2 INTERACTIONS OF RFR WITH BIOLOGICAL ENTITIES

Interactions of electromagnetic fields with biological entities often have been loosely characterized in the RFR-bioeffects literature as "thermal" or "nonthermal," usage that led to considerable confusion and controversy. It is therefore appropriate to provide working definitions of these terms at this point, with the understanding that the boundary between these types of interaction may not be sharp in some cases.

The interaction of an agent (e.g., RFR) with an entity (biological or nonbiological) can be characterized as thermal if the energy absorbed by the entity is transformed at the absorption site into heat. Absorption of heat, in turn, is defined in classical thermodynamics as either an increase in the mean random speed (or kinetic energy) of the molecules at the site (a local increase in temperature), or as an increase in the disorder or randomness of the molecular motion (entropy) unaccompanied by an increase in mean kinetic energy (a first-order phase change, such as the process involved in ice melting at 0 deg C), or a combination of the two processes.

An entity can also absorb energy at specific discrete frequencies in the form of energy packets or quanta, with each quantum yielding up energy to the entity in proportion to the discrete frequency of that quantum. Although large numbers of molecules can be involved, quantum absorption is essentially a microscopic phenomenon in that the constituents and configurations of the various molecular species comprising the entity determine the specific frequencies or characteristic spectra at which such absorption can occur. The kinds of interactions involved in such absorption are of varying degrees of complexity and the quantum energies involved cover a very wide range.

Interactions that require relatively large quantum energies include bond disruption within or between molecules and excitation of molecules or atoms to states of higher electron energy (including ionization). The activation energies for such interactions range from about 0.08 electron volts (eV) for hydrogen-bond disruption to about 10 eV for ionization. The corresponding quantum frequencies extend from about 19,000 GHz to 2.4 million GHz (Cleary, 1973). However, an electromagnetic quantum, at, say, 100 MHz, has an energy of 0.42-millionths of an eV or about 5-millionths of the energy required for hydrogen-bond disruption, so RFR quanta, even in great numbers, are most unlikely to be involved in such interactions.

On the other hand, changes of molecular orientations and configurations that do not alter the basic identities of the molecules require much lower quantum energies. Indeed, cooperative interactions occur among subunits of molecules within biological cells, in membranes and other cellular structures, and in extracellular fluids; in such interactions, the energy absorbed at one specific site in a structure (in a membrane or in a biological macromolecule, for example) may not be sufficient to disrupt a bond but could alter a process at the site or elsewhere in the structure, or trigger a function of the structure as a whole by release of the energy stored in the structure, thereby producing biological amplification of the incident quantum of energy. Therefore, it has been postulated that such interactions could occur at frequencies that extend down into the RFR range as defined herein.

All quantum interactions such as those mentioned above are nonthermal, particularly if essentially all of the energy absorbed by the entity is used for such processes. If most of the absorbed energy is subsequently transformed locally into heat (as defined above), however, the thermal-vs-nonthermal distinction becomes blurred. Pragmatically, therefore, to characterize an RFR interaction with a biological entity as nonthermal requires that the interaction produce an effect sharply dependent on the specific RFR frequency and experimentally distinguishable from heating effects due to thermalization of the RFR energy absorbed.

It has been widely believed that since interactions of an incident field on a complex macromolecular structure such as a membrane are non-linear, thermal equilibrium would be quickly attained by distribution of the incoming energy among the many macromolecular resonant normal modes of the structure. However, some current theories suggest that the incoming energy can be periodically exchanged among a few resonant modes for a relatively long time before being thermalized in the sense above, and thereby give rise to effects not ascribable to heat per se.

On the other hand, the mean thermal energy that corresponds to the physiological temperature 37 deg C is about 0.027 eV, with a relative maximum at about 6,500 GHz and a classical spectral distribution that encompasses the frequency range for cooperative processes. Therefore, as a counterargument to the manifestation of such nonthermal effects in vivo, it has been suggested that such effects would be swamped by those spontaneously induced thermally in vivo. Alternatively, separation of such RFR interactions from those thermally induced may require that the rates of occurrence of the former exceed the rates for the latter. This requirement implies that for manifestation of such effects of RFR, the intensity of the incident field must exceed minimum values or thresholds related to the specific processes.

2.1 THERMAL INTERACTIONS AND SPECIFIC ABSORPTION RATES (SARs)

The relative magnetic permeability of most organic constituents is about unity. Therefore, thermal interactions (as defined previously) of RFR with a biological entity are dependent on the complex-dielectric and thermal properties of its constituents and their distribution within the entity, as well as on the characteristics of the RFR.

Measurements of such properties were made some years ago for various mammalian tissues, blood, cellular suspensions, protein molecules, and bacteria over the spectral region from about 10 Hz to 20 GHz by Cook (1951, 1952) and by Schwan and coworkers (Schwan and Li, 1953; Schwan and Piersol, 1955; Schwan, 1957; 1963). In general, the dielectric constants were found to vary inversely with frequency, with distinct dispersion regions ascribed to three different predominant relaxation mechanisms. In the low and intermediate frequency ranges (from about 10 Hz to 100 MHz), the properties of cell membranes predominate because of their large specific capacitances (about 1 microfarad/sq cm). In the range above about 10 GHz, membrane impedances are negligible, and the behavior of the water and electrolyte content are most predominant.

In the frequency range from 3 to 30 MHz, the dielectric constant of muscle varies from about 360 to about 110. The values for skin, blood, and other tissues with high water contents are comparable. The values for fat, bone, and other tissues with low water content are about an order of magnitude smaller and are sensitive to the amount of water the tissues contain. From about 300 MHz to about 10 GHz, the dielectric constants of skin, muscle, and blood vary little with frequency. The mean dielectric constants of these three constituents are about 40, 50, and 60, respectively; the differences in values are largely ascribable to the proportion of water in each constituent, since the dielectric constant of water is about 80.

