BIOLOGICAL EFFECTS OF RADIOFREQUENCY RADIATION

United States Environmental Protection Agency Health Effects Research Laboratory Research Triangle Park NC 27711 epa- 600883026B June 1983 External Review Draft Research and Development Biological Effects of Radiofrequency Radiation Part 2 of 3 Review Draft (Do Not Cite or Quote) ------- EPA-600/8-83-026A June 1983 External Review Draft Biological Effects of Radiofrequency Radiation Edited By Daniel F. Cahill and Joe A. Elder Health Effects Research Laboratory Research Triangle Park, North Carolina 27711 NOTICE This document is a preliminary draft. It has not been formally released by EPA and should not at this stage be construed to represent Agency policy. It is being circulated for comment on its technical accuracy and policy implications. HEALTH EFFECTS RESEARCH LABORATORY OFFICE OF RESEARCH AND DEVELOPMENT U.S. ENVIRONMENTAL PROTECTION AGENCY RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711 ------- DISCLAIMER This report is an external draft for review purposes only and does not constitute Agency policy. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. i i ------- FOREWORD The many benefits of our modern, developing, industrial society are accompanied by certain hazards. Careful assessment of the risk of existing and new man-made environmental hazards is necessary to establish sound regula- tory policy. Environmental regulations enhance the quality of our environment in order to promote the public health and welfare and the productive capacity of our nation's population. The Health Effects Research Laboratory conducts a coordinated environ- mental health research program in toxicology and clinical studies. These studies address problems in air pollution, radiofrequency radiation, environ- mental carcinogenesis and the toxicology of pesticides as well as other chemical pollutants. The Laboratory participates in the development and revision of air quality criteria documents on pollutants for which national ambient air quality standards exist or are proposed, provides the data for registration of new pesticides or proposed suspension of those already in use, conducts research on hazardous and toxic materials, and is primarily responsible for providing the health basis for radiofrequency radiation guidelines. Direct support to the regulatory function of the Agency is provided in the form of expert testimony and preparation of affidavits as well as expert advice to the Admi ni strator. The intent of this document is to provide a comprehensive review of the scientific literature on the biological effects of radiofrequency radiation. The purpose of this effort is to evaluate critically the current state of knowledge for its pertinence and applicability in developing radiofrequency- radiation exposure guidelines for the general public. F. G. Hueter, Ph.D. Director Health Effects Research Laboratory i i i ------- ABSTRACT This document presents a critical and comprehensive review of the avail- able literature on the biological effects of radiofrequency (RF) radiation through 1980. The objective is to determine whether the existing data base can contribute to the formulation of RF-radiation exposure guidance for the general public. The frequency range of concern in this document is 0.5 MHz to 100 GHz, which includes all the significant sources of population exposure to RF radia- tion. Research reports that are judged to be credible according to a set of objective criteria are examined for the relation between the RF energy ab- sorbed and the presence or absence of biological effects. The reported conse- quences of the interaction between RF radiation and biological systems are examined from four perspectives by 1) RF-energy-induced core temperature increases, 2) whole-body-averaged specific absorption rate (SAR), 3) the exogeneous energy burden as a percentage of resting metabolic rate, and 4) actual human experiences documented in epidemiological studies. The existing data base does lead to tentative conclusions about the relation between RF-radiation exposure and biological effects that may serve as unadjusted upper limits for population exposure guidance; however, the un- certainties and unknowns in the current state of knowledge are potentially signi ficant. iv ------- CONTRIBUTORS Ernest N. Albert* Joseph S. Ali John W. All is Ezra Berman Carl F. Blackman+ Daniel F. Cahill Joe A. Elder Michael I. Gage Christopher J. Gordon Doreen Hill William T. Joines James B. Kinn K William P. Kirk5 Charles G. Liddle James R. Rabinowitz Ralph J. Smialowicz Ronald J. Spiegel Claude M. Weil U.S. Environmental Protection Agency Office of Research and Development Office of Health Research Health Effects Research Laboratory Research Triangle Park, NC 27711 Department of Anatomy The George Washington University Medical Center Washington, DC 20037 ^Current Address: Carolina Power and Light Company Harris Energy and Environmental Center Route 1, Box 327 New Hill, NC 27562 ^Office of Health Research U.S. Environmental Protection Agency Washington, DC 20406 ^Current Address: Three Mile Island Field Station U.S. Environmental Protection Agency 100 Brown Street Middletown, PA 17057 v ------- CONTENTS Foreword i i i Abstract iv Figures ix Tables xii Acknowledgment xv 1 Introduction 1-1 (Daniel F. Cahill^ 1.1 Ground Rules and General Assumptions 1-1 1.2 General Approach 1-2 1.3 Specific Approach Used in This Review 1-3 2 Summary and Conclusions 2-1 (Daniel F. Cahill and Joe A. Elder) 3 Physical Principles of Electromagnetic Field Interactions . . . 3-1 3.1 Electromagnetic Field Theory 3-1 (William T. Joines) 3.1.1 Electromagnetic spectrum 3-1 3.1.2 Wave propagation 3-5 3.1.3 Wave modulation 3-11 3.2 RF-Field Interactions with Biological Systems 3-15 (Claude M. Weil and James R. Rabinowitz) 3.2.1 Scattering and absorption of electromagnetic waves 3-15 3.2.2 RF dosimetry definitions 3-31 3.2.3 Analytical and numerical RF electromagnetic interaction models 3-34 3.2.4 Mechanisms of RF interaction with biological systems 3-52 3.3 Experimental Methods 3-67 (Claude M. Weil and Joseph S. Ali) 3.3.1 Exposure methods used in biological experimentation 3-67 3.3.2 Animal holders 3-97 3.3.3 Densitometric instrumentation 3-102 3.4 Dosimetric Methods 3-115 (James B. Kinn) 3.4.1 Whole body dosimetry 3-115 3.4.2 Regional dosimetry 3-119 3.4.3 Unresolved questions 3-122 vi i ------- Page 4 Effect of RF-Radiation Exposure on Body Temperature 4-1 4.1 Thermal Physiology 4-1 (Christopher Gordon) 4.1.1 Temperature regulation 4-1 4.1.2 Ambient temperature vs. RF-radiation exposure . . . 4-3 4.1.3 Mechanisms of heat gain during RF-radiation exposure 4-5 4.1.4 Local thermal responses: Effect of RF-radiation exposure on blood flow 4-11 4.1.5 Whole-body thermal response to RF-radiation exposure and dependence on ambient conditions 4-12 4.1.6 Heat stress and the general adaptation syndrome . . 4-17 4.1.7 Thermal physiology of humans 4-19 4.1.8 Thermoregulatory state and physiological responsiveness to RF radiation 4-24 4.1.9 RF-radiation exposure and behavioral temperature regulation 4-26 4.1.10 Unresolved questions 4-27 4.2 Numerical Modeling of Thermoregulatory Systems in Man and Animals 4-33 (Ronald Spiegel) 4.2.1 Heat-transfer models 4-33 4.2.2 RF-radiation-heat-transfer models 4-37 4.2.3 Numerical results 4-42 4.2.4 Unresolved questions 4-45 5 Biological Effects of RF Radiation 5-1 5.1 Cellular and Subcellular Effects 5-1 (John W. Allis) 5.1.1 Effects on molecular systems 5-3 5.1.2 Effects on subcellular organelles 5-8 5.1.3 Effects on single cells 5-13 5.1.4 Unresolved questions 5-23 5.2 Hematologic and Immunologic Effects (Ralph J. Smialowicz) 5-31 5.2.1 Hematology 5-33 5.2.2 Immunology 5-46 5.2.3 Unresolved questions 5-71 5.3 Reproductive Effects 5-73 (Ezra Berman) 5.3.1 Teratology 5-73 5.3.2 Reproduction efficiency 5-99 5.3.3 Testes 5-103 5.3.4 Unresolved questions 5-112 v i i i ------- Page 5.4 Nervous System 5-117 (Ernest N. Albert) 5.4.1 Morphologic observations 5-119 5.4.2 Blood-brain barrier studies 5-122 5.4.3 Pharmacological effects 5-126 5.4.4 Effects on neurotransmitters 5-129 5.4.5 Unresolved questions 5-131 5.5 Behavior 5-137 (Michael I. Gage) 5.5.1 Introduction 5-137 5.5.2 Summary 5-139 5.5.3 Naturalistic behavior 5-140 5.5.4 Learned behavior 5-145 5.5.5 Interactions with other environmental stimuli . . . 5-159 5.5.6 Unresolved questions 5-163 5.6 Special Senses 5-171 (Joe A. Elder) 5.6.1 Cataractogenic effects 5-171 5.6.2 Unresolved questions 5-184 5.6.3 Auditory effects 5-186 5.6.4 Unresolved questions 5-199 5.6.5 Human cutaneous perception 5-203 5.6.6 Unresolved questions 5-206 5.7 Other Physiological and Biochemical Effects 5-209 (Charles G. Liddle) 5.7.1 Clinical chemistry and metabolism 5-209 5.7.2 Endocrinology 5-219 5.7.3 Growth and development 5-225 5.7.4 Cardiovascular system 5-228 5.7.5 Calcium ion efflux 5-233 5.7.6 Unresolved questions 5-237 5.8 Genetics and Mutagenesis 5-241 (Carl F. Blackman) 5.8.1 Introduction 5-241 5.8.2 Effects on genetic material of cellular and subcellular systems 5-243 5.8.3 Effects on genetic material of higher-order biological systems 5-252 5.8.4 Unresolved questions 5-256 5.9 Life Span and Carcinogenesis 5-267 (William P. Kirk) 5.9.1 Life span 5-267 5.9.2 Carcinogenesis 5-273 5.9.3 Prausnitz and Susskind study 5-276 5.9.4 North Karelia connection 5-279 5.9.5 Moscow Embassy study 5-280 5.9.6 U.S. Navy study 5-281 5.9.7 Unresolved questions 5-282 ix ------- Page 5.10 Human Studies 5-285 (Doreen Hill) 5.10.1 Occupational surveys 5-285 5.10.2 Mortality studies 5-291 5.10.3 Ocular effects 5-297 5.10.4 Reproductive effects 5-303 5.10.5 Unresolved questions 5-307 6 Assessment 6-1 (Daniel F. Cahill and Joe A. Elder) 6.1 Core Temperature 6-2 6.2 Specific Absorption Rate 6-5 6.3 Resting Metabolic Rate 6-13 6.4 Human Studies 6-18 References R-l Glossary G-l x ------- FIGURES Number Page 3-1 Far-field electromagnetic wave at a particular instant in time 3-8 3-2 Power density vs. distance along axis from antenna aperture 3-12 3-3 Interaction of RF radiation with electrical conductors, biological tissue, and electrical insulators 3-16 3-4 Energy distribution in proximity to man at 1 GHz at the chest plane contour presentation 3-19 3-5 Dielectric data for tissues in RF range 0.01 to 10 GHz 3-22 3-6 Illustration of object size vs. wavelength dependence 3-24 3-7 Whole-body average SAR vs. frequency for three polarizations in a prolate spheroidal model of a human 3-26 3-8 Absorption dependence on various ground and multi-path factors 3-30 3-9 ARD distribution in core of 6-cm radius multi-layered sphere at 1650 MHz 3-42 3-10 Curve fitting of SAR data for a prolate spheroidal model of man for three basic orientations 3-44 3-11 Effect of a capacitive gap on average SAR between the man model and the ground plane 3-46 3-12 A realistic block model of man 3-48 3-13 Absorption for man block model standing on ground plane .... 3-50 3-14 The real component of the complex permittivity of muscle as a function of frequency 3-55 3-15 EPA 2450-MHz anechoic chamber facility or 2.45-GHz far-field facility 3-72 xi ------- Number Page 3-16 EPA 2450-MHz anechoic chamber facility: diagram of the microwave exposure facility 3-73 3-17 Diagram of absorber-1 ined horn 3-74 3-18 Diagram of a point-source compact range 3-75 3-19 Miniature anechoic chamber facility 3-77 3-20 Photograph of tapered exposure chamber at EPA facility .... 3-78 3-21 Facility for simultaneous exposure of 10 animals with minimal inter-animal interaction 3-79 3-22 Monopole-over-ground plane irradiation facility 3-81 3-23 Coaxial air-line system for high power exposures of cell cultures 3-86 3-24 Parallel-plate (microstrip) exposure system 3-87 3-25 Block diagram of complete RF near-field synthesizer 3-87 3-26 EPA 100-MHz rectangular strip line or Crawford cell 3-89 3-27 Circularly polarized 915-MHz waveguide facility 3-91 3-28 Photograph of exposure chamber with associated instrumentation or the 970-MHz circularly polarized waveguide facility at EPA . 3-92 3-29 VHF resonant cavity facility 3-95 3-30 Multimodal cavity facility for primate irradiation 3-96 3-31 Water-supply system for exposure chamber 3-102 3-32 Samples of commercially available survey meters for measuring RF electric-field strength 3-106 3-33 Microprocessor-controlled twin-well calorimeter 3-118 4-1 Simple neural model of thermoregulation in a mammal 4-4 4-2 Simple model of temperature gradients in a homeotherm 4-6 4-3 Power densities at 2450 MHz necessary to raise rectal temperature 1 °C in 60 min 4-8 4-4 SAR as a function of body weight 4-10 xi i ------- Number Page 4-5 Idealistic response of a homeotherm's metabolism and body temperature to changes in ambient temperature 4-13 4-6 SAR at 2450 MHz, ambient temperature effects, and EHL in mice 4-16 4-7 Effect of an increasing THI on the lethal dose of RF radiation in mice 4-18 4-8 Effects of increasing skin temperature on sweating in humans 4-21 4-9 Effect of exercise on heat loss, metabolic rate, and tympanic temperature of humans 4-22 4-10 Effects of 5-HT injections on mice 4-25 4-11 Block diagram for one segment of the thermal model 4-40 4-12 Incident power density vs. exposure duration to obtain a hot spot 4-45 5-1 Arrhenius plot of Na+ efflux 5-18 5-2 Summed incidence of abnormal and nonviable chick embryo eggs exposed to radiation 5-78 5-3 Cross-sectional sketch of the human and the rabbit eye 5-174 5-4 Time and power density threshold for cataractogenesis in rabbits 5-177 2 5-5 Distribution of energy absorption rate per mW/cm incident power density in the rabbit's eye and head exposed to 2450-MHz radiation 5-178 2 5-6 Distribution of energy absorption rate per mW/cm incident power density in the rabbit's eye and head exposed to 918-MHz radiation 5-179 6-1 Steady-state temperature vs. SAR and power density 6-6 6-2 Relationship of body mass to the SAR necessary to alter the activity of various physiologic systems 6-17 xi i i ------- TABLES Number Page 3-1 Radiofrequency Bands 3-4 3-2 Proposed System of RF Dosimetric Quantities, Definitions, and Units 3-32 3-3 Range of Resonant Frequencies of Man and Animals Irradiated by Plane Waves in Free Space at 1 mW/cm2 With Long Axis Parallel to Electric Field 3-37 3-4 Dielectric Permittivities for Various Tissues 3-54 3-5 Energy Units for RF Radiation 3-58 4-1 Partitioning of Heat Loss in Humans as a Function of Ambient Temperature 4-20 4-2 Summary of Studies Concerning RF-Radiation Effects on Thermoregulation 4-30 4-3 Summary of Studies Concerning RF-Radiation Effects on Blood Flow 4-32 4-4 Steady-State Temperatures in a Human Body after Exposure to 80- and 200-MHz RF Fields 4-43 5-1 Classification of Cellular and Subcellular Experiments 5-2 5-2 Summary of Studies Concerning RF-Radiation Effects on Molecular Systems 5-5 5-3 Summary of Studies Concerning RF-Radiation Effects on Subcellular Systems 5-9 5-4 Summary of Studies Concerning RF-Radiation Effects on Single Cells 5-14 5-5 Summary of Studies Concerning Hematologic Effects of RF-Radiation Exposure 5-34 5-6 Summary of Studies Concerning Immunological Effects (Li Vivo) of RF-Radiation Exposure 5-50 xiv ------- Number Page 5-7 Summary of Studies Concerning Immunologic Effects (In Vitro) of RF-Radiation Exposure 5-66 5-8 Conversion of J/g to W/kg 5-82 5-9 Summary of Studies Concerning Teratologic Effects of RF-Radiation Exposure 5-100 5-10 Summary of Studies Concerning Reproductive Effects of RF-Radiation Exposure 5-103 5-11 Summary of Studies Concerning Reproductive Effects of RF-Radiation Exposure in the Rat 5-112 5-12 Summary of Studies Concerning RF-Radiation Effects on the Nervous System 5-133 5-13 Summary of Studies Concerning RF-Radiation Effects on Behavior 5-164 5-14 Summary of Studies Concerning Ocular Effects of Near-Field Exposures 5-172 5-15 Summary of Studies Concerning Ocular Effects of Far-Field Exposures 5-173 5-16 Summary of Studies Concerning Auditory Effects of RF Radiation in Humans 5-188 5-17 Summary of Studies Concerning Threshold Values for Auditory- Evoked Potentials in Laboratory Animals 5-201 5-18 Summary of Studies Concerning Human Cutaneous Perception of RF Radiation 5-204 5-19 Summary of Studies Concerning RF-Radiation Effects on Clinical Chemistry and Metabolism, Endocrinology, and Growth and Development 5-210 5-20 Summary of Studies Concerning RF-Radiation Effects on Various Aspects of Cardiac Physiology 5-234 5-21 Summary of Studies Concerning Genetic and Mutagenic Effects of RF-Radiation Exposure 5-263 5-22 Summary of Studies Concerning RF-Radiation Exposure Effects on Life Span/Carcinogenesis 5-268 5-23 Summary of Prausnitz and Susskind Data 5-271 xv ------- Number Page 5-24 Distribution of Years of Exposure for 226 Radar Workers 5-287 5-25 Age Distribution of 226 Microwave Workers and 88 Controls 5-287 5-26 Microwave Exposure Levels at the U.S. Embassy in Moscow 5-292 5-27 Number of Deaths from Disease and Mortality Ratios by Hazard Number 5-296 5-28 Classification by Military Occupation of World War II and Korean War Veterans With and Without Cataracts 5-299 5-29 Estimated Relative Risk of Cataracts Among Army and Air Force Veterans 5-300 5-30 Paternal Radar Exposure Before Conception of Index Child 5-305 5-31 Summary of Selected Human Studies Concerning Effects of RF-Radiaton Exposure 5-311 6-1 Studies Reporting "No Effects" at SAR's < 10 W/kg Grouped by Biological Variable 6-10 6-2 Studies with Reported "Effects" at SAR's < 10 W/kg Grouped by Biological Variables 6-11 6-3 Equivalent Power Density for Five Ages and Mass Sizes at Resonant Frequency for SAR = 0.4 W/kg 6-14 6-4 The Increased RMR and Equivalent SAR and Power Density Predicted to be Associated with the Onset of Human Thermoregulatory Response 6-18 6-5 Summary of Estimates of Unadjusted Limits for RF-Radiation Exposure at Resonant Frequencies 6-21 xv i ------- ACKNOWLEDGMENT The authors wish to thank Jan Parsons and Carole Moussali, Northrop Services, Inc., for editorial review of this document, and Linda Jones and Connie Van Oosten, Northrop Services, Inc., for the word processing. Wanda Jones, Bonnie Waddell, and Barbara Queen are also to be commended for their word processing assistance on earlier versions of this document. xv ii ------- SECTION 5 BIOLOGICAL EFFECTS OF RF RADIATION 5.1 CELLULAR AND SUBCELLULAR EFFECTS John W. All is The investigation of RF-radiation effects on cellular and subcellular systems is principally an attempt to eludicate specific biochemical mechanisms for the interaction of the radiation with macroscopic biological systems. This approach takes advantage of the relative simplicity of in vitro systems, the ability to control variables, the rapid and economical way with which results may be obtained, and the perceived ease of mechanistic interpretation of the results. In practice, however, these advantages are not always realized. The divisions between molecular, subcellular, and cellular systems are somewhat arbitrary, but for convenience, results presented in this section will be discussed in this format. Examples of each classification are presented in Table 5-1. When investigating molecular and subcellular systems, one normally looks at a single key structure or function where, one hopes, all other variables may be held constant. In these cases, thorough understanding 5-1 ------- of the end point under study allows the investigator to interpret results more easily in terms of a detailed biochemical mechanism. For instance, if a change in the kinetics of an enzyme-catalyzed reaction is found, the type of inhibition can be determined, the particular molecules involved can be identified, and a hypothesis for the effect can be constructed. TABLE 5-1. CLASSIFICATION OF CELLULAR AND SUBCELLULAR EXPERIMENTS Classification Type of System Studied Examples of Experimental Measures Molecular Purified enzyme preparations Enzyme activity, binding to small molecules Molecular Purified DNA preparations DNA melting Subcellular organelles Membrane-bound enzymes Enzyme activity Subcellular organelles Cell membranes, phospholipid vesicles Infrared, Raman spectra Subcellular organelles Isolated mitochondria Mitochondrial function Cellular Red blood cells Ion transport Cellular Bacteria Growth, survival, infectivity Cellular Isolated neurons Neuronal firing, membrane potential In some experiments on intact cells, these conditions can also be approached. An example of this is sodium (Na+) and potassium (K+) transport across the red-blood-cell (RBC) membrane. The RBC's limited metabolic activity is geared 5-2 ------- primarily to performing its main function of transporting oxygen from lungs to tissue. In this case, ion transport across the membrane is linked to the mode of energy production in the cell and very little else, and studies on this system are relatively simple to interpret. Experiments on most living cells are complicated by complex biochemical systems, many of which have multiple or alternate pathways. Rather than isolated functional changes, broad end points such as growth or genetic changes are often assayed. Here, the investigator tries to use "biological amplification" where, for example, the cell's growth is highly sensitive to disruptions of a critical but unspecified metabolic step, or where a mutant appears that must reproduce over several generations before sufficient numbers are present to be measurable. The investigator also recognizes that the redundancies in a complex system like a living cell may negate any efforts to detect the putative effect. This inherent complexity makes it difficult to ascribe a detailed biochem- ical mechanism to effects found. Such effects must be viewed as a starting point from which a mechanism may eventually be deduced. On the other hand, effects on cells can also be used to rationalize or interpret—perhaps even pre- dict—effects in the intact animal. 5.1.1 Effects on Molecular Systems To obtain insights into possible direct interactive mechanism(s) of micro- wave radiation at the molecular level, several investigators have attempted 5-3 ------- to measure changes in biologically important macromolecules exposed i_n vitro (Table 5-2). Enzyme kinetics studies form a majority of the reports, but a few examine changes in macromolecular structure. In one of the latter, Hamrick (1973) measured DNA melting curves after exposure of calf thymus DNA to 2.45-GHz (CW) radiation for 16 h (SAR = 67 W/kg) and at dose rates to 160 W/kg for 1 h. Temperature was controlled during all exposures, usually at 37 °C, but for some experiments it was maintained at 40, 45, and 50 °C. The intent of the study was to determine if microwaves disrupted hydrogen bonding between the DNA strands of the double helix, thereby affecting the melting curve. The assay was carried out after exposure. In one experiment, any disruption of hydrogen bonding between the strands was preserved by performing the microwave exposure with formaldehyde in the buffer. However, all melting curves were virtually identical to those of unexposed, temperature-matched controls. In another study of macromolecular structure (Allis 1975), the protein bovine serum albumin (BSA) was exposed to 1.70- and 2.45-GHz (CW) radiation (SAR's ranging from 30 to 100 W/kg). The author recognized that structural changes due to microwave exposure may be reversible. In this study, the problem was attacked by developing an exposure apparatus in which ultraviolet (UV) and visible spectrophotometry measurements could be performed during ex- posure. The difference in UV absorption between the exposed and unexposed samples was measured directly by using a double-beam spectrophotometer. Differences of this type reflect small changes in the surroundings of certain UV-absorbing amino acids and, in this case, could be interpreted as changes in the structure of the protein. The temperature of the exposed samples was 5-4 ------- TABLE 5-2. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON MOLECULAR SYSTEMS Exposure Conditions End Point Measured/ Effects Experimental System Exposure Frequency Duration Facility SAR (GHz) (nin) (Type) (WAg) Reference No change in UV difference spectra measured over pH range 2 5-5.5 UV spectra and binding constants for nononucleotides showed no difference from controls No change in enzyme activity No difference in melting curves Inactivation of enzyme, probably temperature inhooogeneity effect at very high doses Heat inactivation of enzymes found at highest SAR (T = 50 CC), corresponded closely to heat- treated controls Heat inactivation of enzyme found at SAR's > 165 WAg BSA Ribonuclease Glucose-6-phosphate dehydrogenase; adenylate kinase, NADPH cytochrome C reductase DMA 1 70 (CW) 2 45 (CW) 1 70 (CW) 2 45 (CW) 2 45 (CW) 2 45 (CW) Horseradish peroxidase 2 45 (CW) Glucose-6-phosphate 2 8 (PW) dehydrogenase, lactate dehydrogenase; acid phosphatase, alkaline phosphatase Lactate dehydrogenase 3 0 (CW) 30 30 60, 960 5, 10, 20, 30, 40 4 5, 18 5 20 Waveguide 30-100 Waveguide 39 Waveguide 42 Allis (1975) Allis et al. (1976) Ward et al (1975) Far field 67, 160 Hamrick (1973) Waveguide 62,500- Henderson et al^ (1975) 375,000 Waveguide 200-500 Belkhode et al (1974a,b) Waveguide 33-960 Bini et al (1978) BSA = bovine serum albumin ------- controlled during exposure, and it was matched by the temperature of the con- trol sample. Temperatures ranged from 24 to 32 °C depending on the SAR value. Spectra were measured immediately upon beginning exposure, and again 30 min later with continuous exposure. The study results showed that no changes in the UV spectrum could be found over a variety of structural states of the protein, implying that structural changes due to microwave exposure could not be inferred. The ability of microwave radiation to alter enzyme activity has been studied by several workers. Measurements were performed during microwave exposure by two groups, each using spectrophotometry measures of enzyme activity. Ward et al_. (1975) examined three enzymes (glucose-6-phosphate dehydrogenase, adenylate kinase, and NADPH-cytochrome reductase) exposed to 2.45-GHz (CW) radiation (SAR = 42 W/kg). All exposed and control samples were maintained at 25 °C. Exposure durations were ~ 5 min, during which the enzyme activity was measured. No differences between exposed and control samples were found. Bini et aL (1978) followed the activity of lactate dehydrogenase exposed to 3.0-GHz (CW) radiation (SAR's between 33 and 960 W/kg). They demonstrated that the changes found in the enzyme activity were entirely consistent with calculations of thermal inactivation of the enzyme at the temperatures attained. The sample exposed at 33 W/kg was not different from the unexposed control; all other exposures (SAR's ranging from 165 to 960 W/kg) showed evidence of enzyme inactivation. Belkhode et al_. (1974a,b) reported the effect of 2.8-GHz square-wave modulated (1-kHz) radiation on four enzymes: glucose-6-phosphate dehydrogenase, 5-6 ------- lactate dehydrogenase, acid phosphatase, and alkaline phosphatase. Enzyme preparations were analyzed after exposures (SAR's ~ 200 to 500 W/kg on average). Exposures were conducted at 37, 46.7, and 49.7 °C; enzyme activities were compared to the activity of sham-exposed samples at the same temperatures. Activities of the exposed enzymes at each temperature were indistinguishable from the shams. Henderson et aK (1975) reported a change in enzyme activity that was interpreted as an indicator of direct interaction on the enzyme by microwaves. In this experiment, horseradish peroxidase was subjected to 2.45-GHz (CW) radiation (SAR's between 62,500 and 375,000 W/kg) with the sample exposed in a tube (4.7-mm ID) that protruded through a waveguide. The sample tube was surrounded by a concentric cooling jacket, through which an organic coolant was pumped continuously to maintain the temperature at 25 °C. Thermocouples were placed in the sample tube so that the sample temperature could be monitored from positions just outside the waveguide. The total volume of the exposed sample was ~ 0.8 ml. A marked decrease in enzyme activity was found at 62,500 W/kg after 30 min of exposure, and at 187,500 W/kg after 20 min of exposure, even though the temperature was reported never to exceed 35 °C. It is possible that the very high fields present at these SAR's could produce field-specific effects. However, it appears likely that very high local heating occurred in the sample that was responsible for enzyme inacti- vation. This likelihood is substantiated by the work of Harrison et aK (1980) who performed liquid-crystal thermography under similar exposure conditions. Temperature rises of as much as 0.3 °C were recorded within a micropipette 5-7 ------- suspended in a waveguide and cooled with water circulating through the wave- guide. Henderson et aH. (1975) exposed their samples at 5 to 10 times the levels used by Harrison et a_L ; also, the latter researchers used water as a coolant, which would attenuate the energy reaching the sample much more strongly than would the organic solvent used by Henderson et aK A study by Livingston et art. (1979) graphically illustrates the effect of temperature gradients during microwave exposure. The binding of small substrate-1ike molecules to the enzyme ribonuclease was studied to determine if the binding relationship between an enzyme and substrate could be affected by microwaves (Allis et aH. 1976). In this work, UV-absorption spectra were measured during exposure to 1.70- and 2.45-GHz (CW) radiation (SAR = 39 W/kg). Measurements were performed immediately upon beginning irradiation and after 30 min of exposure. Neither structural changes in the enzyme-substrate complex nor changes in the binding constants were found. In sum, no consistent effects on molecular systems exposed i_n vitro to RF radiation have yet been demonstrated. 5.1.2 Effects on Subcellular Organelles There has been relatively little research on the effects of microwave radiation on subcellular organelles (Table 5-3). The reports included in this document range from work with phospholipid bilayers (i.e., synthetic analogs of cell membranes) to experiments using intact mitochondria, the energy-producing system in eukaryotic (e.g., mammalian) cells. Most of the work to be discussed 5-8 ------- TABLE 5-3. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON SUBCELLULAR SYSTEMS Exposure Conditions End Point Measured/ Effects Experimental System Exposure Frequency Duration Facility SAR (GHz) (oin) (Type) (W/kg) Reference Increase in exchange of strongly bound aside hydrogens in membrane protein measured ty IR spectra for SAR > 10 W/kg, No change in o-helix or p-sheet content of proteins No change in activity of membrane bound enzymes measured spectrophotometries!ly No change in activity of meobrane bound enzyae measured spectro- photometrically No difference in respiratory activity No difference in respiratory activity No change in formation of microtubules No change in migration of proteins within axonal membrane No changes in IR spectra of proteins and nucleic acids in E coli exposed before drying r RBC membrane RBC membrane, mito- chondrial inner membrane 1 0 (CW) 30 2 45 (SW) 10 Endoplasmic reticulua 2 45 (CW) Mitochondria 2 45 (CW) 30 to 210 2-4 (Swept) 10 3.4 (CW) Mitochondria Tubulin (rabbit brain) 3.1 (PW) 15 Vagus nerve cell 3 1 (PW) 24 h Dried film of E coll cells 3 2 (CW) 8, 10, 11 h Stripline 5-45 Waveguide 26 Waveguide 42 Anechoic chamber far field Coaxial Airline 17 5, 87 5 Isoailov (1977) AllTS et al (1979) Ward et al (1975) Elder and All (1975) 16-2.3 Elder et al (1976) 41 Far field 112-430 Paulsson et al^ (1977) Far field ~ 10-100 Paulsson et al (1977) Waveguide 20 Corelli et al (1977) SW = Sine-wave modulated tRBC = Red blood cell #IR = Infrared ------- here has not demonstrated effects of microwave exposure at dose rates ranging from 1 to 430 W/kg. Two reports that have indicated an effect do not meet all criteria for inclusion here and are, therefore, discussed with other reports that present unresolved questions. Two reports describe work with enzymes bound to biological membranes. The study by Ward et (1975), discussed above, focused on the enzyme NADPH- cytochrome C reductase, which is loosely bound to the membrane of the endoplasmic reticulum of rat liver cells. All is et al^ (1979) studied adenosine tri- phosphatase (ATPase) in RBC membranes and cytochrome oxidase in the inner mitochondrial membrane of rat liver cells. The latter two enzymes are thought to be integral parts of membranes. Conditions of the exposure were identical for all three enzymes in that the assay was performed spectrophotometrically during exposure to microwaves. In the latter study the dose rate was 26 W/kg, and the 2.45-GHz radiation was sinusoidally modulated at 16, 30, 90, and 120 Hz. Enzyme activity was not measurably affected by these exposures; a 10- to 15-percent change in enzyme activity would have been required to detect a reliable microwave effect. Ismailov (1977) investigated the infrared (IR) absorption spectra of pro- teins in RBC membranes exposed to 1.009-GHz fields (SAR's up to 45 W/kg) and maintained at 25 °C. The samples were exposed for 30 min in aqueous suspension in a stripline, then were dried to a thin film to obtain the IR spectra. No change in or helix or p-sheet content of the membrane proteins was noted. However, when D2O (heavy water) was added to the suspension before beginning 5-10 ------- exposure, application of microwaves was found to increase the degree to which strongly bound amide hydrogens were exchanged. This effect was pronounced at SAR = 45 W/kg but disappeared when the SAR was below 10 W/kg. The increase in accessibility of the poorly exchangeable amide hydrogens indicates that the microwaves disturb the relationship of the protein to its neighboring membrane lipid. A quantitative estimate of the sensitivity of this experiment cannot be made from the spectra shown in the paper. IR spectra of Escherichia coli after exposure to microwaves were also mea- sured (Corelli et al_. 1977). After 12 h of exposure to 3.2-GHz (CW) radiation (SAR = 20 W/kg), E. coli were dried to a film, and spectra were measured in the protein and nucleic acid absorption regions. No differences were found. These results are comparable with those of Ismailov (1977) that indicated no changes, but they do not address the conditions for which Ismailov did report effects. The functional properties of the microtubule assembly system extracted from rabbit brain cells were studied after exposure to 3.1-GHz fields by Paulsson et aL (1977). The binding of a drug, colchicine, to the microtubule precursor protein tubulin was measured after exposure to PW microwaves for 15 min at average dose rates of 112 and 243 W/kg (pulse-repetition frequency of 200 Hz, pulse duration of 1.4 [is). Colchicine normally blocks the formation of microtubules, which halts cell division. The normal assembly of microtubules from tubulin exposed for 10 min to PW microwaves, as above, at 430 W/kg was also studied. No noticeable effect on either process was found. The data indicate that a change of about 15 percent in the colchicine binding and about 5-11 ------- 10 percent in the microtubule assembly measurement would have been noted. Paulsson et a^. also studied the migration of proteins within the axonal mem- brane of the rabbit's vagus nerve. In this case the samples were exposed for 24 h (the SAR estimated from their data was 10 to 100 W/kg) at a pulse-repetition rate (PRR) of 100 Hz and a pulse duration of 1.4 ps. The distribution of tritium-labeled protein in the axonal membrane was found to be the same in exposed and control samples. A difference of > 20 percent would have been required for detection in this experiment. Two papers (Elder and Ali 1975; Elder et aK 1976) present results of exposure of rat liver mitochondria to microwave radiation. Both papers examine oxygen utilization by measuring respiratory activity (e.g., respiratory control ratio, ADP to oxygen ratio) under various conditions. These parameters are functional indicators of the energy production system in eukaryotic cells. The earlier paper tested mitochondria kept at 0 °C, or inactive state, during exposure in the far field at 2.45 GHz. The mitochondrial functions were examined after exposure in the active state at 25 °C. (Exposures for periods up to 3.5 h were conducted at SAR's of 17.5 and 87.5 W/kg.) No changes in mitochondrial activity were seen. In the later paper, the disadvantage of ex- posing inactive mitochondria was overcome by use of a novel flow-through system that coupled a coaxial airline and an oxygen electrode. The mitochondrial suspension was cycled continuously between the airline, where exposure was accomplished, and the oxygen electrode. Samples were exposed to 2.45-, 3.0-, and 3.4-GHz (CW) radiation (SAR = 41 W/kg) and also to radiation at swept frequencies between 2 and 4 GHz (SAR's from 1.6 to 2.3 W/kg). As in the earlier work, no effects of microwave exposure were detected under any 5-12 ------- condition. In general, a five-percent change would have been sufficient for detection. Other workers have presented data on microwave exposure of mito- chondria, either in a form too incomplete for inclusion or in oral presenta- tions. Their findings do not differ from those described here. 5.1.3 Effects on Single Cells Effects on single cells have been investigated by several researchers (Table 5-4). Transport and related properties of RBC membranes have been studied by three groups. In each case K+ transport was used as an end point. In the RBC, active (or energy requiring) Na+ and K+ transport across the membrane is by the enzyme Na+-K+ ATPase (the same enzyme discussed in § 5.1.2, Effects on Subcellular Organelles), and passive transport is through channels in the membrane. Ismailov (1971) exposed human RBC's to 1.0-GHz (CW) micro- waves at 45 W/kg and found an increased efflux of K+ and a concomitant increased influx of Na+. The Na+ influx was twice as large, ion per ion, as the K+ efflux. Exposures were carried out in a coaxial stripline for 30 min, with analysis of the ion content of the supernatant performed afterwards. These results indicate either a reversal of the normal action of the Na+-K+ ATPase, or an inhibition of the enzyme, which permitted ion leakage across the membrane to change the Na+/K+ ratio. Hamrick and Zinkl (1975) and Peterson et al. (1979) have performed similar experiments by exposing RBC's at 2.45 GHz in the far field (SAR's were 3 to 57 W/kg, and ~ 200 W/kg, respectively). Neither study found a difference between the K+ efflux from microwave-exposed RBC's and conventionally heated RBC's with similar histories of temperature elevation, although Peterson et aK did find a difference between unexposed 5-13 ------- TABLE 5-4. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON SINGLE CELLS End Point Measured/ Effects Experimental Systen Exposure Conditions Frequency Duration (GHz) (Bin) Exposure Facility SAR (Type) (W/kg) Reference tn l Increase In RBC electrophoretlc nobility 30«1n post-exposure (SAR > 10 W/kg) Increase In K+ efflux and Na+ influx K+ transport no different from heat-treated controls; no change in osnotlc fragility K transport no different from controls at corresponding temperatures; no difference in heooglobln release Passive transport of Na+, Rb Increased at transition temperature Rapid response in change of firing rate of pacemaker neurons which does not correlate with tempera- ture changes In ninority of trials No change in growth, CFU,* of ex- posed cultures No change in growth, CFU, of vari- ous strains of exposed cultures under several growth conditions Ho change in survival curves (treasuring CFU) of exposed cultures RBC' RBC RBC RBC RBC Isolated neuron from ftplysla E coll P aeruginosa E col i E coli B subtlUs spores 1 0 (CV) 4, 8. 15, 30 1 0 (CW) 30 2.45 (CW) 60, 120, 180, 240 2 45 (CW) 45 Stripline 5-45 2 45 (CW) 60 1 5, 2 45 3 (CW and PW) 2 45 (CW) 720 2 45 (CW) 240 2 45 1 Stripline Monopole far field Anechofc chamber far field Waveguide Stripline Far field flnecholc chamber far field Microwave oven 45 3-57 200 100, 190 390 1-100 29-320 0 0075- 75 ~ 400 Ismailov (197B) Ismailov (1971) Hamrlck and Zinkl (1975) Peterson et al^ (1979) Olcerst et a^ (1980) Wachtel et al (1975), Seaaan and RIchtel (1978) Haonck and Butler (1973) Blackman et al. (1975) Goldblith and Wang (1967) (continued) ------- TABLE 5-4. (continued) Exposure Conditions End Point Measured/ Effects Experimental System Frequency (GHz) Duration (min) Exposure Facility (Type) SAR (W/kg) Reference Growth rate slowed, morphological changes found Chinese hamster lung cells, V79 2 45 (CW) 20 Waveguide 1059 Chen and Lin (1976) No change in light emission of photoactive bacterium P fischeri 2 6-3 0 (CW) - 22 Waveguide 660 to 5300 Barber (1962) No effect on colony-forming ability E coli 2 6-4 0 (CW) 8 h Waveguide 29 Corelli et al (1977) Temporary decrease in virulence (> 6 h) of bacteria for its host cells; recovery within 24 h at 37 °C A tumefaciens 10 (CW) 30, 60 230 Cavity ~ 1 Hoore et al (1979) CFU = Colony Forming Unit tRBC = Red Blood Cell. ------- rabbit cells maintained at 25 °C compared with those maintained at 37 °C. This difference did not occur with human cells. Ismailov also used temper- ature controls but did not describe the solution parameters for the cell suspensions. Under certain conditions of chemical concentration, it is pos- sible to reverse the Na+-K+ ATPase; however, Ismailov's controls behaved normally, indicating that chemical concentrations in the controls were not unusual. The origin of the discrepancy between these studies is not clear. Hamrick and Zinkl (1975) and Peterson et al_. (1979) measured other end points as well. The former looked at osmotic fragility of the RBC's and concluded that there was no difference. The latter paper gives data on hemo- globin release from RBC's, an indicator of membrane fragility. Again, no differences were found between irradiated and heat-treated RBC's; however, as for K+ efflux, the unexposed rabbit cells released less hemoglobin at 25 °C than at 37 °C. In a separate paper, Ismailov (1978) reported increases in the electro- phoretic mobility of human RBC's exposed under conditions identical to those in his previously discussed study. The electrophoretic mobility was measured at 10-min intervals after cessation of exposure. The mobility was found to peak at 30 min post-exposure and to return to base line ~ 60 min after expos- ure. The peak mobility decreased with shorter exposure durations (30, 15, 8, and 4 min). Also, the mobility change decreased as a function of dose rate, disappearing between 5 and 10 W/kg. Although a change in the counter-ion distribution around the cell and possible conformational changes in the 5-16 ------- membrane proteins were discussed or suggested as possible causes, it remained unclear why these phenomena peaked 30 min after exposure. Passive ion transport in rabbit RBC's was examined by Olcerst et al^- (1980) after exposure to 2.45-GHz (CW) radiation (SAR's at 100, 190, and 390 W/kg). The cells were treated with ouabain to inhibit active transport of Na+ and K+ by Na+-K+ ATPase, the important enzyme discussed previously in several papers. Exposures took place in a waveguide system in which the sample was placed parallel to the E-field in a cylindrical tube. An organic coolant of low dielectric constant was circulated around the sample in a larger con- centric cylinder to maintain the temperature of the sample under exposure. The SAR's were computed from readings of forward and reflected power. The RBC's were pre-incubated with radioactive Na+ or Rb+ (the latter is a K+ sub- stitute). Samples were exposed or heat-treated for 1 h, and the suspending medium was analyzed for radioactivity. Graphs of the logarithm of the efflux vs. inverse temperature were identical except at three transition tempera- tures, where the slope of these plots changes sharply (see Figure 5-1). At these points, the exposed samples exhibited considerably higher efflux than the heated samples; however, no consistent difference between exposure levels could be established. These results imply that the cell membrane structures responsible for passive ion transport are sensitive to microwave exposure at temperatures where transitions between two states are taking place. Living cells depend on closely regulated ion concentrations for many processes, and serious disruption of these balances could be lethal. 5-17 ------- o c xf 3 LU <0 Z LU 39.3 40 -l 29.8 TEMPERATURE, °C 20.9 12.5 4.6 20- 10- o CO CC "c 4H 2- 1- 0.4- 0.2- 0.1 ]— 3.2 I 3.3 —r~ 3.4 ~1~ 3.5 —r~ 3.6 103/T,°K Figure 5-1. Arrhenius plot of Na efflux. Unirradiated temperature con- trols are represented by open symbols; irradiated samples are represented by darkened symbols. Specific absorption rates are expressed as watts per kilogram (• = 100, n = 190, ~ = 390). Control samples had an average standard error of 0.98 percent. Irradiated samples had an average standard error of 4 percent. (Olcerst et al. 1980). 5-18 ------- Many studies have been conducted that focus on a major function of a cellular species as a generalized end point. The reasoning for this approach is that if microwaves disrupt an important metabolic step, the result will be a decline in the ability of the cell to perform its major function. As mentioned earlier, this assumes that no compensating mechanism will operate. A broad spectrum of such end points has been investigated for cells exposed i_n vitro. Some of these are discussed elsewhere: phagocytosis and blastic transformation in lymphocytes are presented in § 5.2, Hematologic and Immunologic Effects; the induction of the antibiotic colicin, along with point mutations in single cells, is discussed' in § 5.8, Genetics and Mutagenesis. Perhaps the ultimate test of a cell's functional ability is growth and survival. Several workers have concentrated on this end point and have inves- tigated the frequency range between 1 and 4 GHz. In general, the results have proved negative. Far-field exposures were conducted (Hamrick and Butler 1973; Blackman et aH. 1975) on several strains or mutants of E. coli and on Pseudomonas aeruginosa. Samples were exposed in T-flasks or petri dishes, principally at 2.45-GHz (CW) radiation. Growth was measured by assays of colony-forming units (CFU). In both experiments the duration of exposure was sufficiently long (12 and 4 h) that the average cell divided at least once. Blackman et al. examined several growth conditions such as lag, log and stationary phases of growth, rich and minimal media, and a "normal" as well as a mutant amino-acid-requiring strain. In each case, no differences between the exposed samples and temperature-matched control samples were found. Hamrick's SAR's O ranged from 29 to 320 W/kg with power densities of 60 to 600 mW/cm , and Blackman's from 0.0075 to 75 W/kg with power densities of 0.005 to 50 mW/cm^. 5-19 ------- Corelli et aK (1977) exposed E. coli at the end of a waveguide to microwaves swept from 2.6 to 4.0 GHz for 8 h at a dose rate of 29 W/kg. No effect of the microwave exposure was found on colony-forming ability of the bacteria. Goldblith and Wang (1967) exposed E. coli and Bacillus subti1is spores in a microwave oven at 2.45 GHz for periods to 1 min (SAR estimated at 400 W/kg). In this case, microwave irradiation and conventional heating were found to have identical effects on survival. Chen and Lin (1978) exposed Chinese hamster lung cells, V79, in a waveguide fitted with a micropipette that contained the cell suspension. Temperature was regulated by circulating cooling water through the waveguide around the micropipette. Samples were exposed to 2450-MHz (CW) radiation for 20 min, and the cells were allowed to grow in cultures for 12 days after exposure. The exposed cells were observed to divide at a slower rate and to exhibit a fibro- blastic type of growth, in contrast with the controls. These cells were 2 exposed at 400 mW/cm (SAR = 1059 W/kg) under conditions similar to those in which Harrison et al^. (1980) found temperature elevations in the sample that were as much as 0.3 °C higher than the cells in the cooling bath. Chen and Lin (1978) state that temperature-treated controls at 38 °C (1 °C higher than the coolant temperature during microwave exposure) did not display the changes observed in the microwave-exposed cells. However, it is not entirely clear whether the changes in the microwave-treated cells were caused by elevated temperatures within the micropipette. The emission of light by a photoactive bacterium, Photobacterium fischeri, has been used as the end point of one study (Barber 1962), where the bacterial 5-20 ------- suspension was circulated through a waveguide. The bacteria were undergoing exposure in the waveguide for approximately half the time during a typical 43-min experiment. Bacteria were exposed at several frequencies between 2.6 and 3.0 GHz; the assay was performed 24 h later. In spite of extremely high dose rates, 660 to 5300 W/kg, there were no differences between microwave- irradiated and conventionally heated samples that received parallel treatment. In the single experiment conducted at 10-GHz frequency, a transient effect was found in the virulence of Agrobacteriurn tumefaciens towards its normal hosts, potato and turnip disks (Moore et aK 1979). This bacterium produces a plasmid, which it injects into the host cells and which is responsible for turning these cells into uncontrolled tumor cells. In this experiment, a suspension of A. tumefaciens was exposed in a petri dish for 30, 60, and 230 min. The longer exposures produced a decrease of virulence near 60 percent with no essential change in the number of viable cells. The effect was unchanged 6 h post-exposure, but virulence returned to normal 23 h post- exposure when the bacteria were maintained at 27 °C. An additional experiment, in which treated and untreated cells were added to the host simultaneously, indicated that the exposed cells were able to compete effectively with untreated cells for binding sites on the host cells. The treated bacteria were always maintained at or below 27 °C during irradiation. From the temperature data in the paper, a dose rate of ~ 1 W/kg and power density of 0.58 mW/cm can be estimated. A possible explanation for this effect, which was not offered by the original authors, is that the plasmid DNA of the A. tumefaciens was incorporated into the major DNA of the bacteria during microwave exposure, 5-21 ------- preventing injection into the host. Normal growth may have subsequently allowed the plasmid to return to its original state, restoring activity. In two papers, Wachtel and coworkers (Seaman and Wachtel 1978, Wachtel et al. 1975) have examined the effects of microwave irradiation on the firing rate of isolated neurons from the marine gastropod Aplysia. Neurons were exposed in a stripline at 1.5- and 2.45-GHz (CW and PW) fields (0.5- to 10-ps duration, 1000 to 15,000 pulses/s). The firing rate of pacemaker neurons and the burst rate of bursting cells were measured during microwave exposure at dose rates between 1 and 100 W/kg. In the majority of cases, the firing rate of pacemaker cells increased with an increase in temperature, and decreased with a decrease in temperature. In a minority of cases, 13 percent, for the pacemaker cells, the microwave irradiation reversed the normal change in firing rate; i.e., the rate decreased or stopped with a microwave-induced increase in temperature. The authors were able to detect slow and rapid components. The slow component, occurring in 30 to 60 s, was correlated with the slow rise of temperature associated with exposure. The rapid component, occurring within 1 s, appeared to correlate with the presence of the microwave field. The rapid component was always found to be a decrease in firing rate in the presence of the field and was never produced by convective heating. Similar but more variable effects were found for the bursting cells. The threshold for the slow component was ~ 7 W/kg, but in one case the rapid component was found at an SAR as low as 1 W/kg. In all cases where effects were found, the firing rates returned to normal when the radiation was terminated and when the temperature was returned to normal. The authors hypothesized that at a dose rate of ~ 1 W/kg, conversion of 0.1 percent of the microwave 5-22 ------- energy into a polarizing current density across the cell membrane would be sufficient to affect the firing rate of pacemaker neurons. 5.1.4 Unresolved Questions Several questions remain unresolved in the area of cellular and subcellular effects of microwave radiation. Nearly all of the relevant research is concentrated in a narrow frequency band, between 1.0 and 4.0 GHz. No acceptable studies have been reported for a large portion of the frequency spectrum of concern, 0.5 MHz to 100 GHz. A related question is that of the role of structured water in the cell. This is presently an active area of research. Questions such as how much water in a cell is structured to a greater degree than "bulk" water, and whether this structure plays an important role in cell metabolism, are as yet unanswered. Three monographs (Alfsen and Berteaud 1976; Drost-Hansen and Clegg 1979; Grant et al^. 1978) summarize knowledge in this area. Experiments attempting to establish the presence and extent of structured water within biological systems are described, and the dielectric data indicating a possible frequency range for the structured-water resonance are presented. According to the limited information now available, RF radiation is most effective in modifying the state of structured water at frequencies below 1 GHz. As yet no experiments have been conducted that define whether absorption of RF energy by structured water leads to a measurable change in a biological system. 5-23 ------- Three of the reports discussed in § 5.1.3, Effects on Single Cells, also raise questions. The explanation offered by Ismailov (1978) for the change in electrophoretic mobility of exposed RBC's is speculative and does not account for the peaking of the phenomenon 30 min post-exposure. These results must be considered an effect of unknown origin until more information is available. The results of experiments by Moore et aK (1979), in which the virulence of A. tumefaciens was decreased for more than 6 h after exposure to microwave radiation, suggest reversible functional changes in the organism. In this case, the implications for cellular function after microwave exposure would be broad. However, no other worker has noted a similar effect with other single- cell organisms, and this experiment has not been independently confirmed. Therefore, its significance is unknown at this time. The results of Wachtel and coworkers (Seaman and V/achtel 1978; Wachtel et al. 1975) are potentially highly significant because they indicate the possibil of a direct interaction between the microwave field and the functioning of the pacemaker neuron (i.e., the rapid effect). Three other workers have obtained supportive results. The documentation in these reports is not sufficiently complete for inclusion in the preceding sections, but they bear mentioning here. Yamaura and Chichibu (1967) found results strikingly similar to those of Wachtel in ganglia of crayfish and prawn exposed at 11 GHz. The regular firing rate of the ganglia decreased rapidly during microwave exposure, rebounded to a higher than normal rate when the radiation was removed, and then returned to normal. Temperature controls showed only an increased firing 5-24 ------- rate as the temperature was increased. The authors stated that the SAR was ~ 100 W/kg but did not describe the method of measurement. Arber (1976) found a hyperpolarization of the resting potential of giant neurons of the mollusk Helix pomatia during exposure to 2.45-GHz (CW) radiation. In this experiment, the ganglion was isolated and mounted in a stripline. The cell potential was measured, as in Wachtel's experiment, by inserting micro- electrodes into the neuron. Exposure to microwaves (SAR near 15 W/kg) for 1 h produced a 5- to 10-percent increase in resting potential, followed by a stabilization or slight additional increase over 1 h post-exposure. This result is presumably caused by a change in the Na+-K+ balance in the cell. When Arber treated the cells with ouabain (post-exposure), which inhibits Na+-K+ ATPase, he found that part of the hyperpolarization could be accounted for by the action of this enzyme under the influence of microwaves. The remainder was attributed to changes in passive ion transport. Pickard et al_. (1980) have recently presented results in which single cells from Chara braunii and Nitella flexilis, plants from the family Characeae, exhibited a large step increase in voltage when exposed to 0.1- to 5-MHz (PW) radiation (pulses were of 250-ms duration, pulse interval was 6.3 s). These authors also found a fast and slow component where the slow component correlated with a temperature rise in the sample. The fast component was frequency dependent and disappeared abruptly at ~ 10 MHz. The authors suggest that the fast component is produced by rectification of the oscillating electric field by the cell membrane. The fast component disappeared into noise at ~ 667 V/m and may have been an effect of intense fields. 5-25 ------- The results of these four studies indicate a possibility of a direct microwave interaction with the electric potential across the cellular membrane of all living cells. This kind of interaction would have broad significance in the functioning of all cells and, in particular, cells of the nervous system. Additional studies have been conducted recently that have not yet been sufficiently documented to include in the previous sections. However, because of their potential importance, they are discussed here. The first two reports concern microwave effects on phospholipid bilayers, whereas the third is concerned with changes in growth of the yeast Saccharomyces cerevisiae exposed at frequencies between 41 and 42 GHz. Tyazhelov et aJL (1979a) have exposed a phospholipid membrane formed between two chambers containing solutions of NaCl or KC1. An antibiotic that forms pores through the membrane was added to facilitate passage of Na+ or K+. Conductance was measured across the membrane while 4-s pulses of 900-MHz radiation were delivered at between 125 and 280 V/m (field strength in the aqueous medium). The results showed a change in conductance under exposure that is consistent with temperature rises of 12 °C, but the temperature of the NaCl or KC1 solutions did not vary by > 0.5 °C. However, information concerning the exposure system is insufficient to judge whether serious inhomo- geneities in energy deposition, and thus similar inhomogeneities in the temperature distribution, were likely. It is not possible to estimate an SAR from the information presented. 5-26 ------- Sheridan et aK (1979) have presented at meetings, but so far have not published, results of Raman spectroscopy on single- and multi-lamellar phospholipid vesicles exposed to 2.45-GHz (CW) radiation. No change was found in the Raman bands of single layered vesicles. In contrast, the data for multi-lamellar vesicles indicate that the hydrocarbon tails of the phospholipids were undergoing a temperature-dependent phase transition at a point where the bulk temperature was too low for the transition to have begun. The change was reported to be equivalent to a temperature difference of ~ 2 °C at an exposure of 25 niW/cm . The bulk temperature of the sample in these ex- periments was measured by a unique method. Small ruby crystals were suspended in the sample, and the shifts in the Raman bands of the crystal were measured and compared with calibration curves. This appears to be an accurate method for temperature determination. The origin of this effect is unknown, but if a similar effect were found in naturally occurring membranes, it could have an impact on the functioning of the biological membrane. Keilmann and coworkers (Grundler et aT. 1977; Keilmann 1978; Grundler and Keilmann 1980) have demonstrated that S. cerevisiae exhibits an enhanced or inhibited growth rate when exposed at certain closely spaced frequencies between 41.60 and 41.80 GHz. For instance, they found a 10- to 15-percent increase in growth rate at 41.64 and 41.68 GHz, and a 20-percent decrease at 41.66 GHz. The experiments were conducted using a unique waveguide termination that was dipped into a suspension of yeast cells. In a typical experiment, 24 W were dissipated in the yeast-cell suspension, and the authors estimated a maximum exposure intensity of about 10 mW/cm . However, because of the unusual 5-27 ------- nature of the waveguide termination and the high attenuation of high-frequency radiation by aqueous samples, SAR values are not determinate. Sample temperature was monitored and was within 0.5 °C of the desired 32 °C. Equivalent temperature controls were performed, and the authors believe that changes of the observed magnitude could not be purely temperature effects. This contention appears reasonable, since the growth rate both increased and decreased at the same levels of incident energy but at different exposure frequencies. The decrease is difficult to explain, based on the author's reports of no observable temperature rise. If, however, a localized temperature of > 37 °C (> 5 °C above the controlled temperature) were attained close to the waveguide termi- nation, then the decrease could be explained solely on the basis of temperature. There are scanty reports from the Soviet literature that present results similar to those of Keilmann and coworkers. In summary, the available literature on cellular and subcellular effects of microwaves does not yet definitely establish whether effects unrelated to elevation of temperature exist at dose rates on the order of 1 W/kg. Several investigators have reported effects, but a majority have not found effects unrelated to temperature variations. In some cases, the results conflict. In other cases, the effects found are equivocal. Effects of elevated temperature may not be clearly eliminated, or, as in the case of neuronal firing rate, the change in the rate may occur only in a minority of cases (Seaman and Wachtel 1978). The question of heat-induced temperature rises must always be carefully considered when dealing with the biological effects of exposure to microwave radiation. In cellular systems, a principal effect is to increase the rates of all biochemical reactions. This includes the rate of denaturation of 5-28 ------- proteins and DNA (Lehninger 1975) and the rate at which bases are removed (Lindahl and Nyberg 1974) and mutations occur (Bingham et al_. 1976) in DNA. When temperature is raised to a certain level, often ~ 43 °C, cell functions become so disrupted that the cell's capacity to repair the damage is exceeded and the cell dies. The effect of a rapid heating rate is less clear. Some disruption of cell function may be expected from a rapid temperature rise, even if the critical temperature, e.g., 43 °C, has not been reached. Whether this disruption is fully reversible has not been well documented. RF- radiation exposure can produce such high heating rates, especially from high peak, low average power pulses. Effect of nonuniform heat deposition is probably the most difficult aspect to account for when evaluating the effects of exposure to RF radiation. The effects presented in this section are sufficiently well established to warrant continued concern and effort. The effects of microwave radiation on the electrical properties of the cell are potentially the most significant. Four separate experiments have demonstrated these effects. All cells use the electrical potential across the cell membrane in their life functions, and perhaps the most important cells are those of the nervous system. Microwave effects have been found in the functioning of the nervous system and in behavior, as will be discussed in following sections. At this time, nearly all of the effects documented for cellular and sub- cellular systems are observed in intact cells. This could imply that a living cell is required for the necessary interaction with microwave radiation to occur. Or, it may be that the right questions are not being asked about the 5-29 ------- subcellular level, perhaps because the level of understanding and instru- mentation capabilities are as yet too limited. Although the primary mecha- nism(s) of interaction other than heating the water medium have not yet been defined, some useful directions are indicated by the results reviewed here. There also appear to be valuable correlations between some of the effects at the cellular level with those at higher levels of organization. 5-30 ------- 5.2 HEMATOLOGIC AND IMMUNOLOGIC EFFECTS Ralph J. Smialowicz Over the years, a considerable number of reports has appeared in the literature dealing with the effects of RF radiation on the hematologic and immunologic systems of animals. For the most part, the responsible investi- gators have been motivated by a concern for the possible adverse health effects of exposure to RF radiation. Studies in which animals have been exposed at different frequencies and intensities have shown inconsistent changes in elements of both biological systems. In some instances, a thermal burden to the exposed animal has been credited with the observed changes, whereas in others, a "nonthermal" (i.e., lack of measurable elevations of temperature) or direct (i.e., athermal or field-specific effects) interaction of RF radiation with the blood and blood-forming systems has been suggested as the causative mechanism for the observed effects. In any case, the final interpretation of RF-induced changes must consider many variables that affect the interaction of RF radiation with the biological entity. As has been noted previously, variables such as body shape and mass, radiation frequency, duration of exposure, field intensity, specific absorption rate, energy distribution, orientation of the body in the field, ambient environmental conditions, area of the body exposed, and field modulation may all influence the final results. Variability of response among species and strains as well as between sexes must also be considered. This section critically reviews the reported effects of RF radiation on the hematologic and immunologic systems of laboratory animals. From this review the following general conclusions can be drawn: 5-31 ------- • Partial or whole-body exposure of animals to RF radiation may lead to a variety of changes in the hematologic and immunologic systems. One of these changes most consistently found is an increase in lymphocyte formation and activity. • Transient changes in peripheral blood composition, probably caused by redistribution of blood cells and hemoconcentration, were reported; effects on the immune system have often turned out to be transient. • Several reports describe a causal relationship between increased core temperature and hematologic and immunologic changes. • Many of the reported effects of RF radiation on the hematologic and immune systems are similar to those (a) resulting from a stress response involving the hypothalamic-hypophyseal-adrenal axis, or (b) following administration of glucocorticoids. • That an increase in rectal temperature is not observed after exposure to RF radiation does not preclude a thermal interaction that the animal is able to compensate for and control. • Localized heating or "hot spots" in organs critical to the hematopoietic and immune systems may occur following the production of thermal gradients that are unique to RF-energy absorption by biological specimens. • There is a lack of convincing evidence for a direct interaction of RF radiation with hematologic and immunologic systems in the absence of some form of thermal involvement. • There is presently no convincing evidence from animal studies for adverse alterations in the hematopoietic or immune systems at RF- radiation intensities comparable to average environmental levels. For convenience, this section is divided into two general topics: hematologic effects and immunologic effects. These two topics are further subdivided into reviews of studies in which cellular components of these systems have been exposed jjn vitro and of studies dealing with j_n vivo ex- posures to RF radiation. 5-32 ------- 5.2.1 Hematology Hematology is the study of the anatomy, physiology, and pathology of the blood and blood-forming tissues. The hematopoietic system is comprised of a variety of cells and cell products. In fetal life, the production of blood cells occurs in the liver, spleen, and bone marrow. After birth this function is limited largely to the bone marrow, which produces red cells (erythrocytes), white cells (neutrophilic, eosinophilic, and basophilic granulocytes; lympho- cytes; and monocytes), and platelets. Each of these cell types performs specific functions that are essential to life. For example, mature erythrocytes transport Og and CO2 to and from tissues, granulocytes and monocytes phagocytize invading microorganisms, and lymphocytes are involved in immune responses. These functional cells are all descendants of progenitors (stem cells) that reside within the bone marrow. Blood-cell formation consists of two essential processes, proliferation and differentiation; progenitor cells proliferate in the bone marrow and then differentiate into red and white cells. As the process of cell differentiation progresses, the capacity for cellular prolif- eration decreases. Impairment of either of these processes may lead to dysfunctions in the hematologic system that may be life threatening. 5.2.1.1 In Vivo Studies— A paucity of information on the health effects of RF-radiation exposure from clinical and epidemiological studies has led to studies of effects on the hematologic systems of laboratory animals (Table 5-5). Many of the early investigations on the blood-forming system of laboratory animals employed 2 power densities of 10 mW/cm and higher. For example, Deichmann et al_. (1964) 5-33 ------- TABLE 5-5. SUMMARY OF STUDIES CONCERNING HEMATOLOGIC EFFECTS OF RF-RADIATION EXPOSURE Exposure Conditions Effects Species Frequency Intensity Duration SAR References (MHz) (mW/cn2) (days x o1n) (W/kg) Increased. WBC, lymphs, PKN, Rat 24,000 (PW) 10 1 x 180 f Deichaann et al (1964) RBC, Hct, and Hgb 20 1 x 420 3 0T No change Dog 24,000 (PW) 24 400 x 400-900 i' Deichmann et a^ (1963) No change Rat 3,000 (PW) 10 216 x 60 A Kitsovskaya (1964) Decreased RBC, WBC and lymphs 40 20 x 15 81 Increased Ptt< 100 6x5 20 Decreased lymphs and eosln Dog 2,800 (PW) 100 1 x 360 4; Michaelson et al^ (1964) Decreased WBC, PHN, and eosin 2,800 (PW) 165 1 x 120 61 Decreased- WBC, lymphs, and eosin 1,280 (PW) 100 1 x 360 4 5 Increased. PHN + Decreased: lymphs 200 (CW) 165 1 x 360 25 Increased pm No change Mouse 800 43 175 x 120 12 9^ Spalding et al^ (1971) Increased lymphs and mitotic Guinea pig 3,000 (CW or 3.5 120 x 180 0.5t Baranski (1971 and 1972a,b) index of lymphoid cells PW) Increased RBC, Hct, and Hgb Rat 2,400 (CW) 10 30 x 120 2* Djordjevic and Kolak (1973) No change Rat 2,400 (CW) 5 90 x 60 1* Ojordjevic et al (1977) Increased eosinophiIs Rabbit 2,450 (CW) 10 180 x 1380 1 5 McRee et al (1980a) Increased- WBC, CFU House 2,450 (CW) 100 1x5 70t Rotkovska and Vacek (1975) Decreased: s8Fe uptake (continued) ------- TABLE 5-5. (continued) Exposure Conditions Effects Species Frequency (HHz) Intensity Duration (nW/cm2) (days x nin) SAR (W/kg) References (Ji l CO in Accelerated recovery following x-irradiation, increased erythropoiesis and nyelopoiesis Accelerated recovery from x-irradiation Increased- PMN and RBC Decreased lymphs Accelerated recovery room x-irradiation Increased lymphs Decreased- PMN (Not reproduced consistently) No change No change No change Decreased Hct, WBC, and lymphs House Dog Chinese hamster Rat (Perinatal exposure) Rat (Perinatal exposure) Rat (Perinatal exposure) Quail egg Rat (young) 2,450 (CV) 2,800 (PW) 2,450 (CW) 425 (CW) 2,450 (CW) 100 (CW) 2,450 (CW) 2,736 (PW) 100 100 60 10 46 1x5 1 x 3600 1 x 30 47 x 240 57 x 240 57 x 240 30 1 x 1440 24 4 35 x 240 70' 28' 3-7 1-5 2-3 14 5-25t Rotkovska and Vacek (1977) Hichaelson et al. (1963) Lappenbusch et a^ (1973) Soialowicz et al (1982) Saialowicz et al. (1979a) Soialowicz et a^ (1981a) Hamrlck and McRee (1975) Pazderova-Vejlupkova and Josifko (1979) (continued) ------- TABLE 5-5. (continued) Exposure Conditions Effects Species Frequency (MHz) Intensity (raW/cB2) Ouration (days x mm) SAR (W/kg) References Ho change House 2,450 (CW) 30 22 x 30 22 Smtalowicz et a^ (1979b) in i Decreased lymphs Increased PMN House 26 (CW) 8610 13* Liburdy (1977) co CTl Decrease in CFU for erythroid and granulocyte-macrophage series House 2,450 (CW) 15 9 x 30 10 Huang and Hold (1980) Reduction in CFU granulocyte- macrophage precursors exposed in vitro House 2,450 (CW) 60-1000 1 x 15 120-2000 Lin et a! (1979b) WBC - white blood cell, PMN = polymorphonuclear leukocytes, RBC = red blood cell, Hct = hematocrit, Hgb = hemoglobin, and CFU = colony-forming unit SAR estimated ------- reported significant leukocytosis, lymphocytosis, and neutrophilia in rats following 7 h of exposure to 24,000-MHz (PW) microwaves at an average power O density of 20 mW/cm (SAR estimated at 3 W/kg). One week following exposure, 2 peripheral blood values returned to normal. Rats exposed for 3 h at 10 mW/cm displayed the same changes and returned to normal after 2 days (SAR estimated at 1.5 W/kg). Increases in circulating erythrocytes, hemoglobin concentration, and hematocrit were observed in two of three strains of rats (Osborne-Mendel 2 and CFN) exposed to 24,000-MHz fields at 10 or 20 mW/cm . However, in Fischer rats exposed under the same conditions, there was a reduction in the number of circulating erythrocytes and a reduction in hematocrit and hemoglobin concentration. In another experiment, Deichmann et cH. (1963) exposed two 2 dogs to 24,000-MHz (PW) fields at an average power density of 24 mW/cm (SAR estimated at 1 W/kg). One dog was exposed for 20 months, 6.7 h/day, 5 days/ week; whereas the second dog was exposed for 20 months, 16.5 h/day, 4 days/ week. No significant changes were observed in blood volume, hematocrit, hemoglobin, erythrocytes, total and differential leukocytes, blood cholesterol, or protein-bound iodine. The only symptom attributed to the exposure was a slight loss of body mass. Kitsovskaya (1964) exposed rats to 3000-MHz (PW) radiation at 10, 40, or 2 100 mW/cm for various periods of time (SAR estimated at 2, 8, and 20 W/kg, 2 respectively). No changes were found in rats exposed at 10 mW/cm ; at 40 and 2 100 mW/cm , however, the number of peripheral blood erythrocytes, leukocytes, and lymphocytes decreased, while granulocytes increased. These blood changes did not return to normal until several months after cessation of exposure. 5-37 ------- The apparent discrepancy between the results of Deichmann et a^. (1964) and Kitsovskaya (1964) may be partially explained by the work of Michael son et al^ (1964). These investigators reported that the hematopoietic effects of 2800- and 1280-MHz (PW) fields depend on the frequency, intensity, and duration of exposure. For example, dogs exposed to 2800-MHz fields showed a marked decrease in circulating lymphocytes and eosinophils after 6 h at 100 mW/cm (SAR estimated at 4 W/kg). This exposure resulted in a 1 °C mean increase of rectal temperature. Neutrophils remained slightly increased at 24 h, while eosinophils and lymphocyte values returned to normal levels. After a 2-h exposure at 165 mW/cm^ to 2800-MHz fields (SAR estimated at 6 W/ kg), there was a slight leukopenia, neutropenia, and definite hemoconcentration. These changes were accompanied by a rectal temperature rise of 1.7 °C. Eosinopenia was still evident 24 h after this exposure. General leukocytic changes were more apparent following exposure of dogs to 1280-MHz (PW) fields or to 200-MHz (CW) radiation. After exposure of dogs at 1280 MHz for 6 h at 100 mW/cm (SAR estimated at 4.5 W/kg), a leukocytosis and neutrophilia were observed. After 24 h the neutrophil level was still increased above pre- exposure levels. Lymphocyte and eosinophil values were slightly depressed following exposure, but at 24 h they were slightly higher than initial values. A 6-h exposure to 200-MHz (CW) fields at 165 mW/cm^ (SAR estimated at 25 W/kg) caused a marked increase in neutrophils and a slight decrease in lymphocytes. After 24 h this trend was more evident. Michael son et aH. (1964) suggested that the results indicated a stress response of the exposed animals in the hypothalamic and/or adrenal axis that was brought about by a thermal stimu- lation from RF-radiation exposure. 5-38 ------- Spalding et al_. (1971) exposed mice to 800-MHz fields at an average power 2 density of 43 mW/cm (SAR estimated at 10.7 W/kg) for 2 h/day, 5 days/week, for a total of 35 weeks. These investigators found no changes in blood erythrocytes, leukocytes, hematocrit, or hemoglobin concentrations. It is interesting that these investigators did not detect changes in the peripheral blood picture of exposed mice, despite the thermal burden that was being placed on these animals. Four mice died from "thermal effects" following the 33rd and 34th RF-radiation exposures. 2 Effects produced at levels at or below 10 mW/cm (SAR estimated at 0.5 to 2.0 W/kg) have also been reported. For example, Baranski (1971, 1972a,b) ex- posed guinea pigs and rabbits to 3000-MHz (CW and PW) microwaves at an average power density of 3.5 mW/cm (SAR estimated at 0.5 W/kg) for 3 months, 3 h daily. At this power level, the body temperature of the animals was not elevated. Observations were made of increases in absolute lymphocyte counts in peripheral blood, abnormalities in nuclear structure, and mitosis in the erythroblastic cell series in the bone marrow and in lymphoid cells in lymph nodes and spleen. No changes were observed in the granulocytic series in peripheral blood. Shifts in peripheral blood cells were found to correlate with changes in the cellularity of the spleen and lymph nodes. An increase in the mitotic index and in the percentage of cells incorporating H-thymidine was observed in the spleen and lymph nodes of exposed animals. Djordjevic and Kolak (1973) exposed rats to 2400-MHz (CW) fields at 2 10 mW/cm (SAR estimated at 2 W/kg) 2 h/day for 10 to 30 days. Body temperature in rats exposed under these conditions increased by 1 °C within the first 5-39 ------- 30 min of exposure and remained at this level throughout the exposure period. Hematocrit, hemoglobin concentration, and circulating erythrocytes increased during the 30-day exposure. Fluctuations in the various leukocyte populations were also observed. The authors suggested these changes were caused by the thermal effect of microwaves. In a more recent study, Djordjevic et al^. (1977) found no significant difference in any of several hematologic end 2 points for rats exposed to 2400-MHz (CW) microwaves at 5 mW/cm (SAR estimated at 1 W/kg) for 1 h/day during a 90-day period. Recently, McRee et aK (1980a) reported significant decreases in eosinophils and a lowering of albumin and calcium in blood from rabbits immediately following chronic exposure to 2450-MHz fields. In this study, rabbits were exposed 23 h daily for 180 consecutive days to 2450-MHz (CW) radiation at a power density 2 of 10 mW/cm (SAR =1.5 W/kg). No change in hematologic parameters was observed 30 days after termination of exposure (i.e., depression in eosinophils seen immediately following exposure had normalized); however, a significant decrease in albumin/total globulin ratio was observed in the blood of exposed rabbits at this time. The authors contend that, because only 3 of the 41 blood- chemistry parameters measured immediately after exposure were significantly different (P < 0.05), and because this observation is close to that expected by chance, further validation of these changes is warranted. Rotkovska and Vacek (1975) reported changes in hematopoietic cell popu- lations of mice following a single 5-min exposure to 2450-MHz (CW) radiation 2 at an intensity of 100 mW/cm (SAR estimated at 70 W/kg). The response of microwave-exposed mice was compared with that of mice placed in a warm-air 5-40 ------- chamber at an ambient temperature of 43 °C for 5 min. Both treatments caused a rise in rectal temperature > 2 °C. A leukocytosis occurred in mice under both conditions; however, the time course for the leukocytosis and the response of hematopoietic stem cells differed between the two treatments. Following RF-radiation exposure, a decrease in the total cell volume of the bone marrow and spleen was observed, and the number of hematopoietic stem cells in bone marrow and spleen, as measured by the colony-forming unit (CFU) assay, in- 59 creased. Incorporation of Fe in the spleen decreased 24 h after RF-radiation exposure. On the other hand, the exposure to heat caused a decrease in the 59 CFU's in bone marrow and spleen and an increase in the percentage of Fe incorporation. Rotkovska and Vacek concluded that the different effects of RF radiation and externally applied heat on the hematopoietic stem cells indicate that biological effects caused by high intensities of RF radiation may not necessarily be related only to increases in internal temperature. They indicated that their results suggest a possible "direct" effect. This study is significant because it demonstrates a marked difference in the kinetic response of the hematopoietic system to two forms of heat stress. Consequently, these differences must be considered in the interpretation of RF-radiation-induced changes in the hematopoietic system. Subsequently, Rotkovska and Vacek (1977) studied the effect of microwaves on the recovery of hematopoietic tissue following exposure to x-irradiation. Mice exposed to x-rays at 300 to 750 rads were then exposed to 2450-MHz (CW) 2 microwaves for 5 min at 100 iriW/cm (SAR estimated at 70 W/kg). The combined treatment resulted in an accelerated recovery of hematopoietic tissue, a heightened erythropoiesis and myelopoiesis, and an increased survival rate 5-41 ------- compared with x-irradiated mice. The increase in the number of endogenous hematopoietic colonies in the spleens of the x-irradiated mice after microwave exposure supports Rotkovska and Vacek's earlier (1975) observation of an elevation in the number of stem cells in the spleens of intact mice after microwave exposure alone. These investigators suggested that RF radiation may influence the mechanisms that activate the pool of stem cells, either by improving the repair of sublethal radiation damage or by increasing the proli- ferative capacity of stem cells that survive x-irradiation. The authors concluded that this acceleration of the processes of repairing radiation damage in hematopoietic cells after thermogenic doses of RF radiation depended on the stage of intracellular repair at the time of RF-radiation exposure. In earlier work, Michael son et aK (1963) reported that simultaneous exposure to 2 x-rays and microwaves (2800 MHz, PW modulated, 100 mW/cm , SAR estimated at 4 W/kg) caused an accelerated recovery of the hematopoietic function in dogs. Thomson et ajk (1965) reported that pretreatment of mice with RF radiation (2800 MHz, PW modulated, 100 mW/cm^, SAR estimated at 70 W/kg) reduced the mortality after x-irradiation (800 rads). The 30-day lethality was 40 to 55 percent among mice given single or multiple RF treatment prior to x-irradia- tion, compared with 76 percent lethality in mice not pretreated with RF radi- 2 ation. Exposure of Chinese hamsters to RF radiation (2450 MHz, CW, 60 mW/cm , SAR estimated at 28 W/kg, for 30 min) 5 min after x-irradiation (725 to 30 950 rads) significantly increased the x-ray LD^q dose compared with exposure of animals to x-rays only or with exposure to RF radiation before irradiation (Lappenbusch et al_. 