BIOLOGICAL EFFECTS OF ELECTROMAGNETIC FIELDS

M. A. Stuchly
Department of Electrical and Computer Engineering,
University of Victoria
Victoria, BC,  CANADA, V8W 3P6

Abstract: Electromagnetic fields are present in our environment, and are used in several medical applications. The most prevalent environmental fields are at power line frequencies (50 or 60 Hz), and at radio frequencies (300 MHz - 10 GHz), including the frequencies associated with wireless communication. Understanding of interactions of electromagnetic fields with human tissue is essential for prevention of harmful effects, as well as the use of these fields in diagnostic and therapeutic medical applications.  Current state of knowledge of the basic interaction mechanisms, dosimetry (determination of the fields in tissue) and experimental evidence of biological effects are reviewed.

INTRODUCTION

Biological effects of electromagnetic fields have been the subject of active research, and at times of public concern regarding their safety. Several reviews have been published within the last decade concerning extremely low frequency (ELF) fields [1-4], and radio frequency (RF) fields [5-10]. Exposure to electromagnetic fields results in electric fields and currents in tissues. Depending on the field characteristics, particularly its frequency and strength, the tissue induced fields may result in an interaction leading to a biological effect. Even the same effect may be harmful or beneficial. For instance, for a person working in strong magnetic fields in close proximity of a power line, stimulation of peripheral nerves may result in an injury. On the other hand, pulses of a strong magnetic field are used in clinical settings to stimulate nerves in the human cortex for diagnostic purposes. 

In this overview, electrical properties of tissues are defined and their role in interactions with electromagnetic fields is outlined, methods of evaluation of induced fields in tissue (dosimetry) are explained, an overview of biological effects is given, and unresolved issues are highlighted. Since the dosimetry and biological effects are considerably different at ELF and RF, they are described in separate sections.

TISSUES IN ELECTROMAGNETIC FIELDS

Electrical Properties

Human tissues are characterized by their permittivity defined as:

                                                        (1)

where is the dielectric constant, which defines the energy stored in the material, and   is the power loss (due to material conductivity and dielectric relaxation [11]. Both components vary with frequency and depend on the tissue type. Biological tissues are non-magnetic, and their permeability is the same as that of air.

Physical Interaction Mechanisms

External fields induce fields and currents in the human (and other) biological bodies. At ELF, the magnetic fields in tissue are the same as the exposure fields, but at RF both electric and magnetic fields in tissue are altered. Internal fields are highly non-uniform within the human body. The magnitude and the spatial distribution of the fields in the body depends on the body anatomy (including shape and size), and several parameters of exposure field, most notably the frequency, and orientation with respect to the body.

At ELF, the human body significantly perturbs an external electric field. To the contrary, a magnetic field is not altered, because of the quasi-static nature of the fields and the permeability of the human tissue being the same as that of free space. At RF, electric and magnetic fields are coupled, and the body or any other object perturbs the electromagnetic field.

Electric and magnetic fields at ELF, including 50 and 60 Hz, are well-known to cause stimulation of some tissues such as nerves, muscles and the heart, once the electric field in the tissue exceeds a threshold value.  Membranes surrounding nerve and muscle cells can be excited once a membrane potential threshold is reached.  The threshold action potential (the potential across a membrane that initiates propagation along the nerve or muscle cell) depends on the type and dimensions of the cell, and the frequency and waveform of the imposed signal.  Cell excitation and action potential propagation are complex non-linear processes. However, the behavior prior to the onset of excitation can be described by a cable equation, and thresholds have been well established. The excitation threshold for a given neuron id defined by the derivative of the electric field along the neuron. Thus, even in a spatially uniform electric field in tissue can excite a neuron. depending on its shape.  Stimulation of neurons in the retina results in visual phosphenes.