Subsequent in-vitro measurements were done with more advanced techniques and instrumentation as they became available. Among such studies were those of Lin (1975), Bianco et al. (1979), Schwan and Foster (1980), Foster and Schepps (1981), and Foster et al. (1982). Burdette et al. (1980) and Stuchly et al. (1981) measured the dielectric properties of various animal tissues in vivo at frequencies up to 10 GHz. Differences in dielectric constant and electrical conductivity found between in-vivo and in-vitro measurements of similar tissues were ascribed primarily to differences in water content. Stuchly and Stuchly (1980) tabulated data for the range from 10 kHz to 10 GHz.

Because the refractive index of a material is related to its dielectric constant, electromagnetic fields are reflected and refracted at the air-surface interface and at internal boundaries between constituents of widely different dielectric properties (for example, at the interfaces between connective and fatty tissues or between a body cavity and the adjacent tissues), thereby affecting the internal field distributions. At an air-muscle interface, for example, only about 22% of the incident power density of 100-MHz RFR is transmitted (the rest being reflected), and similarly, about 41% and 46% are transmitted for 1-GHz and 10-GHz RFR, respectively. The corresponding values for the air-skin interface are approximately the same.

At an air-surface interface, the fraction of incident energy that is not reflected enters the body and undergoes partial or complete absorption. The attenuation constant of any material (rate of energy absorption per unit distance along the propagation direction) is proportional to the square-root of the electrical conductivity of the material. For muscle, skin, blood, and other constituents of the body, conductivities increase slowly with frequency up to about 1 GHz, and rapidly from that frequency upward. The concept of "penetration depth" (the inverse of attenuation constant) is often used. For homogeneous isotropic planar specimens, the penetration depth is defined as the distance at which the electric-field amplitude is 1/e (37%) of its value or the power density is 1/(e-squared) of its value just within the surface. At 1 GHz, for example, the penetration depth for muscle is about 2.4 cm, whereas at about 10 GHz and higher, field penetration is confined to the skin. Thus, in the latter frequency range, RFR penetration is much like that of sunlight.

In the RFR-bioeffects literature, absorption of energy from an incident electromagnetic field by a biological entity is generally quantified by the "specific absorption rate" (SAR). The SAR of a small volume at any locale within an entity is defined as the rate of energy absorption per unit volume divided by the mean mass density of the constituents in that volume, and is usually expressed in units of W/kg or mW/g (1 mW/g = 1 W/kg). The local value of SAR thus defined at any site within an entity depends on: the characteristics of the incident RFR (carrier frequency, modulation, amplitudes and directions of its components); the spatial distribution of complex-dielectric and thermal properties of the entity (including those of the site and its location within the entity); and the configuration of the entity and its orientation relative to the RFR.

For entities that have complex shapes and large spatial variations of constituents, distributions of local SAR are difficult to determine by experiment or by calculation. Thus, the concept of "whole-body SAR," which represents the spatial mean value for the body (in any specified configuration and orientation) is useful because it is a quantity that can be measured experimentally--e.g., by calorimetry--without the need for information on the internal SAR distribution.

Although SAR (local or whole-body) by this definition appears to be a measure of RFR-absorption as heat (true in most cases), it is also used as a measure of internal field intensities in studies characterized as nonthermal, in which the heat generated by the RFR is negligible.

After the SAR concept gained acceptance, many investigators calculated whole-body SARs for relatively simple geometric models (homogeneous and multilayered spheroids, ellipsoids, and cylinders) having masses and dimensions representative of various species (including humans), which were assumed to be exposed (in free space) to linearly polarized plane waves in various orientations. In some studies, such calculations were verified experimentally.

Many important results of such theoretical and experimental studies were embodied in a series of handbooks (Johnson et al., 1976; Durney et al., 1978, 1980), in which plots are presented of whole-body SARs (normalized to an incident power density of 1 mW/sq cm) vs frequency for three major orientations. Such plots have proved useful for approximate frequency-scaling and interspecies comparisons of whole-body SARs. A significant finding of such studies is that the largest value of whole-body SAR is obtained when the longest dimension of each model is parallel to the electric-field component (polarization direction) of the RFR and when the wavelength of the RFR is about 2.5 times the longest dimension. The frequency corresponding to this wavelength is often referred to as the "resonant frequency" for that species. The whole-body SAR for each such model at its resonant frequency also varies inversely with the dimension of the model that is perpendicular to the polarization and propagation directions of the field. Thus, the model has absorption characteristics somewhat similar to those of a lossy dipole antenna in free space.

Resonances would also occur for circularly polarized RFR. Such RFR can be resolved into two mutually perpendicular components, each having half the total power density. Therefore, an entity exposed to circularly polarized RFR would have lower resonant SAR values than it would have if exposed to linearly polarized RFR of the same total power density.

Plots of whole-body SAR versus RFR frequency for the prolate-spheroidal homogeneous model of an "average" or "standard" man, about 5 ft 9 inches (1.75 m) tall and weighing about 154 lb (70 kg), are given in Durney et al. (1978), p. 78, for three orientations defined as follows: the "E-orientation," in which the long axis of the spheroid is parallel to the electric vector and perpendicular to the magnetic vector and propagation direction; the "H-orientation," in which that axis is parallel to the magnetic vector and perpendicular to the electric vector and propagation direction; and the "K-orientation," in which that axis is parallel to the propagation direction. The resonant frequency (E-orientation) for this model of a man is about 70 MHz; at this frequency, the SAR is about 0.2 W/kg for an incident power density of 1 mW/sq cm, or about 1/6 of his resting metabolic rate, or 1/21 to 1/90 of his metabolic rate for exercises ranging from walking to sprinting (Ruch and Patton, 1973).

Similar whole-body-SAR plots were presented in Durney et al. (1978), pp. 81 and 84 respectively for prolate-spheroidal models of the "average" woman and 10-year-old child exposed to 1 mW/sq cm of plane-polarized RFR. The average woman is assumed to be about 5 ft 3 inches (1.61 m) tall and to weigh about 135 lb (61.14 kg). Her resonant frequency is about 80 MHz and her whole-body maximum SAR is about the same as for the man. The child is assumed to be about 4 ft 6 inches (1.38 m) tall and 71 lb (32.2 kg). Its resonant frequency is still higher, about 95 MHz; its whole-body maximum SAR is about 0.3 W/kg, somewhat larger than for the adults. It is noteworthy that all three maximum SARs are smaller than the 0.4-W/kg value used as the basis for the current ANSI standard (ANSI, 1982).