1973). Lappenbusch et al_. reported that the radio- protective effect of RF radiation appears to be associated with a delayed decrease in the number of circulating white blood cells, reduced period of 5-42 ------- decreased cell density, and complete replenishment of white blood cells within 30 days following the dual treatment. Exposure to RF radiation either alone or combined with x-ray exposure increased the relative number of neutrophils, reduced the relative number of lymphocytes, and slightly increased the number of circulating red blood cells. On the other hand, animals exposed first to RF radiation and then to x-rays demonstrated a more severe leukocyte picture than hamsters x-irradiated only; in these animals, leukocyte counts decreased faster, and the animals developed leukopenia. The effect of exposure to RF radiation on circulating blood cells of de- veloping rats has been studied by Smialowicz et aK (1979a, 1982). Rats were exposed pre- and post-natally to 425-MHz (CW) fields at 10 mW/cm , 4 h daily up to 41 days. Because the animals were growing, SAR's ranged from 3 to 7 W/kg. No consistent changes in blood values were observed in exposed com- pared to sham-irradiated control rats (Smialowicz et al^. 1982). Rats exposed under the same regimen but to 2450-MHz (CW) fields at 5 mW/cm (SAR estimated at 1 to 5 W/kg) also showed no difference in circulating erythrocyte count, leukocyte and differential counts, or hematocrit and hemoglobin concentration when compared with sham-irradiated controls (Smialowicz et aj. 1979a). Rats exposed to 100-MHz fields at 46 mW/cm^ (SAR estimated at 2 to 3 W/kg) pre- and post-natally as above showed no change in blood parameters compared with con- trols (Smialowicz et aj. 1981a). Recently Pazderova-Vejlupkova and Josifko (1979) reported decreases in the hematocrits, number of leukocytes, and absolute numbers of lymphocytes in young rats exposed to 2736-MHz (PW, 395 Hz, 2.6-ps pulse width) microwaves 5-43 ------- 2 at 24.4 mW/cm for 7 weeks (5 days/week, 4 h/day). The means of body mass of rats at the beginning and at the end of the 7-week exposure period was 65 and 350 g, respectively (SAR estimated at 5 to 25 W/kg). These changes disappeared within 10 weeks after termination of exposure. The activity of alkaline phosphatase in neutrophils increased during the first week of irradiation but decreased transiently after the irradiation. In a similar experiment by the same authors (data not given), in which adult rats were exposed for 14 weeks 2 to 3000-MHz (PW, 300 Hz, 2.5-\|js pulse width) microwaves at 1 mW/cm (SAR esti- mated at 0.2 W/kg), no difference was observed in hematologic parameters between exposed and control rats. Hamrick and McRee (1975) examined the effect of RF radiation on developing birds. Quail eggs were exposed for 24 h during the second day of incubation to 2450-MHz (CW) fields at 30 mW/cm^ (SAR = 14 mW/g). At 24 to 36 h after hatching, quail chicks were examined for gross deformities, changes in organ weight, and hematologic changes. No significant effects due to RF exposure were detected. In another study, Smialowicz et cH. (1979b) exposed mice to 2450-MHz (CW) fields at 30 mW/cm^ (SAR = 22 W/kg) for 30 min on 22 consecutive days. These mice showed no significant difference in circulating-erythrocyte count, leukocyte and differential counts, or hematocrit and hemoglobin concentration compared with sham-irradiated controls. In this experiment, mice were maintained in an environmental chamber where temperature, humidity, and air flow were continuously controlled. Under the conditions of this study, the RF radiation did not significantly elevate rectal temperatures of exposed mice. In contrast, when 5-44 ------- mice were exposed to thermogenic levels (2 to 4 °C rise in rectal temperature) of 26-MHz (CW) radiation, 8610 mW/cm^ (SAR estimated at 13 W/kg), a decrease in the number of circulating lymphocytes and an increase in circulating neutrophils were observed immediately after exposure (Liburdy 1977). These mice were held in a chamber that lacked a continuous turnover of air. Liburdy (1977) reported that this shift reached its peak 3 h after exposure. The num- bers of circulating lymphocytes and neutrophils were reported to return to normal, pre-exposure levels 55 to 96 h after exposure. On the other hand, mice exposed at high ambient temperatures (79 °C) in a vented, dry-air oven showed an increased number of circulating lymphocytes and neutrophils for a 12-h period after exposure. It appears, therefore, that the response of circulating leukocytes to exogenous thermal loading depends on the means by which the body is heated. These results are similar to those reported by Rotkovska and Vacek (1977) and indicate that the heating properties of RF radiation differ from those of conventional modes of tissue heating. Recently Huang and Mold (1980) reported that bone marrow (cultured in vitro) from mice exposed to 2450-MHz (CW) fields at 15 mW/cm^ (SAR = 10 W/kg) for 30 min on 9 consecutive days had significantly fewer (P < 0.05) CFU's of both the erythroid and granulocyte-macrophage series. No data were presented on the peripheral blood counts of any of these blood cells that would confirm or expand the information gathered in the CFU assay. 5.2.1.2 In Vitro Studies— Lin et ah (1979b) reported a reduction in the number of CFU's (granulocyte and macrophage precursor cells) formed by bone marrow cells exposed j_n vitro 5-45 ------- to 2450-MHz fields at 60 to 1000 mW/cm2 (SAR's at 120 to 2000 W/kg). This reduction in colony formation was reported to be dose dependent and occurred without a significant rise in the temperature of the cell suspension. The authors indicate that their results point to a direct effect of microwave radiation on these hematopoietic precursor cells. Although these results are interesting, the in situ application of fields as intense as those required to produce the observed effects would certainly cause gross thermal injury to the tissue. In summary, levels of RF radiation that cause an increase in body tempera- ture elicit changes in the hematopoietic system that can for the most part be ascribed to a thermal stress response. Changes in the blood of animals exposed to RF radiation at intensities below those that cause an increase in core temperature suggest a similar stress-response mechanism. The failure to record an increase in core temperature does not preclude that the animal is compensating for the added thermal energy by thermoregulatory mechanisms. Indeed, lack of a temperature change indicates that thermoregulation is operat- ing. The response elicited by RF-radiation-induced heating, however, appears to differ from that of conventional heating because of the differing internal heating patterns of this form of radiation. 5.2.2 Immunology The immune system is comprised of myriad mechanical, cellular, and humoral components that act as the body's defense against various pathogenic microor- ganisms, viruses, and neoplasias. The immune system is divided into the humoral 5-46 ------- element, such as antibodies and complement, and the cellular elements, which are composed of the lymphoid and phagocytic cells. The cellular elements of the immune system are also part of the hematologic system. The phagocytic cells responsible for engulfing and digesting certain microorganisms are the neutrophils or polymorphonuclear leukocytes (PMN's) and the monocytes or macro- phages. In the presence of antibodies and complement, neutrophils are aided in engulfing and digesting invading organisms. The monocyte is also a phago- cytic cell. Monocytes move into an area in which an infection has begun and then differentiate into macrophages. Macrophages can be "activated" to kill certain microorganisms (e.g., intracellular, facultative bacteria such as Mycobacteria tuberculosis and Listeria monocytogenes, viruses, and fungi) through the interaction of certain subpopulations of lymphocytes, such as the T lymphocytes. The other cellular components of the immune system are the lymphocytes. These cells are broadly divided into two groups, the B lymphocytes and the T lymphocytes. Although these cells are similar morphologically, they are different functionally; B and T lymphocytes can be distinguished by the pres- ence of unique antigens or receptors on their membrane surface. Both T and B lymphocytes are believed to originate in the bone, marrow and then to proceed through various stages of development and differentiation, maturing into functional cells of the immune system. The B lymphocyte, or bursa-equivalent lymphocyte, is responsible for humoral immune responses. The B lymphocytes, after appropriate stimulation by 5-47 ------- antigens, proliferate and undergo morphological changes and develop into plasma cells that actively synthesize and secrete antibodies. The T lymphocyte, or thymus lymphocyte, is processed through the thymus after leaving the bone marrow. Classically cell-mediated or T-lymphocyte responses include protection against viruses, fungi, and several bacteria. T lymphocytes are also involved in reactions such as delayed hypersensitivity or contact hypersensitivity and rejection of tumors and foreign tissues such as transplants (allografts). Cell-mediated reactions are so named because these reactions, which operate by specifically sensitized T lymphocytes, can be transferred by these cells to normal animals. B-lymphocyte-mediated humoral responses, in contrast, are transferable by serum. Each element of the immune system—the T and B lymphocytes and macrophages—plays a cooperative role in defending the host against infection and disease. A delicate balance exists wherein the immune system is prevented from reacting to its own tissues to avoid autoimmune reactions. The alteration or dysfunction of any of these elements may lessen the host's ability to combat infection or may lead to autoimmune disease. However, because of adaptability and redundancy in the immune system, the host can survive subtle perturbations. Consequently, although subtle effects on the immune system may be generated by physical or chemical agents, all such effects may not lead to clinically significant immune dysfunctions. 5-48 ------- 5.2.2.1 In Vivo Studies-- A summary of in vivo studies concerning immunologic effects of RF-radiation exposure is presented in Table 5-6. Effects on adult animals—One of the most consistently found RF-radiation- induced changes in the hematopoietic system is the increase in lymphocyte formation and activity following exposure of animals of several species to RF radiation at various frequencies (Baranski 1971, 1972a,b; Czerski 1975). Con- sequently, there have been several studies of the effects of RF radiation on lymphocytes and the immune system. In a study reported by Czerski (1975), mice were exposed 2 h daily to 2950-MHz (PW modulated) microwaves at 0.5 mW/cm (SAR estimated at 0.5 W/kg) for 6 to 12 weeks. After 6 weeks, there was a large increase in the relative number of lymphoblasts in the lymph nodes of exposed mice. In another series of experiments (Czerski 1975), rabbits were exposed 2 h/day, 6 days/week for 6 months to 2950-MHz (PW) microwaves at 2 5 mW/cm (SAR estimated at 0.8 W/kg). After culturing for 7 days ifi vitro, peripheral blood lymphocytes from these animals were found to undergo an increase in "spontaneous lymphoblastoid transformation." Maximal increases occurred after 1 to 2 months of exposure; the transformation rate then returned to base line and rose again 1 month after an irradiation had been terminated. Miro et aJL (1974) continuously exposed mice to 3105-MHz (PW) microwaves over a 145-h period at an average power density of 2 mW/cm (SAR estimated at 2 W/kg. No description was given of how animals were fed or watered. An in- crease in lymphoblastic cells in the spleen and lymphoid areas of exposed mice was observed. A somewhat similar response was observed (Huang et al_. 1977) in lymphocytes cultured from Chinese hamsters that were exposed to 2450-MHz 5-49 ------- cn i CJ1 o TABLE 5-6. SUMMARY OF STUDIES CONCERNING IMMUNOLOGIC EFFECTS (IN VIVO) OF RF-RADIATION EXPOSURE Exposure Conditions Effects* Species Frequency (mi) Intensity (mW/cB) Duration (days x nin) Increase in lytnphoblast: in Mouse 2,950 (PW) lyaph nodes and increased response to SRBC Increase in "spontaneous" Rabbit 2,950 (PW) lymphoblast transformation of cultured lymphocytes Increase in lymphoblasts in House 3,105 (PW) spleen and lymphoid tissue Increased transformation of Chinese 2,450 (CW) unstimulated cultured lymphocyte hanster and decreased altosis In PHA- stimulated lymphocyte cultures Transient decrease and Increased Mouse 2,450 (CW) response of cultured lynphocytes to PHA, Con A, and LPS Increased mitosis of PHA- Rhesus 10-27 (PW) stimulated lymphocytes Donkey Increase in CR+, Fc+, and lg+ Mouse 2,450 (CW) spleen cells Increased response to B-cell mitogens Decrease in primary response to SRBC 0 5 5 or 15 1320 42 x 120 6-8700 5, 15, 30 5 x 15 or 45 1-17 x 30 1 x 30 1 or 3 x 30 SAR (W/kg) 0 5 References 24-48 x 120 0 8' Czerskl (1975) Czerskl (1975) Mlro et al (1974) 2 3, 6 9, Huang et al (1977) 13 8, or 20.7 3.6 or 10 Huang and Hold (1980) 0 4-2 0t Prince et aj (1972) 14 Wiktor-Jedrzejczak et al (1977a,b,c) (continued) ------- TABLE 5-6. (continued) Exposure Conditions Effects* Species Frequency (mi) Intensity (oM/ca2) Ouratlon (days x otn) SAR (W/kg) References Increase In Cr* and fc. spleen eel Is Mouse 2,450 (CW) — 1 x 15 1 x 30 11 8 5 Sulek et al (1980) Increase In response of cultured lyophocytes to T- and B-cell mitogens Rat 2,450 (CW) 425 (CW) 5 10 57 x 240 47 x 240 (Perinatal exp) 1-5 3-7 Saialowicz et a^ (1979a, No change Mouse 2,450 (CW) 5-35 1-22 x 15 or 30 4-25 Smalowicz et al (1979b) No change Rat 100 (CW) 46 57 x 240 2-3 Saialowicz et al. (1981a) Increase in T and B lynphocytes in spleen Decrease in DTH Mouse 26 (CW) 800 1 x 15 or 10 x 15 5.6 Liburdy (1979) Reduction of lymphocyte traffic froo lung to spleen House 2,600 (CW) 5 or 25 1 x 60 3.8 or 19 Liburdy (1980) Decreased response to PWM Rabbit 2.450 (CW) 10 1B0 x 1380 1 5 McRee et ak (1980a) No change Quail 2,450 (CW) 5 12 x 1440 4 03 Hamrick et a^ (1977) Decrease in tumor developoent Mouse 2,450 (CW) (near-field application) 11-14 day of gestation or 11-14 & 19-45 x 20 35 Preskorn et al. (1978) Decreased granulocytic response Rabbit 3,000 (CW) 3 (continued) 42 to 84 x 360 0 5 Szmigielski et al (1975) ------- TABLE 5-6. (continued) Exposure Conditions Effects* Species Frequency (HHz) Intensity (sM/cm2) Duration (days x tain) SAR References (W/kg) Timor regression and increase in antitumor antibodies and anti-BSA Rabbit 13.56 (near-field application) 1 x 10-15 (local Shah and Dickson (1978b) hyperthermia) Tumor inhibition and Immune stimulation Rat 2,450 (CW) 200W 3 or 6 x 45 (local Szmlgielski et aj (1978) hyperthermia) Increased tumorlcidal activity in lymphocytes and macrophages Mouse 1,356 600-900 1x5 (local Maroor et al (1977) hyperthermia) Tumor regression House 3,000 (CW) 40 1-14 x 120 28* Szmlgielski et a^ (1977) Increase in lung cancer colonies and inhibition of contact sensitivity to oxazolone House 2,450 (CW) 50 4, 7, 10 or 14 x 120 36* Roszkowski et a^ (1980) Decrease 1n response to BSA Rabbit 1,356 (near-field applicatlon) 3 x 60 (local Shah and Dickson (1978a) hyperthermia) SRgC = sheep red blood cells, PHA+= phytoheoagglutinin, Con A = concanavalin A, LPS = lipopolysaccharide, CR = complement-receptor positive, Ig = immunoglobulin positive, Fc = Fc portion of ionunoglobulln, DTH = delayed-type hypersensitivity, PWH = pokeweed mitogen, BSA = bovine serum albumin SAR estimated ------- 2 (CW) microwaves for 15 min on 5 consecutive days at 5 mW/cm (SAR = 2.3 W/kg) without a detectable rise of rectal temperature. Increased transformation to lymphoblastoid forms in the absence of mitogens was maximal in cultures from hamsters exposed at 30 mW/cm (SAR = 13.8 W/kg). Irradiation at this power density caused a 0.9 °C rise in rectal temperature of exposed hamsters. Mitosis of lymphocytes cultured in the presence of the mitogen phytohemag- glutinin (PHA), however, was depressed in cells obtained from hamsters exposed 2 at 5, 15, 30, or 45 mW/cm . These effects were reported to be transient and reversible; control levels were again observed after 5 to 10 days (Huang et aHL 1977). More recently, Huang and Mold (1980) reported an oscillating response of spleen cells to PHA, concanavalin A (Con A), and lipopolysaccharide (LPS) from mice exposed to 2450-MHz fields at 15 mW/cm^ (SAR = 10 W/kg) for 30 min. Mitogen responsiveness decreased significantly after 2 days of exposure, returned to normal after 4 days of exposure, and was significantly increased for all mitogens after 9 days of irradiation. Responsiveness to mitogens tended to return to normal or to fall to subnormal levels after 17 days of 2 exposure. In contrast, when mice were exposed at 5 mW/cm for 30 min on 5 consecutive days, a significant increase in the response to LPS was observed, 2 whereas no change was observed in LPS responsiveness to exposure at 15 iriW/cm for 5 days. When macrophages were removed from spleen-cell suspensions of mice irradiated at 15 mW/cm for 9 days, the responsiveness to LPS was greater than the already increased spleen-cell responsiveness without macrophage removal. However, addition of macrophages from the RF-irradiated mice to spleen-cell cultures from nonirradiated mice caused a significant decrease in 5-53 ------- responsiveness to LPS. The authors suggest that macrophage activation (macrophages from irradiated mice displayed increased spreading and increased phagocytosis of latex particles iji vitro) by microwaves may be responsible for inhibiting LPS responsiveness. Prince et jH. (1972) reported the opposite effect in rhesus monkeys. These investigators found an enhanced mitotic response of peripheral blood lymphocytes stimulated i_n vitro with PHA from monkeys 3 days following a 2 30-min exposure to 10.5-MHz (PW) radiation at 1320 mW/cm (SAR estimated at 0.4 W/kg). Enhancement of mitosis of cultured lymphocytes from monkeys similarly exposed to 19.27- and 26.6-MHz were also reported. Also reported were increases in circulating lymphocytes that ranged from 4 to 47 percent above pre-exposure levels. At a frequency of 26.6 MHz (SAR estimated at 2 W/kg), the rectal temperature of monkeys following exposure was reported to increase by 2.5 °C above pre-exposure levels. The particular susceptibilities of lymphocytes to RF radiation has led to an examination of the effects of this form of radiation on the immune response. For example, Czerski (1975) reported that mice exposed for 6 weeks to 2950-MHz (PW) microwaves at 0.5 mW/cm^ (SAR estimated at 0.5 W/kg) had i significantly greater numbers of antibody-producing cells and higher serum antibody titers following immunization with sheep red blood cells (SRBC's). Interestingly, mice that were exposed for 12 weeks did not show this increased responsiveness. More recently, Wiktor-Jedrzejczak et aJL (1977a,b,c) exposed mice in a rectangular waveguide to 2450-MHz radiation for 30 min at an average dose rate near 14 W/kg. At 3, 6, 9, and 12 days after a single or multiple 5-54 ------- exposures, mice were tested for (1) the relative frequency of T and B splenic lymphocytes, (2) the functional capacity of spleen cells to respond to T- and B-cell-specific mitogens, and (3) the inability to respond to SRBC's (a T-dependent antigen) or dinitrophenyl-lysine-Ficoll (DNP-lys-Ficoll, a T-independent antigen). A single 30-min exposure induced a significant increase in the proportion of complement-receptor-positive (CR+) lymphocytes in mouse spleens that peaked 6 days after exposure. This effect was further enhanced by repeated (three) exposures, which also produced a significant increase in the proportion of immunoglobulin-positive (Ig+) spleen cells (Wiktor-Jedrzejczak 1977a). A significant increase in the proportion of Fc-receptor-positive (FcR+) cells in the spleens was also observed 7 days after a single 30-min exposure (SAR - 13.7 W/kg). However, no change in the number of Ig+ cells in spleens of these mice was observed (Wiktor-Jedrzejczak 1977c). The type and combination of surface receptors (CR, Ig, Fc) expressed on splenic B cells represent different maturational stages in B-cell develop- ment. Wiktor-Jedrzejczak et al. (1977a,b,c) were unable to demonstrate any change in the total number of theta-antigen-positive (0+) T cells in the spleens of mice following a single or multiple exposure to 2450-MHz microwaves. No change was detected in the iji vitro spleen-cell response to stimulation by the T-cell-specific mitogens PHA and Con A or by pokeweed mitogen (PWM), which stimulates both T and B cells (Wiktor-Jedrzejczak et _aL 1977a). The response to the B-cell-specific mitogens lipopolysaccharide (LPS), polyinosinic polycyti- dylic acid (Poly I C), and purified protein derivative (PPD) of tuberculin, however, was significantly increased over controls following a single exposure. These results clearly correlate with the observed changes in the proportion of 5-55 ------- cells bearing different surface markers. These authors (1977a) noted that RF irradiation did not stimulate lymphoid-cel1 proliferation j>er se but rather appeared to act as a polyclonal B-cell activator, which led to an early maturation of noncommitted B cells. They also found a significant decrease in the primary immune response to SRBC's, a thymus-dependent antigen, in mice that had been immunized just prior to the first exposure to RF radiation. They suggested that this decreased response may result from the nonspecific stimulation of some cells by RF radiation to mature before they are activated by antigen (SRBC's), thereby increasing the proportion of unresponsive cells. Recently, Sulek et al^. (1980) corroborated this increase in CR+ and Fc+ spleen cells in mice 6 days following a 30-min exposure to 2450-MHz fields (average SAR = 12 W/kg). The kinetics for increased frequency of CR+ cells in the spleens of irradiated mice showed an initial increase 3 days following exposure that persisted for 5 to 6 days and then returned to normal within 9 to 10 days. The threshold for this effect was determined by varying the time of exposure with a constant forward power (0.6 W), or by maintaining a constant time of exposure while varying the forward power (0.1 to 0.78 W). It was shown that a minimum of a single 15-min exposure (11.8 W/kg) or a single 30-min exposure (5 W/kg) caused significant increases in CR cells 3 or 6 days post-exposure. The effect of absorption of multiple subthreshold quantities of microwaves was found to be cumulative only if the exposures occurred within 1 h of one another. The increase in CR+ cells was found at dose rates ranging from 10 to 18 W/kg for mice within a weight range of 18 to 25 g. The rectal temperature of exposed mice was found to be elevated no more than 0.6 °C above that of sham-irradiated mice. No change in 6+ (T lymphocytes) cells was 5-56 ------- observed. In contrast, Huang and Mold (1980) reported a significant increase in 0+ cells but no change in B cells (Ig+) of mice exposed to 2450-MHz fields 2 at 15 mW/cm (SAR at 10 W/kg) for 30 min on 9 consecutive days. Smialowicz et al_. (1979b) exposed mice to 2450-MHz (CW) fields under far-field conditions for 15 or 30 min daily for periods to, 22 consecutive days at power densities from 5 to 35 mW/cm (SAR's at 4 to 25 W/kg). Splenic lymphocyte function was assayed by the jjn vitro mitogen-stimulated response as 3 measured by H-thymidine incorporation after culture in the presence of T (PHA, Con A, PWM) or B (LPS, PWM, PPD) mitogens. The proportions of T (0+) and B (CR+) splenic cells and the primary immune response of mice to SRBC's were also studied. No difference in the response to mitogens or to SRBC's or in the frequency of T or B cells in spleens was observed in RF-irradiated compared with sham-irradiated mice. Liburdy (1979) reported that changes in splenic lymphocyte populations similar to those observed by Witkor-Jedrzejczak et al_. (1977a,b,c) can be produced by exposure of mice to thermogenic levels of 26-MHz radiation. When 2 mice were exposed to 26 MHz at an intensity (800 mW/cm , SAR at 5.6 W/kg) that produced a 2 to 3 °C rise in rectal temperature, a relative increase in splenic T and B lymphocytes was observed. Similar responses (i.e., increase in T and B cells) were induced following administration of the synthetic glucocorticoid methyl prednisolone sodium succinate. These results indicate that these RF- radiation-induced changes might represent some form of stress. It is difficult to understand how a 2 to 3 °C rise in rectal temperature would occur in mice irradiated for 15 min (SAR = 5.6 W/kg). A possible explanation is that these 5-57 ------- mice were restrained in perforated acrylic cages and held in an exposure chamber in which the air was not circulated during this period of irradiation. In examining further possible mechanism(s) for the shift in lymphocyte populations in the spleens of mice exposed to RF radiation, Liburdy (1980) ex- amined the circulation of. lymphocytes in microwave-exposed mice. Mice injected 51 with Cr-labeled syngeneic spleen cells were exposed for 1 h at either 5 or 25 mW/cm2 to 2600-MHz fields (SAR at 3.8 and 19 W/kg, respectively). Controls included sham-irradiated mice, mice held in a 63 °C warm-air oven for 1 h, and mice injected with methyl prednisolone sodium succinate. The distribution of injected cells was determined for the lung, liver, spleen, and bone marrow at o 1, 6, and 24 h post-exposure. Exposures at 25 mW/cm caused a 2.0 °C increase in core temperature. This regimen led to a 37-percent reduction in lymphocytes leaving the lung and migrating to the spleen. Also, a threefold increase in spleen lymphocytes entering the bone marrow occurred in this group of mice. A similar pattern of lymphocyte circulation was observed in the steroid-treated group. No change in lymphocyte traffic was observed in mice of the 5-mW/cm or warm-air groups. Liburdy (1980) concluded that these results support the idea that steroid release associated with thermal stress and attempts by the animal to thermoregulate during exposure to RF radiation are responsible for effects on the immune system. Effects on young animals—RF-radiation effects on the development of the immune response have been studied. Smialowicz et al.. (1979a) exposed rats on day 6 of gestation through day 41 to 2450-MHz (CW) fields at 5 mW/cm2 (SAR at 1 to 5 W/kg). The young animals absorbed microwaves at a higher rate (5 W/kg) 5-58 ------- than the adults (1 W/kg). In this study the exposed rats had lymphocytes that responded to a significantly greater extent than those from control animals following in vitro stimulation by T- or B-cell mitogens. A similar increase in lymphocyte responsiveness was seen in another study in which rats were exposed pre- and postnatally to 425-MHz radiation (SAR at 3 to 7 W/kg, with the neonates absorbing at the latter SAR) for periods to 41 days postpartum (Smialowicz et ah 1982). However, lymphocytes from rats exposed peri natal ly to 100-MHz fields at 46 iriW/cm^ (SAR at 2 to 3 W/kg) showed no change in mito- genic responsiveness (Smialowicz et ah 1981a). The results of the two former studies indicate that long-term exposure of developing (especially neonatal) rats to RF radiation at absorption levels higher than those achieved in the latter study may give rise to increased responsiveness of cultured lymphocytes. These results are similar to other reported changes in mammalian lymphocyte responsiveness following RF-radiation exposure (Czerski 1975; Prince et ah 1972; Wiktor-Jedrzejczak et ah 1977a,b,c). Although the benefits are not known concerning this increased responsiveness to mitogens by lymphocytes from animals exposed peri natally to RF radiation, a recent report by Preskorn et al_. (1978) indicates that irradiation at this time during development may be beneficial. These investigators exposed mice to 2450-MHz fields (SAR = 35 W/kg) for 20 min either on days 11 through 14 of gestation, or on days 19 through 45 postpartum, or during both periods. On the 16th day postpartum, all mice were implanted with a lymphoreticular cell sarcoma. Mice irradiated i_n utero only (colonic temperature increase of 2.2 °C in dams) showed a lower incidence of tumors (13 percent vs. 46 percent for sham-irradiated mice) 93 days postpartum. In mice irradiated _i_n utero and postnatal ly, tumors initially developed at a lower rate compared with controls; however, after 2.5 months, no difference 5-59 ------- was observed in tumor incidence between groups. At the end of 4 months, the tumor incidence in irradiated mice was slightly greater than controls (46 vs. 40 percent, respectively). An interesting finding, however, was that both tumor-bearing and tumor-free animals that had been irradiated only j_n utero lived longer on the average than their respective controls. This result is somewhat similar to that reported by Prausnitz and Susskind (1962), who found that mice briefly irradiated hundreds of times by a highly thermogenic level of RF radiation survived longer than controls. McRee et al^. (1980a) reported that 30 days after termination of a 6-month 23-h daily irradiation to 2450-MHz fields (SAR = 1.5 W/kg) spleen cells from rabbits showed a decreased responsiveness to PWM. Decreased responsiveness to PHA and Con A by these spleen cells was also reported, but responses to PHA and Con A were not statistically different from those of controls. Although these results are interesting, they are not conclusive, because only four exposed and four sham-irradiated rabbits were employed. It should also be noted that both irradiated and sham-irradiated rabbits were transported from one laboratory to another between the termination of RF exposure and spleen-cell assay. Hamrick et al^. (1977) examined the avian humoral-immune response in Japanese quail exposed to RF radiation during embryogenesis. Fertile quail eggs were continuously exposed to 2450-MHz (CW) radiation at 5 mW/cm (SAR = 4.03 W/kg) throughout the first 12 days of development. At 5 weeks of age, quail were immunized with SRBC's, and the levels of anti-SRBC antibodies were determined. No difference was observed in the antibody titers of exposed and 5-60 ------- sham-exposed quails. The masses of the bursa of Fabricius (site of B-cell production in birds) and spleen were not altered significantly by exposure to RF radiation. Effects on phagocytosis—RF-radiation-induced effects on phagocytic leukocytes of animals have been reported by Szmigielski et ah (1975). Rabbits were exposed to 3000-MHz fields for 6 h daily for 6 to 12 weeks at 2 3 mW/cm (SAR estimated at 0.5 W/kg). After the last exposure to RF radiation, rabbits were infected with an intravenous injection of virulent Staphylococcus aureus. At periods before and after infection, functional tests of granulo- poiesis were performed. The investigators reported a decreased production of mature granulocytes in infected, RF-radiation-exposed rabbits, which was mani- fested as a more serious illness in these animals. Summary--Exposure of laboratory animals to RF radiation can lead to changes in the functional integrity of lymphocytes. These cells play an important role in the immune-defense system of man and animals. The significance of the changes caused by RF radiation is difficult to interpret. Although some studies indicate that RF radiation causes an increased responsiveness of lymphocytes (Czerski 1975; Smialowicz et ah 1979a, 1982; Prince et ah 1972; Wiktor-Jedrzejczak et ah 1977a,b,c) and a potentiation of the immune response to antigen (Czerski 1975), others indicate a depression in responsiveness (Huang et ah 1977; Wiktor-Jedrzejczak et a2- 1977a; Liburdy 1979; Szmigielski et ah 1975). In most cases these alterations can be attributed to a stress response, since qualitatively similar but quantitatively more pronounced changes are observed at obviously stressful thermogenic levels 5-61 ------- of RF radiation (Huang et aK 1977; Huang and Mold 1980; Prince et ah 1972; Liburdy 1979) or after administration of glucocorticoids (Liburdy 1979). It is well known that stress alters physiological systems that regulate immunolog- ical function. Both immunosuppression and immunoenhancement have been observed to result from stress (Monjan 1981; Blecha et aL 1982), which is not inconsis- tent with the reported effects of RF radiation on the immune system. Many questions remain about the role of stress in the modulation of the immune sys- tem, not the least of which, in the present discussion, deal with RF-radiation- induced stress. Effects caused by RF-induced hyperthermia--A1terations in the immune system can be produced by RF-radiation-induced hyperthermia. Whole-body microwave-induced hyperthermia has been reported to serve a therapeutic role either alone (LeVeen et ah 1976) or in combination with ionizing radiation (Nelson and Holt 1978). In many cases, the direct destruction of malignant tissues by RF-radiation-induced heating is the ultimate goal. However, in some cases, hyperthermia has led to changes in the immune response. For example, Shah and Dickson (1978b) reported that following local heating of VX2 (carcinoma) tumor-bearing rabbits by a 13.56-MHz field, tumor regression and host cure were observed in 70 percent of the rabbits. Intratumoral tempera- tures of 47 to 50 °C were achieved within 30 min. Along with tumor regression, cell-mediated immunity—as measured by skin reactivity to tumor extract and dinitrochlorobenzene—markedly increased. A hundredfold increase in serum levels of antitumor antibody and increased response to the antigen bovine serum albumin (BSA) were also observed. In contrast, whole-body hyperthermia led to temporary reduction of tumor growth, followed by a return to an exponential 5-62 ------- increase in tumor volume and rapid death of the rabbit. This course of events following whole-body hyperthermia was accompanied by abrogation of the enhanced cellular and humoral immune responsiveness, observed following local RF-induced heating. Szmigielski et al^. (1978) reported that local heating (43 °C) of the Guerin epithelioma in Wistar rats by 2450-MHz (CW) radiation inhibited tumors and stimulated the immune reaction against the tumor. Other immune reactions stimulated by this treatment were the antibody response to BSA, high reactivity of spleen lymphocytes to the mitogen PHA, and increased serum lysozyme levels as a measure of macrophage activity. Tumor-specific reactions observed were increased cytotoxicity of spleen cells and peritoneal macrophages to cultured tumor cells. Similar results were reported by Marmor et ah (1977), who exposed tumors in mice to focal 1356-MHz radiation. The EMT-6 tumor was found to be extremely sensitive to RF heating. The cure rate was a function of temperature and duration of exposure. A 5-min exposure at 44 °C reduced the tumors by almost 50 percent. To determine the effectiveness of RF-induced heating on tumor regression, tumor-cell survival was studied by treating EMT-6 tumors i_n situ. Cell inactivation by RF-radiation-induced heating was similar to that for heating by a hot water bath. The results indicated that direct cell killing could not account for the observed cures, and these investigators suggested that hyperthermia (RF or convection-induced) may stimulate a tumor- directed immune response. Szmigielski et al_. (1977) exposed mice bearing transplanted sarcoma-180 tumors to 3000-MHz radiation, 2 h daily on the 1st through 14th day after 5-63 ------- 2 transplantation, whole-body at 40 mW/cm (SAR estimated at 28 W/kg). This exposure led to a 3 to 4 °C increase in rectal temperature and resulted in a reduction of tumor mass by ~ 40 percent, a reduction enhanced when microwave hyperthermia was combined with Colcemide, Streptolysin S, or both. Colcemide enhances the inhibiting effect of hyperthermia on proliferation of cells _in vitro (Szmigielski et al^ 1976), and Streptolysin S is an antineoplastic agent. Szmigielski et al_. (1978) suggested that immunostimulation is important in the complex inhibition of tumor growth by increased temperature. While many have heralded local and systemic hyperthermia as a possible cancer treatment, either alone or in combination with drugs or ionizing radiation, there is evidence that hyperthermia may enhance the dissemination of certain cancers and abrogate the immune response. For example, Roszkowski 2 et aK (1980) reported that exposure of mice to 2450-MHz radiation at 50 mW/cm for 4, 7, 10, or 14 days, 2 h daily (SAR estimated at 36 W/kg) caused an in- creased number of lung-cancer colonies and an inhibition of contact sensitivity to oxazolone (a measure of T-lymphocyte activity) with increased duration of hyperthermia. Shah and Dickson (1978a) exposed normal rabbits either to RF-radiation-induced (13.56 MHz) or to watercuff-local hyperthermia of thigh muscles, which were maintained at 42 °C for 1 h on 3 consecutive days. No alteration in the response to dinitrochlorobenzene challenge was observed. However, the humoral immune response to BSA was significantly depressed. This response was independent of the method and degree of heating. The results indicate that B lymphocytes might be more susceptible to hyperthermic damage than are T lymphocytes. 5-64 ------- The above results indicate that if the applied microwaves are of suffi- cient intensity to cause heating of tissue or of the whole animal, changes in the immune system will follow. And, with heating to any extent, the hypotha- lamic- hypophyseal -adrenal axis plays a major role in the responses elicited. It is well known that endogenous or exogenous adrenal glucocorticoid hormones affect the immune response. But what of the RF-induced responses reported in the absence of measurable temperature increases or at rates of energy absorption well below that of the resting or basal metabolic rate? Observed changes in the immune response under these conditions are more difficult to explain on a thermal-stress basis, primarily because of a lack of sensitive techniques to detect subtle stress responses. The fact that no increase of rectal tempera- ture occurs after exposure to RF radiation, however, does not preclude that the animal's capablility to compensate for added thermal energy by thermoregulatory mechanisms. RF radiation may also cause focal heating (thermal "hot spots") in organs critical to the immune response. 5.2.2.2 J_n Vitro Studies— Among the studies in this area (Table 5-7), several have involved attempts to determine if in vitro exposure of lymphocytes to RF radiation leads to "direct" changes in the metabolic or functional state of these cells. In an early study, Stodolnik-Baranska (1967) exposed human lymphocytes in culture to 2 3000-MHz (PW) microwaves at 7 or 14 mW/cm . Some lymphocytes were irradiated 2 2 4 h/day at 7 mW/cm for 3 to 5 days, while those exposed at 14 mW/cm were irradiated 15 min daily for 3 to 5 days. After 5 days in culture, the micro- wave-exposed cells were found to have undergone a fivefold increase in lymphoblastoid transformation compared with controls. Czerski (1975) attempted 5-65 ------- TABLE 5-7. SUMMARY OF STUDIES CONCERNING IMMUNOLOGICAL EFFECTS (IN VITRO) OF RF-RADIATION EXPOSURE* Exposure Conditions Effects Species Frequency (HHz) Intensity Ouration (iiM/cb2) (days x min) SAR (W/kg) References tn l Increased blastogenesis of exposed lymphocytes jn vitro Increased blastogenesis No change in Mitogen response to PHA, Con A or LPS No change in mitogen response to PHA No change in viability or growth Decreased macrophage phagocytosis Liberation of intracellular hydrolytic enzymes and Increased death Human 3,000 (PW) lymphocytes Hunan 10,000 lymphocytes House spleen 2,450 (CV) cells Rat blood 2,450 (CW) lymphocytes Human 2,450 (CW) lyophoblast cell lines (Daudi and hsb2) House 2,450 (CW) macrophage Rabbit 3,000 (CW) granulocytes 7 14 5-15 10 10-500 50 1 or 5 3-5 x 240 t 3-5 x 15 Observed effect t only when culture temperature ap- proached 38 °C 1 x 60, 120, or 19 240 5, 10 or 20 1 x 240, 1440, or 2640 1 x 15 1 x 30 1 x 15, 30, or 60 0 7, 1 4 or 2 8 25-1200 Stodolnlk-Baranska (1967) Baranskl and Czerski (1976) Saialowicz (1976) Hamrick and Fox (1977) Lin and Peterson (1977) 15 J/min Mayers and Habeshaw (1973) t Szmigielski (1975) PHA = phytoheoagglutinin, Con A = concanavalin A, and LPS = lipopolysaccharlde ^Unable to calculate SAR ------- without success to repeat this experiment. But, in a more recent study, Baranski and Czerski (1976) reported that exposure of human lymphocytes to 10,000-MHz fields at power densities between 5 and 15 mW/cm could induce lymphoblastoid transformation (SAR not given). At power densities below 2 2 5 mW/cm , this effect was not observed, whereas at power levels above 20 mW/cm , cell viability decreased. The induction of blastic transformation depended on 2 termination of irradiation (5 to 15 mW/cm ) at the moment when the temperature of the medium reached 38 °C. These results indicate that the microwave-induced blastic transformation might be caused by a thermal mechanism. Similar increases in the lymphoproliferative response of cells exposed to temperatures > 37 °C have been reported. Ashman and Nahmias (1978) reported that human lymphocytes, when cultured at 39 °C with the mitogens PHA or Con A, 3 showed an enhancement and earlier onset of H-thymidine incorporation compared with cultures incubated at 37 °C. In a similar study, Roberts and Steigbigel (1977) reported that the iji vitro human lymphocyte response to PHA and the common antigen streptokinase-streptodornase was enhanced at 38.5 °C relative to 37 °C. Smith et al^ (1978) reported that the in vitro response of human lymphocytes to PHA, Con A, PWM, and allogeneic lymphocytes was markedly enhanced by culture at 40 °C compared with 37 °C. These studies demonstrate the need to monitor and to control the temperature of cultures exposed to RF radiation. Without adequate temperature data, it is virtually impossible to accept _in vitro effects as due to RF radiation itself. The proliferative response of lymphocytes exposed iji vitro to RF radiation appears to be related to culture temperature. Smialowicz (1976) exposed 5-67 ------- murine splenic lymphocytes to 2450-MHz (CW) radiation for 1, 3, or 4 h at 10 mW/cm (SAR = 19 W/kg). Immediately after irradiation the temperature of the exposed cultures did not differ significantly from that of controls, and cell viability was unchanged. These cells were then cultured for 72 h in the presence of T- or B-cell mitogens, and the proliferative response was measured by H-thymidine incorporation. No difference was found in the blastogemc response of microwave-exposed and sham-exposed spleen cells to any of the mitogens employed. In a similar experiment, Hamrick and Fox (1977) exposed rat lymphocytes to 2450-MHz (CW) radiation for 4, 24, or 44 h at 5, 10, or O 20 mW/cm (SAR's at 0.7, 1.4, or 2.8 W/kg, respectively). The transformation 3 of unstimulated or PHA-stimulated lymphocytes was measured with H-thymidine. No significant differences were found in the proliferative capacity of lympho- cytes from exposed and control cultures. The effects of RF radiation on the growth and viability of cultured human lymphoblasts was studied by Lin and Peterson (1977). Human lymphoblasts (cell lines Daudi and HSf^) were exposed to 2450-MHz (CW) radiation in a waveguide for 15 min at incident power densities of 10 to 500 mW/cm (corresponding SAR's were 25 to 1200 W/kg). No temperature increase was observed, even at the highest power density in the capillary tube that held the cell suspension in the waveguide. No change was observed in the viability or growth of microwave-exposed lymphoblasts compared with controls. These studies provide further evidence that in the absence of heating, no change in lymphocyte activity occurs following RF- radiation exposure in vitro. In vitro exposure of macrophages to 2450-MHz fields has been reported by Mayers and Habeshaw (1973) to depress phagocytosis. Monolayer cultures of 5-68 ------- mouse peritoneal macrophages were perfused with suspensions of human erythro- cytes while being simultaneously exposed to 2450-MHz radiation at 50 mW/cm . The rate of energy absorption in the sample was 15 J/min. The phagocytic index of exposed cultures was significantly lower than that of the control after a 30-min exposure. Macrophage phagocytic activity returned to normal levels if RF irradiation was discontinued. During irradiation, a 2.5 °C temperature increase was observed; however, the final temperature in the culture vessel in any given experiment reportedly did not exceed 36.2 °C. The investigators concluded that the observed depression of phagocytosis in the irradiated cultures was not thermally induced. The 2.5 °C rise in temperature during irradiation, according to the authors, would have been expected to enhance rather than depress phagocytosis, since optimal phagocytosis occurs at 38.5 °C. The mechanism by which the observed effect occurs is not known; however, heating effects are difficult to dismiss because of an observed 2.5 °C rise in culture temperature. Although the temperature of the suspension medium did not exceed 36.2 °C, thermal gradients of much higher temperature would be expected at the macrophage-glass interface. An RF-induced effect on granulocyte integrity and viability was reported by Szmigielski (1975). Rabbit granulocytes were exposed in vitro to 3000-MHz 2 (CW) radiation at 1 or 5 mW/cm (SAR not given) for 15, 30, or 60 min. Cultures 2 exposed at 5 mW/cm for 30 or 60 min showed increased numbers of dead cells as demonstrated by an increase in nigrosine staining and an enhanced liberation of lysosomal enzymes. Exposure at 1 mW/cm did not cause increased cell death but did lead to a partial liberation of hydrolase enzymes. No change was observed in the temperature of the irradiated cultures. The liberation of 5-69 ------- acid phosphatase and lysozyme from granulocytes was observed in cell suspensions 2 exposed at 1 or 5 mW/cm , both suspensions exhibiting a time- and dose-dependent response. In summary, exposure of laboratory animals to RF radiation may lead to changes in the functional integrity of leukocytes, which play important roles in the immune-defense system. The significance of the changes caused by RF radiation is difficult to interpret, since many observed effects are transient and reversible. Furthermore, some studies indicate that RF radiation causes immunopotentiation, whereas others indicate immunosuppression. In many cases the observed alterations in the immune system can be attributed to thermal stress, because qualitatively similar but quantitatively greater changes are observed at obviously stressful (highly thermalizing) levels of RF radiation or following the administration of glucocorticoids. A possible explanation for the immunomodulating effects of RF radiation arises from the timing of the measurements of immune responsiveness after subjecting an animal to stress. For example, corticoid-induced impairment of immune responsiveness is commonly followed by homeostatic recovery, then subsequent overcompensation, which may be associated with immunoenhancement. As for those reports in which measurable elevations of temperature from RF-radiation-induced heating are not detected, a possible role by RF-radiation- induced thermogenesis cannot be dismissed. The failure to detect a measurable increase in tissue or core temperature in RF-irradiated experimental animals through the use of conventional techniques indicates that the animal was able to compensate for the added energy. The role that thermoregulating mechanisms 5-70 ------- play in affecting the immune response needs further study. There is presently no convincing evidence for a direct effect of RF radiation on the immune system in the absence of a thermal (heating) effect. 5.2.3 Unresolved Questions Several questions remain unanswered regarding the effects of RF radiation on the hematologic and immunologic systems. Perhaps the most perplexing question is what to make of the many Soviet reports on immunology. In most cases these reports lack sufficient technical detail for adequate critical assessment of reported results. Alterations in the hematopoietic and immunologic systems have been reported in animals exposed to RF radiation at and below 10 mW/cm over periods of weeks to months. Nevertheless, no con- vincing evidence has been presented to demonstrate a direct effect of RF radiation in the absence of thermal involvement; well defined and planned, 2 chronic (months-to-years), low-level (< 1 mW/cm ) studies have not been carried out to investigate Soviet claims of possible induction of autoimmune reactions following chronic exposure. Investigations of the possible hematologic and immunologic effects of PW vs. CW irradiation have not been undertaken in re- sponse to the claims by Eastern European investigators that PW modulation is more effective than CW irradiation in causing alterations in these systems. Investigations are lacking that would define possible synergistic effects of other agents or drugs with RF radiation on the hematopoietic and/or immune systems. RF radiation (at hyperthermic doses) has been shown to provide a protective effect against damage by ionizing radiation to the hematologic 5-71 ------- system. Although this is a beneficial effect of RF radiation, it is not known if the combination of drugs or other physical agents with RF radiation is detrimental. Another question has been raised that is based on the recent work of Szmigielski et ah (1980), who described the increased incidence of cancer development in mice exposed chronically to 2450-MHz (CW) fields at 5 and 2 15 mW/cm . This is an extremely important piece of work that merits much additional consideration. 5-72 ------- 5.3 REPRODUCTIVE EFFECTS Ezra Berman This section on the reproductive effects of RF radiation is organized into three categories: Teratology (5.3.1), where the treatment is admin- istered to the pregnant dam and observations are then made on the embryo, fetus, neonate, or older offspring; Reproductive Efficiency (5.3.2), where the end point of the experiment occurs in the primary and secondary sex organs of the parent; and Testes (5.3.3), where testicular morphology and function of testes are examined for alterations. RF radiation has been examined intensively for its potential reproductive effects for a variety of reasons: 1) many reproductive toxicologic tests can be carried out in a simple and inexpensive experimental design after insult by RF radiation, 2) the laboratory techniques used are conventionally acceptable as toxicologic assays, and, especially, 3) detrimental reproductive changes carry an emotional impact. 5.3.1 Teratology While making judgments on the potential of RF radiation for teratologic manifestations, it is important that the science of teratology and its under- lying principles be kept in mind. Wilson (1973) has derived general principles of teratogenesis. These principles are rephrased here to famil- iarize the reader with the guidelines one should keep in mind when evaluating the available literature, and so that their relevance to RF radiation teratologic investigations is more easily understood. 5-73 ------- (1) Susceptibility to RF radiation teratogenesis depends on species, strain, and stage of development at the time of exposure. (2) There are four indications of abnormal development: death, malformation, retarded growth, and deficient function. (3) These indications increase in incidence and degree with in- creasing dosage. Other aspects of teratology, as a branch of toxicology, are presented in discussing effective experimental design and interpretation. The interpre- tation of teratologic phenomena is still an art, and there is not full agreement on the significance of certain fetal manifestations after potentially toxic agents are administered. The discussion posits that the list of possible deviations in the conceptus is an order of decreasing degree of severity (death > malformation > growth retardation > functional deficit). Further, the significance that any laboratory animal or system may have as a model of human exposure is not universally accepted among scientists. An attempt is made to derive from available data three aspects of terato- logic toxicology of RF radiation: the presence of a teratologic effect, the generalization of that effect across species, and the dose-response character of that effect. From the evidence presented, we think that the reader will agree with the following statements: (1) RF radiation can cause teratogenesis in all the mammalian species studied adequately so far if sufficiently high power densities or SAR's are obtained. 5-74 ------- (2) Reduced fetal weight seems to occur consistently in rodents exposed gestational!}? to teratogenic or subteratogenic doses of RF radiation. (3) Gestational exposure to RF radiation can cause a functional deficit in later life if sufficiently high SAR's are obtained. 5.3.1.1 Nonmammalian Models-- The potential for human teratology is usually sought in mammalian models. But other models that are lower on the phylogenetic scale or cell-culture models are also often elucidative. Workers at the Bureau of Radiological Health of the Food and Drug Administration have used an insect to demonstrate the reproductive alterations due to RF radiation. Pay et a2. (1978) examined the egg production of female fruit flies (Drosophila melanogaster) in response to RF radiation. Using a 2450-MHz waveguide exposure system housed in an environmentally controlled chamber (24 °C, 50 percent relative humidity), these workers determined the survival rate of IL melanogaster pupae at SAR's ranging from 400 to 800 W/kg. Approximately 70 percent of the pupae did not survive a 10-min exposure when the rate of energy absorption was 640 W/kg; the temperature of the agar surface on which the pupae rested was 45 °C. If pupae were incubated instead in a 44 °C environment without RF radiation, the death rate was not as severe (50 percent). The potential for RF-radiation-induced teratology in birds has been examined using the fetal form of birds, the egg, as a model. This is a reasonably popular model since using bird eggs is more economical than using mammals. Studying the bird egg can provide insight into fetal effects independent of maternal parent influence. The egg is an object that can be 5-75 ------- placed in almost any desired position by the experimenter; it will remain there, a distinct advantage over free-ranging mammals. The shape of the egg represents a technical advantage over the mammal in that it is symmetrical externally, lending itself more easily to estimates of local and total absorption. Theoretically, the bird egg gives no advantage to mammalian models in the study of teratogenic potential of RF radiation because concepts of organo- genesis apply equally to avian fetuses and mammalian fetuses. To dismiss the considerable research on the teratologic effects of RF radiation in birds' eggs on the basis that the egg is poikilothermic whereas the mammal's egg is homeothermic would be an error. The mammalian conceptus is also poikilo- thermic; it has little, if any, control of its own temperature since it is entirely surrounded by placental fluids. It can remove itself of thermal energy only by radiating into surrounding maternal tissues because its capacity to rid itself of thermal energy is totally dependent on the gradient between itself and surrrounding tissues. If the mammalian conceptus' tissue temperature is increased because of RF-radiation absorption, it also must radiate that absorbed energy into its surroundings (dam), just as the egg in similar exposure conditions must radiate energy into the air of its incubator. If air or maternal temperatures are detrimentally high, then the loss of thermal energy from the egg or mammalian fetus is affected similarly. Carpenter et al_. (1960a) have studied RF radiation as an inducer of cataracts. These investigators have also likened the growth in the lens to the process of embryonic development and growth, where proliferation and 5-76 ------- differentiation take place concurrently. To further understand the RF- radiation-induced effects that appear in the lens, Carpenter et aK performed experiments using the egg of the domestic chicken. They irradiated almost 500 chick embryos at the 48-h stage and examined the eggs 48 h after irradiation, scoring incidences of survival and structural abnormalities. The irradiation was conducted in an anechoic chamber at 2450 MHz and at power densities of 2 200, 280, and 400 mW/cm . Estimates of SAR's for these power densities are 70, 98, and 140 W/kg, respectively (Durney et aK 1978). Power densities were determined by calorimetry using a saline-filled egg. The exposure durations ranged from 1 to 15 min. The eggs were incubated at 39 °C air temperature before, during, and after irradiation. At the 96th hour of development, all eggs were opened, and the embryos were removed, fixed, stained, and examined as whole-mounts. Experimental results relevant to this section are presented in Table 3 of the Carpenter et aK 1960a publication. We have used the data in that table to calculate the relative summed incidences of dead and abnormal 96-h egg embryos for each experimental group (power density:time) and present the incidences in Figure 5-2. The teratologic end points observed by Carpenter et aK (1960a) were death and abnormal morphology, which can be viewed as a continuum of one effect, interference with the normal processes of development. When abnormal structure is so severe as to interfere greatly with the continuation of development, the embryo cannot develop or maintain sufficiently normal 5-77 ------- structural and physiologic systems. The relationships between these systems collapse and the embryo dies. As stated earlier, indications of abnormal development increase in incidence and degree as the dosage increases. One mechanism for the increase in death is that, as the dosage increases, more embryos affect more changes in more systems, resulting in more deaths. 2 o "o £ 1/3 a o CO < > z o z a z < 100- 90- 80- 70- 60- 50- 40- < S tr i 20 oa < 30- 10- 0- ~r 10 ~r 14 "T" 15 11 12 13 DURATION OF EXPOSURE (min) Figure 5-2. Summed incidence of abnormal and nonviable chick embryo eggs (percent of total) exposed to 2450-MHz radiation for varying durations at 70, 98, or 140 W/kg (Carpenter et al. 1960a). Figure 5-2 contains three curves, one for each group of eggs exposed at dose rates of 140, 98, or 70 W/kg for varying durations. The data are plotted 5-78 ------- against the summed relative incidence of affected (dead or morphologically abnormal) eggs. The slopes of the curves are steep, indicating that small increases in the exposure duration caused disproportionate increases of effect. For example, the incidence in the group exposed for 10 min at 140 W/kg was 9 percent. An increase of only 30 s in exposure time caused a ninefold increase (80 percent) in incidence. An additional characteristic of these slopes is the disproportionate spacing along the exposure axis of the three curves. Studies of RF radiation effects using the Japanese quail have been conducted exclusively by workers at the National Institute of Environmental Health Sciences. In their first attempt to induce terata in the quail egg, Hamrick and McRee (1975) exposed fertilized eggs in an anechoic chamber to 2450-MHz (CW) fields producing an SAR of 14 W/kg in the egg. The eggs were exposed for 24 h beginning 2 days after laying, then were returned to their incubators until 24 to 36 h after hatching. Several experimental replications occurred so that the sample numbers were 102 in the control group and 110 in the RF-irradiated group. After hatching, observations were made of cellular elements of blood and of selected organ weights. No significant differences between the two groups were seen in hatchability, blood cell parameters, body weights, or organ weights. These authors also reported a similarly designed experiment on Japanese quail eggs, in which the eggs were exposed during the first 12 days of development at SAR's of 4.03 mW/g (McRee and Hamrick 1977). As before, the eggs were allowed to hatch and were then examined for teratologic alterations and blood-cell parameters. When the experiment was conducted at an 5-79 ------- environmental temperature of 35.5 °C, there were no significant changes in body weights, organ weights, or blood-cell parameters. The authors also demonstrated the importance ambient temperature can play in experiments in which RF radiation is the toxic agent under investigation. Absorption of RF radiation energy produces increased temperature. In most studies of reproductive effects reviewed in this document, fields were at sufficiently high power densities to cause significant depositions of energy. Temperatures in the animal subjects were usually increased by 0.5 °C or more. Therefore, the ambient temperature in which the subject is maintained or treated can be a significant determinant of a resulting trend. Levels of ambient temperature can be especially relevant to experimental subjects that do not have a significant thermoregulatory capability. For example, the report by McRee and Hamrick (1977) included the results of an experiment where Japanese quail eggs were in an environment at a temperature only slightly higher (+1.5 °C) than in the experiment discussed above (35.5 °C). The authors began their study using an ambient temperature of 37 °C because that is the temperature at which quail eggs are conventionally incubated. But this ambient temperature, with the addition of a 4 W/kg SAR, was sufficient to cause a 93 percent death rate of the fertilized eggs. The eggs that were irradiated at the same SAR, but at an ambient temperature lower by 1.5 °C, sustained only a 42 percent death rate. The results of this seemingly slight (1.5 °C) increase in incubation temperature during RF irradiation shows that for this poikilo- thermic model, the quail egg, the temperature at which the egg is maintained during experimentation is critical. 5-80 ------- 5.3.1.2 Mammalian Models-- Morphologic teratology—The mouse is commonly used in the laboratory for determining the teratogenic potential of toxic agents and also appears to be a useful model for determining the teratogenic potential of RF radiation. The mouse's size is conducive to this kind of experimentation because the apparent resonance of mice is 2450 MHz, a frequency at which irradiation equipment is commonly and economically available. Rugh et aK (1974, 1975) conducted several experiments on the mouse at 2450 MHz using the experimental techniques from work on the teratology of ionizing radiation to explore teratologic effects of microwaves. The in- vestigators used a waveguide exposure system in exposing individual pregnant CF1 mice. All mice were exposed once on the 8th day of pregnancy to CW fields at 2450 MHz for 2 to 5 min at a forward power that delivered a measured average energy dose ranging from 2.45 to 8 cal/g; there were no control or sham-irradiated mice in this study. These and a few other authors have reported dose parameters in calories or Joules per gram, which we have converted to the SAR unit (W/kg) used in this document. Each of the texts of the two papers by Rugh et a^. (1974, 1975) refers the reader to the other for more detailed methodology, but neither adequately describes exposure factors to convert accurately the dose values to dose rate (SAR). Our best solution was to convert to a range. The range of doses given by Rugh et a2 is 10.3 to 33.5 J/g delivered over 2 to 5 min. The four possible combinations (Table 5-8) of two extremes of dose 5-81 ------- TABLE 5-8. CONVERSION OF J/g TO W/kg* Average Dose Exposure Duration SAR (J/g) (min) (W/kg) 10.3 5 34 10.3 2 85 33.5 5 112 33.5 2 280 W/kg = J/g x 1000 t seconds (from Rugh et aT.. 1974, 1975). (10.3 and 33.5 J/g) and two extremes for exposure duration (2 and 5 min) result in SAR's ranging from 34 to 280 W/kg for this study. The most probable range of values is 85 W/kg (10.3 J/g for 2 min) to 112 W/kg (33.5 J/g for 5 min). The authors clearly demonstrated the teratogenicity of RF radiation in mice, especially the potential for death and generation of anomalies. By determining the energy dose in each pregnant mouse, and because varying doses were absorbed during the 2 to 5 min of exposure, the effect observed in each litter could be assigned to a definite whole-body average dose. This is the strong point in the Rugh et al^. studies. Rugh et al. (1974, 1975) showed that resorption of fetuses is a character- istic of the levels of RF-radiation exposure. Resorptions appeared at a dose rate of approximately 85 W/kg. Both the incidence of resorptions within litters and the incidence of litters with resorptions increased as the average 5-82 ------- dose increased. Near the maximum dose (33.5 J/g, approximately 110 W/kg) almost all the litters were affected. Even at 25 J/g, there were litters with 100 percent resorption rates. Anomalies "... were various, but the one used for this study, and which occurred most frequently, was exencephaly (brain hernia) " The report also included data on resorptions, dead, and stunted fetuses. In lieu of numerical data, the incidences of brain hernia (also termed encephalocele, exencephaly, or cranioschisis) are given by the authors in one figure. Data, as percentage of affected litters, were grouped into cells 8.4 J/g in width that ranged from 10.9 to 35.9 J/g. Approximately 30 percent of the litters given an average dose of 15 J/g (approximately 90 W/kg) had fetuses with brain hernias. Of those litters given an average dose of 31.8 J/g (estimated as 109 W/kg) almost all contained fetuses with some form of brain herniation. Though sham, cage-control, or historical control value for the incidence of brain hernia in this mouse strain (CF1) is not given, the authors state "... since this aberration rarely if ever occurs without trauma, it is signi- ficant when even a single exencephaly occurs.". However, one report of the incidence of spontaneous brain hernia in CF1 mice showed an occurrence of 11 exencephalic litters in a total of 90, approximately 12 percent of litters (Flynn 1968). Apparently, spontaneous incidences of anomalies are quite variable. A report by Rugh and McManaway (1977) included data on mice exposed once for 4 min to 2450-MHz RF radiation (SAR = 100 to 114 W/kg) on days 0 to 11 (10 5-83 ------- mice per day). In this group of mice (we cannot differentiate these data from those reported in Rugh et ah 1975 and Rugh 1976a,b) are uneven incidences of stunted fetuses (irradiated on day 4 of gestation) and death (irradiated on day 3, 8, or 10 of gestation). Rugh considers small fetuses as stunted and believes that this is a teratologic symptom. Rugh's experiments included using a single, but high field intensity for a very short time (i.e., 4 min on the 8th day of gestation). A longer period (100 min daily during entire organogenesis) was used by Berman et ah (1978) in the exposure of pregnant CD1 mice to 2450-MHz CW fields at power densities of 0 to 28 mW/cm in an anechoic chamber. Between 15 and 25 mice were irradiated simultaneously, each in a small ventilated container. Mean rectal temperature at the end of exposure to the highest power density (28 mW/cm ) was 38.2 °C. SAR's were determined by twin-well calorimetry and averaged 2.0 2 2 W/kg at 3.4 mW/cm , 8.1 W/kg at approximately 14 mW/cm , and 22 W/kg at 28 mW/cm^. This study incorporated the examination of several variables of fetal response to toxic agents. Values were tabulated by power density for pregnancy rates, incidences of live and dead fetuses, live fetal weights, and incidences of anomalies. Values were expressed using the litter as the experimental unit. This is a conservative approach in experiments in which the potential fetal toxicity of an agent is determined (Atkinson 1975). Berman et ah (1978) also "corrected" the mean fetal weight of each litter by a factor representing the potential influence that litter size had upon fetal 5-84 ------- weight. This technique is also endorsed by Atkinson, because it allows greater confidence in the interpretation of data. The study by Berman et ah (1978) demonstrated that daily exposure to 2450-MHz fields at intensities producing an SAR of 22 W/kg during the entire period of organogenesis caused a 10-percent decrease in body weight of mouse fetuses. The practical significance of this result is difficult to determine without experimental confirmation of the permanency or lack of reversibility of the smaller fetal weight. Until it can be shown that the effect is permanent (stunting) or temporary (delayed growth), the effect remains hygienically unresolved. The effect will have more practical significance if it can be shown that stunting is the result. Berman et ah (1978) also reported on the incidences of anomalies. There did not appear to be a significant increase of individual or total anomalies at any SAR. The C3H-HeJ strain of mouse was used by Chernovetz et ah (1975) in an attempt to determine the effects of fetal irradiation with 2450-MHz RF radiation. Pregnant mice were irradiated in a multimodal cavity for 10 min on the 11th, 12th, 13th, or 14th day of gestation at an SAR (determined by calorimetry) of 38 W/kg. The fetuses were examined near term (near the end of the gestation period) without any difference seen between the RF-irradiated and the sham-irradiated groups of fetuses. The variables observed in these fetuses related to the morphologic and the lethal aspects of the RF irradiation. Fetal body weights were not observed in these animals. 5-85 ------- The scientific reports on the teratologic potential of RF radiation in the rat represent the major fraction of rodent work. The main reason for this, perhaps, is that many of the studies that have looked for morphologic changes have also included functional teratology as one of the end points. For that kind of study, especially behavioral studies, the rat appears to be preferred over the mouse. Chernovetz is perhaps the most published author on the teratologic potential of microwaves in the rat. She reported (Chernovetz et ah 1977) the results of exposures of pregnant rats to 2450-MHz fields in a multimodal cavity (most likely the same device used to irradiate mice in experiments discussed above in Chernovetz et ah 1975) at an SAR calculated by calorimetry of approximately 31 W/kg. Exposures were conducted during 1 of 7 days, from the 10th through 16th day of gestation, and the fetuses were examined on the 19th day of gestation. Each single exposure lasted 20 min. The experiment included a total of 74 time-bred female rats, each assigned to 1 of 7 days of gestation and to 1 of 3 treatment groups (RF-irradiated, infrared-irradiated, or sham-irradiated). The result is an experimental design with 21 different cells into which 74 rats were distributed. By the end of the 20-min exposure period, the mean rectal temperature in the surviving pregnant rats was 42 °C. Of the original group of 30 bred females that were used in the RF-irradiation group, 7 (or 23 percent) died, while none died in the sham-irradiated group. Therefore, the application of 2450-MHz RF radiation to bred rats for this duration at this SAR is lethal for a significant portion of the animals involved. 5-86 ------- No numerical data on the gross teratologic consequences of this RF irradiation were given in the paper. However, comments were made in the text regarding the lack of any structural abnormalities. The number of resorptions was significantly higher (approximately 6 times) in the RF-irradiated than in the sham-irradiated fetuses, especially in the dams that had been irradiated on the 11th day of gestation. Fetal weights were also altered by the irradi- ation regimen; there was a small but statistically significant decrease in fetal weight. The alterations seen in the fetuses, in this case decreased fetal weights and death (in the form of resorptions), reflect the teratogenic potential of RF radiation exposure in the rat. It is important to remember that the fetal alterations were found from RF-radiation exposure regimens where rectal temperatures rose above 40 °C. Chernovetz et al. (1979) reported on the effect of 2450-MHz RF radiation in pregnant rats, this time at SAR's of 0, 14, or 28 W/kg. The same multi- modal cavity was used, through which air at 22 ± 1.5 °C flowed at 0.75 m/s. Pairs of bred female rats were exposed for a 20-min period on the 8th, 10th, 12th, or 14th day of gestation and then were examined on the 18th day of gestation. With cage controls added to each of these cells, the complete experimental design used 4 different days of treatment and 4 treatments (cage control, 0 W/kg [shams], 14 W/kg, and 28 W/kg), and 6 bred females per cell, for a total of 72 bred females. Those rats irradiated at 28 W/kg developed rectal temperatures of approximately 42 °C, temperatures that were similar to the author's previous experiments in which bred female rats were dosed at 31 W/kg. The rats irradiated at 14 W/kg had a mean peak rectal temperature near 40 °C. 5-87 ------- The values used by the authors in the statistical analyses were individual fetal values, not litter mean values. At the higher SAR level (28 W/kg), only ~ 10 percent below the sublethal level used in their previous report, the authors did not demonstrate lethality in the dams due to RF-radiation exposure. The 20-min exposure period at other levels of SAR produced no gross morphologic alterations in the fetuses or any severely edematous fetuses; there were also no statistically significant changes in resorption rates. The results of fetal weights appear to be most interesting. In the previous study (Chernovetz et aK 1977), the authors reported that exposure for 20 min at approximately 31 W/kg on odd-numbered days of gestation produced a significant decrease in fetal weight. In the 1979 study, a similar result was also obtained from exposure at 14 W/kg on even-numbered days (12 and 14), or 28 W/kg on the 8th day of gestation. Unexpectedly, exposure at 28 W/kg on the 12th or 14th day produced an increased body weight. The unexpected in- creased body weight is an interesting observation, especially when it is viewed against the decreased fetal body weight documented in the same authors' previous paper and other reports of RF-radiation-induced fetal alterations in rats and mice. Berman et al^. (1981) were not able to elicit any fetotoxic or teratologic responses (body weight; numbers alive and dead; external, visceral, or skeletal morphology) in fetal rats irradiated daily (100 min/day) during gestation, even with a large number of litters. The dams were exposed to 2450-MHz radiation at power densitites of 0 or 28 mW/cm (4.2 W/kg). These conditions 5-88 ------- caused a mean rectal temperature of 40.3 °C by the end of the 100-min exposure. Kaplan (1982) did not find differences in the survival of squirrel monkeys up to 10 months of age after daily i_n utero and postnatal exposures to 2450-MHz in multimodal cavities. Pregnant squirrel monkeys were irradiated for 3 h daily, beginning in the first trimester; after birth, the daily irradiation of the dams and their young were continued, and from 6 to 10 months of age only the young monkey was irradiated. The SAR was determined by calorimetry to be 3.4 W/kg. In 21 irradiated and 22 sham live-born, dif- ferences were not seen in sex ratio, the number dead by 10 months of age, or in the mean age at death. Postmortem examinations did not reveal any tendency to specific causes for death. The results of this study did not confirm the results of suspected increased lethality in squirrel monkeys from a previous study. Functional teratology—Functional teratology is a field of reproductive toxicology that has allowed increasingly greater insight into the toxicology of environmental agents. The field contains a wide variety of specialists who examine the functional capacity of neonatal or growing animals after J_n utero exposure. The survival rate of fetuses can be used as an indicator of immediate effects on the fetus. The rate of survival of neonatal, infant, or older off- spring can be also used as an indicator of delayed alterations in functional 5-89 ------- capacity long after insult during the fetal stage. For example, Kaplan et al_. (1982) reported a study of the postnatal effects of jrn utero exposure (at dose rates to 3.4 W/kg) to 2.45-GHz RF radiation in squirrel monkeys. One effect the authors observed was an increase in neonatal deaths. However, a more intensive examination (Kaplan 1981) did not demonstrate any change in death rate and did not confirm the earlier study. A simple test to determine whether RF irradiation iji utero would alter the survival of mice after birth was conducted by Rugh (1976a). In this experiment, pregnant CF1 mice were exposed on the 9th, 12th, or 16th day of gestation to a regimen of approximately 4 min of irradiation at 2450 MHz, which resulted in a mean dose of ~ 25 J/g (104 W/kg). The survivors of this sublethal exposure were allowed to give birth and their young to mature to 2 months of age. At that time, the offspring were again irradiated in the same device until each one was dead. The time-to-kill and the mean dose-to-kill was determined as a measure of the radiosensitivity of the offspring irrad- iated originally as fetuses. At the time of death, the rectal temperatures of these mice ranged from 40 to 51 °C. Because of body weight differences in the sexes at 2 months of age, the analyses were carried out by sex. Time-to-kill was shortened in males irrad- iated originally during the 12th or 16th day of gestation. The males irradiated originally on the 12th day of gestation had a lower mean dose- to- kill. Therefore, Rugh (1976a) has shown that irradiation by RF radiation on select days iji utero can alter subsequent (and far removed in time) effects of re-irradiation with microwaves. 5-90 ------- As noted, one effect observed commonly in fetuses exposed to RF radiation is decreased body weight. There are difficulties associated with interpreting altered fetal weight as an indicator of toxicity, especially when the permanency of decreased fetal weight (delayed growth vs. stunting) is not yet determined. But Rugh (1976a) gives evidence that mice irradiated as fetuses weigh less at 2 months of age. In that study, the mean weights of 2-month-old male offspring of dams irradiated during the 9th, 12th, or 16th day of gestation were all lower than concurrent sham-irradiated males (P < 0.05). Females irradiated during the 16th day of gestation were also smaller than their controls (P < 0.05), but not females from the 8th or 12th day groups. These results are evidence of a permanent change (stunting) in mice caused by RF irradiation. Chernovetz et a_L (1975) carried out an experiment to observe alterations in functional capabilities after _in utero irradiation. Pregnant mice were exposed once during the 14th day of gestation at an SAR of 38 W/kg, and the offspring were examined until weaning for survival. There were 15 dams each in the RF-irradiated and sham-irradiated groups. The statistical analyses, done by litters, show that there was a slight (approximately 12 percent) in- crease in the survival rate of the RF-irradiated litters as compared to the sham-irradiated controls. Some experimental designs continue irradiation past birth. One example of chronic administration at 2450 MHz is reported by Smialowicz et al. (1979a), where pregnant rats were irradiated daily for 4 h/day at a power density of 5-91 ------- 2 4 mW/cm from the 6th day of gestation to term; irradiation of the offspring continued through 40 days of age. The SAR's were determined by twin-well calorimetry for several ages of the animals. Pregnant rats weighing 300 to 350 g had a mean SAR of 0.7 W/kg; offspring 1 to 5 days of age and 6 to 10 g in weight absorbed approximately 4.7 W/kg. There was no significant dif- ference between the mean body weights of the males (female offspring were not used in this experiment) in the 12 sham-irradiated litters when compared to the mean body weights of the males in the 12 RF-irradiated litters. Shore et aL (1977) reported decreased body weight and decreased brain weight in postnatal rats exposed in utero to 2450-MHz RF radiation. In this experiment, pregnant rats were exposed at an average power density of 10 mW/cm^ (SAR estimated, from Durney et aL 1978, to be 2.2 W/kg) for 5 h per day repeatedly on days 3 through 19 of gestation. Rats were allowed to de- liver naturally and were observed frequently thereafter. Rats were grouped by E- or H-field orientation during exposures, and only the 3-day-old offspring that had been exposed parallel to the E-field were significantly different, having lowered body and brain weights. We cannot easily explain this prefer- ential decrement on the basis of either age or orientation during exposure. Our expectation is that there is no significant difference due to orientation in whole-body SAR in rats at this frequency (Gage et aL 1979). In experiments conducted by Michaelson et aK (1978), some aspects of functional teratology of RF radiation were explored in Long-Evans rats. In these experiments, rats were exposed to 2450-MHz RF radiation at power 5-92 ------- 2 densities of 10 mW/cm for 1 h on the 9th and 16th day of gestation, or 2 40 mW/cm for 2 h on the 9th, 13th, 16th, or 20th day of gestation. According 2 to data based on water-calorimetry, an exposure at 40 mW/cm produced an SAR 2 of approximately 10 W/kg and approximately 2.5 W/kg at 10 mW/cm . Exposure at 2 40 mW/cm caused an increase in the mean of rectal temperature of approximately 2 1 to 2 °C over that of the sham-exposed animals. The exposure at 10 mW/cm (2.5 W/kg) on the 9th or 16th day of gestation caused a significant increase (approximately 0.5 to 1 °C) in the temperature of dams at the 16th day of gestation. There were no statistical differences in the sizes of born litters as a consequence of exposure during gestation. There also appeared to be no difference in the growth rates up to 21 days of age of the rat pups that were exposed to RF radiation as compared with shams, nor in their relative brain weights. At the symposium where Michaelson described his work, Johnson et aK (1978) reported on a study, then in progress, on the functional teratologic effects in rats exposed to 918 MHz for 20 h per day for 19 days of gestation at a power density of 5 mW/cm . The experiment is especially interesting because it was conducted at a frequency close to the resonance of the experi- mental subject (rat), and it is the only reported study conducted with almost continuous exposure. The SAR in this study is approximately 2.5 W/kg, determined by calculations from thermographic scans. The eight RF-irradiated and eight shamirradiated bred dams were maintained in the waveguide exposure apparatus under environmental conditions of 22 °C, 45 percent relative humidity, and ad libitum access to rat chow and water. The exposure was begun 5-93 ------- on the first day of pregnancy, the day on which the copulatory plug was first seen. In this experiment, the dams remained in individual waveguides or sham condition until day 20 of gestation, at which time they were removed and placed in delivery cages. At 4 days of age, the litters were culled to four males and four females per litter, which were then observed through 91 days of age. There appeared to be no differences in the litter means for the number of pups born or for number dead during the first day after birth. There were no pups in any of these litters (including RF-irradiated, sham-irradiated, and cage-controls) with any visible physical defects. Body weights at birth, at 28 days of age, and at 91 days of age were analyzed for significant dif- ferences on an individual pup basis. The only statistically significant differences in body weights were found between the RF-irradiated and cage- control females at 28 days of age. A difference also existed in males at 91 days of age, when the RF-irradiated and the sham-irradiated males weighed less than the cage-control males. Another parameter of development used as an assay of functional tera- tology in this study was the age at eye-opening. The phenomenon of eye- opening is a little understood but frequently used indicator of developmental maturation. There appeared to be a significant shift to earlier eye-opening in animals exposed to RF radiation during gestation, and this maturation was earlier by approximately 1 day. 5-94 ------- Jensh reported a series of experiments designed to examine pre- and postnatally the effects of 915-MHz, 2450-MHz, and 6-GHz irradiation of rat fetuses (Jensh et al_. 