At RF, the induced electric field causes tissue heating, which is proportional to the square of the electric field in tissue and its conductivity. A dosimetric measure that has been widely adopted is the specific absorption rate (SAR) defined as "the time derivative of the incremental energy (dW) absorbed by, or dissipated in an incremental mass (dm) contained in a volume element (dV) of a given density ": 

                           (2)

For harmonically varying electromagnetic fields, eq. (2) can be expressed as:

                             (3)

where σ is the tissue conductivity in S/m, ε_o is the dielectric constant of free space   (ε_o = 8.85 x 10^-12 F/m), ε" is the loss factor, ω = 2πf, f is the frequency in the Hz, E_i is the peak value of the internal electric fields in V/m.  The SAR is expressed in watts per kilogram (W/kg) or their derivatives (i.e., mW/g).  The average SAR is defined as a ratio of the total power absorbed in the exposed body to its mass.  The local SAR refers to the value within a defined unit volume or unit mass, which can be arbitrarily small.

In principle, the SAR can be considered both as a thermal and a non-thermal dosimetric measure.  For thermal biological effects that depend on the electric field strength, the SAR can be considered as an appropriate measure.  The rate of temperature increase is directly proportional to SAR, and equal to:

                                                            (4)

where T is the temperature, t is the time and C is the specific heat capacity.

DOSIMETRY

Human Body Models      

Figure 1. External view, major organs and blood vessels and skeleton of a model used in electromagnetic dosimetry.

Until the last few years, the human body was approximated by relatively simple models such as spheroids, ellipsoids, circular cylinders, or at best as composites of a few cylinders with a body-like shape with a few major organs represented in a similarly simple fashion.  Currently, a number of laboratories have developed heterogeneous models of the human body with an anatomical shape and numerous tissues identified.  Most of these models have been developed by computer segmentation of data from magnetic resonance imaging (MRI), and allocation of proper tissue type.  Special care has been taken to make these models anatomically realistic. Figure 1 illustrates a typical model.

Extremely Low Frequency Fields

Various computational methods have been used to evaluate induced electric fields in the high-resolution models for exposures to the externally applied electric and magnetic fields. Because of the low frequency of interest exposures to the two fields should be considered separately and the induced vector fields added, if needed. Detailed description of the methods is given in [3].

Extensive data are available on induced electric fields in various tissues [3]. Examples of such data are given in Figs. 2 and 3, for exposures to a uniform magnetic field and electric field, respectively.

Computed electric fields in tissues can be used for comparison with the levels that have been shown to cause biological effects. For effects that are related to the tissue electric field, there is no difference between the electric and magnetic field.

Figure 2. The average and maximum induced electric fields in human body tissues. Exposure to a magnetic field of 1 mT (Tesla) at 60 Hz, in three orthogonal orientations.

Figure 3. The average and maximum induced electric fields in human body tissues. Exposure to an electric field of 1 kV/m at 60 Hz, for the body in free space. 1.4 cm above ground and in contact with perfect electric ground.

Radio Frequency Fields

Numerical methods and realistic high-resolution models of the human body have recently been used to compute SAR in tissue. In addition to the whole body dosimetry, localized SAR values have been evaluated for various handheld telephones. The most often used numerical method is the finite-difference time-domain method (FDTD). Its main advantage is the compatibility with the heterogeneous voxel model of the human body and computational efficiency.

The FDTD method and an anatomically correct model of the human body were recently used to compute the whole body average and averages in horizontal body cross-sections for adults and children in free space and in contact with a perfect conductor ground plane [12].  Data were obtained for frequencies from 1 MHz to 1 GHz. For exposures with the E-polarization (electric field along the long body axis) in free space, a resonant frequency of 65 MHz was obtained and of 35 MHz for the model grounded.  Both of these values are in good agreement with previously predicted values for highly simplified models.  There are currently in progress further computations.