To illustrate how the concept of whole-body SAR could be interpreted, consider the standard model man. Absorption of RFR energy as heat by exposure of such a model man at his resonant frequency (70 MHz) in the E-orientation to an average power density of 1 mW/sq cm (SAR 0.2 W/kg) for 1 hr would produce a mean body temperature rise of only about 0.2 deg C if no heat removal mechanisms were present and if no first-order phase changes were involved.

In general, the whole-body SAR at frequencies (f) below resonance in the E-orientation is approximately proportional to f-squared; at frequencies above resonance, the whole-body SAR is approximately proportional to 1/f for about one decade of frequency and exhibits smaller relative maxima (secondary resonances) at higher frequencies. The data from which such plots were derived can be used to calculate, by simple proportion, the incident power densities necessary to produce an SAR of 0.4 W/kg. Plots of power density versus frequency derived in this manner are higher than the limits of the current ANSI standard, indicating that the ANSI limits are somewhat more stringent than such data.

Similar data for a prolate-spheroidal model of a "small" rat (0.14 m long and weighing 0.11 kg) are presented in Durney et al. (1978), p. 92. Not only is the resonant frequency in the E-orientation (about 900 MHz) higher than any of the values for humans, but the resonant SAR is also larger (about 1.1 W/kg for the rat, compared with about 0.2 W/kg for the adult human, per mW/sq cm). Therefore, in scaling experimental results from animals to humans, such differences of whole-body SAR as well as frequency must be considered.

The presence of a ground plane or other reflecting surfaces shifts the resonant frequencies downward and can produce higher values of whole-body SAR at the lower resonant frequencies (Gandhi, 1975; Gandhi et al., 1977; Hagmann and Gandhi, 1979). Hagmann and Gandhi (1979) showed that for a homogeneous block model (see below) of "standard" man standing in electrical contact with a perfectly conducting infinite ground plane, the whole-body resonant frequency (in the E-orientation) is shifted from the free-space value of 77 MHz to 47 MHz. Moreover, the whole-body SAR at 47 MHz is 32.5% higher than at 77 MHz. However, they also noted that such ground-plane effects are largely eliminated if conductive contact with the ground is removed.

The foregoing discussion of whole-body SARs is also largely applicable to modulated RFR (including pulsed RFR) at the corresponding carrier frequencies and time-averaged incident power densities.

Numerical calculations of internal spatial distributions of SAR have been done for "block" models. In such models, the shape of the body is approximated by an appropriate arrangement of many rectangular cells of various sizes, and each cell is assumed to be biologically homogeneous and to have constant internal field over its volume when the model is exposed to RFR. In addition, the biological properties ascribed to each cell are selected to approximate those of the tissues in corresponding locations of the body (Chen and Guru, 1977; Hagmann et al., 1979a, 1981; Chatterjee et al., 1980). More accurate values of whole-body SAR have been obtained with such models than from simpler ones.

Block models, as well as homogeneous and multilayered spheroidal and cylindrical models that have appropriate electromagnetic and thermal characteristics have been used also to represent various parts of the body, such as the head and limbs (Joines and Spiegel, 1974; Weil, 1975; Lin, 1975; Kritikos and Schwan, 1975, 1976; Neuder et al., 1976; Wu and Lin, 1977; Rukspollmuang and Chen, 1979; Massoudi et al., 1979; Hagmann et al., 1979a, 1979b, 1981; Janna et al., 1980; Spiegel et al., 1980; Kritikos et al., 1981).

An early, very significant finding for spherical models of the isolated head assumed to be exposed to plane-wave RFR was the discovery of local regions of relative maximum SAR values. The locations of such regions depend on the size of the head, the electromagnetic characteristics of its layers, and the wavelength of the incident field. These regions have been dubbed "hot spots," even for combinations of incident power density and exposure duration that would produce temperature increases at such spots that are biologically insignificant. An analysis of a homogeneous lossy spherical-head model (Kritikos and Schwan, 1975) showed that in the frequency range from about 300 MHz to 12 GHz, there are hot spots inside spheres having radii between 0.1 and 8 cm; there are also internal hot spots for larger radii and other frequencies, but the hottest spots are at the front surface (facing the RFR source).

Similar results were obtained for multilayered spherical models (Weil, 1975; Kritikos and Schwan, 1976). Specifically, Kritikos and Schwan (1976) analyzed two such models, one with a radius of 5 cm and the other, 10 cm. For the 5-cm head, the hot spots are internal over the approximate frequency range from 400 MHz to 3 GHz. The highest relative maximum SAR occurs near the center of the head at about 1 GHz, and has a value of about 9 W/kg for an incident power density of 1 mW/sq cm. (Of course, the whole-head SARs are considerably lower.) By contrast, for the head of 10-cm radius (about that for an adult human head), no deep internal hot spots are produced at any frequency; the hot spots are always at or just beneath the front surface.

Rukspollmuang and Chen (1979) obtained qualitatively similar results for a block model of an isolated multilayered spherical head. They then studied, at 918 MHz and 2.45 GHz, a block model with shape and internal structure more closely approximating that of the human head (including eyes, nose, skull bone, and brain), and found that much of the energy within the head would be absorbed by the skull. In particular, frontal exposure of the more accurate model at 918 MHz would yield a maximum SAR for the brain region about one-third that for the brain region of a 7-cm-radius multilayered spherical model. Also, for frontal exposure of the model to 2.45 GHz, the induced field is concentrated primarily near the proximal surface, and therefore energy dissipation within the brain would be relatively low.

Hagmann et al. (1979b) calculated SAR distributions in the attached head of a block model of the human, and derived the whole-head and whole-body SARs for three orientations of the model relative to the source of RFR. For front-to-back propagation with the long axis of the body parallel to the electric vector, they found a broad head resonance at about 350 MHz, with a whole-head SAR of about 0.12 W/kg per mW/sq cm; the corresponding whole-body SAR is about 0.05 W/kg per mW/sq cm. For propagation in the head-to-toe direction, a sharper head resonance at 375 MHz was obtained, with whole-head and whole-body SARs respectively approximately 0.22 and 0.07 W/kg per mW/sq cm.

