Free HealthE Newsletter

Do Electric and Magnetic Fields Pose A Health Risk?

The scientific evidence suggesting that ELF-EMF exposures pose any health risk is weak. The strongest evidence for health effects comes from associations observed in human populations with two forms of cancer: childhood leukemia and chronic lymphocytic leukemia in occupationally exposed adults. While the support from individual studies is weak, the epidemiological studies demonstrate, for some methods of measuring exposure, a fairly consistent pattern of a small, increased risk with increasing exposure that is somewhat weaker for chronic lymphocytic leukemia than for childhood leukemia. In contrast, the mechanistic studies and the animal toxicology literature fail to demonstrate any consistent pattern across studies although sporadic findings of biological effects (including increased cancers in animals) have been reported. No indication of increased leukemias in experimental animals has been observed.

The lack of connection between the human data and the experimental data (animal and mechanistic) severely complicates the interpretation of these results. The human data are in the "right" species, are tied to "real-life" exposures and show some consistency that is difficult to ignore. This assessment is tempered by the observation that given the weak magnitude of these increased risks, some other factor or common source of error could explain these findings. However, no consistent explanation other than exposure to ELF-EMF has been identified.

Epidemiological studies have serious limitations in their ability to demonstrate a cause and effect relationship whereas laboratory studies, by design, can clearly show that cause and effect are possible. Virtually all of the laboratory evidence in animals and humans and most of the mechanistic work done in cells fail to support a causal relationship between exposure to ELF-EMF at environmental levels and changes in biological function or disease status. The lack of consistent, positive findings in animal or mechanistic studies weakens the belief that this association is actually due to ELF-EMF, but it cannot completely discount the epidemiological findings.

The NIEHS concludes that ELF-EMF exposure cannot be recognized as entirely safe because of weak scientific evidence that exposure may pose a leukemia hazard. In our opinion, this finding is insufficient to warrant aggressive regulatory concern. However, because virtually everyone in the United States uses electricity and therefore is routinely exposed to ELF-EMF, passive regulatory action is warranted such as a continued emphasis on educating both the public and the regulated community on means aimed at reducing exposures. This is described in greater detail in the section, Recommended Actions. The NIEHS does not believe that other cancers or non-cancer health outcomes provide sufficient evidence of a risk to currently warrant concern.

Scientific Evidence Supporting This Conclusion

The reports from the Science Review Symposia (9-11) and the Working Group (12) provide detailed reviews of the literature in this area of science. What follows is a brief synopsis of this evidence. The reader should refer to the individual reports for greater detail.

Background on the Limitations of Epidemiology Studies

Epidemiological studies are used to investigate the associations between health effects and exposure to a presumed disease agent. A well-designed and conducted epidemiological study involves several steps including identification of a study population, definition of the exposure to bestudied, choice of the type of study to conduct (e.g. cohort study versus case-c ontrol study) and description of the period over which the exposure is relevant. All of these factors influence the quality of a study and the limits thatmust be placed on interpretation of a study’s findings.

In carefully controlled laboratory and clinical investigations, study subjects are typically assigned to a treatment or exposure regimen. In epidemiological investigations, the inability to randomly assign exposures means that investigators must design their study so that the individuals who develop the disease of interest (cases) resemble the individuals who are disease-free (controls) in all aspects except for exposure; this is intended to limit possible bias. Bias due to improper selection of cases and controls is introduced if exposure is related to characteristics that would make cases more or less likely to be sampled than controls, or once sampled, to participate.

In the Nordic countries, comprehensive national population registries are generally used for selecting controls. If all persons are listed in these population registries and participation rates are high, bias due to selection of improper controls is unlikely even if exposure is related to participation. In countries such as the United States where population registries do not exist, other methods must be used to study rare diseases like leukemia for which existing cohort studies are inadequate. These methods lead to difficulties in identifying, contacting and recruiting controls that match the cases in all aspects other than exposure. For example, controls are sometimes identified through stratified random sampling of individual telephone numbers (random-digit dialing). Random-digit dialing may not properly identify controls of low socioeconomic status that do not have telephones; this could bias the results found in studies of childhood leukemias (13).

It is also possible to introduce bias through the selection of cases. For example, case selection bias may occur in studies that are based on mortality records (death certificates) if the survival rates of the exposed and unexposed subjects differ. This may occur if, for example, the exposure is related to socioeconomic status, and different socioeconomic groups have different survival rates for the studied disease (this might be due to a difference in the ability of cases to receive medical care). In addition, for diseases that are easily cured or allow patients to survive with the disease for a long period of time, persons who contract the disease and are treated properly may die of other causes and not appear as cases.

The inability to randomly assign exposures also introduces the possibility of confounding. Confounding occurs when the exposure of interest is associated with another factor that can increase (or decrease) the risk of getting the disease of interest (14). For example, smoking increases the risk of oral cancer; smoking is also associated with alcohol consumption, and there is a greater proportion of smokers among alcohol drinkers than among non-drinkers. Because smoking increases the risk of oral cancer and alcohol drinkers are more likely to smoke than non-drinkers are, alcohol drinkers will have a greater risk of oral cancer simply as a consequence of the greater percentage of smokers among alcohol drinkers. Thus, any study showing an increased risk of oral cancer associated with alcohol drinking will overstate that risk (resulting in a positive bias) if the effect of smoking is not carefully evaluated. Confounding can produce bias in either direction, artificially increasing or decreasing risks, depending on the direction of the association between the exposure, the disease and the confounder. When known, confounding can be controlled through statistical methods. Because there are very few known causes of childhood leukemias and chronic lymphocytic leukemia, it is difficult to identify and control potential confounders in these studies.

Another limitation of epidemiological studies is that exposure occurs through the natural course of events rather than being assigned and controlled by the investigator. Thus, a determination of the degree of exposure can be incorrect leading to what is known as "exposure misclassification." Exposure misclassification may distort measures of association observed in a study. For example, in epidemiological studies aimed at exposures received on the job (occupational studies), it is common to define exposures by the type of job a person performs. Errors may occur in assigning job titles or the jobs themselves may have markedly different exposures for different individuals. It is also possible that the exposure assignment may differ for diseased and non-diseased subjects. Information on exposure can be obtained either prospectively (before the disease has occurred) or retrospectively (after the disease has occurred). In the case where exposure is determined prior to disease onset, there is a reduced potential for misclassification of the exposure. In the case where exposure is determined after the onset of the disease, especially where it is obtained from questioning individuals with the disease, the recall of exposure may be influenced by the fact that the patient has a disease and is influenced by previous descriptions of potential causes of that disease.

