Analysis of the Exposure Levels and Potential Biologic Effects of the PAVE PAWS Radar System (1979)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

CHAPTER 2 BIOLOGIC EFFECTS RELATED TO EXPOSURE TO RADIOFREQUENCY WAVES The introduction of new physical and chemical factors into the environ- ment or an increase in the presence of existing ones dictates an analysis of potentially adverse biologic and ecologlc effects, especially potential effects on the health of exposed human populations. The purpose of this chapter is to review information pertinent to the assessment of the potential effects of human exposure to electromagnetic radiation in the radiofrequency range, particularly microwave radiation and to the extent possible radiation similar to that of the PAVE PAWS radar. It must be noted that the present state of knowledge of the effects of radiofrequency exposure on living systems, such as the exposure that may be anticipated from PAVE PAWS, is not adequate to define clearly the risks involved, although some general conclusions may be drawn relative to specific biologic end points. It should be emphasized that such limitations on the determination of risks of human exposure to microwave or to other radio- frequency (RF) radiation are not limited to the PAVE PAWS radar, but pertain in general to all sources of such exposure. The inherent uncertainties in the evaluation of the biologic effects of microwave and other RF radiation are due in part to the absence of an adequate data base, especially with respect to the effects of chronic or long-term exposure to low radiation intensities. The application of avail- able data, most of which have been derived from studies of acute exposure of other species, to the assessment of human effects is difficult because of, among other things, the manner in which such radiation is absorbed, which results in complex, wavelength-dependent, nonuniform energy distribu- tion in the bodies of experimental animals. Recent investigations have revealed that microwave and other RF absorption patterns depend on the frequency or wavelength of the radiation and the size, shape, and orienta- tion of the body with respect to the radiation field. It is therefore difficult to apply results of experimental-animal studies at a given radiation frequency to other species, such as man, exposed at the same or a different frequency. Although the consequences of nonuniform micro- wave-energy absorption in mammalian systems are not well understood, it is likely that such effects are involved in some types of alterations to be described in this chapter. The effects of microwave radiation on living systems may be arbi- trarily categorized as effects of high field intensities or low field intensities. High-field-intensity effects may be defined as effects associated with a degree of absorption of microwave energy that results in readily detectable increases in whole-body average temperature. Expo- sures of laboratory animals to such field intensities have been shown to result in a variety of physiologic alterations; prolonged exposure to -42-

such fields will lead to irreversible changes, such as cataract induction, and death as a result of excessive thermal stress. Low-intensity-field effects may be defined as effects that are not associated with detectable increases in whole-body average temperature. On the basis of most of the available data, the known or suspected low-intensity-field effects are reversible: conditions revert to normal soon after exposure is terminated. As previously noted, microwave-energy absorption depends on the physical and geometric characteristics of the absorber, as well as on the presence of reflecting surfaces in the surroundings. Consequently, it is not generally possible to specify ji priori whether a given microwave frequency and intensity will result in a low- or a high-intensity-field effect in a given mammalian species. However, on the basis of data derived from both theoretical calculations and experiments with various species exposed to microwave radiation, a field intensity of around 1.0 mW/cm will be used in this report to demarcate low- and high-intensity fields.* In the frequency range of 10-10,000 MHz, exposure to field intensities of 1 mW/cm will result in maximal energy absorption of approximately 1 mW/g, which is equivalent to the quantity of heat generated by normal human metabolism when averaged over the total body mass in the sleeping state. Absorption of microwave intensities of around 1 mW/g will thus add a thermal load equivalent to the whole-body average basal metabolic rate. Such absorption may result in temperature increases in specific tissues equiva- lent to those resulting from normal light physical activity. In view of such factors as the previously noted nonuniformity of body heating induced by microwave absorption, it is not possible to evaluate potential biologic effects of microwave radiation solely on the basis of comparisons with the heat generated by basal metabolic processes. However, if a time-averaged exposure intensity of 1 mW/cm is used to differentiate high- from low-field intensity effects, the class of effects to be anticipated from exposure to PAVE PAWS radiation may be specified. The available theoretical and exper- imental data relative to the field intensities anticipated in areas of public access in the vicinity of PAVE PAWS suggest that effects in humans will be restricted to the class of low-intensity effects, inasmuch as exposures apparently will be at intensities less than 1 mW/cm . To provide the necessary perspective for assessing the potential expo- sure effects of PAVE PAWS radiation, the Panel has reviewed the known bio- logic effects of microwave radiation, with particular attention to low- field-intensity effects. The available data are not extensive enough to permit specific conclusions to be drawn, but some general conclusions may be arrived at on the assumption that effects will be limited to low- field-intensity microwave effects. In contrast with the effects of exposure to low doses of ionizing radiation, exposure to low-intensity microwave or RF radiation has not been reported to result in detectable increases in the incidence of irreversible somatic or genetic effects, such as cancer or genetic mutation. This statement is consistent with the *"Field intensities" in this report refers to time-averaged intensities, unless otherwise indicated. -43-

known mechanisms of interaction of low-intensity microwave or RF radiation with biomolecules, which consist primarily of induced molecular rotations that are not known to be associated with irreversible molecular structural or conformational alterations. The genetic and somatic effects of ioni- zing radiation, however, are known to be a consequence of the substantially greater energy quanta of this type of radiation, which endow it with the ability to ionize or disrupt bonds in biomolecules and so lead to irreversi- ble molecular structural alterations that are manifested in irreparable damage to the organism. In general, therefore, the documented low-field-intensity microwave effects are not associated with irreversible alterations or increased morbidity or mortality. Such effects as do occur appear to be reversible and have been more prominently associated, by some investigators, with psychologic alterations in mood or attitude; this has led to the suggestion that it is principally the central nervous system that is sensitive to low-intensity microwave exposure. At present, except as noted below, there is no well-defined theoretical basis for the sensitivity of the cen- tral nervous system to microwave radiation, but experimental data suggest that neuronal membranes may be involved. Reversible alterations in neural systems, which have been reported to occur at field intensities that do not result in significant whole- body heating, appear to be related to the pulse modulation rate, as well as to the instantaneous or peak field intensity. Although specific effects, such as the audible perception of pulse-modulated microwave fields (the so called "microwave hearing" phenomenon), have been tentatively attributed to very rapid but slight heating of structures in the head, most of the reported central-nervous-system effects have not been adequately characterized or theoretically explained. In the absence of knowledge of the basic mechanisms of such effects and of verified data on their incidence in humans, it is not possible to predict the exposure conditions under which such effects will occur, nor is it possible to assess their conse- quences if they are, in fact, induced by exposure to low-intensity micro- wave fields. Again, in contrast with human exposures to ionizing radiation, where dose-response relationships for well-defined and quantifiable end points may be specified and relative or absolute risk determined, low- field-intensity microwave effects, because of their characteristics and the present lack of data, cannot be evaluated by such means. In view of the present limitations on the ability to assess the risks involved in human exposure to low-intensity microwave radiation, the approach used here in considering the potential exposure effects related to PAVE PAWS is to characterize, to the greatest extent possible, the anticipated exposures and to contrast these exposures with those encountered in other situations where humans are exposed to microwave or other RF radiation. In characterizing exposures to PAVE PAWS radar, all the factors that are shown to affect incident microwave exposure intensities and ab- sorption in humans must be taken into account to provide an indication of the most probable degrees of exposure, as well as the maximal exposures (or "probable worst-case" situations) that could theoretically be encountered. -44-

Although, as indicated above, it is not possible to translate the antici- pated exposures into determinations of risks to humans, a review of pertinent biologic effects is presented here to provide some perspective for the future consideration of the risks and benefits of the radiation exposure to be encountered as a result of the operation of the PAVE PAWS radar system. PHYSICAL FACTORS AFFECTING MEASUREMENT OF ABSORBED RADIOFREQUENCY DOSE IN BIOLOGIC SYSTEMS STATE OF KNOWLEDGE OF ELECTROMAGNETIC ABSORBED DOSE IN MAN AND ANIMALS The dose of microwave or radiofrequency radiation absorbed by biologic material depends on such physical factors as the size and shape of the irradiated object and the wavelength and frequency of the radiation. These physical factors are discussed here in the context of human exposure to radiation of the PAVE PAWS type, and then the information is applied to estimation of the dose that humans might absorb as a result of exposure to PAVE PAWS frequencies. Free-Space Irradiation The experimental condition that has been used in most studies is free- space irradiation of single animals. The whole-body absorption of electro- magnetic waves by biologic bodies depends strongly on the orientation of the electric field relative to the longest dimension (L) of the body. The highest rate of energy deposition occurs in fields that are polarized parallel to the longest dimension of the body (E|| L) and at such frequen- cies that the longest dimension is approximately 0.36-0.4 times the free- space wavelength (\ ) of radiation. Peaks of whole-body absorption for the other two configurations—longest dimension oriented along the direction of propagation TK|| L) or along the vector of the magnetic field TH|I L)— have also been reported for approximately twice the weighted average cir- cumference of the animals. On the basis of prolate spheroidal and ellipsoidal equivalents of biologic bodies, theoretical calculations have recently been given in a dosimetry handbook for frequencies up to jind slightly beyond^the resonant region for the aforementioned orientations--E|| L, K|I L, and H|I L. Nu- merical calculations with a realistic model of man have shown a more pro- nounced frequency dependence in the whole-body absorption at frequencies higher than the whole-body resonant frequency. Minor peaks in the suprare- sonant region are ascribed to maximums of energy deposition in the various body parts, such as the arm and the head. For the supraresonant region, the E|I L orientation has been studied most extensively.. In this region, whole-body absorbed dose is experimen- tally observed to be inversely proportional to frequency (F) for frequencies up to l.6res times the resonant frequency fr, where Sreg is -45-