Results of theoretical analyses of whole-body SARs and SAR distributions have been verified experimentally. Physical models of simple geometry or of human- or animal-figurine shape were constructed from synthetic biological materials of electromagnetic characteristics about equivalent to their corresponding biological constituents. The models were exposed to RFR at power densities sufficient to produce accurately measurable temperature increases, which were measured immediately after exposure.

To use available sources of RFR that provide only specific frequencies and to avoid problems of exposing large full-scale models, smaller models are often chosen by use of scaling relationships so that results of exposing such smaller models at the available frequencies can be extrapolated to obtain results on full-size models at other frequencies of interest. Using this approach, Guy et al. (1976) exposed homogeneous human figurines having lengths of 37.6 and 26.5 cm (as well as spheres and ellipsoids) at approximately 143 MHz to simulate exposures of full-size figurines (1.74 m in length) at 31.0 and 24.1 MHz. In their study of head resonances, Hagmann et al. (1979b) exposed human figurines with lengths of 20.3, 25.4, 33.0, and 40.6 cm at 2.45 GHz, to correspond with full-size figurines exposed respectively at scaled frequencies of 284.5, 355.6, 462.3, and 569.0 MHz.

Atechnique widely used to determine the SAR distributions in physical models or animal carcasses is to embed such an object within Styrofoam, section the object along the parting planes of interest, reassemble the object, and expose it to RFR. Immediately after exposure, the spatial distribution over each parting plane is measured with scanning infrared thermography. However, such spatial temperature distributions should not be regarded as in direct correspondence with the in-vivo internal temperature distributions, because such carcasses and physical models have heat transfer characteristics much different from those of live animals and do not possess thermoregulatory mechanisms. Instead, such measured temperature distributions represent reasonable approximations to the distributions of internal field or SAR.

Guy et al. (1976, 1977) discovered that exposure of a full-size human figurine to an electric field parallel to its length at frequencies in the HF band yields relatively high SARs in regions of the body where the cross section perpendicular to current flow is relatively small, such as in the neck, knees, and ankles. In addition, exposure of the figurine to a magnetic field perpendicular to the frontal plane at frequencies in the same range produces eddy currents that yield relatively high SARs where such currents are forced through relatively small cross-sectional areas or are diverted by sharp angular changes, such as in the groin and along the sides of the body near the ribs. These results are especially pertinent to near-field exposure situations, for which it is necessary to measure the spatial variations of the electric and magnetic fields separately because their amplitude ratio may vary from point to point, and the field components may not be perpendicular to one another and to the propagation direction.

Whole-body (and detached whole-head) SARs, as well as (attached) part-body SARs, were measured by calorimetry alone or in conjunction with scanning infrared thermography by Hunt and Phillips (1972); Kinn (1977); Allen and Hurt (1979); Hagmann et al. (1979b); Olsen et al. (1980). Whole-body SARs were also determined in waveguide exposure systems by measuring the input, output, and reflected values of RFR power without and with the object present and performing the requisite arithmetic (Ho et al., 1973). The experimental results are in qualitative agreement with those derived from the theoretical models.

Burr and Krupp (1980) measured real-time temperature increases induced by 1.2-GHz RFR at 70 mW/sq cm in homogeneous spheres (3.3-cm radius); in Macaca mulatta cadaver heads (attached to the body and detached); and in living, anesthetized (attached) heads of the same species. They used a sensitive, accurate temperature probe (Bowman, 1976) that essentially does not perturb, or is not perturbed by, the RFR. The bodies of the animals were exposed with their longest axes parallel to the electric or magnetic component of the incident RFR (respectively the E-orientation or H-orientation). The results indicated that temperature distributions in attached cadaver heads vary strongly with body orientation. They also found that the temperature distribution in the head of the live animal is affected by blood flow in a complex manner not adequately predicted by current theoretical models.

Olsen (1984) measured SARs in a full-size model of a human exposed to far-field 2.0-GHz RFR, using a nonperturbing temperature probe and a gradient-layer calorimeter. Local SARs were much higher in the limbs than in the trunk, and the whole-body SAR was about threefold larger than the value estimated from a prolate-spheroidal model. The author suggested that resonant interactions involving the limbs may account for the disparity.

Hill (1984a) measured the whole-body SARs of five male volunteers at frequencies in the range 3-41 MHz. The subjects were exposed at about 0.01 mW/sq cm in a very large transverse electromagnetic (TEM) cell, with the long body axis parallel to the electric vector, and either in profile (EKH-polarization) or frontally (EHK-polarization) relative to the propagation direction. The exposures were done with the subjects ungrounded or grounded.

With the humans ungrounded, the SARs were found to be about 40% higher for the EKH-polarization than the EHK-polarization, but only about 6% higher with the humans grounded. Results were also presented, showing that the method used in making electrical contact between the feet and the ground plane is unimportant, that even use of a sheet of paper to eliminate direct conductive contact did not affect those SARs. Also reported was that for a grounded average human exposed to 3-41 MHz in the EKH-polarization to the power densities specified in ANSI (1982), i.e., 100 to 1 mW/sq cm in the range 3 to 30 MHz and 1 mW/sq cm for 30-41 MHz, the SAR over most of the frequency range is 0.58 +/- 0.14 W/kg, a value slightly higher than the 0.4 W/kg underlying ANSI (1982). Hill (1984b), however, found that for people with ordinary footwear, the SARs are about half those for grounded humans at frequencies below resonance and about 20% lower near resonance, and thus do not exceed 0.4 W/kg.

Much of the dosimetry work discussed thus far was done for actual or assumed exposures to far-field (planar) RFR. Because of concern with possible hazards from the use of RFR in broadcast, industrial, and biomedical applications in which personnel are occupationally in the near field of the RFR, research has also been done to determine SAR distributions induced in such exposure situations (Chatterjee et al., 1980, 1981, 1982; Iskander et al., 1980; Karimullah et al., 1980; Spiegel, 1982; Stuchly et al., 1985a,b). As expected, the results are sensitive to the type, location, and orientation of the RFR source, as well as to the characteristics of the RFR and the properties of the biological entity exposed.