Epidemiological studies have used various methods for estimating past ELF-EMF exposure to provide scientific evidence concerning the possibility of health effects from exposure to ELF-EMF. Residential exposures to ELF-EMF have been conducted in five basic ways: wire codes that are essentially based upon distance to major structures used for delivering electrical energy (e.g. high tension power lines and transformers); calculated magnetic fields that are based upon a theoretical calculation of the magnetic field emitted by certain types of power lines using historical electrical loads on those lines; spot measurements that generally give a single, instantaneous measurement of the magnitude of the magnetic field in one or more spots in a residence; average measured fields that are essentially spot measurements taken repeatedly every few seconds for 24 hours and averaged over time; and personal average measured fields where the subject wears a monitor and measurements are taken repeatedly every few seconds for 48 hours and averaged over time.

The validity of individual exposure assessment methods has been examined and each has its limitations (12, 15-20). Wire codes and calculated fields have the advantage of remaining fairly consistent over time making them more likely to be correctly determined during the time of cancer onset. However, their main disadvantage over measured fields is a lack of consideration of all possible sources of exposure, in particular fields from in-home appliances and ground currents. The relationship of wire codes to direct magnetic field measurements has been examined; the reliability of wire codes as a quantitative measure of magnetic field exposure is variable(15, 17, 19, 20).

Childhood Cancers

The hypothesis generated by the seminal study of Wertheimer and Leeper (1) used wire codes to evaluate residential exposures in children. Four additional epidemiological studies in which wire codes were used to assess exposure to ELF-EMF are of sufficient quality to be used in the evaluation of a causal association between the risk of childhood leukemia and exposure to magnetic fields. Two of the studies reported an association (21, 22), and two studies reported no association with the risk for childhood leukemia (23, 24). A trend of increasing risk with wire codes classification implying increased fields was observed in the two positive studies(21, 22). All of these studies, including the seminal study, could have been affected by the types of biases described earlier including exposure bias (1), control selection (all five studies), and confounding fromother risk factors (all five studies). In addition, the seminal study and the four subsequent studies differed in their groupings of leukemias ranging from evaluating all types of leukemias (1, 21, 22, 24) to evaluating only acute lymphoblastic leukemia(23, 24), the most common form of the disease in children. The most recent U.S. study(23) is the largest of the four subsequent studies for evaluating ELF-EMF exposure. Even though this study (23) shows a negative association when comparing Wertheimer-Leeper wire codes with leukemia risks, when combined with the remaining studies (21, 22, 24) in a meta-analysis (a form of statistical analysis in which like studies are combined to get a single answer), the results indicate a marginal association for the highest exposure group versus the lowest exposure groups. Removal of any of the three remaining studies (21, 22, 24) diminishes this association substantially. After removal of the one follow-up study with the most severe design limitations (21), the association is no longer present. Another study (25) was not included in the meta-analysis due to study limitations; this study showed no effect of wire codes.

Four epidemiological studies (26-29) assessed exposure using calculated fields; all four studies were conducted in Nordic countries. Three of the studies observed an increased leukemia risk in one or more exposure group (26-28) although only one (26) achieved statistical significance. All four studies were population-based, with minimal potential for selection bias both in terms of control selection and participation rates. The main limitations of all four studies are the small number of cases overall and the small number of cases and controls in the high exposure group. The general trend of these studies provides marginal support for a small, increased risk (30).

Four studies in which spot measurements were used to assess exposure to magnetic fields are clearly of greater quality than the remaining studies(21, 22, 26, 31). Two of these studies(21, 22) observed increased risks of marginal significance in one or more exposure groups and the other two(26, 31) showed no risk. Overall, spot measurements do not show an appreciable excess risk for leukemia when the four studies are combined (30).

Four studies used 24-hour measured magnetic fields to assess exposure (22-24, 31)1. The studies examined three different classifications of childhood leukemias: acute lymphocytic leukemia(23, 24), acute leukemia (31) and leukemia including nonlymphocytic leukemia (22, 24). The results of three of the studies showed an increased risk for children in higher exposure class(es); in two studies there were no statistically significant differences(22, 24), in the largest study only one experimental category out of many was statistically significant (23), and depending on the grouping, the fourth study achieved statistical significance (31). The data reported for the largest study (23) suggest an exposure—response relationship that the original authors did not consider important. The pattern of dose versus response in this study was considerably different from the pattern in the other two studies with multiple dose groups (22, 24). The results of these studies, when combined, provide weak evidence for an association between exposure based on 24-hour measured magnetic fields and a small, increased incidence of childhood leukemia (30).

1This publication (24) only provides a single odds ratio from their analysis of the 24-hour measurements. Additional information was obtained from the principal author.

One study (24) assessed exposure using 48-hour personal monitors that measured both magnetic fields and electric fields. Analyses were done for all childhood leukemias and separately for acute lymphocytic leukemia. The general trend in the data indicated a negative association for both magnetic fields (current or predicted two years prior to diagnosis) and electric fields. No statistically significant positive associations were observed. This study, using personal exposure meters, does not support an association between ELF-EMF exposure and childhood leukemia.

Several of the same studies described earlier also looked at electrical appliance use and the risk of childhood leukemia (22, 32, 33). The results do not fit a coherent pattern.

None of the individual epidemiological studies provides convincing evidence linking magnetic field exposure with childhood leukemia. Hence, in making an assessment, one must rely upon the evaluation of the data as a whole using expert judgment and the meta-analyses as a guide. The pattern of response, for some methods of measuring exposure, suggests a weak association between increasing exposure and increasing risk. The small number of cases in these studies makes it impossible to firmly demonstrate this association. This level of evidence, while weak, is still sufficient to warrant limited concern.

Two other childhood cancers have been sufficiently studied to warrant comment. Two early studies observed an increased risk of brain cancers using wire codes as the exposure measure (1, 21). Later studies using wire codes (34, 35), calculated fields (26-28, 36) and measured fields (35) failed to support this finding. The association between exposure to ELF-EMF and childhood lymphomas was considered in several epidemiological investigations (1, 21, 26-28, 36). In all studies, the number of cases of lymphoma in the high exposure groups was too small for any reliable inference to be drawn. In general, these data do not support the concern that exposure to magnetic fields may increase the risk of brain cancers or lymphomas in children.

