy appears incapable of enhancing the healing of osteotomies, ingrowth of bone into a defect, bone elongation, or graft healing and, in at least one case (ingrowth), may inhibit the normal process. The results obtained in a model of endochondral ossification after exposure of whole animals suggest, however, that magnetic field therapy can be effective. Conversely, magnetic fields in animals appear to have a strong, reproducible effect on the process of appositional (surface) bone growth and on inhibition of bone resorption.
The Working Group could not reach a conclusion about whether exposure to electric and magnetic fields affect nervous and non-bone connective tissue repair and adaptation in vertebrates. This conclusion was supported 12 Working Group members abstaining; there were 10 votes for 'moderate' evidence, 6 votes for 'weak' evidence, and 1 absent.
There is weak evidence that short term human exposure to ELF EMF causes changes in heart-rate variability. This conclusion was supported by 13 Working Group members; there was 1 vote for 'moderate' evidence, 2 votes for 'no' evidence, 8 abstentions, and 5 absent.
There is weak evidence that short term human exposure to ELF EMF causes changes in sleep disturbance. This conclusion was supported by 15 Working Group members; there were 9 abstentions and 5 absent.
There is weak evidence that short term human exposure to ELF EMF causes suppression of melatonin. The conclusion for effects on melatonin was supported by 16 Working Group members; there was 1 vote for 'moderate' evidence, 2 votes for 'no' evidence, 5 abstentions, and 5 absent.
There is no evidence that such exposure has other effects on the biological end-points studied in the laboratory. This conclusion was supported by 12 Working Group members; there were 2 votes for 'weak' evidence, 11 abstentions, and 5 absent. The tie vote was broken by the Chair.
There is weak evidence that exposure to electric and magnetic fields alters the levels of melatonin in rodents. This conclusion is supported by 14 members of the Working Group; there were 9 votes for moderate support; 4 abstentions; 2 absent.
There is no evidence that exposure to electric and magnetic fields alters the levels of melatonin in sheep or baboons. This conclusion is supported by 14 members of the Working Group; there were 13 abstentions; 2 absent.
There is no evidence in experimental animals for effects of ELF EMF on hematological system from exposure to ELF electromagnetic fields. This conclusion was supported by 17 members of the Working Group; there was 17 abstention and 11 absent.
There is strong evidence that electric fields can be perceived. This conclusion was supported by 18 members of the Working Group; 2 abstentions; 9 absent.
A majority of the Working Group concluded that classification
of ELF EMF as possibly carcinogenic (Group 2B) is a conservative,
public-health decision based on limited evidence of an increased
risk for childhood leukemias with residential exposure and an
increased occurrence of CLL associated with occupational exposure.
For these particular cancers, the results of in vivo, in
vitro, and mechanistic studies do not confirm or refute the
findings of the epidemiological studies. The overall body of evidence
has, however, laid a foundation for furthering our understanding
of the biological effects, mechanisms, and exposure circumstances
that may be related to the possible carcinogenicity and other
adverse human health effects of exposure to ELF EMF.