1979; Jensh 1979, 1980). These studies included irra- diation up to 8 h daily during most of gestation with power densities not 2 sufficient to cause increased core temperatures (915 MHz, 10 mW/cm , 2 W/kg; 2450 MHz, 20 mW/cm^, 3 W/kg; 6 GHz, 35 mW/cm^, 3.5 W/kg; SAR estimates from Durney et al^. 1978, p. 96). Offspring were examined pre- and postpartum in a complex series of observations meant to locate behavioral or morphological alterations. None were noted. In lieu of any other workable hypothesis supportable by experimental data, the most plausible available theory that can account for RF radiation effects is that, upon absorption, microwaves deposit energy that is converted to thermal energy. There remain differences between theories that attribute individual effects to a general input of energy, or the heating of local (hot-spot) areas, or some more subtle contribution of microwaves not charac- teristic of conventional methods of heating animals. Comparisons between RF heating and infrared or convective heating have been conducted in the study of hyperthermia as a teratogenic agent. Edwards is a frequently cited investigator of the effects of hyperthermia on the develop- ment of mammals. His extensive work on this agent as a cause of congenital malformations is perhaps best summarized in one of his reviews (Edwards 1974), which describes his experiments on the teratologic manifestations of hyper- thermia in the guinea pig. It should be noted that the guinea pig is an unusual laboratory animal in its long gestational development, approximately 5-95 ------- 65 days. During the latter two thirds of gestation in the guinea pig, the un- born offspring is in a stage of development similar to that which occurs after birth in the mouse and the rat. In itself, a long gestation period does not prevent use of the guinea pig as a model for teratogenesis. But the reader is forewarned that, when reading about teratologic experimental methodology, the administration" of the agent at 40 or 50 days of gestational age in the guinea pig has no comparison in rodent fetuses and that this stage of development occurs postpartum in mice and rats. Studies of the teratogenic potential of conventially induced hyperthermia in almost all the species used show some general changes that can be seen in RF-radiation-treated animals as well. The symptom of hyperthermia, whether induced by RF radiation or otherwise, causes varied anomalies (such as kinked tail, microphthalmia, exencephalia, cleft palate, general edema). This is evidence that hyperthermia is a general teratogen. Perhaps the only case where RF-radiation-induced and conventionally induced teratogenesis is not equal is in the development of smaller brain weights in fetuses heated by non-RF-radiation-heating agents. Except for Shore et aK (1977), perhaps the literature in RF-radiation-induced teratogenesis has not been sufficiently developed to show this effect of decreased brain weight. Conventionally induced hyperthermia, like RF-radiation hyperthermia, appears to affect most species if sufficient energy is applied and proper timing of the agent is obtained. A sufficiency is usually manifested by a rise in maternal rectal temperature, usually over 40 °C, but often in a range of 41 to 43 or even 44 °C. At these temperatures, the teratogenic potential 5-96 ------- of heat applied by conventional means or by RF radiation does not appear to show any significant differences. The authors of one study attempted to differentiate teratogenic effects of microwaves from the more conventional methods of heating (in this case, infrared irradiation). Chernovetz et al^. (1977) exposed pregnant rats to 2450-MHz fields at an SAR of approximately 31 W/kg, or to infrared radiation. The rectal temperatures of the RF- irradiated and the infrared-irradiated groups were similar (approximately 42 °C). The results of the exposures showed little difference in effects due to infrared heating or RF-radiation heating. The symptom of decreased body weight of fetuses is a general develop- mental effect. This response to microwaves is commonly seen in experiments, even those experiments that have not otherwise demonstrated specific morphologic changes (e.g., O'Connor 1980). Body weight changes alone have also been seen in mice fetuses at temperatures < 40 °C (Berman et al_. 1978). Body temperatures reported in the above studies were those of the dam, usually measured in the dam's large intestine. Measurements of fetal tempera- tures after RF-irradiation have not been reported. A report by Morishima et al_. (1975) provides some understanding of the physiologic changes that occur in the fetus as a result of conventionally induced hyperthermia. Morishima et al^. used pregnant baboons and simultaneously monitored temperature and physiologic changes in the dam and the fetus during hyperthermia. The temperatures in these dams were raised to 41 to 42 °C by infrared lamps and warming pads. While the pregnant dam was maintained in an analgesic state using nitrous oxide inhalation anesthesia and intravenous succinylcholine 5-97 ------- chloride infusions, thermistors and catheters were placed at comparable positions in the dam and the fetus. Two groups of animals were used; one group was kept at 38 °C, and the temperature of the other was allowed to elevate gradually to 41 or 42 °C. At the more normal temperature, 38 °C, there was a slight but steady increase (0.5 °C) in the temperature of the fetus over that of the dam. When the temperature was increased to 40 °C in the dam, the fetal temperature rose accordingly and remained 0.75 °C higher than the dam's temperature. The hyperthermia produced increased uterine activity and some acidosis in the dam, and a profound acidosis and hypoxia in the fetus. If conditions comparable to those seen in the baboon are also seen in the mouse, then we might expect the mouse fetus to have temperatures greater than that of its dam. That the teratogenic potential of RF radiation might be dependent only on the deposition of energy as heat was shown experimentally by Rugh and McManaway (1976). They exposed pregnant mice to highly teratogenic (high incidence of fetal death) levels of RF radiation with and without pentobarbital anesthetic. The mice exposed during anesthesia had normal rectal temperatures and normal incidences of fetal death. Therefore, the anesthetic protected against the primary temperature effect of the radiation by reducing "... the body tempera- ture to a degree equivalent to the rise in temperature expected from the conditions used." This is clear evidence that the increased fetal t abnormalities are strongly, if not solely, associated with increased maternal temperature. 5-98 ------- Further evidence for a direct relationship between fetal effects and hyperthermia induced by RF radiation is shown in several reports (Dietzel and Kern 1970; Dietzel et a^. 1972; Dietzel 1975). In these experiments, groups of pregnant rats were subjected to 27-MHz radiation to raise rectal temperature to 42 °C. From these data, the thresholds for malformation rates appear to be between core temperatures of 39.0 °C maintained for 5 min and 40.5 °C for 10 min. Observed teratological effects are summarized in Table 5-9. 5.3.2 Reproductive Efficiency This section discusses aspects of reproduction where the conceptus is not "insulted" directly by the agent, although the fetus may be involved in demon- strating the effect. Generally, reproductive efficiency is the capacity of the dam or sire to effect conception and to bear and rear offspring. Changes in this capacity might be due to alterations in behavior, physiology, or morphology. For example, reproduction efficiency might be affected by changes in maternal cyclical hormone secretions. Tests for reproductive efficiency are not conducted as frequently as those for teratologic effects, usually because the male and female reproductive systems require considerable altera- tion by a toxic agent to cause significant fetal wastage or increased reproductive efficiency. The testes as factors in reproductive efficiency are discussed separately in the next section. 5-99 ------- TABLE 5-9. SUMMARY OF STUDIES CONCERNING TERATOLOGIC EFFECTS OF RF-RADIATION EXPOSURE Exposure Conditions* Effects Species Intensity (mW/co2) Duration (days x «in) SAR (W/kg) 30% survival of pupae D. aelanoaaster 1 X 10 640 Eobryonic LD^q Chicken 200 1 X 12 70 280 1 X 7 98 400 1 X 4 140 Decreased postnatal survival House 1 X 4 104 Teratogenesis House 1 X 2-5 85-112 No change in teratogenesis House 1 X 10 38 Increased postnatal survival House 1 X 10 38 Maternal lethality, resorptions, Rat 1 X 20 31 decreased fetal weight Decreased fetal weight House 28 12 x 100 22 No change post-hatching, hatchability, Japanese quail 1 X 1440 14 heoograo, body or organ weights No change Rat 1 X 20 14 No change Rat 40 1 X 120 10 No change House 3.4-14 12 x 100 2-8 Teratogenesis House 12 4 No change Rat 5 Hany x 240 0 7-4 7 No change Rat 28 12 x 100 4 2 No change Squirrel ¦onkey Hany x 180 3 4 No change Rat 10-35 Hany 1-3.5 No change Rat 5 19 x 1200 2.5 Decreased body and brain weight Rat 10 16 x 300 2 2 " Frequency used was 2450 MHz, except for Jensh (914, 2450, and 6000 Wz), and Johnson 1978 (918 MrizJ ------- We have previously discussed the effects of 2450-MHz RF radiation on D. melanogaster reported by Pay et al.. (1978). The fruit flies that were the subjects of that study were irradiated before production of ova. The SAR used in the study ranged from 400 to 800 W/kg. Reproductive efficiency, in this case the production of eggs, was significantly reduced in both conventionally heated and RF-irradiated females as compared with the shams. But, there appeared to be no significant difference in the production of ova from females exposed to RF radiation compared to those exposed to the conventional source of thermal energy. The egg production of chickens can also be used to provide an index of reproductive efficiency or reproductive wastage. A single, intense exposure of day-old chicks was made in the experiment reported by Davidson et ajL (1976). Described in one of four experiments is the reproductive efficiency of 28-week-old hens that had been irradiated as chicks on day 1 of age for 2 4.5 s at a power density of 800 W/crrl in a multimodal cavity. The estimated SAR was 2770 W/kg. No differences appeared between the control group and the irradiated group in the production of eggs during 100 days of laying. The authors also stated that a dose rate of 2500 W/kg for 9 s at 2450 MHz is a lethal dose in day-old chicks and that 42 percent of the chicks dosed for 6 s at 2810 W/kg died. The animals that survived were unconscious for up to 5 min. The day-old chicks had approximately the same sublethal SAR (2770 W/kg for 4.5 s) used in the examination for latent reproductive effects, but there were no deaths that could be directly attributed to the RF-radiation exposure. 5-101 ------- From these experiments, it is concluded that the massive doses of this experi- mental regimen left no latent alterations in reproductive efficiency. Rugh et a^. (1975) reported an experiment in which they examined dif- ferences in average lethal dose of RF radiation (2450 MHz) during the estrous cycle of CF1 mice. The weights of the mice ranged from 29 to 31 g. There was a prior 20-min acclimation to the waveguide exposure chamber, and environ- mental conditions were 23.5 °C, 50 percent relative humidity, and an airflow of 38 L/min (0.38 km/h). There was a significant decrease in the average lethal dose for females in estrus as compared to the average lethal dose for those in diestrus (P < 0.01). The forward power was 8.24 W. This experiment shows that changes brought about during the reproductive cycle can affect radiosensitivity. In summary, it appears that the efficiency of the female reproductive system is not easily altered. Only irradiation at extremely high dose rates has made changes in reproductive patterns. Small alterations from normal values, though detectable by modern scientific and statistical methodology, do not seem able to sway the general outcome of the reproductive cycle. Observed reproductive effects are summarized in Table 5-10. 5-102 ------- TABLE 5-10. SUMMARY OF STUDIES CONCERNING REPRODUCTIVE EFFECTS OF RF-RADIATION EXPOSURE Exposure Conditions* Effects Species Duration (days x min) SAR (W/kg) Reference Decreased ova production L. melanogaster 400-800 Pay et al.. (1978) No change in egg production Gallus 1 x 0.08 2770 Davidson et al. (1976) Lethality changes with estrous cycle Mouse Rugh et aK (1975) Frequency was 2450 MHz in all cases. 5.3.3 Testes Considerable scientific work has been done to determine effects of RF radiation on the testes. The testes are so placed anatomically that they can be conveniently irradiated without irradiating the remainder of the body. Tests for testicular function are also conducted easily. Quantitative changes in sperm concentration can be conveniently assessed using repeatable laboratory techniques. It is known that when the mammalian testes, which have a normal temperature of 33 to 35 °C, are heated to temperatures approaching abdominal temperature (37-38 °C) sterility can occur. Sterility consequent to high testi- cular temperatures can be viewed as a purposeful contraceptive agent or as an unintended toxic agent. 5-103 ------- A report by Muraca et al. (1976) describes the results of irradiation of rat testicles with 2450-MHz radiation. Each animal was anesthetized, placed in an anechoic chamber, then irradiated in a free field at a power density of 80 mW/cm^. The SAR for adult rats was 16 W/kg (Durney et aK 1978). But, the philosophical aim of this experiment was to examine the effects of RF radiation on the testes, and it may be appropriate to use the SAR of the testes alone as a more adequate expression of absorption. For that reason, we have also esti- mated (Durney et aK 1978, Figure 41, model of a quail egg) that the average SAR to the testes might have been 60 W/kg. In the experiment, the authors irradiated rat testes to produce increases of intratesticular temperature to 36, 38, 40, and 42 °C; the testes were then maintained at these temperatures. The technique of irradiation included implantation of a thermistor into one testis of the anesthetized male, with the temperature of that testis acting as the control of the on-off-on sequence of the RF irradiator. The authors irradiated both testicles, so that temperatures to which the testes were brought were within ±0.5 °C of the levels mentioned above. The duration of each irradiation at a power density of 80 mW/cm varied from 10 to > 70 min. The animals were irradiated once or repeatedly during 5 consecutive days. Five days after the treatment ended, the animals were killed and their bodies were infused with solutions to preserve the testis that was not "invaded" by the thermistor. The microscopic appearance of the testicular tissue was categorized as follows: apparently normal; appearance of early inflammatory or degenerative changes in the spermatogenic epithelium without well developed necrosis; or severe degeneration of the majority of 5-104 ------- seminiferous tubules with coagulated cellular elements. This study also included a report comparing the effects of immersing the scrotum in a water bath. 2 When male rats were exposed once (80 mW/cm ; 2450 MHz; 10 to 73 min; temperature of the testis maintained at 40 °C), there appeared to be no signi- ficant change in the incidence of apparently abnormal testicular tissue. After a single exposure of 10 min, during which the temperature of the testis was allowed to reach 42 °C, the number of animals that had some abnormal changes tripled. Multiple (five) exposures, even where the temperature reached only 36 °C for 60 min, caused all the testes to have some changes in the spermatogenic epithelium. When temperatures were allowed to reach 40 °C in the testes for repeated short periods (10 to 27 min, five times), severe degeneration in the spermatic tubules was seen in all the testicular samples. This study points to two important factors of irradiation of the testes with 2450-MHz RF radiation: that a minimum temperature (> 40 °C) must be reached in an acute exposure, and that repetitive treatments are much more effective. These two factors interplay, as well, so that an acute temperature excursion, even as high as 40 °C for over 70 min, is not as deleterious as a lower temperature (36 °C) reached repeatedly (five times). Fahim et aK (1975) conducted an experiment to compare the contraceptive capability of microwaves with that of other methods of heating. The authors used male Sprague-Dawley rats (Holtzman strain) and irradiated them with 2450-MHz radiation (from a diathermy unit of maximum output of 100 W) during 5-105 ------- pentobarbital anesthesia. The applicator of the diathermy unit was 7.5 cm from the testicles of the rat and provided near-field exposures that do not allow a good estimate of the associated SAR. By varying the power output of the unit and by varying the exposure time from 1 to 15 min, the authors developed four subgroups of animals: those in which testicular temperature reached 65 °C for 5 min, 45 °C for 15 min, or 39 °C for 1 or 5 min. The males were allowed to mate with normal females 24 h after treatment, and every 5 days thereafter until pregnancy was observed in the females. "The endpoint for fertility was the amount of time required for every surviving male in the treatment group to impregnate a female." Later, the sexual organs were weighed, and histological examinations were made of the testes and secondary sex organs. Raising the temperature of the testes to 65 °C for 5 min or to 45 °C for 15 min caused complete infertility in the males for 10 months. When these testes were examined histologically, there was no observable spermatogenesis. When the temperature was raised to only 39 °C, 70 percent of the males retained normal breeding capability, whereas the remaining 30 percent recov- ered their fertility within 2 wk. Histological sections indicated normal spermatogenesis. There appeared to be no differences in testicular weights or secondary sex-organ weights, even in the group (45 °C, 15 min) that was sterile 10 months after the RF-radiation exposure had ended. This experiment demonstrates that a temperature of 45 °C caused by RF- radiation exposure must be attained in the rat testes to produce permanent 5-106 ------- sterility. Temperatures that ranged to 39 °C, somewhat above rectal temper- ature, for only a short period (5 min) were not effective and produced only temporary, if any, infertility. Along with the permanent sterility induced by the high temperatures (minimum of 45 °C), there was damage to the tubules and a lack of spermatogenesis. Sterility can occur in rats following RF-radiation exposure. At high temperatures, epithelial and tubular structures of the testes can be damaged permanently. At lower temperatures (approximately normal body temperature) for short durations, the temperature in the testicles due to RF-radiation exposure resulted in temporary sterility. This last condition implies that the spermatogenic cells themselves are not damaged permanently. The rat's spermatogenic cycle has a duration of about 10 wk, i.e., it takes about 70 days from the first meiotic divisions of sperm generation in the germinal epithelium until the time when the sperm are fully matured, in place, and ready for ejaculation. During the last 2 wk of spermatogenesis, sperm are located in the tubules of the testes and in the epididymis and require very little additional maturation. There is a constant stream of maturing sperm, because various seminiferous tubules are in different stages of this long cycle. The most mature stages of the spermatogenic cycle are most sensitive to heat (Van Demark and Free 1970). The permanent (10-month) sterility seen in the experiment above is due not only to loss of the most mature and most heat-sensitive stages, but also to the loss of sperm at all stages, including those stages involving the germinal epithelium. The 2-wk temporary infer- tility seen in the less affected males was caused by the loss of only the most mature and most heat-sensitive sperm. 5-107 ------- A study to demonstrate an alteration in functional fertility was conducted by Berman et al^. (1980). The authors' report concentrated on chronic exposures of rats to 2450-MHz radiation in an anechoic chamber. There were three separate experiments: 2 (1) Exposure at a power density of 5 mW/cm for 4 h/day, daily, from the 6th day of gestation through 90 days of age post- partum; p (2) Exposure at a power density of 10 mW/cm for 5 h/day for 5 days, beginning on the 90th day of age postpartum; and 2 (3) Exposure at 28 mW/cm for 4 h/day, 5 days a week, for 4 continuous weeks, beginning on the 90th day of age postpartum. The main purpose of the experiments was to evaluate potential mutagenic effects of RF radiation on the germ cells of the male rat. Male rats were bred to untreated female rats shortly after the end of the treatment period. No mutagenic effect was demonstrated. The breeding data of the assays used in these experiments were used to evaluate the effects on the spermatogenic func- tion in rats. In the first experiment, the animals were exposed daily from the 6th day 2 of gestation to 90 days of age, at a power density of 5 mW/cm . In this experimental group, the SAR varied inversely with the growth of the animals, i.e., from approximately 4.5 W/kg in neonates to approximately 0.9 W/kg at 90 days of age. Twin-well calorimetry was not used in the other two experiments, p and SAR's are estimated at 2 W/kg at a power density of 10 mW/cm , and 5.5 W/kg at 28 mW/cm^. 5-108 ------- In this study, temperatures of the testes of rats exposed up to 90 min to 2 2450-MHz fields at 28 mW/cm showed an increase from a pre-exposure level of 34 °C to almost 38 °C after exposure. Simultaneously recorded rectal tempera- tures were ~ 3 °C higher than temperatures of the testes during the entire period. Sham-irradiated animals had testicular temperatures ranging from almost 32 °C to approximately 35 °C. The exposure at 28 mW/cm during the 90-min period produced a temperature in the testis equivalent to the normal rectal temperature. 2 Chronic exposure at 5 mW/cm from the 6th day of gestation through 90 days of age appeared to have no effect on the reproductive efficiency of the male rats when bred with normal females. Also, exposure at a power density of 10 mW/cm (SAR ~ 2 W/kg) at 90 days of age for a period of 5 days had no effect on the reproductive efficiency of the males. It was only the most severe regimen, the exposure of adult male rats at 28 mW/cm for ~ 4 wk (estimated SAR ^5.5 W/kg), that caused any alteration in reproductive function. This exposure produced a severe decrease in the reproductive ability of the males; only 50 percent of the females that were available to the males for breeding became pregnant during the week immediately following the exposure period. The breeding returned to normal beginning at the 3rd week after irradiation (the next period of test breeding). The animals bred normally thereafter. No examination was made of the testes 2 of animals exposed at 28 mW/cm . The temperatures reached in the testes at 2 the highest power density (28 mW/cm ) were similar to those reported by Fahim et al_. (1975) where histological changes were seen. 5-109 ------- Cairnie et aL (1980) studied the effects in mouse testes after a range of whole-body exposures of up to 30 days for 16 h/day at 2450-MHz and power 2 densities up to 36 mW/cm (average whole-body SAR = 7 W/kg; average testicular SAR = 14 W/kg). These exposures caused no measurable increase in testicular temperature. No changes were seen in the number of dead testicular cells, the number of epididymal sperm, or in the percentage of abnormal sperm after exposure, even up to 8 wk after exposure. Four strains of mice were used, and no strain susceptibility was observed. Varma and Traboulay (1975) were able to cause severe scrotal and testi- cular burns in anesthetized mice after exposure for 20 min to 1.7 GHz and 200 2 mW/cm (whole-body SAR estimated at 80 W/kg). Exposures of < 100 min to 10 2 mW/cm at this frequency caused little or no damage. But a 100-min exposure caused general testicular damage without changing Sertoli and interstitial 2 cells. No severe changes occurred at 3 GHz, 50 mW/cm , for 20 min (SAR = 20 W/kg). In the studies reported by Ely et al_. (1964), 2880-MHz radiation was used to irradiate only the testicular area in groups of dogs. The animals were anesthetized during exposures while temperatures were measured in the testis. Irradiation of the testicular area continued until a peak temperature was reached; the RF radiation was then turned off manually so that the testes could cool, at which point the RF radiation was turned on once again. This cycling of exposure was completed many times so that the animal's testes could be kept at a "steady" temperature for a considerable length of time. The 5-110 ------- normal temperature of the testes of the dogs was ~ 33 °C, 5 °C lower than that of the rectum. 2 According to this report, exposure of the testes at 20 mW/cm caused an 2 2 elevated testicular temperature of 36 °C at 33 mW/cm , 38 °C at 45 mW/cm and up to 40 °C. As 38 °C is the approximate body temperature in the dog, bringing the testes to body temperature can expect to produce sterility if the elevation 2 is sufficiently chronic. It appears then, that 45 mW/cm is required to produce this type of effect. There is a large body of literature on the fertility effects caused by temperature increases, such as when the testicles are clothed to prevent the normal thermal radiation. Even high environmental temperatures, when sustained, can produce infertility in male rats. One example of this is a report (Pucak et al_. 1977) describing deaths in a large rat-production colony in which room temperatures accidentally reached as high as 31.6 °C for 2 days and as high as 37.7 °C in individual cages. Under these conditions, approximately 3,000 of 14,000 Sprague-Dawley rats died from heat prostration. When examined 18 days after the incident, 25 percent of the surviving males showed bilateral atrophy of the testicles, which were approximately half the normal size. Histological examination showed atrophy of the spermatic tubules and failure of spermato- genesis. The proportion of affected testicles ranged from 50 to 75 percent. Five weeks after the incident the animals with small testes were still sterile. In comparison with the temperatures seen in this study, raising the tempera- tures of testes to 42 °C by RF-radiation exposure appears to be an extreme / experimental regimen. Observed testes effects are summarized in Table 5-11. 5-111 ------- TABLE 5-11. SUMMARY OF STUDIES CONCERNING REPRODUCTIVE EFFECTS OF RF-RADIATION EXPOSURE IN THE RAT Exposure Conditions Effects Intensity Duration SAR Reference (days x min) (W/kg) Spermatogenic tissue 80 1 x 10-73 16 Muraca et al_. (1976) abnormal 5 x 10-73 16 No change 5 many x 240 0.9-4.5 Berman et aK (1980) No change 10 5 x 360 2 Temporary sterility 28 20 x 240 5.5 ^Frequency was 2450 MHz in all cases. 5.3.4 Unresolved Questions 5.3.4.1 Teratology-- A threshold of teratogenic effects due to RF radiation has not been determined. The measurement of threshold will have to include the entire range of teratogenic effects: lethality, anomaly production, decreased body weight in fetuses, and alteration of postnatal function. Even at this stage of development of the data relating to RF-radiation-induced teratogenesis, a picture is appearing in which the degree of response is dependent on the level of whole-body SAR, the duration of exposure, the timeliness of the exposure, and the species, all of which complicate the threshold determination. It is not yet clear if threshold can be based on a finite whole-body SAR for all species or is some adjustment or proportion will be necessary for species size, metabolic rate, thermoregulatory capacity, etc. So far, the only 5-112 ------- physiologic variable that can be supportably associated with teratogenic effects is the colonic temperature of the dam during or at the end of exposure. The mouse and rat, two extensively studied species, show a minimum temperature of approximately 40 °C in the dam is associated with teratologic symptoms. However, whole-body SAR's required to cause temperature excursions like this are much higher in the mouse than in the rat. Therefore, an adequate relationship of teratogenesis to SAR alone is not apparent. The maternal colonic temperature, then, is the only available indicator of a threshold for teratogenesis in mammals., Published reports that meet the criteria for consideration in this document have limited their examination of the fetal results of gestational exposure to a gross morphological change or one that might be seen under low magnification (15X). There has been no organized attempt to examine in histologic detail fetuses that have been irradiated jfrv utero. Authors of one study (McRee et a^. 1980b), however, made a detailed examination of embryonic hearts but could not demonstrate changes in morphologic, ultrastructural, or enzymatic activity. The subjects were Japanese quail which had been exposed daily for 8 days to 2450-MHz radiation at SAR's of 4 and 16 W/kg. There are classifications of fetal changes that represent no gross structural deficit, but nevertheless represent some variation of structure. An example of a variation that may not be considered by all teratologists as a "deficit" is a small but normal fetus. However, if the incidence of this variation is consistently increased because of the application of a toxic agent, it could be considered an expression of embryotoxicity. The decreased 5-113 ------- body weight so often seen in offspring exposed to RF radiation (Berman et al. 1978; Chernovetz et aj[. 1977, 1979; Rugh 1976a) might otherwise be considered as a lesser category of "structural variation without deficit" were it not that decreased weight is so consistent in these experiments. Whether this decreased fetal weight is temporary (i.e., only a delayed growth that will disappear in the neonatal stage) or permanent (i.e., a stunting of the fetus that will persist) has not yet been scientifically resolved. There is one aspect of the literature on the teratogenic potential of RF radiation that deserves further discussion. More than half the papers in this document on teratogenesis report experiments with the rat. Chernovetz et al. (1977) raise an interesting point about using the rat in determining the teratogenic potential of RF radiation. The authors argue that because levels of microwave exposure which are associated with high mortality rates of dams do not also produce fetal structural abnormalities in the rat, "... that the teratogenic threshold of microwave radiation is higher than the dam's threshold of mortality," and hence "... that maternal mortality is more probable than malformation of the fetus, irrespective of the dose." The problem of using animal models in determining teratogenic potential in humans is that there is no assurance that any of these models is a real estimate of effects in humans. The concepts supporting the use of animals as models of human beings require that the response be demonstrated in a number of species, so that the resultant generality can be more confidently extended to humans. Therefore, we seek among species some generality of effects of 5-114 ------- microwaves on the fetus. Mice and rats are the two laboratory animals upon which rest almost all of the data of RF-radiation-induced teratogenesis. Any difference between the two species in their teratogenic response to RF radiation, therefore, becomes important. 5.3.4.2 Reproductive Efficiency and Testes-- The testes contain, besides sperm-producing tissues, interstitial cells that secrete testosterone, the male hormone. Gunn et cH. (1961) described the effect of 24-GHz radiation on the morphology and function of the interstitial cells in rats. They exposed rats once at this frequency for a period of 5 min o at a power density of 250 mW/cm , which caused the temperature to rise to 41 °C in the testes. As a result, there were scrotal burns, severe edema, and spermatic tubular degeneration, but no interstitial-cell pathology. The testosterone secreted by the interstitial cells of the testes regulates zinc uptake by the dorsolateral prostate. When Gunn examined zinc uptake in RF-irradiated rats that had no interstitial cell pathology, he found a decreased uptake. Gunn related the decreased function of this secondary sex organ to decreased testosterone secretion. The effects seen by Gunn were produced at a frequency of 24 GHz. This is such a short wavelength that, theoretically, there would not be any sig- nificant penetration to affect the dorsolateral prostate directly. The tests using the dorsolateral prostate as an indicator of sexual function in the rat have been commonly used by Gunn and others in other types of experimental situations. 5-115 ------- It is not known why and how the dorsolateral prostate lost its capacity to function normally when there was no observable change in the interstitial tissues, the other testicular damage appeared to be minimal, and the energy could not reach the prostate. This appears to deserve additional scientific attention and exploration. The literature we have cited does not lend itself to extrapolation of effects of RF radiation that cause only small increases in the temperature of the testes. The lack of data on RF-radiation effects at lower power densities (which cause lower testicular elevations of temperature than have been cited in the articles above) is especially important in light of suggestions that thermal energy, in the form of absorbed RF radiation, can be used as a contra- ceptive in men (Medical World News 1974; Arehart-Treichel 1974). One report by Rugh (1976a) contains data on survival in RF-radiation fields that are different in males and females. In this study, mice of both sexes and of three ages (weanlings, young mature, and aged) were irradiated with 2450-MHz RF radiation until dead. Of the three variables (age, sex, body weight) examined for their contribution to RF-radiation lethality, Rugh found that "The overall conclusion would be that no matter at what age ... absorbed dose to death ... is different for the two sexes. Females are slightly more radiosensitive ...." These results are not explained easily. Although they are not apparently related to sexual function, they were included here on the basis of sexual differences. 5-116 ------- 5.4 NERVOUS SYSTEM Ernest N. Albert The nervous system in higher forms of life is the result of an evolution- ary development in response to the environment, and has reached a very high stage of development in man. The body experiences many environmental stimuli and responds to these conditions either voluntarily or involuntarily. These responses are mediated by the nervous system. Briefly, specialized receptors enable the human being to perceive visual, auditory, thermal, mechanical, chemical, electrical, pain, and pressure stimuli, and send the sensory information to the central nervous system (CNS, i.e., brain and spinal cord) via the peripheral nervous system (PNS). In addition, the brain is the site of mental operations such as mathematical integration, thought, memory, will, anger, and love. The brain integrates all of the sen- sory information and provides a response appropriate to maintaining homeostasis within the entire organism. These functions of the nervous system are carried out by its two major divisions: (1) the somatic or voluntary, and (2) the autonomic or involuntary. The somatic or voluntary division is primarily involved in controlling con- scious activities. On the other hand, the autonomic or involuntary division of the nervous system is more involved with those functions of the body that an individual cannot control (e.g., regulation of blood pressure and heart rate). When the body is suddenly stressed, the heart begins to beat faster; respiration rate and blood pressure increase involuntarily. 5-117 ------- The basic structural units of all parts of the nervous system consist of cells (neurons and glia) and their processes (axons, dendrites, and termi- nals). There are many types of neurons and glial cells as well as many modifications of their axons and dendrites. These cells and their processes communicate with each other, as well as with muscles and glands of the body, via chemical and electrical means in integrating information and maintaining homeostasis. When an environmental condition fluctuates within acceptable limits, the nervous system responds by compensating within the body's physiological limits with no apparent harm. However, extreme environmental conditions stress the nervous system beyond these limits, resulting in abnormal physiological func- tions. Such abnormal functioning could result from exposure to RF radiation. Therefore, it is important to investigate and evaluate the interactions and thresholds of interactions between RF radiation and the nervous system. A critical review of the literature in this area permits the following general statements to be made about the effects of RF radiation: (a) Morphology. There is ample evidence that acute and chronic exposure of animals at sufficiently high intensities of RF radiation (CW or PW) produces morphological alterations in the nervous system. The morphological changes are qualitatively similar after acute and chronic exposure (at high and low densities), but quantitatively there are greater alterations in neuronal structure at higher (> 10 mW/cm2) power densities and after chronic exposures. In general, PW radiation produces greater changes than CW radiation. Almost all reported changes due to RF-radiation exposure are reversible. (b) Blood-brain barrier (BBB). There is an ongoing controversy whether RF radiation at low power densities alters the BBB of experimental animals. At present, no definitive conclusions 5-118 ------- are possible concerning RF-radiation effects on the BBB at low power densities (< 10 mW/cm2). (c) Pharmacologic effects on CNS. PW radiation appears to have a potentiating effect on drugs that affect nervous system function. Whether RF radiation also has an inhibiting effect on neuropharmacologic drugs is not certain. (d) Effect on neurotransmitters. There are no data on this subject at low power densities (< 10 mW/cm2). However, at high power densities and SAR's (> 24 W/kg) there appears to be an effect on the release of neurotransmitters after exposure. Whether there is an increased or a decreased release depends on the specific neurotransmitter. 5.4.1 Morphologic Observations The nature of morphologic changes in the nervous system of exposed animals depends on the frequency, power density, duration, and modulation characteristics (e.g., PW or CW) of the radiation. Gordon (1970) and Tolgskaya and Gordon (1973) report that severe damage to the brain occurs when 2 RF radiation of various frequencies at high power densities (> 40 mW/cm ), producing measurable AT's, is used. These changes consist of hemorrhages, edema, and vacuolation of neurons after a 40-min exposure to 3000- or 10,000-MHz PW or CW radiation at 40 to 100 mW/cm^. At 20 mW/cm^ similar but less severe effects were observed. Further, 3000-MHz radiation produced more marked changes than 10,000 MHz of equal power density. However, Austin and Horvath (1954) did not observe similar changes in brains of rats that became convulsive and hyperthermic (brain temperature ~ 43 °C) due to a single exposure to 2450-MHz radiation. These authors observed only mild pyknosis and hyperemia in some areas of the brain. Albert and DeSantis (1975, 1976) did not observe hemorrhage, gliosis, or focal necrosis in Chinese hamsters exposed 5-119 ------- to 2450-MHz (CW) radiation at 50 mW/cm^ (SAR estimated at 17.5 W/kg) for 30 to 120 min, but they did observe swollen neurons with frothy cytoplasm in the hypothalamic and thalamic regions of the brain. Such observations were not seen in the cerebellum, pons, or spinal cord. Histologic change in the brains of rats have also been reported at low power densities (< 10 mW/cm ) of 3000-MHz microwaves after multiple 30-min exposures. The histological alter- ations include cytoplasmic vacuolation of neurons, axonal swelling and beading, and swelling in and decreased numbers of dendritic spines (Tolgskaya and Gordon 1973; Albert and DeSantis 1975; Gordon 1970). All of the above- mentioned changes were reversible after 3 to 4 weeks. Baranski (1972b) noted that exposure of guinea pigs and rabbits to 3000-MHz (PW and CW) radiation at power densities of 3.5 to 25 mW/cm^ (SAR estimated at 0.4 to 2.5 W/kg) resulted in myelin degeneration and metabolic alterations in glial cells. Qualitatively, the morphological effects on the CNS are similar at 10 to 50 mW/cm power densities, but quantitatively they are greater at higher power densities. Most scientists agree that irradiation at these power densities can raise the body temperature. Soviet scientists have reported similar 2 morphological changes at 1 mW/cm after chronic irradiation, and they do not 2 consider these alterations at 1 mW/cm of thermogenic origin. Tolgskaya and Gordon (1973), Baranski (1972b), and Baranski and Edelwejn (1968) further state that morphological effects are more marked after PW or chronic exposure than after CW or acute exposure. Most Eastern European studies claim full recovery by irradiated animals in 1 to 3 weeks after exposure at less than 10 mW/cm . Albert and DeSantis (1975, 1976) found continued existence of 5-120 ------- neuronal cytopathology in animals 2 weeks after exposure. Perhaps a longer recovery period in the latter study would have shown complete reversibility. Switzer and Mitchell (1977) described the existence of myelin figures in brain dendrites of rats exposed to 2450-MHz (CW) fields for 550 h over a 110-day period (average SAR =2.3 W/kg) and allowed 6 weeks for recovery. These authors did not observe any gliosis, perivascular edema, or synaptic pathology. However, irradiated rats in this study exhibited marked disruption of discriminative performances during exposure. Takashima et al_. (1979) exposed male rabbits to 1- to 30-MHz fields amplitude modulated at ~ 15 Hz or 60 Hz during 2- to 3-h single exposures or during 4 to 6 weeks of chronic exposure. The electric field strengths ranged from 60 to 500 V /m. Acute and chronic EE6 recordings were obtained from rms animals under sodium pentobarbital anesthesia. There was no temperature rise in the exposed animals. The EEG recordings after acute exposures at field strengths of 60 to 500 Vrms/m or after chronic exposures at strengths up to 70 V /m showed no differences between control and experimental animals, rms However, chronic irradiation at higher field strengths was associated with abnormal patterns. These consisted of bursts of high amplitude spindles at 90 \L /m as well as suppression of activity at 500 V„ /m. All brain activi- rms rms ties returned to normal a few hours after irradiation. The results in this study appear to be free of electrode artifacts because chronic exposures were made without electrodes, and all recordings were made when the fields were switched off. The effects of anesthesia are not clear, and the natural fluctuations of brain activity also complicate interpretation of the results. 5-121 ------- Although the occurrence of high-amplitude spindles in irradiated animals in this study is similar to that described by Bawin et afL (1973), it is difficult to compare the two results because, in the latter study, chronically implanted electrodes may have interfered with the imposed fields. Takashima et a^. (1979) confirmed the existence of artifacts during irradiation when implanted electrodes were used. We can conclude that RF radiation of centimeter wavelengths causes morphological changes in the CNS of experimental animals following acute or 2 chronic exposures at low power densities (< 10 mW/cm ). Note that, in most of the low-exposure studies, total recovery was observed after 3 to 4 weeks. 5.4.2 Blood-Brain Barrier Studies In the past few years, contradictory reports have been published concern- ing the effects of RF radiation on the permeability of the BBB. (For reviews, see Albert 1979a and Justesen 1980.) Some of these reports indicate that direct RF-radiation effects in experimental animals might result in increased perme- ability of the BBB (Frey et aL 1975; Oscar and Hawkins 1977; Albert and DeSantis 1976; Albert 1979b). Other reports indicate that increased perme- ability might be mediated by hyperthermia induced by intense RF fields (Sutton and Carroll 1979; Merritt et aL 1978; Lin and Lin 1980). Preston et al. (1979) and Preston and Prefontaine (1980) reported negative findings at lower power densities. Some of these discrepancies may be attributed to differences in techniques employed to assess changes of permeability. These methodologies consist of gross examination of brain slices, single-passage isotopic tracers, 5-122 ------- fluorometric measurements, arid electron microscopic tracers. All of these techniques have some inherent shortcomings, either in quantitation or sensi- tivity. Thus, unless the changes of permeability are diffuse and significant, positive results may not be readily apparent. Further complications are reports that average power density, peak power, and pulse width may be important parameters affecting the BBB permeability (Frey et al_. 1975, Oscar and Hawkins 1977). Therefore, one must consider the limitations of the tech- niques and the exposure parameters before reaching conclusions regarding effects of RF-radiation on the BBB. Frey et aK (1975) were the first authors in the United States to report a permeability increase in the BBB of the rat after RF-radiation exposure. 2 They observed that a 30-min 1200-MHz exposure (CW) at 2.4 mW/cm (SAR esti- mated at 1.0 W/kg) resulted in a statistically significant increase in fluo- rescein in brain slices of experimental animals over controls. Most of the fluorescein appeared to be concentrated in the vicinity of the lateral and third ventricles. Some dye also was detected in the metencephalon. These authors also reported similar but heightened alterations in the permeability 2 of the BBB when rats were irradiated with PW radiation at 2.1 mW/cm peak 2 and 0.2 mW/cm average power density (SAR estimated at 0.8 W/kg). Their results also indicated that PW radiation was more effective in altering brain permeability than CW—once again, the theme of PW radiation producing greater biological effects than CW surfaces in these studies. Merritt et jfL (1978) were unable to replicate the fluorescein studies of Frey et al^. (1975). However, increased brain permeation of fluorescein-albumin 5-123 ------- (mol. wt. 60,000) was produced in rats heated to 40 °C by hot air or by RF radiation. They concluded that hyperthermia per se, and not field-specific effects of RF radiation, is the essential determinant of increased perme- ability. Using sodium fluorescein and Evan's blue, Lin and Lin (1980) also found no change in BBB permeation after a single 20-min focal exposure within the rat head at 0.5 to 1000 mW/cm^ (local SAR's ranged from 0.04 to 80 W/kg) at 2450 MHz (PW). Albert (1977, 1979b) and Albert and Kerns (1981), using electron microscopic tracer methodology, followed the movement of horseradish peroxidase (mol. wt. 40,000) in rat and Chinese hamster brains after 2450-MHz (CW) irradiation in the far field at 10 mW/cm^ (SAR estimated at 0.9 to 2.0 W/kg). The authors reported focal areas of increased permeability in brains of 35 per- cent of the irradiated animals in contrast with 10 percent in controls. Areas with increased permeability were seen with greater frequency in the thalamus, hypothalamus, medulla, and cerebellum than in the cortex or hippocampus. Using the same parameters of exposure and species as above, Albert (1979b) and Albert and Kerns (1981) showed that within 2 to 4 h after irradiation, there was no evidence of increased BBB permeability, thus demonstrating complete reversi- bility of the RF-radiation effect. It was also shown that the increased perme- ability of the BBB appeared to be due to increased pinocytotic transport of the tracer rather than to opening of the endothelial tight junctions (Albert and Kerns 1981). Sutton and Carroll (1979) observed increased permeability of the BBB to horseradish peroxidase when brain temperatures of rats were elevated to 40 to 45 °C for a few minutes by 2450-MHz (CW) microwaves. A significant observation was that the integrity of the BBB and survival times at 45 °C 5-124 ------- (cerebral temperature) could be prolonged if core temperature was maintained at 30 °C. They hypothesized that perfusion of vessels with cool blood had a protective effect. This study indicates that severe hyperthermia induced by RF radiation might result in increased permeability of tracer proteins in rat brains. In a physiological study using radioisotope tracers, Oscar and Hawkins (1977) exposed rats to 1.3-GHz (CW or PW) radiation for 20 min. Using the 2 technique of Oldendorf (1970), they found after CW irradiation at 1 mW/cm (SAR estimated at 0.4 W/kg) that there was a significantly greater uptake of mannitol (mol. wt. 182) and inulin (mol. wt. 5000), but not dextran (mol. wt. 60,000), in brains of exposed animals. Similar, but greater, uptake of these 2 compounds was observed after PW irradiation (average power density 0.3 mW/cm ; SAR estimated at 0.1 W/kg) than after CW irradiation. Uptake of mannitol by the brain was quite dependent upon power density, pulse width, and number of pulses per second. Merritt et aK (1978), who also used the Oldendorf tech- nique, reported that there was no significant change in uptake of mannitol or inulin in rats exposed to RF radiation under conditions similar to those used by Oscar and Hawkins (1977). Preston et al_. (1979), who also used the Oldendorf technique, exposed rats to 2450-MHz (CW) fields for 30 min at 0.1 to 30 mW/cm^ (SAR estimated at 0.02 to 6 W/kg); no change in uptake of mannitol in the brain was found. They also speculated that the changes reported by Oscar and Hawkins (1977) may have been due to changes in blood flow. Later, Oscar et a^. (1981) measured the blood flow in several brain regions during exposure to 2800-MHz (PW) fields 5-125 ------- o at 15 mW/cm (average) for 5 to 60 min and found increased local blood flow. They then suggested that previously reported BBB-permeability changes (Oscar and Hawkins 1977) may be less in magnitude than originally indicated. A recent report by Preston and Prefontaine (1980) described studies on BBB permeability in rats exposed to 2450-MHz (CW) radiation both in the near and the far field. The far-field exposure took place in an anechoic chamber at 1 or 10 mW/cm^ (SAR estimated at 0.2 to 2.0 W/kg) for 30 min. In the second study, a microwave applicator was placed on the rat head for a near-field ex- posure. The exposure consisted of a single 25-min irradiation (SAR's at 0.08 to 1.6 W/kg). In this second study the exposure took place after the tracer ^C-sucrose was injected into the animal so that BBB function during irradia- tion could be examined. No change in permeation was found in either study. 5.4.3 Pharmacological Effects RF radiation has been reported to alter effects of drugs that influence CNS functions. Baranski and Edelwejn (1968) noted altered EEG tracings and decreased tolerance to Cardiasol, a CNS stimulant, in persons working in micro- wave fields. To better understand their observations of humans, they conducted animal experiments. In their investigations on rabbits, these authors found that administration of Phenactil, a depressant of cortical activity, followed by 3000-MHz (PW) irradiation at 20 mW/cm^ (SAR estimated at 3.0 W/kg) for 20 min, resulted in desynchronization of the EEG; i.e., the synchronizing of the EEG by Phenactil was reversed by irradiation, which indicates a stimu- latory effect by microwaves on the brain-stem reticular formation. Under the 5-126 ------- same exposure conditions, these authors demonstrated that half the dosage of Cardiasol was required to obtain the same EEG reaction that was observed in control animals without previous irradiation; the authors concluded that microwaves potentiate the effects of Cardiasol. Chronic exposure of rabbits to 3000-MHz (PW) radiation at 7 mW/cm^ (SAR estimated at 1.0 W/kg), 3 h/day for a total of 70 to 80 h resulted in violent convulsions after injection of the same dose of Cardiasol (3 mg/kg) as in controls, thus indicating that tolerance to Cardiasol was decidedly lower in chronically exposed animals than in controls. Chronic exposure also resulted in desynchronization and high- amplitude recording potentials in the EEG. In this study, thermal effects of microwaves were considered unlikely, and the authors suggested that microwaves may exert a stimulating effect on specific areas of the CNS. It should be noted that the EEG records were obtained with screw electrodes implanted into the skull of the rabbit and that the pulse modulation characteristics were not specified. Servantie et al_. (1973) investigated the combined effects of microwave radiation and Pentetrazol, a convulsant drug. Mice were subjected to 3000-MHz (PW) radiation (maximum peak power = 600 kW and maximum mean power = 350 W), presumably for a few minutes per day for 8, 15, 20, 27, and 36 days. The mean 2 power density measured in the absence of the animal was 5 mW/cm . At the end of exposure, 50 mg/kg of Pentetrazol was administered intraperitoneally. They observed that RF radiation affected the convulsive time and mortality rate. Specifically, they noted that there was no difference between controls and ex- perimental animals through 8 days of exposure. After 15 days of irradiation, the animals exhibited delayed appearance of convulsions and were less 5-127 ------- susceptible to the epileptogenic action of Pentetrazol. However, after 27 to 36 days of exposure, the effect was reversed. The results indicate a biphasic response of mice over 36 days. Decreased incidence of mortality was observed in drug-injected animals if they were irradiated more than 8 days. Servantie et al. (1973) also investigated the effect of curare-like drugs on iji vivo and in vitro preparations of the rat's sciatic nerve. They found that irradiated rats were less susceptible to paralyzing drugs. Similar findings were noted in sciatic-nerve preparations from rats. In addition, sciatic nerves isolated from irradiated rats were paralyzed to a lesser extent and recovered sooner than those of control rats. Servantie et aL (1973) suggested that, based on their observations, the effect of microwaves is more likely at the neuro- muscular junction (synapse) and less likely due to secondary effects on the metabolism or binding of the drugs. Goldstein and Sisko (1974) investigated the effects of pentobarbital and o 9.3-GHz (CW) microwaves at a power density of 0.7 to 2.8 mW/cm (SAR estimated at 0.1 to 0.3 W/kg) on the behavior and EEG patterns of rabbits. They reported that there was no difference in EEG patterns between control (pentobarbital- injected) rabbits and rabbits treated with pentobarbital plus RF irradiation (0.7 mW/cm ) during the the first 5 min. However, after a latent period, cycles of intense arousal and sedation accompanied by other behavioral changes occurred in animals exposed to RF radiation after the barbiturate injection, but not in controls. Similar but more pronounced effects were seen at 2 mW/cm (SAR estimated at 0.2 W/kg). In this study, the EEG was recorded with the aid of implanted stainless steel electrodes. 5-128 ------- 2 Recently, Thomas et a_h (1979) reported that acute, low level (av 1 mW/cm ) 2450-MHz (PW) radiation in the near field (SAR estimated at 0.2 W/kg) poten- tiates the behavioral response to chlordiazepoxide (a tranquilizer) in rats. This was primarily a behavioral study and is discussed in § 5.5, Behavior. It can be concluded from such pharmacologic studies that low level microwaves may interact in a specific manner on the nervous system in con- junction with drugs. Such interactions may prove potentially to be both useful and harmful as more information becomes available. 5.4.4 Effects on Neurotransmitters Specific neural systems that contain various neurotransmitters are known to affect the inhibitory or excitatory states of the brain. The relative firing rates of these neuronal systems is reflected in the turnover of their neurotransmitters. Since RF radiation has been reported to stimulate and depress the CNS, several scientists have investigated the effects of RF radiation on CNS neurotransmitters. Snyder (1971) made neurochemical measurements in rats exposed to 3000-MHz (CW) radiation. He observed that a 1-h exposure at 40 mW/cm (SAR estimated at 8 W/kg) resulted in a significant increase of 5-hydroxyindolacetic acid (5-HIAA) and 5-HT (serotonin) in discrete nuclei of the brain. The opposite effect, that is, reduced levels of 5-HIAA and 5-HT, was found in rats exposed ? at 10 mW/cm (SAR estimated at 2 W/kg) 8 h/day for 7 days. The body tempera- 2 ture in the rats exposed at 10 mW/cm rose by 1 to 2 °C during irradiation, 5-129 ------- and the animals showed signs of moderate heat stress. To compare the effects p of microwave exposure (10 mW/cm ) to those of conventional radiant heat loads, rats were placed in an incubator maintained at 34 °C by thermostatically con- trolled incandescent lights for 8 h/day for 7 days. Body temperature was 2 raised to the equivalent of that observed after irradiation at 10 mW/cm . No difference was found in the turnover rate of norepinephrine or 5-HT or steady state levels of 5-HIAA between heated and control animals. Snyder concluded that exposure to RF-radiation produced distinctly different effects on 5-HT and 5-HIAA in rat brains than conventional heating. Zeman et aK (1973) investigated the effects of acute and chronic ex- posure of rats to 2860-MHz (PW, average power 300 W) radiation on rat brain gamma-aminobutyric acid (GABA). Chronic exposures were conducted for 4 to 2 6 weeks, 4 to 8 h/day, at 10 mW/cm (SAR estimated at 2 W/kg). Acute ex- 2 posures were conducted at 40 to 80 mW/cm power density (SAR estimated at 8 to 16 W/kg) for 5 or 20 min. No significant change in body temperature was 2 noted during chronic exposures at 10 mW/cm , while rectal temperature in- 2 creased by 3 °C after acute exposures at 40 to 80 mW/cm . There was no significant difference in whole-brain GABA levels between control and irradiated animals in acute or chronic experiments. These results would be more meaningful if the investigation had focused on discrete areas of the brain rich in GABA, rather than whole brains; however, others have used the procedure of whole-brain assays successfully. Merritt et al^. (1976) reported a decrease in norepinephrine, dopamine, and serotonin in discrete areas of rat brains after whole-body exposure for 5-130 ------- 2 10 min to 1.6-GHz microwaves at 80 mW/cm (SAR estimated at 24 W/kg). These exposures raised rectal temperatures by 3.7 °C. Hypothalamic epinephrine decreased significantly in irradiated and hyperthermic controls, although serotonin levels decreased in hippocampi of irradiated rats but not in those of control hyperthermic rats. There were no differences in the serotonin contents in the hypothalamus and striatum between hyperthermic controls and irradiated animals. Merritt et ah concluded that hyperthermia was responsible for the effects on neurotransmitters. In a separate study, Merritt et al_. (1977) observed a decrease in rat hypothalamic norepinephrine and dopamine after a 10-min exposure to 1.6-GHz 2 2 microwaves at 20 mW/cm (SAR estimated at 6.0 W/kg), but not at 10 mW/cm . The higher power density was associated with increased temperature, whereas the lower was not. Serotonin in the hypothalamus was unaffected by microwaves even at power densities as high as 80 mW/cm . No comment was made on hippo- campal levels of serotonin in this study. The effect of RF radiation on calcium ion efflux from brain and other tissues is discussed in § 5.7.5, Calcium Ion Efflux. 5.4.5 Unresolved Questions The effects of low-level exposure to RF radiation on the development of the CNS need much greater attention. Thus far, only gross brain teratologic and brain-weight studies have appeared in the literature. Investigations on subtle effects of RF radiation on CNS development are lacking. The effects 5-131 ------- on the growth and maturation of the nervous system at the cellular and sub- cellular levels need to be examined. Albert et al_. (1981) reported permanent loss of cerebellar Purkinje cells in rat pups after exposure of pregnant dams to 100-MHz (CW, SAR = 2.77 W/kg) and 2450-MHz (CW, SAR = 2 W/kg) radiation. These authors reported a reversible decrease in cerebellar Purkinje cells if newborn rat pups were exposed to 2450-MHz radiation. Synergistic and antagonistic effects of RF radiation with alcohol, x-rays, and other agents should be investigated, since some findings already indicate that this area should not be ignored. Nervous system effects are summarized in Table 5-12. There appear to be ample data that suggest effects of RF radiation on the nervous system of animals and humans. On the other hand, data on the BBB are controversial. Long-term low level studies on the adult nervous system are conspicuously lacking in the U.S. In addition, information on the effects of chronic, low level RF-radiation exposure on neurotransmitters is lacking. In summary, the state of the art does not permit one to assume that ex- posure to low level RF radiation produces a significant effect on nervous system morphology, the blood-brain barrier, CNS-active drugs, or neurotrans- mitters. Since chronic low level studies are lacking in this country, great care must be taken when making judgments until long-term studies are completed. 5-132 ------- TABLE 5-12. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON THE NERVOUS SYSTEM* Effects Species Frequency (MHz) Exposure Conditions Intensity (bM/co2) Duration (days x oin) SAR (W/kg) Reference Desynchronized EEG - Rabbit 3000 (PW) Lowered tolerance to CMS Rabbit 3000 (PW) stimulating drug Biphasic effect on the potency Mice 3000 (PW) of a convulsive drug Changes in EEG patterns of Rabbits 9300 (CW) anesthetized aninals Potentiation of drug response Rats 2450 (PW) Decreased brain NE, DA, Rats 1600 (CW) serotonin in hyperthermic aninals Decreased hypothalamic NE, Rats 1600 (CW) DA, serotonin No effect on neurotransaltter Rats 1600 (CW) levels No effect on GABA content Rats 2860 (CW) 20 (av) 7 (av) 0 7-2 8 1.0 (av) 80 20, 80 10 80 40 10 1 x 20 24-26 x 180 8-36 x Unknown 1x5 1 x 30 1 x 10 1 x 10 1 x 10 1 x 50 1 x 20 24-56 x 240-480 3 0 (est) 1 0 (est) 0 1-0 3 (est) 0 2 (est) 24 (est) 6-24 (est) 3 0 (est) 16 0 (est) 8 0 (est) 2 0 (est) Edelwejn (1968) Edelwejn (1968) Servantle et al. (1973) Goldstein and Sisko (1974) Thosas et al (1979) Merritt et al (1976) Herritt et al (1977) Merritt et al (1977) Zeoan et al. (1973) (continued) ------- TABLE 5-12. (continued) Exposure Conditions Effects Species F requency (HHz) Intensity (¦W/co2) Duration (days x mm) SAR (V/kg) Reference Swollen neurons in specific brain regions Chinese hamsters 2450 (CV) 25-50 25 1 x 30-1440 22 x 840 3 5-17 5 (est) 8 8 Albert and DeSantis (1975) Swollen neurons in hypo- thalamus and subthalamus Chinese hamsters 1700 (CV) 10, 25 1 x 30-120 5, 12.5 (est) Albert and DeSantis (1976) Myelin figures in dendrites 6 weeks post-exposure Rats 2450 (CV) 23 110 x 300 2 3 Switzer and Mitchell (1977) Increased permeability of BOB to fluorescein Rats 1200 (CV) 1200 (PV) 2.4 0.2 (av) 1 x 30 1 x 30 1 0 (est) 0 08 (est) Frey et a^ (1975) Myelin degeneration and metabolic alterations in glial cells Guinea pigs Rabbits 3000 (CV) 3000 (CV) 3.5-25 5 0 5-3 5 (est) 0 4-2 8 (est) Baranski (1972b) Focal areas of increased BBB permeability to peroxidase Chinese hamsters Rats 2450 (CV) 2450 (CV) 10 10 1 x 120-480 1 x 120-480 2.0 0 9 Albert (1979b) Brain temperature elevation (40-45 °C); Increased perme- ability of BBB Rats 2450 (CV) 80 V 1 x 10-30 " Sutton and Carroll (1979) Increased permeability of BBB (mannitol and inulln) Rats 1300 (CV) 1300 (PV) 1.0 0 3 (av) 1 x 20 1 x 20 0 4 0 1 Oscar and Hawkins (1977) en i M co (continued) ------- TABLE 5-12. (continued) Effects Species Frequency (MHz) Exposure Conditions Intensity (nW/ca) Duration (days x Bin) SAR (V/kg) Reference in i Co ------- 5.5 BEHAVIOR Michael I. Gage 5.5.1 Introduction Behavior has been defined as "anything an organism does" (Catania 1968), or as "the actions or reactions of persons or things under specified circum- stances" (Morris 1976). Behavior may also be defined as the actions of an organism in relation to its environmental stimuli. Behavioral effects of RF radiation have been extensively studied for several reasons. The first is the reports of experiments (not fully docu- mented with methods and detailed data analysis) from the Soviet Union and other European countries that behavioral effects of microwaves are seen at 2 relatively low levels, i.e., 500 pW/cm and below at 2375 MHz in rats (Dumansky and Shandala 1974; Shandala et aK 1977). The second is that behavior can serve as an index of total organism functioning, especially by elucidating the integrative status of the nervous system as well as that of other organ systems of the body. Moreover, behavior can be analyzed in a non terminal fashion, without resort to surgically or biochemically invasive pre- paratory techniques. Behavior may be separated into two major categories: naturalistic and acquired. Spontaneous or naturally occurring behavior may be innate and is often species-specific in frequency of occurrence. Examples of naturally occurring behavior include locomotor activity, eating, drinking, and mating. 5-137 ------- There are two general categories of acquired or learned behavior—respon- dent and operant, which are distinguished by the procedures used in the acquisition or conditioning of the behavior. Respondent conditioning occurs as a consequence of the temporal contiguity between stimuli. Stimuli paired in time with another stimulus, which reflexively elicits a response, gradually elicit the response. Examples of acquired respondent behavior are salivation and hunger pangs when passing a restaurant or eye blinks to acoustic stimuli. Responses conditioned by respondent procedures are usually measured by their occurrence and their magnitude. Much complex behavior of human beings and other higher animals in the course of daily activities can be viewed as emitted or operant behavior. Operant conditioning occurs as a consequence of reinforcement that follows the emission of a response. Reinforcers may be positive (such as food or water), or negative (such as termination of electric shock or intense radiant energy). All reinforcers maintain or increase the frequency of response. By definition, a positive reinforcer is a stimulus that increases or maintains the probability of emission of an operant by its presentation. A negative reinforcer is a stimulus that increases or maintains this probability by its removal after emission of the operant. Reinforcement need not follow the occurrence of each response. It may be intermittent according to a schedule (i.e., a schedule of reinforcement). Responses are said to be conditioned when they have a highly predictable probability of occurrence. This prob- ability is often expressed by the average occurrence within a period of time, or the response rate. Examples of operant behavior include eating with a 5-138 ------- knife and fork, driving a car, or writing a review on behavioral effects of microwaves. Operant conditioning may be used to answer specific questions about be- havior that may be similar across species. For example, by reinforcing the responses in the presence of light of one wavelength but not reinforcing responses in the presence of light of any other wavelength, animals can learn to respond selectively in the presence of light of the first wavelength. By presenting light at wavelengths close to the one that is reinforced, the threshold for discriminability of colors can be obtained. Specific types of behaviors investigated in behavioral research are so numerous that no attempt can be made to describe all of them here. Descrip- tions will be given of behaviors that have been studied as a function of microwave exposures. For a more complete description of behavior as studied by ethologists and psychologists, the reader is referred to several standard books (Brown 1975; Hinde 1970; Honig and Staddon 1977; Kling and Riggs 1971; Konorski 1967; and Pavlov 1960). 5.5.2 Summary Some general statements can be made regarding the effects of RF-radiation exposure on behavior: o Some microwave effects have been reported on various kinds of animal behavior. Although most of the studies have used rats as subjects, a few have used mice, squirrel monkeys, and rhesus monkeys. 5-139 ------- • Changes in locomotor behavior have occurred after CW exposures at an SAR as low as 1.2 W/kg in rats (D1 Andrea et a^. 1979). Changes in food and water intake or body mass have not been consistently reported at such levels. • Decreases in rates of ongoing operant behavior have been seen during exposures at SAR =2.5 W/kg in rats (de Lorge and Ezell 1980), and cessation of operant behavior has been seen at an SAR of 10 W/kg in rats (D1 Andrea et al_. 1976). • Alterations in operant performance measured after exposure was terminated also occurred with SAR's of 2.5 W/kg or more in rats (Gage 1979a). • The threshold for detection of microwaves may be as low as 0.6 W/kg in rats (King et al. 1971). However, it is not cer- tain that animals avoid or attempt to escape from CW microwaves except at very high power levels. • Drug effects on behavior in rats have been augmented after PW-radiation exposures lasting 0.5 h at average SAR = 0.2 W/kg (Thomas et aK 1979). Behavioral thermoregulation has been altered after only several minutes of exposure at SAR =1.0 W/kg in the rat (Stern et al. 1979) and at SAR =1.0 W/kg in the squirrel monkey (AdaTr and Adams 1980b). • Although the same behavioral effects during or after microwave exposure of the same magnitude may not be consistently predict- able, enough behavioral changes have been reported after simi- lar exposures to warrant the conclusion that behavior is disrupted by microwaves with an energy input that approximates one quarter to one half of the resting metabolic rate of many animals. • These behavioral alterations are reversible with time. 5.5.3 Naturalistic Behavior Spontaneous locomotor behavior has been studied with both acute- and chronic-exposure regimens in the rat. Hunt et al. (1975) found decreased initial exploratory activity by male Wistar rats after a 30-min exposure to 2450-MHz fields in a multimodal cavity with power adjusted to produce an SAR of 6.3 W/kg. The activity of exposed animals returned to control values 5-140 ------- within a 2-h period. The decrease in initial exploratory activity was the same whether the rat was placed in the apparatus immediately after or 1 h after exposure was terminated. Core body temperature was increased to 40.3 °C at termination of exposure but dropped to 37.8 °C, within the normal range, 1 h after exposure. Decreased swimming speed in water at 24 °C was also seen in rats that were practiced swimmers after 30-min exposures to 2450-MHz fields at SAR's of either 6.3 or 11 W/kg. The effects were seen only after 1.2 km of swimming following a 6.3-W/kg exposure; after an 11-W/kg exposure, effects were seen immediately for the first few meters and again after 0.6 km of swimming despite the fact that water at 24 °C would reduce persistence of a microwave-related hyperthermia. Roberti et ah (1975), on the other hand, did not see changes in spon- taneous motor activity, as measured in a glass cage, in male Wistar rats after four different exposure conditions in the far field of an anechoic chamber. The locomotor activity measured by Roberti et ah may not have required as much physical effort as continuous swimming. Exposures (estimates of SAR values based on single animal exposures) were either to 10,700-MHz (CW) fields at 0.6 to 0.9 mW/cm2 (SAR estimated at 0.15 W/kg) for 185 continuous hours (24 h daily for 7 2/3 days); to 3000-MHz (CW) fields at 0.5 to 1.0 mW/cm2 (SAR can be estimated at 0.3 W/kg) for 185 continuous hours; to 3000-MHz (PW) fields, 769 pulses/s, at 1.5 to 2.0 mW/cm2 (SAR estimated at 0.6 W/kg) for 185 2 continuous hours; or to 3000-MHz (PW) fields, 769 pulses/s at 24 to 26 mW/cm (SAR estimated at 8.3 W/kg) for 408 continuous hours (24 h daily for 17 days). After this last exposure condition, no change was also observed in running speed during forced runway running by the rats. 5-141 ------- Locomotor activity on a small platform was increased (as compared with five controls) in five female Sprague-Dawley rats during the course of repeated exposures to 2450-MHz (CW) fields in a multimodal cavity (Mitchell et al. 1977). Exposures of 5-h duration occurred 5 days weekly for 22 weeks. The average SAR was determined to be 2.3 W/kg, which is similar to the SAR at p power density of 10 mW/cm in a plane-wave environment. The activity in- creased within the first week of exposure and remained high throughout the course of the exposure period. Decreases in activity, measured by visual observation, were reported during repeated exposures of eight Wistar male rats to 918-MHz (CW) fields in a circular waveguide (Moe et ah 1976). Each exposure lasted 10 h (from 2200 to 0800) and occurred nightly for 3 weeks (total of 210 h), and the range of p whole-body average SAR's was measured as 3.6 to 4.2 W/kg (10 mW/cm average power density). The decreased activity predominantly occurred shortly after the microwave field was activated. The exposed rats were stretched out in a prone position more frequently than control rats in the early morning hours. In addition, exposed rats were reported to consume less food than did the controls over the course of the exposure, even though their body mass was not different from that of controls. A repeat of this experiment where eight male Wistar rats received an average SAR of 0.9 to 1.0 W/kg (average power density of 2.5 mW/cm ; daily 10-h exposures lasting for 13 weeks) resulted in no differences in food consumption or in activity measured during the eleventh week (Lovely et al. 1977). The two studies indicate that a dose- rate-related threshold for these effects might be somewhere between 1.0 and 3.6 W/kg. Unfortunately, in the circular waveguide, the power density is 5-142 ------- twice the average on axis and falls off rapidly toward the wall, resulting in the possibility of fluctuating and uncertain quantitative description of SAR at specific times during exposure (Guy and Chou 1976). Effects on spontaneous behavior of rats were reported in two other chronic exposure experiments. Both experiments used 15 exposed and 15 control male Long-Evans rats irradiated from 0900 to 1700 (8-h), 5 days weekly for 16 weeks in an anechoic chamber with a central monopole antenna and ground plane. The rats were adapted to the exposure and testing apparatus for 4 or 8 weeks before irradiation. In one experiment, exposures were to 2450-MHz (CW) fields 2 at an average SAR of 1.2 W/kg (power density, 5 mW/cm ) (D'Andrea et a2- 1979); and in the other, exposures were to 915-MHz (CW) fields at an average SAR of 2.5 W/kg (power density, 5 mW/cm^) (D1Andrea et aK 1980). In the study at 2450 MHz, rats showed decreases in activity as measured on a stabilimetric platform throughout the course of exposure but increased running-wheel activity overnight through the course of exposure (this latter effect was not significant). No significant differences were seen in food and water intake and in body mass. In the study at 915 MHz, exposed rats showed increased activity as measured both in the running wheels and on the stabi1imeter. Again, no significant changes in food and water intake or body mass were observed. Rudnev et a^. (1978) reported effects of exposure of 25 male albino rats to 2375-MHz (CW) fields at 0.5 mW/cm^ (SAR estimated at 0.1 W/kg for indi- vidual animal exposure) for 7 h daily for 1 month. Open field activity was 5-143 ------- measured in irradiated rats and in 25 controls as the number of squares crossed in 3 min on 2 successive days. The count on the first day was defined as exploratory activity, whereas that on the following day was defined as motor activity. Shock-induced aggression was measured by observing battles between an exposed and control rat. Maintenance of balance on a rotating treadmill (dynamic-load endurance) and on an inclined rod (static-load endurance) was measured, as was the amount of food consumed in 20 min after 23 h of deprivation. Electrodermal skin sensitivity was measured as the voltage of 100-Hz square-wave electrical stimulus that was needed to elicit paw withdrawal from the metal bars on the cage floor. Measurements were made prior to irradiation on the 10th, 20th, and 30th days of exposure, and every 15 days for 3 months after exposure. A significant reduction in food intake, time on the treadmill, and in time on the inclined rod was seen by the 10th day of exposure. Exploratory activity was significantly decreased, and shock sensitivity was increased after 20 days of exposure. At the end of exposure, exploratory and motor activity, time on the inclined bar, and shock sensitivity were significantly decreased. The latency to start eating was increased for 30 days after exposure ended. Time on the treadmill was reduced for 15 days after the termination exposure. Exploratory activity was increased throughout a 3-month postexposure period. Sensitivity to electric shock was reduced significantly on the 30th and 60th day after exposure ended and was still below control levels on the 90th day. 5-144 ------- 5.5.4 Learned Behavior 5.5.4.1 Respondent Conditioning— Currently, there appear to be no reports by U.S. investigators of micro- wave exposure altering respondent behavior, although an experiment using microwaves as an unconditional stimulus has been published. Some studies using respondent conditioning techniques that report changes in behavior as a consequence of microwave exposure appear in the Soviet literature (e.g., Dumansky and Shandala 1974; Lobanova 1974). Unfortunately, details regarding exposure conditions or behavioral methodology and results are too sketchy to permit inclusion in this review. An attempt was made to use microwave ex- posure as an unconditional stimulus in one experiment (Bermant et al_. 1979). A 30-s presentation of a 525-Hz tone that preceded 2450-MHz sinusoidally modulated microwaves that were either presented for 10 s (SAR = 420 W/kg) or 30 s (SAR = 220 W/kg) or that preceded an electric shock to the tail produced rises in rectal temperature of 0.7 C, 0.5 C, or 0.37 °C, respectively, during a base-line period over the course of 10 conditioning sessions. Control female Sprague-Dawley rats presented with the tone alone showed a decrease of 0.47 °C in rectal temperature during this period. The microwave exposure itself was designed to produce an increase in rectal temperature of 1.5 °C. Aside from this study there are no reports of microwaves altering respondently conditioned behavior in which exposure parameters allow clear determinations of the SAR. 5-145 ------- 5.5.4.2 Operant Conditioning— There is a large body of literature examining alterations of operant behavior produced by microwaves for which SAR values have been, or can be, determined. Reduction in rates of responding on an operantly conditioned task has been observed during the course of microwave irradiation. Rats were trained to lever-press for food pellets on a random-interval, 30-s schedule. After behavior on this task stabilized, the rats showed a response rate that was in general linearly uniform and typical of random- or variable-interval schedule- controlled performance. The rats were then exposed to microwaves under each of several conditions in sessions lasting 25 min, or until the rat's response rate fell below one third of its base-line control rate. In an initial experi- ment (D'Andrea et al_. 1976), six male Long-Evans rats were exposed to 360-, 480-, or 500-MHz (CW) fields in a parallei-piate apparatus at an incident o power density of 25 mW/cm (the long axis of the rat was parallel either to the electric-field vector or to the vector of wave propagation). Responding was reduced only during exposures to 500-MHz fields when the long axis of the body was parallel to the electric-field vector. Behavior stopped abruptly after ~ 11 min of exposure, and upon removal from the apparatus the animals appeared flaccid, wet, and, according to the authors, heat stressed. At this exposure the SAR was computed from the measured 0.16 °C/min rise in rectal temperature to be ~ 10 W/kg. Exposures that produced no change in behavior had SAR's ranging from 5 to 6 W/kg. In a second experiment (D1 Andrea et al_. 