Adult Cancers

Epidemiological reports of diseases associated with occupational exposure to ELF-EMF preceded concerns about residential exposure. Reports of various health problems in high-voltage substations in the former USSR initially focused attention on ELF electric fields (37). Initial studies in the United States (38, 39) led to over 100 epidemiological investigations of workplace exposure to ELF-EMF and various diseases. The early studies were based on workers in jobs assumed to entail exposure, and more recent studies used measured fields.

Recent studies evaluating the association between exposure to magnetic fields and chronic lymphocytic leukemia (40-44) show mixed results. The two studies in the United States (43, 44) reported no association, but one (44) used death certificates to identify the cases (chronic lymphocytic leukemia has a rather long survival time that can confound the diagnosis of the cases). One of the remaining studies (42) indicated increased risk, which did not achieve statistical significance, and the two Scandinavian studies (40, 41) showed significantly elevated risks in one or more exposure groups. Both of the Scandinavian studies had consistently increasing risks with increasing exposure. Each of these studies has its limitations and the limitations are different across studies, as are the designs and exposure assessment methods. Taken together, the studies provide weak evidence for an association between occupational exposure to magnetic fields and chronic lymphocytic leukemia.

Acute myelogenous leukemia was considered in these same epidemiological studies. The results, which were observed from these studies, are not sufficiently compelling to support an association.

The association between exposure to magnetic fields and a variety of other cancers has also been considered in occupational settings. Included are brain cancers, breast cancers (in both males and females), testicular cancers, cancers in offspring of workers, lymphoma, multiple myeloma, melanoma, non-Hodgkin’s lymphoma, thyroid cancers and many others. Some evidence exists for an association between brain cancers and exposure to ELF-EMF and between female breast cancers and ELF-EMF exposure; however, the studies evaluating these associations are inconsistent and have limits to their interpretation making them inadequate for supporting or refuting an eff ect. In the remaining cases, the evidence supporting an association is negative or too weak to warrant concern.

The risks of adult cancer based on residential exposure to ELF-EMF have been evaluated in a number of studies. Risks of leukemia (of all types and of specific sub-types) from residential exposures were evaluated in several recent studies (40, 45-50). The calculated field studies (40, 47-50) showed mixed results for the different sub-types of leukemia studied and for changes in the definition of the exposure category. Specifically, when chronic lymphocytic leukemias was examined separately (this was done in only two of the studies), the results were inconsistent with one study(40, 48) showing no increased risk and with the other (49) showing fairly consistent dose-response with increasing cumulative exposure. The remaining studies, using wire codes (46) and measured fields (46, 48), demonstrated no increased risk. These data are inadequate for evaluating the association between exposure to ELF-EMF and leukemias. Specifically, for chronic lymphocytic leukemia, which demonstrated a weak association in the occupational studies, there are mixed results for adults in the residential studies.

The risk for leukemia associated with use of electrical appliances was also considered in two studies (45, 51). These studies resulted in inconsistent findings and generally do not support an association between appliance use and increased leukemia risk.

Limited data are available on risks of male and female breast cancer associated with residential exposure to ELF-EMF. A small, non-significant association between use of electric blankets and the risk for breast cancer was observed in one, large U.S. study(52) but not in another(53). Both found no evidence for an association with duration of exposure. Three studies, using exposure measured by calculated fields (50, 54, 55), identified no association between exposure to magnetic fields and the risk of breast cancer. These same scientists(40, 47, 48, 50, 55) also looked at exposures to ELF-EMF and cancers of the central nervous system (such as brain cancers); no associations were found.

None of the associations between cancer and residential exposure to magnetic fields in adults were indicative of a positive association. However, the specific adult cancer showing weak evidence of a positive association with occupational exposure to ELF-EMF, chronic lymphocytic leukemia, was inadequately studied in residential settings. It cannot, therefore, be concluded that there is no association.

Non-Cancer Findings in Humans

The relationship between spontaneous abortion and exposure to ELF-EMF has been considered in several studies. Recent occupational and residential studies were the focus of this assessment. In the first occupational study (56), no association was observed. In a second occupational study (57), a significant association was found with exposure to high ELF-EMF; however, the response rate was very poor, particularly among controls, which could have biased this result upward. Pregnancy loss was investigated in two residential cohort studies (58, 59). In one study (58), an increased risk was observed in the highest exposure category but not in the intermediate category. In the other (59), no association was observed for any measure of exposure. In a carefully designed prospective study in the United States (60) , no association was reported between measured fields (including personal exposure monitoring) and intrauterine growth, birth weight or gestational age.

Low birth weight (60, 61), intrauterine growth retardation (60), preterm birth (61) and congenital anomalies arising from the father’s exposure (62) were not associated with occupational exposures to ELF-EMF. The risk for congenital anomalies in relation to the mother’s use of heated waterbeds and electric blankets around the time of conception was evaluated in three studies(63-65); no association was observed for heated waterbeds in any study, and inconsistent results were reported for electric blanket use.

The association between occupational exposure to ELF-EMF and Alzheimer’s disease was considered in five studies (66-70). All five studies showed increases in one or more exposure groups with four studies (66-69) showing statistically significant increases and one (70) showing non-statistically significant increases. All of these studies suffer from design limitations that make it inappropriate to use them for addressing a causal association between ELF-EMF exposure and Alzheimer’s disease. Two of these (66, 67) are based on diagnoses from death certificates (Alzheimer’s disease is not consistently noted on death certificates). Two studies (68, 69) used different groups of cases and controls; some of the control groups included persons with other types of dementia, and proxy information was used to define the exposure of cases. The one remaining study(70) was evaluated using data for twins and also suffered many limitations. These data are inadequate for interpreting the possibility of an association.

The association between exposure to magnetic fields and amyotrophic lateral sclerosis was assessed in three studies(66, 71, 72). One study(71) showed an increased risk in the highest exposure group and the other two studies were negative. Adequate adjustment could not be made for known risk factors (electric shocks or a family history of amyotrophic lateral sclerosis) making these studies difficult to interpret.

