Electric and magnetic phenomena have been recognized since ancient times, but the means to measure, generate, control, and use the phenomena to develop practical devices became adequately understood only in the past 200 years. In little more than a century since the invention of the light bulb, society has become dependent on electricity and the myriad devices that are driven by it. It is relied on in nearly every aspect of everyday life. Electrically driven devices ease the workload in factories, farms, offices, and homes. Electricity is used to control the indoor climate, to clean clothes, to store and prepare food, and to perform many other tasks in the home and workplace. Electric devices are used in such diverse applications as medical imaging, cardiac pacemakers, cancer therapy, and communication. So widespread is the use of electricity that it is impossible to avoid exposure to the electric and magnetic fields produced in the transmission and distribution of electric power or to those fields generated by devices used in homes and workplaces.
Although the hazard of shocks and burns from coming into contact with energized electric conductors has been known since the first application of electric current, only during the past 15 years or so has public concern been raised about the more subtle effects of exposure to the fields generated by electric devices. To help determine whether a potential health risk from exposure to low-strength, low-frequency electric and magnetic fields might exist, the U.S. Department of Energy (DOE) asked the National Academy of Sciences to conduct a review. In response to the request, the Committee on Possible Effects of Electromagnetic
Fields on Biologic Systems was convened by the Board on Radiation Effects Research of the National Research Council's Commission on Life Sciences. The committee was to perform the appropriate review and report to the board on its findings.
SCOPE OF THE STUDY
The committee was asked to focus its attention on electric and magnetic fields typical of household frequencies and on the possible adverse health effects of cancer, reproductive and developmental abnormalities, and neurobiologic dysfunction, such as learning and behavioral disabilities. Those effects are the health-related end points most often suggested to be associated with exposure to power-frequency electric and magnetic fields. The committee also was asked to examine the scientific evidence for the effects of the electric and magnetic fields of household frequencies on biologic systems to determine if sufficient scientific data of adequate quality exist to perform a health risk assessment.
The DOE charge to the committee included the following:
Review and evaluate the existing scientific information on the potential effects of exposure to electric and magnetic fields on cancer incidence, reproduction and development, and learning and behavior.
Critically examine epidemiologic and laboratory data relating to those topics and assess potential health effects.
Focus on electric-and magnetic-field frequencies and exposure modalities found in residential settings.
Produce a report that contains a review of pertinent information on the effects of electric and magnetic fields, identification of research areas in which data are needed to better understand any potential health hazard, and recommendations for research in those areas and strategies for implementing research that would enhance understanding. If data of appropriate quality are available, include a health risk assessment of power-frequency electric-and magnetic-field exposures.
AREAS OF CONCERN
Concern about the possible health effects of exposure to electromagnetic fields first arose when military personnel were exposed to fields of relatively high strength from high-frequency radar systems and video screens during World War II. Since then, claims have risen of adverse health effects purportedly associated with high-frequency sources, such as radar units used by police, antenna systems used by the military, cellular phones used for communication, and microwave ovens and other appliances used in homes. Recently, attention has also focused on the potential for adverse health effects of low-frequency
sources, such as transmission and distribution lines and electric appliances, including shavers, hair dryers, water beds, and electric blankets. It must be emphasized, however, that the effects of exposure to different sources of electric and magnetic fields can be quite different, depending on their frequency and strength. Possible effects of the fields generated by high-voltage transmission lines or electric blankets, operating at 60 hertz (Hz),1 might be quite different from those generated by high-frequency (megahertz or gigahertz)2 devices.
Of primary interest to this committee is the concern about sources of low-frequency electric and magnetic fields associated with the generation, distribution, and use of electric power, including transmission lines, substations, distributions lines, and numerous electric devices ranging from personal computers to electric clocks.
Questions of the possible adverse health effects of exposure to electric and magnetic fields from 60-Hz power lines were first raised by Wertheimer and Leeper (1979). They reported epidemiologic data suggesting an association between the configuration of power lines near homes and the incidence of leukemia and other types of childhood cancer. Similar studies have been published in succeeding years in the United States and in numerous other countries. The results of these studies have increased the scrutiny of the possible association between raised levels of electric and magnetic fields in residences, as one site of exposure, and the incidence of cancer—the adverse health effect of most concern. Much of the early laboratory research on biologic effects of very low-frequency electric and magnetic fields focused on the study of electric fields, but results of epidemiologic and other studies have caused a gradual shift of interest toward magnetic fields as a possible cause of disease.