the relative absorption cross section (defined as absorption cross section divided by physical cross section) at the resonant frequency. Empirical equations have been derived for the mass-normalized rate of electromagnetic energy deposition (specific absorption rate, SAR) for E| | L orientation. These equations are as follows: Peak absorption or resonant frequency: fr = (ll,400/Lcm) MHz. (9) For subresonant region — 0.5 fr<f<fr: SAR in mW/g for 1 mW/cm2 = 0.52 L2cm / f \2.75 incident plane waves mass in g ( f I (10) For supraresonant frequency region — fr<f<1.6 SAR in mW/g for 1 mW/cm2 = 5950 Lcm , (11) incident plane-wave fields ^MH mass i° 8 where L is the long dimension of the body in centimeters and Sres = 0.48 L,. • (12) ).48/ L Y mas mass in g Because human subjects cannot be used for experimentation, the empirical equa- tions have been checked by experiments with six animal species from 25-g mice to 2,250-g rabbits and found to be fairly accurate. For reasons not as yet understood, the measured SAR values for experimental animals are approximately 59% higher than those given by Equations 10 and 11, which were derived from experiments with figurines. For free-space E|| L irradiation, SARs considerably higher than the whole-body average are obtained for the neck, the legs, and the elbows (Gandhi ^t^ _al^. and Gandhi and E. L. Hunt, unpublished data), with the lower torso receiving SARs comparable with the average and the upper torso receiving SARs lower than the average. The deposition rates at the locations of maxi- mal absorption, or hot spots, may be 5-10 times the whole-body-averaged SARs given by Equations 10 and 11. Electromagnetic Absorption in Humans and Animals in the Presence of Nearby Ground and Reflecting Surfaces QulY highly conducting (e.g., metallic sheet) ground and reflecting sur- faces ' of infinite extent have been studied in attempts to determine absorption of electromagnetic waves in the presence of reflecting surfaces. For a standing-man model with feet in conductive contact with a perfect ground, there is a drastic alteration in SAR as a function of frequency. For E|| L -46-

orientation, the new resonant frequency is roughly half that given by Equa- tion 9. At this lower resonant frequency, the SAR is about twice that at the peak absorption frequency for free-space irradiation. For feet in conductive contact with the ground, the highest SARs are observed for the ankles and legs. The deposition rates at the hot spots, again, are 5-10 times larger than the whole-body averaged SAR under these conditions. The nature of the ground effects on SAR (for E|| L orientation) is such that even a small separation, '' from ground (breaking the conductive contact) is sufficient to eliminate much of the ground effect. For separa- tions from ground of more than about 3-4 in. (7.6-10.2 cm), the total energy deposition and its distribution are identical with those for free-space irradiation conditions. Even for a human model in conductive contact with a perfect ground, the energy deposition in the supraresonant region (f > 2-3 f ) is comparable with that for free-space irradiation. Other orientations and finite-conductivity ground effects on SAR have not been studied. For highly conducting (metallic sheet) reflecting surfaces of flat and 90°-corner types, increases in SAR by factors as large as 27 have been observed for the E|| L orientation. Most of the work done so far has concentrated on frequencies close to the resonant region. The observed increases are ex- plained on the basis of antenna theory. Indeed, for incident-plane waves for~~E|| L orientation, most of the observed results imply that the irradiated target acted like a pickup half-wave dipole with reflecting surfaces close by. Finite-conductivity, finite-size reflecting surfaces and other orien- tations have not been considered. Results for frequencies higher than about 8.5 times the resonant frequency (550 MHz for man) have not been obtained, even for highly conducting reflectors. Enclosed structures, such as rooms, may act as lossy resonators with electromagnetic fields being coupled from the windows. If such structures have highly reflecting walls, field enhancements by one or two orders of magnitude may indeed be possible. However, because walls typically en- countered are not very reflecting, power-density increase by a factor of more than about 5-10 may not be realistic. Further research into the reflec- tion characteristics of these structures is needed in order to describe precisely the nature of field enhancement. Head Resonance Gandhi et^ jil. have recently identified a frequency region for the highest rate of energy deposition in the head. The head resonance '' occurs at such frequencies that the head diameter is approximately one- fourth of the free-space wavelength. For the intact (adult) human head, the resonant frequency is estimated to be around 350-400 MHz. At head resonance, the absorption cross section for the head region is approximately 3.0 times the physical cross section with a volume-average SAR that is about 3.3 times the SAR averaged 9Ygr the whole body. Both values greatly exceed numbers reports earlier '^ for spherical models of the isolated -47-

human head. Numerical calculations based on 144 cubical cells of various sizes to fit the shape of the human head (340 cells for the total body) yielded local SARs at hot spots (above the palate area and the upper part of the back of the neck) about 5 times the average values for the head. Multianimal Effects55 For resonant biologic bodies close to one another, antenna theory may be used to predict the modification in SAR relative to free-space values. For two resonant targets separated by 0.65 to 0.7A, the highegt SAR (150% of the free-space value) can result for man and animals in an E|I L orienta- tion for frontally (broadside) incident plane waves. For three animals in a row with an interanimal spacing of 0.65 A, the central-animal SAR would be roughly 2 times that for an isolated animal, and the end-animal SARs would be approximately 1.5 times that for an isolated animal. Full implications of the multibody effects on SAR are not completely understood, even though pilot experimental studies with anesthetized rats showed that the above-mentioned increases may also occur for sub- resonance and supraresonance regions. Other orientations, irregular spac- ings, and non-free-space exposure conditions have not been considered. Multilayer Effects Most of the results outlined above were obtained for homogeneous models of man with tissue properties averaged on the basis of 65% muscle, skin, and tissues with high water content and 35% fat, bone, and tissues with low water content. Some of these results were checked by experimentation with small laboratory animals. Gandhi's group (P. W. Barber, 0. D. Gandhi, M. J. Hagmann, and I. Chatterjee) has recently studied the effects of tissue layering on energy deposition in a multilayer model of the whole body of man. Their calculations showed that layering in humans can alter the results somewhat for frequencies above 500 MHz, but the homogeneous model was found to be quite appropriate at frequencies lower than 500 MHz. Calculations for the Electromagnetic Absorption in Humans for PAVE PAWS Frequencies The Panel used the dosimetric information summarized above to estimate the electromagnetic absorption in man for PAVE PAWS frequencies. These calculations were based on the following assumptions: a 70-kg, 1.75-m average- sized man; a 32.2 kg, 1.38-m, 10-yr-old child; and a 3.5-kg, 0.4-m average- sized infant. The average head diameters for the man, child, and infant were taken to be 21, 12, and 8.7 cm, respectively. A free-space incident power density of 0.1 mW/cm was assumed in calculating the mass-normalized rates of electromagnetic energy disposition (SAR) given in this section. Table 13 shows calculated SARs for the c|I L orientation for the whole body and the head and estimates of expected deposition rates for local areas of the body and head. Resonant frequencies for the body and head are also described. Table 13A describes SARs for free-space exposure conditions, and Table 13B for non-free-space conditions. Table 14 summarizes the results of calculations of the "probable worst-case" SARs predictable under conditions of exposure to PAVE PAWS radiation in buildings with unscreened windows facing the antenna, where field power density may be increased. -48-

TABLE 13 Mass-Normalized Rates of Energy Deposition (Assumed Incident Power Density, 0.1 mW/cm ) A. Free-space irradiationfE | | L Orientation 70-kg, 1.75-m adult man I 32.2-kg, 1.38-m 10-yr-old I 3.5-kg, 0.4-m j child I ~ infant From Equation 9, whole-body resonant or peak absorption frequency (fr) in MHz = ll,400/Lcm, where Lcm is length in centimeters. fr = 65.2 MHz I I f_ = 82.6 MHz i L I I fr - 285 MHz From Equation 11, the whole-body averaged SAR = 0.0035 mW/g at 420 MHz 0.0033 mW/g at 450 MHz 0.0061 mW/g at 420 MHz 0.0057 mW/g at 450 MHz 0.016 mW/g at 420 MHz 0.015 mW/g at 450 MHz Deposition rates 5-10 times higher than whole-body average SARs are expected for the neck, the legs, and the elbows for these irradiation conditions. Head resonance frequency (fn r) "7,500/d^, where dn is average head diameter in centimeters fh,r * 357 MHz Head*averaged SAR:72 0.019 mW/g at 420 MHz 0.016 mW/g at 450 MHz fh =625 MHz n,r Head-averaged SAR:55 0.014 mW/g at 420 MHz 0.011 mW/g at 450 MHz h 862 MHz Head-averaged SAR:55 0.0075 mW/g at 420 MHz 0.014 mW/g at 450 MHz Deposition rates about 5 times higher than head-averaged SARs may occur at hot spots above the palate area and the upper part of the neck. -49-

TABLE 13 (Continued) B. Non-free-space lrradiation E| | L Orientation The above values should also be applicable for conditions of electric contact with high-conductivity ground. SARs increased by factors of 5-20 may be encountered in the presence of highly conducting (metallic sheet) reflecting surfaces of flat and 90°- corner types. Smaller increases are expected for imperfectly reflecting surfaces. Even though enclosed structures, such as rooms, have not been studied at length, field increases by factors of some 5-10 have been reported, presumably owing to reflections from walls, of electromagnetic energy coupled in from windows. Increase in SAR by a factor of no more than 2 may occur, owing to proximity to other human beings. Other Orientations The SAR values for other orientations are likely to be no more than those for E|| L orientation given above. -50-

TABLE 14 Rates of Energy Deposition for "Probable Worst-Case"3 Exposure Conditions (Assumed Incident Power Density, 0.1 mW/cm ) 70-kg, 1.75-m adult 32.2-kg. 1.38-m 10-yr-old I 3.5-kg, 0.4 m man I child I infant I I Whole-body-averaged SAR 435 MHz, corresponding to average PAVE PAWS frquency: I I 7.5x0.0034= I 7.5 x 0.0059 = 0.044 mW/g I 7.5x0.0155= 0.026 mW/g I 0.12 mW/g SAR at hot spots: I I =0.26 mW/g | -0.44 mW/g | -1.16 mW/g Head-averaged SAR: I I :7.5 x 0.0175 = -7.5 x 0.012 = 0.09 mW/g -7.5 x 0.0108 0.13 mW/g j I 0.08 mW/g SAR at hot spots in the head: I I = 0.66 mW/g "0.47 mW/g -0.4 mW/g aField increase by factor of 7.5 assumed in these calculations. Such a factor is likely to be among highest that may be encountered in buildings. Field increases of this type would be fairly localized. Furthermore, metallic screen at windows facing antenna installation would keep most fields out. -51-