Suicide and depression were studied in three occupational epidemiological studies(72-74). These studies do not support an association with ELF-EMF exposure.

Two occupational studies (75, 76) assessed possible adverse cardiovascular outcomes that may result from exposure to magnetic fields. In the first study (75), a significant decrease in risk using a broadly defined cardiovascular grouping was observed. In the second (76), data from five utilities were examined. This study was motivated a priori by a biological hypothesis based on the results of human clinical studies on heart rate variability (77) for increased numbers of deaths due to arrhythmia and acute myocardial infarct. Significant, exposure-dependent associations were reported. Lacking additional epidemiological studies to collaborate these results, these data are inconclusive regarding an association between cardiovascular disease and exposure to ELF-EMF.

Human clinical studies of ELF-EMF exposures were carried out mainly through three major research initiatives. These include a long series of studies of utility workers begun in the 1960s in the former USSR (37), human laboratory research conducted in the 1970s in Germany(78, 79) and the human laboratory research program started in 1982 at the Midwest Research Institute in the United States(80). Dedicated facilities for human exposure testing were designed and constructed in Australia (81), Canada (82), England(83), France(84), Germany (78), New Zealand (85), the Russian Federation (86) and the United States (87, 88). Research with human volunteers is currently under way in many of these facilities.

A large number of clinical end-points were evaluated in these laboratories. Several effects reported at high exposures warrant little concern as health dangers such as hair standing on end in very strong electric fields and flickering visual sensations in very strong magnetic fields. However, a number of measurements potentially linked to health effects have been studied. The central nervous system was one of the first areas investigated as a potential site of interaction with ELF-EMF. Studies of changes in brain wave patterns (electroencephalography) during waking hours were generally negative showing little or no effect of ELF-EMF, especially in the range of power-line frequencies (79, 80, 86, 89-94). Several studies (95-97) showed decreased sleep and reduced sleep efficiency during ELF-EMF exposure. These studies all had deficiencies (e.g. disturbance of subjects by drawing blood and incomplete adaptation of study subjects to the laboratory environment) making them inconclusive.

Changes in human pulse as a function of exposure to ELF-EMF fall into two categories: changes in the number of beats per minute (pulse rate) and changes in the variability of the electro-chemical signals going to the heart (heart-rate variability). Two research groups examined changes in pulse rate following exposure to ELF-EMF (80, 91-93, 98, 99). All five clinical studies (80, 91-93, 99) from the same laboratory showed a decrease in pulse rate in at least one exposure group; however, all exposures represented rather large, combined electric and magnetic fields (6 to 12 kV/m and 10 to 30 mT, respectively). The remaining study (98) was a field trial under a high-tension power line and no effect was observed. The biological mechanism is unknown, and the general effect is very small making it unlikely that this is a health risk at lower doses.

Changes in heart-rate variability were evaluated in a retrospective analysis of three previous studies (77). Some changes in heart-rate variability were observed, which according to the authors, could indicate a potential for increased risk of sudden cardiovascular death. However, even though decreased heart-rate variability is associated with increased risk of cardiovascular death, it is not clear that transiently induced changes in healthy individuals will carry any risk. While these findings are inconclusive, the recent epidemiological result (76) discussed earlier suggests this area may warrant additional study.

Two possible mechanistic explanations for cancer findings from exposure to ELF-EMF, changes in melatonin (a hormone associated with sleep) and changes in the immune system, have been studied. The potential for ELF-EMF exposure to alter nighttime melatonin levels was addressed in 11 studies (81, 84, 96, 100-106). The clinical studies (81, 84, 96, 102, 103) demonstrated no consistent pattern of melatonin reduction (one study saw a marginal effect in men with already reduced melatonin levels and one saw a reduction in onset of the nightly increase in melatonin). In the occupational studies (100, 101, 105, 106), some changes were reported in urinary excretion of melatonin metabolites (the result of degradation of melatonin in the body) following workplace exposure (when melatonin levels are generally low), but not in evening melatonin levels. In the one residential study (104), significant dose-related reductions were associated with measured fields in bedrooms, but not with other measures (e.g. wire codes and total 72-hour exposure). All combined, these studies provide little support that exposure to ELF-EMF is altering melatonin levels in humans. A number of other hormones were also studied such as testosterone, thyroid hormones and several stress hormones; no effects of ELF-EMF exposure on these levels were observed.

Few laboratories studied the effects of ELF-EMF on the immune system. Three studies investigated effects of ELF-EMF exposure on the immune system (80, 107, 108) and all were negative.

Finally, there have been a number of case reports of mood changes and hypersensitivity thought attributable to ELF-EMF exposure (manifested as physiological reactions, disturbed sleep, fatigue, headaches, loss of concentration, dizziness, eye strain and skin problems). These symptoms generally seem to be intermittent and difficult to study clinically. Several carefully designed studies (109-113) were performed to evaluate the response of persons with these symptoms to ELF-EMF. In general, these studies were negative with the exception of one (112) that reported an increased incidence of skin rashes in persons exposed to high ambient electric fields (>31 V/m) relative to control fields (<10 V/m). These data are insufficient to support an association between ELF-EMF and hypersensitivity.

Animal Cancer Data

Animal carcinogenicity studies are routinely used to identify environmental agents that may increase cancer risk in humans. Many areas of biological investigation are more efficiently studied in animal models than in human beings, because the agent can be studied invasively and under carefully controlled environmental conditions. The use of animal models in studying effects of ELF-EMF exposure is limited by two problems: extrapolation of experimental findings across species and extrapolation of laboratory exposure patterns to environmental exposure patterns. Animal carcinogenic studies of ELF-EMF were done at levels of exposure generally much higher and having greater uniformity in frequency and intensity than would appear in environmental settings. These experimental conditions were chosen to maximize the ability of a researcher to detect an effect, if one exists, for a clearly defined exposure.

The laboratory data in animal models are inadequate to conclude that exposure to ELF-EMF alters the rate or pattern of cancer. There are some sporadic findings (including increased cancers) with no clear interpretation; however, it is noteworthy that these data provide no support for the reported epidemiological findings (discussed earlier) of increased risk for leukemia for ELF-EMF exposure.