2.2 QUANTUM INTERACTIONS AND NONTHERMAL EFFECTS

The literature on theories of direct interactions of RFR with biological entities at the microscopic level (such as neurons and other cells) is extensive. Mechanisms of interaction have been proposed to account for various reported effects on cellular membranes, microtubules, DNA, and other intracellular structures and constituents, and on the transport of various ions and/or biomolecules across cell membranes. Topics such as these have been addressed in various symposia and review articles. The proceedings of several symposia have been published, e.g., Taylor and Cheung (1978, 1979), Illinger (1981), and Adey and Lawrence (1984).

Arecent symposium for which the proceedings were not yet available at this writing is "Radiation Field Effects as Signal Transducing Mechanisms in Biological Systems," held at the University of Texas Health Science Center at San Antonio, TX, under the sponsorship of its Department of Pharmacology and of the Air Force Office of Scientific Research during 11-12 December 1985. Review articles by Cleary (1973, 1977, 1979), Adey (1981), Lawrence and Adey (1982), Froehlich (1980, 1982), and Taylor (1981) are representative.

Theories of direct RFR interaction have been proposed and/or used to account for various experimental results deemed by the investigators in many cases as nonthermal. Such studies will be discussed later herein under several appropriate topic headings (Auditory Effects, Blood-Brain-Barrier Effects, Calcium Efflux, Erythrocyte Studies, Cellular and Subcellular Effects, and possibly others).

2.3 INTERACTIONS OF MODULATED RFR

2.3.1 SINUSOIDAL AMPLITUDE MODULATION

The effects of sinusoidally-amplitude-modulated RFR that are ascribable to average power density per se (i.e., to the RFR-induced heat) do not differ from those for CW RFR or for frequency-modulated-CW (FMCW) RFR at the same carrier frequency and power density. However, studies have been conducted with sinusoidally-amplitude-modulated RFR of specified characteristics, in which effects were described that were attributed to the frequency of the amplitude modulation per se, as well as the average power density, notably the calcium-efflux phenomenon. This phenonemon will be discussed in Section 3.4.4.

2.3.2 RFR PULSES AT LOW DUTY CYCLES

The temperature increase of any given region within a biological entity due to the arrival of a single RFR pulse would be small because of the relatively large thermal time constants of biological materials and the operation of heat-exchange mechanisms, unless the total energy within the pulse is extremely large (high enough peak power density and pulse duration). However, if the region contains a boundary between layers of widely different dielectric properties, the temperature gradient (rate of change of temperature with distance) can be large at such a boundary even if the mean temperature increase in the region is small.

Awell known phenomenon that occurs in vivo is perception of single pulses, or of repetitive pulses of RFR that are short relative to the duration between pulses (low duty cycle), as apparently audible clicks, usually called the "microwave-hearing" or "RFR-auditory" effect. The interaction mechanisms involved are not yet fully understood. However, most of the experimental results tend to support the theory that pulse perception occurs because of transduction of the electromagnetic energy into sound pressure waves in the head at an interface between layers of widely different dielectric properties (e.g., at the boundary between the skull and the skin or cerebrospinal fluid). The energy in a pulse arriving at such a boundary is believed to be converted into an abrupt increase in momentum that is locally thermalized, producing a negligible volumetric temperature increase but a large temperature gradient across the boundary. Under such conditions, rapid local differential expansion would occur and create a pressure (sound) wave that is detected by the auditory apparatus.

The RFR-auditory effect has been characterized as nonthermal because the average power density can be minuscule. Specifically, the time-averaged power density for any two successive pulses is inversely proportional to the time interval between the arrival of the pulses at the perceiver and this interval can be indefinitely long without affecting the perception of each pulse. For this reason, the time-averaged power density has no relevance to perception. This effect and postulated mechanisms for its occurrence will be discussed more fully in Section 3.1.4.2.

Pulsed RFR has also been reported to produce other effects, such as alterations of the blood-brain barrier and behavioral changes. Such effects will be discussed in Sections 3.4.1 and 3.7.1, respectively.

REFERENCES:

Adey, W.R.
TISSUE INTERACTIONS WITH NONIONIZING ELECTROMAGNETIC FIELDS
Physiol. Rev., Vol. 61, pp.435-514 (1981)

Adey, W.R. and A.F. Lawrence (eds.)
NONLINEAR ELECTRODYNAMICS IN BIOLOGICAL SYSTEMS
(Proceedings of the International Conference on Nonlinear Electrodynamics in Biological Systems, held at Pettis Memorial Veterans Hospital under sponsorship of the Veterans Administration, Loma Linda, CA, during 5-9 June 1983), Plenum Press, New York (1984)
[2.1]

Allen, S.J. and W.D. Hurt
CALORIMETRIC MEASUREMENTS OF MICROWAVE ENERGY ABSORPTION BY MICE AFTER SIMULTANEOUS EXPOSURE OF 18 ANIMALS
Radio Sci., Vol. 14, No. 6S, pp. 1-4 (1979)

ANSI, C95.1-1982
SAFETY LEVELS WITH RESPECT TO HUMAN EXPOSURE TO RADIO FREQUENCY ELECTROMAGNETIC FIELDS, 300 KHZ TO 100 GHZ
Published by the Institute of Electrical and Electronics Engineers, New York (1982)

Bianco, B., G.P. Drago, M. Marchesi, C. Martini, G.S. Mela, and S. Ridell
MEASUREMENTS OF COMPLEX DIELECTRIC CONSTANT OF HUMAN SERA AND ERYTHROCYTES
IEEE Trans. Instr. & Meas., Vol. 28, No. 4, pp. 290-295 (1979)

Bowman, R.R.
A PROBE FOR MEASURING TEMPERATURE IN RADIO-FREQUENCY-HEATED MATERIAL
IEEE Trans. Microwave Theory Tech., Vol. 24, No. 1, pp. 43-45 (1976)

Burdette, E.C., F.L. Cain, and J. Seals
IN VIVO PROBE MEASUREMENT TECHNIQUE FOR DETERMINING DIELECTRIC PROPERTIES OF VHF THROUGH MICROWAVE FREQUENCIES
IEEE Trans. Microwave Theory Tech., Vol. 28, No. 4, pp. 414-427 (1980)