1977), these results were con- firmed and extended. Exposures were conducted in a monopole-above-ground 5-146 ------- radiation chamber and lasted for up to 55 min, or until responding on the random-interval schedule fell below one third of the base-line rate. Exposures of five male Long-Evans rats for periods up to 55 min to 400-, 500-, 600-, and 700-MHz (CW) fields at 20 mW/cm2 power density and 22 ± 1 °C ambient temperature yielded a U-shaped function of time to the criterion of reduction in rate, with the minimum of ~ 23 min at 600 MHz when rectal temperature increased 0.09 °C/min (SAR estimated at 16.4 W/kg). At the time they stopped responding all rats appeared heat stressed and were engaged in spreading saliva on their fur. Six additional rats, exposed at 600-MHz (CW) fields at 2 power densities of 5, 7.5, 10, and 20 mW/cm , showed decreased times to stop 2 their responding, at power densities above 7.5 mW/cm when rectal temperature 2 increased 0.04 °C/min or more. At 10 mW/cm (SAR estimated at 7.5 W/kg) the rats stopped responding after ~ 45 min, and rectal temperature increased 2 0.04 °C/min. Three rats exposed to PW microwaves with 170 mW/cm peak and 2 5.10 mW/cm average power density showed no change in performance. As in the earlier experiment of D'Andrea et al_. (1976), response reduction was abrupt and was correlated with the rectal temperature increase. A series of experiments by de Lorge (1976, 1979) also examined altera- tions in operant performance during exposure to microwaves. In most of these experiments, the schedule of reinforcement was used to measure observing and detection responses in a vigilance task. Two levers were present in a testing chamber or in front of a primate restraint chair made of Styrofoam and sheet plastic for optimal transparency to microwaves. Responses on the right lever, called observing responses, produced either a low- or a high-frequency acoustic signal. The high-frequency signal was scheduled to occur on a variable inter- 5-147 ------- val of either 30 or 60 s. If the animal made a response on the left lever when the high-frequency signal occurred—a detection response—it received a food pellet as a reinforcer. Observing responses occurred at fairly linear response rates, a pattern similar to that seen on variable interval schedules of reinforcement. In the first experiment (de Lorge 1976) five male rhesus monkeys (Macaca mulatta) showed no change in behavior on this schedule during 1 or 2 h of exposure to 2450 MHz in the far field of an anechoic chamber when power 2 densities ranged to 16 mW/cm (SAR estimated at 1.2 W/kg). The field was 100 percent amplitude modulated at 120 Hz. Three of these monkeys showed reduced 2 rates of observing response during 1-h exposure to this field at 72 mW/cm (SAR estimated at 5.0 W/kg) but not at 32, 42, 52, or (in two monkeys) 62 2 2 mW/cm . At 72 mW/cm , rectal temperatures rose ~ 2 °C, and were still increasing at the end of the hour-long session. The animals moved more in their chairs after 20 min, and were observed to take short naps after ~ 30 min at this power density. When observing responses decreased, reaction time to respond on the left lever increased. In the second experiment (de Lorge 1979) four male squirrel monkeys (Saimeri sciureus) were tested and exposed to 2450-MHz 120-Hz modulated microwaves under conditions similar to those described above at 10, 20, 30, 40, 50, 60, and 70 mW/cm . No reduction in the mean rate of right-lever observing responses > 1 standard deviation below the mean of the control rate was seen at any power density during the 30-min exposures. But brief pauses in responding were seen at microwave onset and offset at 50 mW/cm and above 5-148 ------- (SAR estimated at 2.75 W/kg). At this power density rectal temperature rose > 1 °C. Three of the monkeys were also given 1-h exposures. Observing response rate seemed to decrease more, and, according to an impression gained from performance of one monkey (de Lorge 1979, Figure 6), this decrease began earlier in sessions at higher power densities. One of the three monkeys exhibited an increase of response rate as power densities increased. After microwave exposure, all monkeys also showed decreased responding that was directly related to the power density. All monkeys showed increase in frequency of incorrect responding on the left lever for food, which was a direct function of power density. Reliable changes in behavior were considered to occur only at power densities of 40 to 50 mW/cm at a time when rectal temperatures had risen by > 1 °C. In a third experiment (de Lorge and Ezell 1980) eight male Long-Evans rats were exposed to 5620-MHz (PW) fields at 662 pulses/s, and then to 1280-MHz (PW) fields at 370 pulses/s. They were tested on the behavioral task of vigilance during exposures (SAR's were reported at 0.19 W/kg per mW/cm at 5620 MHz, and 0.25 W/kg per mW/cm^ at 1280 MHz). Rates of observing responses 2 during exposures to 1280-MHz fields decreased somewhat at 10 mW/cm (average 2 power density), and decreased markedly after 15 to 20 min during 15 mW/cm . At 5620 MHz, rates of observing responses decreased only during exposures at o 26 mW/cm and above. Behavior decrements occurred at SAR =2.5 W/kg at 1280 MHz but required SAR =4.9 W/kg at 5620 MHz. In a related experiment (Sanza and deLorge 1977), four male Sprague- Dawley rats were trained to respond on a fixed-interval 50-s schedule for 5-149 ------- food pellets as reinforcers. With this schedule, the first response emitted 50 or more seconds after arrival of the last food pellet produced another food pellet. Exposures for 60 min to 2450-MHz, 120-Hz modulated at fields at 2 37.5 mW/cm in the far field of an anechoic chamber produced decreases in response rates that had a fairly abrupt onset. The response decrements were seen only in two rats that had high base-line rates and not in two rats with low response rates. Exposures at 8.8 and 18.4 mW/cm produced no decrements in performance. All rats spent more time at the wall opposite the food cup 2 2 during 18.4- and 37.5-mW/cm exposures than during sham or 8.8-mW/cm ex- 2 posures. The SAR was not given but is estimated at 3.7 W/kg at 18.4 mW/cm and 7.5 W/kg at 37.5 mW/cm^. Responding of three male rhesus monkeys trained to a high level of pro- ficiency on a visual tracking task was not disrupted by exposures at 10 or 20 mW/cm^ to 1200-MHz (CW) fields (reported SAR estimated at 0.8 and 1.6 W/kg) during behavior sessions lasting ~ 2 h (Scholl and Allen 1979). Some other studies have looked at changes in previously learned operant behavior at the termination of single or multiple exposure periods. Thomas et al. (1975) trained four male Sprague-Dawley rats to respond on a multiple schedule. One component was fixed ratio 20 (FR20): every 20th response on the right lever was reinforced by a food pellet. The other component was a differential-reinforcement-of-low-rate of 18 s with a limited hold of 6 s (DRL 18 LH 6): Responses on the left lever separated by at least 18 s, but by no more than 24 s, were reinforced. Components alternated on an irregular basis. Only one of the schedules was in effect during a given time period. 5-150 ------- Exposures to microwaves lasted for 30 min, and behavioral testing began 5 to 10 min after exposure. Exposure parameters were as follows: 5, 7, 15, and 20 mW/cm2 at 2450-MHz (CW) fields; 5, 10, 15, and 20 mW/cm2 at 2860-MHz (PW) 2 fields at 500 pulses/s and 1-ps pulse width; and 1, 5, 10, and 15 mW/cm at 9600-MHz (PW) fields at 500 pulses/s and 1-ys pulse width. All rats were exposed to all parameters while restrained in the far field of an anechoic chamber. In general, response rates increased on the DRL schedule and de- creased on the FR schedule. Increased rates on the 0RL schedule were seen 2 following exposures to 5 mW/cm and above at 9600-MHz fields (SAR estimated at 1.5 W/kg), 7.5 mW/cm2 at 2450-MHz (CW) fields (SAR estimated at 2.0 W/kg), and 10 mW/cm2 and above at 2860-MHz fields (SAR estimated at 2.7 W/kg). De- creased response rates on the FR schedule were measured following exposures 2 to 5 mW/cm and above at all frequencies (SAR estimated at 1.5 W/kg for 9600 MHz, and 1.4 W/kg for 2450 and 2860 MHz). Increased response rates 2 during time-out periods between components were seen following 5 mW/cm at all three frequencies. Time-out responses peaked and then dropped after ex- posures at higher power densities. In a second study, Thomas et al_. (1976) trained four male Sprague-Dawley food-deprived rats on a fixed-consecutive-number-eight (FCN 8) schedule. With this schedule, at least eight presses had to be made on the right lever before depression of the left lever would yield a food pellet reinforcer. If the rat made fewer than eight consecutive responses on the right lever before switch- ing, the count was restarted at zero. Well trained rats were tested after 2 30-min exposures in the near field at 5, 10, and 15 mW/cm to 2450-MHz (PW) fields at 500 pulses/s, and 1-ps pulse width. Because exposures were in the 5-151 ------- near field, SAR cannot be precisely estimated but may be assumed at 0.4 W/kg 2 for each mW/cm (Durney et aK 1980). The percentage of eight or more con- secutive responses on the right lever (reinforced runs) decreased, and the length of these runs decreased after all exposures. Those decreases were direct functions of power density. However, the overall rate of responding and the running rate (i.e., the rate after the first response following rein- forcement) hardly decreased during these postexposure sessions. Gage (1979a) reported decreases in response rates of eight adult male Sprague-Dawley rats trained to alternate between levers either 11 or 33 times for a food-pellet reinforcer. The decrements occurred in sessions following o overnight, 15-h exposures to 2450-MHz (CW) fields at 10, 15, and 20 mW/cm , 2 but not after exposures at 0.5 and 1.0 mW/cm . Only very small decrements 2 were seen after 55-min exposures at power densities to 30 mW/cm (the SAR 2 measured in rats under similar exposure conditions was 0.3 W/kg per mW/cm ). Vigilance discrimination was tested immediately after 30 min of irradiation in 10 well trained, young adult male Wistar rats (Hunt et 1975). Exposures produced SAR's of 0.0, 6.5, and 11.0 W/kg at 2450 MHz (modulated in a quasi - sinusoidal fashion at 120 Hz in a multimodal cavity). All rats were exposed at both SAR values, but the sequence of exposures was varied. The task required the rats to obtain saccharin-flavored water reinforcers by pressing a bar when a light flashed, but by not pressing a bar when a sonic stimulus was presented. One of the two stimuli was presented every 5 s, and the light was presented 12.5 percent of the time. Rats at both SAR's had an increased number of omis- sion errors (i.e., failures to respond to the light after exposure), but there 5-152 ------- were no increases in errors of comission (i.e., responding wrongly when the noise was presented). The failure to correctly respond was more frequent at the start of the session, as well as after exposure at the higher SAR. Schrot et aK (1980) experimented with three male albino rats trained to learn a new sequence of pressing three levers for food reinforcers daily in repeated-acquisition procedure. The rats increased their number of errors and decreased their rate of sequence completions when tested immediately after 30-min exposures to 2800-MHz (PW) fields (500 pulses/s, 2-(js pulse width) at 5 and 10 mW/cm average power density. No effects were seen at lower o average power densities of 0.25, 0.5, and 1 mW/cm . Peak powers were 0.25, 0.5, 1, 5, and 10 W/cm in these exposures. The rats were exposed in a sleeve holder with the electric-field vector perpendicular to the long axis of the animal's body. The reported SAR's based on temperature measurements were 0.7 o and 1.7 W/kg at 5 and 10 mW/cm , respectively. Operant behavior has also been examined during or following the course of chronic microwave exposures in some reports also described in §5.5.3, Natural- istic Behavior. Mitchell et a^. (1977) measured performance on two schedules in separate groups of rats pretrained before the start of 22 weeks of 5-h exposures, 5 days/week, to 2450-MHz fields (average SAR at 2.3 W/kg) in a multimodal cavity. Five exposed and five control female Sprague-Dawley rats were tested for 30-min sessions on a schedule of multiple FR5 extinction for 15 s (MULT FR5 EXT 15 s). When a white lamp was on, every fifth response was reinforced by a food pellet, but during periods when the lamp was off, no responses were reinforced. Although the control rats had higher response 5-153 ------- rates during the FR component, this difference was not significant. Exposed rats had higher response rates than controls during the extinction component, and this difference was statistically significant. The ratio of response rate during the MULT-FR5 component to response rate during the EXT-15-s component was significantly different in exposed as compared with control rats: Exposed rats exhibited a higher ratio, indicating poorer discrimination over the course of exposure, although they had a value like that of the control rats before irradiations began. Another schedule, Sidman avoidance, was used to test escape and avoidance response. Five exposed and four control rats were trained to postpone an unsignaled 2.0-mA foot shock for 15 s (response-shock interval) by pressing a lever. During 30-min sessions, the animals received a shock every 0.5 s (shock-shock interval) until they pressed the lever. No significant effects of microwave exposures were seen with this schedule, although improvement in avoidance was seen in all rats both within and over the course of testing sessions. Conditioned taste aversion was studied in experiments reporting chronic 2 exposures of rats to 918-MHz fields in circular waveguides at 10 mW/cm (Moe et a2. 1976) and 2.5 mW/cm^ (Lovely et aH. 1977). Rats were given a saccharin solution to drink in place of water during the microwave exposure period. Presumably, if the saccharin were drunk in conjunction with an agent, such as microwave radiation, which made the rat sick or produced some undesirable sensations, this connection would be learned and the solution would be avoided in the future. Preference was measured after exposure by allowing a water- deprived rat to choose between drinking the saccharin solution and water for p 20 min. In the experiment at 2.5 mW/cm (Lovely et aK 1977), saccharin 5-154 ------- preference was tested only from the 9th to the 13th week of exposure. No difference between exposed and control rats was seen in amount of saccharin 2 solution consumed either at 2.5 mW/cm (average measured SAR at 1.0 W/kg) or o at 10 mW/cm (average measured SAR at 3.9 W/kg). Several studies have investigated the ability of animals to detect or to take behavioral action to minimize, avoid, or escape from microwaves. In an early paper, King et al_. (1971) showed that three irradiated and three control male albino rats could respond to 2450-MHz microwaves doubly modulated at 60 and 12 Hz as the conditioned stimulus in a measure of conditional suppression. In this experiment, rats were reinforced with sugar water on a random interval schedule for licking at a water tube. At various times during each 2-h session either a 525-Hz tone or the microwaves were presented for 1 min, and an unavoidable 0.5-s electrical foot-shock followed. Conditioned suppression to the tone was reliably indicated by no licks being emitted during the tone. Microwave dose rates of 0.6, 1.2, 2.4, 4.8, and 6.4 W/kg were substituted for the tone in some sessions. During irradiation at SAR = 0.6 W/kg one of three rats suppressed responding, at SAR =1.2 W/kg two rats suppressed, and at SAR =2.4 W/kg all three suppressed reliably. Johnson et aK (1976) trained two male Wistar-derived rats to nose poke in a restraint for food pellets on an FR5 schedule in the presence of an acoustic-pulse stimulus of 7.5 kHz, 10 pulses/s, 3-ps pulse duration for 3-min periods, which alternated with 3-min periods of no stimulation during which nose pokes were not reinforced (extinction). When microwaves at 918 MHz (10 pulses/s, 10-ps pulse durations) were presented for 30-s intervals during 5-155 ------- an extinction period, or when microwaves were substituted for the auditory stimulus during reinforced periods, response rates were observed that were similar to those seen in periods when the acoustic stimulus was present. The 2 energy density per pulse of microwaves was 150 pJ/cm , and the average power o density was 15 mW/cm . (The SAR would be near 7.5 W/kg if the rats were in the far field.) Detection of microwaves does not imply that there are affective proper- ties of this stimulus, i.e., that they hurt or feel good. In fact, such detection may be further evidence of the RF hearing phenomenon (Frey and Messenger 1973). Indeed, in several experiments in which PW microwaves are presented during exposure, the alteration in behavior of the exposed animal might be due to effects of acoustic stimulation by the microwave pulses. Such experiments should have as a control the presentation of pulsed auditory stimuli. Frey et al^. (1975) experimented with female Sprague-Dawley rats exposed to 1200-MHz (PW) fields at 1000 pulses/s (0.5-s pulse duration) at an average 2 power density of 0.2 mW/cm (SAR estimated at 0.2 W/kg) and a peak power 2 density of 2.1 mW/cm . Six rats spent only 30 percent of a 30-min period of exposure in an unshielded half of a Styrofoam shuttle box during the last 2 of 4 successive daily exposures. Six other rats exposed to 1200-MHz (CW) fields 2 at 2.4 mW/cm (SAR estimated at 2.2 W/kg) spent 52 percent of the time in the unshielded half of the box on these days. Six other control rats spent 64 percent of the 30-min period in the unshielded half. Only the rats exposed to PW microwaves could be said to escape from the stimulus or exhibit a modest 5-156 ------- preference, which would decrease their exposure. A similar finding was also reported to occur in male rats during exposures to a 1200-MHz (PW) field at 2 100 pulses/s at 0.4 or 0.9 iriW/cm average power densities (SAR estimated at 2 0.4 and 0.81 W/kg, respectively) and at 133 or 300 mW/cm peak power densities (Frey and Feld 1975). These rats spent an average of only 29 percent of their time in the unshielded half in seven 90-min sessions, and this side preference was maintained throughout all 7 days. Sham-irradiated rats spent an average of 57 percent of their time in the unshielded half of the box. Two groups of eight male Wistar rats spent more than half of each of nine weekly hour periods in the side of a shuttlebox when occupancy of that side kept a PW microwave field turned off (Hjeresen et ah 1979). The 2880-MHz 2 field was pulsed at 100 pulses/s (3.0-\|js pulse width) with a 9.5-mW/cm o average and 33-mW/cm peak power density (SAR calculated at 2.1 W/kg). The exposure was in the far field of an anechoic chamber. A group of eight rats that could not extinguish the field and another group that received no micro- waves showed no side preferences. Preference for occupancy of the side that extinguished the field increased across each weekly session. Substitution of a 37.5-MHz (PW) acoustic stimulus for the microwaves in one session resulted in the rats' spending most of the session in the side that kept the acoustic stimulus off. A continuously occurring broadband "pink" noise in the anechoic chamber prevented appearance of side preference in two other groups. In addition to confirming that rats avoid or escape from pulsed microwaves, this experiment suggests that microwaves may be detected as an auditory stimulus. 5-157 ------- Monahan and Ho (1976) showed that male CF1 mice irradiated for 10 or 15 min at 2450-MHz (CW) fields in a waveguide when forward-power levels were stabilized at 0.4, 0.8, 1.6, 2.4, 3.2, 4.0, or 4.8 W exhibited a decline in energy absorption rate at 2.4 W and above. This decline usually occurred within the first 5 min of exposure and stabilized toward the latter part of the period. Monahan and Ho interpret the results as a behavioral action to minimize exposure to the microwaves. However, they did not directly measure any behavior. During the first 5 min of exposure the SAR's measured in this apparatus ranged from 7.7 to 65.7 W/kg. The lowest SAR at which the mice clearly altered their rates of energy absorption was 28 W/kg. In a second study Monahan and Ho (1977) showed that the SAR associated with reduction of absorption during 20 min of irradiation decreased reliably from 43.6 W/kg when ambient temperature was 20 6C, to 0.6 W/kg, when ambient temperature was 35 °C. Videotape observations of rats and mice exposed to 2450-MHz (CW) fields 2 at 15 mW/cm for 1 h in the far field below a radiating horn in an anechoic chamber did not reveal any preferential behavior that minimized whole-body absorption rates through parallel orientation to the magnetic- as opposed to the electric-field vector (Gage et aj. 1979). These observations occurred at ambient temperatures of 22 or 28 °C while the animal was held in a cuboid or cylindrical enclosure. Usually, the animals assumed a curled, sleep-like posture in the hour before the microwaves were turned on and maintained that posture throughout most of the exposure period; an exception occurred in mice exposed at 28 °C, in which case they more often assumed positions that were oriented parallel to both field vectors. The SAR of the rats without any 5-158 ------- 2 enclosure was 3.3 W/kg at 15 mW/cm , independent of their orientation. The SAR of the mice when parallel to the electric-field vector was 12.3 W/kg, and the SAR when the mice were parallel to the magnetic-field vector was 6.2 W/kg, without any enclosure. The animals changed positions more frequently when some difference in SAR existed relative to the position assumed. In a recent article, Carroll et aK (1980) reported that 20 female Long-Evans rats did not go to a marked-off floor area in a multimodal cavity to reduce the intensity of 918-MHz microwaves (modulated at 60 Hz with 3-Hz mode stirrer modulation) from 60 W/kg to either 40, 30, 20, or 2 W/kg. The mean number of entries into the marked-off "safe" area and the percentage of time spent there during 22-min sessions were not different when the microwaves were on (for five 2-min periods) or off. SAR's > 60 W/kg were associated with lethality within 8 min of continuous exposure in tests of a separate group of rats. However, in the same apparatus 10 rats quickly learned to escape from an 800-pA electric foot shock by going to this marked-off area within a 22-min session. Shocked rats would remain in this area over 90 percent of the session time. Sanza and de Lorge (1977) noted the fact that their rats first tried to jump out of a testing chamber and then assumed a stationary position after 2 failing to escape during exposures to 2450 MHz at 37.5 mW/cm (SAR = 7.5 W/kg). 5-159 ------- 5.5.5 Interactions With Other Environmental Stimuli Interactions of microwaves with two types of environmental stimuli have been reported to affect behavior. One of these stimulus types is chemical, specifically, several commonly used psychoactive drugs; the other is physical, e.g., ambient temperature during exposure. Response rates of rats performing on a fixed-interval 1-min schedule of reinforcement beginning 0.5 h after 2.5- to 20-mg/kg dosages of chlordiazepoxide (Librium) (Thomas et al^ 1979) were increased over control values. These increases were further augmented if an 2 exposure at an average power density of 1 mW/cm of 2450-MHz (PW) fields (500 2 pulses/s, 2-\|js pulse width, 100-mW/cm peak power density) occurred in the half hour between the injection and behavioral testing. In a second study (Thomas and Maitland 1979), the dose-response function of rats given d-amphetamine sulfate was shifted to the left after exposure to microwaves having the same parameters and conditions of exposure as those mentioned above. A shift to the left of a dose-response function indicated an increased potency of the drug. Rats in this second study were reinforced for responses separated by more than 18 s (DRL 18 s). Drug doses producing increases and decreases in response rates were lower after microwave exposures than those after no microwave exposures. Although the rat was in the near field in these studies, the dose rate measured in a Styrofoam-insulated water model was 0.2 W/kg. Thomas and Maitland (1979) also showed that 0.5-h exposure to microwaves, at the parameters indicated above, 4 days/week when amphetamine was not administered, shifted the dose-response function of this drug to the left on the fifth day when a microwave exposure did not occur. In both of these studies, exposure to microwaves alone or after saline injection had no 5-160 ------- effect on behavior. The shift to the left of a dose-response function indi- cated that an exposure to microwaves, which by itself did not affect behavior, acted synergistically with chlordiazepoxide or amphetamine to increase sensitivity of the organism to the drug. Similar results have not been seen with chlorpromazine or with diazepam (Valium), an analog of chlordiazepoxide, in experiments in the same laboratory (Thomas et al_. 1980). In a related experiment, Monahan and Henton (1979) showed that chlordiaz- epoxide and, with less consistency, chlorpromazine and d-amphetamine altered response rates of mice trained in an operant-conditioning procedure to escape from or to avoid 2450-MHz (CW) fields at an average dose rate of 46 W/kg. While this experiment showed that drug effects can interact to alter microwave exposure effects, the design of the experiment does not allow any conclusion to be drawn about interactions between drugs and microwaves on performance. Interactions with ambient temperature and humidity were predicted by Mumford (1969). Monahan and Ho (1977) showed that reduction in the rate of energy absorption by mice exposed in a waveguide to 2450-MHz (CW) fields occurred at a low dose rate of 0.6 W/kg when ambient temperature was 35 °C but required higher doses at lower ambient temperatures. Although the mice were not directly observed, the authors presumed the mice reoriented in the micro- wave field to reduce the amount of absorbed energy. 2 Gage (1979b) showed that overnight exposures at 5, 10, or 15 mW/cm to p 2450-MHz (CW) fields (SAR =0.2 mW/g per mW/cm ) at an ambient temperature of 28 °C reduced operant response rates of rats measured the morning after ex- posure was terminated. Similar exposures when ambient temperature was 22 °C 5-161 ------- did not result in reduced response rates except after exposure at the highest power density. Two reports have indicated mammals will alter thermoregulatory behavior 2 in the presence of as little as 5 mW/cm of 2450-MHz (CW) microwaves. In one (Stern et aK 1979), six male rats were trained to press a lever to switch on an infrared (IR) heat lamp for 2s in a chamber at an ambient temperature of 3.9 to 5.3 °C. Microwaves at power densities as low as 5 mW/cm reduced the rate of responding for the heat lamp. The reduction in response rate that occurred shortly after microwave onset was a direct function of power density 2 in the range of 5 to 20 mW/cm and returned to base-line values when micro- waves were switched off. The rats were exposed in the near field (SAR =0.2 W/kg per mW/cm ). Adair and Adams (1980b) showed that thermoregulatory behavior of the squirrel monkey (a new world primate found in equatorial jungle areas) was significantly altered at power densities as low as 6 mW/cm in the far zone of a 2450-MHz (CW) field within 10 min of onset of exposure. In this study, the monkeys were trained to select a preferred ambient air temperature by making an operant response to obtain 15 s of 55 °C air when the air was otherwise 15 °C or to obtain 15 s of 15 °C air when the air was otherwise 55 °C. Preferred air temperatures ranged between 35 and 36 °C without microwaves and decreased as a direct function of the microwave power density during exposure. (The SAR in this experiment was determined calorimetrically on saline models to be 0.2 W/kg per mW/cm .) As in the experiment by Stern et aJL (1979) with rats, preferred temperatures returned to base-line levels when the microwave field was extinguished. 5-162 ------- The above-mentioned studies of behavioral effects caused by microwave exposure are summarized in Table 5-13. In conclusion, there is ample evidence to suggest that microwaves alter a variety of unlearned and learned behaviors occurring during and after ex- posures. In most cases the behavior change can be described as a reduction in the level of ongoing activity. However, there are some situations in which increases in activity have been seen. When measured, the magnitude of behavioral change seems to be related to the power density or SAR of the ex- posure. Behavioral changes usually revert to base-line levels after removal of the microwave field. 5.5.6 Unresolved Questions Most unresolved questions regarding behavioral effects of microwaves arise because observed findings have not been verified within the same laboratory or in other laboratories. Repeating a study involves high cost and the risk of failure to confirm a finding due to small unnoticed differences, e.g., between standard procedures in two laboratories. Possibly for these reasons verification has often not been attempted. However, verification and sys- tematic replication would allow determination of the limits of conditions within which a behavioral effect may be expected, as well as definition of the range of conditions adequate to observe a threshold of effect. 5-163 ------- TABLE 5-13. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON BEHAVIOR Exposure Conditions Effects Species Frequency Intensity Duration SAR Reference (Weight, g) (PWz) (riW/ca2) (days x «in) (W/kg) Decreased exploratory activity Young sale and swinging speed, AT increase rat 2.5 °C No effect on spontaneous activity Hale rat or activity and forced running (160-180) Increased locosotor activity Feoale rat (307) Decreased spontaneous activity and Male rat food intake (360-410) No effect on spontaneous activity Male rat or food intake (316-388) Decreased activity on stabillaetric Hale rat platform, no significant increase (350-375) In wheel running Increased activity on stabllinetric Hale rat platfonn and 1n wheel running (350-375) 2450 (PW, 7 ¦ultlaodal cavity 120 Hz, AH) 10700 (CW) 3000 (CW) 3000 (PW) 3000 (PW) 2450 (CW, aulti- ¦odal cavity) 918 (CW) 918 (CW) 2450 (CW) 915 (CW) 0.6-0.9 0 5-1.0 1.5-2 0 24-26 (av) 10 10 2.5 5 1 x 30 7.7 x 1440 7 7 x 1440 7 7 x 1440 17 x 1440 110 x 300 21 x 600 91 x 600 80 x 480 80 x 480 6.3 0.2 0.3 0.6 8.3 2 3 3.6 1 0 1 2 2 5 Hunt et al (1975) Robertl et aK (1975) Hitchelletal (1977) Hoe et al (1976) Lovely et al (1977) D'Andrea et al^ (1979) D1Andrea et al (1980) (continued) - ------- TABLE 5-13. (continued) Exposure Conditions Effects Species Frequency Intensity Duration SAR* Reference (Weight, g) (MHz) (wU/cw2) (days x mm) (W/kg) Decreased time on treadmill and Hale rat inclined rod, decreased exploratory activity, increased then decreased shock sensitivity Decreased activity and shock sensitivity persisted 90 days after exposure Rectal temperature rise = 0 37 °C Female before start of test, AT = 1 5 °C rat with microwaves Response decreased during exposure on random interval schedule (lowest intensity for effect, AT = 1 8 °C) Response decreased during exposure (maximum effect) on random interval schedule, AT = 1 8 °C Decreased observing responses on vigilance task, AT = 2 °C No effect on observing responses Decreased observing responses on vigilance task No effect on observing responses Hale rat (350-380) Hale rat (357-382) Hale rhesus monkey (4 kg) 2375 (CW) 2450 (PW, multi- modal cavity, 60 and 12 Hi AH) 0 5 30 x 420 0 1 Rudnev et al (1978) 500 (CW) 600 (CW) 2450 (120 Hz, AH) Hale squirrel 2450 (120 Hz, monkeys AH) (850-950) 25 10 72 16 50 10 x 0 17 10 x 0 5 1 x 11 1 x 55 1 x 60 1 x 20 1 x 30 1 x 60 1 x 60 420 Bermant et aj^ (1979) 220 10 7 5 5 0 1 1 2 8 0 6-17 0'Andrea et al (1976) D'Andrea et al (1977) de Lorge (1976) de Lorge (1979) (continued) ------- TABLE 5-13. (continued) Exposure Conditions Effects Species (Weight, g) Frequency (MHz) Intensity (mW/cm2) Duration (days x mm) SAR* (W/kg) Reference Decreased observing responses on vigilance task Hale rat (362-400) 1280 5620 (PW) (PW) 10 26 1 x 40 1 x 40 2 5 4 9 de Lorge and Ezell (1980) Response rate decreased on fixed interval schedule in rats with high baseline rates, spending time away from lever No effect on response rate Hale rat (290-340) 2450 AH) (120 Hz, 37 5 8 8-18 4 1 x 60 1 x 60 7 5 1 8-3 7 Sanza and de Lorqe (1977) VJ1 1 o No effect on visual tracking task Response rate decreased on FR and increased on DRL schedules Hale rhesus monkey (6 2-7 9 kg) Male rat (120 days, 150') 1200 2450 2860 9600 (CW) ------- TABLE 5-13. (continued) Exposure Conditions Effects Species (Weight, g) Frequency (MHz) Intensity Duration (mW/cm2) (days x mm) SAR* (W/kg) Reference Increased response rates in extinction, decreased stimulus control, no effect on Sidman avoidance No effect on flavor aversion test No effect on flavor aversion test Microwaves detected as stimulus Microwaves detected as stimulus Spending more time in shielded vs unshielded compartment Spending equal time in shielded vs unshielded compartment Spending more time in shielded vs unshielded compartment (occurred in first of 7 sessions) Spending more time in unirradiated compartment Female rat Male rat (360-410) Male rat (316-388) Male rat (409-427) Male rat (300-350) Female rat Female rat Male rat (250) Male rat 2450 (CW, multimodal cavity) 918 (CW) 918 (CW) 2450 (PW, 120 Hz, AM, multimodal cavity) 918 (PW) 1200 (PW) 1200 (CW) 1200 (PW) 2880 (PW) 10 10 2 5 15 0 2 2 4 0 4 9 5 (continued) 110 x 300 21 x 600 91 x 600 1 x 1 1x05 4 x 30 3 x 30 1 x 90 9 x 60 2 3 3 9 1 0 0 6-2 4 7 5 0 2 2 2 0 4 2 1 Mitchell et al^ ( 1977) Moe et a^ (1976) Lovely et aj (1977) King et a\| (1971) Johnson et a^ (1976) Frey et a^ (1975) Frey et aj[ (1975) Frey and Feld (1975) Hjeresen et a\| (1979) ------- TABLE 5-13. (continued) Effects Species (Weight, g) Frequency (MHz) Exposure Conditions Intensity (mW/cm2) Ouration (days x mm) SAR* (W/kg) Reference Decrease in SAR at 24 °C Decrease in SAR when ambient temperature increased from 20 °C to 35 °C No preferential orientation of rats or mice in far field of plane wave Cannot take specific action to reduce intensity of irradiation Augmentation of increased response rates produced by chlordiazepoxide Shift to left of dose-response curve for d-amphetamine in DRL schedule No effect on dose response curve for chlorpromazine or diazepam Hale mouse (30-34) Hale mouse (30-34) Hale rat (200-360) Hale mouse (25-33) Female rat (290) Hale rat (325-375) Hale rat (250-300) Hale rat (360-380) 2450 (CW) 2450 (CW) 2450 (CW) 2450 (CW) 918 (PW, 60 Hz, AM, multimodal cavity) 2450 (PW) 2450 (PW) 3800 (PW) 15 15 1 x 15 1 x 20 1 x 50 1 x 60 5 x 2 1 x 30 1 x 30 4 x 30 1 x 30 28 Monahan and Ho (1976) 43 6-0 6 Monahan and Ho (1977) 3 3 Gage et a^ (1979) 6 2-12 3 (depending on orientation) 60 Carroll et al (1980) 0 2 0 2 0 2 0 2 Thomas et al (1979) Thomas and Maitland (1979) Thomas et al (1980) (continued) ------- TABLE 5-13. (continued) Exposure Conditions Effects Species (Weight, g) Frequency (MHz) Intensity (mW/cm2) Duration (days x mm) SAR* (W/kg) Reference Chlordiazepoxide reduced responses, decreased avoidance responses, and increased escape responses to microwaves Male mouse (35-44) 2450 (CW) 1 x 30 46 Monahan and Henton (1979) Increased temperature during exposure caused greater response rate decreases after exposure Hale rat (315-365) 2450 (CW) 10 1 x 930 2 0 Gage (1979b) Reduced responding for heatlamp in a cold room Hale rat 325-450) 2450 (CW) 5 1 x 15 1 0 Stern et a! (1979) Selection of a lower ambient air temperature Squirrel monkey (750-1100) 2450 (CW) 6 1 x 10 1 0 Adair and Adams (1980b) If measured SAR was not reported, SAR was estimated when possible ------- There is no unifying hypothesis to explain all the observed behavioral changes. The research on interaction between microwaves and chemicals has not been verified in independent laboratories, and it has not been extended to see if the interaction is limited to particular classes of compounds. Delineation of differences between effects of single and multiple exposures would help determine if effects of chronic exposure are qualitatively different from repeated measurement of effects of each single exposure. Behavioral experiments have used only a limited sample of microwave exposure conditions. Exploration of frequency spectrum, modulations, wave- forms, and interactions between waves of different parameters has hardly begun. Most behavioral work has used rats as subjects. Animals more like humans in physical size and shape have only infrequently been studied to help extrapolate findings to humans. Specifications of conditions of exposure other than the microwave stimulus such as ambient temperature and humidity have not been consistently controlled and reported so that their influence on behavior may be determined. There is no information on exposure effects of specific or limited areas of the body in comparison to total body exposure to evaluate the effects of localized energy absorption. 5-170 ------- 5.6 SPECIAL SENSES Joe A. Elder 5.6.1 Cataractogenic Effects A cataract is an opacity in the crystalline lens of the eye. These lens defects may be clinically insignificant or may cause partial or total blind- ness. The following conclusions may be drawn from animal experiments on the cataractogenic potential of RF radiation (Tables 5-14 and 5-15): 1. High-intensity RF radiation is cataractogenic. 2. For single acute exposures, the threshold intensity for cataract production exceeds 100 mW/cm2. Multiple exposures at intensities near threshold values for single acute exposures will result in lens opacities. 3. The cataractogenic potential of RF radiation varies with frequency; the most effective frequencies appear to be microwave frequences in the 1-to 10-GHz range. 4. Similar ocular effects are produced by CW and PW radiation of the same average intensity. 5. In contrast with the above conclusions, which are based on acute, near-field exposures to the eye or head, no cataracts have been reported in unrestrained animals after far-field exposure at power densities near lethal values. 5.6.1.1 Microwave Radiation in Experimental Cataractogenesis-- The rabbit eye has been the experimental model most often used in animal studies because of its similarity in size and anatomy to the human eye (Figure 5-3). Its diameter is ~ 75 percent that of the human eye; the cornea is as large, and although the lens is thicker, its diameter is the same 5-171 ------- TABLE 5-14. SUMMARY OF STUDIES CONCERNING OCULAR EFFECTS OF NEAR-FIELD EXPOSURES Exposure Conditions Effects Species Frequency Intensity Duration SAR Reference (MHz) (raW/cm2) (days x mm) (W/kg) Cataract Rabbit 5500 (CW and PW) 470-785 1 x 2-100 300-500t Birenbaum et al (1969a) Cataract Rabbit 800 (CW) 785* 1 x 25 50ot Birenbaum et a_[ (1969b) 4200 (PW) 785* 1 x 17 5oo; 4600 (PW) 785* 1 x 15 500 + 5200 (PW) 500-785* 1 x 5-12 350-5001 5400 (CW and PW) 500-785* 1 x 3-4 300-500J 5500 (CW and PW) 500-785* 1 x 2-3 300i500 6300 (PW) 785* 1x5 500 Cataract Rabbit 2450 (CW) 180 1 x 240 Carpenter (1979) 120-180 20 x 60 No cataract Rabbit 2450 (CW) 75 20 x 60 Carpenter (1979) Cataract Rabbit 2450 (CW) 150 1 x 100 138* Guy et al (1975a) Cataract Rabbit 2450 (CW) 295 1 x 30 Hagan and Carpenter (1976) 10,000 (CW) 375 1 x 30 No cataract, keratitis Rabbit 35,000 ~ 40 1 x 60 s 175# Rosenthal et al (1976) (inflammation of cornea) 107,000 - 40 1 x 60 > 238 iEstimate calculated by dividing the microwave power by the irradiated area (d = 1 27 cm) of the eye lEstimate based on the assumption that all the incident power was absorbed by the eye (2 g) ^Maximum SAR in the eye Estimated SAR values for the cornea (See text for discussion of Rosenthal et ^ studies of frequency specificity ) ------- TABLE 5-15. SUMMARY OF STUDIES CONCERNING OCULAR EFFECTS OF FAR-FIELD EXPOSURES Exposure Conditions Effects Species Frequency (MHz) Intensity (mW/cm2) Duration (days x mm) SAR (W/kg) Reference No ocular effects, including no lenticular changes Rabbit 3000 (CW) 100, 200 1 x 15, 30 14, 28* Appleton et aj (1975) un I Acute ocular changes, e g , hyperemia of lids and conjunctiva, meiosis, anterior chamber flare, engorgement of iris vessels, and periorbital cutaneous burns, no lenticular changes 300, 400, 500 1 x 15 42, 56* 70* Death 300 500 1 x 30 1 x 15 42* 70* No cataracts Rabbit 385 (CW) 385 (CW). 468 (CW)T 60 30 60 10 x 15 10 x 90 10 x 20 48* 24* 8 1 Cogan et a\| (1958) No cataracts Rabbit 2450 (CW) 10 5 x 480 (x 8-17 weeks) 1 5* Ferri and Hagan (1976) No ocular effects Monkey (H mulatta) 9310 (PW) 150 30-40 x 294-665** McAfee et al (1979) JEstimated average whole-body SAR values (Ourney et 1978, figure 31) Iwaveguide average whole-body exposure Total exposure time in minutes for the entire 30- to 40-day experimental period ------- CILIARY BODY RETINA CHOROID CILIARY BODY SUSPENSORY LIGAMENT OF LENS SCLERA CORNEA LENS IRIS TOVEA OP1 ic NtRVE LENS CORNEA RIS ANTERIOR CHAMBER POSTERIOR CHAMBER OPTIC NERVE CENTRAL ARTERY AND VEIN OF RETINA POSTERIOR CHAMBER VITREOUS BODY VITREOUS BODY CONJUNCTIVE Figure 5-3. Cross-sectional sketch of the human (left) and the rabbit (right) eye (from Birenbaum et a^. 1969a, Figure 1). as that of the human eye. In a typical study (Carpenter et ak 1960b), one eye of an anesthetized rabbit was exposed to microwave radiation in the near field, and the nonirradiated eye served as the control. The eyes were then examined at various intervals using an ophthalmoscope, slit-lamp biomicroscope, or both instruments. The earliest positive reaction in this type of study, occurring within 24 to 48 h after a cataractogenic exposure, is the appearance of one or two narrow translucent or milky bands in the posterior cortex of the lens, just under the capsule, which extends no further than the lens equator. These bands can be seen only by slit-lamp examination with an angled beam. If the ocular injury is minimal, no further change occurs, and the cortical 5-174 ------- banding disappears within a few days. Otherwise, in 2 to 4 days after exposure small granules, appear in the region of the suture of the posterior lens. If a more intense reaction occurs, larger numbers of granules appear over a larger area within the next few days, and small vesicles may develop. These early changes may develop further and become either well-defined circumscribed or diffuse cataracts. These lens changes remain as permanent ocular defects. In general, it has been found that microwave cataracts in rabbits involve only the posterior cortex of the lens, unless the exposure is so intense that the opacity extends throughout the lens. At exposure levels that cause cataracts, other ocular reactions also occur, but these are transient and differ in severity with the intensity and duration of the exposure. Examples include swelling and chemosis of bulbar and palpebral conjuctivae, pupillary constriction, hyperemia of iris and limbal vessels, and vitreous floaters and filaments (Carpenter 1979). 5.6.1.2 Exposure Threshold Values for Cataracts-- Following the publication of reports demonstrating that microwaves can cause cataracts in experimental animals (Richardson et al. 1948; Daily et aj. 1950a,b), three laboratories (Williams et a^. 1955; Carpenter et al^. 1960b; Carpenter and Van Ummersen 1968; and Guy et cH. 1975a) published time vs. power-density threshold curves for cataract induction in rabbits by a single exposure to near-field 2.45-GHz radiation. The time vs. power-density thres- hold curves originally published by the three laboratories are similar in shape but are quantitatively different. The more recent studies found lower threshold values than those reported by Williams et al. (1955). This probably 5-175 ------- reflects differences in the irradiation method and in techniques used to measure power density. In fact, Carpenter (1979) has determined that his power densities were 50 percent higher than originally published. His corrected data are plotted in Figure 5-4 along with the results of Williams et al. (1955) for comparison. Guy et ah (1975a) replicated Carpenter's work for single acute exposures with essentially the same results, and also quantified the threshold of cataractogenesis in terms of SAR (see Figure 5-4). For example, these workers found the cataractogenic threshold for a 100-min exposure, the longest period of irradiation, to be 150 mW/cm (138 W/kg) peak absorption (cf. Figure 5-5). The cumulative effect of microwave radiation on cataractogenesis in the rabbit has been examined by repeatedly exposing the eye at power densities below the threshold for single acute exposures (Carpenter 1979). For example, 2 daily 1-h exposures at 180 mW/cm for 13 to 20 days were found to be catar- actogenic in 8 of 10 animals, whereas single exposures at this power density 2 were not effective. At 150 mW/cm , 4 of 10 rabbits gave a positive response after 18 to 32 daily exposures. No cataracts were observed after 20 daily 1-h 2 exposures at 75 mW/cm . If the entire body of an unanesthesized rabbit is exposed at power densities similar to those that cause cataracts under near-field conditions, the animal exhibits acute stress; Appleton et ah (1975) reported that rabbits became heat stressed and struggled out of the field during a 15-min exposure 2 at 100 mW/cm . This report is described in more detail in § 5.6.1.5. 5-176 ------- 700 -600 600 - -500 500- -400 400- -300 WILLIAMS ef a/ (1955) z 300 - CARPENTER et a! (1960b 1979) -200 200- GUY eial (1975a) -100 100 - 100 80 60 40 20 0 EXPOSURE TIME (min) Figure 5-4. Time and power-density threshold for cataractogenesis in rabbits exposed to near-field 2450-MHz radiation; values of maximal SAR are also given (from Guy et a^. 1975a, Figure 7). 5.6.1.3 Frequency Specificity— As indicated above, most studies of experimental cataractogenesis were conducted at 2.45 GHz, but opacities in rabbit eyes have been reported after near-field exposures at 0.8, 4.2, 4.6, 5.2, 5.4, 5.5, 6.3, and 10 GHz (Birenbaum et al. 1969a,b; Hagan and Carpenter 1976). In several studies, the cataractogenic potential of different frequencies was addressed. For example, after Hagan and Carpenter (1976) determined the relative effects of 2.45- and 10-GHz (CW) radiation on the rabbit eye, they concluded that the cataractogenic potential 5-177 ------- 1 25-i E 100- 5 E E 0 75 - cn 0 50- < > u cc RETINA 0 25- 0 00 POSTERIOR POLE — ANTERIOR POLE CORNEA DEPTH FROM CORNEAL SURFACE Figure 5-5. Distribution of energy absorption rate (W/kg) per mW/cm incident power density in the rabbit's eye and head exposed to 2450-MHz radiation (from Guy et al. 1974, Figure 3). for single acute exposures is greater at the lower frequency. At both frequencies, the opacities were characteristically located in the posterior subcapsular cortex of the lens, although the initial appearance and subsequent development differed. These differences probably reflect differences in the pattern of absorbed microwave energy in the eye due to the different depths of penetration of the radiation at these two frequencies. Guy et ak (1974, 1975a) measured the distribution of absorbed energy in rabbit eyes exposed to 918- and 2450-MHz radiation and found the patterns to be significantly different. At 2450 MHz, energy absorption was maximal in the vitreous body at a point midway between the posterior surface of the lens and 5-178 ------- the retinal surface (Figure 5-5). The locus of peak absorption thus cor- relates well with the observation of irreversible changes in the posterior cortical lens. Exposure to 918-MHz fields in a specially devised cavity resulted in relatively uniform absorption in the eye, but maximal absorption was only ~ 25 percent of the peak absorption at 2450 MHz (Figure 5-6). There- fore, one would expect the threshold for cataractogenesis in rabbits exposed to 918 MHz to be considerably higher than the threshold at 2450 MHz. But more importantly, at 918 MHz, peak absorption in the rabbit brain was 36 percent higher than in the eye. It is possible that lens effects or ocular changes may not occur before more severe damage occurs in other sensitive tissues, such as the brain, during exposure to 918-MHz fields. 0 30 -i 5 E 0 25* cc o 0 70 - ~c O > o CC 0 15 - n HEAD RETINA LENS 2 3 DEPTH FROM CORNEAL SURFACE (cm) Figure 5-6. Distribution of energy absorption rate (W/kg) per mW/cm inci- dent power density in rabbit's head and eye exposed to 918-MHz radiation (from Guy et a]_. 1974, Figure 27). 5-179 ------- Both Hagan and Carpenter (1976) and Guy et a^. (1975a) used exposure sys- tems that applied microwave energy across an air space to the eye, and both reported lenticular cataracts in the posterior subcapsular cortex. Birenbaum et al. (1969b) produced cataracts in the anterior cortex of rabbit lens with an exposure system that applied PW microwaves to the corneal surface. Further- more, as the frequency decreased from 6.3 to 5.2 to 4.6 to 4.2 GHz, longer exposure times at a constant field strength were required to produce lens defects. Under similar experimental conditions, even longer exposure times were required to induce cataracts at a lower frequency, i.e., 0.8 GHz (CW) radiation. Furthermore, Birenbaum et £l. (1969a) found no substantial differ- ence in the cataractogenic threshold values for CW and PW 5.5-GHz radiation and concluded that the average, not the peak, rate of energy absorption deter- mines whether lens injury will occur. Rosenthal et a_h (1976) examined the effects of 35- and 107-GHz radiation on the rabbit eye. At both frequencies keratitis (inflammation of the cornea) occurred at lower intensities than required to produce any other demonstrable ocular effect, such as lens injury (cataract) or iritis. Irradiation at 107 GHz was more effective in producing immediate corneal damage, but this change was generally gone by the next day. Effects at 35 GHz were persistent, were almost always present the next day, and were associated with marked injury to the corneal epithelium. Effects on the cornea correlate well with the pattern of microwave-energy absorption because most of the energy at these high frequencies is absorbed in the outer regions of the eye. The earliest stage of keratitis or minimal corneal stromal injury occurred after a 30-min exposure at an incident power of 50 mW or after a 60-min exposure at 25 mW. Estimates 5-180 ------- of the rate of energy absorption in the eye at the lower incident power (25 mW) are 35 mW at 35 GHz and 47.5 mW at 107 GHz (Rosenthal et ah 1976); there- fore, the average SAR's of a rabbit eye weighing 2 g are 17.5 and 23.8 W/kg, respectively. Since maximal absorption occurs in the outer structures of the eye, the SAR in the cornea is estimated to exceed the average SAR of the eye by more than one order of magnitude. Although the above data cannot be compared directly because of widely varying experimental procedures, these results indicate that the potential for cataract induction in rabbits is higher in the frequency range between 1 and 10 GHz than at either lower or higher microwave frequencies. Power densities that would cause cataracts at frequencies below 1 GHz and above 10 GHz give rise to insult in other ocular sites and extra-ocular tissues. Rosenthal et al. (1976) found that 35- and 107-GHz radiation primarily affected the outer structure of the eye, for example, the cornea; and at 918 MHz, Guy et ah (1974) showed that maximal energy absorption occurred in the brain, not in the eye. The effects of near-field microwave exposures on the rabbit eye have been summarized by Cleary (1980) as follows: "The induction of cataracts in experimental animals, principally New Zealand white rabbits, has been described by a number of investi- gators using a variety of microwave exposure modalities. Generally microwave field intensities necessary for cataract induction in the rabbit are such that acute whole body exposures would be lethal due to hyperthermia. Most cataract studies have, therefore, employed focused or near-zone fields which limit exposures to the head or eyes. Localized thermal trauma is still of such a magnitude to necessitate the use of general or local anesthesia. Corneal irri- gation with physiological saline solutions have also been used to 5-181 ------- prevent corneal damage due to tissue dehydration during microwave exposure Anesthesia and corneal irrigation, as well as air temper- ature and humidity, may significantly affect the temperature of ocular structures." The experimental results strongly suggest that radiation-induced tempera- ture elevation may be essential for the cataractogenic effect of microwaves. Additional evidence for this position has been provided by Kramar et ak (1975), who reported that rabbits kept under general hypothermia during irradiation at known cataractogenic levels of 2.45-GHz (near-field) radiation did not develop cataracts. This study is described more fully in § 5.6.2. 5.6.1.4 Far-field Exposure Studies-- In contrast with the acute, near-field exposures that can cause cataracts and other ocular effects, cataracts have not been produced in rabbits whose entire bodies were exposed to radiation in the far field (Table 5-15). Appleton et al. (1975) exposed anesthetized rabbits to far-field radiation at 3000 MHz 2 for 15 or 30 min at 100 or 200 mW/cm . No ocular changes were observed during or immediately after exposure. Fourteen daily examinations and four weekly examinations, followed by monthly examinations for 1 year revealed no 2 lenticular changes. During exposure at higher levels (300, 400 or 500 mW/cm for 15 min) animals exhibited acute ocular changes consisting of hyperemia of lids and conjunctiva, meiosis, anterior chamber flare, engorgement of iris vessels, and periorbital cutaneous burns. Subsequent examinations revealed no morphologic lenticular abnormalities. The authors concluded that "It is noteworthy that one year after a single microwave exposure, sufficient in intensity to cause both thermal cutaneous and acute gross ocular effects, no lens changes or cataracts were observed." It is also noteworthy that these 5-182 ------- power levels and durations were well above the cutaneous sensation level, because unanesthetized animals became heat stressed and struggled out of the 2 2 field during a 15-min exposure at 100 mW/cm . Exposures at 300 mW/cm for 2 30 min or 500 mW/cm for 15 min were lethal to some of the rabbits. Ferri and Hagan (1976) exposed unanesthetized rabbits to 2450-MHz (CW) 2 radiation in the far field at 10 mW/cm , 8 h/day, 5 days/week for 8 to 17 weeks. Weekly examinations of the eye showed no abnormal changes during the study, and no post-irradiation changes were observed during the following 3 months. Cogan et aK (1958) found no cataracts in rabbits 4 weeks after exposure 2 at 385 MHz; the rabbits were irradiated twice weekly for 5 weeks at 60 mW/cm 2 for 15 min or 30 mW/cm for 90 min. Six weeks after exposure in a waveguide, 2 no cataracts were observed in rabbits irradiated at 468 MHz, 60 mW/cm (SAR = 8.1 W/kg), for 10 days (20 min daily). Although the authors reported that the exposure levels at both frequencies were near lethal values, the 8.1 W/kg SAR is considerably lower than the value other investigators have found to be sublethal to rabbits (see Table 5-15). The environmental conditions (temperature, airflow, etc.) within the waveguide were not given; therefore, one must assume that the ambient conditions were significantly different from normal values or that the measured SAR is in error. McAfee et aL (1979) trained unfettered monkeys (Macaca mulatta) to 2 expose their faces and eyes to 9.31-GHz (PW) radiation at 150 mW/cm average power density. Over a period of about three months, the animals were irradiated 5-183 ------- for a total of 294 to 665 min during 30 to 40 daily sessions. No cataracts or corneal lesions were observed in these monkeys during a 1-year period follow- ing irradiation. 5.6.2 Unresolved Questions Exposure of the rabbit eye to microwave radiation at sufficient power densities and durations causes an immediate increase in intra-ocular temper- ature, and, after a latent period of a few days, opacities develop in the posterior subcapsular cortex of the lens (Kramar et a^. 1975). This sequence of events has led to the assumption that microwave-induced lens opacities are thermally caused. Several experiments have been designed to test directly the cause-and-effect relationship between temperature increase and cataract formation. Kramar et aK (1975) exposed rabbit eyes to cataractogenic levels of microwave radiation while the animal's body was submerged in cold water. By this means, microwave-induced intra-ocular temperature was limited to < 41 °C, and no lens opacities developed. In a later experiment, Kramar et aK (1976) used heated water to produce ocular and rectal temperatures characteristic of those in rabbits exposed to a cataractogenic level of microwaves. Although the vicinity of the lens was heated to temperatures above those known to be associated with microwave cataracts, no lens opacities were observed; however, the rate of ocular temperature rise was about one-tenth the rate of increase with micro- waves. Kramar et aK (1976) concluded that a combination of a sharp temper- ature gradient and rapid rise in temperature following irradiation may be more traumatic to the lens than a critical temperature per se. 5-184 ------- Carpenter et ak (1977) reported no cataracts of the posterior cortex of the lens in rabbits after 6 weeks of treatment in which "the eye was heated at the same rate, to the same extent, and for an equal period of time as it would experience during a cataractogenic microwave exposure, the difference being that the equal heating was provided by other means, namely, direct application of heat to the surface of the eye." In addition, elevating retrolental and rectal temperatures to values characteristic of a cataractogenic microwave exposure, through a combination of restricted body heat loss and irradiation of one eye to power densities slightly below the cataractogenic threshold, produced cataracts in only 3 of 10 rabbits. According to their hypothesis, if cataracts were solely of thermal origin, all animals given the two treatments should have developed cataracts. Carpenter et a_h (1977), therefore, concluded that the increase in intra-ocular temperature occurring during microwave irradiation is not the sole causative factor in microwave cataractogenesis. The reason for the apparent disagreement between the conclusions of Carpenter et aK (1977) and Kramar et cH. (1976) probably rests with the difficulty of duplicating by nonmicrowave heating techniques the temporal and spatial temperature profiles induced by microwave irradiation of the eye. Note that the temperature at a single site in the eye was the basis for evidence of duplicating microwave heating of the eye at the same rate and to the same extent. Furthermore, even though retrolental and rectal temperatures char- acteristic of cataractogenic microwave exposures were produced, the overall effects on the rabbits were more traumatic than a near-field microwave exposure to one eye. Within 24 h after heated water was applied directly to the eye, almost all the eyes exhibited hemorrhage into the anterior cavity, so that 5-185 ------- observation of the lens became impossible. Only 21 of 32 rabbits survived the heat treatment. Of these, five developed lenticular cataracts of the anterior cortex, and only three developed cataracts of the posterior cortex of the lens that are characteristic of microwave-induced opacities (Carpenter et al. 1977). As mentioned above, the difficulty of duplicating by nonmicrowave tech- niques the time-temperature profile of microwave energy absorption in the eye is probably responsible for the unsuccessful attempts to prove that an elevation of temperature is responsible for microwave cataracts (cf. Kramar et ah 1976; Carpenter et ah 1977). On the other hand, strong evidence for thermalization being the causative factor in microwave cataracts is provided by the experiment of Kramar et ah (1975), which showed that cataracts were not produced in hypothermic rabbits receiving a cataractogenic microwave exposure. At present, it is generally understood that intense localized exposure of the eye for p substantial durations (i.e., 150 mW/cm for 100 min) is necessary to induce cataracts in laboratory animals, and that such acute exposures cause death by hyperthermia if the entire animal is irradiated. 5.6.3 Auditory Effects When the human head is exposed to PW RF radiation, an audible sound described as a click, buzz, chirp, or a knocking sensation is perceived by some individuals; the sound appears to originate from within or behind the head. This auditory phenomenon is called "RF sound" or "RF hearing." Our present knowledge of RF hearing is summarized below. 5-186 ------- 1. The RF sound occurs only upon exposure to PW sources; it appears to originate from within or near the back of the head regardless of orientation of the head in the RF field, and the sound varies with pulse width and pulse repetition rate. 2. The effective frequency range is 216 to 6500 MHz. The effect has been found to occur at average power densities as low as 0.001 mW/cm2 with peak power densities in the range of 100 to 300 mW/cm2. Effective pulse widths vary from 1 to 1000 ps. (See Table 5-16.) 3. The ability to perceive RF pulses has been shown to be related to bone-conduction hearing and to the ability to hear high fre- quency acoustic waves above 5 to 8 kHz. 4. The available data support the conclusion that the RF auditory effect is evoked by a mechanism similar to that for conventional acoustic stimuli and that the primary site of interaction is peripheral to the cochlea. 5. The most generally accepted mechanism responsible for the RF auditory sensation is thermoelastic expansion. That is, the absorption of the energy in a brief RF pulse causes a small temperature rise (~ 10"6 °C) in a short time (~ 10 ps), which results in thermoelastic expansion of matter within the head, which then launches an acoustic pressure wave that is detected by the hair cells in the cochlea via bone conduction. 5.6.3.1 Human Perception of Pulsed RF Radiation-- Frey (1961) was the first to study systematically the human auditory response to pulse-modulated radiation. The subjects, who were more than 30 m from an enclosed antenna, reported hearing a transient buzzing sound upon ex- posure to the intermittent rotating beam. The apparent location of the sound, which was described as a short distance behind the head, was the same no matter how the people were oriented in the RF field. When an RF shield (aluminum flyscreen) was placed between the subject and the RF source, the subject did not perceive RF sounds (Frey 1973). When earplugs were used, a reduction in the ambient noise level and an increase in the RF sound level were reported. The sensitive area for detecting RF sounds was later described 5-187 ------- TABLE 5-16. SUMMARY OF STUDIES CONCERNING AUDITORY EFFECTS OF RF RADIATION IN HUMANS Exposure Conditions Effect Comment Number of Subjects Frequency (MHz) Pul se Repetition Rate (s1) Pulse Width (MS) Peak Intensity (mW/cm2) Average Intensity (mW/cm2) Energy/ Pul se (J/cm2) Noise Level (dB) References RF hearing "distinct" cltcks" Threshold values 8 3000 0 5 5 10 15 2500 225-2000 300-1000 0 006 0 001-0 01 0 002-0 007 12 5 2 3-20 0 4 5-15 0 45 (+ plastic foam ear muffs) Cain and Rissmann (1978) RF hearing buzz heard at PRR > 100, individual pulses heard at PRR < 100 3 3000 6500 <100-1000 <100-1000 1-2 1-2 2500-50,000 2500-50,000 5 5 40 Constant (1967) No auditory response 9500 0 5 Constant (1967) RF hearing "buzzing sound" 4 1245 1245 50 50 10 70 370 90 0 19 0 32 Frey and Messenger (1973) RF hearing "buzz, clicking, hiss, or knocking" Threshold values Not given 216 425 425 425 425 27 27 27 27 125 250 500 1000 670 263 271 229 254 4 0 1 0 1 9 3 2 7 1 70-90 (+ ear stopples) Frey (1962, 1963) No auditory response 8900 400 2 5 25,000 25 70-90 (+ ear stopples) Frey (1962) SF hearing "buzzing sound" Threshold values 8 7 1310 2982 244 400 6 1 267 5000 0 4 2 70-80 (+ ear plugs) Frey (1961) RF hearing "clicks, chirps" Threshold values 1 2450 3 1-32 1250-40,000 0 1 40 45 Guy et a! (1975) RF hearing Polytonal sound 18 800 1000-1200 10-30 > 500 40 ( + ear (stopples) Tyazhelov et al (1979b) Calculated peak-absorbed-energy-denslty per pulse is 16 mJ/Kg ------- as a region over the temporal lobe of the brain, because the placement of a small piece of metal screen (5x5 cm) over this area completely stopped the RF sound (Frey 1962). Guy et aK (1975b) described the effect of PW radiation on two of the coinvestigators. Three pulses, 100 ms apart, were presented each second to keep the average power density below 1 mW/cm . Each individual pulse was heard as a distinct and separate click, and short pulse trains were heard as chirps with the tone corresponding to the pulse repetition rate (PRR). The RF sound appeared to originate from within or near the back of the head. This report also included the note that transmitted digital codes could be accurately interpreted by the subject when the pulse generator was keyed manually. Guy et aK (1975b) also reported that the threshold for RF hearing was lower when ear plugs were used. In a study by Constant (1967), the RF sound was described as being in the area of the ear on the side opposite to the one that was irradiated. All three of his subjects readily detected 2-ps pulses, whereas 0.5-ps pulses were not perceived. All three experienced a buzzing sensation at PRR's greater than 100/s, whereas individual pulses were heard when subjects were exposed to PRR's below 100/s. Five of eight human subjects reported hearing distinct clicks either inside the head or behind the head when exposed to 15-ps pulses (Cain and Rissmann 1978). The remaining three people heard faint clicks when the pulse width was increased to 20 ps. 5-189 ------- 5.6.3.2 Effective Radiation Parameters— In the initial report by Frey (1961), human subjects perceived PW radiation at frequencies of 1310 and 2982 MHz. Although the peak of power- density thresholds for RF hearing was 266 mW/cm^ for 1310-MHz and 5000 mW/cm^ 2 for 2982-MHz fields, the average power density thresholds were 0.4 mW/cm and 2 2 mW/cm , respectively. When ear plugs were used to attenuate the ambient noise level of 70 to 80 dB, the subjects reported an increase in the RF sound levels. In the following year, Frey (1962) reported that humans could perceive PW 2 radiation at 425 MHz with an average power density threshold of 1 mW/cm ; the 2 peak of power density was 263 mW/cm . A frequency of 8900 MHz was not effective 2 even at an average power density of 25 mW/cm ; the peak of power density was 2 25,000 mW/cm . At 216 MHz, the lowest effective frequency reported in the 2 literature, the average power density threshold was 4 mW/cm ; the peak of 2 power density was 670 mW/cm (Frey 1963). Later, in a study of four subjects exposed to 1245-MHz PW radiation, Frey and Messenger (1973) concluded that perception was dependent primarily upon peak power and secondarily upon pulse width. They reported that the peak power for perception was somewhat less 2 than 80 mW/cm for pulse widths of 10 to 70 ps. In the study by Constant (1967), three human observers were exposed to PW 2 radiation at 3, 6.5, and 9.5 GHz at an average power density of 5 mW/cm (pulse width was 0.5 to 2.0 ps; PRR was up to 1000/s). Only two of the three observers perceived 3- and 6.5-GHz radiation; none experienced a response to 5-190 ------- the highest frequency. Cain and Rissmann (1978) reported that all eight of their subjects heard RF pulses at 3 GHz. In this study, plastic foam earmuffs were worn to attenuate the ambient noise, which was 45 dB. The average 2 threshold energy density per pulse was 10.6 pJ/cm (range of values was 3.4 to 17.5 pJ/cm2). The range of microwave pulse widths varied from 1 to 32 ps in the study by Guy et al^. (1975b) on one human subject. The results indicate that regard- less of the peak power of the pulse or the pulse width, the threshold for RF 2 hearing of 2450-MHz radiation was related to an energy density of 40 pJ/cm per pulse, or energy absorption per pulse of 16 pJ/g, as calculated with the aid of a spherical model. The background noise of the exposure chamber was 2 45 dB. When earplugs were used, the threshold level decreased to 28 pJ/cm . The threshold for a second subject, who had a hearing deficit, was approxi- 2 mately 135 pJ/cm . Guy et ah (1975b) stated that two pulses which occurred within several hundred microseconds of each other were perceived as a single pulse with energy equal to the sum of the two pulses. The human studies cited above indicate that the highest effective fre- quency of RF hearing is between 6.5 and 8.9 GHz and that the lowest effective frequency is 216 MHz. Also, the results describe other radiation parameters (peak power density, energy density per pulse, and pulse width) that are important in determining the threshold for RF hearing in humans. 5-191 ------- 5.6.3.3 Dependence of RF Hearing on Acoustic Hearing-- Standard audiograms measure hearing thresholds for air conduction at acoustic frequencies of 250 to 8000 Hz and for bone conduction to 4000 Hz. Cain and Rissman (1978) measured the hearing ability of eight subjects over the frequency range of 1 to 20 kHz in addition to determining their standard audiograms. They found that, although there was no apparent correlation between the ability to perceive pulsed microwaves at 3000 MHz and hearing ability as measured by standard audiograms, there was a strong correlation between microwave-hearing threshold and hearing thresholds to air-conducted acoustic signals above 8 kHz. For example, three of the subjects who had normal hearing below 4 kHz could not hear microwave pulses of less than 20-ps duration under conditions in which the other subjects could perceive RF sounds. All three had a hearing deficit at frequencies above 8 kHz. Frey (1961) compared human acoustical hearing and RF hearing and reported that a necessary condition for perceiving the RF sound was the ability to hear audiofrequencies above approximately 5 kHz, although not necessarily by air conduction. This conclusion was based on one subject who had normal air- conduction hearing but failed to hear the microwave pulses. The person was subsequently found to have a substantial loss in bone-conduction hearing. On the other hand, a subject with good bone-conduction hearing but with poor air-conduction hearing perceived the RF sound at approximately the same power density that induced threshold perception in subjects with normal hearing. The studies by Cain and Rissmann (1978) and by Frey (1961) show RF hearing to be dependent on high-frequency hearing above 8 kHz and bone-conduction hearing at lower acoustic frequencies. 5-192 ------- 5.6.3.4 Similarity of Auditory Response to Microwave and Conventional Acoustic Stimuli-- Taylor and Ashleman (1974) measured the electrical response in three successive levels of the cat auditory nervous system (eighth cranial nerve, medial geniculate nucleus, and primary auditory cortex) to both acoustic and pulsed-microwave (2450-MHz) stimuli. They concluded that the microwave- induced auditory effect on the animal is exerted similarly to that of con- ventional acoustic stimuli. Furthermore, these authors reported that inactivation (perforation of the round window and aspiration of perilymph) of the cochlea, the known first stage of transduction for acoustic stimuli, affected the central nervous system (CNS) response to acoustic and microwave energy in the same way; i.e., the evoked electrical activities of all three sites were abolished by cochlear destruction. These results indicated that the locus of the initial interaction of pulse-modulated microwave energy with the auditory system might reside peripheral to the cochlea. In an experiment in which the thresholds of evoked electrical responses from the medial-geniculate body in cats were determined as a function of back- ground noise, Guy et aK (1975b) found that as the noise level (50 to 15,000-Hz bandwidth) increased from 60 to 80 dB, there was only a negligible increase in the threshold for the 2450-MHz microwave stimuli, a moderate increase in the threshold for a piezoelectric bone-conduction source, and a large increase in the threshold for loudspeaker-produced stimuli. The finding that the evoked response to microwave stimuli did not increase in relation to background 5-193 ------- noise, which included acoustic frequencies to 15,000 Hz, indicates that microwaves may interact with the high-frequency portion of the auditory system. Guy et tH. (1975b) also demonstrated that potentials evoked by microwave stimuli could be recorded at CNS sites other than those that correspond to the auditory nervous system. This finding indicates that elicited potentials recorded from any CNS location could be misinterpreted as indicating a direct microwave interaction with the particular system in which the recording is made. Prior to 1970, Frey (1962) had suggested that RF hearing might be a result of direct cortical or neural stimulation. He based this suggestion on (1) his observations that the perception of RF pulses was instantaneous and occurred at low average-incident-power densities and on (2) the failure to record cochlear microphonics at power densities much, higher than those required to elicit auditory nerve responses. Cochlear microphonics are electrical potentials that mimic the sonic waveforms of acoustic stimuli; they are the signature of mechanical distortion of cochlear hair cells, the first stage of sound transduction. The failure to observe microwave-induced cochlear microphonics had led to the suggestion that pulsed microwaves, unlike conventional acoustic stimuli, may not act on any sensor prior to acting directly on the inner ear apparatus (Frey 1967). 5-194 ------- In 1975, Chou et a_L reported their success in overcoming the technical problems that had prevented investigators from recording cochlear microphonics from microwave-irradiated animals. The cochlear microphonics of guinea pigs exposed to 918-MHz (PW) radiation were found to be similar to those evoked by acoustic stimuli. The results of the above studies of evoked electrical potentials in the auditory system, including the demonstration of pulsed-microwave-evoked cochlear microphonics, strongly indicate that the microwave-induced auditory sensation is detected similarly to conventional sound detection and that the site of conversion from microwave to acoustic energy resides peripheral to the cochlea. However, it is not known what structure(s) in the head transduce(s) the microwave energy to acoustic energy. 5.6.3.5 Mechanism of RF Hearing— As mentioned above, Frey (1967) had suggested that RF hearing might be a result of direct cortical or neural stimulation because of the failure to record cochlear microphonics and because the perception of RF pulses was instantaneous and occurred at low average-incident-power densities. The latter points were evidence against a radiation-pressure/bone-conduction hypothesis (see also Guy et aK 1975b). Sommer and von Gierke (1964) had suggested that radiation pressure exerted by the RF pulse impinging on the surface of the head could launch an acoustic signal of sufficient amplitude to be detected by the inner ear via bone conduction. Other types of pressure much greater than radiation pressure can be produced in tissue exposed to RF 5-195 ------- pulses. These include thermal expansion forces that are proportional to the square of the electric field in the material. For example, Gournay (1966) has shown that pressures greatly exceeding radiation pressure result when visible light from a laser is converted to acoustic energy by thermal expansion due to absorbed energy in various liquids. Foster and Finch (1974) extended Gournay1s analysis to a physiological solution exposed to microwave pulses similar to those that produce RF hearing in humans. They showed both theoretically and experimentally that radiation- induced pressure changes would result from the absorption of RF pulses and could produce significant acoustic energy in the solution. In fact, audible sounds were produced by rapid thermal expansion, resulting from only a 5 x 10 k °C temperature rise in the physiological solution, due to the absorption of the energy in the RF pulse. The authors concluded that the RF sounds involve perception, via bone conduction, of the thermally generated sound transients caused by the absorption of microwave pulses. The pulses must be moderately intense (typically 500 to 5000 mW/cm at the surface of the head). However, they can be sufficiently brief (< 50 ps) such that the maximum increase in tissue temperature after each pulse is very small (< 10~5 °C). The hypothesis of Foster and Finch (1974) predicts that the RF hearing effect is related to thermoelastically induced mechanical vibrations in the head. Vibrations of this type can be produced by other means such as a laser 5-196 ------- pulse or by a pulsed piezoelectric crystal in contact with the skull (Chou et al. 1976). Frey and Coren (1979) used a holographic technique to test whether the skull and the tissues of the head of an animal have the predicted vibra- tions when exposed to a pulsed RF field. No displacements were recorded, but, subsequent to this report, Chou et a2- (1980) demonstrated that the sensi- tivity of the holographic technique used by Frey and Coren (1979) was three to four orders of magnitude too low to detect displacements related to vibrations from microwave-induced thermoelastic expansion in biological tissues. Tyazhelov et al_. (1979b) conducted a series of psychophysical experiments with 18 subjects to evaluate the adequacy of the thermoelastic hypothesis and to study the perceptual qualities of RF-induced sounds. Audiofrequency sig- nals were presented alternately to or concurrently with microwave pulses (see Table 5-16) under conditions in which the subject could adjust the amplitude, frequency, and phase of the audio signal. The authors concluded that the thermoelastic hypothesis adequately explained some of their findings for microwave pulses of high peak power and short width (< 50 ps), but other results were interpreted as inconsistent with a thermoelastic mechanism for RF hearing. For example, pulse widths greater than 50 ps, which increased the mean power level, produced increases in loudness that rose more rapidly than predicted by the thermoacoustic model. In addition, suppression of RF sounds by the audio signal was reported to be inconsistent with the model. 5-197 ------- Lebovitz and Seaman (1977) compared the response to pulsed microwaves (915 MHz) with the response to acoustic clicks by monitoring single auditory neurons in the cat. A response of these neurons to pulsed microwaves was predicted by earlier studies that demonstrated subjective auditory perception (Frey 1962), auditory evoked potentials (Taylor and Ashleman 1974), and cochlear microphonics (Chou et al^. 1975). Furthermore, the thermoelastic model predicts that a mechanical wave of pressure stimulates the inner ear via bone conduction. Thus, the response of the neurons in the auditory pathway to pulsed microwaves should be similar to their response to transient mechanical stimuli such as acoustic clicks. The results indicated that mechanical factors within the cochlea are similarly involved in determining both the acoustic and the pulsed microwave response (Lebovitz and Seaman 1977). Other data in this report (Lebovitz and Seaman 1977) appeared to be inconsistent with the thermoelastic model that predicts a high-frequency component such as the microwave-induced cochlear microphonic recorded by Chou et ak (1975). That is, Lebovitz and Seaman (1977) observed a decrease in sensitivity of high-frequency auditory units to microwave pulses. However, they used long pulses of 250 to 300 ps in duration to obtain maximal energy per pulse. More recently, Tyazhelov et al^ (1979b) reported that long pulses (~ 100 ps) result in a lower pitch of the RF sound in humans. Two of their observers who had a high-frequency auditory limit of 10 kHz could not hear short RF pulses but could hear long pulses. Thus, the results of single unit recordings in cats are consistent with human perception of RF pulses when the pulse widths are long. 5-198 ------- Lin (1977) developed a theoretical model that estimates the characteristics of acoustic signals induced in laboratory animals and humans by microwave pulses; his model is based on thermal expansion in spherical heads irradiated by pulsed microwave energy. The frequency of the induced sound was found to be a function of head size and of acoustic properties for brain tissues; hence, the acoustic pitch perceived by a given subject will be the same regardless of the RF frequency. The calculations of Lin show that the fundamental frequency predicted by the model varies inversely with the radius of the head; i.e., the larger the radius, the lower the frequency of the perceived RF sound. He estimated the fundamental frequency of vibration in guinea pigs, cats, and adult humans to be 45, 38, and 13 kHz, respectively. The frequency for an infant head was estimated to be about 18 kHz. These results appear to be in good agreement with the measurements of Chou et al. (1975), who found cochlear microphonics in guinea pigs to be 50 kHz. In addition, the calculations provide further evidence that a necessary condition for auditory perception by adult humans is the ability to hear sound above 5 to 8 kHz (Frey 1961, Rissmann and Cain 1975). However, the prediction that the pitch perceived by a given subject will be the same regardless of the RF frequency is not in agreement with the experimental finding that long pulses result in a lower pitch of the RF sound in humans (Tyazhelov et al_. 1979b). Other unresolved questions are discussed below. 5.6.4 Unresolved Questions Several investigators have tried to determine the thresholds for the RF-induced auditory sensation in human beings and in laboratory animals 5-199 ------- (Table 5-17). Human studies and animal experiments, in general, are few and have used small sample sizes, different frequencies, and different experimental procedures. However, the radiation parameters that are most important in determining the threshold for the auditory response in humans (discussed in § 5.6.3.2, Effective Radiation Parameters) and in laboratory animals (discussed below) are being characterized. The threshold for an auditory response in cats exposed to 918- and 2450-MHz (PW) radiation was studied by recording electric potentials from the medial geniculate body (Guy et ah 1975b). As the pulse width increased from 0.5 to 32 \|js, the threshold value for the peak of power density decreased proportion- ately for pulse widths below 10 \|js. The thresholds of average power density and energy density per pulse also increased with pulse width, but these para- meters did not show the strong proportional relationship with pulse width as did the peak of power density. Guy et al^. (1975b) concluded that the thresh- old for the evoked auditory response was related to the incident energy density per pulse, at least for pulse widths less than 10 jjs. The threshold energy density per pulse was found to be about one-half of that which had produced a sensation in one human subject. Later, one of the coauthors (Lin 1978) stated that one cannot easily rule out the possible connection between the pulsed- microwave-evoked auditory responses and other radiation parameters, including peak incident power density, absorbed energy, and pulse width. In a similar experiment conducted at frequencies between 8670 and 9160 MHz, Guy et sH. (1975b) found that the threshold values of power density and of energy density per pulse, which include the auditory response of cats, were 5-200 ------- TABLE 5-17. SUMMARY OF STUDIES CONCERNING THRESHOLD VALUES FOR AUDITORY-EVOKED POTENTIALS IN LABORATORY ANIMALS Exposure Conditions Effect Species (n) Frequency (MHz) Pulse Repetition Rate (s-1) Pulse Width (ps) Pulse Intensity (raW/cio2) Average Intensity (mW/craz) Incident Energy Density Per Pulse (pj/cm2) Peak Absorbed Energy Density per Pulse (pj/g) References Response obtained with scalp electrodes Cat (2) 3000 0 5 5 10 15 2200,2800 1300 580 11,14 13 8 7 Cain and Rissmann (1978) Response obtained from round window with carbon lead Guinea" pig (5) 918 100 1-10 A A 20 Chou et a\| (1975) Response obtained from medial geniculate with glass electrode Cat (2) 918 2450 8670-9160 1 1 1 3-32 0 5-32 32 800-5800 600-35600 14800-38800 0 017-0 028 0 015-0 047 0 472-1 24 17 4-28 3 15 2-47 0 472-1240 12 3-20 0 8 7-26 7 Guy et a^ (1975b) Response obtained Cat from individual (not auditory neurons given) with glass electrode 915 < 10 25-250 <10 4-40 Lebovitz and Seaman (1977) 'Direct comparison of power density in the waveguide exposure system to free-field power density is improper because the efficiency of energy coupling is ten times higher than for free-field exposure (See Chou et al[ 1975, p 362 ) ------- an order of magnitude higher than those required at 918 and 2450 MHz (Table 5-17). Furthermore, no auditory response was obtained at 8670 to 9160 MHz until the brain was exposed by enlarging the hole in the skull that served as the electrode access port. Cain and Rissmann (1978) determined the threshold for auditory responses in animals exposed to 3000-MHz (PW) radiation. Although their results are in general agreement with Guy et a\|. (1975b) in that the threshold energy density per pulse was relatively constant for pulse widths of 5, 10, and 15 ps, their results are confounded by the use of scalp electrodes and only few animals of three species (two cats, two chinchillas, and a dog). Lin (1978) concluded that the studies on microwave-induced auditory sen- sations in laboratory animals strongly indicate a threshold, but the exact numerical value must await further experimentation. Furthermore, if the available threshold data are analyzed as a function of microwave frequency, it becomes clear, both in terms of the peak of power density and peak rate of energy absorption, that the threshold differs for different frequencies even in the same animals (Lin 1978). Recently, Wilson et ah (1980) described an autoradiographic technique 14 using [ C]2-deoxy-D-glucose to map auditory activity in the brain of rats exposed to acoustic stimuli and to PW and CW microwave radiation. With this technique, i_n vivo determination of metabolic activity, i.e., glucose utili- zation and associated functional activity in the brain, can be visualized. 