Only a few lifetime bioassay studies (114-116) have been performed for ELF-EMF exposure. These studies exposed large groups of animals generally for periods of up to two years at magnetic field intensities considerably higher than elevated residential exposures. No consistent effects of ELF-EMF exposure on cancer rates in bioassay animals were found. The most comprehensive study conducted through the National Toxicology Program (115) used four exposure groups (control, 2, 200 and 1000 µT continuous exposure for 18.5 hours per day and 1000 µT intermittent exposure) and four gender/species groups. There were no exposure-related clinical findings for rats or mice. The two-year study found no evidence of carcinogenicity in female rats and male or female mice at any exposure level and equivocal evidence for carcinogenicity in male rats based upon an increased incidence of thyroid gland C-cell tumors.

A similar study (114) was conducted in female rats where exposure to 60 Hz linearly polarized magnetic fields (control, 2, 20, 200 and 2000 µT continuous exposure) began in utero two days before birth and continued for 20 hours per day for two years. No consistent, exposure-related clinical findings or evidence of carcinogenic activity from 60 Hz magnetic fields were reported. In another study (116) male and female rats were exposed to control, 500 or 5000 µT 50 Hz magnetic fields for 22.6 hours per day for two years. No differences in cancer rates between field-exposed and sham-exposed animals were found.

Epidemiological findings have suggested a possible association between magnetic field exposure and breast cancer in men(117, 118) or women (119). In addition, a hypothesis was proposed that magnetic field exposure might lower nocturnal melatonin levels that could increase risk for breast cancer (120). Animal studies using chemically induced mammary cancer followed by magnetic field promotion of carcinogenesis were undertaken to test whether mammary cancer was affected by ELF-EMF exposure.

Following an initial report that magnetic fields promoted mammary tumor development in rodents (121), a comprehensive series of studies on ELF-EMF exposure and mammary tumor initiation and promotion in the rodent model was conducted (122-124). In these studies, female Sprague-Dawley rats were used and cancer was initiated by intragastric administration of four weekly doses of 7,12-dimethylbenz[a]anthracene (DMBA) followed by promotion with 50 Hz ELF magnetic fields, 24 hours per day for 13 weeks. One of the early studies in this series (122), where the data were subsequently examined histologically (125), provided evidence that magnetic fields of low flux density (100 µT) promoted increased growth and size of mammary tumors but did notaffect tumor incidence. The same laboratory repeated this work, and in additional studies testing different magnetic flux densities, examined the question of whether a dose-response relationship exists with field intensity (126-128). Over the range of 10 to 100 µT magnetic fields (50 Hz), a higher (not statistically significant) number of total tumors was found in the field-exposed groups. Magnetic field exposure was not associated with more tumors per tumor-bearing animal. Effects on tumor latency and size were not consistent across the studies.

The National Toxicology Program (129) conducted similar studies. Animals were exposed to magnetic fields at both European frequency (50 Hz, 100 or 500 µT) and American frequency (60 Hz, 100 µT) 18.5 hours per day, seven days per week for 13 weeks following intragastric administration of four weekly doses of DMBA as the initiator. There was no difference in size or incidence of mammary gland tumors between control and exposed groups. However, the tumor incidence was high in all groups, and sensitivity was reduced for detecting a promoting effect of magnetic fields. The study was repeated at a lower dose of DMBA. Tumor incidence, latency and size, total number of tumors and number of tumors per tumor-bearing animal were not affected by magnetic field exposure; in the exposure groups there were slightly fewer total mammary neoplasms (not statistically significant) than in controls. A 26-week study, where animals received a single initiating dose of DMBA, gave similar results (129); there were significantly fewer tumors for the two exposed groups. However, the tumor incidence was high in all groups, and sensitivity was reduced for detecting promoting effects of magnetic fields. This collection of studies (129) provides strong evidence of no effect of magnetic fields on the promotional development of mammary cancer.

Another laboratory (130) also examined the effects of magnetic field exposure, which included transients, on mammary tumor development in female Sprague-Dawley rats. This study differed slightly in experimental design from the ones described earlier, but used DMBA as initiator and examined similar magnetic fields, 250 and 500 µT, at 50 Hz. No effects of magnetic fields were observed.

The explanation for the observed difference among these studies is not readily apparent. However, within the limits of the experimental rodent model of multistage mammary carcinogenesis, the findings do not provide consistent evidence for a promoting effect of ELF-EMF on chemically induced mammary cancer.

Animal models of skin carcinogenesis are well established for the study of the initiation, promotion and progression of cancer (131). Several laboratories examined whether 50 and 60 Hz magnetic fields promoted or co-promoted development of cancer using this model (132-137). Skin tumors were initiated by topical treatment of the animals with a known chemical carcinogen (e.g. DMBA) followed by exposure to various intensities of magnetic fields or combinations of magnetic fields plus a known chemical promoter (e.g. 12-O-tetradecanoyl phorbol 13-acetate, TPA). The findings from these studies demonstrated no significant promotional effect of magnetic fields on skin tumor development.

Rat liver is a most commonly used experimental model for investigating multistage carcinogenesis in tissues other than the skin (138). Several experiments from a single laboratory used this model to investigate ELF-EMF exposure effects and reported no evidence of a promotional or co-promotional role of magnetic fields in cancer development(139, 140).

Several epidemiological studies have suggested a possible association between ELF-EMF exposure and an increased risk for leukemia. Two types of animal models were used for determining whether magnetic fields can alter the time of onset or incidence of leukemia: 1) initiation with X-rays or chemical carcinogen followed by ELF-EMF exposure and 2) progression of leukemia by injection of leukemia cells into the animal followed by ELF-EMF exposure.

The largest ELF-EMF study using an agent to initiate disease involved over 2000 mice with different doses of ionizing radiation to initiate lymphoma followed by either exposure to 1400 µT magnetic fields or no exposure for up to 30 months. Exposure to magnetic fields did not affect the incidence or time of onset of leukemia/lymphoma, the rate of death among animals with leukemia/lymphoma or the leukemia sub-types (141). In another study (142), no promotional effects of a 1000 µT 50 Hz magnetic field in mice were found following initiation of lymphoma/leukemia with DMBA.

A study of leukemia progression was conducted in Fischer rats inoculated with large granular lymphocytic leukemia cells(143, 144). In the first study(144), treatment with a 1000 µT continuous 60 Hz magnetic field did not significantly alter the clinical progression of the disease in exposed versus ambient-field controls. In the second study(143), an additional, lower inoculum of leukemia cells was included to increase sensitivity as well as intermittent magnetic field presentation (3 min on, 3 min off). No significant effects were observed for the continuous field exposure at either inoculum; however, with intermittent fields at the higher inoculum, latency to disease was slightly decreased.