The SARs calculated in Table 14, for the "probable worst case" (maximal SAR at a "hot spot" in the head of 0.66 mW/g) for an adult human being, may be compared with the typical metabolic rates for an adult human being ' Whole-body averages Basal metabolic rate 1.08 mW/g Light-activity metabolic rate 1.66 mW/g Metabolic rate for slow walking . 3.32 mW/g Local values Brain metabolic rate 11 mW/g Heart-muscle metabolic rate 33 mW/g LOCALIZED POWER ABSORPTION DUE TO ATTACHED INSTRUMENTATION AND IMPLANTS When conducting objects, wires, or electrodes (such as surgical pins or pacemakers) are brought into contact with or implanted in biologic tissues exposed to electromagnetic fields, high-intensity fields may be induced locally where the conductors contact tissue. These fields can be much greater than the fields that would normally be present without the conductors. Although it is beyond the scope of this discussion to analyze quantitatively the fields induced in tissue owing to various implants under all exposure conditions, some examples based on simplistic analyses can be discussed. 69 Guy e£ _al:. have demonstrated some possible situations by using simple first-order analytic determinations of the field increase that occurs in tissues as a result of the presence of conductors. These sim- ple examples illustrate the increase due to wires that connect external instrumentation to electrodes in contact with the tissues, implanted en- capsulated instrumentation (such as pacemakers), and implanted conductors (such as surgical pins and prosthetic joints). In all these cases, currents are Induced in the conductor portions of the instrumentation or implants and result in field increase and in an increased SAR (We), which is much greater than the normal SAR (W). The value of W will in general increase with the length of the leads or the implant. The factor of SAR increase due to an external conductor of length L and radius a in contact with an exposed subject was found to be: (13) W -52-

when L is assumed to be smaller than the wavelength, £t> is the permittivity of free space, e * is the complex dielectric constant of the tissue,*1*** is the electric conductivity of the tissue, and f is the frequency. When the length of the wire is appreciable, compared with the wavelength, the intens- ification factor would be even greater than that for short leads. For muscle tissue and an exposure frequency of 450 MHz, |£m*| is equal to 78, and the electric conductivity of*vnis 1.43 S/m. Inserting these values in Equation 13, we obtain W — = 1.36 W With L/a = 10, Wg/W 3 -2 6 a (14) 1.67 x 10J; with L/a = 100, Wg/W = 4.43 x 10°. Thus, the presence of the wire results in large SAR intensification at the point of contact. The SAR increase due to metallic implants in tissues may be illustrated by considering the conducting prolate spheroid with major axis L and minor axis 2a imbedded in tissue in the presence of an electric field. The enhancement factor discussed by Guy is: !?e = [u2-l) (ucoth-1u-l]-2 , (15) where u cosh[tanh 2a]. (16) For a spherical implant where L/2a approaches unity, the factor is 9. When the ratio is 5, the factor is 321; when the ratio is 10, the factor is 2.43 x 103; and when the ratio is 100, the factor is 5.4 x 10. With in- creasing ellipticity, the SAR increases substantially, but the volume of tissue affected is small, because the intensification region extends to a distance of about the radius of the conductor. For an implanted insulated wire, with conductor radius a, insula- tion radius b, insulation dielectric constant £j , and the end of the con- ductor in direct contact with the tissue, the enhancement factor may be expressed as 2 U W V mfeoed o.ln(Wa) B (17) . - „ For a frequency of 450 MHz with£j = 2.25 and b/a = 2, the enhancement factor given above may be expressed in terms of the ratio of wire length of radius: W _§. = 3.23 x 10 W (18) -53-

When L/a = 5, the factor is 2; when L/a = 10, the factor is 3.2; and when L/a = 100, the factor is 3.23 x 105. This indicates that substantial field increases can occur even for the implanted insulated conductors. However, such increases probably result in highly localized transient heat- ing in the immediate vicinity of conductive implants in tissue exposed to microwave fields with time-averaged intensities greater than lyW/cm . CARDIAC-PACEMAKER INTERFERENCE Mitchell reported an extensive study on the interference of cardiac pacemakers from radar-like pulses, including those operating at frequencies of 450 MHz. The study indicated that field intensities of radar-like pulses above some threshold can disrupt normal pacemaker function and create a potential hazard for the user; that the pacemakers tested were most sus- ceptible to interference at frequencies of 500-1,000 MHz, with the inter- ference threshold inversely proportional to pulse width; that fields with a pulse-recurrence rate of 1-10 pulses/s and a peak above the pacemaker's interference threshold would stop the operation of a pacemaker; and that when the effective pulse-recurrence rate was greater than some inherent value (which depended on the particular device), the pacemaker would revert to an interference-rejection mode by operating at a fixed rate. It was noted in the report that operation of a pacemaker at a fixed rate is generally judged [.unhazardous, whereas inhibition of the output of the de- vice is judged hazardous. Pacemakers were tested with the Association for the Advancement of Medical Instrumentation technique of simulating implantation by placing the pacemaker in an 80-cm x 40-cm x 20-cm container made of 5-cm-thlck plastic foam and filled with 0.03 m saline solution. The pacemaker and leads are placed so that there is 1 cm of solution between the pacemaker and the wall of the container. The implanted pacemaker was exposed under controlled laboratory conditions to circularly polarized 450-MHz fields with electric-field strengths up to 292 V/m. The pulse widths and pulse-repetition rates were varied between 1 ; s and 1 ms and between 2 and 40 pulses/s, respectively. Table 15 summarizes the results of Mitchell's report pertaining to 450 MHz inference. An adverse effect is defined when the pacemaker rate falls below 50 beats/min (bpm) or exceeds 125 bpm as a direct result of electromagnetic interference. The data in the table illustrate the wide range of sus- ceptibility to electromagnetic interference from 8 V/m to over 300 V/m of the 23 pacemaker models tested. Results show a dramatic improvement of new models, such as the American Optical 281143 over the older models AO 281003 and 281013 and the Starr Edwards new model 8116 over the older model 8114. The report indicated that, with design lmprovements in the newer models, the interference problem should be eliminated. However, older, maximally susceptible pacemakers may be affected by exposure to PAVE PAWS radiation fields, especially near the exclusion area, where in- stantaneous field strengths in excess of 10 V/m may be encountered. The scanning mode of the radar beam would, however, be expected to induce only transient pacemaker interference, rather than a complete cessation of operation or a continual increase in rate exceeding 125 bpm. -54-

0) O CM O O O O 1 "" I rH I l " """"" ^ l l "" I ^ \ t l I 00 « CO O O rHOOO iH mcMOOOOO 0 O CO rH| -* O O -*OOO ** rHrHOOOOO O O ** H5P CMCOCOrHCOCOCO rH COCOCOCOCOCO CO CM COCO AA AAA AAAAAA A <4HS .a o 0I O 4J 0 0) h C ^-^ 4J 4-i M CO cfl O V e H V rH <H rH X^ X"N X"-\ X™\ x™\ X"N X^ X^ X^ V-l <1) 4J O O l I l O l 11 O O O O . l O l j O 0) e • invo o co rHincM<^o o o ococdo ,£ 60 4J rHCM O st -* rH rH rH O iH O O-*S CO *O • • o CO CM rH CO COCOCMOI4J •H B CM A A A A y rH * a CO 3 IH A a. co 4J 0) • y co tt C M 0) rH •H X"N X"N ^^ X^N /^N ^^ X^ ^^ ^N Q) rH OO^u^O O O OO-C-u M PM OJ ^X ^*/ OJ fsj ^X ^^ ^^ ^^ ^->" U O w •H rH 0) 1*4 • • cfl X^> ^^ -rj Id 0> M Cd III O OomCTi )^-rH rH |l|vO||O l IO -3- l-» OI-H III O OO-H O-3- ** IlicMlirH 1 IO rH O CTJ "SJ 2 0 O COCOCO CMrHlH COrHrHCO O rH AAA A •H CO o *H CO u m 4J CO Cfl (0 <H co cd Pu 4_> m a> co E y X-N CL CO V a 0) J3 m -H o a Ptt X^ ^-^ X"N X™\ X-N X^ X"*s X™\ X™\ X"N X™\ X^ X^ X^ X"N, X-N fj iH •O 0) 43 OO O OOOOOOOOO O O OOrHin U cd 3 ^^ 0) S«X ^—> N^r X^ N^ NMX V^ N«X V_^ x^* *-— ' *— ' ^— ' X^ ^— ^ V-^ O CM er CO COv^OO i— 1 OO°^OO ONCO ON incMCOvOOOO O CO O r^ CO X rH rJ U 0) 8 rHCMO -* OOrH CMCM CM iHrHCMCMOOrH O CMO O ON 00 (4 ^ 3 COrHCOCO COCOCOCOrH Ol CO < n PM <x CM AAA A A A A r4 *O SH O X 0) MH a 4J 0) o y ca <H 4-) ^ •o 4J y cu <r4 0) o ^3 'j_. V- .e C MH O ca O 0) 0) u l '"""*• u 4-I 0) B CO & H e M ^O ££ 9 (U N-X y Q. T3 OJ 0) e ^1 4-J *w O 3 <4H | <: CM C •9 CO CO CO *^ vD O <XJ O rH <t NO rH CM in < •H w M e 4-l Ol OOi— i rHr»-r~r» i H CO CO ^i iHrHrHr^WIHOO i H r4 (4 rH 3 CO OO OC CJlHCOO CMCM • Ol • o i 9 4J CMCMCM coo-*y<!<! -*vo H rH OL. Ol ^ S M COyrH-H I-HrHrH Oi rH y •o •H CO cd n rHrHr-j iH-Hfiyy OrHl-H l CO Q 0) s t M p cfl -H -H rHOOOO O O) DO C <4H 0) "" c 1 3 43 yyyst M4JO4J hH CMCMCO*^Oi-HO i -* y cfl IH <4H •H-HIH PL. o <! O co 4-j 4_i ^ ** *^ s t m m o o coco i— i i 4-) Ol Ol O CO 3 4-j ij 4_i Q y I irH y y ^ oo O^ c^ 0^ <y* C?N o PQ *rj *X3 CM P^ *H 43 u_ S Z a. cx ex M -H-HC•H cuoi O inminminmcTi u u oo M s I-l r7 \4 rH cDOO MCCOCrHrHOO hCflCOcOS _y 4Jg4Jg uu) I yyyyyyy o 33 B m 4J o e O Ol CO 0) 0) SCC -H <!OWO CO -H-H-H-H-H-H-H 4-) TaTa 43 C «cd C r-jrH CCCCCCC u M w y O r* T3 h C -H H cd o yyy o cocococa c d eg wi o o o o o o o o> oi >-. MH Ol 9 S S •H-H*H l_i -H-n*HtH I-i I-i O lj M I-. K. |_i J-i u CO t-t U 4J 4J oi y Cn 0) I-l I-l 1^ 4J T3 T3 T3 T3 OlOI y 4-i 4-i 4-l 4-l 4J 4-l 4-l <1) I-l I-l £ CO cd s -H y -o OlOlOl O V-i i-j p u C C *O *^*OvOT)*O*X3*O y COCO *H 4J 4J 4J 4-l CO C 6 S B •H OOOO OlOI Ol Q)0)ci)OIOIOIOI CO 4J4J 4J *H cd a> oi p^ Qj ^<< m uC5oC5 oo z ssssxss PL. cncn cn > Q 43 C id -55-