5-202 ------- Prior to exposure to the acoustic stimuli or to microwaves, one middle ear was ablated. First, the authors showed the expected bilateral asymmetry of radio- active tracer uptake in the auditory system of rats exposed to acoustic clicks or weak background noise. Second, in contrast, a symmetrical uptake of tracer was found in animals exposed to PW radiation. Thus, the autoradiographic results confirmed the finding that RF hearing does not involve the middle ear (Frey 1961; Chou and Galambos 1979). Unexpectedly, Wilson et aK (1980) found similar patterns of radioactive tracer uptake in the auditory system of rats exposed to CW radiation (918 MHz; 2.5 and 10 mW/cm^) and to PW radiation (2450 MHz). This result, indicating an auditory response to CW microwaves, was unexpected, because no report of a direct hearing sensation due to exposure to CW microwaves had appeared in the literature. Since the thermoelastic hypothesis of RF hearing is based on the properties of PW radiation, this observation suggests that another mechanism may be involved in the interaction of RF radiation with the auditory system of the brain. 5.6.5 Human Cutaneous Perception Exposure of the human body to microwave radiation can cause heating that is detectable by the temperature sensitive receptors in the skin. As shown in Table 5-18, several investigators have experimentally determined the microwave intensities that cause sensations of warmth and pain in human subjects. 5.6.5.1 Frequency Specificity— 2 Hendler and colleagues (1963, 1968) exposed a circular area (37 cm ) of the forehead to 3- and 10-GHz (PW) radiation and to infrared (IR) radiation. 5-203 ------- TABLE 5-18. SUMMARY OF STUDIES CONCERNING HUMAN CUTANEOUS PERCEPTION OF RF RADIATION Effect Frequency (MHz) Intensity (mW/cm2) Duration (s) SAR* (W/kg) Reference Sensation of warmth on forehead 3000 (PW) 10,000 (PW) — 1-5 1-5 20-40 25-35 Hendler (1968) Hendler et al. (1963) Sensation of warmth on forehead 2880 74 56 15-73 50-180 Schwan et al^ (1966) Sensation of warmth on inner forearm 3000 (PW) 300-2500 1 6 — Vendrick and Vos (1958) Sensation of pain on inner forearm 3000 2500 1000 30 130 2000 Cook (1952) SAR value is estimated. The forehead was selected for a study of warmth sensations because the temperature receptors in the skin of the forehead are relatively numerous and are evenly distributed, so that the area constitutes a low-threshold region of uniform temperature sensitivity. The lower-frequency microwaves (3 GHz) had a higher intensity threshold for warmth sensation. The higher-frequency IR radiation was more effective than either microwave frequency, because the IR energy was absorbed more effectively in the outer skin layers containing the thermal sensors. 5-204 ------- 5.6.5.2 Temperature Thresholds-- For both microwave and IR radiation at intensities producing warmth sensations, a threshold of warmth was experienced when the temperature of a more superficial layer of subcutaneous tissue ~ 0.2 mm below the skin's surface was increased ~ 0.01 to 0.02 °C over the temperature of a deeper layer in the skin lying ~ 1 mm below the surface. In this study it was also noted that there was a persistent sensation of warmth for ~ 7 s after cessation of the exposure, which indicated the continued existence of an effective tempera- ture difference between the subcutaneous tissue layers (Hendler 1968). Schwan et a2- (1966) exposed a small area of the forehead (7-cm diameter) equivalent to the area exposed in Hendler's studies to 2.88-GHz radiation and measured the length of time that elapsed before the person was aware of a sensation of warmth. The times for four subjects varied from 15 to 73 s at 2 2 74 mW/cm and from 50 to 180 s at 56 mW/cm . The authors found that the reaction times were not linearly proportional to the reciprocal of the incident power density and concluded that subjective awareness of warmth was not a reliable indication of microwave hazard. 2 Vendrik and Vos (1958) exposed a 13-cm area of the inner forearm to 3-GHz (PW) radiation (300 to 2500 mW/cm^) and found the threshold for tempera- ture changes to be 0.4 to 1.0 °C. Skin temperature increases that were kept below 1 °C were linear with microwave intensity for six exposure durations. In contrast to the regularity of skin temperature changes induced by microwaves, the reports of temperature sensations were variable. Sensations of warmth 5-205 ------- occurred < 0.5 to 3.5 s after rapid rises in skin temperature. The sensations did not cease when the skin temperature began to drop. In this study, microwave radiation (3 GHz) was found to be a factor of 10 less effective in producing a temperature elevation than was IR radiation at a similar intensity. Cook (1952) determined the pain threshold in six subjects who were exposed to 3-GHz radiation at five different sites on the body's surface. The initial skin temperature ranged from 31.5 to 33.5 °C. Pain resulted when a critical skin temperature (~ 46 °C) was reached rather than from a critical temperature 2 rise (AT). For an inner forearm area of 9.5 cm , the power density pain 2 2 thresholds varied from 2500 mW/cm for a 30-s exposure to 1000 mW/cm for an exposure of 130 s. The pain threshold was lower for a larger exposed area of 2 53 cm . The skin temperature corresponding to burning pain was found to be independent of the area of exposure, radiation intensity, exposure time, and anatomical site. At high intensities, the exposure time needed to produce pain was an inverse function of radiation intensity. The author reported that the sensations of warmth and pain with microwave heating differed little from those resulting from IR radiation. 5.6.6 Unresolved Questions The few studies on pain and warmth sensations in human beings exposed to frequencies in the range of 3 to 10 GHz provide useful data on exposure levels that are clearly undesirable for the general population. Cutaneous perception, however, may be a reliable indicator of an unsafe exposure level only at RF 5-206 ------- frequencies with wavelengths small in comparison to the length of the exposed body, i.e., wavelengths comparable to or smaller than the thickness of skin. Under these conditions, most of the energy is absorbed in the outer tissue layers containing thermal sensors. At lower frequencies, which have wavelengths approximately equal to or longer than the human body, modeling studies have shown that much of the energy is absorbed within the body below the superficial skin layers. One should make note of the fact that, in all mammals tested, the threshold temperature (~ 42 °C) of cellular injury for sustained elevations (seconds to tens of seconds) is below the threshold (~ 46 °C) of pain. These results strongly indicate that cutaneous perception of RF energy is not a reliable sensory response that will protect against potentially harmful levels of RF radiation over the broad frequency range of 0.5 MHz to 100 GHz. 5-207 ------- 5.7 OTHER PHYSIOLOGICAL AND BIOCHEMICAL EFFECTS Charles G. Liddle 5.7.1 Clinical Chemistry and Metabolism An individual's response to many stresses manifests itself through changes in some of the clinical chemistry indices. Serum calcium and phos- phate levels normally increase initially and then decrease below normal in response to stress; whereas serum glucose, blood urea nitrogen, and uric acid levels increase following stress. These blood chemical responses are con- sistent with induction of release of adrenocortical hormones in response to stress. Physiological responses, including thermal responses, to RF-radiation exposure may occur at exposure levels too low to produce large changes in thermoregulatory behavior or colonic temperature. In such instances there may be changes in various metabolic parameters. There are few clinical chemistry and metabolism reports where exposures were sufficiently defined to relate results to the power density or SAR. These studies are described below, and are outlined in Table 5-19. A study of the effects of microwaves on serum chemistry values was reported by Lovely et ah (1977), who exposed rats in a circularly polarized waveguide to 918-MHz (CW) radiation. The animals were irradiated 10 h/day for 2 13 weeks at a power density of 2.5 mW/cm (SAR ~ 1.04 W/kg). They reported no change in Na+, K+, ion gap, CI , blood urea nitrogen, or glucose values compared to sham-irradiated animals. They did report a significant difference in the serum calcium values at the end of the 12-week exposure period. 5-209 ------- TABLE 5-19. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON CLINICAL CHEMISTRY AND METABOLISM, ENDOCRINOLOGY, AND GROWTH AND DEVELOPMENT Exposure Conditions Effects Species Frequency (MHz) Intensity (raW/cmJ) Duration (days x mln) SAR (W/kg) At (°C) Reference No effect on serum chemistry values Rat 918 (CW) 2 5 91 x 600 1 0 0 Lovely et a^ (1977) Increase In serum glucose Rabbit 2450 (CW and PW) 5, 10, 25 1 x 120 0 8-4 0 (est) 0. 0, 0, 0, 1 7 (PW) 2 9 (CW) Wangemann and Cleary (1976) Increase in blood urea nitrogen Rabbit 2450 (CW) 25 1 x 120 4 0 (est) 2 9 No increase 1n blood urea nitrogen Rabbit Rabbit 2450 (CW) 2450 (PW) 5 and 10 5, 10, 25 1 x 120 1 x 120 0 8, 1 6 (est) 0 8-4 0 (est) 0 0, 0, 1 7 Increased In uric acid values Rabbit 2450 (CW and PW) 10, 25 Neg 5 1 x 120 1.6, 4 0 (est) Neg 0 8 0, 1 7 (PW) 0, 2 9 (CW) 0 No effect on other serum chemistry values Rabbit 2450 (CW and PW) 5, 10, and 25 1 x 120 0 8-4 0 (est) 0, 0, 0, 0, 1 7 (PW) 2 9 (CW) Increased Iron and manganese levels In brain Rat 1600 (CW) 80 1 x 10 48 (est) 4 5 Chamness et a2 (1976) Decrease in specific metabolic rate (Ambient T = 24 °C) House 2450 (CW) ... 1 x 30 10.4 (Neg 5 5) Ho and Edwards (1977b) Increase in specific metabolic rate (Ambient T = 35 °C) Mouse 2450 (CW) ... 1 x 30 8 6 (Neg 3 6) Ho and Edwards (1979) Increased NADH fluorescense Decreased ATP Rat (exposed 591 (CW) 13 8 1x05 0 36-2 2 0 Sanders et al (1980) l ro Decreased CP brain) (continued) ------- TABLE 5-19. (continued) Exposure Conditions Effects Species Frequency Intensity Duration SAR At Reference (~Wz) (iiW/cb2) (days x mm) (W/kg) (°C) (Ji l ro Decreased ATP Decreased CP No effect on blood coagulation Increased thyroxine and tri- iodothyronine No effect on thyroid gland or thyroid hormone No effect on thyroid function Decrease in seruoi-protein-bound iodine, thyroxine and thyroxine/ serum ratio Increase in thyroid hormone Decrease in serum thyroxine levels Increase in corticosterone levels Decrease in thyrotropin levels Increase in corticosterone levels Decrease in thyrotropin levels Rat (exposed brain) Human plasma Dog Rat Rat Rat Rabbit Rat Rat 591 (CV) 2450 (CW) 2450 (CW) 2450 (CW) 2450 (CW) 2450 (CW) 3000 (PW) 2450 (CW) 5 10-280 72-236 1,10. 100 1. 10 10, 20, 25 15 20 Neg 1-10 2450 (120 Hz 40-70 AM) Neg 1-20 10, 40-70 Neg 1-5, 20 1x05 1 x 30 1 x 120 1 x 10-45 56 x 480 1 x 240-960 2 5 x 1440 48 x 180 1 x 240-480 1 x 60-480 1 x 60 1 x 240 10-40 Neg 1-5 25, 40 5 2, 8 4 Neg 1-20 0 02-4 2 0 13-0 8 1 3-38 (est) 58-190 0 25-25 0 25-2 5 2.5, 5. 6 25 (est) 3 8 (est) 0 25-0 75 (est) 0 25-25 8 4-14 7 Neg 0 21-4 2 2 1-14 7 Neg 0 21-4 2 2 1-8 4 Neg 0 2-1 0 0 6-2 1 0-1 Not reported Not reported S 10 mW/cm2 = 0 100 mW/cm2 i 5 S 20 mW/cm2 = 0 25 mW/cm2 =17 Not reported 0-0 6 0 6-1 4 1 3-3 0 0-0 6 0-3 0 0-0 6 0 3-2 1 0 Sanders et a^ (1980) Boggs et a\| (1972) Magin et a2 (1977a,b) Hilroy and Hichaelson (1972) Parker (1973) Parker (1973) Baranski et a^ (1972) Lu et a! (1977) Lu et al (1981) (continued) ------- Effects Species No effect on thyroid, pituitary, Rat or adrenal gland weight or growth hormone levels No effect on thyroid, anterior Rat pituitary gland, adrenal, prostate or testes weights, no change in follicle stimulating hormone or ^ gonadotropic hormone levels N> Increase in leutinizing hormone Rat Increased adrenal weights and Infant significant adrenal responses rat Increased plasma corticosterone Rat levels No effect on serum corticosterone Rat levels No effect on weight gain Rat TABLE 5-19. (continued) Exposure Conditions Frequency Intensity Duration SAR (MHz) (mW/cm2) (days x mm) (W/kg) 2450 (CV) 1-20 1 x 60-480 0 25-25 2860-2880 (CW) 10 36 x 360 At Reference (°C) 0-1 4 Lu et al (1977) Not reported Mikolajczyk (1976) 1-2 (est) 2860-2880 (CW) 10 36 x 360 1-2 (est) Not reported Mikolajczyk (1976) 2450 (CW) 40 6x5 20-60 (est) 1 5-2 5 Guillet and Michaelson (1977) 2450 (CW) 50, 60 1 x 30-60 11 5-13 8 (est) 13 mW/cm2 =05 Lotz and Michaelson (1978) Neg 13-40 Neg 3 0-9 2 20 mW/cm2 =07 (est) 30 raW/cm2 =09-14 20-40 1 x 120 4 6-9 2 (est) 40 mH/cm2 =13-14 Neg 13 Neg 13 50 mW/cm2 =16-2 4 60 mW/cm2 =25-29 918 (CW) 2 5 91 x 600 10 0 Lovely et al (1977) 2450 (CW) 5 55 x 240 0 7-4 7 Not reported Smialowicz et al (1979a) (continued) ------- TABLE 5-19. (continued) Exposure Conditions Effects Species Frequency (MHz) Intensity (oW/cmJ) Duration (days x mm) SAR (W/kg) at (°C) Reference No effect on growth, neuro- logical or lmauoological develop- ment or mutagenicity Rat 100 KHz (CW) 46 112 x 240 2 8 Not reported Smialowicz et a^ (1981a) Possible decrease in brain acetylcholinesterase activity 37, 55 x 240 No effect on growth House 2450 (CW) 10 24 x 48 6-8 (est) Not reported HcAfee et al (1973) No effect on body weights Infant rat 2450 (CW) 40 6x5 20-60 (est) 1 5-2 5 Guillet and Hichaelson (1977) No effect on growth House 10 5, 19 27, 26 6 (CW) 8900 1 x 20 0 9, 1 8, 3.6 (est) Not reported Stavinoha et a^ (1975) House 19 17,000- 114,000 5 x 40 6 3 (est) No effect on weight gain House 148 (CW) 0 5 50 x 60 0 013 Not reported Lin et §2 (1979a) Neg = Effect not found at value indicated ------- However, this is most probably a spurious result because the calcium levels in the sham-irradiated animals were decreased from previous values, while the levels in the irradiated animals remained unchanged from earlier values. These results, therefore, mean that 918-microwaves at 2.5 mW/cm do not alter the measured clinical chemistry values. Colonic temperature measurements were made on the animals, and no detectable changes were observed. Somewhat differing from these results, Wangemann and Cleary (1976) reported an increase in serum glucose levels in rabbits exposed for 2 h to 2 2450-MHz microwaves (CW and PW) at power densities of 5, 10 and 25 mW/cm (SAR's estimated at 0.8, 1.6, and 4.0 W/kg, respectively), and an increased 2 blood urea nitrogen value in animals exposed to 25-mW/cm microwaves only. 2 2 Increased uric acid levels were found at 10 and 25 mW/cm but not at 5 mW/cm (both CW and PW). Levels of calcium, phosphorus, cholesterol, total protein, alkaline phosphatase, lactic dehydrogenase, and serum glutamic oxalic trans- aminase were unaffected at the three power densities. The authors stated that the PW and CW results could not be compared directly because the exposure con- ditions were different. The results were those that would be expected from 2 heat stress. Animals exposed at 25 mW/cm for 2 h showed a significant rectal temperature increase of 1.7 and 2.9 °C for PW and CW exposures, respectively. 2 Those exposed at 10 mW/cm showed evidence of mild heat stress, such as peripheral vasodilation, but no significant increase in rectal temperature. One possible explanation for the difference between these results and those of Lovely et al_. is the difference in absorbed energy patterns in rats and rabbits at 2450 MHz. 5-214 ------- Brains from rats exposed at 1600 MHz for 10 min at a power density of 80 mW/cm (SAR estimated at 48 W/kg) were analyzed for selected heavy metals by Chamness et aK (1976). Iron levels were increased in all the areas of the brain (hypothalamus, corpus striatum, midbrain, hippocampus, cerebellum, medulla, cortex). Manganese was increased in the cortex and medulla, and copper was increased in the cortex; while calcium, zinc, sodium, and potassium levels were unchanged. The observed changes were probably a result of hyper- thermia since most alterations seen were also observed in rats subjected to a hyperthermal environment, and the irradiated animals showed a rectal tempera- ture increase of 4.5 °C. Platelet-rich human plasma was irradiated i_n vitro with 2450-MHz micro- waves at power densities of 10 to 280 mW/cm^ (SAR's of 1.3 to 38 W/kg) for 0.5 to 24 h by Boggs et aK (1972), and effects on blood coagulation were analyzed. They reported no significant changes in platelet count, coagulation time, or clot strength at these power densities. The plasma temperature was maintained below 37 °C during the exposures. They also conducted studies on the effect of heating on coagulation time and clot strength. Samples were heated either by microwaves or by radiant heating, and the results were com- pared to unheated control samples. The relative coagulation time for the microwave-heated samples remained unchanged throughout the temperature range studied (34, 37, 39 and 42 °C), while the samples heated by radiant energy showed increases in the relative coagulation time (1.58 and 2.03 times the non-heated samples) at 39 and 42°C, respectively. A similar, but reverse effect was seen in the relative clot strength. The samples heated to 42 °C by microwaves showed a decrease to 0.74 times the control samples, while the 5-215 ------- radiant-heated samples showed a decrease to 0.31 times the control values. A possible explanation for these differences is the difficulty of producing the same heating rate and pattern within a sample using different sources of heat. Ho and Edwards (1977b) measured the rate of oxygen consumption of mice exposed to 2450-MHz microwaves in a waveguide for 30 min at dose rates from 1.6 to 44.3 W/kg at an environmental temperature of 24 °C. They found a significant decrease in the specific metabolic rate (SMR) at 10.4 W/kg or higher, but not at 1.5 or 5.5 W/kg. There was no detectable rectal tempera- ture increase at or below 10.4 W/kg, but there was a 0.5 °C increase at 23.6 W/kg and a 1.0 °C increase at 44.3 W/kg. These results indicate that the mouse compensates for large dose rates of microwave energy by adjusting its metabolic rate downward to compensate for the thermal load. In a more recent report, Ho and Edwards (1979) reported a continuation of the previous study. Exposure conditions were identical to those of the pre- vious study except that environmental temperatures of 20, 30, and 35 °C were also used. At 20 °C they found a significant decrease during exposure in the SMR at 12.1 W/kg and above, but not at 6.0 W/kg or below. A significant decrease was found in the SMR during the 30-min post-exposure period at 45.1 W/kg, but not at 27.0 W/kg or less. For the exposures at 30 °C there was no change in the SMR at any of the dose rates used (1.4 to 23.7 W/kg) either during exposure or in the post-exposure period. , At 35 °C there was a signi- ficant increase in the SMR at 8.6 W/kg and higher, but not at 3.6 W/kg or lower during exposure, and no changes were observed during the postexposure period. The authors re-evaluated their first report (Ho and Edwards 1977b) by 5-216 ------- comparing the results during exposure to the pre-exposure values instead of comparing the values of the animals during exposure with those of sham-exposed animals. With this method of evaluation, the SMR values during exposure at 24 °C were not significantly different from the pre-exposure values. They re- ported that this method of comparison accounted for the lack of uniformity among animals at the beginning of each experiment and therefore was a better method of comparison. The general trends indicated by their studies are that microwave exposure at the lowest ambient temperature resulted in a reduction in the SMR, exposure at the highest ambient temperature resulted in an increase in the SMR, and no significant changes occurred at the intermediate temperatures. A possible interpretation of these trends is that at the lower ambient temperatures the animals are producing heat to maintain thermal neutrality; addition of the microwave heating reduces the animals' demand for additional heat. At the higher amibent temperatures the animals may be actively trying to dissipate the additional heat from the microwaves, thereby increasing their SMR. A study on the effects of microwaves on energy metabolism of the rat brain was reported by Sanders et aj. (1980). First, a small area of the brain of anesthetized animals was surgically exposed. Then a horn antenna was positioned so that exposures were in the far field and only the exposed sur- face of the brain was irradiated with the electric field parallel to the body axis. Animals were exposed to 591-MHz (CW) radiation for 0.5, 1, 2, 3, or 2 2 5 min at 13.8 mW/cm or for 0.5 or 1 min at 5 mW/cm . (Calculated SAR for 5-217 ------- 2 5 mW/cm = 0.13 W/kg using a 2-cm sphere model or 0.8 W/kg using a semi-infinite 2 plane model.) During exposures at 13.8 mW/cm they found an increase in nicotinamide adenine dinucleotide (NADH) fluorescence to a maximum of 4.0 to 12.5 percent above pre-exposure control levels. In addition, adenosine tri- phosphate (ATP) levels were significantly decreased at all exposure times, as 2 were creatine phosphate (CP) levels. At 5 mW/cm , ATP and CP levels were also significantly decreased following 0.5- and 1-min exposures. The ATP and CP 2 changes at 5 mW/cm were not significantly different from those seen at 13.8 2 mW/cm . There were no changes in rectal temperature at any of the exposures and no significant difference in brain temperature between exposed and sham- exposed animals. The authors concluded that the results (increased NADH fluorescence; decreased ATP and CP levels) support the hypothesis that RF radiation inhibits mitochondrial electron transport chain function and that the changes cannot be attributed to general tissue hyperthermia. In summary, changes in clinical chemistry values have been reported at dose rates as low as 0.8 W/kg in rabbits, and negative results have been reported at exposures as high as 1 W/kg in rats. The clincial chemistry changes that have been reported are those that would be expected from heat stress. In other studies, effects on the rate of oxygen consumption of mice have been reported at 9.6 W/kg but not at 5.5 W/kg, and changes in brain energy metabolism have been found at 0.13 to 0.8 W/kg (Table 5-19). 5-218 ------- 5.7.2 Endocrinology The endocrine system in coordination with the nervous system is a major regulatory system in the body. Alterations in the function of the neuroendocrine system often reflect the efforts on the part of the animal to maintain homeostasis when subjected to stressful internal and external stimuli. Detection of changes in the endocrine system is a sensitive means of analyzing the animal's responses either to direct stimulation of the endocrine organs themselves or to stimulation of the CNS. The thyroid gland plays a basic role in regulating basal metabolism as well as in generating metabolic heat in the tissues. Changes in thyroid activity can result from changes in thyroid- stimulating hormone from the hypothalamic-hypophyseal system and/or increased metabolic activity of the thyroid gland due to heating. Direct interaction with the CNS could also produce changes in thyroid activity. Moderate or gradual heating results in a reduction of thyroid hormone, while rapid or marked elevation of body temperature results in an increase in thyroid activity. The results of RF-radiation-effects studies on the endocrine system are discussed below and summarized in Table 5-19. Thyroid function of rats following exposure to 2.45-GHz microwaves at 10, 2 2 20, and 25 mW/cm (SAR's estimated at 0.25 W/kg per mW/cm ) for 4 or 16 h was studied by Parker (1973). No effects on thyroid gland function were found at 2 these exposures. However, at 15 mW/cm (SAR estimated at 3.75 W/kg), exposure for a longer period (60 h) was reported to produce a significant decrease in serum-protein-bound iodine, thyroxine and thyroid/serum iodine ratio. A 2 significant rectal temperature increase (1.7 °C) was reported at 25 mW/cm , but not at lower power densities. 5-219 ------- Magin et aJL (1977a, b) irradiated the surgically exposed thyroid gland of anesthetized dogs with 2450-MHz (CW) microwaves using a waveguide appli- 2 cator at power densities from 72, 162, and 236 mW/cm (SAR's from 58 to 190 W/kg) for 2 h. One thyroid was irradiated while the other was used as a control. Tissue temperatures of 39, 41, and 45 °C were maintained in the thyroid at the three power densities. They reported an increased release of thyroxine (TH) and triiodothyronine (T3) at all power densities, demonstrating that the thyroid gland in the dog can be stimulated directly by microwave heating. Milroy and Michaelson (1972), who exposed rats to 2450-MHz microwaves at 1, 10, and 100 mW/cm^ (SAR = 0.25 W/kg per mW/cm^) for single exposures of 10, 2 20, 30, and 45 min and at 1 and 10 mW/cm 8 h/day for 8 weeks, reported no effect on T3 levels, thyroxine levels, or on the uptake of radioactive iodine. 2 No rectal temperature increase was observed at 10 mW/cm or less. At 100 mW/ 2 cm there was a constant rise in rectal temperature throughout exposure, up to 42 °C at the end of the exposure period. Histopathological examination of the thyroid glands also showed no effect from the exposure. An increased production of thyroid hormone in rabbits as measured by in- 131 creased incorporation of I and increased radioactivity per gram of thyroid (verified by autoradiography) was reported by Baranski et al^. (1972). The animals were exposed for 3 h/day for 4 months to 10-cm (3-GHz, PW) micro- 2 waves at an average incident power density of 5 mW/cm (SAR estimated at 0.25 to 0.75 W/kg). They reported no increase in body temperature or thyroid temperature. (The pulse parameters were not given, so peak power density cannot be calculated.) 5-220 ------- Lu et a^. (1977) reported serum thyroxine levels in rats exposed to 2450-MHz microwaves at 1, 5, 10, and 20 mW/cm^ (0.25, 1.25, 2.5, and 5 W/kg) 2 for 1, 2, 4, or 8 h. Decreased thyroxine levels were observed at 20 mW/cm following 4- and 8-h exposures. The thyroxine values at shorter exposures and lower power densities were not significantly different from the sham 2 values, except for an increase after 4 h at 1 mW/cm , which appears to be a chance variation, since at both higher and lower power densities and at longer and shorter periods of exposure no effect was detected. There were small but 2 statistically significant rectal temperature increases at 1 mW/cm after 4 h, 2 2 at 5 mW/cm after 1 and 2 h, and at 10 mW/cm after 2 and 4 h of exposure. The increases were in the 0.2 to 0.56 °C range. The lack of correlation between power density, exposure time, and rectal temperature increase, along with the small rectal temperature change, suggests that the effects may not have been due entirely to the microwave exposure but possibly from the stress 2 of confinement. At 20 mW/cm , however, the rectal temperature was signifi- cantly elevated at all four exposure times and tended to increase with longer exposures. At 1 and 2 h the changes were small (0.64 and 0.54 °C), but larger (1.01 and 1.35 °C) increases occurred after 4 and 8 h. Serum corticosteroid 2 levels were decreased at 20 mW/cm after 8 h only, which the authors report as a shift in the circadian periodicity. Serum growth hormone measurements showed no change at any of the power densities or times reported. They also measured no change in the mass of the thyroid, pituitary, and adrenal glands following irradiation. In another study on the effects of microwaves on acute endocrine responses in rats, Lu et aJL (1981) exposed animals to 2.45-GHz microwaves (AM at 120 Hz) 5-221 ------- for 1 and 4 h, and measured colonic temperature, thyrotropin, and cortico- sterone levels immediately after exposure. The 1-h exposures were conducted 2 at power densities of 1, 5, 10, 20, 40, 50, 60, or 70 mW/cm , and the 4-h exposures at 0.1, 1, 5, 10, 20, 25, and 40 mW/cm^ (measured SAR = 0.21 W/kg 2 per mW/cm ). The results for the 1-h exposure showed an increase in colonic temperature with increasing power density with a significant increase at 2 2 20 mW/cm and above, but not at 10 mW/cm or below. Corticosterone also showed an increase with increased power density with evidence of a threshold 2 between 20 and 40 mW/cm . Thyrotropin values showed a decrease with increasing 2 power densities, significant results at 40 mW/cm and above, and equivocal 2 2 results in the 10 to 20 mW/cm range (10 mW/cm values were significantly 2 lower while those at 20 mW/cm were not). Similar trends were observed fol- lowing 4-h exposures, with a significant increase in colonic temperatures at 2 2 10 mW/cm and above, and no increases at 5 mW/cm or below with the exception 2 of one small group at 1 mW/cm . The trends for the increase in corticosterone levels and the decrease in thyrotropin levels were not significant; however, 2 the corticosterone level at 40 mW/cm was significantly increased, but not at 2 2 25 mW/cm or lower values. Also, the thyrotropin values at 25 mW/cm and 2 2 40 mW/cm were significantly decreased, whereas those at 20 mW/cm and below were not. Comparing the effects of 1- and 4-h exposures on the three param- eters, the change in colonic temperature was similar for both exposures; however, the stimulatory effect on serum corticosterone levels was less for the 4-h exposures than for the 1-h exposures, while the depression of serum thyrotropin was more pronounced after 4 h than after 1 h. The authors con- cluded that the hormonal changes probably represented a general nonspecific 5-222 ------- stress reaction that was related to the intensity and duration of the stress- ing agent, rather than to the nature of the agent itself. No change in thyroid weight was seen in a study by Mikolajczyk (1976), where rats were exposed to 2860 to 2880-MHz microwaves at 10 mW/cm , 6 h/day, 6 days/week for 6 weeks. The SAR is estimated to be 1 to 2 W/kg for a single animal exposure, but the animals were exposed close together in a box with dividers every 10 cm, which should give a somewhat higher SAR value. The author did find a significant increase in luteinizing hormone from the anterior pituitary gland but no change in follicle-stimulating or gonadotropic hormone levels. The weights of the anterior pituitary, adrenal, prostate, or testes were not affected by the exposure. Other investigators also have reported the effects of microwaves on adrenal function. An increase in blood-corticosteroid concentration above that which would normally occur at that time of day in the absence of a stimulus is considered by many to be an indicator of an animal's response to stress. This results when an internal or external stimulus, either chemical, physical, or emotional, excites neurons of the hypothalamus to release corticotropin-releasing hormone, which drives the pituitary to release adrenocorticotropic hormone (ACTH). This hormone then stimulates the adrenal cortex to secrete corticosteroids. Among the strongest stressful stimuli are surgery, anesthesia, cold, narcosis, burning, high environmental temperature, and rough handling or restraint. Additional studies examining the effects of microwaves on adrenal function are described below. 5-223 ------- 2 Lovely et a2. (1977) exposed rats to 918-MHz microwaves at 2.5 mW/cm (SAR ~ 1.04 W/kg), 10 h/day for 13 weeks, and observed no changes in serum corticosterone levels. In a study by Guillet and Michaelson (1977), rat pups 2 were exposed to 2450-MHz microwaves at 40 mW/cm (SAR estimated at 20 to 60 W/kg), 5 min/day for 6 days beginning at day 1 postpartum; no change in basal corticosterone levels was found. There was a significant adrenal response to the microwaves, but this was the same as seen following ACTH administration, indicating a stress reaction. Rectal temperatures of the exposed animals averaged 1.5 to 2.5 °C higher than those of the sham-exposed animals. Adrenal mass of the irradiated animals was significantly greater than those of the control animals. In another study, Lotz and Michaelson (1978) irradiated rats with 2 2450-MHz microwaves at power densities of 13, 20, 30, and 40 mW/cm for 30, 60, or 120 min and at 50 and 60 mW/cm^ for 30 or 60 min (SAR estimated at 0.21 2 W/kg per mW/cm ). Plasma corticosterone levels were increased at power 2 2 densities at or above 50 mW/cm , but not at 40 mW/cm or less, for the 30- and 60-min exposures. At the longest exposure time (120 min), increased levels 2 2 were seen at or above 20 mW/cm but not at 13 mW/cm . Graphs were presented of the rectal temperatures taken at the completion of exposures; estimates of the rectal temperature increases taken from the graphs for 50 mW/cm were 1.6 O and 2.4 °C and for 60 mW/cm were 2.5 and 2.9 °C for 30 and 60 min, respectively. 2 The temperature increases for 40 mW/cm were 1.3 and 1.4 °C for 30 and 60 min. 2 For the 120-min exposures, rectal temperature increased 0.5 °C at 13 mW/cm and 0.7 °C at 20 mW/cm^. 5-224 ------- In summary, microwave effects on thyroid function have been reported at SAR's as low as 2.1 W/kg, and negative results have been reported at values as high as 25 W/kg. The duration of exposure as well as the exposure rate appears to be important, as demonstrated by the study of Parker (1973), where exposures at 6.25 W/kg for 16 h produced no changes, while 3.75-W/kg exposures for 60 h resulted in a decrease in serum thyroxine levels. Changes in corti- costerone levels have been reported from microwave exposure at SAR's as low as 10 W/kg but not at 6.25 W/kg, and adrenal responses have been reported at levels as low as 4.6 W/kg but not at 3 W/kg. 5.7.3 Growth and Development Few investigators have reported the effects of a combination of pre- and postnatal exposure or postnatal exposure only to RF radiation on the growth of laboratory animals. Smialowicz et al^. (1979a) exposed rats for 4 h per day, beginning on day 6 of gestation through 40 to 41 days postpartum to 2450-MHz (CW) microwaves at 5 mW/cm (SAR's = 0.7 to 4.7 W/kg), and reported no dif- ference between the weight gains of the exposed and sham-exposed animals. In another study, Smialowicz et a^. (1981a) reported the growth and development of rats exposed to 100-MHz (CW) microwaves at an incident power density of 46 mW/cm (average SAR =2.8 W/kg) for 4 h/day from day 6 of gestation through 97 days of age. There was no consistent difference between the body weights of the exposed and sham-exposed animals, though the exposed animals tended to be larger than the sham-exposed animals. Some of the animals were tested for neurological development, and no differences were observed in 5-225 ------- the development of a startle response or righting reflex. There was a sig- nificant difference in the age of eye opening, with the mean age of eye open- ing in the sham-irradiated animals occurring almost one day later than the irradiated animals. The authors stated that the change probably did not represent an acceleration of eye opening, as the age of eye opening in the irradiated animals was similar to that normally seen in control animals in other experiments, but also mentioned that eye opening was delayed in the sham-irradiated animals. Tests for development of motor activity at 35 and 84 days of age showed no difference between exposed and sham-exposed animals. No difference was observed for complete blood counts, mitogen-stimulated response of lymphocytes, frequency of T- and B-lymphocytes, or antibody response to Streptococcus pneumoniae capsular polysaccharide between the ex- posed and sham-exposed animals. No mutagenic effect was observed on the sperm cells after 20 days using the dominant lethal assay. Seven regions of the brain were weighed at 22, 40, and 97 days of age, and there was a significant increase in the weight of the medulla in the irradiated animals at 40 days of age but not at 22 or 97 days of age. There were no differences in the weights of any other brain region. There were also no differences in the brain pro- tein concentrations for the seven regions. Brain acetylcholinesterase (AChE) activity, however, was reduced in the striatum and medulla at 22 days of age and in the midbrain at 40 days of age. No effects on AChE activity were noted in the other regions at 22 and 40 days of age, and no differences were noted at 97 days of age. The differences were small, and there appeared to be no pattern to the changes; the sample size was small (3 to 5 animals), and out of 21 comparisons made at the 5 percent confidence level, one significantly different result would be expected. Consequently, without replication with 5-226 ------- a larger sample size, it is difficult to ascribe these changes to the micro- wave exposure. The effect of postnatal exposure to microwaves on growth and development was studied by McAfee et al^ (1973). Weanling mice were exposed to 2450-MHz microwaves at 10 mW/cm , 2 min each hour for 24 days; no effect on animal growth was found. There was no elevation of body temperature in the exposed animals. The authors also stated that a previous study, in which a stimula- tory effect on growth from microwaves was claimed, was probably in error because of inaccuracies in weighing the animals (see Nieset et ah 1958). Guillet and Michael son (1977) exposed neonatal rats to 2450-MHz micro- waves at 40 mW/cm (SAR's at 20 to 60 W/kg), 5 min/day for 6 days beginning on day 1 postpartum. No effect on body mass was found, though the authors did report adrenal changes, as discussed in the previous section, § 5.7.2, Endocri nology. Stavinoha et a2- (1975) irradiated 4-day-old mice at a field intensity of 5800 V/m (8.92 W/cm^) for 20 min with 10.5-, 19.27-, or 26.6-MHz radiation, and observed no change in gr.owth rate to 22 days of age compared to the control animals (SAR's estimated at 0.9, 1.8, and 3.6 W/kg). Mice were also exposed to 19-MHz microwaves for 40 min/day for 5 days using a near-field 2 synthesizer to deliver an E field of 800 V/m (17 W/cm ) and an H field of 2 55 A/m (114 W/cm ); no change in body mass through 120 days of age was found (SAR estimated at 6.3 W/kg). 5-227 ------- In a chronic study, Lin et aj. (1979a) exposed 4- to 7-day-old mice to 148-MHz microwaves at 0.5 mW/cm^ (SAR = 0.013 W/kg), 1 h/day, 5 days/week for 10 weeks and reported no significant effect on weight gain. These studies indicate that RF exposures at SAR's < 8 W/kg have no effect on growth and development of animals exposed pre- and postnatally or post- natal ly only. Exposures at 20 to 60 W/kg postnatally for short exposure times also did not alter the growth rate of animals even though adrenal changes were seen. 5.7.4 Cardiovascular System 5.7.4.1 Whole-Body Exposures— Some of the initial studies of biological effects of microwave radiation were made on the cardiovascular system. Presman and Levitina (1962) reported that whole-body or ventral exposure to CW microwaves at 2400 MHz, 7 to 2 12 mW/cm , promoted a decreased heart rate (bradycardia) in the rabbit. However, exposing only the head to microwaves at the same power densities resulted in an increased heart rate (tachycardia). Kaplan et al_. (1971) attempted to replicate the tachycardia reported by Presman and Levitina by exposing the head of rabbits to 2400-MHz (CW) micro- 2 waves at an incident power density of 10 mW/cm (SAR estimated at 2 W/kg). The exposure conditions differed in that Kaplan et aK exposed the animals in the far field, while Presman and Levitina used near-field exposures. Kaplan et al. reported no change in heart rate. They then exposed rabbits at 5-228 ------- 2 increasing power densities (20, 40, 60, 80, and 100 mW/cm ) and measured respiration rate, body temperature, and heart rate. Respiration increased at 2 2 40 mW/cm and greater, and body temperature was elevated at 80 and 100 mW/cm p (0.5 °C at both power densities). Heart rate increased at 100 mW/cm only. Birenbaum et a\_. (1975) exposed the heads of unanesthetized rabbits to 2.4-GHz (CW) microwaves for 60 min at 20, 40, 60, and 80 mW/cm^ (SAR estimated at 3 to 12 W/kg), and found increases in heart rate, respiration rate, and subcutaneous temperature (lower back) at all four power densities, with greater increases at higher exposure levels. They also compared 2.8-GHz (CW and PW, o 1000/s, 1.3 ps) microwaves at 20 mW/cm for the same three parameters and found no differences in the responses to CW or PW irradiation. Phillips et aK (1975b) exposed adult rats to 2450-MHz microwaves pulsed at 120 Hz for 30 min (SAR's = 4.5, 6.5, and 11.1 W/kg) and measured colonic and skin temperatures and heart rate after exposure. They reported increases in colonic and skin temperatures with increased exposure levels immediately after irradiation. At the highest exposure, the colonic temperature dropped below normal at 1 h after exposure and remained subnormal for 4 h. There was no post-exposure effect on heart rate at 4.5 W/kg; however, there was a mild bradycardia at 6.5 W/kg and a more pronounced decrease in heart rate at 11.1 W/kg. They attributed the effect to the heat stress induced by the microwaves. 5-229 ------- Hamrick and McRee (1980) assessed the effects of body temperature on the heart rate of embryonic quail. They exposed the embryos to 2450-MHz (CW) microwaves for 5 to 10 min (SAR's = 3, 6, 15, and 30 W/kg) at incubation temperatures from 35 to 38 °C, and to 2450-MHz (PW) microwaves (10-ps pulses, varied from 10 to 50 pulses/s, SAR's at 0.3, 1.5, and 3 W/kg) at incubation temperatures of 35 to 39 °C. There were no significant differences between the heart rates of the exposed and control embryos in any of the groups at any of the temperatures used. They did observe that the embryonic heart rate increased ~ 23 beats/min for each °C rise in incubation temperature in the 36 to 39 °C range. 5.7.4.2 Isolated Heart Preparations— Although the data supporting a microwave-induced bradycardia in the in- tact animal are equivocal, some researchers have exposed isolated heart preparations and reported an effect of microwaves on heart rate. Tinney et al. (1976) attempted to determine the locus of microwave-induced bradycardia in the isolated turtle heart using 960-MHz (CW) microwaves. Although the heart was isolated from the CNS, it was still capable of responding to neuro- humoral agents. They found that exposing the isolated heart to microwaves at 2 8 mW/cm caused bradycardia. When the heart was treated with atropine, blocking the sympathetic system, tachycardia resulted. However, when the heart was treated with propranolol, blocking the parasympathetic system, the heart rate decrease was more significant than that following microwave exposure alone, indicating that microwaves may have affected the para- sympathetic reflexes. With both blocking agents added to the heart, microwave exposure had only slight effects on the heart rate. Tinney et a^. postulated 5-230 ------- that microwave exposure of the isolated heart equally enhances the release of acetylcholine and norepinephrine; however, since the activity of the former usually predominates over the latter (Johansen 1963), the net effect of micro- wave exposure is a decrease in heart rate. A similar drug-microwave interaction has been observed in the isolated rat heart by 01 sen et al^ (1977). They exposed the heart to 960-MHz (CW) microwaves for 4 min at 1.3 and 2.1 W/kg and observed a decrease in heart rate. The decrease was greater at 1.3 W/kg than at 2.1 W/kg. They also conducted studies at 2.1 W/kg with drugs to block the sympathetic and para- sympathetic nervous system and obtained the same results as Tinney et al. (1976). Paff et al_. (1963), working with the isolated embryonic chicken heart, were unable to detect changes in heart rate during exposure to 24,000-MHz fields. They did, however, detect effects on the electrocardiogram (ECG), 2 including abnormal P and T waves from 3-min exposures at 74 mW/cm . Frey and Seifert (1968) showed that 10-ps pulses at a carrier frequency of 1.425 GHz given at a synchronous period with the ECG (220 ms after the P wave) resulted in tachycardia or heart arrhythmia in the isolated frog heart. 2 2 The peak power density was 60 mW/cm (average power density -0.6 pW/cm ). Liu et a2. (1976) reported no effect on heart rate with isolated frog hearts or in hearts irradiated i_n situ in a similar study. The iji situ hearts were exposed to 100-ps pulses of either 1.42 or 10 GHz, and the isolated frog hearts were exposed to 100-ps pulses of 1.42 GHz. The pulse was delivered 5-231 ------- on the rising phase of the R-wave from the ECG, which was somewhat similar to, but not exactly the same as, the 200-ms delay following the P-wave used by Frey and Seifert. (The R-wave follows the P-wave by around 200 ms.) The peak and average power densities of 320 mW and 32 ^iW were also considerably higher than those used by Frey and Seifert. These factors, plus differences in the manner of preparing the isolated hearts (Liu et al_. curarized the frogs, whereas Frey and Seifert decapitated the frogs), make it difficult to compare the results of the two studies. Clapman and Cain (1975), however, tried to replicate the study of Frey and Seifert using similar pulse widths (10 ps), peak and average power densities 2 2 (60 mW/cm and 0.6 pW/cm ), carrier frequency (1.42 GHz), and method of isolating the frog heart; they reported no change in heart rate. When they conducted studies with a different peak power (5.5 W/cm ), frequency (3 GHz), and pulse widths (2 and 150 ps), they also found no heart rate changes. Clapman and Cain were able to produce an increased heart rate with 20-mA current pulses synchronized 200 ms after the P-wave peak. The results of microwave exposure on the cardiovascular system indicate that whole-body exposures of sufficient intensity to produce heating will also produce an increase in heart rate similar to that which would be expected from heating alone. In the isolated heart there appears to be a stimulation of the autonomic nervous system from microwave exposure at levels where little heat- ing would be expected (1 to 2 W/kg). Low levels of synchronized PW microwaves (0.6 to 32 mW/kg) apparently are ineffective in producing detectable alterations 5-232 ------- in heart rate. A summary of the RF-radiation effects on cardiac physiology is given in Table 5-20. 5.7.5 Calcium Ion Efflux Calcium ions play a prominent role in many biochemical and biophysical processes. They are involved in maintaining cellular membrane integrity and function, as a cofactor in many enzyme reactions, as a putative second mess- enger for the conduction of extracellular signals to the nucleus of the cell, and in neural tissue for excitation and for the secretion of transmitter substances at synapses. Several investigators have reported the effects of electromagnetic fields on the response of calcium ions in biological systems, principally brain tissue. Bawin et aK (1975) showed that a 20-min in vitro exposure of chicken brain tissue to a 147-MHz field, at a power density of 1 2 to 2 mW/cm , could cause enhanced efflux of calcium ions from the samples, but only if the field was sinusoidally amplitude modulated with frequencies of 6, 9, 11, 16, or 20 Hz. Maximal efflux was measured at 16 Hz. Modulation fre- quencies of 0, 0.5, 3, 25, or 35 Hz were ineffective. This result was corrob- orated and extended by Blackman et al_. (1979, 1980a), who showed that the 2 response occurred only within a narrow intensity range around 0.83 mW/cm . Sheppard et a^. (1979) confirmed this intensity response with a 450-MHz carrier wave amplitude modulated at 16 Hz; calcium efflux was enhanced at 0.1 2 ? and 1.0 mW/cm but not at 0.05, 2.0, or 5.0 mW/cm . Blackman and co-workers also examined the influence of 50-MHz fields, amplitude modulated at 16 Hz, and found that the response could be elicited 5-233 ------- TABLE 5-20. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON VARIOUS ASPECTS OF CARDIAC PHYSIOLOGY Effects Species Frequency (MHz) Exposure Conditions Intensity Duration SAR (mW/cm2) (total mm) (W/kg) Reference cn i ro CO Bradycardia develops after whole-body exposure along with hyperthermia Exposure to head promotes tachycardia, exposure to back raises respiratory rate but not heart rate Increased respiration Increased heart rate from dorsal exposure of the head Alterations in ECG (shortening of QT interval, increased height of T-wave, appearance of U-wave) No effect on heart rate that cannot be attributed to microwave heating Pulses synchronized with each R-wave do not affect heart rate Synchronized pulses with QRS complex causes increase in heart rate with some arrhythmias Rat Rabbit Rabbit Rabbit 72-h chick heart Quail embryo Frog Frog 2,450 (PW) 28 & 48 2,450 (CW and 20 PW) 2,400 2,400 24,000 (PW) 2,450 (CW and PW) 1,420-10,000 (PW) 1,425 (PW) 40-100 100 74 NA NA 0 6 jjW/cm2 30 60 20 20 5 to 10 100-ps pulses 10-MS pulses 6 5 & 11 1 Phillips et al (1975b) Birenbaum et al (1975) 8-20 Kaplan et al (1971) 20 Kaplan et a2 (1971) NG Paff et al (1963) 0 3-30 Hamrick and McRee (1980) 12 or 16 pW Liu et a! (1976) per heart NG Frey and Seifert (1968) (continued) ------- TABLE 5-20 (continued) Exposure Conditions un Effects Species Frequency (MHz) Intensity (mW/cn2) Duration (total mm) SAR (W/kg) Reference PO oo vn Low power levels cause bradycardia in the isolated turtle heart Turtle 960 (CW) NA 60 2-10 T inney et al^ (1976) Causes slight decrease in the isolated heart Rat 960 (CW) NA 5 to 10 1 3 and 2 1 Olsen et a! (1977) Synchronized exposures with ECG have no effect on heart rate Frog 1420-3000 (PV) 0 0006 2-, 10-, 150-ys pulses NG Clapman and Cain (1975) * NA = Not appltcable ------- 2 within two intensity regions (between 1.44 and 1.67, and at 3.64 mW/cm ) 2 separated by power densities of no effect including 0.72 mW/cm (Blackman et al. 1980b). The apparent discrepancy in the calcium ion efflux data at dif- ferent power densities at the three different carrier frequencies (50, 147, and 450 MHz) has been resolved by the finding that the efflux is dependent upon the electric field strength within the sample and not on the incident power density (Joines and Blackman 1980). The empirical model used to trans- form the incident power density to internal field strength has also been used to predict additional values of intensity that produced both alteration and no alteration in calcium ion efflux (Joines and Blackman 1980). It should be noted that the 147-MHz and 50-MHz exposures caused no generalized heating of the sample. The maximum temperature rise was calculated to be less than 0.0004 °C, and calculated SAR's at each carrier frequency were less than 0.0014 W/kg (Blackman et aK 1980b). Although outside the range of the frequencies of interest for this review, frequencies within the same 6- to 20-Hz region applied directly, without a carrier wave, were demonstrated to alter the efflux of calcium ions from chicken brain tissue (Bawin and Adey 1976). In this case, the efflux of cal- cium ions was reduced by the exposure. At the intensities used, there is no heating of the sample exposed to 6- to 20-Hz fields because energy absorption is very small. No change in calcium ion efflux from brain tissue has been reported by Shelton and Merritt (1981), who used procedures somewhat different from those described by Bawin et a^. (1975), Bawin and Adey (1976), and Blackman et a]_. 5-236 ------- (1979, 1980a). They irradiated brain tissue from the rat with 1-GHz radiation, 2 square-wave, pulse-modulated at 16 Hz (0.5, 1.0, 2.0, or 15 mW/cm ). In their original study, Bawin et aK (1975) also exposed skeletal muscle to the same exposure conditions that produced calcium ion efflux from chick brain tissue, but no effect was observed. An increase of calcium ion efflux from rat pancreatic tissue exposed at ? 2 mW/cm for 1 to 2.5 h at 147 MHz, sinusoidally amplitude modulated at 16 Hz (estimated SAR < 0.075 W/kg), has been reported by Albert et fH. (1980). However, the efflux was not accompanied by a change in protein secretion, which is normally associated with calcium mobilization in the pancreas. The authors attributed the lack of protein secretion to the relatively small volume of medium available to the tissue slices for normal metabolic activity. Although the efflux of calcium ion from chick brain tissue has been docu- mented in two laboratories, the mechanism of the effect and physiological significance are still highly uncertain. These points are addressed in more detail in Unresolved Questions. 5.7.6 Unresolved Questions No report has yet described a mechanism of action in sufficient detail to identify all the relevant conditions necessary and sufficient to unequivocally explain the calcium ion response in the brain or pancreas. The response to specific intensities is unusual and presently unexplained. Any critical 5-237 ------- evaluation of this phenomenon, particularly in brain tissue, must also con- sider corollary supporting studies with brain tissue from cats (Bawin and Adey 1977), cat brain tissue i_n vivo (Kaczmarek and Adey 1973), and specific brain rhythms in cats (Bawin et aj. 1973), as well as behavioral changes in monkeys (Gavalas-Medici and Day-Magdaleno 1976). This response to modulation of an RF-carrier wave or to sub-ELF signals alone may be a true field effect at a very low SAR and at biologically relevant frequencies, i.e., in the range of frequencies normally present in the electroencephalogram (EEG). Other areas where there are unanswered questions is in comparisons of CW vs. PW microwaves under identical exposure conditions. Such studies would help determine if the differences seen by Wangemann and Cleary (1976) were due to different exposure conditions or to the irradiation parameters (CW or PW). There is also a paucity of information on the effects of RF radiation at different frequencies, particularly at frequencies of environmental impor- tance. Studies at different frequencies would help to determine the reasons for differences in effects at similar SAR's. Such studies might help explain why Wangemann and Cleary (1976) reported serum chemistry changes in rabbits at 0.8 W/kg (2450 MHz), and why Lovely et al^. (1977) reported no change in serum chemistry values in rats at 1 W/kg (918 MHz). There are also data such as those by Boggs et al^. (1972) where the results from microwave heating to a predetermined temperature are different from those resulting from the same temperature produced by other means of 5-238 ------- heating. Perhaps there are differences in the uniformity of heating or the rate of heating that would account for these differences. Additionally, a study by Deficis et aK (1979) reported elevated serum triglyceride and P~1ipoprotein levels in mice exposed to 2450-MHz microwaves at power densities 2 2 of 1.5, 3.3, and 4 mW/cm , but not at 1 mW/cm . Because the exposures were conducted in a multimodal cavity, SAR values were not reported and cannot be predicted. If this study is repeated, particular attention should be given to dosimetry. An alternative is to make or report dosimetric measurements in the exposure system used. The reported effects of thyroid function at 3.75 W/kg for 60 h contrasted with no effect at 6.25 W/kg for 16 h (Parker 1973) suggests that the total amount of energy absorbed may also be an important consideration. Additional studies could define further the relative importance of dose rate com- pared with total dose. 5-239 ------- 5.8 GENETICS AND MUTAGENESIS Carl F. Blackman 5.8.1 Introduction Genetics is the branch of biology that deals with the heredity and variation of organisms. The biochemical basis of heredity lies in the sequence of bases found in the nucleic acids, deoxyribonucleic acid (DNA), and in a few cases, ribonucleic acid (RNA). All living cells have the biochemical machinery to detect the sequence of bases in the DNA and to transcribe sections of DNA in- formation into similar sequences of bases in RNA. The RNA molecules then move to other locations inside the cell where their information is translated into various series of amino acids joined together in sequences that were precisely defined in the original DNA molecule. These specifically arranged amino acids form proteins, some of which are enzymes. Enzymes, in turn, catalyze biochemical reactions that ultimately result in the growth and propagation of intact organisms. Hereditary (genetic) material can be either nucleic acids alone, as found in bacteria and viruses, or a nucleic acid in association with proteins, which form the chromosomes found in more complex organisms, including man. Heat is a physical agent that can disrupt genetic material by causing the temperature to rise above a normal physiological range. The effect of heat, or temperature rise, has been studied in many biological systems. As examples, heat has been shown to cause: physicochemical damage in isolated DNA preparations (Lindahl and Nyberg 1974; Ginoza et a2_ 1964; Ginoza and 5-241 ------- Miller 1965) and in bacteria (Pellon et ah 1980); chromosome changes in Drosophila melanogaster (Grell 1971) and in the locus (Buss and Henderson 1971), including a change from dipolid to haploid in pollen from maize (Mathur et ah 1980); enhanced sensitivity to other agents in mammalian cell cultures (Ben-Hur et ah 1974) and in D. melanogaster (Mittler 1979); reduced fertility in rats (Bowler 1972; Fahim et ah 1975); and mutations in bacteriophages (Bingham et ah 1976), in bacteria (Zamenhof and Greer 1958) and in D. melanogaster (Muller and Altenburg 1919). Because absorbed RF energy is usually dissipated as heat, all reports of genetic and mutagenic changes caused by exposure to RF radiation must be examined closely to determine whether temperature rise, or some other mechanism, is the causative agent. Mutations are relatively permanent changes in the hereditary material involving either a physical change in chromosomal relations or a biochemical change in the sequence of nucleic acid bases that make up the genes. These changes can be passed on to future generations of cells. Two different types can be affected: germ cells, which are egg or sperm or their antecedent cells; and somatic cells, which form all other tissues in the body. Mutations in germ cells can be passed on to future generations of the organism, whereas mutations in somatic cells of an organism may lead to impairment of organ function and, in the extreme case, to cancer. Most mutations are considered harmful because they disturb the biochemical processes necessary for the sur- vival of the organism and for the propagation of the species. Evaluation of the mutagenic potential of RF radiation has generally fol- lowed the procedures established for testing the mutagenic activity of chemicals 5-242 ------- (Hollaender 1971). This evaluative scheme utilizes • Molecular systems, to detect changes in the hereditary material • Single-cell organisms, to detect changes in structure and func- tion that are transmissible to the next generation • Multi-cellular systems, including plants and animals, to detect changes in reproductive potential • Infra-human primates, to detect changes in reproductive potential The evaluation scheme starts with simple, well-defined genetic systems that enable rapid analysis and identification of suspected agents meriting further evaluation. The more complex biological systems used in further testing require substantially larger investments of time and resources. As discussed below, the experiments reviewed in this section have demon- strated that (1) RF irradiation of low-to-moderate intensity does not cause mutations in biological systems, and (2) the only exposures that are potentially mutagenic are those at high power densities of CW or PW radiation—exposures that result in substantial thermal loading or in extremely high electric-field forces within cells. 5.8.2 Effects on Genetic Material of Cellular and Subcellular Systems 5.8.2.1 Physical and Chemical DNA and Chromosome Studies -- Mechanisms for the absorption of RF radiation by DNA molecules are treated elsewhere in this document (see § 3.2, Mechanisms of RF-Field Interactions with Biological Systems; and § 5.1, Cellular and Subcellular Effects). This 5-243 ------- section will address the experimental studies that have examined changes in genetic material induced by RF radiation. Purified DNA or DNA extracted from the testes of exposed mice has been examined to determine if the physical properties of the molecule can be altered by the exposure to RF radiation. Hamrick (1973) examined the thermal-denaturation profiles of aqueous solutions of isolated DNA exposed iji vitro at 37 °C to 2450-MHz (CW) radiation (SAR = 67 W/kg). No changes were found. Even elevated temperatures (to 50 °C) produced by microwaves (1 h at SAR < 160 W/kg) did not cause a difference in the thermal denaturation profile. (See § 5.1 for details.) Varma and Traboulay (1976) reported radiation-induced changes in thermal denaturation profiles (i.e., shift to lower temperature in the midpoint of the transition curve [Tm] and reduction in the maximum hyperchromicity) as well as changes in the base composition of testicular DNA extracted from anesthetized mice whose testes were irradiated in the near field. Ten animals were exposed 2 individually either to 1.7-GHz radiation at 50 mW/cm for 30 min (SAR estimated 2 at 2.4 W/kg for testes alone) or at 10 mW/cm for 80 min (SAR estimated at O 0.48 W/kg for testes alone), or to 0.985-GHz radiation at 10 mW/cm for 80 min (SAR estimated at 0.26 W/kg for testes alone). The animals exposed to 1.7-GHz 2 2 fields at 50 mW/cm or to 0.985-GHz fields at 10 mW/cm were given a 1-day recovery period before they were subjected to euthanasia, and the DNA was ex- 2 tracted. The animals exposed to 1.7 GHz-fields at 10 mW/cm were used in another test for 8 weeks before euthanasia, and the DNA was extracted. Identical results p are reported for the 1.7-GHz exposure at 50 mW/cm in another publication (Varma and Traboulay 1977); they are assumed to describe the same experiment rather than a duplicate experiment because it is unlikely that all the experimental measure- s' 244 ------- ments upon repetition would have given identical values. They conclude "that biological damage may be due to the combined effect of thermal and nonionizing radiation." The experimental results reported by Varma and Traboulay must be examined with careful attention given to the control experiments and to the potential size of the thermal insult. Although the percent adenine/thymine is higher in the DNA extracted from exposed animals, which could be responsible for the drop in Tm, it is highly unlikely that identical values would be obtained for each DNA extract from the three groups of control animals. Based on the variability normally inherent in these biochemical measurements, a more probable explanation is that the control values were a pool of all the control groups or represent just one of those groups. Since the normal experimental variability in the measurements is not described—either for repeated measures on one DNA-extract preparation or between extract preparations—it is impossible to conclude that the differences cited between control and exposed samples are significant; nor can it be concluded that the differences result from the radiation exposure directly, rather than from the extraction procedures used. Similarly, the crude SAR values estimated for testicular exposures appear to be quite low to produce the types of damage cited by these authors. A more likely explanation is that the assumptions underlying either the SAR estimates or the exposure conditions are in error, and that elevated temperatures produced by the radiation exposures were the causative agent. (For example, it is possible that the temperature of the absorbing material used to shield the animals was highly elevated, and it may have elevated the testicular temperature.) In an anesthetized animal ex- posed in the near field to 1.7-GHz radiation at 50 mW/cm for 30 min and shielded 5-245 ------- with urethane foam except for the testes (SAR estimated at 2.4 W/kg), a 1 to 2 °C rectal temperature rise was recorded following exposure (Varma and Traboulay 1977). In another report (Varma and Traboulay 1975), animals were exposed under the same conditions, except that the exposure time varied between 30 and 40 min, and the testes were examined histologically. The authors report, "the lumens were empty with complete disintegration of spermatids, Sertoli cells and the delicate connective tissue which surrounds the seminiferous tubules." In the same report, with respect to animals exposed to 1.7-GHz fields at 10 mW/cm , the authors state: "there was little or no damage to testes, except when the time of exposure was increased to 100 min, then severe changes in morphology were observed." There was no explanation why little or no damage suddenly became severe damage at 100 min of exposure. Anesthetized animals have been shown to lack homeothermic capacity (Cairnie et aK 1980). It is probable that 2 even exposures to 10 mW/cm deposited sufficient energy in these anesthetized animals to cause the temperature elevation responsible for many of the changes that are attributed to microwave-specific effects. Thus, an evaluation of the existing evidence from physical studies on DNA indicates that RF irradiation at low-to-moderate intensities not accompanied by temperature rise causes no changes in DNA bases, the fundamental unit of the genetic code. However, if substantial elevations of temperature occur during exposure, they may cause disruptions in the pairing of the two complementary strands as well as other damage. Other researchers have used cytogenetic techniques to examine some physi- cal and chemical properties of chromosomes in intact cells to determine if the 5-246 ------- relationship of various parts of the genetic material is altered by RF-radiation exposure. Huang et ah (1977) reported no RF-induced chromosomal aberrations in white blood cells from Chinese hamsters exposed to 2450-MHz radiation at power densities up to 45 mW/cm (SAR = 20.7 W/kg) for 15 min a day on 5 con- secutive days. McRee et ah (1978) in a preliminary report found no sister chromatid exchanges in bone-marrow cells of mice exposed to 2450-MHz fields at 20 mW/cm^ (SAR = 15.4 W/kg), for 8 h daily, 28 days total. Alam et ah (1978) showed, in great detail, that chromosomal aberrations occur in a Chinese- hamster-cell line (CH0-K1) exposed 30 min to 2450-MHz radiation from a diathermy applicator, but only if the temperature of the culture is allowed to rise to 49 °C during exposure. These authors demonstrated that irradiation of cell lines at high (> 200 mW/cm ) power densities (SAR estimated at 360 W/kg) would cause no detectable cytogenetic effects, provided proper temperature control was maintained. Thus, heating seems to account for the observed cytological changes. Authors of one detailed study used a frequency well below the 1.7- to 2.45-GHz range. McLees et ah (1972) exposed rats treated to undergo liver regeneration either to 13.12-MHz CW (4.45 kVp_p/m) fields or to PW [44.1 kVp_p/m, 200-jjs pulse width, 50-Hz pulse repetition rate (PRR)] radiation for 28 to 44 h. They examined the effects of radiation on liver cell mitotic activity, that is, the percentage of cells in mitosis, and the number of chromosomal aberrations. Rat liver cells may be very sensitive to exogenous stresses during this regenerative process because they are normally non- dividing cells in an intact animal that have been challenged to divide in situ. However, the authors found no RF-induced alterations in chromosomal 5-247 ------- morphology (SAR estimated at 1.2 to 1.3 W/kg). This result indicates that no cytogenetic changes would be expected in this range of frequencies for low- intensity exposures. Thus, although there are reports indicating that exposure to RF radiation can cause cytogenetic changes, these changes probably result from radiation-induced elevations of temperature. 5.8.2.2 Biological Studies of DNA and Chromosomes- Bacteria have been used to study the mutagenic potential of RF radiation, because the single-cell system is simple, easy to culture, quick to test, and relatively sensitive to the action of mutagenic agents. By using the bacterial system and the biological amplification it provides for any DNA change, molecular biologists have been able to decipher the genetic code and to identify a change in as few as one DNA subunit out of 10^ to 10"^ subunits. In contrast, biologists using standard physical techniques such as DNA melting curves usually can detect changes in no better than 0.1 percent of the DNA. Blackman et al_. (1976) exposed growing cultures of the bacterium Escherichia coli either to 1.7- or to 2.45-GHz (CW) radiation for 3 to 4 h. Exposure at 1.7 GHz was in the near field at 88 V /m or ~ 250 V„ /m (SAR = rms p-p 3 W/kg). The 2.45-GHz exposures were in the far field at either 10 or 50 2 mW/cm (SAR = 15 or 70 W/kg, respectively). Although exposure of growing cells provided enhanced sensitivity to mutagenic agents, no mutagenic activity was detected. A positive control, ultraviolet (UV) light, caused mutations and was used to demonstrate the sensitivity of the assay method. Dutta et aH. (1979a) exposed growing cultures of various bacterial strains of Salmonella typhimurium, commonly used in the Ames testing procedures to detect chemical mutagens (Ames et aK 1975), to 2.45-GHz (CW) radiation for 90 min at 5-248 ------- 20 mW/cm^ (SAR = 40 W/kg) and to 8.6-, 8.8-, 9.0-, 9.2-, 9.4-, and 9.6-GHz 2 (PW) radiation (1-ps pulse width, 1-kHz PRR) at 10 and 45 mW/cm average power 2 densities, and 10,000 and 45,000 mW/cm peak power densities. (The SAR at 45 2 mW/cm is estimated at 80 W/kg.) No mutagenic activity was observed under any of these exposure conditions. Another approach with bacterial systems is to test for radiation-induced alterations in genetic processes, including cell death. Since most mutations are detrimental, they might be detected indirectly by this method. Corelli et al. (1977) exposed cultures of E. coli to microwaves at frequencies swept between 2.6 and 4.0 GHz for 8 h (SAR = 19 W/kg). Although at 26 °C these cultures were probably growing slowly, no change was noted in the number of colony-forming units (CFU) in the cultures following irradiation, indicating no detectable lethal events because of the exposure. These workers also examined the infrared (IR) spectrum of these cells when exposed to 3.2-GHz radiation for 11 to 12 h (SAR is either 21 or 16 W/kg). There was no observable effect on the molecular or conformational structure of these cells, in contrast to results obtained using a positive control, ionizing radiation. Two strains of E. coli, one deficient in an enzyme needed to repair damaged DNA, were tested (Dutta et a2- 1979b) for survival following microwave exposure. The exposure conditions were,8.6-GHz (PW) radiation ( 1-jjs pulse width, 1-kHz PRR, SAR = 12 W/kg) for 1, 2, 4, or 7 h. There was no significant change in the relative growth patterns of these strains that could be attributed to microwave-induced DNA damage repairable in one strain, but not in the other. Blackman et ak (1975) exposed a different strain of E. coli in log phase (actively dividing) and in lag phase (undergoing metabolic activities pre- 5-249 ------- paratory to division) at 32 °C for 4 h to 2.45-GHz radiation at 0.005, 0.5, 5.0, or 50 mW/cm^ (at 50 mW/cm^ power density the SAR = 75 W/kg). Additional experiments were conducted at 5 mW/cm and 25 °C to test for the influence of cold stress, and in two-culture media at 30 and 35 °C to compare the relative influence of a rich medium with a minimal medium, the latter requiring greater utilization of the genetic apparatus of the cell for growth to occur. They found no change in the colony-forming ability of the cultures due to the 2 exposure except for enhanced growth at 50 mW/cm , which was attributed to slight temperature rises in the exposed cultures. Few researchers have used single-cell systems more complex than bacteria specifically to look for RF-radiation-induced mutagenesis. Dutta et a_[. (1979a) exposed a diploid strain of the yeast Saccharomyces cerevisiae, a 2 primitive eukaryote, to 2.45-GHz (CW) radiation for 2 h at 20 mW/cm (SAR = 40 W/kg). They found essentially no change in the number of mutations at either of two loci affecting the nutritional requirements for adenine or tryptophan. These investigators conducted additional tests at 8.5-, 8.6-, 8.8-, 9.0-, 9.2-, 9.4-, and 9.6-GHz (PW) radiation (1-ps pulse width, 1-kHz PRR) for 2 h at average power densities of 1, 5, 8.9, 10, 15, 30, 35, 40, or 2 45 mW/cm . Although no measurements of SAR are cited, thereby making com- parisons with CW exposures difficult, the highest power density was reported to raise the temperature of the culture by 12 °C, indicating substantial 2 absorption of the radiation. (At 45 mW/cm power density the estimated SAR = 80 W/kg). In no case did the exposures cause a change in the frequency of genetic events, altering the requirements for either adenine or tryptophan, in the treated population as compared with the control population. Dardalhon et 5-250 ------- al. (1980) exposed a diploid strain of S. cerevisiae at 20 °C in the near field of 9.4-GHz (CW) radiation (SAR estimated at < 2) for 1 to 5 h, or in the near field of 17-GHz (CW) radiation at either of two power densities (SAR's estimated at 28 or < 6 W/kg, respectively) for various times to 24 h. No significant changes attributable to the irradiation were observed in the percent survival, in the induction of cytoplasmic "petite" mutations, in the induction of mitotic recombinations, or in sporulation. Thus, no changes were detected in complex single-cell systems used to examine directly the mutagenic potential of microwaves. Some work has been done with bacteria and yeast cultures that compares the lethal and mutagenic effects of microwaves with those induced by con- ventional heating. Dutta et aK (1980) examined the responses of various strains of the bacteria S. typhimurium and E. coli and those of a diploid strain of the yeast S. cerevisiae, exposed to 8.6-, 8.8-, or 9.0-GHz (PW) radiation (1-kHz PRR, l-\|js pulse width) at average power densities to 45 mW/cm (SAR estimated maximum at 80 W/kg). The bacteria were exposed 90 min at an ambient temperature of 37 °C, whereas the yeast was exposed 2 h at 30 °C. The results of these irradiation treatments were compared with those obtained from treatments at elevated temperatures produced by conven- tional heating. The comparison indicated that conventional heating could produce cellular damage Reading to reduced survival in a manner similar to the changes caused by microwave-induced elevations of temperature (to 10 °C above normal growth temperatures). No mutational events occurred in bacteria to 42 °C, or in the yeast to 40 °C; however, at 45 °C, a slight increase in mutational events occurred in the yeast, and at 47 °C, in S. typhimurium. 5-251 ------- Dardalhon et	------- pectively) and mated each of the surviving males individually with two virgin females during 15 consecutive days. No changes were observed in the genera- tion time or sex-ratio pattern of the offspring. These offspring were then mated and observed for possible sex-linked lethal mutations in their offspring. No such mutations were found at frequencies > 1 percent, a detection limit based on the small number of chromosomes (< 800) actually evaluated. An additional caveat by the authors was that the most sensitive stages to detect recessive sex-linked lethal mutations, late spermatocytes and early meiosis, were not tested in their study, because few sperm were still in meiosis at the time of exposure. Mittler (1976), who exposed adult males from various strains of D. melanogaster to 29-MHz (CW) fields at 600 Vrms/m (SAR roughly estimated at 0.024 W/kg) and to 146-MHz (CW) fields at 62.5 Vrms/m (SAR roughly estimated at 0.015 W/kg) for 12 h, mated them with virgin females for 12 h every 2 days in production of 4 or 5 broods. In these experiments, brood 4 was produced from sperm irradiated "in or about meiosis." There were no mutations induced by these treatments as evidenced by the lack of chromosome loss, nondisjunction, or sex-linked recessive lethals. In addition, Mittler (1977) observed no mutagenic effects (recessive lethals) when exposing adult females to 98.5-MHz frequency-modulated fields (composed of standard commercial broadcast audio frequencies) at 0.3 V /m (SAR ^ rms estimated at 0.0004 W/kg), 134 h per week for 32 weeks. Although the results with D. melanogaster are difficult to extrapolate to the human condition, this test system as a qualitative index was not able to detect any mutagenic alterations by RF radiation over a wide range of frequencies. 5-253 ------- 5.8.3.3 Vertebrates— Experiments with vertebrates address the impact of the genetic effects of RF radiation most directly because of the general biological similarities between man and other vertebrates. However, there are very few such experi- ments directed at this subject. Three studies have been reported that are directly focused on the mutagenic potential of microwaves in mice and rats. Varma et a^. (1976) used the dominant-lethal test to investigate the effects of 2.45-GHz (CW) radiation. The testes of anesthetized mice were treated either by a single 10-min exposure at 100 mW/cm (SAR estimated at 11.4 W/kg), or by three exposures of 10 min each at 2-h intervals on the same day at 2 50 mW/cm (SAR estimated at 5.7 W/kg), or by four exposures of 10 min each at o 3-day intervals over 2 weeks at 50 mW/cm (SAR estimated at 5.7 W/kg). Varma and Traboulay (1976) in a similar study exposed the testes of anesthetized 2 mice to a single dose of 1.7-GHz radiation, at 10 mW/cm for 80 min (SAR estimated at 0.48 W/kg) or at 50 mW/cm^ for 30 min (SAR estimated at 2.4 W/kg). Following a 24-h recovery period, the males were bred with virgin females—one group of females per week for 6 to 8 weeks. In the 2.45-GHz study, the authors used the week-6-group results for comparison; they found a higher mutagenicity 2 index in the results of the 100-mW/cm group bred at week 1, and in the results 2 of the 50 mW/cm group exposed three times in one day and bred at week 4. The authors concluded that a single, intense exposure or multiple exposures during one day induced a number of significant mutations in the mice. The results of the 1.7-GHz studies at 50 mW/cm for 30 min indicated a radiation-induced increase in infertility, in pre-implantation losses, and in the mutagenicity index of the groups bred at 3, 4, 5, and 6 weeks. In addition, for the group 2 exposed to 1.7-GHz fields at 10 mW/cm for 80 min, an increase was seen in the 5-254 ------- mutagenicity index of the groups bred at 1, 2, 3, and 6 weeks. The authors concluded that the "biological damage may be due to the combined effect of thermal and nonionizing radiation." Both of these reports should be evaluated on the basis described in § 5.8.2.1, Physical and Chemical DNA and Chromosome Studies, for the physical and chemical studies reported by these authors. It is not possible to determine the extent of the biological damage caused by elevated temperature in either of these reports, especially because anesthetized mice lack an effective temperature-regulation capability (Cairnie et al^ 1980). However, if the histological damage found in the testes of animals exposed under similar conditions (Varma and Traboulay 1975) is reflected in damage to sperm, it could account for many if not all of the changes in the mutagenic index, fertility, and pre-implantation loss reported by those authors in later studies. The conclusion that radiation in this frequency range could cause mutagenic changes in mice was challenged in a detailed study by Berman et aK (1980), who exposed unanesthetized male rats to 2.45-GHz (CW) radiation under three treatment regimens: 4 h/day from day 6 of gestation to 90 days of age at 2 5 mW/cm (SAR varied from 4.7 W/kg to somewhat less than 0.9 W/kg at day 90, because growth of the animals changed their energy absorption efficiencies), 2 5 h/day for 5 days beginning on day 90 at 10 mW/cm (SAR estimated at 2 W/kg), 2 and 4 h/day, 5 days/week, for 4 weeks beginning on Day 90 at 28 mW/cm (SAR estimated at 5.6 W/kg). During selected weekly periods after treatment, the exposed males were bred to untreated females that were examined in late pregnancy by the dominant-lethal assay. No significant germ-cell mutagenesis was detected under any treatment condition, even though significant increases in rectal and 5-255 ------- 2 testicular temperature were observed during the 28-mW/cm exposure and were associated with a concomitant decrease in incidence of pregnancy during some of the breeding periods, which indicates temporary sterility. In addition, these authors reexamined the data of Varma and concluded they had been interpreted incorrectly because the effects of litter size on fetal mortality had been ignored and the differences between treated and control groups had been overemphasized when the control values were not representative of normal values. Berman et aK (1980) concluded "it still remains to be demonstrated that microwaves, even at near-lethal doses, can cause a dominant lethal mutagenic effect in the mouse or the rat." Thus, well designed experiments with biological systems having complex genomes similar to man's have demonstrated no mutagenic effects from low- to moderate-intensity RF radiation when the temperature of the biological system is maintained in the physiological range. Most of the reported changes are consistent with the effects expected from elevations in temperature, caused by CW exposure to high-intensity RF radiation. The other changes cited in this section may result from incomplete experimental design. 5.8.4 Unresolved Questions \ Heating, which raises the temperature of biological samples above normal physiological ranges and may result in genetic and mutagenic changes, is a well known result of exposure to high-intensity RF radiation. The choice of exposure conditions to study the genetic and mutagenic effects of moderate- to low intensity radiation is arbitrary because no other well defined mechanism of RF interaction has been developed to address this biological problem. The experimenter must select the frequency, intensity, and duration of exposure, as 5-256 ------- well as the type and characteristics of modulation. To compound the problem further, there are an almost limitless number of biological systems that could be selected for study. There are several guides that can be developed from recent experimental findings to help delineate a future experimental plan. Although the conclusion has been stated above that no strong evidence exists in the cited reports to demonstrate mutagenic activity for low intensity RF radiation per se, it should be qualified by the following observations: 1. These experiments are limited to small portions of the RF region of interest, i.e., 0.5 MHz to 100 GHz. This 106-Hz frequency range is so large that if detrimental effects occurred over very narrow frequency ranges, they might have gone undetected. 2. Many of the experiments were conducted with CW radiation. If effects exist, they may occur with amplitude-, frequency-, or PW-modulated radiation at frequencies associated with ongoing biological or biochemical processes. An exposure to a short-duration, high-intensity PW radiation (radar-like) may not produce the same biological response as a CW exposure of the same average power density. Thus, the concept of averaging the intensity of PW radiation over time to determine an effective intensity may not be generally valid. 3. The general approach in most experimental studies has been to assume that the higher the radiation intensity, the more pro- nounced the biological effect. It is possible that several mechanisms are operative during radiation exposure and that mechanisms with thresholds at relatively high intensities (e.g., heating) mask the effects being caused by the more subtle mechanisms. 4. RF radiation, as a co-stressor, may enhance the effects of known mutagenic agents. In environmentally relevant situations, biological systems are exposed simultaneously to many agents, both chemical and physical. 