The findings from the lifetime bioassay study ((115), discussed earlier) with ELF-EMF exposure are also consistent with the absence of an effect on leukemia/lymphoma. When animals exposed to a range of magnetic fields for up to two years were examined, no increases in leukemias or lymphomas were found in the 16 gender/species groups.

Two studies were conducted in genetically altered mice that are prone to leukemia (145, 146). These studies showed no evidence of magnetic field effects on lymphoma incidence.

Based upon some evidence from occupational and residential studies suggesting an increased risk for brain cancer with ELF-EMF exposure, several animal studies examined this question. Rodent models are relatively insensitive to the induction of brain cancer by chemicals, and as such, caution should be used in interpreting the findings from studies with ELF-EMF exposure. The lifetime studies in rodents (114-116) demonstrated no effect of magnetic field exposure on brain cancer. In the large initiation/promotion leukemia study in female mice ((141), discussed earlier), sections of the brain were prepared and reviewed for primary proliferative lesions (147). No evidence of an effect of magnetic field exposure on primary brain tumors was found.

Non-Cancer Health Effects in Experimental Animals

A number of non-cancer end-points were investigated for possible adverse effects of ELF-EMF exposure. In general, the experimental models used to study interactions with ELF-EMF have been guided by methods and end-points that were developed to assay the effects of other physical and chemical agents such as drugs, chemicals and ionizing radiation.

The effects of ELF-EMF exposure on the immune system were investigated in multiple animal models including baboons and rodents, and there is no consistent evidence in experimental animals for effects from ELF-EMF exposure. Reports of effects in baboons (148) were not confirmed when the study was repeated. Some studies had methodological difficulties making interpretation of the findings difficult (127, 149). Other studies found no or inconsistent effects of ELF-EMF exposure on immune system indices and function (150, 151).

Seven studies examined standard measurements of hematological and clinical chemistry indices following ELF-EMF exposure(152-158); several included a limited number of animals and were of short duration. These studies provide no evidence that exposure to ELF-EMF affects hematological or clinical chemistry parameters in rodents.

A variety of animal models including non-human primates, pigeons and rodents were exposed to high intensity electric or magnetic fields to study the behavior and physiology of the nervous system. Detection of electric fields by animals is a well-established phenomenon, and the sensitivity thresholds for animals appear to be similar.

Various neuro-behavioral responses including avoidance and aversion and learning and performance were tested for effects from exposure to ELF-EMF. The data from studies including baboons and rodents suggest that exposure to strong electric fields can be perceived (159-162), but there is no evidence that these fields are harmful at environmental intensities. The addition of a magnetic field to the electric field appears to modulate the acute behavioral response of animals to perceptible electric fields (163, 164).

Relatively little evidence is available for evaluating whether exposure to ELF electric fields can affect performance of learned behavior. The studies in baboons(160, 161) suggest that any effects are minimal. In contrast, exposure to ELF magnetic fields was associated with several effects: adverse (165, 166), beneficial (167) or absent (168, 169) depending upon the task being performed and the timing of the magnetic field exposure. Studies in non-human primates with combined exposure to electric fields and magnetic fields detected no impact on operant performance (164, 170).

Epidemiological studies have addressed the question of whether ELF-EMF exposure affects reproduction and development. Studies using avian species were conducted, but their relevance to mammalian systems is not clear. Studies examining teratogenic and reproductive end-points were also done in mammalian systems. An extensive evaluation of magnetic field exposure (control, 2, 200 and 1000 µT continuous exposure and 1000 µT intermittent exposure) on fetal development and reproductive toxicity in the rodent was conducted (171). There was no evidence of any maternal or fetal toxicity or malformation. A further study examined multi-generational reproductive toxicity using a continuous breeding experiment. The results suggested no evidence of altered reproductive performance or developmental toxicity in the rat(172).

At the onset of the EMF-RAPID Program, one hypothesis was that magnetic fields acting through the retina as a sensitive receptor reduce melatonin levels. It was thought that this depression might act as a risk factor for cancer (120, 173). Studies examining effects of ELF-EMF exposure on circulating melatonin levels were conducted in a variety of mammalian species. Overall, the experimental evidence is lacking in consistency and quality across the studies. The data in rodents is weak, but suggests that when effects do occur, the result is a decrease in melatonin concentration. There is no evidence for ELF-EMF effects on melatonin in sheep and baboons. These findings parallel those reported from clinical investigations in humans and population studies (discussed earlier).

Long-term exposure to electric fields decreases melatonin concentrations slightly in rats(174-177); the biological significance of this effect is not understood. In a series of studies of acute magnetic field exposure in hamsters(178-180), a suppression of pineal and plasma melatonin levels reported in the earliest study was not replicated in later studies. Studies in rats with different magnetic field exposures, field intensities and times of exposure relative to the dark cycle have not shown consistent effects of magnetic fields on melatonin levels. Some laboratories reported that long-term exposure to magnetic fields in rats can reduce nocturnal pineal or blood concentrations of melatonin (123, 181-184), but other laboratories did not find similar results (127, 129, 185, 186). Interpretation of the findings from this large data set is complicated by variability across studies in confounding factors such as species, strain, gender, co-exposure tochemicals, field characteristics and measured outcomes. Long-term studies of ELF-EMF exposure in lambs (187, 188) and baboons (189) showed no effects on melatonin levels.

Studies of Cellular Effects of ELF-EMF

The number of cellular components, processes and systems that can possibly be affected by ELF-EMF is large. Historically, testing of potentially toxic substances has relied on the use of carefully controlled in vitro experimental systems. In an attempt to identify potentially carcinogenic or toxic effects of an agent, these studies have typically exposed cells to the agent over a range of doses including levels above those encountered in the environment. Measurements are then made of cellular end-points as a means to detect alterations in processes such as differentiation, proliferation, gene expression and signal transduction pathways. This toxicological approach was applied to ELF-EMF in general through exposure of cultured cells over a range of doses. Because nothing is known about the potential mechanistic action of ELF-EMF on biological end-points, careful consideration must be given to the range over which the experimental doses of ELF-EMF is varied. The extrapolation of observed effects to lower field intensities may be inappropriate as ELF-EMF may have different mechanistic actions over different patterns of field intensity. Likewise, the actual agents responsible for the ELF-EMF "dose" to which individuals are exposed are not clear. Environmental ELF-EMF exposure is complex being composed of not only pure 60 Hz electric fields and magnetic fields, but also possibly transients (intermittent spikes and changes in the frequency of the field) and harmonics (multiples of the pure 60 Hz exposure: 120, 180, 240, etc.). To understand this complexity, careful control of laboratory exposure conditions also becomes important to ensure that the exposure being tested is known.