BIOLOGIC EFFECTS EFFECTS ON CENTRAL NERVOUS SYSTEM AND OTHER NERVOUS TISSUE Stimuli to the nervous system may be evaluated on the basis of their behavioral effects or the concurrent physiologic or biochemical changes induced in the brain or other nervous tissue. In some instances, there may be alterations in tissue ultrastructure, discernible microscopically. Effects of low-level, nonionizlng electromagnetic fields on nervous tissue are related in part to RF carrier-wave frequencies and to patterns of amplitude modulation. Brain tissue may be sensitive to imposed fields only within narrow ranges of incident energy. Electroencephalographic Alterations Both acute and chronic exposures to microwave fields have been re- ported to change EEG patterns in humans and animals. According to Baranski and Edelwejn, technicians repeatedly exposed to presumed high- level fields in radar-repair shops may eventually exhibit flat EEG records associated with subjective complaints of headaches and copious sweating, but no precise data on field characteristics are available. s 1 "\ In animal studies, Baranski and Edelwejn noted high-amplitude "desynchronized" records in rabbits after 3-4 months of exposure for 2 hr a day to fields of 2,950 MHz, average power of fields with 5.0 mW/cm2 and 1.0- ys pulses at 1,200 pulses/s under far-field conditions in an anechoic chamber. Major changes in rabbit EEG patterns have also been reported by U.S. investigators after 4-6 weeks of exposure to radiofrequeuey fields at much lower frequencies (3-5 MHz) with 14-Hz modulation. Two opposing effects were noted. With fields at 90-150 V/m, there was increased activity at higher EEG frequencies (10-15 Hz). Fields at 500 V/m increased the effects of lower frequencies (4-5 Hz), and the increased low-frequency activity was associated with suppression of high-frequency activity. It is unlikely that these effects are attributable to thermal changes in brain tissue. Facilitation of microwave-induced desynchronization by small amounts of pentobarbital has also been reported. ' Brain sensitivities have been studied at a carrier frequency of 450 MHz, with sinusoidal amplitude-modulation frequencies of 5-30 Hz, which are modulation components used in long-range radar systems. Findings in laboratory studies with similar RF and modulation characteristics may disclose bioeffects likely to occur near long-range radar transmitters with appropriately scaled field intensities. Sensitivity of the EEG modulation frequencies on a low-level (0.8- mW/cm ) VHF (147-MHz) carrier was demonstrated by Bawin ££ jl. in cats; imposition of this field increases occurrence of a brief burst of -56-

EEG waves in particular brain structures, when the modulation and EEG wave-burst frequencies are similar. This EEG burst can be trained as a conditioned behavioral response to an environmental light or sound stimulus. The rate of correct responding is then substantially increased, and extinction of the response is delayed for many days in the presence of the RF field. After 20 min of exposure to 450-MHz fields, with a 1.0-mW/cm amplitude modulated at 16 Hz, neonate chicks have sharply higher EEG power spectral density at frequencies from 14 to 25 Hz for 2 h or more. 6 Chronic exposure of rats to 3.0-GHz, 5.0-mW/cm fields with 500 pulses/s is followed by a persistent 500-Hz component in the EEC spec- trum. These studies support the concept that microwave fields modulated at EEG frequencies may actually entrain neuroelectric activity of brain structures that generate intrinsic electric rhythms at the same or closely related frequencies. Microwave fields modulated at much higher frequencies, typically at radar pulse-repetition frequency (PRF) rates between 300 and 1,500 pulses/s, may also modify EEG activity and produce effects not seen with continuous-wave (CW) fields of the same average power. Fields pulsed in the audible range (2850 MHz; 2.7 ys; repetition frequency, 357 Hz; average power density, 30 mW/cm ) diminished sensitivity to audiogenlc seizures in rats exposed for 4 h a day for the first 10 weeks of life. A distinction should be drawn between microwave fields that appear to influence brain-tissue excitability directly and fields that are sufficiently intense to induce thermal stress. The latter may be associated with endo- crine changes, and EEG patterns may then shift as an element in the stress response. This distinction is clearly justified for frequency-specific low-frequency modulations cited above. Those fields raise the isolated brain temperature by less than 0.1°C, and brain-tissue dosimetry shows gradients of 100 mV/cm for incident 450-MHz fields of 1.0 mW/cm . ° Indeed, EEG entrainment has been reported for 7-Hz extremely low-frequency (ELF) fields, which produce far weaker tissue gradients of around 10" V/cm Comparison and neuronal firing thresholds of single invertebrate ganglion cells (Aplysia californica) in microwave fields and during warming indi- cate that both alter firing rates with shifts around 0.1°C, although qualitatively different responses may occur in the same cell with compara- ble small temperature increments induced by the two stimuli. ' Absorbed radiation doses in brain tissue of small mammals may be modi- fied in a complex way by resonances attributable to whole-body geometry as a function of the wavelength of the incident field. The cranial cavity itself may be resonant, with sharply higher absorption of 2,450 MHz for the monkey. For the larger head of man, this resonance occurs between 350 and 400 MHz, and for a child, between 600 and 850 MHz56—a factor to be considered in establishing safety guidelines near radar trans- mitters operating in this frequency range. -57-

Microwave Effects on Blood-Brain Barrier The passage of ions and molecules from cerebral blood vessels to the environment of brain cells is normally restricted or prevented by barriers associated with the blood vessels themselves. A continuous layer of endothelial cells line cerebral capillaries. Electron microscopy shows so-called "tight junctions" joining these cells, which appear to lack pinocytotic vesicles that provide a mechanism for trans- membrane transport of molecules into the interior of other cells. Oscar has described microwave-induced changes in brain uptake and distribution of blood-carried proteins, electrolytes, and large water-soluble non- electrolyte molecules in rats, Chinese hamsters, and rabbits. Oscar and Hawkins noted an apparent increase in permeability of the blood-brain barrier arising from both pulsed and CW microwaves at power levels considerably below 10 mW/cm , with evidence of increased effects of pulsed over CW fields at similar average power. Studies by Oscar, Albert, and Merritt et^ ^1. ' have shown increases in blood-brain exchange of horseradish peroxidase, sodium fluorescein, radiolabeled saccharides, and labeled dibasic phosphate in rats, Chinese hamsters, and rabbits. These substances were scattered in random parts of the brains and showed no consistent pattern of localization of any functional region, at least for substances with molecular weights of 40,000 or less. Fields in these studies ranged from 918 to 2,480 MHz at incident ener- gies of 0.5-10 mW/cm^. A thermal basis of these effects has not been clearly established. I 1 Q Oscar pointed out that these alterations in the blood-brain barrier may be reversible. This may indicate changes in tight-junction integrity or pinocytotic transport, rather than cell-membrane destruc- tion. The observed changes in barrier permeability may also be related to altered cerebral circulation induced by neural stimulation. It is known that the barrier may be reversibly opened by convulsive epi- sodes, concussion, hypercapnia, acute hypertension, and ionizing radia- tion. Effects of Neurotransmltters in Nervous Tissue Brain neurotransmitters now include a much broader range of substances than was suggested by initial experiments with cholinergic and adrenergic agents three decades ago. Known or suspected brain neurotransmitters now include amines (serotonin, dopamlne, and norepinephrine), amlno acids (gamma-aminobutyric acid, glutamic acid, and taurlne), and a variety of peptides believed to exercise neurohormonal influences at neuronal-membrane receptor sites, apart from any direct influence that they may exercise as transmitters released at synaptic junctions—including peptlde hormone molecules and fragments of these molecules. Both western and eastern European studies have therefore focused on possible effects of chronic exposure to low-intensity microwaves on responses to neurotropic drugs that have effects on specific classes of transmitter substances. -58-

Baranski and Edelwejn tested rabbits previously exposed chronically to 3.0-GHz pulsed microwave fields at a power density of 7 mW/cm (total irradiation, 200 at 3 h/day). The authors described this exposure as "sub- thermal." The irradiated animals showed a rapid, desynchronized EEC pattern, which may be attributable to chronic activation of the midbrain reticular formation. Chlorpromazine (at 4 mg/kg) produced a regular rhy- thmic EEG wave pattern that was consistent with the known action of chlor- promazine in reducing midbrain reticular activity by adrenergic blocking. However, the convulsant hydrazine drug pentylenetetrazole, known to act on the thalamic part of the reticular formation and motor cortex, produced greater effects in irradiated animals than in controls. Baranski and Edelwejn compared the effects of 2,950-MHz CW fields and pulsed fields (1,200 Hz with 1- ys pulses) at 5-7 mW/cm on phosphorus-32 incorporation into brain lipid and nucleic acid fractions. Turnover indexes were lower for irradiated than control animals, and greater for pulsed than CW fields at the same mean power density and carrier frequency. Brain concentrations of norepinephrine, dopamine, serotonin, and 5- hydroxyindoleacetic acid were sharply reduced in animals rendered hyper- thermic with 1,6-GHz radiation at 80 mW/cm , with focal absorption in the hypothalamus; but the possible susceptibility of brain neurotransmitter release to much weaker microwave fields was not addressed in these studies. Fields of 10 m W/cm caused marked slowing of serotonin turnover and a decrease in the firing rate of individual serotonin-dependent neurons. It has been suggested that, inasmuch as these neurons are known to partici- pate in the regulation of sleep and wakefulness, as well as in body-tempera- ture regulation, these findings may account for the behavioral and functional effects of low-intensity microwaves reported by Soviet and eastern European researchers. ° For the rabbit, a dose-dependent decrease in duration of pentobarbital-induced sleep followed exposure to 1.7- and 2.45-GHz microwave radiation at intensities of 5-50 mW/cm ; this suggests thermal stress as a possible, but by no means exclusive, mechanism. There is little direct evidence of modification of cholinergic trans- mission in brain tissue by microwave fields. But Soviet workers reported that blood concentrations of cholinesterase were sharply reduced in rabbits exposed chronically to 50-MHz, 2.5-GHz, and 10-GHz fields. Pulse rates of the microwave fields were either 20 or 1,000 pulses/s, and field intensi- ties ranged from 0.5 to 10 yW/cm . Cholinesterase activity dropped by almost half during 50-MHz, 10 yW/cm exposures 90-120 days after the beginning of exposure. A reduction of 25% occurred in similar exposures to the microwave fields at 10 yW/cm . Gordon and colleagues corrobor- ated these findings, at least for fields down to 1.0 mW/cm . Gordon pointed out that the biochemical technique measures "nonspecific" serum cholinesterase and that changes in tissue activity of this enzyme after whole-body irradiation result not from a direct action on molecular struc- tures, but from changes in tissue enzyme concentration attributable to disturbances in neurohormonal regulation of metabolic processes. -59-