5. Since no mechanism has been developed to describe the interaction of RF radiation with complex biological structures and genetic material, it could be argued that standard procedures that have been optimized to identify ionizing and chemical mutagens are not necessarily applicable to this form of radiation. Thus, in view of this lack of knowledge, no test system should be overlooked 5-257 ------- as potentially sensitive for detection of mutagenic activity. For example, plant systems have been shown to be extremely sensitive indicators of environmental pollution. Recent experimental developments support some of the suggestions made above. Possible frequency-specific responses were reported by Smolyanskaya and Vilenskaya (1973), who described the induction of colicin in E. coli by extremely high-frequency radiation, near 37 GHz, at power densities of 0.001, 2 0.01, 0.1, or 1.0 mW/cm for 0.5, 1, and 2 h. Colicin is a protein usually not made by the cell. Its induction, while not well understood, requires the transcription of a new portion of the DNA molecule. This induction demonstrates that specific microwave frequencies can induce changes in the genetic processes of cells. More recently, Grundler et al_. (1977) described frequency-dependent growth responses in the yeast S. cerevisiae exposed for periods to 11 h to 2 microwave frequencies between 41 and 42 GHz at 1 to 3 mW/cm (average SAR estimated at 4 to 11 W/kg). (See § 5.1, Cellular and Subcellular Effects, for details.) Although this work is controversial, it supports the frequency- specific-response concept. The importance of examining the effects of RF radiation as a co-stressor with other agents is emphasized in the report by Dardalhon et al. (1980) that was discussed earlier with respect to exposure either to heat or to microwave radiation. These authors exposed haploid and diploid strains of S. cerevisiae 2 for 1 h to 17-GHz (CW) radiation, near field, at 50 mW/cm (SAR estimated at < 6 W/kg) and 20 °C, followed by exposure to various doses of UV light (254 nm), which is known to cause damage directly to the DNA. Both haploid and diploid strains were affected. The exposure combination caused a tendency— 5-258 ------- but apparently not a statistically significant change—toward diminished survival, increased mitotic recombination, and increased cytoplasmic "petite" mutations, when compared with the effects caused by exposure to the same doses of UV light alone. This microwave-radiation enhancement of effects due to UV-light exposure is unusual, because enhancement was observed only at high doses of UV light, where, presumably, additional repair systems are operative. Similar trends were obtained by replacing microwave-radiation exposure with temperature elevation to 46 °C before treatment with UV light. No influence from exposure to microwave radiation, followed by UV light, was 7 2 observed for 17-GHz fields at 2.5 mW/cm , or for 9.4-GHz fields at 5 mW/cm . Because the microwave-radiation treatment caused only a 1 °C global temperature rise in the sample that was 10 °C below its normal growth temperature, the authors reasoned that if temperature rise is the basic agent enhancing the UV-1ight-induced response, it must be occurring selectively at specific sites within the biological system other than on the DNA itself, causing change perhaps in metabolic processes or in the structural integrity of cytoplasm, or of membranes. In this case, there may be a differential sensitivity among the various repair systems induced by the UV light. This type of investigation potentially can reveal biological processes, including repair of damage or detoxification of chemicals, that may be particularly sensitive to microwave- radiation exposure. In another area, radiation-induced efflux of calcium ions from jm vitro brain tissues has been shown to depend on the modulation frequency of a carrier wave (Bawin et aK 1975; Blackman et a^. 1979). The effective modulation frequencies were in the same frequency range as the natural biological rhythms 5-259 ------- associated with the EEG in the intact animal, and thus the radiation may have coupled with an existing oscillatory system. Further studies by Blackman et al. (1979; 1980a,b) and Sheppard et aK (1979) indicated that higher and lower intensities of radiation can lead to the disappearance of the effect. This suggests that examining biological processes only at a single power density may provide misleading information. Three additional, incomplete reports have been included in this section. These studies are included either because they present positive findings at frequency ranges or intensities seldom studied or because of a strong claim of the authors. Heller (1970) and Mickey and Koerting (1970) reported chromosomal aberrations in cultured Chinese-hamster-1ung cells exposed for 30 min to 19- or 21-MHz fields, although no changes were observed after exposure to 15- or 25-MHz (PW) fields (100-Hz PRR, 50-ps pulse width) at field intensities to 300 kVp_p/m. It is very likely either that the extremely high peak voltages (up to 300 kVp_p/m) were producing intense, rapid heating within the cells, or that the field per se was causing major stresses within the system. These reports provide insufficient descriptions to allow an estimate of the SAR values, and, thus, these results cannot be compared with those found by others. It is possible that fields at extremely high intensities can account for some cytogenetic changes. Manikowska et a^. (1979) reported a non-dose-dependent increase in chromosomal translocations and in chromosomal pairs remaining as univalents at Metaphase I in the sperm cells of mice exposed 1 h/day, 5 days/ week, for 2 weeks to 9.4-GHz (PW) radiation at 200-, 1000-, 2000-, or 20,000 mW/cm^ peak power densities (0.5-ps pulse width; 1-kHz PRR; 0.1-, 0.5-, 1-, and 10-mW/cm average power densities, respectively). (Average SAR is 5-260 ------- roughly estimated at 5 W/kg, and peak SAR roughly at 9000 W/kg for the highest power density used.) The authors do not describe the relation of the animals' testes to the incident field, thereby raising the question of the actual quantities of energy coupled to the target cells at 9.4 GHz, where tissues exhibit large attenuation coefficients. A description of the environmental conditions during exposure is also absent. These omissions essentially prevent any critical assessment of the results. Furthermore, because of the small number of animals used in the study, the authors state that "the findings ... obviously need confirmation on larger numbers of animals." In another study, based on bacteria exposed to extremely high temperatures, an incomplete analysis of the distribution of temperatures within the samples led to an unjustified conclusion. Blevins et al^. (1980), who exposed several strains of S. typhimurium commonly used in the Ames testing procedures (Ames et al. 1975) to 2.45-GHz radiation in a microwave oven at a calculated power density of 5100 mW/cm for periods of 2 to 23 s, examined the cultures for lethality and for mutation induction. These exposures caused extremely large elevations of temperature. Corollary heating experiments were performed in high-temperature water baths to determine the extent to which temperature change alone contributed to lethality and mutations. Their conclusion that microwave radiation is a potent mutagen was not supported by their data, because the authors failed to demonstrate that the uniformity of microwave heating was duplicated by the water bath experiments. The authors acknowledge that "differences in the kinetics of water bath and microwave heating are possible" but fail to evaluate this possibility further. Because of the large temperature increases in their culture systems—apparently as large as 5-261 ------- 46 °C in 14 s, more definitive temperature distribution work must be done to establish accurately the contribution that temperature change makes to induction of mutations before additional mutagenic properties are assigned to microwaves. Blevins et aK (1980) reported using a power density 500 times greater than the current U.S. occupational guidelines, and, because the apparent mechanism is based on such high, nonphysiologic temperatures, the authors' conclusions are not applicable to the other cited studies, which have used 2.45-GHz radiation. However, the study may support a reevaluation of the con- cept that brief, high-intensity exposures can be averaged over a longer time period to define the average intensity. For example, an exposure to 5100 2 2 mW/cm for 2 s is 28 mW/cm if averaged over 6 min, and an exposure for 14 s 2 becomes 200 mW/cm if averaged over 6 min. The Blevins et cfL study would also support the concept that high-intensity pulses of microwaves may affect biological systems differently from convection or conduction heating. Although no solid evidence exists (i.e., there are no independently verified reports) to indicate that low-intensity RF radiation is mutagenic, sophisticated concepts and designs are just beginning to appear in experiments concerning the biological effects of RF radiation. Thus, although it is pre- mature to proclaim unequivocally that RF radiation is not mutagenic, the vast majority of evidence presently indicates that in the absence of temperature elevation this statement is true. The above studies of the biological effects of RF radiation that relate to genetics and mutagenesis are summarized in Table 5-21. 5-262 ------- TABLE 5-21. SUMMARY OF STUDIES CONCERNING GENETIC AND MUTAGENIC EFFECTS OF RF-RADIATION EXPOSURE Exposure Conditions Effects Species F requency (MHz) Intensity Duration (mW/cm2) (days x rain) SAR (W/kg) Reference tn i NJ CTi U> No change in thermal denaturation profile, except at elevated temperature Change in thermal denaturation profile and hyperchromicity of ONA extracted from testes following exposure No chromosome aberrations in white blood eel Is No sister chromatid exchange in bone marrow cells No chromosome aberrations in CH0-K1 cell line if temperature maintained No chromosome aberrations or change in mitotic activity in regenerating liver cells in rat No mutation induction No mutation induction observed in Ames tester strains DNA House 2450 (CW) 1700 (CW) 98b (CW) Chinese hamster 2450 (CW) House 2450 (CW) Chinese hamster 2450 (CW) Rat 13 12 (CW) 13 12 (PW) £ coll 2450 (CW) i coTT 1700 (CW) S t.yphimurmm 2450 (CW) 8600-9600 (PW) 134 < 50 10 5-45 20 < 200 1 x 960 1 x 80 5 x 15 28 x 480 1 x 30 4 45 kV /mix 1680-2640 44 1 kVP p/ra p-p 10 or 50 1 x 180-240 250 V /m 1 x 210 p-p 20 1 x 90 10, 45 1 x 90 67 (est) Hamrick (1973) <24 (est) Varma and Traboulay < 0 26 (1976, 1977) 21 Huanq et (1977) 15 McRee et a^ (1978) < 360 (est) Alara et a^ (1978) 1 3 (est) McLees et a] (1972) 15 or 70 3 40 18, 80 (est) 81ackman et a^ (1976) Dutta et al (1979a) (continued) ------- TABLE 5-21. (continued) Effects Species Frequency (MHz) Exposure Conditions Intensity (mW/cm2) Duration (days * nin) SAR (W/kg) Reference Reduction in survival concomitant with rise in sample temperature E col i 8600-9000 (PW) typhi murium 8600-9000 (PW) cerevisiae 8600-9000 (PW) 1-20 < 45 < 45 1 x 90 1 x 90 1 x 120 > 50 < 80 < 80 Outta et al (1980) No reduction in survival or mutational events S cereytsiae Ui ro CTi No detectable lethal events due to no change in CFU's No observable change in molecular structures because no change in IR spectrum No repairable DNA damage No change in growth pattern, enhanced colony-forming activity No change in mutation frequencies at either of two loci controlling requirements for adenine and tryptophan No mutagenic effects in exposed embryos No changes in generation time, sex ratio, or sex-linked lethal nutations in offspring No mutations in adult males as evidenced by chromosome loss, nondisjunction, or sex- 1 inked recessive lethals coll coll co) 1 co) i cerevisiae 9 4 (CW) 17 (CW) 2600-4000 (CW) 3200 (CW) 8600 (PW) 2450 (CW) 2450 (CW) 8500-9600 (PW) 0 melanoqaster 2450 (CW) D melanoqaster 2450 (CW) 0 melanoqaster 29 (CW) 146 (CW) 0 005-50 20 1-45 4600-6500 1 x 300 1 x 1440 1 x 480 1 x 660-720 1 x 60-420 1 x 240 1 x 120 1 x 120 1 x 360 1 x 45 < 28 (est) 19 21 or 16 12 0 008-75 40 < 80 (est) 100 Oardalhon et al (1980) Corel 1 i et aj[ (1977) Corelli et al (1977) Outta et al^ (1979b) Blackman et a^ (1975) Outta et al (1979a) Hamnerius et aj (1979) 600 V /m 1 x 720 62 5 /m 1 x 720 ruts 150-210 (est) Pay et al (1972) 0 024 (est) Mittler (1976) 0 015 (est) (continued) ------- TABLE 5-21. (continued) Exposure Conditions Effects Species Frequency (KHz) Intensity Duration (rtf/cra2) (days x mm) SAR (W/kg) Reference cn i rvj CP No mutagenic changes (recessive lethals) in adult females No significant germ-cell mutagenesis in weekly breedings No significant germ-cell mutagenesis in weekly breedings Same, except decrease in pregnancies, indicating temporary sterility caused by elevated testicular temperatures Induction of a repressed protein, colicin, indicating a change in the genetic processes Change in growth rate that was very fre- quency specific, indicating an alteration in the processes of the cell Chromosome aberrations in lung cells hi vitro at two frequencies but not at two closely related frequencies, 15 or 25 MHz Increase in chromosome translocations in sperm eel Is Increased mutations and lethality 0 aelanoqaster 98 5 (CW) Rat 2,450 (CW) Rat 2,450 (CW) Rat 2,450 (CW) E col i Chinese hamster 37,000 (CW) S cerevisiae 41,000-42,000 (CW) 19 (PW) 21 (PW) Mouse 9,400 (PW) S typhimurium 2,450 (CW) 0 3V /m, 224 x 1140 (FM) £?$audio 5 106 x 240 10 28 0 001-1 1-3 up to 300 kV /m p-p 0 1-10 5100 5 x 300 20 x 240 1 x 30-120 1 x 660 1 x 30 10 x 60 1 x 0 03-0 48 0 0004 (est) Mittler (1977) 4 7-0 9 Herman et a]^ (1980) 2 Berman et a\| (1980) 5 6 Berman et al (1980) Smolyanskaya & Vilenskaya (1973) 4-11 Grundler et a^ (est, av) (1977) Heller (1970) Mickey and Koerting (1970) 0 05-5 (est) Manikowska et aj (1979) Blevins et aj (1980) est = estimated, av = average ------- 5.9 LIFE SPAN AND CARCINOGENESIS William P. Kirk 5.9.1 Life Span There are few data (Table 5-22) and no definitive studies on which to judge the effects of RF-radiation exposure on human longevity. Pazderova-Vejlupkova and Josifko (1979) mention some previous studies in which various clinical para- meters were examined in 82 employees of radio transmitting stations (0.3 to 30 MHz; 14.4-year mean exposure; mean intensity = 80 V/m) and 58 employees of TV transmitting stations (30 to 300 MHz; 7.2-year mean exposure; mean intensity = 2.9 V/m). No signs of damage were found in these personnel; however, the study did not directly address mortality. More recently, information has been developed on the mortality experience of U.S. Government employees assigned to the Moscow Embassy during the period 1953 to 1976, when Soviet microwave irradiation of the U.S. Embassy was taking place. Comparisons were made with a group of employees who had been stationed at other U.S. Embassies in Soviet Bloc cities (Budapest, Leningrad, Prague, Warsaw, Belgrade, Sophia, and Zagreb). The comparison group was chosen to be as similar as possible to the 1800 employees in the Moscow group for selection (i.e., posting) criteria and environmental influences, except that the posts were not subject to microwave exposure (Lilienfeld et al^ 1978). Exposures to 2.56- to 4.1-GHz radiation at 2 a maximum power density of 5 pW/cm were documented from August 1963 to May 1975; the frequency range was expanded to 0.6 to 9.5 GHz in May 1975, and the 2 maximum power density was increased to 18 pW/cm from August 1975 to February 1976. The duration of exposure ranged from 0.5 h/day to 20 h/day (United 5-267 ------- TABLE 5-22. SUMMARY OF STUDIES CONCERNING RF-RADIATION EXPOSURE EFFECTS ON LIFE SPAN/CARCINOGENESIS Exposure Conditions Effects Species Frequency (MHz) - Intensity (mW/cm2) Duration (days x nin) SAR (W/kg) References No effect on life span or cause of death Human adult male & female 2560-4100 600-9500 0 005 (max) 0 01B (max) 8030 x 600 180 X 1200 -4 2 x 10 . (max) 7 x 10 (max) Lilienfeld el al^ (1978) No effect on mortality in a military population followed for 20 years Human adult male 200-5000 (est) (PW) ~ 1 (routine) 100 (occasional) 730 x 480 (est) <0 05 (est) <5 (est) Robinette et (1980) Slight increase in mean life span Adult mouse 800 43 175 x 120 12 9 (est) Spalding et al^ (1971) Increased mean and maximum life span for "irradiated mice with tumors " increased mean life span but no change in maximum life span of non-tumor-bearing mice Delay in development of tumors in irradiated mice but no change in ultimate number of tumors Infant mouse 2450 4 x 20 (In utero) 35 Preskorn et aj[ (1978) Increased mean life span in irradiated mine (concurrent infection-pneumonia) Adult mouse 9270 (PW) 100 4 4 mm/day 5 days/week 59 weeks 40 (est) Prausnitz and Susskind (1962) The duration of 8030 days equals the number of years (22) of irradiation of the embassy and length of the study period, but the average exposure of individuals is estimated to be 2 to 4 years ------- States Senate Committee on Commerce, Science, and Transportation 1979). No evidence was found that the Moscow group had experienced any higher mortality or any differences in specific causes of death up to the time of the report. However, investigators noted that, because the population was relatively young, it is too early to detect long-term mortality effects except for those serving in the earliest period of the study. This important study is covered in more detail in § 5.10 Human Studies. Robinette et al_. (1980) examined the records of a group of 40,000 U.S. Navy personnel who enlisted during the period 1950 to 1954. Approximately 20,000 had job classifications with maximum potential for exposure to radar (i.e., electronics technicians, fire control technicians, and aviation elec- tronics technicians). A similar number of radio and radar equipment operators believed to have a minimal potential for exposure were used as a comparison group. No specific data on exposure or frequencies were provided; however, the authors state that the low exposure group of radio and radar operators 2 were exposed to levels well below 1 mW/cm , whereas the high exposure occu- 2 pations involved average exposure levels below 1 mW/cm during duty hours combined with infrequent higher exposures, which on occasion may have exceeded 2 100 mW/cm . The authors found no apparent difference in the long-term mortality patterns between the two exposure groups more than 20 years post exposure. This study is also covered in more detail in § 5.10. Although there have been many experiments to determine the lethal effects in animals exposed to high levels of RF radiation, data on life span effects of experimental animals exposed to low power levels are scarce. To our knowledge 5-269 ------- there is only one report of a study in which animals were exposed to microwave radiation as the sole stressor and observed throughout their life span. Spalding et al_. (1971) exposed 24 adult female mice to 800-MHz fields at a 2 power density of 43 mW/cm , for 2 h/day, 5 days/week for 35 weeks. A waveguide apparatus was used to irradiate the animals, which were cooled by forced air. No temperature or relative humidity data are reported. The whole-body-averaged SAR is estimated to be 12.9 W/kg. Four exposed mice died of "thermal effects" during the experiment, and a fifth is reported to have died during exposure because it became too obese for the exposure chamber. The mean life span of the remaining 19 exposed mice was 664 ± 32.2 (SEM) days, while that of the 24 sham-irradiated mice was 645 ± 32.2 (SEM) days. Prausnitz and SUsskind (1962) studied pathological and longevity effects on male Swiss mice exposed to 9270-MHz (PW) radiation (duty cycle = 0.001) at an average power density of 100 mW/cm for 4.5 min daily, 5 days/week for 59 weeks. The daily exposure produced an average body temperature rise of 3.3 °C. This daily dose is stated to be half the LD^q for the mice. Originally there were 200 irradiated mice and 100 control mice. Five percent of each group was killed for pathological and hematological examination 7 months after the beginning of irradiation, and an additional 10 percent of each group was killed at 16 months. The latter series was done within a month of the final irradiation. Because of a partial contradiction in pathological findings in the animals killed at 7 and 16 months, all surviving mice (19 controls and 67 irradiates) were killed 19 months after the beginning of irradiation (4 months after final irradiation). The data from this experiment are summarized in Table 5-23. The authors state that the longevity of the mice did not appear to 5-270 ------- TABLE 5-23. SUMMARY OF PRAUSNITZ AND SUSSKIND DATA (1972) Controls (100 total) Irradiates (200 total) When Sacrificed No X of No with % This Series No % of No with % This Series Series (months) Animals Total "Leucosis" with "Leucosis" Animals Total "Leucosis" with "Leucosis" Comment tn i (V) --J Sacrifice Series 01 7 Sacrifice Series #2 16 10 (1 month post- irradiation) Sacrifice Series #3 (all animals surviving at 19 months) Longevity Series 19 19 (4 months post- lrradiation) As 40 occurred Spontaneous Deaths No data given 26 10 19 40 26 None apparently given 10 21 10 10 20 10 None apparently given 30 67 33 5 12 60 30 21 43 22 18 35 Reported as negative for "leucosis" Reported as positive for "leucosis" Not statistically significant X2 (1 df) = 1 49 Reported as negative for "leucosis" See text Autolysis too advanced for diag- nosis of cause of death ------- be affected under the prevailing conditions. The data appear to support a stronger statement, i.e., significantly more irradiated mice than controls survived until the termination of the experiment 19 months after the beginning of irradiation. The difference in survival is not statistically significant at 14 months 2 into the experiment (x ~ 2.2, using data from Fig. 3, Prausmtz and Susskind 1962) but is significant at 19 months (x^ =4.9, 0.05 > P > 0.025). The data pro- vided are not sufficiently detailed to apply more refined methods of analyzing survival (Peto et al_. 1976, 1977). The authors suggested that, in view of problems with pneumonia experienced in both control and irradiated mice in the last few months of irradiation, "conceivably microwave irradiation in this modality, with periodically induced slight artifical fevers, is of some benefit to the animal in combatting disease." There have been some reports of altered and possibly enhanced immunological competence after microwave exposure. These are discussed in detail in § 5.2, Hematologic and Immunologic Effects. In a relevant study, Preskorn et aK (1978) compared tumor development and longevity in female mice prenatally exposed to 2450-MHz microwaves with sham-irradiated controls. The irradiated mice were offspring of dams exposed for 20 min/day on days 11, 12, 13, and 14 of gestation in a multimodal cavity (SAR = 35 W/kg). Both groups were injected with sarcoma cells at age 16 days. Both the mean and maximum survival time of "irradiated mice with tumors" exceeded those of "non-irradiated mice with tumors" (P < 0.05). The mean survival time of "irradiates without tumors" was also greater than "non-irradiated mice without tumors." When 50 percent of 5-272 ------- the controls had died, 67 percent of the irradiates were alive. The difference was not significant (0.05 < P > 0.1), however, and maximum life span did not di ffer. Rotkovska and Vacek (1972, 1977) reported increased survival and increased LD^o_3q in m1ce for sublethal (600 to 750 R) acute exposures to 200 kV x-rays if the x-ray exposures were preceded (1, 3, 14 days) or followed (30 min) by a 2 5-min exposure to 2450-MHz microwaves at 100 mW/cm . In their initial experi- ment, the 30-day survival after a 600-R x-ray exposure was increased from 14 percent (controls) to 53, 88, or 90 percent when such exposure followed the microwave irradiation by 1, 3, or 14 days, respectively. In their second experiment, the LD^q_^q was increased by 100 R with all microwave-irradiated groups having significantly higher survival at 30 days than their control counterparts. These data illustrate the plethora of potentially confounding variables that may have to be considered in determining RF-radiation exposure effects. 5.9.2 Carcinogenesis Carcinogenesis is the process of inducing cancer or malignant neoplasia. Neoplasia is uncontrolled growth or cell division in a tissue and is considered malignant if cells from the affected tissue (tumor) enter the blood and travel to other parts of the body to colonize and form new tumors (metastases). A tumor that does not metastasize is considered benign, although it may well grow to sufficient size to be 1ife-threatening. Consideration of the various theories of carcinogenesis is beyond the scope of this discussion, except to 5-273 ------- note that most authorities believe that the initiating event on the cellular level involves physical or chemical alterations of a cell such that its descendent cells are abnormal and may override the body's cellular proliferation control processes. The role that the immune system plays in primary carcinogenesis has not been well documented and is somewhat controversial. However, it is thought that neoplasia may arise through an epigenetic mechanism in which immunosuppression allows the development of tumors initiated by a genetic event. The specter of potential carcinogenicity has been periodically raised in connection with RF-radiation exposure since 1953, when J. R. McLaughlin, a medical consultant to the Hughes Aircraft Corporation, submitted a report to the military, listing leukemia as one of the possible effects of radar exposure (McLaughlin 1953). The relevant literature, which is sparse, has been re- viewed by Baranski and Czerski (1976), Justesen et cH. (1978), and Dwyer and Leeper (1978) with little supportable evidence that RF exposures are likely to be carcinogenic. However, the subject remains controversial for many reasons, including (1) Widespread fear of cancer in large segments of the population (2) The proven carcinogenicity of ionizing radiation (3) The failure of the public and the media to distinguish between ionizing and nonionizing (i.e., RF) radiation (4) The existence of several anecdotal reports of association of cancer with RF-radiation exposure in humans (5) The publication of an early animal study (Susskind 1962, Prausnitz and Susskind 1962) reported by the authors to show leukemogenesis in mice chronically irradiated with microwaves (6) The unsupported reports by Zaret (1976) of an increased incidence of cancer in North Karelia followed by the misinterpretation of 5-274 ------- the Zaret article by Dwyer and Leeper (1978), coupled with their editorial transformation of the North Karelia Project from a study of cardiovascular disease to a study of cancer (7) The recent reports by Szmigielski et a^. (1980) that exposure to 2450-MHz microwaves accelerated the development in mice of spontaneous breast cancer and 3,4-benzopyrene-induced skin cancer (8) The postulation that, since heat is a known mutagen and there is a strong correlation between mutagenesis and carcinogenesis, under certain conditions of relatively high-intensity exposure, RF radiation that causes significant tissue heating could theoreti- cally induce cancer (9) The lack of well designed human or animal studies that have sufficient exposure data and are statistically adequate to provide credible data that would support a reliable conclusion. Reviewed in depth herein are four published reports that have received extensive publicity, including (1) the Prausnitz and Siisskind (1962) report of leukemogenesis in mice chronically irradiated with microwaves, (2) the "North Karelia Connection" suggested by Zaret (1976), (3) the irradiation of U.S. Government employees stationed at the U.S. Embassy in Moscow (Lilienfeld et aK 1978), and (4) the health status of U.S. Navy personnel exposed to radar during the Korean War (Robinette et aJL 1980). Brodeur (1977) gives anecdotal accounts of increased incidence of cancer in several groups of defense con- tract personnel working in RF-radiation research and development, but no reliable reports on these incidents have appeared in the scientific or medical literature. Since the cutoff date for general literature review, two letters to the editor have appeared in the New England Journal of Medicine suggesting an association of polycythemia vera with occupational microwave exposure (Friedman 1981) and leukemia with occupational exposure to a variety of elec- tric and magnetic fields (Milham 1982). Neither of these letters would meet 5-275 ------- the criteria established for this review but should lead to more elaborate studies to resolve the questions raised. 5.9.3 Prausnitz and Susskind Study In 1962, Prausnitz and Susskind published reports on an experiment con- ducted to determine the pathological and life-shortening effects of chronic microwave exposure of mice. In 200 male Swiss mice irradiated with 2 9270 MHz (PW) at an average power density of 100 mW/cm for 4.5 min daily, an average body temperature increase of 3.3 °C was produced. Exposures were conducted 5 days/week for 59 weeks, and survival, body mass, blood parameters, and certain postmortem pathological findings were compared with those of a con- currently sham-irradiated group of 100 animals. Animals were sacrificed at 7, 17, and 19 months after the beginning of irradiation. In addition, 23 percent of the animals were lost to the study by spontaneous death and autolysis prior to necropsy. The remaining 100 animals (40 controls, 60 irradiates) formed a "longevity" study group. One of the findings reported in this group was "cancer of the white cells," defined as monocytic or lymphocytic leucosis, or lymphatic or myeloid leukemia. Leucosis was defined as a noncirculating neoplasm of the white cells, whereas leukemia was defined as a circulating leucosis. Data were grouped and reported as "leucosis." Although the liver and spleen were removed for examination at necropsy, they were not considered in determining lymphoid infiltration, even though these organs are usually involved in this kind of reaction. Data describing the differential composition of the blood cell types are not given. A summary of the fate of the 200 irradiates and 100 controls is shown in Table 5-23. 5-276 ------- The results as reported are confusing. The authors performed a third sacrifice series at 19 months to resolve a perceived contradiction in the "leucosis" findings from the first and second sacrifice series. Yet, had they used even a crude 2x2 contingency table analysis to test the prevalence of leucosis in the control and irradiated animals sacrificed in the second series, 2 they would have arrived at a x of only 1.49 (uncorrected for continuity) with one degree of freedom (P > 0.20). Thus, the prevalence of leucosis in irradiated animals was not significantly different from that in the control animals for any of the "sacrifice" groups. While leucosis in the "longevity" series is highly significant (P < 2 0.005, x >8.0), these data are severely compromised by (1) the marked loss of animals that would have fallen in this group due to autolysis, (2) the differential loss of control animals due to infection, and (3) the premature sacrifice of all remaining animals to resolve a nonexistent paradox. Com- paring data from two groups of animals that selected themselves (by dying prior to 19 months and being found in time for necropsy) is not an acceptable statistical analysis of experimental data. An acceptable method would be to compare the prevalence rates of leucosis in the controls and irradiates in combined groups that include all the animals at risk in the original longevity group (longevity and third sacrifice series). If this is done, the following results are obtained: Group With Leucosis Without Leucosis Total Control 8 (a) 51 (b) 59 (a+b) Irradiated 33 (c) 94 (d) 127 (c+d) Total 41 (a+c) 145 (b+d) 186 (n) 5-277 ------- To test whether the leucosis rates are different in controls and irradiates, the following statistic may be used: S = n(ad - bc)^/[(a+b) (c+d) (a+c) (b+d)] If the animals comprising the table represent a random sample from a larger 2 population, S approximately follows a x distribution with one degree of 2 freedom. Upon substitution, the value of 3.72 is obtained. Since x >3.62 (P - 0.06), the rates would not be declared different at the 5 percent significance level. Furthermore, if a continuity correction were used in computating this 2 statistic, the resultant x = 2.93, which is significant only at the 9 percent level. There are now better statistical methods available to analyze survival data (Peto et al^. 1976, 1977), but they require detailed documentation of the fate of each animal during the course of the experiment. Given (a) the loss of histopathological data on animals because of autolysis and infection, (b) the further complication that different pathologists were used for different sacrifice series, and (c) the absence of historical data on the incidence of leucosis in the mouse strain used, any attempt to apply these methods retro- actively would not be beneficial. In summary, because of the previously described problems with the bio- logical protocol, the lack of sound statistical methodology in experimental design and data analyses, and the questionable significance of what was reported, 5-278 ------- the study by Prausnitz and Susskind is of limited value in defining the pathological effects of chronic whole-body microwave radiation. As noted previously, the data on survival times are of considerable interest due to similar findings by Spalding et a^. (1971) and Preskorn et al^. (1978). 5.9.4 North Karelia Connection In 1971, the Finnish government, in conjunction with the World Health Organization, began a program known as the "North Karelia Project" (Puska et aK 1978). Finland has one of the highest rates of mortality from cardio- vascular disease (CVD) in the world. Within the country, the eastern provinces had higher rates of CVD than those in the west. North Karelia, an eastern province that borders the Soviet Union, was identified in the 1950's as having a particularly high incidence of CVD. The objectives of the project were to decrease CVD morbidity and mortality by identifying the causative factors, devising means for primary prevention, and strengthening treatment and secondary prevention. Cigarette smoking, elevated blood pressure, and serum cholesterol levels were postulated as three major risk factors. The proximity of North Karelia to the Soviet border prompted an American physician, Dr. Milton Zaret, to speculate that RF radiation from Soviet communications or radar might be contributing to the incidence of CVD. In a presentation at the symposium Biologic Effects and Health Hazards of Microwave Radiation (Warsaw, October 15 to 18, 1973), he suggested that microwave radiation from the Soviet Union might be a factor. His remarks were later withdrawn and do not appear in the record of the Symposium. However, he restated this hypo- thesis in a published article (Zaret 1976), in which he additionally stated, 5-279 ------- without data or reference, that "a new finding, an increased incidence of cancer, also appears to be emerging in North Karelia." The major thrust of the article was directed to the possibility of exploiting the North Karelia Project to investigate the role of RF radiation in the evolution of CVD. He also noted that the potential role of RF radiation in carcinogenesis should be studied. In their literature review on carcinogenic properties of microwave and RF radiation, Dwyer and Leeper (1978) misinterpreted the 1976 Zaret article. They editorially transformed the North Karelia Project into an epidemiological project designed to test the hypothesis that RF radiation caused or contributed to heart attacks or cancer. The Finnish Project (1972 to 1977) caused significant changes in diet and lifestyle, which correlated with a significant reduction in CVD mortality. Further, information provided by the Finnish Institute of Radiation Protection (K. Jokela, personal communication to M. Hattunen, Scientific Counselor of Embassy of Finland, 2133 Wisconsin Avenue, N.W., Washington, D.C. 20007, November 22, 1978) specifically denies the existence of an abnormal cancer incidence in eastern Finland and knowledge of any possible linkage to microwave fields. 5.9.5 Moscow Embassy Study The long-term microwave irradiation of the American Embassy buildings in Moscow was highly publicized in 1976. This led to a comprehensive study of 5-280 ------- the morbidity and mortality of the Moscow Embassy employees and dependents as compared with employees and dependents of U.S embassies in other Eastern European locations (Lilienfeld et a^. 1978). The Moscow group experienced less overall mortality for all causes of death than did the comparison group. However, the death rate in females from malignant neoplasms was slightly, but not significantly, higher than expected in the Moscow group; and the incidence of malignant neoplasms, other than of the skin, was significantly higher in the Moscow female group. In both cases, the numbers were very small. In the former case, there were 7 different cancer sites involved in 8 cases, which, according to the authors, virtually eliminates a single causal factor; in the latter case, there were 10 cases involving 7 different sites, and, when exposure to microwaves was considered, it was found that the death rate was highest for those who had the least exposure to micro- waves. 5.9.6 U.S. Navy Study A study of the effects of occupational exposure to radar during the Korean War on U.S. Navy enlisted personnel was recently published (Robinette et al. 1980). Approximately 20,000 men whose occupations possibly involved relatively more frequent exposures to higher intensity fields were compared with 20,000 men in other occupational classifications. One of the end points evaluated was the effect of exposure from 1950 to 1954 on the relative inci- dence of cancer during 1950 to 1976. No statistically significant differences 5-281 ------- between the two groups were evident for malignant neoplasms as a cause of death from 1950 to 1974, or as a cause of hospitalization in Navy or VA hospitals. Within the high exposure group, subgroups were created, depending on Hazard Number. The Hazard Numbers were measures of potential individual exposures, dependent upon the type and power rating of the radar and upon the ship and the time spent in service on the ship. Hazard Number categories were 0, 1 to 5000, and > 5001. In considering the causes of death within the highest exposure group, there are significantly more malignant neoplasms of the respiratory tract in the subgroup with Hazard Number > 5001 than in the subgroup with the Hazard Number < 5000. A significant trend for increased death from all diseases in the > 5001 subgroup was also noted. However, informa- tion relevant to respiratory neoplasms such as smoking histories was not available because the study was an analysis of cause of death from death certificate data. Also, in a study where so many statistical comparisons are done using a P = 0.05 criterion, one or two positive findings would be expected by chance. 5.9.7 Unresolved Questions Because few RF-radiation studies in man or animals have employed life span or cancer as end points, and none has had sufficient statistical power and adequate quality control to place an upper limit of risk at less than two times control incidences, the questions of microwave carcinogenesis or life shortening are still open. Conversely, none of the complete reports in the literature presents a convincing case for the existence of a significantly increased risk of cancer or life shortening in exposed populations. A critical issue is the difficulty of developing exposure data or information in human population studies. 5-282 ------- The interaction of RF radiation with other moieties (such as ionizing radiation) in the induction, prevention, or treatment of cancer has not been fully characterized. Microwave-induced hyperthermia has been demonstrated to increase the effectiveness of x-ray treatment of various cancers. In nonthera- peutic situations, however, both inhibition and enhancement of tumor development in microwave-irradiated mice have been reported. In their paper, discussed in the section on longevity (§ 5.9.1, Life Span), Preskorn et a_h (1978) found that the development of tumors following injection of sarcoma cells into 16-day-old CFW mice was significantly delayed if the mice had been exposed to 2450-MHz microwaves jn utero on day 11, 12, 13, and 14 of gestation (20 min/day, SAR = 35 W/kg). There was no difference, however, in the ultimate number of tumors. Conversely, Szmigielski et al^. (1980) found that repeated exposure to 2450-MHz microwaves 5 or 15 mW/cm^ (SAR 2 to 3 or 6 to 8 W/kg), 2 h/day, 6 days/week for up to 10 months significantly accelerated the appearance of spontaneous breast tumors in C^H/HeA mice, and the appearance of skin cancer in BALB/c mice treated with 3,4-benzopyrene during or after the microwave exposure. It is noted that chronic stress due to deliberate overcrowding 2 produced essentially the same effect as the 5 mW/cm exposure. In both ? experiments, 15 mW/cm produced greater acceleration in tumor production than 2 did 5 mW/cm or overcrowding. Workers in this field have only begun to discover the effects of RF-radiation on animals under different immunologic and hormonal equilibrium levels, e.g. animals subjected to prolonged stress. 5-283 ------- 5.10 HUMAN STUDIES Doreen Hill The general approach used in this document for the evaluation of the health effects literature is stated in the Introduction. Rigorous criteria were established and applied to the reports of experimental results in order to establish what is believed to be a credible data base. It is, however, difficult to apply these selection and review criteria uncompromisingly to the human studies or epidemiological literature, largely because the research method is population-based and observational rather than experimental. The papers included here were selected because they were judged to present rela- tively more information on exposure parameters and/or more rigorous or ana- lytical study designs. The amount of detail in reporting and the degree of specification of study methods and procedures (use of controls, statistics, control of confounding variables, and so forth) were considered important. Case reports are not included. 5.10.1 Occupational Surveys/Clinical Studies The majority of reports in the literature concern people occupationally exposed in military or industrial settings. A wide variety of conditions, symptoms, and clinical measurements are usually evaluated. The health condit- ions investigated usually are pre- or subclinical instead of overt or diag- f nosed disease. Studies of this type that focused on single rather than multi- ple end points are described later. 5-285 ------- It is in occupational surveys and industrial health surveillance programs that the potential for neurological and behavioral changes has been investi- gated. The Soviet and Eastern European literature frequently describes a collection of symptoms reported to occur in personnel industrially exposed to microwaves. These collective symptoms, which have been variously called the "neurasthenic syndrome," the "chronic overexposure syndrome," or "microwave sickness," are based on subjective complaints that include headaches, sleep disturbances, weakness, decrease of sexual activity (lessened libido), impotence, pains in the chest, and general poorly defined feelings of non- well-being (Baranski and Czerski 1976, p. 157). Also described are a set of labile functional cardiovascular changes including bradycardia (or occasional tachycardia), arterial hypertension (or hypotension), and changes in cardiac conduction, and this form of neurocirculatory asthenia is also attributed to nervous system influence (Silverman 1980). Barron et a^. (1955) presented results of a study conducted to evaluate changes in various physical and functional characteristics of radar personnel employed by an airframe manufacturer. A total of 226 exposed workers were initially included in the medical surveillance program. The radar workers were characterized by their duration of exposure as seen in Table 5-24. Controls totaling 88 subjects, stated to have had no industrial radar exposure, were also examined. Methods of selection of cases or controls were not specified. The age distribution of all subjects ranged from 20 to more than 50 years, with the majority under 40 years of age. The controls were, however, substantially older, as can be seen in Table 5-25. 5-286 ------- TABLE 5-24. DISTRIBUTION OF YEARS OF EXPOSURE FOR 226 RADAR WORKERS Years of Exposure No. of Workers Percent of 226 0 to 2 106 47 2 to 5 83 37 5 to 13 37 16 Data from Barron et al. 1955. TABLE 5-25. AGE DISTRIBUTION OF 226 MICROWAVE WORKERS AND 88 CONTROLS Age Radar Controls % of 226 Cumulative % % of 88 Cumulative % 20 to 29 34 34 14 14 30 to 39 49 83 40 54 40 to 49 13 96 27 81 50+ 4 100 19 100 Data from Barron et al. 1955. A decrease in polymorphonuclear cells was reported in 25 percent of the radar workers vs. 12 percent of the controls. An increase in monocytes and eosinophils was also observed for the exposed group, but in a later report (Barron and Baraff 1958) these effects were attributed to a technical error. Platelet counts and urinalyses were similar in the two groups. Ophthalmo- logical examinations revealed ocular anomalies of several diverse types among 5-287 ------- 12 exposed persons vs. 1 case in the control group. The medical surveillance program was extended to permit periodic reexaminations (Barron and Baraff 1958). No significant changes in health status were noted. The radar bands of exposure include S-band (2880 MHz) and X-band (9375 MHz). Exposure times and power densities for individuals could not be developed, but zones at various distances from the antenna were specified and used to estimate three ranges of power densities. The minimal average power 2 2 density in Zone A was 13.1 mW/cm . Zone B ranged from 3.9 to 13.1 mW/cm . 2 Zone C was <3.9 mW/cm . The authors stated that because of the relatively low power densities, personnel working in Zone C were eliminated from the study. The majority of personnel worked with or around APS-45 and AN/APS-20-B and -E radars. By establishing these zones of potential exposure, steps were taken to limit entry of personnel to Zone A. Zone B was judged to be safe for occasional but not constant exposure. As a result, most subjects continuing in or added to the examination program are believed to have had only incidental exposures to power densities less than 13.1 mW/cm . The strengths of this study are the attempts to estimate potential exposures and to periodically reexamine the workers. But, there are also some major problems with study methods and analyses. For example, there are dis- proportionate numbers of exposed vs. control subjects as well as questionable comparability because of age differences (Table 5-25). None of the observed results were tested statistically. 5-288 ------- Czerski, Siekierzynski, and co-workers in Poland evaluated 841 men occupationally exposed to pulse-modulated microwave radiation (Czerski et al_. 1974; Siekierzynski et a^. 1974a, 1974b). The radiation frequencies were not explicitly stated, but one can infer from the references to pulse-modulation and radio-location that the working environment dealt with radar frequencies. The age distribution of the men ranged from 20 to 45 years. The men had worked varied periods of time, with some employed over 10 years. An unexposed control population comparable in general working conditions and socio-economic status could not be established; therefore, the study group was subdivided into two groups on the basis of level of microwave exposure. One group con- 2 sisted of 507 men exposed to mean power densities greater than 0.2 mW/cm , 2 with short-term exposures estimated to reach 6 mW/cm . The other group was 2 334 men exposed to mean power density levels less than 0.2 mW/cm . The health end points evaluated covered three major categories, e.g., neurotic syndrome, digestive tract functional disturbances, and cardio- circulatory disturbances with abnormal electrocardiogram (ECG) findings. According to Polish occupational exposure criteria, these conditions are considered contraindications for work in a microwave environment. The neurotic syndrome was defined by a variety of symptoms such as fatigue, headaches, sleep disturbances, and difficulties in memorizing and con- centrating. Psychologic examinations were given. Comparisons were made between and within exposure groups according to age and duration of occupational exposure. The two groups were found to be similar with respect to the distribution of these symptoms and conditions. There was no dependence on the duration of exposure. 5-289 ------- In 1979, Djordjevic et a^. reported on medical evaluations of radar workers, aged 25 to 40 years, with a work history of 5 to 10 years. Specific frequencies of exposure were not cited but were stated to be within the whole range used in radar operations. Evaluation of the working environment was undertaken, including power density measurements. While the environmental analyses are not given in the paper, it was concluded that the workers were exposed to pulsed microwaves within a wide range of intensities but generally ? at levels less than 5 mW/cm . The lower limit of this exposure may have been 2 1 mW/cm , but it is not clear from the discussion whether this estimate refers to the workers included in this study or to radar station personnel in general. The control group consisted of 220 persons reported to be similar in age, working environment, and socioeconomic status. The controls did not have work experience with microwave sources. Selection criteria or further descriptive information was not given for either the cases or controls. Ten major end points or diagnoses were covered in the clinical eval uation, including ophthalmologic examinations. These were conducted by clinical specialists following the same scheme for classification of abnormalities. The two groups did not differ with respect to the 10 factors. The type of procedure was not reported. The groups also did not differ statistically in terms of electrocardiogram results or on multiple biochemical and hematologic indicators. Radar workers did demonstrate more subjective complaints, including headache, fatigue, irritability, sleep disturbance, inhibition of sexual activity, and impairment of memory. Based on their survey of working conditions, the authors attribute the latter result to specific problems such as lighting or poor ventilation in the environment 5-290 ------- of the radar workers. If true, it appears that either the subject and control groups were not as comparable with respect to working environment as originally stated, or some other factor(s) is (are) operating to produce more subjective complaints. One possibility is a selective or reporting bias on the part of the radar workers, e.g., enhanced awareness of the possible "microwave sickness1' syndrome. Some other environmental factor may have been present that may not be measurable in the type of environmental survey conducted or detectable through these clinical evaluations but microwave exposure cannot be clearly excluded. 5.10.2 Mortality Studies Lilienfeld and assodiates (1978) completed a broad survey of the mortality and morbidity experience of Foreign Service employees and their dependents to assess the potential health consequences of microwave irradiation of the American Embassy in Moscow. Foreign Service employees and those from other agencies who had served in the Moscow Embassy during 1953 to 1976 were compared with employees at eight other embassies or consulates in Eastern Europe over the same time period. The microwave irradiation of the Moscow Embassy was first detected in 1953 and subsequently varied in intensity, direction, and frequency over time. The frequencies ranged from 0.6 to 9.5 GHz (United States Senate, Committee on Commerce, Science, and Transportation 1979; Pollock 1979). The measured average power densities over time are given in Table 5-26. 5-291 ------- TABLE 5-26. MICROWAVE EXPOSURE LEVELS AT THE U.S. EMBASSY IN MOSCOW Time Period Exposed Area of Chancery Power Density and Exposure Duration 1953 to May 1975 West Facade 2 Max of 5 pW/cm 9 h/day June 1975 to Feb. 1976 South and East Facade 18 pW/cm2 18 h/day Since Feb. 7, 1976 South and East Facade 2 Fractions of a pW/cm 18 h/day Data from Lilienfeld et al. 1978. Extensive efforts were launched to identify and trace the populations. Information on illnesses, conditions, or symptoms were sought from two major sources: (1) employment medical records, which were fairly extensive, given examination requirements for foreign duty and (2) a self-administered health history questionnaire. Questionnaire responses were validated for a stratified sample by review of hospital, physician, and clinic records. Death certificates were also sought, although other sources also were used to ascertain mortality status. Standardized mortality ratios for various subgroups were developed for each cause of death, were standardized for age and calendar period, and were specific for sex. Similar procedures were applied to develop summary indices of morbidity. 5-292 ------- A total of 4388 employees and 8283 dependents was studied. More than 1800 with 3000 dependents were employed at the Moscow Embassy and 2500 with more than 5000 dependents worked at the comparison posts. Ninety-five percent of the employees were traced. Receipt of completed questionnaires was less successful, with an overall response rate of 52 percent for State Department personnel. Hundreds of comparisons between the Moscow and control posts were made from information in the medical records. Various health problems were generally similar with two exceptions. Moscow employees had a threefold greater risk of acquiring protozoal infections than comparison-post employees. In general, both sexes in the Moscow group had somewhat higher frequencies of most of the common kinds of health conditions reported. The authors stated, "However, these most common conditions represented a very heterogeneous collection and it is difficult to conclude that they could have been related to exposure to microwave radiation since no consistent pattern of increased frequency in the exposed group could be found." Some excesses were reported by Moscow employees in the health history questionnaire. Both sexes reported more eye problems due to correctable refractive errors. More psoriasis was reported by men and anemia by women. The Moscow employees, especially males, reported more symptoms such as irritability, depression, difficulties in concentration, and loss of memory. It is possible, however, that a bias due to awareness of potential adverse effects is operating, since the strongest differences were present in the subgroup with the least exposure. 5-293 ------- The observed mortality was less in both male and female employees than expected, based on U.S. mortality rates; the male employees had more favorable experience than female employees. In both sexes, cancer was the predominant cause of death. The Moscow and comparison groups did not differ appreciably in overall and specific mortality. The authors noted, however, that the population was relatively young, and it may be too early to detect long-term mortality effects. The authors concluded that no convincing evidence was discovered to implicate microwaves in the development of adverse health effects at the time of the analysis. But they also carefully discussed the limitations inherent in the study. These included uncertainties associated with the reconstruction of the employee populations and dependents, difficulties of obtaining death certificates, the low percentage of responses for the questionnaire, and the statistical power of the study. The limitation most critical for consid- eration in a document such as this relates to ascertainment of exposure. Problems relative to individual mobility within the embassy and variation of field intensities within the building are present in this study as in any other. No records were available on where employees lived or worked, so one had to rely on questionnaire responses to estimate an individual's potential for exposure. The highest exposure levels (18 pW/cm ) were recorded for only 6 months in 1975-76, thus the group exposed to the most intense fields had the shortest cumulative time of exposure and of observation in the study. Robinette and Silverman (1977) and Robinette et al^- (1980) examined mortality and morbidity among U.S. naval personnel occupationally exposed to 5-294 ------- radar. Records of service technical schools were used to select subjects for the study. Exposure categorizations were made on the basis of occupational specialty. The exposure group (probably highly exposed) consisted of technicians involved in repair and maintenance of radar equipment. The controls (probably minimally exposed) were involved in the operation of radar or radio equipment. It was estimated from shipboard monitoring that radiomen and radar operators (in the low-exposure group) were generally exposed at less 2 than 1 mW/cm , while gunfire control and electronics technicians (in the high exposure group) were exposed to higher levels during their duties. Over 40,000 veterans were included in the study, with about equal numbers in these two major exposure classifications. In conjunction with naval personnel, an effort was also undertaken to develop an index of potential exposure, termed Hazard Number, for a limited portion of the population. This number was based on the duty months multiplied by the sum of the power ratings (equipment output power) of gunfire-control radars (ship) or search radars (aircraft) where technicians were assigned. Medical information was obtained through Navy and Veterans Administration records. Records were searched for information on four major end points: (1) mortality; (2) morbidity via in-service hospitalizations; (3) morbidity via VA hospitalizations; and (4) disability compensation. Mortality was ascertained through the VA beneficiary system. Strokes, cancers of the digestive tract and respiratory system, and leukemias were elevated for the high exposure group, but none of the increases was statis- tically significant. The authors note that the differences in mortality from 5-295 ------- malignant neoplasms of the lymphatic and hematopoietic system are not statistically significant. As seen in Table 5-27, comparisons were also made within the high exposure group across Hazard Number categories. In this case, only two comparisons were statistically significant: (1) the difference in TABLE 5-27. NUMBER OF DEATHS FROM DISEASE AND MORTALITY RATIOS3 BY HAZARD NUMBER: U.S. ENLISTED NAVAL PERSONNEL EXPOSED TO MICROWAVE RADIATION DURING THE KOREAN WAR PERIOD Number of Deaths International Classification ^ High Exposure, by Hazard No. Diseases Low Cause of Death (8th Rev.) Exposure Total 0 1-5000 5000+ All diseases 000-796 325 309 63 160 86 (1.04) (0.96) (0.82) (0.91) (1.23) Malignant neoplasms 140-209 87 96 22 45 29 (0.96) (1.04) (0.99) (0.90) (1.44) Digestive organs 150-159 14 20 6 11 3 (0.85) (1.14) (1.49) (1.14) (0.78) Respiratory tract 160-163 16 24 4 10 10 (0.85) (1.14) (0.82) (0.86) (2.20) Lymphatic and 200-209 20 26 6 12 8 hematopoietic system (0.83) (1.18) (1-09) (1.04) (1.64) Other malignant Residue 37 26 6 12 8 neoplasms (1.19) (0.82) (0.78) (0.70) (1.17) Diseases of circulatory 390-458 167 150 36 73 41 system (1.07) (0.93) (0.94) (0.83) (1.17) Other diseases Residue 71 63 5 42 16 (1.08) (0.92) (0.30) (1.13) (1.08) aMortality ratio (in parentheses) standardized for year of birth; the combined experience of the low and high exposure groups is taken as the standard. ^Data from Robinette et al. 1980. 5-296 ------- respiratory tract cancer between those with a Hazard Number less than 5000 vs. more than 5000 and (2) the test for trend for all diseases combined. However, information relevant to respiratory neoplasms such as smoking histories was not available because the study was an analysis of cause of death from death certificate data. Differential health risks with respect to hospitalized illness around the period of exposure were not apparent. Subsequent VA hospitalizations and disability awards provided incomplete information. Since the study focused largely on the use of automated VA record systems, it was not possible to determine non-Navy or non-VA hospitalizations, nonhospitalized conditions, reproductive histories, or subsequent employment histories. Actual individual exposure could not be reconstructed retro- spectively, thus the information represents the potential exposure of the men. 5.10.3 Ocular Effects Many surveys on ocular effects have been conducted. The potential of RF radiation to induce cataracts and the production of lens defects and opacities have been cited in both American and foreign literatures. Ocular effects have, in fact, been the major end point of study in U.S. research, primarily in military populations. For this health end point more than others, much attention has been devoted to careful, detailed clinical examinations and to use of standard procedures or protocols. Population selection criteria, however, is not well elaborated; therefore, the possibility of selection bias cannot be excluded for most of these surveys. Ocular studies have largely taken the form of cross-sectional clinical surveys in actively working popula- 5-297 ------- tions and, as such, have focused on minor lens changes. Longitudinal studies of populations with known exposure, with followup of either a prospective or retrospective nature, are nonexistent despite the identification of potential cohorts from the cross-sectional surveys. The research to date has not yet fully examined the question of whether lens changes, if observed, lead to overt clinical eye disease or conditions, e.g., the functional and clinical significance of minor lens changes in general and over time. Thus, while it appears reasonable that RF radiation at high levels could influence cataract- ogenesis, it cannot be confirmed without more rigorous followup studies in already identified exposed populations. This would be a reasonable line of future research if such epidemiological studies are feasible. Utilizing Veterans Administration hospital records and military personnel records, Cleary et aj. (1965) conducted a case-control study to examine cataract formation among Army and Air Force veterans. This is the only case- control study focused on clinically diagnosed cataracts. Cases were defined as white males born after 1910 who were treated for cataracts between 1950 and 1962, based on diagnoses given in VA hospital records. Controls were drawn from the same sources by selecting men with adjacent hospital register numbers; therefore, the control group had random diagnoses made in the same hospitals at the same time the cataract cases were diagnosed, and the controls were also born after 1910. Military occupational specialties, as abstracted from service records, were used to denote radar or nonradar workers. Job classification was then used as an indicator of potential exposure. The distribution of cases and controls according to radar exposure is seen in Table 5-28. The frequency of microwave exposure as denoted by radar-work 5-298 ------- history was similar between cases and controls; the odds ratio was less than one (0.67) and was not statistically significant. The distribution over all age groups is seen in Table 5-29. Excess risk was noted for Air Force veterans, but the numbers were small. Only 40 radar workers were found in over 5000 cases and controls. TABLE 5-28. CLASSIFICATION BY MILITARY OCCUPATION OF WORLD WAR II AND KOREAN WAR VETERANS WITH AND WITHOUT CATARACTS, BASED ON DISCHARGES FROM VA HOSPITALS* Veterans' Cataract Status Occupation Yes No Total Radar Workers 19 21 40 Nonradar Workers 2625 1935 4560 TOTAL 2644 1956 4600 Odds Ratio = OR = [21] (2625) = °'67 Data are from Cleary et aK 1965. Cleary and Pasternack (1966) examined and scored subclinical or minor lens changes in 736 microwave workers and 559 controls. A relative exposure index or score was developed by weighing various parameters such as power output and distance from the microwave generating source. The work environ- ment in terms of microwave parameters was not specified. Utilizing regressi techniques, radar workers were found to have more lens changes than controls and the number of defects was correlated with duration of exposure and expo- sure score among the workers. The authors suggested that the subclinical changes may indicate accelerated aging of lens tissues. 5-299 ------- TABLE 5-29. ESTIMATED RELATIVE RISK OF CATARACTS AMONG ARMY AND AIR FORCE VETERANS Cataracts Radar Workers Nonradar Workers Odds Ratio 2 X Total Yes No 19 21 2625 1935 0.67 1.26 NS 20-39 Yes No 3 4 418 522 0.94 0.08 NS 40-49 Yes No 7 13 1517 1101 0.39 3.39 NS 50-59 Yes No 9 4 699 316 1.02 0.09 NS Data are from Cleary et aK 1965. Majewska (1968) studied the eyes of 200 Polish workers employed from 6 months to 12 years at installations with microwave-generating equipment that operated from 600 MHz to 10.7 GHz (2.8 to 50 cm). Although cited as "high intensity" by the author, intensity levels were not specified. Two hundred comparably aged unexposed controls were also examined. No methods of subject selection were reported, nor was the sex of the participants reported. After dilation of the subjects' pupils, lenses were examined with an ophthalmoscope and a slit lamp. It was not stated whether procedures ("blinding") were applied to mask the group assignment of the subjects for the examiners. Lens changes were noted in 168 of the subjects vs. 148 controls. This difference was calculated as statistically significant. The result was presented as a summary measure over all ages; age-specific differences were not presented. The authors stated that the controls were age-matched. 5-300 ------- In the same study, the effects of longer-term exposure were evaluated by comparing 100 controls with 102 employees, drawn from the original group, who had worked with high-frequency-electromagnetic-wave generators for more than 4 years. Subjects were graded on a five-point scale for degree of lens opacity. The mean grade of opacities in the exposed group was greater than in controls in each 5-year age-group interval for ages ranging from 20 to 50 years. Among microwave workers, the mean grade of lens changes, uncontrolled for age, also showed an increase with length of employment. Although this part of the study was stated to be focused on employees with four or more years of work experience, the data on lens scores and duration of employment list results for persons with less than four years of employment. This evaluation also suffers from a lack of quantitative measures of exposure. In 1972, Appleton and McCrossan reported a clinical survey conducted among 226 military personnel stationed at Fort Monmouth, New Jersey. Microwave-exposed workers were defined as those who worked with Signal Corps electronic communication, detection, guidance, and weather equipment. Controls did not report such a work history. Likely intensities or frequencies were not specified. There were 135 controls and 91 exposed personnel. Personnel were drawn from the post's Occupational Vision Program, but sampling or selection procedures were not documented. Ophthalmologists based diagnoses on si it-lamp examination of dilated pupils. The presence or absence of opacities, vacuoles, and posterior subcapsular iridescence was noted. The examination results were similar in the two groups over all age groupings, although no statistical tests were applied. 5-301 ------- The survey was later extended to six other installations with results reported by Appleton in 1973. The exposure of military personnel was classified as "likely" vs. "unlikely." The latter group served as controls. Blind procedures were utilized for the examining ophthalmologists; that is, they were not aware of the exposure classification of subjects. The same team of examiners performed all tests, except at one location, in an attempt to minimize inter-observer variation. The same three end points used in the 1972 study were also used in the 1973 study. Older age groups demonstrated a trend of greater opacities among exposed personnel, but, since the numbers in some age groups were small, the validity of this result is questionable. Odland (1973) reported results of a survey of ocular anomalies in personnel from eight military installations. The population consisted of 377 exposed individuals and 320 controls. Exposure conditions were not specified. Exposed personnel were defined as individuals whose primary duties involved the operation or maintenance of radar equipment. The selection criterion for controls was duty that did not permit actual or potential exposure to radar. The actual work assignments of control persons were not stated. Medical histories were taken, and ophthalmic examinations were performed in a double- blind manner. The occurrence of lens anomalies was similar in the two groups; however, the frequency of anomalies between control and exposed groups was different for individuals who had a family history of diabetes, nontraumatic cataract, glaucoma, or defective vision. Lens changes were noted in 29 percent of the exposed individuals with such a history vs. 17 percent in the controls with a family history of eye problems. No statistical tests were applied to any of the reported frequency distributions. 5-302 ------- Shack!ett et aK (1975) reported eye examinations of 817 military and civilian personnel. There were 477 persons with a history of microwave exposure and 340 controls without exposure drawn from eight Air Force bases between November 1971 and December 1974. The authors stated that detailed work histories were recorded (including time with different types of equip- ment), but information on typical exposure settings is not given. Local unit commanders selected the subjects by using criteria established by the examining team. Standard diagnostic criteria were established. The same ophthalmologists performed all examinations and were not aware whether a subject was considered as exposed or a control. No differences were noted between the two groups in the frequency of opacities, vacuoles, and posterior subcapsular iridescence. Differences in results were not statistically significant, but the type of test was not stated. An age-dependent increase in lens changes was noted in both groups. The study by Siekierzynski et aT. (1974b) discussed earlier also compared 2 lens opacities in the two exposure groups (less than 0.2 mW/cm and 0.2 to 6 2 mW/cm ). Ophthalmologic examinations were performed. Lens translucency was assessed with a slit lamp after pupil dilatation and according to criteria established for five grades. No differences were reported between the exposure groups nor within the groups for duration of exposure. 5.10.4 Reproductive Effects In 1965, Sigler et aK reported a history of occupational exposure to radar and more military service among fathers of children with Down's Syndrome. 5-303 ------- The association with radar exposure was an ancillary observation; this study was specifically directed toward examination of the relationship between ionizing radiation exposure and Down's Syndrome. A case-control approach was used, and 288 children born with Down's Syndrome in Baltimore between January 1946 and October 1962 were identified for inclusion in the study. Controls were selected by matching each case of Down's Syndrome with another normal birth for (1) hospital of birth, (2) sex, (3) date of birth, and (4) maternal age at birth of child. Of the original 288, 216 matched pairs were available for final analysis. Eliminations occurred for various reasons such as non- cooperation or equivocal diagnoses. Occupational histories and other data were obtained by interview of the parents. The fathers of the children with Down's Syndrome had more military service experience than control fathers, but the difference was not statistically significant. A greater history of radar exposure, as radar technicians or operators, was reported by case fathers. This difference was statistically significant. It should be noted that radar operators are not necessarily "exposed," since the place of operation may be away from the power-generating equipment or the microwave source. Cohen et aK (1977) extended the study. The case series was expanded to include 128 additional verified cases and their matched pairs born through 1968. To serve as an independent replication, essentially the same procedures were applied along with the following expansions: (1) more extensive questions on microwave/radar exposure and military service, (2) validation of exposure histories by searching armed service records, and (3) chromosomal studies. The previously noted differences disappeared in the extended analysis as seen in Table 5-30. Results of the cytogenetic analyses are not yet available. 5-304 ------- TABLE 5-30. PATERNAL RADAR EXPOSURE BEFORE CONCEPTION OF INDEX CHILD (FROM INTERVIEW AND/OR NAS)a Down1 s Cases Controls Case Series'3 No. % No. % 1. Current Exposed 20 15.7 27 21.3 Exposure Known 127 100.0 127 100.1 2. Original Exposed 36 18.6 30 15.2 Exposure Known 194 100.1 198 100.0 3. Combined Exposed 56 17.4 57 17.5 Exposure Known 321 99.9 325 100.0 aData are from Cohen et aK 1977. ^1: 128 pairs; 2: 216 pairs; 3: 344 total pairs. The authors stated that although the extended study did not confirm excess radar exposure among Down's case fathers, the possible relationship of such exposure to increased risk of Down's offspring cannot completely be ruled out. They added further that the most challenging aspect of the investigation was the definition of radar "exposure." In discussing explanations for lack of confirmation of a radar exposure factor (if one does in fact exist) the authors offer several possibilities. As mentioned above, inaccurate exposure estimates could distort results. The role of maternal factors in Down's Syndrome is so important that paternal factors could be masked. It was also suggested that the most severely impacted males WitK the highest risk of abnormal offspring also experienced germinal tissue damage, eliminating them from reproductive experience or ascertainment through live birth cases. It 5-305 ------- was further suggested that a prospective approach might, then, prove more fruitful. Although not discussed by the authors, radar equipment and military occupational specialties could have changed over time in such a way as to lower risk. It has also been suggested that since microwave-generating equipment, especially of an older vintage, may have emitted more ionizing radiation than modern equipment, the ionizing radiation presents the actual risk factor, with microwaves possibly operating as a co-variable. Lancranjan et cH. (1975) studied 31 adult males with a mean age of 33 years and a mean exposure of 8 years (a range of 1 to 17 years) to electro- magnetic fields that "frequently were in the range of tens to hundreds of 2 pW/cm ." The frequencies were defined as microwaves of wavelengths between 3 and 12 cm and frequencies between 10,000 and 3,600 MHz. No details on the RF source(s) were provided. A group of 30 men of similar mean age and no known exposure to microwaves served as a control for the analysis of spermatic fluids and hormones. Statistical analysis of the results showed no differences in urinary content of 17 keto-steroids (as an indirect measure of Leydig cell function) or total gonadotropin between the exposed and controls. Statistically significant decreases were reported for exposed personnel in the number of sperm per milliliter of semen, percent of motile sperm in the ejaculate, and the percent of normal sperm. The goal of this study was to try to apply objective measures to assess the subjective reports of decreases in libido or of other sexual disturbances. This was accomplished, but the exposures are poorly defined, and the number of 5-306 ------- men evaluated was very small. Spermatogenesis improved in two-thirds of the subjects after cessation of exposure. The investigators felt this supported an argument for microwave influence on the observed alterations. It is interesting to note that the values obtained by the semen analyses for both exposed and control groups might be considered to be low-normal or below normal values. Table 5-31 summarizes those studies on human beings for which SAR values could be estimated. 5.10.5 Unresolved Questions There are some general issues surrounding the use and applicability of epidemiological research methods to study the effects of environmental agents, including radiofrequency radiation. Examples of such issues include the ability of epidemiological studies to detect low-level risks, to separate the effects of multiple causes, and to identify and control confounding factors. But, in addition, there are specific problems seen in the literature on human beings exposed to RF radiation. The reports in the literature have some important and common problems that limit interpretation and use of the data in determining population exposure levels. These problems are briefly discussed i bel ow. 5.10.5.1 Exposure Assessment-- This is perhaps the largest single problem in conducting epidemiological studies on RF radiation. It is difficult to determine the actual exposure of individuals, and even of groups, for a variety of reasons such as difficulties 5-307 ------- in developing methods of personal dosimetry for this type of radiation. There may be no continuous surveillance programs in the workplace, for example, that could yield data for use in epidemiological studies. In many cases, health studies, especially of chronic conditions, are conducted at a point in time when it is hard, if not impossible, to reconstruct exposure patterns. The study by Robinette and Silverman (1980) is a good example of attempts to deal with the difficulties of retrospective exposure assessment, e.g., two approaches (job type vs. Hazard Number) were used and an exposure gradient was obtained with the Hazard Number which simultaneously considered the RF sources and the length of service assignment. Despite these efforts, the results are still estimates of potential rather than actual exposure. On the other hand, the levels and frequencies of exposure are either not known, not estimated, or not reported in many other studies. If well-developed exposure data is not available, it is difficult to develop dose-response relationships, to inter- pret the significance of human health studies, and to use the data in establishing protective exposure limits. 5.10.5.2 Documentation and Methods-- Another problem in the RF radiation literature on human beings relates to documentation of methods and procedures. Degree of detail in reporting seems to be a major difference between studies done in the U.S. and those conducted in other countries. The paper "Guidelines for Documentation of Epidemiologic Studies" (Epidemiology Work Group 1981) suggests the types of topics that are useful to document when reporting an epidemiologic study, especially one used to support regulatory decisions. These major elements include background and objectives, study and comparison subjects, data collection procedures, and 5-308 ------- analysis. Few reports on human beings exposed to RF radiation adequately address these topics. It is frequently hard to tell whether certain research methods were applied or simply not reported. The selection and use of control groups is an example. The criteria for selecting controls are frequently not stated. Controls are often said to be comparable in all respects except exposure but analyses or data to support such statements may not be supplied. Also, practices such as development of standardized rates or use of procedures to control confounding variables, e.g., age adjustment, are not common in the literature. Statistical power is rarely evaluated or discussed; however, it is hard to estimate this if the underlying prevalence or incidence of the disease under study is not well known. This is especially true for some of the conditions and symptoms, e.g., functional disturbances, studied in relation to RF radiation exposure. 5.10.5.3 Health End Points: Design and Populations— Another issue is the medical significance of any changes that may be induced by exposure to RF radiation. For example, the studies on ocular effects usually have examined a subclinical end point, e.g., lens opacities, which may not necessarily be an early marker or risk factor for cataracto- genesis and visual problems (Silverman 1979). Further, the study populations have not been followed to develop incidence or longitudinal data; thus, most studies present only single measures at one point in time. Future studies should try to determine whether symptoms and subtle changes lead to any sub- sequent disability or disease; longitudinal data needs to be developed. A similar concern surrounds the problematic information on functional changes and nervous system effects reported in some of the Eastern European 5-309 ------- literature. Not yet substantiated, the potential for neurological or behavioral effects needs to be more thoroughly and rigorously evaluated. In addition, standardized questionnaires and more objective medical measurements need to be applied in such studies. Most studies concern occupational groups of relatively young healthy males. It cannot be presumed that sensitivity, or lack thereof, to radio- frequency radiation exposure would be the same in the general population as in working groups. The general population is more diverse with the full range of ages, sexes, races, and other factors that could influence health status or the development of disease. To resolve this issue, specific exposed popula- tions could be identified and evaluated if such research appears to be feasible. 5.10.5.4. Summary— There are some serious methodological problems in the human studies literature that meke results equivocal. These problems are not necessarily insurmountable. But, at present, the data on human beings exposed to RF radiation is not yet sufficiently developed to be very useful in determining exposure limits for the general population. 5-310 ------- TABLE 5-31. SUMMARY OF SELECTED HUMAN STUDIES CONCERNING EFFECTS OF RF-RADIATION EXPOSURE Exposure Conditions Effects Species Frequency (MHz) Intensity (mW/cm2) Duration (Years) SAR (V/kg) (est) Reference No significant change in health status of exposed personnel No differences in three major diagnostic categories between the two groups of microwave workers No differences observed in clinical evaluations More subjective complaints in exposed group No effect on life span or cause of death No effect on mortality in a military population followed for more than 20 years Decreased number of sperm/ml of ejaculate Reduced percentages of normal and motile sperm in ejaculate Human adul t male Human adult male Human adul t male Human adult male and female Human adul t male Human adul t male 400 (PV) 2,880 (PW) 9,375 (PW) Radar Radar 2,560-4,100 600-9,500 -4 -4 (avg) -4 (avg) <0 2 (avg) >0 2 (avg) 6 0 (max) <5 3,600-10,000 Tens to hundreds jjW/cm2 0-13 1-10 5-10 0 005 (max) 22 0 018 (max) 0 5 200-5000 (est, PW) ~1 (routine) 100 (occasional) 1-17 8 (avg) ~0 16 ~0 12 -0 12 <8x10 -3 n"3 >8 x 10 24 (max) <0 2 -4 2 x 10 (max) . 7 x 10 (max) <0 05 <5 Barron and Baraff 1958 (See also Barron et M 1955) Czerski et at 1974, Siekierzynski et 1974a, 1974b Djordjevic et aj^ 1979 Li 1lenfeld et al 1978 Robinette et al 1980 0 3-4 x 10 2 Lancranjan et a^ 1975 'Number of years of irradiation of the embassy and length of the study period, but the average exposure of individuals is estimated to be 2-4 years -------