BIOLOGICAL EFFECTS

Extremely Low Frequency Fields

There is well-established evidence that strong fields in tissue cause neural stimulation. However, vast majority of available data on stimulation thresholds have been obtained with electrodes of limited dimensions and given in terms of current or current density. It can be estimated that fields of the order of a few V/m in tissue are required for the stimulation. As evident from a comparison of the dosimetry data in Figs. 2 and 3, with typical environmental exposure levels neural stimulation is not likely to occur. One example where it may occur, is the peripheral nerve stimulation due to the fast switching magnetic fields in some MRI scanners.

Experimental evidence and thresholds have been determined for stimulation of the visual system [3].  The lowest threshold magnetic field has been found equal to 10 mT (in darkness) at 20 Hz.  The threshold is somewhat higher when illumination increases and increases for higher frequencies.  Visual stimulation resulting in phosphenes has also been obtained with current applied by electrodes placed on the head. 

There is substantive evidence that low-frequency fields accelerate bone-healing [2].  The physical interactions are related to induced electric fields of the order of 30 mV/m at the most effective frequencies of 10 to 30 Hz. These frequencies are associated with those observed in live animals.  Weaker fields of 0.1 – 1 mV/m elicit transient changes in bone cell preparations [2].  Detailed biophysical mechanisms are only partly understood. 

There is evidence that low-frequency fields affect signal-transduction systems in cellular preparations (this does not by itself indicate adverse effects).  However, the interaction mechanisms are very incompletely understood.

There have been several toxicology studies, and none of them indicated carcinogenic effects. Two studies of mammary cancer promotion in mice produced contradictory results. Overall evidence that relatively low strength fields cause detrimental effects is lacking. On the other hand, the epidemiological studies indicate correlation between children exposure and leukemia, but not other cancers [2].

Radio Frequency Fields

Considerable data on biological effects of RF fields have been accumulated and considered in development of health protection standards [6, 8, 10].  The reported biological effects observed for various animals correlate relatively well with the whole-body-average SAR, but not with the power density when different frequencies are considered.  The correlation with the whole-body average SAR is not perfect, since the distribution of the SAR is in most cases highly non uniform within the exposed body and can be considerably different for the same value of the average SAR. Threshold levels in terms of the whole-body-average SAR have been established for various effects.  The average SAR correlates reasonably well with increases in body temperature.  Increases in body temperature of an order of 5^o C cause numerous malformations, temporary infertility in males, brain lesions, changes in blood chemistry.  Even relatively small (approximately 1^o C) increases in body temperature have been observed to lead to such effects as altered production of hormones, suppressed immune response, and changes in behavior.  There are no experimental data on the SAR levels required to increase the body temperature for humans except from the studies of exposure to strong magnetic fields during magnetic resonance imaging (MRI).   Statistically significant increases in the body and peripheral temperatures were observed during MRI procedure [10]. Mathematical modeling and extrapolation from the animal data indicate that about 1 W/kg, particularly at frequencies close to the resonant frequency, i.e., about 70 MHz, is sufficient to produce an increase in body temperature in humans by about 1 degree centigrade.

Effects on the hematological and immunologic systems have been observed to occur at the average SARs > 0.5 W/kg.  A variety of changes have been observed in animals exposed to RF fields under conditions that have not resulted in whole-body temperature increase.  Localized heating and "hot spots" appear to affect immune response.    RF exposure at high SARs (above 15 W/kg) causes teratogenic effects; lower level exposures (above 2 W/kg) can cause other reproductive effects, such as a decreased fetal weight.  These effects are not related to a direct genetic damage but to the thermal load.The effects of RF exposure on the central nervous system and behavior have been controversial. Thresholds for behavior disruption occur at the RF energy deposition rates between 25% to 50% of the resting metabolic rate of the animal.  The behavioral alterations are reversible with time.