DEFINITIONS AND DESCRIPTIONS OF TERMS
Electric and Magnetic Fields
The term ''electromagnetic field," which is commonly used in the literature, applies to alternating fields. The electric and magnetic components of the fields generated by moving charged particles are formally linked and mathematically described by a set of coupled differential equations called Maxwell's equations. Electromagnetic fields are characterized by their wavelength, λ (expressed in meters), and their frequency, f (expressed in hertz); the frequency and wavelength
are related by the velocity of light, c, as λ = c/f. The full range of frequencies (or wavelengths) of natural and anthropogenic electromagnetic fields is described as the electromagnetic "spectrum." The electromagnetic spectrum, described in detail in textbooks on radiation, ranges from extremely low frequencies (ELF),3 which are associated with common household electric current (50-60 Hz),4 to radio waves (106-1010 Hz), microwaves (1010-1012 Hz), infrared radiation (1012-1014 Hz), visible light (1014 Hz), ultraviolet radiation (1015 Hz), and very high frequencies and very short wavelengths of X-rays and gamma rays (> 1017 Hz). In this list, which represents a hierarchy of increasing electromagnetic (or photon) energy, only the radiation with frequencies greater than about 1015 Hz is capable of ionizing atoms and molecules (i.e., producing charged particles from the atoms and molecules with which it interacts). Ionizing radiation (e.g., X-rays and gamma rays) is a well-understood source of damage to biologic systems through the reactions of the products of ionization with critical cellular components. ELF radiation, on the other hand, is nonionizing; it does not have sufficient quantal (photon) energy to produce ionization in the manner of high-frequency radiation, and its mode of interaction, if any, with molecules and biologic systems at low field strengths is speculative. Most equipment used for the generation, transmission, and distribution of electric power in the United States generates ELF (60-Hz) electric and magnetic fields. The components of the electric utility system that generate such fields include power plants (generating stations), which produce the electricity; high-voltage transmissions lines, which carry the electricity to major population centers; substations and their transformers, which reduce the voltage to levels suitable for distribution within a population center; distribution lines (distribution primaries), which commonly carry power along residential streets; distribution transformers, which reduce the voltage to amounts suitable for use in homes; and distribution secondaries (service drops), which carry electricity to individual residences. Transmission and distribution lines are commonly called "power lines," but the term can also include service drops.
Electric power that is used to operate devices in the home and workplace is also associated with the production of electric and magnetic fields. As electric charges move to produce a current, magnetic fields are created. An electric appliance connected to a source of electricity might have an electric field present even when it is turned off. When turned on and operating, a magnetic field is also present.
The exposures of interest in this report are limited to ELF electric fields (expressed in volts per meter) and magnetic fields (expressed in tesla (T) or gauss (G), where 1 T = 104 G) associated with household use of electricity. Because
the range of magnetic fields encountered is usually quite small, the fields are generally described in units of microtesla (1 µT = 0.000001 T) or milligauss (1 mG = 0.001 G). For example, the earth's geomagnetic field is a static field of about 50 µT (0.5 G), and a current of 50 amperes (A) in a straight wire produces a magnetic flux density (magnetic field) of 100 µT at a distance of 10 centimeters (cm). Although household alternating current in the United States has a frequency of 60 Hz, other relatively low-frequency electric and magnetic fields can be induced when the current is used to operate appliances, such as electric razors, hair dryers, video-display terminals, and dimmer switches.
Electric fields from direct exposure to high-voltage power lines and electric appliances induce current on or just within the surface of an exposed person's body. Because the electric fields are perturbed by the tissue conductivity, the fields inside the body are very weak. On the other hand, magnetic fields pass through the body and can induce electric currents throughout the body. Magnetic fields can pass through most common building materials, including thin sheets of metal. However, magnetic materials, such as iron and some metallic alloys, can serve as convenient paths for the conduction of magnetic fields and can be used as magnetic shields in some cases. People can be shielded quite easily from exposure to electric fields, because most materials possess sufficient conductivity to attenuate the fields.
Although electric and magnetic fields are quite different in character, time-varying fields are generally described together as electromagnetic fields. As noted above, time-varying electric and magnetic fields are formally linked and described mathematically by Maxwell's equations. Through coupling, a time-varying magnetic field induces an electric field and vice versa. However, in the limit of unchanging (static) fields, the electric and magnetic fields are independent. At the low frequencies associated with electric-power use, the coupling is extremely weak, and electric fields and magnetic fields can be considered independent to an excellent approximation. In this report, the term electromagnetic field (EMF) is used when the electric and magnetic fields are substantially linked, usually only for high-frequency fields.