Burr, J.G. and J.H. Krupp
REAL-TIME MEASUREMENT OF RFR ENERGY DISTRIBUTION IN THE MACACA MULATTA HEAD
Bioelectromagnetics, Vol. 1, No. 1, pp. 21-34 (1980)

Chatterjee, I., M.J. Hagmann, and O.P. Gandhi
ELECTROMAGNETIC-ENERGY DEPOSITION IN AN INHOMOGENEOUS BLOCK MODEL OF MAN FOR NEAR-FIELD IRRADIATION CONDITIONS
IEEE Trans. Microwave Theory Tech., Vol. 28, No. 12, pp. 1452-1459 (1980)

Chatterjee, I., M.J. Hagmann, and O.P. Gandhi
AN EMPIRICAL RELATIONSHIP FOR ELECTROMAGNETIC ENERGY ABSORPTION IN MAN FOR NEAR-FIELD EXPOSURE CONDITIONS
IEEE Trans. Microwave Theory Tech., Vol. 29, No. 11, pp. 1235-1238 (1981)

Chatterjee, I., O.P. Gandhi, and M.J. Hagmann
NUMERICAL AND EXPERIMENTAL RESULTS FOR NEAR-FIELD ELECTROMAGNETIC ABSORPTION IN MAN
IEEE Trans. Microwave Theory Tech., Vol. 30, No. 11, pp. 2000-2005 (1982)

Chen, K.-M. and B.S. Guru
INTERNAL EM FIELD AND ABSORBED POWER DENSITY IN HUMAN TORSOS INDUCED BY 1-500-MHZ EM WAVES
IEEE Trans. Microwave Theory Tech., Vol. 25, No. 9, pp. 746-756 (1977)

Cleary, S.F.
UNCERTAINTIES IN THE EVALUATION OF THE BIOLOGICAL EFFECTS OF MICROWAVE AND RADIOFREQUENCY RADIATION
Health Phys., Vol. 25, pp. 387-404 (1973)

Cleary, S.F.
BIOLOGICAL EFFECTS OF MICROWAVE AND RADIOFREQUENCY RADIATION
CRC Critical Reviews in Environmental Control, Vol. 8, pp. 121-166 (1977)

Cleary, S.F.
RECAPITULATION: BIOMEDICAL EFFECTS
Bull. N.Y. Acad. Med., Vol. 55, No. 11, pp. 1119-1125 (1979)

Cook, H.F.
DIELECTRIC BEHAVIOR OF SOME TYPES OF HUMAN TISSUES AT MICROWAVE FREQUENCIES
Brit. J. Appl. Phys., Vol. 2, pp. 295-300 (1951)

Cook, H.F.
A COMPARISON OF DIELECTRIC BEHAVIOR OF PURE WATER AND HUMAN BLOOD AT MICROWAVE FREQUENCIES
Brit. J. Appl. Phys., Vol. 3, pp. 249-255 (1952)

Durney, C.H., C.C. Johnson, P.W. Barber, H.W. Massoudi, M.F. Iskander,
J.L. Lords, D.K. Ryser, S.J. Allen, and J.C. Mitchell
RADIOFREQUENCY RADIATION DOSIMETRY HANDBOOK [SECOND EDITION]
USAF School of Aerospace Medicine, Brooks AFB, TX, Report SAM-TR-78-22
(1978)

Durney, C.H., M.F. Iskander, H. Massoudi, S.J. Allen, and J.C. Mitchell
RADIOFREQUENCY RADIATION DOSIMETRY HANDBOOK [THIRD EDITION]
USAF School of Aerospace Medicine, Brooks AFB, TX, Report SAM-TR-80-32
(1980)

Foster, K.R. and J.L. Schepps
DIELECTRIC PROPERTIES OF TUMOR AND NORMAL TISSUES AT RADIO THROUGH MICROWAVE FREQUENCIES
J. Microwave Power, Vol. 16, No. 2, pp. 107-119 (1981)

Foster, K.R., J.L. Schepps, and B.R. Epstein
MICROWAVE DIELECTRIC STUDIES ON PROTEINS, TISSUES, AND HETEROGENEOUS SUSPENSIONS
Bioelectromagnetics, Vol. 3, No. 1, pp. 29-43 (1982)

Froehlich, H.
THE BIOLOGICAL EFFECTS OF MICROWAVES AND RELATED QUESTIONS
In L. and C. Marton (eds.), ADVANCES IN ELECTRONICS AND ELECTRON PHYSICS, Vol. 53, Academic Press, pp. 85-152 (1980)

Froehlich, H.
WHAT ARE NON-THERMAL ELECTRIC BIOLOGICAL EFFECTS?
Bioelectromagnetics, Vol. 3, No. 1, pp. 45-46 (1982)

Gandhi, O.P.
CONDITIONS OF STRONGEST ELECTROMAGNETIC POWER DEPOSITION IN MAN AND ANIMALS
IEEE Trans. Microwave Theory Tech., Vol. 23, No. 12, pp. 1021-1029 (1975)

Gandhi, O.P., E.L. Hunt, and J.A. D'Andrea
DEPOSITION OF ELECTROMAGNETIC ENERGY IN ANIMALS AND IN MODELS OF MAN WITH AND WITHOUT GROUNDING AND REFLECTOR EFFECTS
Radio Sci., Vol. 12, No. 6S, pp. 39-47 (1977)

Guy, A.W., M.D. Webb, and C.C. Sorensen
DETERMINATION OF POWER ABSORPTION IN MAN EXPOSED TO HIGH FREQUENCY ELECTROMAGNETIC FIELDS BY THERMOGRAPHIC MEASUREMENTS ON SCALE MODELS
IEEE Trans. Biomed. Eng., Vol. 23, pp. 361-371 (1976)

Guy, A.W., M.D. Webb, and J.A. McDougall
RF RADIATION ABSORPTION PATTERNS: HUMAN AND ANIMAL MODELING DATA
U.S. Dept. of Health, Education, and Welfare, National Institute for Occupational Safety and Health (NIOSH), Cincinnati, Ohio, Publication PB-274 749 (1977)