Effects of electromagnetic radiation on the three major eye components essential for vision: cornea, lens and retina have been investigated. It was established that lens opacities could form after exposure to microwaves above 800 MHz but below about 10 GHz.  Cataract induction requires sufficiently long exposures at an incident power density exceeding 100 mW/cm². SARs in the lens large enough to cause temperatures in the lens greater than 41^o C are required.  Effects on the retina have been associated with rather high levels of microwave radiation, above 50 mW/cm².

There have been very few long-term chronic animal studies and studies related to cancer. Such studies are currently in progress.

CONCLUSIONS

Physical interactions of electromagnetic fields are well understood. Similarly, recent progress in computational methods and computing facilities facilitate accurate evaluation of internal fields in the human body due to exposure to external fields. At ELF, as well as RF induced fields can be computed in various tissues with a resolution of 1-2 mm. Accurate computations at sub-cellular level have only been done for highly simplified models of cells. However, in principle more complex shapes and morphology of cellular structures can numerically modeled, once the permittivity of various components is known.

Several biophysical interaction mechanisms of the fields and currents induced in tissue are known, e. g. membrane excitation at low frequencies, or energy conversion to heat at RF. Biological effects of fields at ELF and RF have been investigated, and a reasonable database is available. However, some questions related to the biophysical interaction mechanisms and potential for harmful effects remain unresolved. Some controversial finding exist in both frequency ranges, which relate to cancer. At ELF, these uncertainties are related to epidemiological studies. At RF, animal studies related to cancer are still in progress.

REFERENCES

[1] National Research Council, Possible Health Effects of Exposure to Residential Electric and Magnetic Fields, National Academy Press, Washington, DC, 1997.

[2] NIEHS Working Group Report, Assessment of Health Effects from Exposure to Power-Line Frequency Electric and Magnetic Fields, ED: C. J. Portier and M. S. Wolfe, National Institute of Environmental Health Sciences of the National Institutes of Health, Brooklyn Minnesota, 1998.

[3] M. A. Stuchly and T. W. Dawson, “Interaction of low-frequency electric and magnetic fields with the human body,” Proceedings IEEE, vol. 88(5), pp. 643-664, 2000. 

[4] Documents of the NRPB, ELF Electromagnetic Fields and the Risk of Cancer, National Radiological Protection Board, Chilton, Didcot, Oxon, UK, vol. 12(1), 2001.

[5] IEGMP, Mobile Phones and Health, Chairman, Sir William Stewart, Independent Expert Group on Mobile Phones, National Radiological Protection Board, Chilton, Didcto, Oxon, UK, 2000.

[6] D. Krewski, M. McBridge and T. Sale, Guest Editors, “Special Issue on Radiofrequency Fields and Health,” J. Toxiocology and Environmental Health, Part B Critical Reviews, vol. 4(1), pp. 1-143, 2001.

[7] M. H. Repacholi, “Low-level exposure to radiofreqneyc electromagnetic fields: Health effects and research needs,” Bioelectromagnetics, vol. 19, pp. 1-19, 1998.

[8] Royal Society of Canada, A review of the potential health risks of radiofrequency fields from wireless telecom-munication devices,” Expert Panel Report, ISBN 9200064-68-X, 1999.

[9] Z. Sienkiewicz, “Biological effects of electromagnetic fields and radiation,” J. Radiol. Prot., vol. 18, pp. 185-193, 1998.

[10] M. A. Stuchly, “Biomedical concerns in wireless communications.” Crit. Review Biomed. Engineering, vol. 26, pp. 117-151, 1998.

[11] K. Foster and H. Schwan,  “Dielectric properties of tissues”, Chapter 1, in Handbook of Biological Effects of Electromagnetic Fields, 2^nd ed., C. Polk and E. Postow, EDS., CRC Press, New York, NY., 1996.

[12] P. J. Dimbylow, “FDTD calculations of the whole-body averaged SAR in an anatomically realistic voxel model of the human body from 1 MHz to 1 GHz”, Phys. Med. Biol., vol. 42, pp. 479-490, 1997.