Very-low-frequency electric and magnetic fields are known or suspected to interact with biologic systems in a number of ways. Some biologic effects at high field strengths, such as nerve stimulation and tissue heating, are well understood and have been used to set standards for occupational and public exposure to fields. Other reported effects, particularly at low field strengths, are not as well understood; those include effects on cell metabolism and growth, gene expression, hormones, learning and behavior, and promotion of tumors. The reality of all those effects is the subject of scientific debate and an issue for discussion in this report.
The term biologic effect is intended to be a neutral term; it implies no judgment about whether an effect is good or bad. Some biologic effects of electric and magnetic fields have already been found to be beneficial; for example, the ability of fields to stimulate tissue and bone growth has been known and used for a number of years to speed the healing of fractures and burns. Other effects might be harmful.
The committee has focused its attention on three kinds of adverse health effects that have become the chief concerns of the public and of health officials. These are cancer, primarily childhood leukemia; reproductive and developmental effects, primarily abnormalities and premature pregnancy termination; and neurobiologic effects, primarily learning disabilities and behavioral modifications. Each of those effects has been reported in epidemiologic studies to be associated with exposure to some indirect estimates of the strength of power-frequency electric and magnetic fields. Childhood leukemia has attracted the most attention because of studies conducted in Denver, Los Angeles, and the Nordic countries that reported an increased risk of the disease in association with various indicators of exposure to electric or magnetic fields.
Exposure Assessment in Epidemiologic Studies
Determining the amounts and types of environmental agents to which an individual is exposed is often difficult. For example, persons exposed to environmental (secondhand) tobacco smoke might also be exposed to byproducts of tobacco as past or present smokers. That example is similar to the case of exposure to power-frequency electric and magnetic fields where there are multiple opportunities for exposure and almost no way to reconstruct the history of the exposure sources associated with any eventual adverse effect.
Multiple sources of possible adverse health effects are also difficult to separate in epidemiologic studies. For example, studies intending to determine if lead in the residential environment is a hazard can be altered by the fact that study subjects might be exposed to lead not only in their homes but also from outside air or in their workplaces. An individual who is exposed to one agent thought to be associated with an adverse effect might also be exposed to other agents that could contribute to the risk of the disease. Lung cancer might be assumed to be caused by an individual's smoking, even though the individual was exposed instead or in addition to another potential causative agent, such as radon or asbestos. Determining exposure to power-frequency electric and magnetic fields can be confounded similarly, and thus it is difficult to associate accurately the purported exposure with health effects in individuals.
Several dissimilar methods have been used to assess exposure to electric
and magnetic fields and adverse health effects in epidemiologic studies. In only a small number of studies, actual electric-field or magnetic-field measurements were made of an individual's exposure. But even then, the measurements were made of present-day field strengths rather than the strengths to which individuals were exposed when they developed health problems or, for cancer, over the years when the cancer might have been induced. One study used records of power-line loads to calculate the average prediagnosis magnetic field. Several studies used the distance from the power line to judge whether a residence should be considered in the high-exposure category. In a number of the key studies used in this report, the association between power-frequency electric and magnetic fields and adverse health effects was made on the basis of exposed persons living in houses with power transmission or distribution lines in particular locations or configurations. The authors of those studies used the terms "wiring configuration" or "wire code" to describe a particular configuration of wires nearby a residence; the configurations were defined in such a way that they were expected to correlate approximately with electric-and magnetic-field measurements.
The formal techniques used for risk assessment were developed primarily for assessment of risk from chemicals in the environment. Risk assessment is generally divided into four distinct steps: hazard identification, dose-response assessment, exposure assessment, and risk characterization (NRC 1983). The last step implies the quantification of the proportion of persons in a population who might be adversely affected. For example, one in 1 million persons might be affected by exposure to a given agent at a certain concentration. Quantitative risk assessments have been the topic of discussion of numerous authors. One recent publication developed for the lay reader by Morgan (1995) describes quantitative assessments that place an upper limit on the lifetime risk of cancer due to exposure to electric and magnetic fields and compares that risk with other common risks.
Risk assessment can be complete (i.e., encompassing all four steps described above and leading to a quantitative risk estimate), or it can be partial (i.e., using only some of the steps). The completeness of risk assessment is governed by the available data as well as the purpose of the risk assessment.
At each step of the risk-assessment process, the assessor makes judgments about the "weight" of particular pieces of the evidence. An epidemiologic study might be deemed well designed and authoritative because the population studied is large, and thus amenable to meaningful statistical analysis, and because the study is free of confounding by other exposures. Such an epidemiologic study will be given more weight than another study that fails in some of these aspects. Laboratory tests using cell systems or whole animals will be given more weight if they have been replicated and validated by other investigators in other laboratories.