The breadth of in vitro data on ELF-EMF produced over the last two decades is enormous. Many of these investigations were done using unique experimental protocols in single laboratories. Under the EMF-RAPID Program, a major focus was research that targeted examination of in vitro effects that might clarify potential mechanistic actions of ELF-EMF in order to explain reported epidemiological associations with magnetic fields. Because of the noted complexity of ELF-EMF exposures, efforts were also made to standardize the exposure systems used in these studies to allow for comparability of findings across laboratories. Through oversight by the DOE, on-site quality assurance evaluations were made of laboratories funded by this program. In addition, four regional ELF-EMF exposure facilities were established and made available for use by investigators (discussed earlier).

Through the EMF-RAPID Program, considerable progress was made in the area of in vitro research on ELF-EMF. Many of these studies of ELF-EMF exposure focused on end-points commonly associated with cancer (e.g. cell proliferation, disruption of signal transduction pathways and inhibition of differentiation). Convincing evidence for causing effects is only available for magnetic flux densities greater than 100 µT or internal electric field strengths greater than approximately 1 mV/m. To date, there is no generally accepted biophysical mechanism by which actions of lower intensity ELF-EMF exposures, including those reported to be of concern in epidemiological studies, might be explained.

Given the concern about whether ELF-EMF exposure is carcinogenic, considerable effort was undertaken to investigate whether ELF-EMF exposures can damage DNA or induce mutations. It has been generally believed that the energy associated with ELF-EMF is not sufficient to cause direct damage to DNA; however, it has been postulated that indirect effects might be possible by ELF-EMF altering processes within cells that could subsequently lead to changes in DNA structure. Overall, there was considerable variability in experimental design and methodology used in these studies resulting in no conclusive evidence that genotoxic effects result from ELF-EMF exposures.

Studies also examined the potential cytogenetic effects of power-frequency sine wave or pulsed magnetic fields using model systems of human cells isolated directly from peripheral blood and amniotic fluid or cultured human lymphocytes and leukemia cells. Overall, the studies varied considerably, and in general, there is no evidence of chromosomal damage even when cells were exposed to relatively strong magnetic fields (190, 191). Chromosomal aberrations were reported in one study (192) using pulsed magnetic fields; however, the exposures tested were within the range of exposures reported in other studies to have no effect.

Relatively few studies have addressed the question of whether ELF-EMF exposures cause genetic mutations(193). Studies using bacteria or yeast cells (194, 195) to investigate possible mutational changes in DNA reported no damage from ELF-EMF exposure at levels less than 1000 µT. However, at higher field strength (400,000 µT, 50 Hz), well above environmental field intensities, enhanced mutagenicity was reported in two cell lines(196, 197). Exposure to ELF-EMF (magnetic field strengths 500 µT) following exposure to ionizing radiation was reported to produce significant enhancement of mutagenicity(197, 198); ELF-EMF exposure alone had no effect. Several investigators examined the ability of ELF-EMF to alter the repair of DNA strand breaks caused by hydrogen peroxide or radiation; no effects with exposure to either magnetic or electric fields were observed(199-201).

The concept that ELF-EMF might be carcinogenic through effects on gene transcription was stimulated by an extensive series of studies in human leukemia cells(202, 203). It was initially reported that high-intensity ELF-EMF exposure increased expression of several genes important in carcinogenesis. The presence of this effect was later reported to occur at field intensities more characteristic of environmental levels(204) and in three types of human cell lines(203, 205, 206). Because some of these genes may have a central role in controlling cancer, these findings were of great significance. Intense efforts by several laboratories failed to confirm the reported findings(207-210). Follow-up studies by the original investigators demonstrated strain-specific responsiveness to ELF-EMF of the cell line(211), although this does not appear to explain the inability of other laboratories to confirm the reported findings(209).

Several investigations were undertaken to determine whether cells might respond to ELF-EMF with transcriptional or translational changes of heat-shock proteins, which are important in control of stress within a cell. Exposure of cells to ELF-EMF was reported from a single laboratory to result in increases in some of these proteins(212-214).

Signal transduction processes aid cells in receiving signals from their environment and from other cells. These signals help to regulate cellular processes such as gene expression, metabolic activity, differentiation and proliferation. Signals received by the cell membrane, which control processes within the cell, have been proposed as a means by which ELF-EMF might affect cellular function. In the case of electrical signals, these are not expected to penetrate the cell’s outer membrane but may signal release of proteins on the cell membrane that could alter cellular function.

Numerous laboratories performed studies to evaluate potential ELF-EMF effects on cellular end-points related to signal transduction pathways, which if altered, might be carcinogenic. Overall the body of evidence suggests that ELF-EMF exposures at magnetic field intensities greater than 100 µT and electric fields greater than 1 mV/m have shown effects on signal transduction pathways. Studies at lower exposures are inconclusive.

Recent studies investigated whether ELF-EMF exposure might play a role in B-cell leukemogenesis (the major form of childhood leukemia) through signaling pathways. A series of studies, which focused on one particular signal (the protein kinase C-linked signaling cascade), provided preliminary evidence that in vitro exposure to ELF-EMF (100 µT) can affect this pathway (215-217). This finding was not reproduced by a second independent laboratory (218).

Because of concern about ELF-EMF possibly being carcinogenic, studies were initiated to investigate whether there were effects on ornithine decarboxylase (ODC), an enzyme activated during carcinogenesis. An early study(219) reported increased ODC activity in three cell lines in response to a sinusoidal 60 Hz electric field (10 mV/cm). Subsequent work by others demonstrated effects of ELF magnetic fields (field strengths 100 µT) on ODC although the experimental conditions (e.g. cell line/tissue, field intensity, time of exposure) varied among laboratories (220-222). One study reported increased ODC activity in mouse lymphoma cells exposed to 10 µT 60 Hz magnetic fields(220). Attempts to reproduce this finding were not successful (223, 224).