Hagmann, M.J. and O.P. Gandhi
NUMERICAL CALCULATIONS OF ELECTROMAGNETIC ENERGY DEPOSITION IN MODELS OF MAN WITH GROUNDING AND REFLECTOR EFFECTS
Radio Sci., Vol. 14, No. 6S, pp. 23-29 (1979)

Hagmann, M.J., O.P. Gandhi, and C.H. Durney
NUMERICAL CALCULATIONS OF ELECTROMAGNETIC ENERGY DEPOSITION FOR A REALISTIC MODEL OF MAN
IEEE Trans. Microwave Theory Tech., Vol. 27, No. 9, pp. 804-809 (1979a)

Hagmann, M.J., O.P. Gandhi, J.A. D'Andrea, and I. Chatterjee
HEAD RESONANCE: NUMERICAL SOLUTION AND EXPERIMENTAL RESULTS
IEEE Trans. Microwave Theory Tech., Vol. 27, No. 9, pp. 809-813 (1979b)

Hagmann, M.J., I. Chatterjee, and O.P. Gandhi
DEPENDENCE OF ELECTROMAGNETIC ENERGY DEPOSITION UPON ANGLE OF INCIDENCE FOR AN INHOMOGENEOUS BLOCK MODEL OF MAN UNDER PLANE-WAVE IRRADIATION
IEEE Trans. Microwave Theory Tech., Vol. 29, No. 3, pp. 252-255 (1981)

Hill, D.A.
THE EFFECT OF FREQUENCY AND GROUNDING ON WHOLE-BODY ABSORPTION OF HUMANS IN E-POLARIZED RADIOFREQUENCY FIELDS
Bioelectromagnetics, Vol. 5, No. 2, pp. 131-146 (1984a)

Hill, D.A.
EFFECT OF SEPARATION FROM GROUND ON HUMAN WHOLE-BODY RF ABSORPTION RATES
IEEE Trans. Microwave Theory Tech., Vol. 32, No. 3, pp. 772-778 (1984b)

Ho, H.S., E.I. Ginns, and C.L. Christman
ENVIRONMENTALLY CONTROLLED WAVEGUIDE IRRADIATION FACILITY
IEEE Trans. Microwave Theory Tech., Vol. 21, No. 12, pp. 837-840 (1973)

Hunt, E.L. and R.D. Phillips
ABSOLUTE PHYSICAL DOSIMETRY FOR WHOLE ANIMAL EXPERIMENTS
Digest of Papers of the Microwave Density Workshop, Atlanta, Georgia, Department of Microwave Research, Walter Reed Army Institute of Research (1972)

Illinger, K.H. (ed.)
EFFECTS OF NONIONIZING RADIATION
American Chem. Soc. Series ACS 157 (1981)

Iskander, M.F., P.W. Barber, C.H. Durney, and H. Massoudi
IRRADIATION OF PROLATE SPHEROIDAL MODELS OF HUMANS IN THE NEAR FIELD OF A SHORT ELECTRIC DIPOLE
IEEE Trans. Microwave Theory Tech., Vol. 28, No. 7, pp. 801-807 (1980)

Janna, W.S., E.P. Russo, R. McAfee, and R.M. Davoudi
A COMPUTER MODEL OF TEMPERATURE DISTRIBUTION INSIDE A LOSSY SPHERE AFTER MICROWAVE RADIATION
Bioelectromagnetics, Vol. 1, No. 3, pp. 337-343 (1980)

Johnson, C.C., C.H. Durney, P.W. Barber, H. Massoudi, S.J. Allen, and
J.C. Mitchell
RADIOFREQUENCY RADIATION DOSIMETRY HANDBOOK
USAF School of Aerospace Medicine, Brooks AFB, TX, Report SAM-TR-76-35, pp. 100-101, (1976)

Joines, W.T. and R.J. Spiegel
RESONANCE ABSORPTION OF MICROWAVES BY HUMAN SKULL
IEEE Trans. Biomed. Eng., Vol. 21, pp.46-48 (1974)

Karimullah, K., K.-M. Chen, and D.P. Nyquist
ELECTROMAGNETIC COUPLING BETWEEN A THIN-WIRE ANTENNA AND A NEIGHBORING BIOLOGICAL BODY: THEORY AND EXPERIMENT
IEEE Trans. Microwave Theory Tech., Vol. 28, No. 11, pp. 1218-1225 (1980)

Kinn, J.B.
WHOLE BODY DOSIMETRY OF MICROWAVE RADIATION IN SMALL ANIMALS: THE EFFECT OF BODY MASS AND EXPOSURE GEOMETRY
Radio Sci., Vol. 12, No. 6S, pp. 61-64 (1977)

Kritikos, H.N. and H.P. Schwan
THE DISTRIBUTION OF HEATING POTENTIAL INSIDE LOSSY SPHERES
IEEE Trans. Biomed. Eng., Vol. 22, No. 6, pp. 457-463 (1975)

Kritikos, H.N. and H.P. Schwan
FORMATION OF HOT SPOTS IN MULTILAYER SPHERES
IEEE Trans. Biomed. Eng., Vol. 23, pp. 168-172 (1976)

Kritikos, H.N., K.R. Foster, and H.P. Schwan
TEMPERATURE PROFILES IN SPHERES DUE TO ELECTROMAGNETIC HEATING
J. Microwave Power, Vol. 16, Nos. 3 and 4, pp. 327-344 (1981)

Lawrence, A.F. and W.R. Adey
NONLINEAR WAVE MECHANISMS IN INTERACTIONS BETWEEN EXCITABLE TISSUE AND ELECTROMAGNETIC FIELDS
Neurolog. Res., Vol. 4, No. 1/2, pp. 115-153 (1982)

Lin, J.C.
MICROWAVE PROPERTIES OF FRESH MAMMALIAN BRAIN TISSUES AT BODY TEMPERATURE
IEEE Trans. Biomed. Eng., Vol. 22, pp. 74-76 (1975)