Experiments that appear to show an effect but have no biologic or physical explanation might be given less weight. In weighing the evidence, some assessors might weigh evidence differently than others in reaching their conclusions.
At the end of the risk-assessment process, the body of evidence is weighed together to reach an overall assessment of a possible hazard. If the results from several areas of research (e.g., epidemiologic studies, tests in cell systems, or whole-animal studies) are consistent and have been replicated and if a biologically plausible mechanism of action for the effect is evident, the evidence for the effect is given great weight. In contrast, a body of evidence that includes inconsistent and conflicting results, no replication of results, and effects that are often at the threshold of detection might be given little weight in reaching a conclusion.
In this report, the committee attempts to explain carefully why individual studies are given more or less weight and how the body of evidence is weighed in reaching its conclusions.
SOURCES OF EXPOSURE
Electric Power Lines
Electric transmission lines are generally built on "rights-of-way" and vary in voltage, height, and configuration of suspension from the towers. Electric power lines commonly observed in typical neighborhoods are not transmission lines as described above, but are lower-voltage distribution lines. The magnetic fields immediately under distribution lines are generally of the order of 0.5 µT (5 mG), although in some densely populated areas, fields as high as 5.0 µT (50 mG) have been measured. The fields decrease rapidly as the distances from the power lines increase.
Electric substations are installations where the voltages used with transmission lines are stepped down to lower voltages used with distribution lines. Electric and magnetic fields produced by substation equipment are generally not appreciable beyond the substation boundaries, but the fields can be somewhat stronger near them than in other parts of the neighborhood, because the power lines converge at the substation and might be closer to the ground as they go in and out of the substation.
Electric appliances in residences and workplaces are all potential sources of exposure to electric and magnetic fields. The strengths of magnetic fields vary widely: the magnetic fields from household appliances might be as low as a few tenths of a microtesla or as high as 150 µT (1,500 mG). The magnetic fields from all sources tend to decrease rapidly as the distance from the source increases. It has been noted that the earth's natural static magnetic field of about 50 µT (500 mG)
is about 100 times stronger than residential magnetic fields normally associated with the alternating current (ac) of power lines and electric appliances; exposure to the earth's field is constant, and exposure to constructed alternating fields is intermittent. However, that comparison might be of little relevance, because it is not known which aspect of the magnetic fields could be of significance to health.
Some transportation systems, including subways and intercity trains, operate on ac current and generate ac electric and magnetic fields. Measurements in the Baltimore-Washington commuter trains indicate exposures to magnetic fields at 25 Hz with peak strengths as large as 50 µT (500 mG) in the passenger areas at seat height. The fields vary greatly with the position in the car as well as with the particular type and model of the car; detailed measurements indicate an average field of approximately 12.5 µT (125 mG).
WHY KNOWLEDGE ABOUT ELECTRIC AND MAGNETIC FIELDS IS IMPORTANT
Because electricity is used so extensively and sources of electric and magnetic fields are everywhere, every person in modern society is unavoidably exposed to them. Thus, understanding any biologic effects that might be associated with exposure to electric and magnetic fields is fundamentally important.
It is easy to see why so much attention has been given to the possibility that power-frequency electric and magnetic fields are associated with adverse effects. People who study how individuals respond to risk have learned that certain types of risks elicit stronger responses than others (Slovic 1987). One of the health effects that has been associated with exposure to electric and magnetic fields is an especially dreaded one, namely, cancer. Children, a group of particular concern, are a reported target for leukemia and possible reproductive and behavioral effects. The sources of the reported electric-and magnetic-field risks are largely imposed on people and not under their control. Furthermore, the fields that are the source of the reported risks are invisible and mysterious to many. All these factors cause many people to respond with concern and anxiety to potential risks associated with exposure to electric and magnetic fields (MacGregor et al. 1994).
When such an indispensable resource as electricity is reported to be associated with adverse health effects, it is not difficult to understand why concerns have arisen. It is also apparent that the potential health effects are only one part of the concern. If extreme steps are taken to reduce exposure to power-frequency electric and magnetic fields, large sums of money will need to be expended (e.g., to bury transmission and distribution lines, to redesign residential wiring and electric appliances, or to retrofit existing ones). Every citizen would contribute to the implementation of these measures through higher utility bills and greater
personal expenditures on appliances. If the concerns are misplaced, but measures are taken to satisfy the public, enormous costs would be incurred unnecessarily. If the concerns are real, those who have called attention to them will have made an important contribution to public health. The charge to this committee is to assess the scientific information that will aid the government agencies, public utilities, and the general public to conduct more fully informed discussions regarding the potential health risks of exposure to power-frequency electric and magnetic fields and ultimately to help shape the appropriate policy.