Abnormal cellular proliferation is a hallmark of carcinogenesis. This complex process is under control of numerous signal transduction pathways. Several laboratories studied in vitro cellular proliferation as an end-point for ELF-EMF effects. Alterations in proliferation were observed in a number of laboratories using a variety of exposure conditions (magnetic fields strengths of 1000 to 5000 µT) and cell lines(225-227). Two studies(228, 229) did not confirm an earlier report(227) of increased colony growth for cells exposed to 60 Hz magnetic fields, although one study(229) used a similar experimental protocol. Another study, which used several methods for independently assessing proliferation, reported increased growth over an exposure range of 50 to100 Hz and 100 to 700 µT (230).

Disruption of the normal circadian rhythm of melatonin, a hormone produced by the pineal gland, has been postulated as a possible mechanism whereby ELF-EMF exposure might increase risk for breast cancer (120). Studies in a human breast cancer cell line(231) showed that cellular proliferation in vitro was decreased by treatment with physiological levels of melatonin; exposure to a sinusoidal ELF magnetic field (1.2 µT) could overcome this effect. These studies were extended and the anti-proliferative effects of tamoxifen (an anti-cancer therapy) were also reported to be reversed by a 1.2 µT field(232). Another laboratory presented similar findings(233). The original laboratory also reported finding comparable effects using a second human breast cancer cell line(234) and a human glioma cell line(235). There is some concern about the experimental design of these studies and further work is underway. In addition, because the observed effect is small, the importance of these findings for human health is not clear(236).

Numerous investigations have examined ELF-EMF exposure effects on markers characteristics of cellular differentiation (e.g. matrix protein synthesis; cell surface characteristics; cell morphology, size and orientation). Several of these studies demonstrated a role of electric fields in affecting cellular behavior. Two investigations of alterations in matrix protein production studied effects of electric fields(237, 238) and found a positive correlation between dose and the differentiated state of the cells. Studies examining ELF-EMF effects on alterations of cell surface markers used a variety of cell types. In two ofthese investigations, the observed cellular effects were attributed to the induc ed electric fields(239, 240). Exposure to 60 Hz electric fields was also found to suppress formation of osteoclast-like cells in marrow culture(241).

Biophysical Theory

The physics governing the interactions of ELF-EMF with matter were elucidated over a century ago and succinctly stated in the Maxwell equations. Years of successful application of these principles for practical advances have left little doubt about our ability to understand and predict electromagnetic biophysical phenomena when details of the system and fields are completely described. Given the complexity, dynamics and organization in living organisms, it is difficult to apply this knowledge. Living organisms function through the use of biochemical and electrical signals carefully controlled by the organism’s structure. Early attempts to explain the biological effects of ELF-EMF focused on simple application of electromagnetic theory to calculate the forces on biological molecules and the energies transferred to them by weak ELF-EMF. The extremely small magnitude of these interactions led many investigators to conclude that they would not occur at normally encountered field strengths. This has not fundamentally changed; calculations still strongly suggest that the small electric fields and magnetic fields associated with ELF-EMF in environmental settings cannot be expected to supply, by themselves, the energies necessary for chemical changes.

The complexity and structure of biological systems make uniform application of these findings difficult. For example, even very small fields might act as control signals to modify processes that depend on metabolically supplied energy. This would be analogous to extremely weak radio signals, such as those transmitted over thousands of miles, that control locally supplied energy or power a loud-speaker or a large-screen television set. The exact nature of biological signal processing systems and their susceptibility to control by time-varying ELF-EMF is of continuing interest. Biological systems contain complex feedback loops and amplification sequences in which very small changes at one point may ultimately lead to very large changes further along the communication chain. In considering ELF-EMF changes on the nature of biological signals, it is essential to recognize that all aspects of a field (frequency, amplitude and pattern) may be involved. These considerations make definitive statements based upon biophysical theory difficult to apply to living organisms.

Several mechanisms for explaining ELF-EMF effects on biological systems have been proposed. One set of theories(242-248) predicts effects of ELF-EMF on chemical reactions due to resonances that depend on complex interactions between constant and oscillating magnetic fields. There is limited experimental support for these theories (12); the validity of the assumptions used in the theories has been questioned(249).

Modification of the transfer of electrons from one molecule to another has also been suggested as a theoretical mechanism for the effects of ELF-EMF (250-255). However, the energies involved in electron binding are many orders of magnitude larger than those contained in weak, externally applied electric fields or magnetic fields (256-260) making these theories difficult to accept.

It is also possible that ELF-EMF could interact with magnetic particles in human cells (261-264). However, work with this theory(263-265) would suggest that such effects can occur only with large magnetic fields and are not applicable to the normal human environment; these conclusions may be premature(12, 266).

Magnetic fields are capable of altering specific types (e.g. radical pair formation) of chemical reactions(267-273). Potential effects of ELF-EMF have been predicted by analytical work(274-278). Such reaction effects have been shownfor strong fields(279), but there are few studies of the effects in biological systems with moderate to low field intensities.

Biochemical and biomechanical processes are generally dynamic. It has been suggested that rather than causing changes in the usual state of the system, ELF-EMF may induce slight changes in the frequency of events that trigger other processes, especially for effects on chemicals that oscillate within cells and between cells and their environments(250, 277, 280-286). Both theoretical(287-291) and biological(292-294) studies exist that support this suggestion. However, there is open debate about whether this phenomenon is applicable for ELF-EMF exposures that are generally found in the human environment.

All of the theories for biological effects of ELF-EMF suffer from a lack of detailed, quantitative knowledge about the processes to be modeled. Nevertheless, theoretical models are useful, even in the absence of critical data, because they can indicate what data are needed, suggest previously uncontemplated experiments, suggest bounds on risks under defined situations and provide nonlinear methods of analysis of critical data based upon presumed mechanisms. The current biophysical theories for ELF-EMF would suggest little possibility for biological effects below exposures of 100 µT. However, considering the complexity of biological systems and the limitations required by the assumptions used to mathematically model these theories, this finding has to be viewed with caution.

Path: Home>Education>Healthy Home Information>Do Electric and Magnetic Fields Pose A Health Risk?