Western workers have now confirmed important aspects of these findings of Soviet investigators. Lovely et al. found that rats exposed to 500 yW/cm , 2,450-MHz CW fields T"h7day for 3 months showed initial de- creases in blood chollnesterase, and then a return to control values. The isolated rat heart was slowed (bradycardia) by 960-MHz CW fields at a speci- fic adsorption rate of 1.5 W/kg (incident energy of approximately 1.5 mW/cnr). When parasympathetlc and sympathetic nerves were simultaneously blocked by atropine and propranolol, respectively, irradiation was without effect. Atropine plus irradiation caused cardiac acceleration (tachycardia), whereas propranolol plus irradiation produced bradycardia. The authors concluded that effects of these weak fields may indicate a microwave-neuron or microwave-synapse interaction by a mechanism other than generalized heating of tissues. Alteration in Cation Fluxes or Binding Excitatory processes in nerve fibers with production of a nerve im- pulse involves a transient loss of membrane potential. In the resting state, this potential is about 70 mV. The interior of the fiber is nega- tive with respect to its surroundings—a condition accounted for by high internal potassium and low internal sodium concentrations. A sudden massive inward sodium current and a slower, later outward potassium current provide a reasonable model of the fiber action potential. Calcium is essential for the stability of the membrane potential. It is present in high concentration (2.4 mM) in cerebral extracellular fluid. By contrast, general cytoplasmic contents are kept low (0.1 yM) by active transport into organelles or back into the membrane. For nerve fibers, calcium ions have been described as a "plug in the bath" with respect of their ability to limit and regulate movement of monovalent sodium and potassium cations through transmembrane channels. For nerve cells, in contrast with nerve fibers, actions of calcium ions on membrane functions are known to be vastly more complex. Cere- bral neurons have dendritic branches typically spreading far away from the body of the cell. Membranes of dendrites and cell bodies participate in the sensing of immunologic, hormonal, and neurotransmitter stimuli. Dendrites may also sense small electrochemical gradients in their immediate environment. Sensing of these weak cell-surface events and their trans- duct ive coupling to the interior of the cell appear to depend on calcium at every stage. As to the structural substrate, available evidence supports a "fluid mosaic" in which protelnaceous particles intrude within a fluid lipid bllayer. These particles have stranded terminals extruded onto the membrane surface, terminating in anionic amlno sugars (siallc acids). The membrane surface thus forms a polyanionic sheet with a strong affinity for cations, particularly calcium and hydrogen. Structural continuity of the protein particle from the exterior to the interior of the membrane provides a communication pathway, which may undergo conformational changes associated with the binding or release of calcium ions. -60-

cm Calcium binding in cerebral tissue has been found extremely sensitive to some weak vhf (147-MHz) and microwave (450-MHz) fields. With an in- cident energy of 0.8 mW/cra , 147-MHz fields with sinusoidal amplitude- modulation frequencies from 0.5 to 35 Hz increase Ca efflux from freshly isolated chick cerebral hemispheres by about 15% for modulation frequencies of around 16 Hz. Increased effluxes were noted over the band of modulation frequencies from 6 to 20 Hz. Unmodulated carrier waves and modulation frequencies above and below this band were without effect. These findings have been confirmed in all essential aspects by two independent groups. ' Similar experiments with a 450-MHz, 1.0-mW/ field, amplitude modulated at 16 Hz, also caused an increase in calcium efflux of more than 10%.21 Studies by Bawin ££ al. '21 and Blackman jet al. cited above have shown that, in addition to a modulation-frequency "window," there is an intensity window within which these modifications in calcium binding occur. Increased calcium efflux was maximal for field intensities around 1.0 mW/cm2 in the studies at 147 MHz by Blackman £t al. and at 450 MHz in the studies by Bawin et al. Bawin et al. defined an effective window between 0.1 and 1.0 mW/cm^. This corresponds to an electric-field gradient between 17 and 55 V/m. Theoretical Models of Microwave Interactions with Tissue The window character of brain-tissue sensitivity in both frequency and intensity domains strongly suggest that these interactions arise in nonequilibrium processes characterized by resonant interactions of macromolecular fixed-charge sites over long atomic distances. *' ' Theoretical and experimental models of these processes currently under review include the following: • Thermal-phase transitions with quantum amplification as a resonant phenomenon in electric-charge dipoles on surface proteins at very low frequencies. • "Pumping" of charge sites on proteins by microwave radiation to produce long-range interactions; oscillations at very low frequencies would occur on the basis of limit-cycle phenomena. ' QO • Limit-cycle interactions in surface proteins. • Tunneling, which may occur with protein carriers in the length of membrane-surface proteins. '' BEHAVIORAL AND SENSORY EFFECTS Very small changes in behavioral characteristics are often the most sensitive indicators of an organism's response to exposure to microwave or radiofrequency electromagnetic radiation. However, these measures, -61-

as reported in the literature, have provided little insight on the ques- tion of the possible hazards of such exposure, because the distinction between "effect" and "hazard," although critical, may be derived differ- ently by different observers. For example, eastern European scientists more readily equate "effect" with "hazard" than do western scientists. The difficulty of isolating and measuring a behavioral effect of low- intensity electromagnetic fields separately from behavioral effects of other, concurrent environmental exposures also complicates the process of interpreting data related to behavioral effects. In the following discussion, confirmed reports of behavioral effects of microwave irradiation are reviewed from the perspective of thresholds (the point at which effects are first measurable). Perception Dose-rate thresholds of the albino rat's perception of 60-Hz sinu- soidally modulated microwaves at 2,450 MHz in a multipath field, as measured by conditional suppression, were found to range from 0.6 to 1 mW/g during 60-s excitations of the field. The associated minimal energy dose (0.6 mW/g x 60 s) is less than 36 mJ/g, because detection of the field occurred well before the 1-min periods of excitation were termi- nated. The estimated energy-flux densities of an incident 2,450-MHz plane wave that would result in an energy-absorption rate of 1 mW/g lie between 2 and 6 mW/cm and would depend on the animal's absorptive cross-sectional area and its orientation with respect to E and H vectors of the incident field.57 Perceptual thresholds for short-duration pulsed microwave fields have not been reported, although early reports ' revealed that human beings can readily hear a train of pulses at peak flux densities near 100 mW/cm and averaged densities near 50 ji W/cm . The RF-hearlng effect is believed to result from thermoelastic expansion; i.e., a microwave pulse that rapidly (rise time in microseconds) deposits less than 10 y J of energy in the head, which would result in a change in temperature of less than 10 K, can produce an elastic wave that is within the range of acoustic sensitivity of the cochlea. ''' The associated average flux density for a threshold response to a single pulse, if averaged over 1 s, would be near 10 HW/cm2. Work Stoppage When microwave irradiation (900 or 2,450 MHz, CW or sinusoidally modu- lated at 60 Hz) is administered continuously while a hungry albino rat is working at a lever-pressing task for a reward of food, cessation of work under standard environmental conditions typically occurs at a dose rate of 7-10 mW/g. ' Corresponding flux densities would be around 10 mW/cm in a plane-wave 900-MHz field (near the adult rat's resonant frequency) and would approximate 25-50 mW/cm in a 2,450-MHz field. The energy-dose -62-

threshold of work stoppage lies between 8 and 10 J/g, at least for CW and sinusoidally modulated microwaves. Work-stoppage thresholds pulsed radia- tion have not been reported. Aversion Several studies have been performed on mice and rats; ' ' although there is excellent agreement that microwave radiation at high flux densities produces easily recognizable behavioral and physiologic signs (e.g., hyper- activity followed by torpor, urination, defecation, spreading of saliva on pelt, and "spread-eagle" posturing), there is reason to doubt that these animals can learn to escape from CW or sinusoidally modulated fields that range from barely perceptible to lethal in intensity. Escape of sorts does occur in pulsed fields, " but the behavior of the albino rat near a thres- hold of averaged flux density approximating 800 y W/cm is not permanent withdrawal, but a shift of preference in which the animal with a choice spends more time away from the field than in it. But frequent reentry by the animal into the field does occur. Experimental animals appear to find the field "slightly noxious." but will reenter it from time to time. Aversive action has been reported to depend on the fact that thermal receptors in biologic tissue continue to sense heat even after exposure to radiation ceases. Convulsion The grand mal seizure occasioned by radiation-induced hyperthermia may be considered the final behavioral indicant consistent with probable survi- 125 val of the organism. ' For frequencies between 915 and 2.880 MHz, for per- iods of exposure that extend from 0.5 ms to 15 min, ' and for pulsed, sinusoidally modulated and CW fields, both plane-wave and multipath, the energy-dose thresholds of convulsion in the albino rat range from 25 to 28 J/g. The corresponding thresholds of flux density vary by orders of magni- tude, inasmuch as the convulsion, at least for periods shorter than 20 min, is the result of a time-intensity product. As indicated earlier, some of the data from behavioral studies in which fields (average flux density, less than 10 mW/cm ) have been used do indicate effects that are not necessarily construed as deleterious in the absence of correlative data on irreversible functional or structural impairment. One recently reported study that does exemplify the inte- grative approach involved intermittent exposure (7 h/day) of rats to plane- wave 2,450-MHz CW microwaves for 3 months at a flux density of 500 y W/cm . The animals were observed during or after the 3-month period for activity, for acquisition of an avoidance habit, and for sensitivity to electric shock to the feet. Reliable differences were observed in these behavioral end points that correlated well with changes in serum electrolytes (Na , K, carbon dioxide content, and ion gap). The changes observed in irradiated animals subsided within a month after cessation of exposures; this could be taken as evidence of reversible influence by the 500- y W/cm field. -63-