United States
Environmental Protection
Agency
Health Effects Research
Laboratory
Research Triangle Park NC 27711
epa- 600883026B
June 1983
External Review Draft
Research and Development
Biological Effects of
Radiofrequency
Radiation
Part 2 of 3
Review
Draft
(Do Not
Cite or Quote)

-------
EPA-600/8-83-026A
June 1983
External Review Draft
Biological Effects of Radiofrequency Radiation
Edited By
Daniel F. Cahill and Joe A. Elder
Health Effects Research Laboratory
Research Triangle Park, North Carolina 27711
NOTICE
This document is a preliminary draft. It has not been formally released by EPA
and should not at this stage be construed to represent Agency policy. It is
being circulated for comment on its technical accuracy and policy implications.
HEALTH EFFECTS RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711

-------
DISCLAIMER
This report is an external draft for review purposes only and does not
constitute Agency policy. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
i i

-------
FOREWORD
The many benefits of our modern, developing, industrial society are
accompanied by certain hazards. Careful assessment of the risk of existing
and new man-made environmental hazards is necessary to establish sound regula-
tory policy. Environmental regulations enhance the quality of our environment
in order to promote the public health and welfare and the productive capacity
of our nation's population.
The Health Effects Research Laboratory conducts a coordinated environ-
mental health research program in toxicology and clinical studies. These
studies address problems in air pollution, radiofrequency radiation, environ-
mental carcinogenesis and the toxicology of pesticides as well as other chemical
pollutants. The Laboratory participates in the development and revision of
air quality criteria documents on pollutants for which national ambient air
quality standards exist or are proposed, provides the data for registration of
new pesticides or proposed suspension of those already in use, conducts
research on hazardous and toxic materials, and is primarily responsible for
providing the health basis for radiofrequency radiation guidelines. Direct
support to the regulatory function of the Agency is provided in the form of
expert testimony and preparation of affidavits as well as expert advice to the
Admi ni strator.
The intent of this document is to provide a comprehensive review of the
scientific literature on the biological effects of radiofrequency radiation.
The purpose of this effort is to evaluate critically the current state of
knowledge for its pertinence and applicability in developing radiofrequency-
radiation exposure guidelines for the general public.
F. G. Hueter, Ph.D.
Director
Health Effects Research Laboratory
i i i

-------
ABSTRACT
This document presents a critical and comprehensive review of the avail-
able literature on the biological effects of radiofrequency (RF) radiation
through 1980. The objective is to determine whether the existing data base
can contribute to the formulation of RF-radiation exposure guidance for the
general public.
The frequency range of concern in this document is 0.5 MHz to 100 GHz,
which includes all the significant sources of population exposure to RF radia-
tion. Research reports that are judged to be credible according to a set of
objective criteria are examined for the relation between the RF energy ab-
sorbed and the presence or absence of biological effects. The reported conse-
quences of the interaction between RF radiation and biological systems are
examined from four perspectives by 1) RF-energy-induced core temperature
increases, 2) whole-body-averaged specific absorption rate (SAR), 3) the
exogeneous energy burden as a percentage of resting metabolic rate, and 4)
actual human experiences documented in epidemiological studies.
The existing data base does lead to tentative conclusions about the
relation between RF-radiation exposure and biological effects that may serve
as unadjusted upper limits for population exposure guidance; however, the un-
certainties and unknowns in the current state of knowledge are potentially
signi ficant.
iv

-------
CONTRIBUTORS
Ernest N. Albert*
Joseph S. Ali
John W. All is
Ezra Berman
Carl F. Blackman+
Daniel F. Cahill
Joe A. Elder
Michael I. Gage
Christopher J. Gordon
Doreen Hill
William T. Joines
James B. Kinn K
William P. Kirk5
Charles G. Liddle
James R. Rabinowitz
Ralph J. Smialowicz
Ronald J. Spiegel
Claude M. Weil
U.S. Environmental Protection Agency
Office of Research and Development
Office of Health Research
Health Effects Research Laboratory
Research Triangle Park, NC 27711
*Department of Anatomy
The George Washington University
Medical Center
Washington, DC 20037
^Current Address:
Carolina Power and Light Company
Harris Energy and Environmental Center
Route 1, Box 327
New Hill, NC 27562
^Office of Health Research
U.S. Environmental Protection Agency
Washington, DC 20406
^Current Address:
Three Mile Island Field Station
U.S. Environmental Protection Agency
100 Brown Street
Middletown, PA 17057
v

-------
CONTENTS
Foreword		i i i
Abstract		iv
Figures		ix
Tables		xii
Acknowledgment 		xv
1	Introduction 		1-1
(Daniel F. Cahill^
1.1	Ground Rules and General Assumptions 		1-1
1.2	General Approach 		1-2
1.3	Specific Approach Used in This Review		1-3
2	Summary and Conclusions 		2-1
(Daniel F. Cahill and Joe A. Elder)
3	Physical Principles of Electromagnetic Field Interactions . . .	3-1
3.1	Electromagnetic Field Theory 		3-1
(William T. Joines)
3.1.1	Electromagnetic spectrum 		3-1
3.1.2	Wave propagation		3-5
3.1.3	Wave modulation		3-11
3.2	RF-Field Interactions with Biological Systems 		3-15
(Claude M. Weil and James R. Rabinowitz)
3.2.1	Scattering and absorption of electromagnetic
waves		3-15
3.2.2	RF dosimetry definitions		3-31
3.2.3	Analytical and numerical RF electromagnetic
interaction models		3-34
3.2.4	Mechanisms of RF interaction with biological
systems		3-52
3.3	Experimental Methods 		3-67
(Claude M. Weil and Joseph S. Ali)
3.3.1	Exposure methods used in biological
experimentation 		3-67
3.3.2	Animal holders		3-97
3.3.3	Densitometric instrumentation 		3-102
3.4	Dosimetric Methods 		3-115
(James B. Kinn)
3.4.1	Whole body dosimetry		3-115
3.4.2	Regional dosimetry		3-119
3.4.3	Unresolved questions		3-122
vi i

-------
Page
4	Effect of RF-Radiation Exposure on Body Temperature 		4-1
4.1	Thermal Physiology 		4-1
(Christopher Gordon)
4.1.1	Temperature regulation 		4-1
4.1.2	Ambient temperature vs. RF-radiation exposure . . .	4-3
4.1.3	Mechanisms of heat gain during RF-radiation
exposure		4-5
4.1.4	Local thermal responses: Effect of RF-radiation
exposure on blood flow		4-11
4.1.5	Whole-body thermal response to RF-radiation
exposure and dependence on ambient
conditions		4-12
4.1.6	Heat stress and the general adaptation syndrome . .	4-17
4.1.7	Thermal physiology of humans 		4-19
4.1.8	Thermoregulatory state and physiological
responsiveness to RF radiation 		4-24
4.1.9	RF-radiation exposure and behavioral temperature
regulation		4-26
4.1.10	Unresolved questions 		4-27
4.2	Numerical Modeling of Thermoregulatory Systems in Man and
Animals		4-33
(Ronald Spiegel)
4.2.1	Heat-transfer models 		4-33
4.2.2	RF-radiation-heat-transfer models 		4-37
4.2.3	Numerical results 		4-42
4.2.4	Unresolved questions 		4-45
5	Biological Effects of RF Radiation 		5-1
5.1	Cellular and Subcellular Effects 		5-1
(John W. Allis)
5.1.1	Effects on molecular systems 		5-3
5.1.2	Effects on subcellular organelles 		5-8
5.1.3	Effects on single cells		5-13
5.1.4	Unresolved questions 		5-23
5.2	Hematologic and Immunologic Effects
(Ralph J. Smialowicz) 		5-31
5.2.1	Hematology		5-33
5.2.2	Immunology		5-46
5.2.3	Unresolved questions 		5-71
5.3	Reproductive Effects 		5-73
(Ezra Berman)
5.3.1	Teratology		5-73
5.3.2	Reproduction efficiency 		5-99
5.3.3	Testes		5-103
5.3.4	Unresolved questions		5-112
v i i i

-------
Page
5.4	Nervous System		5-117
(Ernest N. Albert)
5.4.1	Morphologic observations 		5-119
5.4.2	Blood-brain barrier studies 		5-122
5.4.3	Pharmacological effects 		5-126
5.4.4	Effects on neurotransmitters 		5-129
5.4.5	Unresolved questions 		5-131
5.5	Behavior		5-137
(Michael I. Gage)
5.5.1	Introduction		5-137
5.5.2	Summary		5-139
5.5.3	Naturalistic behavior 		5-140
5.5.4	Learned behavior 		5-145
5.5.5	Interactions with other environmental stimuli . . .	5-159
5.5.6	Unresolved questions 		5-163
5.6	Special Senses		5-171
(Joe A. Elder)
5.6.1	Cataractogenic effects 		5-171
5.6.2	Unresolved questions 		5-184
5.6.3	Auditory effects 		5-186
5.6.4	Unresolved questions 		5-199
5.6.5	Human cutaneous perception 		5-203
5.6.6	Unresolved questions 		5-206
5.7	Other Physiological and Biochemical Effects 		5-209
(Charles G. Liddle)
5.7.1	Clinical chemistry and metabolism 		5-209
5.7.2	Endocrinology		5-219
5.7.3	Growth and development		5-225
5.7.4	Cardiovascular system 		5-228
5.7.5	Calcium ion efflux		5-233
5.7.6	Unresolved questions		5-237
5.8	Genetics and Mutagenesis		5-241
(Carl F. Blackman)
5.8.1	Introduction		5-241
5.8.2	Effects on genetic material of cellular
and subcellular systems 		5-243
5.8.3	Effects on genetic material of higher-order
biological systems 		5-252
5.8.4	Unresolved questions		5-256
5.9	Life Span and Carcinogenesis		5-267
(William P. Kirk)
5.9.1	Life span		5-267
5.9.2	Carcinogenesis		5-273
5.9.3	Prausnitz and Susskind study		5-276
5.9.4	North Karelia connection		5-279
5.9.5	Moscow Embassy study		5-280
5.9.6	U.S. Navy study		5-281
5.9.7	Unresolved questions		5-282
ix

-------
Page
5.10 Human Studies		5-285
(Doreen Hill)
5.10.1	Occupational surveys		5-285
5.10.2	Mortality studies 		5-291
5.10.3	Ocular effects 		5-297
5.10.4	Reproductive effects 		5-303
5.10.5	Unresolved questions 		5-307
6 Assessment		6-1
(Daniel F. Cahill and Joe A. Elder)
6.1	Core Temperature		6-2
6.2	Specific Absorption Rate		6-5
6.3	Resting Metabolic Rate		6-13
6.4	Human Studies		6-18
References		R-l
Glossary		G-l
x