There have been numerous studies of the possible health effects of nonionizing radiation and specifically of the very-low-frequency electric and magnetic fields on which the committee has focused. Perhaps the most quoted of the national and international reviews of the possible adverse health effects from exposure to power-frequency electric and magnetic fields is that published by the Oak Ridge Associated Universities in 1992 (ORAU 1992). The ORAU report concluded that there was "no convincing evidence in the published literature to support the contention that exposures to extremely-low-frequency electric and magnetic fields (ELF-EMF) generated by sources such as household appliances, video display terminals, and local power lines are demonstrable health hazards." They noted that the results of their review did not justify an expansion of the national research effort to investigate the health effects of exposure to electric and magnetic fields and that "in the broad scope of research needs in basic science and health research, any health concerns over exposures to ELF-EMF should not receive a high priority."
In 1989, the Office of Technology Assessment published a 103-page background paper prepared for them by Carnegie Mellon University's Department of Engineering and Public Policy and entitled "Biological Effects of Power Frequency Electric and Magnetic Fields" (Nair et al. 1989). The report called for more research on the potential effects of power-frequency electric and magnetic fields on the central nervous system and on the possibility of cancer promotion. Nair et al. (1989) recommended a policy of "prudent avoidance," which was defined as avoiding exposure by formulating strategies that were prudent from the standpoint of cost and the best understanding of risks.
Great Britain's National Radiological Protection Board (NRPB) published reports on the biologic effects of nonionizing EMF and radiation. These reports contain summaries of experimental investigations. The board's assessment of power-frequency electric and magnetic fields and the risk of cancer was published in 1992 (NRPB 1992). In summary, no firm evidence of a carcinogenic hazard was found from exposure of paternal gonads, the fetus, children, or adults to ELF electric and magnetic fields. A follow-up to that report reaffirmed its earlier conclusions (NRPB 1994).
A panel on EMF and health was established by the Victorian government of Australia in 1991 to review the range of approaches that are taken in relation to power-line fields and, when appropriate, to recommend appropriate courses of action. As a part of its activities, the panel reviewed the literature on health effects of exposure to low-frequency electric and magnetic fields. In its report (Peach et al. 1992), the panel concluded that the uncertainties in the data were so great as to preclude the possibility of establishing an association of risk with exposure. Although it noted that such fields have not been proven scientifically to be harmful, the panel recommended adoption of a policy of prudent avoidance.
ORGANIZATION OF THIS REPORT
In Chapters 2, 3, 4, and 5 of this report, the committee summarizes the research on the biologic effects of exposure to power-frequency electric and magnetic fields. Emphasis is placed on the data that pertain to the focus of the charge, studies of electric and magnetic fields at strengths and frequencies typical of residential settings. In Chapter 2, exposure is discussed: how it is measured, how it is estimated in epidemiologic and laboratory studies, and what types of exposures occur in various environments. This chapter also describes the way ELF magnetic fields interact with biologic systems. In Chapter 3, the committee summarizes what has been learned about the biologic effects of exposure to power-frequency electric and magnetic fields from cellular and molecular studies in the laboratory. Chapter 4 focuses on studies involving whole animals. Chapter 5 describes the epidemiologic data relating to the possible effects of power-frequency electric and magnetic fields on the development of cancer, reproductive and developmental abnormalities, and learning and behavioral disabilities. In Chapter 6, the evidence presented in the preceding chapters is evaluated and synthesized, and the committee's findings and conclusions are presented concerning the possible health risks of exposure to power-frequency electric and magnetic fields. The committee's recommendations for areas of possible research appear in Chapter 7. In chapters in which numerous tables of data were included in the evaluation, the tables were placed in Appendix A to improve the readability of the text. Appendix B contains a detailed description of the wire codes used in the various epidemiologic studies.
In the course of its work, the committee examined a vast amount of scientific literature. The criterion for including a paper in the deliberations was that it be published in a recognized peer-reviewed journal (unless otherwise noted). Technical reports and abstracts of papers delivered at scientific meetings were read as background information, but were not relied upon in forming judgments. The U.S. Environmental Protection Agency and Electric Power Research Institute publications used in Chapter 2 and Appendix B are notable exceptions to this rule; these technical reports are a major source of exposure data and were relied upon after careful review by committee members. As in any committee deliberation,
committee members individually and collectively ascribed different weights to the importance of the papers evaluated and the quality of the research therein. The committee members often had specialized criteria based on their individual disciplines for evaluating the literature; the criteria are described in the report.