NEUROENDOCRINOLOGIC EFFECTS The hypothalamus can be regarded variously as a collection of homeo- stats, as the "head ganglion" of the autonomic nervous system, and as the primary source of nervous-system input into the endocrines. It is largely in its last-named role that studies of the endocrinologic response to microwave radiation have been performed. The hypothalamus is a continuous sensor of influences especially from thalamic and cortical structures and from fibers of both spinal and extraspinal origin. Blood vessels carry a constant stream of chemical messages from the hypothalamus to the endo- crines about external and internal factors that affect the body's function. The hypothalamus is thus a master regulator. Endocrinologic changes ob- served in the intact organism during or after exposure to microwaves must be understood in this context. The "general adaptation syndrome" of Hans Selye provides a convenient framework within which to understand the neuroendocrinologic responses to stress. Intense, sustained microwave irradiation shares with other sources of prolonged stress the evocation of an alarm reaction, development of resistance, and if intense exposure continues, exhaustion and death, ' presumably from thermal stress. At least three systems involving the hypothalamus with the endocrines are implicated in the response to stress: the hypothalamic-pituitary- adrenocortical (HPC) system, the hypothalamic-pituitary-thyroidal (HPT) system, and the sympathoadrenomedullary (SAM) system. The hypothalamus is also believed to regulate growth hormone. Most of the research performed on the neuroendocrines has involved short-term exposures of the albino rat to moderate- or high-intensity irradiation (1-60 mW/cm , typically for less than 2 h at 2,450-MHz CW in the far field). Stimulation of the HPT system resulting in an increase in thyroxine content occurs at power densities that result in measurable increases in body temperature. Stimulation of the HPC system has been reported after irradiation at 50-60 mW/cm for various periods in which colon temperatures are increased by 1-3°C, but the response is "equivocal" at power densities between 30 and 40 mW/cm . The primary indicant of HFC-system activation is increased serum titer of corticosterone (CS). More recent studies have confirmed that CS titer is strongly and positively related to increased body temperature, which in turn is strongly conditioned by intensity (20-40 mW/cm2) of the incident field, at least during relatively short exposures (less than 2 h). '* When, however, rats were exposed once for 8 h to a field at 20 mW/cm , serum CS content was decreased. Because the CS content is an index of the severity of stress, as well as of the progression of the general adaptation syndrome, one might interpret these data as evidence that irradiation at 20 mW/cm does provoke the alarm reaction, but that the stage of resistance develops quickly. Alternatively, the decline in CS titer below control values might be evidence of incipient entry into the stage of exhaustion. That the latter interpretation is unlikely is inferred from other data on rats that were exposed for 10 h/day for 3 weeks to 918-MHz multipath radiation that approximated 10 mW/cm . Basal serum CS content did not differ between irradiated and sham-irradiated rats at the conclusion of irradiation treatments. -64-

1 09 In other studies of rats exposed once to 2,450-MHz plane waves at 36 mW/cm for 90-150 min, growth hormone (GH) was decreased. Whether GH decrease is an acute response or a prolonged sequela of microwave expo- sure is an unanswered question, but one of considerable interest, in the light of conflicting reports that decreased or even increased body mass is a consequence of acute or chronic in utero exposure to microwaves. The lowest power density at which endocrinologic changes have been reported in the eastern literature is 1 mW/cm ; serum thyroxine was transiently increased in irradiated, but not in sham-irradiated, rats after a 4-h treatment with 2,450-MHz microwaves.l01 In none of the studies cited were irreversible changes observed, but the dearth of endocrinologic assays of animals that undergo truly long- term exposures to microwaves or other RF radiation precludes assessment of chronic effects of fields of low power density (less than 1 mW/cm ) or moderate power density (less than 10 mW/cm ). IMMUNOLOGIC EFFECTS Immunologic and hematologic responses to exposure to microwaves have recently been reviewed. This section summarizes that review. Extensive research on the effects of microwave radiation on the mammalian hematopoietic system and its constituents has been carried out, particularly by Eastern European investigators. Kicovskaja (1964, cited in Baranski and Czerski12tp' 137' irradiated rats for 1 h/day for 216 days with 10-cm microwaves at power densities of 10, 40, and 100 mW/cm . Evidence of lymphocytosis, of lymphopenia, and of a slight decrease in the number of red blood cells was found at gower densities equal to or greater than 40 mW/cm . Deichmann et_ al^., in contrast, reported evidence of neutrophilia, lymphopenia, and increased numbers of red blood cells in rats after single and repeated exposures (for 10 min to 7.5 h)) to 1.25-cm microwaves at power densities of 10-24 mW/cm . Baranski exposed large numbers of guinea pigs and rabbits to 10-cm radiation at power densities of 3.5-7 mW/cm for 3 h/day. Overall durations of the intermittent exposures ranged from 2 weeks to 4 months, and both continuous and pulsed waves were used. After 2-3 months, the exposures resulted in leukocytosis due entirely to increased numbers of lymphocytes; there was no effect on the number of granulocytes. Bone- marrow examination revealed no significant changes in the myeloid:erythroid ratios, but there was a marked decrease in both pronormoblastic and baso- philic normoblastic populations of cells and a shift to more mature forms. The mitotic index of the erythrocyte series of cells, determined after administration of colchicine, was severely decreased in microwave-exposed animals. Pronounced mitotic changes were also noted in the nuclei of normoblasts, but no such effects were seen in precursors of granulocytes. Examination of lymph-node and spleen-impress ion smears revealed a marked -65-

increase in lymphoblasts and in reticular cells, and abnormalities of nuclear structure were observed that were similar to those observed in normoblasts. After termination of the irradiation, the blood character- istics gradually returned to normal. The data in Baranski's report appear to indicate a stimulation of the lymphoid system, the response being much more pronounced to pulsed than to CW radiation. Miro et^ al^. exposed mice for 150 h to pulsed, 10-cm micro- waves at nearly 20 mW/cm ; the authors reported an apparent stimulatory effect of irradiation on the reticulohistiocytic system. They made their determinations by histologic analyses and by observing uptake of [ Sjmethionine into proteins of liver, spleen, and thyraus. Czerski et^ al. exposed guinea pigs for 4 h/day for 14 days to 2,950-MHz pulsed microwaves at a power density of 1 mW/cm . These daily exposures were either at 8 a.m. or at 8 p.m., to permit study of possible alterations in the circadian rhythm of bone-marrow mitoses, determined after the arrest of mitosis by colchicine. No effects were seen on precursors of granulocytes, and only minimal effects on the erythrocyte series, but pronounced phase shifts were noted in the pool of stem cells. Included in the latter category were early normo- blasts, myeloblasts, lymphoblasts, and other unidentified blast cells. The study was then extended; a large group of inbred Swiss albino mice were exposed once for 4 h at 0.5 mW/cm to pulsed 2,950-MHz microwaves— a frequency at which the mouse exhibits resonant absorption of micro- waves. Another group of mice were used to determine body temperature immediately after comparable exposures. No significant increases in body temperature were found during the 4-h exposure. The results of the study on mice were similar to those of the experiment on guinea pigs. The diurnal rate of proliferation of the stem-cell population of exposed animals was amplified, and the phase shifted from that of controls. In another study, the investigators exposed rabbits daily for 2 h (for a total of 74 or 158 h) to 2,950-MHz pulsed or continuous waves at 3 mW/cm and found an impairment of red-cell production, de- termined by ferrokinetic studies; the effects of pulsed waves were more pronounced than those of CW radiation. The finding of shifts in the circadian rhythm of the blood-forming system at power densities near 1 mW/cm is a physiologic indicant of responsiveness to relatively weak microwave fields. A similar phase shift in the circadian rhythm of body temperature was observed by Lu et^ _al. 01 in rats exposed at 1 mW/cm to 2,450-MHz radiation for 1-8 h. The implications of such field-induced shifts, which represent perturbations of the biologic clock, are not clear, but are reminiscent of the work of Wever, who argued that man-made electromagnetic radiation can interfere with "natural" fields of solar and terrestrial origin that he postulated to be regulators of circadian biologic rhythms. A number of recent studies have been focused directly on immunologic effects of microwave irradiation. Paradoxically, enhanced immunologic -66-

creased phagocytic activity, increased population of complement- receptor—bearing lymphocytes, and augmentation of antiviral responses. In contrast, Shandala and colleagues reported that a month of daily 7 h exposures of rats to microwave radiation at much lower power densi- ties, near 500yW/cm , resulted in impaired immunologic competence and induction of autoimmune disease. GENETIC, TERATOLOGIC, AND DEVELOPMENTAL EFFECTS The genetic effects of exposure to microwave and RF radiation have been investigated at several frequencies and over a range of field intensities in both animal and plant systems. Mutagenic effects and effects on growth and development have been reported to be induced in experimental animals exposed to microwave field intensities of 5 mW/cm or greater. ' Exposure to low-intensity microwave fields (i.e., less than 1 mW/cm ) has not been shown to result in genetic or developmental alterations in biologic systems. In the few epidemic- logic studies that have been performed, either there has been no re- ported association between exposure and morbidity or, in the case of one study of the incidence of Down's syndrome in the progeny of men who were potentially exposed to microwave radiation, a positive association reported in an initial study was not corroborated in a later study. Microwave irradiation at high (thermal) intensities has been shown to exert a teratogenic effect in insects that differs quali- tatively from effects due to nonmicrowave heating. It has been suggested that the extent of teratologic damage depends more on the total exposure dose than on the microwave field intensity and that such effects are due to field interactions at the microstructural level. Although the teratogenic effects of microwave exposure do not appear to depend solely on the average temperature increase in the test systems, such effects have not been demonstrated to result from low-field-intensity exposures. OCULAR EFFECTS Of the many purported effects of microwave exposure, cataract induction is the only irreversible alteration reported to have occurred in humans as a result of accidental overexposure. In general, accurate reconstruction of the exposure conditions that have resulted in cataract induction has not been possible. Thus, there is some uncertainty in the interpretation of the ocular effects of microwave exposure. But there is no indication that low-field-intensity microwave exposure results in cataract induction or other ocular pathology. This statement is con- sistent with the results of animal experimentation, which have yielded -67-

intensity-duration thresholds for cataract induction for acute exposures. Acute exposure of rabbits to field intensities of greater than 100 mW/cm for duration of greater than 1 h has been shown to result in cataract induction. Repeated exposures at field intensities that were below the threshold for cataract induction for single exposures have been reported to result in cataracts, but again the exposures were at about 100 mW/cm . Although there are uncertainties regarding the detailed mechanism of micro- wave cataract induction, as well as the dependence of this effect on radia- tion frequency and other field characteristics, it is generally assumed that microwave cataracts are a result of thermal damage to lens tissue and thus that they do not appear to constitute a problem in the case of prolonged exposures to field intensities of less than 10 mW/cm . REPORTED HUMAN EFFECTS Reports of human effects of microwave radiation are derived almost exclusively from surveys of workers who were occupationally involved in the fabrication, use, and repair of microwave equipment. Such surveys may be divided into two groups: broadly based general surveys in which a variety of biologic-physiologic responses in exposed and nonexposed (control) groups are assessed and surveys designed to detect effects— e.g., ocular, cardiovascular, auditory, teratologic, gonadal, and genetic— in specific organ systems. A few individual case histories of human ex- posure are also available. A representative number of such surveys have been reviewed to identify those with possible relevance to the radiation conditions produced by the pulse-modulated signals of the PAVE PAWS radar. Un- fortunately, the identity and operational characteristics of most of the sources of the microwave fields mentioned in these surveys are not specified. Furthermore, the radiation fields produced by the many different microwave sources in given work environments are often generalized into expressions of averaged power density, without reference to the radiation frequency or waveform. The difficulty of applying such generalized information to the specific environmental radiation conditions produced by PAVE PAWS becomes immediately apparent. Although a broad attempt is made here to examine representative literature on reported effects in humans, particular attention is given to reported effects at exposures of approximately 1 mW/cm and below, because the PAVE PAWS may be expected to produce accessible radiation well below this. GENERAL SURVEYS There have been a number of studies of personnel who were routinely associated with the operation of microwave-generating equipment or who were present in areas where microwave-radiation exposure was possible. -68-