Massoudi, H., C.H. Durney, and C.C. Johnson

A GEOMETRIC-OPTICS AND AN EXACT SOLUTION FOR INTERNAL FIELDS IN AND ENERGY ABSORPTION BY A CYLINDRICAL MODEL OF MAN IRRADIATED BY AN ELECTROMAGNETIC PLANE WAVE
Radio Sci., Vol. 14, No. 6S, pp. 35-42 (1979)

Neuder, S.M., R.B. Kellogg, and D.H. Hill
MICROWAVE POWER DENSITY ABSORPTION IN A SPHERICAL MULTILAYERED MODEL OF THE HEAD
In C.C. Johnson and M. Shore (eds.), BIOLOGICAL EFFECTS OF ELECTROMAGNETIC WAVES, Vol. II, U.S. Dept. of Health, Education, and Welfare, Washington, D.C., HEW Publication (FDA) 77-8011, pp. 199-210 (1976)

Olsen, R.G., T.A. Griner, and G.D. Prettyman
FAR-FIELD MICROWAVE DOSIMETRY IN A RHESUS MONKEY MODEL
Bioelectromagnetics, Vol. 1, No. 2, pp. 149-160 (1980)

Olsen, R.G.
FAR-FIELD DOSIMETRIC MEASUREMENTS IN A FULL-SIZED MAN MODEL AT 2.0 GHZ
Bioelectromagnetics, Vol. 3, No. 4, pp. 433-441 (1984)

Ruch, T.C. and H.D. Patton
PHYSIOLOGY AND BIOPHYSICS
Vol. III, W.B. Saunders Co., Philadelphia, Pennsylvania (1973)

Rukspollmuang, S. and K.-M. Chen
HEATING OF SPHERICAL VERSUS REALISTIC MODELS OF HUMAN AND INFRAHUMAN HEADS BY ELECTROMAGNETIC WAVES
Radio Sci., Vol. 14, No. 6S, pp. 51-62 (1979)

Schwan, H.P. and K. Li
CAPACITY AND CONDUCTIVITY OF BODY TISSUES AT ULTRAHIGH FREQUENCIES
Proc. IRE, Vol. 41, pp. 1,735-1,740 (1953)

Schwan, H.P. and G.M. Piersol
THE ABSORPTION OF ELECTROMAGNETIC ENERGY IN BODY TISSUE, PART II
Amer. J. Phys. Med., Vol. 34, pp. 425-448 (1955)

Schwan, H.P.
ELECTRICAL PROPERTIES OF TISSUE AND CELL SUSPENSION
In ADVANCES IN BIOLOGICAL AND MEDICAL PHYSICS, Vol. 5, Academic Press, New York, pp. 147-209 (1957)

Schwan, H.P.
ELECTRICAL CHARACTERISTICS OF TISSUES: A SURVEY
Biophysik, Vol. 1, pp. 198-208 (1963)

Schwan, H.P. and K.R. Foster
RF-FIELD INTERACTIONS WITH BIOLOGICAL SYSTEMS: ELECTRICAL PROPERTIES AND BIOPHYSICAL MECHANISMS
Proc. IEEE, Vol. 68, No. 1, pp. 104-113 (1980)

Spiegel, R.J., D.M. Deffenbaugh, and J.E. Mann
A THERMAL MODEL OF THE HUMAN BODY EXPOSED TO AN ELECTROMAGNETIC FIELD
Bioelectromagnetics, Vol. 1, No. 3, pp. 253-270 (1980)

Spiegel, R.J.
THE THERMAL RESPONSE OF A HUMAN IN THE NEAR-ZONE OF A RESONANT THIN-WIRE ANTENNA
IEEE Trans. Microwave Theory Tech., Vol. 30, No. 2, pp. 177-185 (1982)

Stuchly, M.A. and S.S. Stuchly
DIELECTRIC PROPETTIES OF BIOLOGICAL SUBSTANCES--TABULATED
J. Microwave Power, Vol. 15, No. 1, pp. 19-26 (1980)

Stuchly, M.A., T.W. Athey, S.S. Stuchly, G.M. Samaras, and G. Taylor
DIELECTRIC PROPERTIES OF ANIMAL TISSUES IN VIVO AT FREQUENCIES 10 MHZ-1 GHZ
Bioelectromagnetics, Vol. 2, No. 2, pp. 93-103 (1981)

Stuchly, S.S., A. Kraszewski, M.A. Stuchly, G. Hartsgrove, and D. Adamski
ENERGY DEPOSITION IN A MODEL OF MAN IN THE NEAR FIELD
Bioelectromagnetics, Vol. 6, No. 2, pp. 115-129 (1985a)

Stuchly, M.A., A. Kraszewski, and S.S. Stuchly
EXPOSURE OF HUMAN MODELS IN THE NEAR AND FAR FIELD--A COMPARISON
|IEEE Trans. Biomed. Eng., Vol. 32, nO. 8, pp. 609-616 (1985b)

Taylor, L.S. and A.Y. Cheung (eds.)
THE PHYSICAL BASIS OF ELECTROMAGNETIC INTERACTIONS WITH BIOLOGICAL SYSTEMS
U.S. Department of Health, Education, and Welfare, HEW Publication (FDA) 78-8055 (1978)

Taylor, L.S. and A.Y. Cheung (eds.)
THE MECHANISMS OF MICROWAVE BIOLOGICAL EFFECTS
Report of Workshop Held at University of Maryland, College Park, Maryland (May 14-16, 1979)

Taylor, L.S.
THE MECHANISMS OF ATHERMAL MICROWAVE BIOLOGICAL EFFECTS
Bioelectromagnetics, Vol. 2, No. 3, pp. 259-267 (1981)

Weil, C.M.
ABSORPTION CHARACTERISTICS OF MULTILAYERED SPHERE MODELS EXPOSED TO VHF/MICROWAVE RADIATION
IEEE Trans. Biomed. Eng., Vol. 22, pp. 468-476 (1975)

Wu, C.L. and J.C. Lin
ABSORPTION AND SCATTERING OF ELECTROMAGNETIC WAVES BY PROLATE SPHEROIDAL MODELS OF BIOLOGICAL STRUCTURES
IEEE Antenna & Propagation Society Int. Symp. Digest, pp. 142-145 (1977)