As a result of long-term observation of microwave workers who presumably were exposed to a wide variety of CW and pulsed fields, Gordon concluded that prolonged exposure to centimeter-wavelength radiation at power densities of about 0.1-10 mW/cm would produce marked disturbances in cardiac rhythm, such as bradycardia, and persistent hypotonia. Prolonged exposure (years) to centimeter and millimeter wavelengths at power densities of 0.01-0.1 mW/cm reportedly produced the same symptoms as occurred in the higher-exposure group (0.1-10 mW/cm ), but the symptoms were less evident and were reversible. Gordon suggested that chronic exposure to microwaves may lead to the development of a so-called neurocirculatory syndrome. This syndrome reportedly occurs in three stages. The first stage consists of "lability" of cardiac rhythm and blood pressure, both of which are reversible. The second stage is an accentuated form of the first, with EEG changes, thyroid hyperactivity, and an "asthenic state" (headaches, excitability, irritability, fatigue, and pains in the cardiac region). The third stage consists of symptoms that are identical with but more pronounced than those of the first and second stages, with electrocardiographic (ECG) changes. Although Soviet investigators believe that years of exposure to power densities of 0.01-0.1 mW/cm can produce effects in humans, the effects appear to be minimal or reversible. However, a major diffi- culty with these studies is the lack of specification of the modulation characteristics of the microwave sources. The low-intensity pulse- modulated radiation from the PAVE PAWS radar is of primary interest here, and the direct applicability of the Soviet studies to PAVE PAWS is questionable. Personnel engaged in the use, repair, and production of microwave sources were studied by Baranski and Edelwejn. They were divided into low-, moderate-, and high-exposure groups. The designation of "low exposure" usually means average values of "tens of microwatts per square centimeter" in the Polish literature; "moderate exposure" usually means average values of "hundreds of microwatts per square centimeter up to one milliwatt per square centimeter"; and "high exposure" usually means "average values above one milliwatt per square centimeter up to ten milliwatts per square centimeter." Maintenance, repair-shop, and factory personnel were assigned to the high-exposure group. Each of the three groups was subdivided into five categories on the basis of number of years of work associated with microwave sources. Source characteristics (CW, pulse) were not specified. According to the authors, subjective complaints of headaches and sweating were frequent, and personnel with the longest occupational-exposure histories had relatively flat EEC recordings; but no firm conclusions could be drawn, owing to the complexity of environmental and occupational factors and the lack of adequate control groups. An additional difficulty with the study was the lack of definitive environmental-exposure data. The Eastern Europe literature contains many references to what is called a chronic "overexposure syndrome." ' This syndrome is -69-

said to consist of general irritability and complaints of headache, weakness, sleeplessness, decrease in libido, and pains in the chest among those exposed to low-intensity microwave fields. The syndrome is reportedly characterized by periodic recurrences of the same symptoms with intermediate periods of adaptation. However, a more recent appraisal concluded that a microwave-overexposure syndrome remains to be demonstrated. 12 Czerski (in Baranski and Czerski ) described the cases of two long-term radar technicians who were accidentally exposed to microwave power densities of 30-70 mW/cm . These power densities are presumed to represent average values, inasmuch as no peak-power characteristics were mentioned. In the first case, exposure lasted for about 20 min, and in the second, about 5 h. There were no abnormal findings in either subject in followup studies 1 month, 6 months, and 1 yr after the ex- posure. 12 Baranski and Czerski report on a study of several thousand Polish microwave workers who received low, moderate, and high exposures to microwave radiation. The group exposed to low power densities (average, tens of microwatts per square centimeter) was designated the "E" group, the group exposed to moderate power densities (average, hundreds of microwatts per square centimeter up to approximately 1 mW/cm ) was designated the "ES" group, and the group exposed to high power densities (average, 1-10 mW/cm ) was designated the "R" group. No adequate control group was found, because of difficulties in finding persons of the same age working under sufficiently similar environmental, social, and economic conditions. The E group did not have any symptoms after 10 yr of work. Headaches and fatigue disproportionate to effort reportedly occurred in approximately 45% of those in the R group, in 32% in the ES group, and in 30% in the E group during the first year of work. There is no indication as to whether the differences among these percentages are statistically significant. Reportedly, the aforementioned symptoms disappear for 2 yr, recur when an employee has worked for 3-5 yr, and then may reappear in some employees after 5-10 yr of work. Changes in blood pressure reportedly occurred only in the R group. No correlation of heart rate with exposure could be demonstrated; and no EGG changes were found. After 10 yr of work, 10.5% of the R group workers reportedly had absolute lymphocytes is, usually accompanied by monocytosis, the total WBC being over 10,000/mm . After long-term exposure, the ES and R groups reportedly had a decrease in the number and amplitude of alpha waves and an increase in the threshold of stimuation of the senses. The exposures assigned to each group can be interpreted to mean average values pro- duced by a multiplicityof sources, whose individual characteristics (e.g., CW or pulse-modulated) are unidentified. OQ I OQ A study ' of 841 radar shopworkers and installers attempted to determine whether any differences in functional disturbances might exist between a group of 507 persons exposed to mean power densities above -70-

9 9 st 0.2 mW/cm , but not exceeding 6 mW/cm, and a group of 334 persons exposed to mean power densities below 0.2 mW/cm . The pulse-modulation characteristics of the microwave sources were not reported. Functional disturbances were classified as neurotic syndrome (headaches, sleep disturbances, excessive fatigue, emotional instability, difficulties in ability to concentrate or memorize, tremor of hands, sweating, etc.), digestive tract complaints, and cardiocirculatory (including EGG ab- normalities). The difference in exposure between the two groups had no apparent effect on the incidence of functional disturbances, nor did the duration of exposure. However, the incidence of the reported functional disturbances was related to the age of the workers. The findings of these studies do not support the hypothesis of cumulative effects of microwave exposure with increasing duration of exposure, at power densities up to 6 mW/cm . The question of the causal relation- ship between exposure to microwave radiation and the reported functional disturbances should be "left open," according to the authors. A study of some 1,800 employees and more than 3,000 of their dependents who spent time in the U.S. Embassy in Moscow concluded that there was no convincing evidence that microwave radiation had any adverse health effects. Special care was taken to elucidate specific limitations of this epidemiologic study. For example, the identified exposed population in Moscow was too small for the detection of excess risks that were less than twofold for many of the medical conditions studied, except the category "total malignant neoplasms." The health status of the study participants was determined by ques- tionnaire: 59% of those in Moscow study (exposed) group completed questionnaires, and 48% of those in the comparison group, who served at other Eastern European posts. A modest amount of information was available on environmental exposures in various locations in the U.S. Embassy; therefore, a degree of exposure could be assigned to employees who had worked at the embassy earlier if they could recall the exact locations where they had worked. For some unexplained reason, the ratio of the number of deaths due to cancer to the total number of deaths among females employed in the Moscow Embassy (8 cases of 11) during the study period was higher than that in the female comparison group serving in other Eastern European posts (14 of 31). However, no significant difference in total mortality or mortality due to cancer was found in comparisons between the employee group at the Moscow Embassy and those serving at the other Eastern European posts. Some complaints of depression, irritability difficulty in con- centrating, and memory loss were elicited during the study, but the persons in the Moscow Embassy group who registered a greater incidence of such complaints than those in the comparison (control) *The report stated that the power density in the work environment sometimes briefly exceeded 6 mW/cm . -71-

group were the persons who received the lowest exposure to microwaves among the exposed Moscow group. The power densities in the Embassy were reported as approximate maximums that a person could have received by remaining directly in the beam for the entire period of transmission. For the period 1953 through May 1975, the maximal exposure at any point in the Embassy was 5 yW/cm—an intensity that existed for no longer than 9 h/day. From June 1975 to February 7, 1976, the maximum at any point was 15 yW/cm , and the maximal time of beam transmission was 18 h. Beginning on February 7, 1976, the maximum was less than 1 yW/cm , for a maximum of 18 h. Maximal microwave intensities were measured at the outer walls and windows; considerably lower values were found at points away from the windows. In general, the exposure of any given person was much less than the maximum cited (15 yW/cm ). The microwave source was broad-band (0.5-10 GHz), with the highest power density between 2 and 3 GHz. The medical records of a group of U.S. Navy electronics technicians, fire-control technicians, and aircraft electronics technicians—presumably exposed to various amounts of microwave energy during the course of their military assignments—were compared with the medical records of a control group of radiomen, radarmen, and aircraft electrician mates, who pre- sumably received little or no exposure. The assertion that the latter group is a "control" or unexposed group is open to serious question. A small but significantly greater incidence in mortality due to trauma was evident in the exposed group, but there was no significant difference between the two groups in total mortality or mortality from specific diseases—cardiovascular disease, malignant neoplasms, vascular lesions of the central nervous system (strokes), arteriosclerotic heart disease, chronic nephritis, cirrhosis, pneumonia, and leukemia. Major diffi- culties with this study include the absence of data on environmental exposure and on exposure time patterns of personnel and the absence of a valid control (nonexposed) group. One death was alleged to have been due to microwave exposure, ' but the allegation is generally regarded as unproven. Exposure to radiation allegedly occurred from a high-power radar and resulted in necrosis of the stomach and fatal hemorrhage. There is no known proven case of human mortality due to chronic exposure to microwave radiation. OCULAR EFFECTS Cataracts have been produced in experimental animals exposed to microwave radiation. Although there is no fundamental reason to assume that cataracts cannot be similarly produced in humans, the lack of posi- tive proof of microwave induction cataracts in man has led to considerable controversy. -72-

A number of retrospective studies have been undertaken to determine whether a group of persons more likely to have been exposed to micro- wave radiation developed more cataracts or other ocular abnormalities than a control group witn,P9 conceivable exposure to microwaves. In some of these studies, ' a somewhat greater incidence of lenticular imperfections was noted in the microwave workers than in the control groups. Employees in a Swedish radar-equipment facility had a higher incidence of lenticular opacities than the control group. In the same study, 17 of 50 potentially exposed subjects had retinal lesions. Whether microwave radiation can produce retinal lesions is questioned by many investigators,but no definitive study has been conducted. The results of a study of ocular effects among 507 men exposed to microwave at power densities of 0.2-6 mW/cm did not differ from those of a study of 334 men exposed to a mean of less than 0.2 mW/cm . No correlation was demonstrated between the incidence of various grades of lenticular translucence and the duration of exposure to the afore- mentioned radiation, but a clear—cut dependence on age was demonstrated. In a study of possible microwave induction of lenticular changes in U.S. Air Force personnel, Shacklett et^ a\.. found no statistically significant difference in the incidences of opacities, vacuoles, and posterior subcapsular iridescence (PSI) between 447 exposed subjects and 340 control subjects. Similarly, Appleton and colleagues, 3 on the basis of examination of some 1,500 military personnel working with microwave equipment, concluded that there were no differences in lentic- ular opacities, vacuoles, or PSI between microwave workers and unexposed persons of comparable ages. A number of individual case histories of microwave induction of cataract have been reported, '' but in no case, is there reason to suspect that the exposures were not well in excess of 100 mW/cm . One study of possible relevance to PAVE PAWS hints at a lessening of cataractogenic efficiency at the comparatively low frequencies used in the investigation of cataract induction—200, 385, and 468 MHz. In some Soviet investigations of microwave workers (for example, Gordon ), increased incidence of lenticular opacities and sporadic cases of cataract were mentioned, but no firm conclusions were drawn about an association between exposure and cataractogenesis. Further- more, other Soviet investigators (see Petrov ) doubted the causal relationship of lenticular imperfections and microwave exposure. The implication in the soviet work is that lenticular imperfections and cataracts can occur only if exposures are far in excess of the USSR standards, presumably around 10 mW/cm or more. Zydecki examined lenticular transparency in 542 microwave workers whose exposure was supposed to be limited to 0.1-1 mW/cm , but who reportedly received much higher exposures. Baranski and Czerski estimated that the actual exposures probably were in the range of 1-10 mW/cm for 4 h/day, on the average. One conclusion of the Zydecki work was that microwave exposures may cause acceleration of the aging processes of the lens. -73-

Zaret ' reported on more than 50 cases of lenticular changes that he believed characteristic of microwave-radiation exposure. Re- ported preclinical changes consisted of a roughening and thickening of the polar region of the posterior capsule accompanied by the presence of minute areas of opacification and eventual capsular opacification. Other ophthalmologists have been unable to corroborate these diagnostic criteria. Another difficulty with the Zaret reports is the absence of exposure data to compare with the clinical findings. More recently, Zaret and Snyder have reported on nine cases of cataracts induced by hertzian radiation among personnel working in avia- tion environments where they were irradiated at power densities below 1 mW/cm . The nine subjects were three radar technicians, five air-traffic controllers, and one airline pilot. At the time of clinical examination, the ocular lesions had reportedly progressed to capsular cataract, vesicu- lation, and opacification of the proximal subcapsular lenticular substance. According to the authors the slit-lamp examination revealed signs of honeycomb capsulopathy, the earliest clinically recognizable state of "hertzian radiation cataractogenesis." The lack of definitive exposure histories leaves serious doubts that a casual relationship has been established. One Eastern European study of long-term exposure of workers to power densities at or below 1 mW/cm concluded that such exposure does not result in any "reliable deviations of ocular sensitivity." CARDIOVASCULAR EFFECTS Edelwejn et^ a^. concluded in 1974 that no serious cardiovascular disturbances had ever been produced in man or experimental animals by exposure to microwave radiation. Experience after 1974 has tended to support this conclusion, but no definitive work can be claimed. Zaret offered the hypothesis that microwave radiation was responsible for the high incidence of heart disease in North Karelia, Finland, but later reports ' have cited the longstanding concern of public-health authorities over the high-cholesterol diet and obesity of residents as the causal factors. f»N As cited earlier, Gordon claimed that prolonged exposure (for example, to microwave radiation at wavelengths of centimeters and millimeters and at average power densities of 0.1-10 mW/cm can produce marked disturbances in cardiac rhythm (bradycardia) and hypotonia. However, Czerski and Siekierzynski reported that blood pressure of workers routinely exposed to power densities less than 1 mW/cm did not differ significantly from that of unexposed control subjects. 22 Bielski et^ _al. " claimed that exposure to power densities of 1 mW/cm or less may result in "slight but significant cardiovascular alterations" after 10 yr of exposure. However, the exposed workers had a higher resistance to stress. On the basis of available evidence, -74-

the probability is very low that low-intensity microwave radiation has adverse cardiovascular effects on exposed humans. The long-term low- intensity effects reported in some Eastern European publications have no discernible application to exposure conditions associated with the operation of PAVE PAWS. AUDITORY EFFECTS Humans can perceive pulse-modulated electromagnetic energy as sound. The perceived loudness is related to the peak power, not the time-averaged power density. There is strong evidence that the microwave-induced auditory response is due to a mechanical disturbance, rather than a direct effect on the central nervous system. ' The threshold for microwave-induced auditory responses appears to be related to the incident energy per pulse. For humans, the threshold is approximately 40yJ/cm for pulse durations of less than 30 ys. This corresponds to an absorbed energy per unit mass of approxi- mately 16 mJ/kg for the human head. Of particular interest is the fact that the theoretical temperature rise for a single pulse is only 5 x 10 C in the irradiated tissue. Therefore, the temporal character- istics of the pulse and the temporal characteristics of temperature change, albeit minuscule, appear to be the important factors. Given the operating characteristics of PAVE PAWS, it is possible, although unlikely, to experience an auditory response. However, there is no evidence that such an exposure would constitute a hazard to health. GONADAL EFFECTS 93 One report " claimed that 22 of 31 microwave technicians experienced loss of libido and reduced spermatogenesls after an average of 8 yr of exposure to microwave radiation frequencies between 3.6 and 10 GHz at power densities ranging from tens to hundreds of microwatts per square centimeter. Both libido and spermatogenesis returned to normal in most of the technicians 3 months after cessation of microwave exposure. There are no reports of irreversible testicular damage in humans at power densities below 1 mW/cm . The only reported case of injury to the reproductive system caused by the clinical use of microwave radiation involved diathermy treatment of a fallopian tube, with power densities of approximately 100 mW/cm or greater. -75-

TERATOLOGIC EFFECTS A series of 20- to 60-s exposures of gravid mothers to power densities that probably exceeded 100 mW/cm occurred over many years in the clinic of Jose Daels, a Belgian obstetrician. In a personal communication to D. R. Justesen dated January 2, 1978, Dr. Daels reported that more than 10,000 deliveries had been performed with microwave diathermy. Observation of the children had never revealed a harmful effect. OTHER EFFECTS There is no evidence of significant microwave-induced immunelogic, cerebrovascular, or genetic effects in humans, although well-designed studies of sufficiently large groups have never been carried out. DISCUSSION In general, large-scale epidemiologic studies of populations exposed to microwaves have not been conducted. Selected groups of microwave workers (typically less than 1,000 in a group) have been monitored after their initial involvement with microwave sources, but information on the baseline status of their health before employment in a microwave environment (e.g., preplacement medical examination) is often not available, nor are precise data on exposure history (e.g., radiation power densities and exposure time patterns) or information on the extent or anatomic distribution of absorbed energy. It is common to make gross estimates of the average power density in the work area without attempting to characterize the radiation field. The importance of such factors as scattering, radiation frequency, and the presence of a near or reactive field in describing the exposure environment is well known. An overriding difficulty with the literature on human effects is the lack of in- formation on the modulation of pulsed-microwave sources. Indeed, such information may be crucial to a determination of potential biologic effects associated with exposure to radiation at 1 mW/cm or less (PAVE PAWS). The fact that the available exposure data are almost exclusively in terms of average (rms) power density reflects earlier times, when the biologic importance of modulated signals was relatively unappreciated. Therefore, it may be con- cluded that most of the available information on human effects has questionable applicability to PAVE PAWS. The Eastern European literature is replete with references to effects on the central nervous system (CNS) caused by low-intensity microwave exposure. However, most of the publications contain rela- tively few measurements of CNS functions themselves; rather, effects (usually behavioral) are reported and are inferentially regarded as direct effects on the nervous system, e.g., neurasthenic syndrome. -76-

There is an obvious need to determine more precisely the degree of involvement of the CNS when people are subjected to low-intensity microwave radiation. 62 The Eastern European literature (e.g., Gordon ) consistently points out the difficulties of separating the reported biologic effects attributable to microwave exposure from the biologic effects that may result from other factors in the work environment, such as noise, ambient temperature, humidity, illumination, frequent separation from one's family, or isolation from large population centers when servicing or operating microwave equipment in remote areas. Distinguishing the effects of microwave radiation from those of other factors is more difficult when exposure to low power densities (1-5 mW/cm or less) is involved. Even in the case of radar workers who are exposed to little or no microwave radiation, there are complaints about eye fatigue headaches, bradycardia, hypotonia, and general fatigue. Gordon believed that such symptoms may be attributable to illumination that is less than optimal or to the need to stare at video terminals (cathode-ray-tube displays) for long periods. The difficulty of distinguishing signs and symptoms attributable to microwave exposure from those due to alcohol consumption, smoking, overwork, or obesity must also be noted. Although most investigators have attempted to control experimental variables so that discernible differences in individual responses could be attributed exclusively to microwave radiation, there is still appreci- able uncertainty in attempts to control such factors as motivation, personality characteristics, tolerance to assigned work tasks, personal problems, and anxiety. All these could have a major effect on the inci- dence of headaches, irritabililty, and restlessness in study participants. The Johns Hopkins finding of complaints of headache and irritability in U.S. Embassy employees and dependents were remarkably similar to, if not identical with, the neurasthenic syndrome reported by Eastern European investigators, but the Embassy people were only occasionally exposed to microwave radiation and the exposures were in general appreciably below 5 yW/cm . Questions therefore remain as to the suitabililty of present methods for detecting subtle changes in groups of people exposed (and unexposed) to microwave radiation of low power density. An associated difficulty is related to the failure of almost all the published epi- demiologic studies to specify the analytic limitations of their techniques. There is substantial evidence that cataracts can have a thermal origin. A threshold effect has been demonstrated: irradiation of the eyes of animals at 100 mW/cm or greater is required for cataract induc- tion. Although lenticular changes have been discovered in groups of microwave workers, a cause-and-effect relationship between the reported defects and microwave radiation has not been conclusively demonstrated. Considering the radiation frequency and expected power densities associ- ated with PAVE PAWS, the possibility of induction of cataracts in exposed members of the public is very small. -77-

The effects of long-term exposure to microwave radiation at low power densities (e.g., less than 1 mW/cm ) have not been adequately assessed. There is no evidence of a cumulative effect on humans, but the question is unresolved. Taking the literature on reported human effects as a whole and considering the nature and weight of available evidence, as well as the limitations of the investigative techniques used, the Panel believes that the probability of adverse biologic effects on persons exposed to radiation from PAVE PAWS is very low under normal operating conditions and that such effects would be expected to be subtle, transient, and reversible. Additional research is recommended to clarify further the possible effects of long-term exposure to microwave radiation at low power densities. -78-