-------
FIGURES
Number	Page
3-1 Far-field electromagnetic wave at a particular instant
in time	 3-8
3-2 Power density vs. distance along axis from antenna
aperture	 3-12
3-3 Interaction of RF radiation with electrical conductors,
biological tissue, and electrical insulators 	 3-16
3-4 Energy distribution in proximity to man at 1 GHz at the
chest plane contour presentation	 3-19
3-5 Dielectric data for tissues in RF range 0.01 to 10 GHz	 3-22
3-6 Illustration of object size vs. wavelength dependence 	 3-24
3-7 Whole-body average SAR vs. frequency for three polarizations
in a prolate spheroidal model of a human	 3-26
3-8 Absorption dependence on various ground and multi-path
factors	 3-30
3-9 ARD distribution in core of 6-cm radius multi-layered
sphere at 1650 MHz	 3-42
3-10 Curve fitting of SAR data for a prolate spheroidal model
of man for three basic orientations	 3-44
3-11 Effect of a capacitive gap on average SAR between the man
model and the ground plane	 3-46
3-12 A realistic block model of man	 3-48
3-13 Absorption for man block model standing on ground plane .... 3-50
3-14 The real component of the complex permittivity of muscle
as a function of frequency	 3-55
3-15 EPA 2450-MHz anechoic chamber facility or 2.45-GHz far-field
facility	 3-72
xi

-------
Number	Page
3-16 EPA 2450-MHz anechoic chamber facility: diagram of the
microwave exposure facility 		3-73
3-17 Diagram of absorber-1 ined horn		3-74
3-18 Diagram of a point-source compact range 		3-75
3-19 Miniature anechoic chamber facility 		3-77
3-20 Photograph of tapered exposure chamber at EPA facility ....	3-78
3-21 Facility for simultaneous exposure of 10 animals with
minimal inter-animal interaction 		3-79
3-22 Monopole-over-ground plane irradiation facility 		3-81
3-23 Coaxial air-line system for high power exposures of cell
cultures		3-86
3-24 Parallel-plate (microstrip) exposure system 		3-87
3-25 Block diagram of complete RF near-field synthesizer 		3-87
3-26 EPA 100-MHz rectangular strip line or Crawford cell		3-89
3-27 Circularly polarized 915-MHz waveguide facility 		3-91
3-28 Photograph of exposure chamber with associated instrumentation
or the 970-MHz circularly polarized waveguide facility at EPA .	3-92
3-29 VHF resonant cavity facility 		3-95
3-30 Multimodal cavity facility for primate irradiation 		3-96
3-31 Water-supply system for exposure chamber 		3-102
3-32 Samples of commercially available survey meters for
measuring RF electric-field strength 		3-106
3-33	Microprocessor-controlled twin-well calorimeter 		3-118
4-1	Simple neural model of thermoregulation in a mammal 		4-4
4-2 Simple model of temperature gradients in a homeotherm 		4-6
4-3 Power densities at 2450 MHz necessary to raise rectal
temperature 1 °C in 60 min		4-8
4-4 SAR as a function of body weight		4-10
xi i

-------
Number	Page
4-5 Idealistic response of a homeotherm's metabolism and body
temperature to changes in ambient temperature 	 4-13
4-6 SAR at 2450 MHz, ambient temperature effects, and EHL
in mice	 4-16
4-7 Effect of an increasing THI on the lethal dose of RF
radiation in mice	 4-18
4-8 Effects of increasing skin temperature on sweating in
humans	 4-21
4-9 Effect of exercise on heat loss, metabolic rate, and
tympanic temperature of humans 	 4-22
4-10 Effects of 5-HT injections on mice	 4-25
4-11 Block diagram for one segment of the thermal model	 4-40
4-12	Incident power density vs. exposure duration to obtain a
hot spot	 4-45
5-1	Arrhenius plot of Na+ efflux	 5-18
5-2 Summed incidence of abnormal and nonviable chick
embryo eggs exposed to radiation 	 5-78
5-3 Cross-sectional sketch of the human and the rabbit eye 	 5-174
5-4 Time and power density threshold for cataractogenesis in
rabbits	 5-177
2
5-5 Distribution of energy absorption rate per mW/cm
incident power density in the rabbit's eye and head
exposed to 2450-MHz radiation 	 5-178
2
5-6	Distribution of energy absorption rate per mW/cm
incident power density in the rabbit's eye and head
exposed to 918-MHz radiation 	 5-179
6-1	Steady-state temperature vs. SAR and power density	 6-6
6-2 Relationship of body mass to the SAR necessary to alter
the activity of various physiologic systems 	 6-17
xi i i

-------
TABLES
Number	Page
3-1 Radiofrequency Bands 	 3-4
3-2 Proposed System of RF Dosimetric Quantities, Definitions,
and Units	 3-32
3-3 Range of Resonant Frequencies of Man and Animals Irradiated
by Plane Waves in Free Space at 1 mW/cm2 With Long Axis
Parallel to Electric Field 	 3-37
3-4 Dielectric Permittivities for Various Tissues 	 3-54
3-5	Energy Units for RF Radiation	 3-58
4-1	Partitioning of Heat Loss in Humans as a Function of
Ambient Temperature 	 4-20
4-2 Summary of Studies Concerning RF-Radiation Effects on
Thermoregulation 	 4-30
4-3 Summary of Studies Concerning RF-Radiation
Effects on Blood Flow	 4-32
4-4	Steady-State Temperatures in a Human Body after Exposure to
80- and 200-MHz RF Fields	 4-43
5-1	Classification of Cellular and Subcellular Experiments	 5-2
5-2 Summary of Studies Concerning RF-Radiation Effects on
Molecular Systems 	 5-5
5-3 Summary of Studies Concerning RF-Radiation Effects on
Subcellular Systems 	 5-9
5-4 Summary of Studies Concerning RF-Radiation Effects on
Single Cells	 5-14
5-5 Summary of Studies Concerning Hematologic Effects of
RF-Radiation Exposure 	 5-34
5-6 Summary of Studies Concerning Immunological Effects (Li Vivo)
of RF-Radiation Exposure	 5-50
xiv

-------
Number	Page
5-7 Summary of Studies Concerning Immunologic Effects (In Vitro)
of RF-Radiation Exposure 	 5-66
5-8 Conversion of J/g to W/kg	 5-82
5-9 Summary of Studies Concerning Teratologic Effects of
RF-Radiation Exposure 	 5-100
5-10 Summary of Studies Concerning Reproductive Effects of
RF-Radiation Exposure 	 5-103
5-11 Summary of Studies Concerning Reproductive Effects of
RF-Radiation Exposure in the Rat	 5-112
5-12 Summary of Studies Concerning RF-Radiation Effects on the
Nervous System 	 5-133
5-13 Summary of Studies Concerning RF-Radiation Effects on
Behavior	 5-164
5-14 Summary of Studies Concerning Ocular Effects of Near-Field
Exposures	 5-172
5-15 Summary of Studies Concerning Ocular Effects of Far-Field
Exposures	 5-173
5-16 Summary of Studies Concerning Auditory Effects of RF Radiation
in Humans	 5-188
5-17 Summary of Studies Concerning Threshold Values for Auditory-
Evoked Potentials in Laboratory Animals 	 5-201
5-18 Summary of Studies Concerning Human Cutaneous Perception of
RF Radiation	 5-204
5-19 Summary of Studies Concerning RF-Radiation Effects on
Clinical Chemistry and Metabolism, Endocrinology, and
Growth and Development 	 5-210
5-20 Summary of Studies Concerning RF-Radiation Effects on
Various Aspects of Cardiac Physiology	 5-234
5-21 Summary of Studies Concerning Genetic and Mutagenic
Effects of RF-Radiation Exposure	 5-263
5-22 Summary of Studies Concerning RF-Radiation Exposure
Effects on Life Span/Carcinogenesis	 5-268
5-23 Summary of Prausnitz and Susskind Data	 5-271
xv

-------
Number	Page
5-24 Distribution of Years of Exposure for 226 Radar Workers		5-287
5-25 Age Distribution of 226 Microwave Workers and 88 Controls		5-287
5-26 Microwave Exposure Levels at the U.S. Embassy in Moscow		5-292
5-27 Number of Deaths from Disease and Mortality Ratios by
Hazard Number 	 5-296
5-28 Classification by Military Occupation of World War II and
Korean War Veterans With and Without Cataracts 	 5-299
5-29 Estimated Relative Risk of Cataracts Among Army and Air
Force Veterans	 5-300
5-30 Paternal Radar Exposure Before Conception of Index Child 	 5-305
5-31	Summary of Selected Human Studies Concerning Effects of
RF-Radiaton Exposure 	 5-311
6-1	Studies Reporting "No Effects" at SAR's < 10 W/kg Grouped
by Biological Variable 	 6-10
6-2 Studies with Reported "Effects" at SAR's < 10 W/kg Grouped
by Biological Variables	 6-11
6-3 Equivalent Power Density for Five Ages and Mass Sizes at
Resonant Frequency for SAR = 0.4 W/kg	 6-14
6-4 The Increased RMR and Equivalent SAR and Power Density
Predicted to be Associated with the Onset of Human
Thermoregulatory Response	 6-18
6-5 Summary of Estimates of Unadjusted Limits for RF-Radiation
Exposure at Resonant Frequencies 	 6-21
xv i

-------
ACKNOWLEDGMENT
The authors wish to thank Jan Parsons and Carole Moussali, Northrop
Services, Inc., for editorial review of this document, and Linda Jones and
Connie Van Oosten, Northrop Services, Inc., for the word processing.
Wanda Jones, Bonnie Waddell, and Barbara Queen are also to be commended for
their word processing assistance on earlier versions of this document.
xv ii

-------
SECTION 5
BIOLOGICAL EFFECTS OF RF RADIATION
5.1 CELLULAR AND SUBCELLULAR EFFECTS
John W. All is
The investigation of RF-radiation effects on cellular and subcellular
systems is principally an attempt to eludicate specific biochemical mechanisms
for the interaction of the radiation with macroscopic biological systems.
This approach takes advantage of the relative simplicity of in vitro systems,
the ability to control variables, the rapid and economical way with which
results may be obtained, and the perceived ease of mechanistic interpretation
of the results. In practice, however, these advantages are not always
realized.
The divisions between molecular, subcellular, and cellular systems are
somewhat arbitrary, but for convenience, results presented in this section
will be discussed in this format. Examples of each classification are
presented in Table 5-1. When investigating molecular and subcellular systems,
one normally looks at a single key structure or function where, one hopes, all
other variables may be held constant. In these cases, thorough understanding
5-1

-------
of the end point under study allows the investigator to interpret results
more easily in terms of a detailed biochemical mechanism. For instance, if a
change in the kinetics of an enzyme-catalyzed reaction is found, the type of
inhibition can be determined, the particular molecules involved can be
identified, and a hypothesis for the effect can be constructed.
TABLE 5-1. CLASSIFICATION OF CELLULAR AND SUBCELLULAR EXPERIMENTS
Classification
Type of System Studied
Examples of
Experimental Measures
Molecular
Purified enzyme preparations
Enzyme activity, binding
to small molecules
Molecular
Purified DNA preparations
DNA melting
Subcellular
organelles
Membrane-bound enzymes
Enzyme activity
Subcellular
organelles
Cell membranes, phospholipid
vesicles
Infrared, Raman spectra
Subcellular
organelles
Isolated mitochondria
Mitochondrial function
Cellular
Red blood cells
Ion transport
Cellular
Bacteria
Growth, survival,
infectivity
Cellular
Isolated neurons
Neuronal firing,
membrane potential
In some experiments on intact cells, these conditions can also be approached.
An example of this is sodium (Na+) and potassium (K+) transport across the
red-blood-cell (RBC) membrane. The RBC's limited metabolic activity is geared
5-2

-------
primarily to performing its main function of transporting oxygen from lungs to
tissue. In this case, ion transport across the membrane is linked to the mode
of energy production in the cell and very little else, and studies on this system
are relatively simple to interpret.
Experiments on most living cells are complicated by complex biochemical
systems, many of which have multiple or alternate pathways. Rather than
isolated functional changes, broad end points such as growth or genetic changes
are often assayed. Here, the investigator tries to use "biological amplification"
where, for example, the cell's growth is highly sensitive to disruptions of a
critical but unspecified metabolic step, or where a mutant appears that must
reproduce over several generations before sufficient numbers are present to be
measurable. The investigator also recognizes that the redundancies in a
complex system like a living cell may negate any efforts to detect the putative
effect.
This inherent complexity makes it difficult to ascribe a detailed biochem-
ical mechanism to effects found. Such effects must be viewed as a starting
point from which a mechanism may eventually be deduced. On the other hand,
effects on cells can also be used to rationalize or interpret—perhaps even pre-
dict—effects in the intact animal.
5.1.1 Effects on Molecular Systems
To obtain insights into possible direct interactive mechanism(s) of micro-
wave radiation at the molecular level, several investigators have attempted
5-3

-------
to measure changes in biologically important macromolecules exposed i_n vitro
(Table 5-2). Enzyme kinetics studies form a majority of the reports, but a
few examine changes in macromolecular structure. In one of the latter, Hamrick
(1973) measured DNA melting curves after exposure of calf thymus DNA to 2.45-GHz
(CW) radiation for 16 h (SAR = 67 W/kg) and at dose rates to 160 W/kg for 1 h.
Temperature was controlled during all exposures, usually at 37 °C, but for
some experiments it was maintained at 40, 45, and 50 °C. The intent of the
study was to determine if microwaves disrupted hydrogen bonding between the
DNA strands of the double helix, thereby affecting the melting curve. The
assay was carried out after exposure. In one experiment, any disruption of
hydrogen bonding between the strands was preserved by performing the microwave
exposure with formaldehyde in the buffer. However, all melting curves were
virtually identical to those of unexposed, temperature-matched controls.
In another study of macromolecular structure (Allis 1975), the protein
bovine serum albumin (BSA) was exposed to 1.70- and 2.45-GHz (CW) radiation
(SAR's ranging from 30 to 100 W/kg). The author recognized that structural
changes due to microwave exposure may be reversible. In this study, the
problem was attacked by developing an exposure apparatus in which ultraviolet
(UV) and visible spectrophotometry measurements could be performed during ex-
posure. The difference in UV absorption between the exposed and unexposed
samples was measured directly by using a double-beam spectrophotometer.
Differences of this type reflect small changes in the surroundings of certain
UV-absorbing amino acids and, in this case, could be interpreted as changes
in the structure of the protein. The temperature of the exposed samples was
5-4

-------
TABLE 5-2. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON MOLECULAR SYSTEMS
Exposure Conditions
End Point Measured/
Effects
Experimental
System
Exposure
Frequency Duration Facility SAR
(GHz)	(nin) (Type)	(WAg)
Reference
No change in UV difference spectra
measured over pH range 2 5-5.5
UV spectra and binding constants
for nononucleotides showed no
difference from controls
No change in enzyme activity
No difference in melting curves
Inactivation of enzyme, probably
temperature inhooogeneity effect
at very high doses
Heat inactivation of enzymes found
at highest SAR (T = 50 CC),
corresponded closely to heat-
treated controls
Heat inactivation of enzyme found
at SAR's > 165 WAg
BSA*
Ribonuclease
Glucose-6-phosphate
dehydrogenase;
adenylate kinase, NADPH
cytochrome C reductase
DMA
1	70 (CW)
2	45 (CW)
1	70 (CW)
2	45 (CW)
2 45 (CW)
2 45 (CW)
Horseradish peroxidase 2 45 (CW)
Glucose-6-phosphate	2 8 (PW)
dehydrogenase, lactate
dehydrogenase; acid
phosphatase, alkaline
phosphatase
Lactate dehydrogenase 3 0 (CW)
30
30
60,
960
5, 10,
20, 30,
40
4 5,
18 5
20
Waveguide 30-100
Waveguide 39
Waveguide 42
Allis (1975)
Allis et al. (1976)
Ward et al (1975)
Far field 67, 160 Hamrick (1973)
Waveguide 62,500- Henderson et al^ (1975)
375,000
Waveguide
200-500 Belkhode et al (1974a,b)
Waveguide 33-960
Bini et al (1978)
*BSA = bovine serum albumin

-------
controlled during exposure, and it was matched by the temperature of the con-
trol sample. Temperatures ranged from 24 to 32 °C depending on the SAR value.
Spectra were measured immediately upon beginning exposure, and again 30 min
later with continuous exposure. The study results showed that no changes
in the UV spectrum could be found over a variety of structural states of the
protein, implying that structural changes due to microwave exposure could not
be inferred.
The ability of microwave radiation to alter enzyme activity has been
studied by several workers. Measurements were performed during microwave
exposure by two groups, each using spectrophotometry measures of enzyme
activity. Ward et al_. (1975) examined three enzymes (glucose-6-phosphate
dehydrogenase, adenylate kinase, and NADPH-cytochrome reductase) exposed to
2.45-GHz (CW) radiation (SAR = 42 W/kg). All exposed and control samples were
maintained at 25 °C. Exposure durations were ~ 5 min, during which the enzyme
activity was measured. No differences between exposed and control samples
were found. Bini et aL (1978) followed the activity of lactate dehydrogenase
exposed to 3.0-GHz (CW) radiation (SAR's between 33 and 960 W/kg). They
demonstrated that the changes found in the enzyme activity were entirely
consistent with calculations of thermal inactivation of the enzyme at the
temperatures attained. The sample exposed at 33 W/kg was not different from
the unexposed control; all other exposures (SAR's ranging from 165 to
960 W/kg) showed evidence of enzyme inactivation.
Belkhode et al_. (1974a,b) reported the effect of 2.8-GHz square-wave
modulated (1-kHz) radiation on four enzymes: glucose-6-phosphate dehydrogenase,
5-6

-------
lactate dehydrogenase, acid phosphatase, and alkaline phosphatase. Enzyme
preparations were analyzed after exposures (SAR's ~ 200 to 500 W/kg on average).
Exposures were conducted at 37, 46.7, and 49.7 °C; enzyme activities were
compared to the activity of sham-exposed samples at the same temperatures.
Activities of the exposed enzymes at each temperature were indistinguishable
from the shams.
Henderson et aK (1975) reported a change in enzyme activity that was
interpreted as an indicator of direct interaction on the enzyme by microwaves.
In this experiment, horseradish peroxidase was subjected to 2.45-GHz (CW)
radiation (SAR's between 62,500 and 375,000 W/kg) with the sample exposed in
a tube (4.7-mm ID) that protruded through a waveguide. The sample tube was
surrounded by a concentric cooling jacket, through which an organic coolant
was pumped continuously to maintain the temperature at 25 °C. Thermocouples
were placed in the sample tube so that the sample temperature could be
monitored from positions just outside the waveguide. The total volume of
the exposed sample was ~ 0.8 ml. A marked decrease in enzyme activity
was found at 62,500 W/kg after 30 min of exposure, and at 187,500 W/kg after
20 min of exposure, even though the temperature was reported never to exceed
35 °C. It is possible that the very high fields present at these SAR's could
produce field-specific effects. However, it appears likely that very high
local heating occurred in the sample that was responsible for enzyme inacti-
vation. This likelihood is substantiated by the work of Harrison et aK (1980)
who performed liquid-crystal thermography under similar exposure conditions.
Temperature rises of as much as 0.3 °C were recorded within a micropipette
5-7

-------
suspended in a waveguide and cooled with water circulating through the wave-
guide. Henderson et aH. (1975) exposed their samples at 5 to 10 times the
levels used by Harrison et a_L ; also, the latter researchers used water as a
coolant, which would attenuate the energy reaching the sample much more
strongly than would the organic solvent used by Henderson et aK A study by
Livingston et art. (1979) graphically illustrates the effect of temperature
gradients during microwave exposure.
The binding of small substrate-1ike molecules to the enzyme ribonuclease
was studied to determine if the binding relationship between an enzyme and
substrate could be affected by microwaves (Allis et aH. 1976). In this
work, UV-absorption spectra were measured during exposure to 1.70- and 2.45-GHz
(CW) radiation (SAR = 39 W/kg). Measurements were performed immediately upon
beginning irradiation and after 30 min of exposure. Neither structural changes
in the enzyme-substrate complex nor changes in the binding constants were
found. In sum, no consistent effects on molecular systems exposed i_n vitro to
RF radiation have yet been demonstrated.
5.1.2 Effects on Subcellular Organelles
There has been relatively little research on the effects of microwave
radiation on subcellular organelles (Table 5-3). The reports included in this
document range from work with phospholipid bilayers (i.e., synthetic analogs
of cell membranes) to experiments using intact mitochondria, the energy-producing
system in eukaryotic (e.g., mammalian) cells. Most of the work to be discussed
5-8

-------
TABLE 5-3. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON SUBCELLULAR SYSTEMS
Exposure Conditions
End Point Measured/
Effects
Experimental
System
Exposure
Frequency Duration Facility SAR
(GHz)	(oin) (Type)	(W/kg)
Reference
Increase in exchange of strongly
bound aside hydrogens in membrane
protein measured ty IR spectra
for SAR > 10 W/kg, No change in
o-helix or p-sheet content of
proteins
No change in activity of membrane
bound enzymes measured
spectrophotometries!ly
No change in activity of meobrane
bound enzyae measured spectro-
photometrically
No difference in respiratory
activity
No difference in respiratory
activity
No change in formation of
microtubules
No change in migration of proteins
within axonal membrane
No changes in IR spectra of
proteins and nucleic acids in
E coli exposed before drying
r
RBC membrane
RBC membrane, mito-
chondrial inner
membrane
1 0 (CW)
30
2 45 (SW*) 10
Endoplasmic reticulua 2 45 (CW)
Mitochondria
2 45 (CW) 30 to
210
2-4 (Swept) 10
3.4 (CW)
Mitochondria
Tubulin (rabbit brain) 3.1 (PW) 15
Vagus nerve cell	3 1 (PW) 24 h
Dried film of
E coll cells
3 2 (CW) 8, 10,
11 h
Stripline 5-45
Waveguide 26
Waveguide 42
Anechoic
chamber
far field
Coaxial
Airline
17 5,
87 5
Isoailov (1977)
AllTS et al (1979)
Ward et al (1975)
Elder and All (1975)
16-2.3 Elder et al (1976)
41
Far field 112-430 Paulsson et al^ (1977)
Far field ~ 10-100 Paulsson et al (1977)
Waveguide 20
Corelli et al (1977)
*SW = Sine-wave modulated
tRBC = Red blood cell
#IR = Infrared

-------
here has not demonstrated effects of microwave exposure at dose rates ranging
from 1 to 430 W/kg. Two reports that have indicated an effect do not meet all
criteria for inclusion here and are, therefore, discussed with other reports
that present unresolved questions.
Two reports describe work with enzymes bound to biological membranes.
The study by Ward et (1975), discussed above, focused on the enzyme NADPH-
cytochrome C reductase, which is loosely bound to the membrane of the endoplasmic
reticulum of rat liver cells. All is et al^ (1979) studied adenosine tri-
phosphatase (ATPase) in RBC membranes and cytochrome oxidase in the inner
mitochondrial membrane of rat liver cells. The latter two enzymes are thought
to be integral parts of membranes. Conditions of the exposure were identical
for all three enzymes in that the assay was performed spectrophotometrically
during exposure to microwaves. In the latter study the dose rate was 26 W/kg,
and the 2.45-GHz radiation was sinusoidally modulated at 16, 30, 90, and 120 Hz.
Enzyme activity was not measurably affected by these exposures; a 10- to
15-percent change in enzyme activity would have been required to detect a
reliable microwave effect.
Ismailov (1977) investigated the infrared (IR) absorption spectra of pro-
teins in RBC membranes exposed to 1.009-GHz fields (SAR's up to 45 W/kg) and
maintained at 25 °C. The samples were exposed for 30 min in aqueous suspension
in a stripline, then were dried to a thin film to obtain the IR spectra. No
change in or helix or p-sheet content of the membrane proteins was noted.
However, when D2O (heavy water) was added to the suspension before beginning
5-10

-------
exposure, application of microwaves was found to increase the degree to which
strongly bound amide hydrogens were exchanged. This effect was pronounced at
SAR = 45 W/kg but disappeared when the SAR was below 10 W/kg. The increase in
accessibility of the poorly exchangeable amide hydrogens indicates that the
microwaves disturb the relationship of the protein to its neighboring membrane
lipid. A quantitative estimate of the sensitivity of this experiment cannot
be made from the spectra shown in the paper.
IR spectra of Escherichia coli after exposure to microwaves were also mea-
sured (Corelli et al_. 1977). After 12 h of exposure to 3.2-GHz (CW) radiation
(SAR = 20 W/kg), E. coli were dried to a film, and spectra were measured in the
protein and nucleic acid absorption regions. No differences were found. These
results are comparable with those of Ismailov (1977) that indicated no changes,
but they do not address the conditions for which Ismailov did report effects.
The functional properties of the microtubule assembly system extracted
from rabbit brain cells were studied after exposure to 3.1-GHz fields by
Paulsson et aL (1977). The binding of a drug, colchicine, to the microtubule
precursor protein tubulin was measured after exposure to PW microwaves for
15 min at average dose rates of 112 and 243 W/kg (pulse-repetition frequency
of 200 Hz, pulse duration of 1.4 [is). Colchicine normally blocks the formation
of microtubules, which halts cell division. The normal assembly of microtubules
from tubulin exposed for 10 min to PW microwaves, as above, at 430 W/kg was
also studied. No noticeable effect on either process was found. The data
indicate that a change of about 15 percent in the colchicine binding and about
5-11

-------
10 percent in the microtubule assembly measurement would have been noted.
Paulsson et a^. also studied the migration of proteins within the axonal mem-
brane of the rabbit's vagus nerve. In this case the samples were exposed for
24 h (the SAR estimated from their data was 10 to 100 W/kg) at a pulse-repetition
rate (PRR) of 100 Hz and a pulse duration of 1.4 ps. The distribution of
tritium-labeled protein in the axonal membrane was found to be the same in
exposed and control samples. A difference of > 20 percent would have been
required for detection in this experiment.
Two papers (Elder and Ali 1975; Elder et aK 1976) present results of
exposure of rat liver mitochondria to microwave radiation. Both papers examine
oxygen utilization by measuring respiratory activity (e.g., respiratory control
ratio, ADP to oxygen ratio) under various conditions. These parameters are
functional indicators of the energy production system in eukaryotic cells.
The earlier paper tested mitochondria kept at 0 °C, or inactive state, during
exposure in the far field at 2.45 GHz. The mitochondrial functions were
examined after exposure in the active state at 25 °C. (Exposures for periods
up to 3.5 h were conducted at SAR's of 17.5 and 87.5 W/kg.) No changes in
mitochondrial activity were seen. In the later paper, the disadvantage of ex-
posing inactive mitochondria was overcome by use of a novel flow-through
system that coupled a coaxial airline and an oxygen electrode. The mitochondrial
suspension was cycled continuously between the airline, where exposure was
accomplished, and the oxygen electrode. Samples were exposed to 2.45-, 3.0-,
and 3.4-GHz (CW) radiation (SAR = 41 W/kg) and also to radiation at swept
frequencies between 2 and 4 GHz (SAR's from 1.6 to 2.3 W/kg). As in the
earlier work, no effects of microwave exposure were detected under any
5-12

-------
condition. In general, a five-percent change would have been sufficient for
detection. Other workers have presented data on microwave exposure of mito-
chondria, either in a form too incomplete for inclusion or in oral presenta-
tions. Their findings do not differ from those described here.
5.1.3 Effects on Single Cells
Effects on single cells have been investigated by several researchers
(Table 5-4). Transport and related properties of RBC membranes have been
studied by three groups. In each case K+ transport was used as an end point.
In the RBC, active (or energy requiring) Na+ and K+ transport across the
membrane is by the enzyme Na+-K+ ATPase (the same enzyme discussed in § 5.1.2,
Effects on Subcellular Organelles), and passive transport is through channels
in the membrane. Ismailov (1971) exposed human RBC's to 1.0-GHz (CW) micro-
waves at 45 W/kg and found an increased efflux of K+ and a concomitant
increased influx of Na+. The Na+ influx was twice as large, ion per ion, as
the K+ efflux. Exposures were carried out in a coaxial stripline for 30 min,
with analysis of the ion content of the supernatant performed afterwards.
These results indicate either a reversal of the normal action of the Na+-K+
ATPase, or an inhibition of the enzyme, which permitted ion leakage across the
membrane to change the Na+/K+ ratio. Hamrick and Zinkl (1975) and Peterson et
al. (1979) have performed similar experiments by exposing RBC's at 2.45 GHz in
the far field (SAR's were 3 to 57 W/kg, and ~ 200 W/kg, respectively).
Neither study found a difference between the K+ efflux from microwave-exposed
RBC's and conventionally heated RBC's with similar histories of temperature
elevation, although Peterson et aK did find a difference between unexposed
5-13

-------
TABLE 5-4. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON SINGLE CELLS
End Point Measured/
Effects
Experimental
Systen
Exposure Conditions
Frequency Duration
(GHz)	(Bin)
Exposure
Facility SAR
(Type)	(W/kg)
Reference
tn
l
Increase In RBC electrophoretlc
nobility 30«1n post-exposure
(SAR > 10 W/kg)
Increase In K+ efflux and Na+ influx
K+ transport no different from
heat-treated controls; no change
in osnotlc fragility
K transport no different from
controls at corresponding
temperatures; no difference in
heooglobln release
Passive transport of Na+, Rb*
Increased at transition temperature
Rapid response in change of firing
rate of pacemaker neurons which
does not correlate with tempera-
ture changes In ninority of trials
No change in growth, CFU,* of ex-
posed cultures
No change in growth, CFU, of vari-
ous strains of exposed cultures
under several growth conditions
Ho change in survival curves
(treasuring CFU) of exposed cultures
RBC'
RBC
RBC
RBC
RBC
Isolated neuron from
ftplysla
E coll
P aeruginosa
E col i
E coli
B subtlUs spores
1 0 (CV) 4, 8.
15, 30
1 0 (CW) 30
2.45 (CW) 60, 120,
180, 240
2 45 (CW) 45
Stripline 5-45
2 45 (CW)
60
1 5, 2 45 3
(CW and PW)
2 45 (CW) 720
2 45 (CW) 240
2 45	1
Stripline
Monopole
far field
Anechofc
chamber
far field
Waveguide
Stripline
Far field
flnecholc
chamber
far field
Microwave
oven
45
3-57
200
100, 190
390
1-100
29-320
0 0075-
75
~ 400
Ismailov (197B)
Ismailov (1971)
Hamrlck and Zinkl (1975)
Peterson et al^ (1979)
Olcerst et a^ (1980)
Wachtel et al (1975),
Seaaan and RIchtel
(1978)
Haonck and Butler
(1973)
Blackman et al. (1975)
Goldblith and Wang
(1967)
(continued)

-------
TABLE 5-4. (continued)
Exposure Conditions
End Point Measured/
Effects
Experimental
System
Frequency
(GHz)
Duration
(min)
Exposure
Facility
(Type)
SAR
(W/kg)
Reference
Growth rate slowed, morphological
changes found
Chinese hamster lung
cells, V79
2 45 (CW)
20
Waveguide
1059
Chen and Lin (1976)
No change in light emission of
photoactive bacterium
P fischeri
2 6-3 0 (CW)
- 22
Waveguide
660 to
5300
Barber (1962)
No effect on colony-forming ability
E coli
2 6-4 0 (CW)
8 h
Waveguide
29
Corelli et al (1977)
Temporary decrease in virulence
(> 6 h) of bacteria for its host
cells; recovery within 24 h at 37 °C
A tumefaciens
10 (CW)
30, 60
230
Cavity
~ 1
Hoore et al (1979)
*CFU = Colony Forming Unit
tRBC = Red Blood Cell.

-------
rabbit cells maintained at 25 °C compared with those maintained at 37 °C.
This difference did not occur with human cells. Ismailov also used temper-
ature controls but did not describe the solution parameters for the cell
suspensions. Under certain conditions of chemical concentration, it is pos-
sible to reverse the Na+-K+ ATPase; however, Ismailov's controls behaved
normally, indicating that chemical concentrations in the controls were not
unusual. The origin of the discrepancy between these studies is not clear.
Hamrick and Zinkl (1975) and Peterson et al_. (1979) measured other end
points as well. The former looked at osmotic fragility of the RBC's and
concluded that there was no difference. The latter paper gives data on hemo-
globin release from RBC's, an indicator of membrane fragility. Again, no
differences were found between irradiated and heat-treated RBC's; however, as
for K+ efflux, the unexposed rabbit cells released less hemoglobin at 25 °C
than at 37 °C.
In a separate paper, Ismailov (1978) reported increases in the electro-
phoretic mobility of human RBC's exposed under conditions identical to those
in his previously discussed study. The electrophoretic mobility was measured
at 10-min intervals after cessation of exposure. The mobility was found to
peak at 30 min post-exposure and to return to base line ~ 60 min after expos-
ure. The peak mobility decreased with shorter exposure durations (30, 15, 8,
and 4 min). Also, the mobility change decreased as a function of dose rate,
disappearing between 5 and 10 W/kg. Although a change in the counter-ion
distribution around the cell and possible conformational changes in the
5-16

-------
membrane proteins were discussed or suggested as possible causes, it remained
unclear why these phenomena peaked 30 min after exposure.
Passive ion transport in rabbit RBC's was examined by Olcerst et al^-
(1980) after exposure to 2.45-GHz (CW) radiation (SAR's at 100, 190, and
390 W/kg). The cells were treated with ouabain to inhibit active transport
of Na+ and K+ by Na+-K+ ATPase, the important enzyme discussed previously in
several papers. Exposures took place in a waveguide system in which the sample
was placed parallel to the E-field in a cylindrical tube. An organic coolant
of low dielectric constant was circulated around the sample in a larger con-
centric cylinder to maintain the temperature of the sample under exposure.
The SAR's were computed from readings of forward and reflected power. The
RBC's were pre-incubated with radioactive Na+ or Rb+ (the latter is a K+ sub-
stitute). Samples were exposed or heat-treated for 1 h, and the suspending
medium was analyzed for radioactivity. Graphs of the logarithm of the efflux
vs. inverse temperature were identical except at three transition tempera-
tures, where the slope of these plots changes sharply (see Figure 5-1). At
these points, the exposed samples exhibited considerably higher efflux than
the heated samples; however, no consistent difference between exposure levels
could be established. These results imply that the cell membrane structures
responsible for passive ion transport are sensitive to microwave exposure at
temperatures where transitions between two states are taking place. Living
cells depend on closely regulated ion concentrations for many processes, and
serious disruption of these balances could be lethal.
5-17

-------
o
c
xf
3
LU
<0
Z
LU
39.3
40 -l
29.8
TEMPERATURE, °C
20.9	12.5
4.6
20-
10-
o
CO
CC
"c 4H
2-
1-
0.4-
0.2-
0.1
]—
3.2
I
3.3
—r~
3.4
~1~
3.5
—r~
3.6
103/T,°K
Figure 5-1.
Arrhenius plot of Na efflux. Unirradiated temperature con-
trols are represented by open symbols; irradiated samples are
represented by darkened symbols. Specific absorption rates
are expressed as watts per kilogram (• = 100, n = 190, ~ = 390).
Control samples had an average standard error of 0.98 percent.
Irradiated samples had an average standard error of 4 percent.
(Olcerst et al. 1980).
5-18

-------
Many studies have been conducted that focus on a major function of a
cellular species as a generalized end point. The reasoning for this approach
is that if microwaves disrupt an important metabolic step, the result will be
a decline in the ability of the cell to perform its major function. As mentioned
earlier, this assumes that no compensating mechanism will operate. A broad
spectrum of such end points has been investigated for cells exposed i_n vitro.
Some of these are discussed elsewhere: phagocytosis and blastic transformation
in lymphocytes are presented in § 5.2, Hematologic and Immunologic Effects;
the induction of the antibiotic colicin, along with point mutations in single
cells, is discussed' in § 5.8, Genetics and Mutagenesis.
Perhaps the ultimate test of a cell's functional ability is growth and
survival. Several workers have concentrated on this end point and have inves-
tigated the frequency range between 1 and 4 GHz. In general, the results have
proved negative. Far-field exposures were conducted (Hamrick and Butler 1973;
Blackman et aH. 1975) on several strains or mutants of E. coli and on Pseudomonas
aeruginosa. Samples were exposed in T-flasks or petri dishes, principally at
2.45-GHz (CW) radiation. Growth was measured by assays of colony-forming
units (CFU). In both experiments the duration of exposure was sufficiently
long (12 and 4 h) that the average cell divided at least once. Blackman
et al. examined several growth conditions such as lag, log and stationary
phases of growth, rich and minimal media, and a "normal" as well as a mutant
amino-acid-requiring strain. In each case, no differences between the exposed
samples and temperature-matched control samples were found. Hamrick's SAR's
O
ranged from 29 to 320 W/kg with power densities of 60 to 600 mW/cm , and
Blackman's from 0.0075 to 75 W/kg with power densities of 0.005 to 50 mW/cm^.
5-19

-------
Corelli et aK (1977) exposed E. coli at the end of a waveguide to microwaves
swept from 2.6 to 4.0 GHz for 8 h at a dose rate of 29 W/kg. No effect of the
microwave exposure was found on colony-forming ability of the bacteria.
Goldblith and Wang (1967) exposed E. coli and Bacillus subti1is spores in a
microwave oven at 2.45 GHz for periods to 1 min (SAR estimated at 400 W/kg).
In this case, microwave irradiation and conventional heating were found to have
identical effects on survival.
Chen and Lin (1978) exposed Chinese hamster lung cells, V79, in a waveguide
fitted with a micropipette that contained the cell suspension. Temperature
was regulated by circulating cooling water through the waveguide around the
micropipette. Samples were exposed to 2450-MHz (CW) radiation for 20 min, and
the cells were allowed to grow in cultures for 12 days after exposure. The
exposed cells were observed to divide at a slower rate and to exhibit a fibro-
blastic type of growth, in contrast with the controls. These cells were
2
exposed at 400 mW/cm (SAR = 1059 W/kg) under conditions similar to those in
which Harrison et al^. (1980) found temperature elevations in the sample that
were as much as 0.3 °C higher than the cells in the cooling bath. Chen and
Lin (1978) state that temperature-treated controls at 38 °C (1 °C higher than
the coolant temperature during microwave exposure) did not display the changes
observed in the microwave-exposed cells. However, it is not entirely clear
whether the changes in the microwave-treated cells were caused by elevated
temperatures within the micropipette.
The emission of light by a photoactive bacterium, Photobacterium fischeri,
has been used as the end point of one study (Barber 1962), where the bacterial
5-20

-------
suspension was circulated through a waveguide. The bacteria were undergoing
exposure in the waveguide for approximately half the time during a typical
43-min experiment. Bacteria were exposed at several frequencies between 2.6
and 3.0 GHz; the assay was performed 24 h later. In spite of extremely high
dose rates, 660 to 5300 W/kg, there were no differences between microwave-
irradiated and conventionally heated samples that received parallel treatment.
In the single experiment conducted at 10-GHz frequency, a transient
effect was found in the virulence of Agrobacteriurn tumefaciens towards its
normal hosts, potato and turnip disks (Moore et aK 1979). This bacterium
produces a plasmid, which it injects into the host cells and which is
responsible for turning these cells into uncontrolled tumor cells. In this
experiment, a suspension of A. tumefaciens was exposed in a petri dish for 30,
60, and 230 min. The longer exposures produced a decrease of virulence near
60 percent with no essential change in the number of viable cells. The effect
was unchanged 6 h post-exposure, but virulence returned to normal 23 h post-
exposure when the bacteria were maintained at 27 °C. An additional experiment,
in which treated and untreated cells were added to the host simultaneously,
indicated that the exposed cells were able to compete effectively with untreated
cells for binding sites on the host cells. The treated bacteria were always
maintained at or below 27 °C during irradiation. From the temperature data in
the paper, a dose rate of ~ 1 W/kg and power density of 0.58 mW/cm can be
estimated. A possible explanation for this effect, which was not offered by
the original authors, is that the plasmid DNA of the A. tumefaciens was
incorporated into the major DNA of the bacteria during microwave exposure,
5-21

-------
preventing injection into the host. Normal growth may have subsequently
allowed the plasmid to return to its original state, restoring activity.
In two papers, Wachtel and coworkers (Seaman and Wachtel 1978, Wachtel et
al. 1975) have examined the effects of microwave irradiation on the firing
rate of isolated neurons from the marine gastropod Aplysia. Neurons were
exposed in a stripline at 1.5- and 2.45-GHz (CW and PW) fields (0.5- to 10-ps
duration, 1000 to 15,000 pulses/s). The firing rate of pacemaker neurons and
the burst rate of bursting cells were measured during microwave exposure at
dose rates between 1 and 100 W/kg. In the majority of cases, the firing rate
of pacemaker cells increased with an increase in temperature, and decreased
with a decrease in temperature. In a minority of cases, 13 percent, for the
pacemaker cells, the microwave irradiation reversed the normal change in
firing rate; i.e., the rate decreased or stopped with a microwave-induced
increase in temperature. The authors were able to detect slow and rapid
components. The slow component, occurring in 30 to 60 s, was correlated with
the slow rise of temperature associated with exposure. The rapid component,
occurring within 1 s, appeared to correlate with the presence of the microwave
field. The rapid component was always found to be a decrease in firing rate
in the presence of the field and was never produced by convective heating.
Similar but more variable effects were found for the bursting cells. The
threshold for the slow component was ~ 7 W/kg, but in one case the rapid
component was found at an SAR as low as 1 W/kg. In all cases where effects
were found, the firing rates returned to normal when the radiation was terminated
and when the temperature was returned to normal. The authors hypothesized
that at a dose rate of ~ 1 W/kg, conversion of 0.1 percent of the microwave
5-22

-------
energy into a polarizing current density across the cell membrane would be
sufficient to affect the firing rate of pacemaker neurons.
5.1.4 Unresolved Questions
Several questions remain unresolved in the area of cellular and
subcellular effects of microwave radiation. Nearly all of the relevant
research is concentrated in a narrow frequency band, between 1.0 and 4.0 GHz.
No acceptable studies have been reported for a large portion of the frequency
spectrum of concern, 0.5 MHz to 100 GHz.
A related question is that of the role of structured water in the cell.
This is presently an active area of research. Questions such as how much
water in a cell is structured to a greater degree than "bulk" water, and
whether this structure plays an important role in cell metabolism, are as yet
unanswered. Three monographs (Alfsen and Berteaud 1976; Drost-Hansen and Clegg
1979; Grant et al^. 1978) summarize knowledge in this area. Experiments attempting
to establish the presence and extent of structured water within biological
systems are described, and the dielectric data indicating a possible frequency
range for the structured-water resonance are presented. According to the
limited information now available, RF radiation is most effective in modifying
the state of structured water at frequencies below 1 GHz. As yet no experiments
have been conducted that define whether absorption of RF energy by structured
water leads to a measurable change in a biological system.
5-23

-------
Three of the reports discussed in § 5.1.3, Effects on Single Cells, also
raise questions. The explanation offered by Ismailov (1978) for the change in
electrophoretic mobility of exposed RBC's is speculative and does not account
for the peaking of the phenomenon 30 min post-exposure. These results must be
considered an effect of unknown origin until more information is available.
The results of experiments by Moore et aK (1979), in which the virulence of
A. tumefaciens was decreased for more than 6 h after exposure to microwave
radiation, suggest reversible functional changes in the organism. In this
case, the implications for cellular function after microwave exposure would be
broad. However, no other worker has noted a similar effect with other single-
cell organisms, and this experiment has not been independently confirmed.
Therefore, its significance is unknown at this time.
The results of Wachtel and coworkers (Seaman and V/achtel 1978; Wachtel et
al. 1975) are potentially highly significant because they indicate the possibil
of a direct interaction between the microwave field and the functioning of the
pacemaker neuron (i.e., the rapid effect). Three other workers have obtained
supportive results. The documentation in these reports is not sufficiently
complete for inclusion in the preceding sections, but they bear mentioning
here.
Yamaura and Chichibu (1967) found results strikingly similar to those of
Wachtel in ganglia of crayfish and prawn exposed at 11 GHz. The regular
firing rate of the ganglia decreased rapidly during microwave exposure,
rebounded to a higher than normal rate when the radiation was removed, and
then returned to normal. Temperature controls showed only an increased firing
5-24

-------
rate as the temperature was increased. The authors stated that the SAR was
~ 100 W/kg but did not describe the method of measurement.
Arber (1976) found a hyperpolarization of the resting potential of giant
neurons of the mollusk Helix pomatia during exposure to 2.45-GHz (CW) radiation.
In this experiment, the ganglion was isolated and mounted in a stripline. The
cell potential was measured, as in Wachtel's experiment, by inserting micro-
electrodes into the neuron. Exposure to microwaves (SAR near 15 W/kg) for 1 h
produced a 5- to 10-percent increase in resting potential, followed by a
stabilization or slight additional increase over 1 h post-exposure. This
result is presumably caused by a change in the Na+-K+ balance in the cell.
When Arber treated the cells with ouabain (post-exposure), which inhibits
Na+-K+ ATPase, he found that part of the hyperpolarization could be accounted
for by the action of this enzyme under the influence of microwaves. The
remainder was attributed to changes in passive ion transport.
Pickard et al_. (1980) have recently presented results in which single
cells from Chara braunii and Nitella flexilis, plants from the family Characeae,
exhibited a large step increase in voltage when exposed to 0.1- to 5-MHz (PW)
radiation (pulses were of 250-ms duration, pulse interval was 6.3 s). These
authors also found a fast and slow component where the slow component correlated
with a temperature rise in the sample. The fast component was frequency
dependent and disappeared abruptly at ~ 10 MHz. The authors suggest that the
fast component is produced by rectification of the oscillating electric field
by the cell membrane. The fast component disappeared into noise at ~ 667 V/m
and may have been an effect of intense fields.
5-25

-------
The results of these four studies indicate a possibility of a direct
microwave interaction with the electric potential across the cellular membrane
of all living cells. This kind of interaction would have broad significance
in the functioning of all cells and, in particular, cells of the nervous
system.
Additional studies have been conducted recently that have not yet been
sufficiently documented to include in the previous sections. However,
because of their potential importance, they are discussed here. The first two
reports concern microwave effects on phospholipid bilayers, whereas the third
is concerned with changes in growth of the yeast Saccharomyces cerevisiae
exposed at frequencies between 41 and 42 GHz.
Tyazhelov et aJL (1979a) have exposed a phospholipid membrane formed
between two chambers containing solutions of NaCl or KC1. An antibiotic that
forms pores through the membrane was added to facilitate passage of Na+ or K+.
Conductance was measured across the membrane while 4-s pulses of 900-MHz
radiation were delivered at between 125 and 280 V/m (field strength in the
aqueous medium). The results showed a change in conductance under exposure
that is consistent with temperature rises of 12 °C, but the temperature of
the NaCl or KC1 solutions did not vary by > 0.5 °C. However, information
concerning the exposure system is insufficient to judge whether serious inhomo-
geneities in energy deposition, and thus similar inhomogeneities in the
temperature distribution, were likely. It is not possible to estimate an SAR
from the information presented.
5-26

-------
Sheridan et aK (1979) have presented at meetings, but so far have not
published, results of Raman spectroscopy on single- and multi-lamellar
phospholipid vesicles exposed to 2.45-GHz (CW) radiation. No change was found
in the Raman bands of single layered vesicles. In contrast, the data for
multi-lamellar vesicles indicate that the hydrocarbon tails of the
phospholipids were undergoing a temperature-dependent phase transition at a
point where the bulk temperature was too low for the transition to have begun.
The change was reported to be equivalent to a temperature difference of ~ 2 °C
at an exposure of 25 niW/cm . The bulk temperature of the sample in these ex-
periments was measured by a unique method. Small ruby crystals were suspended
in the sample, and the shifts in the Raman bands of the crystal were measured
and compared with calibration curves. This appears to be an accurate method
for temperature determination. The origin of this effect is unknown, but if
a similar effect were found in naturally occurring membranes, it could have
an impact on the functioning of the biological membrane.
Keilmann and coworkers (Grundler et aT. 1977; Keilmann 1978; Grundler and
Keilmann 1980) have demonstrated that S. cerevisiae exhibits an enhanced or
inhibited growth rate when exposed at certain closely spaced frequencies
between 41.60 and 41.80 GHz. For instance, they found a 10- to 15-percent
increase in growth rate at 41.64 and 41.68 GHz, and a 20-percent decrease at
41.66 GHz. The experiments were conducted using a unique waveguide termination
that was dipped into a suspension of yeast cells. In a typical experiment,
24 W were dissipated in the yeast-cell suspension, and the authors estimated a
maximum exposure intensity of about 10 mW/cm . However, because of the unusual
5-27

-------
nature of the waveguide termination and the high attenuation of high-frequency
radiation by aqueous samples, SAR values are not determinate. Sample temperature
was monitored and was within 0.5 °C of the desired 32 °C. Equivalent temperature
controls were performed, and the authors believe that changes of the observed
magnitude could not be purely temperature effects. This contention appears
reasonable, since the growth rate both increased and decreased at the same
levels of incident energy but at different exposure frequencies. The decrease
is difficult to explain, based on the author's reports of no observable
temperature rise. If, however, a localized temperature of > 37 °C (> 5 °C
above the controlled temperature) were attained close to the waveguide termi-
nation, then the decrease could be explained solely on the basis of temperature.
There are scanty reports from the Soviet literature that present results
similar to those of Keilmann and coworkers.
In summary, the available literature on cellular and subcellular effects
of microwaves does not yet definitely establish whether effects unrelated to
elevation of temperature exist at dose rates on the order of 1 W/kg. Several
investigators have reported effects, but a majority have not found effects
unrelated to temperature variations. In some cases, the results conflict. In
other cases, the effects found are equivocal. Effects of elevated temperature
may not be clearly eliminated, or, as in the case of neuronal firing rate, the
change in the rate may occur only in a minority of cases (Seaman and Wachtel
1978). The question of heat-induced temperature rises must always be carefully
considered when dealing with the biological effects of exposure to microwave
radiation. In cellular systems, a principal effect is to increase the rates
of all biochemical reactions. This includes the rate of denaturation of
5-28

-------
proteins and DNA (Lehninger 1975) and the rate at which bases are removed
(Lindahl and Nyberg 1974) and mutations occur (Bingham et al_. 1976) in DNA.
When temperature is raised to a certain level, often ~ 43 °C, cell functions
become so disrupted that the cell's capacity to repair the damage is exceeded
and the cell dies. The effect of a rapid heating rate is less clear. Some
disruption of cell function may be expected from a rapid temperature rise,
even if the critical temperature, e.g., 43 °C, has not been reached. Whether
this disruption is fully reversible has not been well documented. RF-
radiation exposure can produce such high heating rates, especially from high
peak, low average power pulses. Effect of nonuniform heat deposition is
probably the most difficult aspect to account for when evaluating the effects
of exposure to RF radiation.
The effects presented in this section are sufficiently well established to
warrant continued concern and effort. The effects of microwave radiation on
the electrical properties of the cell are potentially the most significant.
Four separate experiments have demonstrated these effects. All cells use
the electrical potential across the cell membrane in their life functions, and
perhaps the most important cells are those of the nervous system. Microwave
effects have been found in the functioning of the nervous system and in behavior,
as will be discussed in following sections.
At this time, nearly all of the effects documented for cellular and sub-
cellular systems are observed in intact cells. This could imply that a living
cell is required for the necessary interaction with microwave radiation to
occur. Or, it may be that the right questions are not being asked about the
5-29

-------
subcellular level, perhaps because the level of understanding and instru-
mentation capabilities are as yet too limited. Although the primary mecha-
nism(s) of interaction other than heating the water medium have not yet been
defined, some useful directions are indicated by the results reviewed here.
There also appear to be valuable correlations between some of the effects at
the cellular level with those at higher levels of organization.
5-30

-------
5.2 HEMATOLOGIC AND IMMUNOLOGIC EFFECTS
Ralph J. Smialowicz
Over the years, a considerable number of reports has appeared in the
literature dealing with the effects of RF radiation on the hematologic and
immunologic systems of animals. For the most part, the responsible investi-
gators have been motivated by a concern for the possible adverse health effects
of exposure to RF radiation. Studies in which animals have been exposed at
different frequencies and intensities have shown inconsistent changes in elements
of both biological systems. In some instances, a thermal burden to the exposed
animal has been credited with the observed changes, whereas in others, a
"nonthermal" (i.e., lack of measurable elevations of temperature) or direct
(i.e., athermal or field-specific effects) interaction of RF radiation with the
blood and blood-forming systems has been suggested as the causative mechanism
for the observed effects. In any case, the final interpretation of RF-induced
changes must consider many variables that affect the interaction of RF radiation
with the biological entity. As has been noted previously, variables such as
body shape and mass, radiation frequency, duration of exposure, field intensity,
specific absorption rate, energy distribution, orientation of the body in the
field, ambient environmental conditions, area of the body exposed, and field
modulation may all influence the final results. Variability of response among
species and strains as well as between sexes must also be considered.
This section critically reviews the reported effects of RF radiation on
the hematologic and immunologic systems of laboratory animals. From this
review the following general conclusions can be drawn:
5-31

-------
•	Partial or whole-body exposure of animals to RF radiation may lead
to a variety of changes in the hematologic and immunologic systems.
One of these changes most consistently found is an increase in
lymphocyte formation and activity.
•	Transient changes in peripheral blood composition, probably caused
by redistribution of blood cells and hemoconcentration, were
reported; effects on the immune system have often turned out to be
transient.
•	Several reports describe a causal relationship between increased
core temperature and hematologic and immunologic changes.
•	Many of the reported effects of RF radiation on the hematologic
and immune systems are similar to those (a) resulting from a
stress response involving the hypothalamic-hypophyseal-adrenal
axis, or (b) following administration of glucocorticoids.
•	That an increase in rectal temperature is not observed after
exposure to RF radiation does not preclude a thermal interaction
that the animal is able to compensate for and control.
•	Localized heating or "hot spots" in organs critical to the
hematopoietic and immune systems may occur following the production
of thermal gradients that are unique to RF-energy absorption by
biological specimens.
•	There is a lack of convincing evidence for a direct interaction of
RF radiation with hematologic and immunologic systems in the absence
of some form of thermal involvement.
•	There is presently no convincing evidence from animal studies for
adverse alterations in the hematopoietic or immune systems at RF-
radiation intensities comparable to average environmental levels.
For convenience, this section is divided into two general topics:
hematologic effects and immunologic effects. These two topics are further
subdivided into reviews of studies in which cellular components of these
systems have been exposed jjn vitro and of studies dealing with j_n vivo ex-
posures to RF radiation.
5-32

-------
5.2.1 Hematology
Hematology is the study of the anatomy, physiology, and pathology of the
blood and blood-forming tissues. The hematopoietic system is comprised of a
variety of cells and cell products. In fetal life, the production of blood
cells occurs in the liver, spleen, and bone marrow. After birth this function
is limited largely to the bone marrow, which produces red cells (erythrocytes),
white cells (neutrophilic, eosinophilic, and basophilic granulocytes; lympho-
cytes; and monocytes), and platelets. Each of these cell types performs
specific functions that are essential to life. For example, mature erythrocytes
transport Og and CO2 to and from tissues, granulocytes and monocytes phagocytize
invading microorganisms, and lymphocytes are involved in immune responses.
These functional cells are all descendants of progenitors (stem cells) that
reside within the bone marrow. Blood-cell formation consists of two essential
processes, proliferation and differentiation; progenitor cells proliferate in
the bone marrow and then differentiate into red and white cells. As the
process of cell differentiation progresses, the capacity for cellular prolif-
eration decreases. Impairment of either of these processes may lead to
dysfunctions in the hematologic system that may be life threatening.
5.2.1.1 In Vivo Studies—
A paucity of information on the health effects of RF-radiation exposure
from clinical and epidemiological studies has led to studies of effects on
the hematologic systems of laboratory animals (Table 5-5). Many of the early
investigations on the blood-forming system of laboratory animals employed
2
power densities of 10 mW/cm and higher. For example, Deichmann et al_. (1964)
5-33

-------
TABLE 5-5. SUMMARY OF STUDIES CONCERNING HEMATOLOGIC EFFECTS OF RF-RADIATION EXPOSURE*
Exposure Conditions
Effects

Species
Frequency
Intensity
Duration
SAR
References


(MHz)
(mW/cn2)
(days x o1n)
(W/kg)

Increased.
WBC, lymphs, PKN,
Rat
24,000 (PW)
10
1 x 180
f
Deichaann et al (1964)
RBC, Hct,
and Hgb


20
1 x 420
3 0T

No change

Dog
24,000 (PW)
24
400 x 400-900
i'
Deichmann et a^ (1963)
No change

Rat
3,000 (PW)
10
216 x 60
A
Kitsovskaya (1964)
Decreased
RBC, WBC and lymphs


40
20 x 15
81

Increased
Ptt<


100
6x5
20

Decreased
lymphs and eosln
Dog
2,800 (PW)
100
1 x 360
4;
Michaelson et al^ (1964)
Decreased
WBC, PHN, and eosin

2,800 (PW)
165
1 x 120
61

Decreased-
WBC, lymphs, and eosin

1,280 (PW)
100
1 x 360
4 5

Increased.
PHN



+

Decreased:
lymphs

200 (CW)
165
1 x 360
25

Increased
pm






No change

Mouse
800
43
175 x 120
12 9^
Spalding et al^ (1971)
Increased
lymphs and mitotic
Guinea pig
3,000 (CW or
3.5
120 x 180
0.5t
Baranski (1971 and 1972a,b)
index of lymphoid cells
PW)




Increased
RBC, Hct, and Hgb
Rat
2,400 (CW)
10
30 x 120
2*
Djordjevic and Kolak (1973)
No change

Rat
2,400 (CW)
5
90 x 60
1*
Ojordjevic et al (1977)
Increased
eosinophiIs
Rabbit
2,450 (CW)
10
180 x 1380
1 5
McRee et al (1980a)
Increased-
WBC, CFU
House
2,450 (CW)
100
1x5
70t
Rotkovska and Vacek (1975)
Decreased:
s8Fe uptake





(continued)

-------
TABLE 5-5. (continued)
Exposure Conditions
Effects
Species
Frequency
(HHz)
Intensity Duration
(nW/cm2) (days x nin)
SAR
(W/kg)
References
(Ji
l
CO
in
Accelerated recovery following
x-irradiation, increased
erythropoiesis and nyelopoiesis
Accelerated recovery from
x-irradiation
Increased- PMN and RBC
Decreased lymphs
Accelerated recovery room
x-irradiation
Increased lymphs
Decreased- PMN
(Not reproduced consistently)
No change
No change
No change
Decreased Hct, WBC, and lymphs
House
Dog
Chinese
hamster
Rat
(Perinatal
exposure)
Rat
(Perinatal
exposure)
Rat
(Perinatal
exposure)
Quail egg
Rat
(young)
2,450 (CV)
2,800 (PW)
2,450 (CW)
425 (CW)
2,450 (CW)
100 (CW)
2,450 (CW)
2,736 (PW)
100
100
60
10
46
1x5
1 x 3600
1 x 30
47 x 240
57 x 240
57 x 240
30	1 x 1440
24 4	35 x 240
70'
28'
3-7
1-5
2-3
14
5-25t
Rotkovska and Vacek (1977)
Hichaelson et al. (1963)
Lappenbusch et a^ (1973)
Soialowicz et al (1982)
Saialowicz et al. (1979a)
Soialowicz et a^ (1981a)
Hamrlck and McRee (1975)
Pazderova-Vejlupkova and
Josifko (1979)
(continued)

-------
TABLE 5-5. (continued)
Exposure Conditions

Effects
Species
Frequency
(MHz)
Intensity
(raW/cB2)
Ouration
(days x mm)
SAR
(W/kg)
References

Ho change
House
2,450 (CW)
30
22 x 30
22
Smtalowicz et a^ (1979b)
in
i
Decreased lymphs
Increased PMN
House
26 (CW)
8610

13*
Liburdy (1977)
co







CTl
Decrease in CFU for erythroid
and granulocyte-macrophage series
House
2,450 (CW)
15
9 x 30
10
Huang and Hold (1980)

Reduction in CFU granulocyte-
macrophage precursors exposed
in vitro
House
2,450 (CW)
60-1000
1 x 15
120-2000
Lin et a! (1979b)
*WBC - white blood cell, PMN = polymorphonuclear leukocytes, RBC = red blood cell, Hct = hematocrit, Hgb = hemoglobin, and CFU = colony-forming unit
*SAR estimated

-------
reported significant leukocytosis, lymphocytosis, and neutrophilia in rats
following 7 h of exposure to 24,000-MHz (PW) microwaves at an average power
O
density of 20 mW/cm (SAR estimated at 3 W/kg). One week following exposure,
2
peripheral blood values returned to normal. Rats exposed for 3 h at 10 mW/cm
displayed the same changes and returned to normal after 2 days (SAR estimated
at 1.5 W/kg). Increases in circulating erythrocytes, hemoglobin concentration,
and hematocrit were observed in two of three strains of rats (Osborne-Mendel
2
and CFN) exposed to 24,000-MHz fields at 10 or 20 mW/cm . However, in Fischer
rats exposed under the same conditions, there was a reduction in the number of
circulating erythrocytes and a reduction in hematocrit and hemoglobin
concentration. In another experiment, Deichmann et cH. (1963) exposed two
2
dogs to 24,000-MHz (PW) fields at an average power density of 24 mW/cm (SAR
estimated at 1 W/kg). One dog was exposed for 20 months, 6.7 h/day, 5 days/
week; whereas the second dog was exposed for 20 months, 16.5 h/day, 4 days/
week. No significant changes were observed in blood volume, hematocrit,
hemoglobin, erythrocytes, total and differential leukocytes, blood cholesterol,
or protein-bound iodine. The only symptom attributed to the exposure was a
slight loss of body mass.
Kitsovskaya (1964) exposed rats to 3000-MHz (PW) radiation at 10, 40, or
2
100 mW/cm for various periods of time (SAR estimated at 2, 8, and 20 W/kg,
2
respectively). No changes were found in rats exposed at 10 mW/cm ; at 40 and
2
100 mW/cm , however, the number of peripheral blood erythrocytes, leukocytes,
and lymphocytes decreased, while granulocytes increased. These blood changes
did not return to normal until several months after cessation of exposure.
5-37

-------
The apparent discrepancy between the results of Deichmann et a^. (1964)
and Kitsovskaya (1964) may be partially explained by the work of Michael son et
al^ (1964). These investigators reported that the hematopoietic effects of
2800- and 1280-MHz (PW) fields depend on the frequency, intensity, and
duration of exposure. For example, dogs exposed to 2800-MHz fields showed a
marked decrease in circulating lymphocytes and eosinophils after 6 h at 100
mW/cm (SAR estimated at 4 W/kg). This exposure resulted in a 1 °C mean
increase of rectal temperature. Neutrophils remained slightly increased at
24 h, while eosinophils and lymphocyte values returned to normal levels.
After a 2-h exposure at 165 mW/cm^ to 2800-MHz fields (SAR estimated at 6 W/
kg), there was a slight leukopenia, neutropenia, and definite hemoconcentration.
These changes were accompanied by a rectal temperature rise of 1.7 °C.
Eosinopenia was still evident 24 h after this exposure. General leukocytic
changes were more apparent following exposure of dogs to 1280-MHz (PW) fields
or to 200-MHz (CW) radiation. After exposure of dogs at 1280 MHz for 6 h at
100 mW/cm (SAR estimated at 4.5 W/kg), a leukocytosis and neutrophilia were
observed. After 24 h the neutrophil level was still increased above pre-
exposure levels. Lymphocyte and eosinophil values were slightly depressed
following exposure, but at 24 h they were slightly higher than initial values.
A 6-h exposure to 200-MHz (CW) fields at 165 mW/cm^ (SAR estimated at 25 W/kg)
caused a marked increase in neutrophils and a slight decrease in lymphocytes.
After 24 h this trend was more evident. Michael son et aH. (1964) suggested
that the results indicated a stress response of the exposed animals in the
hypothalamic and/or adrenal axis that was brought about by a thermal stimu-
lation from RF-radiation exposure.
5-38

-------
Spalding et al_. (1971) exposed mice to 800-MHz fields at an average power
2
density of 43 mW/cm (SAR estimated at 10.7 W/kg) for 2 h/day, 5 days/week,
for a total of 35 weeks. These investigators found no changes in blood
erythrocytes, leukocytes, hematocrit, or hemoglobin concentrations. It is
interesting that these investigators did not detect changes in the peripheral
blood picture of exposed mice, despite the thermal burden that was being
placed on these animals. Four mice died from "thermal effects" following the
33rd and 34th RF-radiation exposures.
2
Effects produced at levels at or below 10 mW/cm (SAR estimated at 0.5 to
2.0 W/kg) have also been reported. For example, Baranski (1971, 1972a,b) ex-
posed guinea pigs and rabbits to 3000-MHz (CW and PW) microwaves at an average
power density of 3.5 mW/cm (SAR estimated at 0.5 W/kg) for 3 months, 3 h
daily. At this power level, the body temperature of the animals was not
elevated. Observations were made of increases in absolute lymphocyte counts
in peripheral blood, abnormalities in nuclear structure, and mitosis in the
erythroblastic cell series in the bone marrow and in lymphoid cells in lymph
nodes and spleen. No changes were observed in the granulocytic series in
peripheral blood. Shifts in peripheral blood cells were found to correlate
with changes in the cellularity of the spleen and lymph nodes. An increase in
the mitotic index and in the percentage of cells incorporating H-thymidine
was observed in the spleen and lymph nodes of exposed animals.
Djordjevic and Kolak (1973) exposed rats to 2400-MHz (CW) fields at
2
10 mW/cm (SAR estimated at 2 W/kg) 2 h/day for 10 to 30 days. Body temperature
in rats exposed under these conditions increased by 1 °C within the first
5-39

-------
30 min of exposure and remained at this level throughout the exposure period.
Hematocrit, hemoglobin concentration, and circulating erythrocytes increased
during the 30-day exposure. Fluctuations in the various leukocyte populations
were also observed. The authors suggested these changes were caused by the
thermal effect of microwaves. In a more recent study, Djordjevic et al^.
(1977) found no significant difference in any of several hematologic end
2
points for rats exposed to 2400-MHz (CW) microwaves at 5 mW/cm (SAR estimated
at 1 W/kg) for 1 h/day during a 90-day period.
Recently, McRee et aK (1980a) reported significant decreases in eosinophils
and a lowering of albumin and calcium in blood from rabbits immediately following
chronic exposure to 2450-MHz fields. In this study, rabbits were exposed 23 h
daily for 180 consecutive days to 2450-MHz (CW) radiation at a power density
2
of 10 mW/cm (SAR =1.5 W/kg). No change in hematologic parameters was observed
30 days after termination of exposure (i.e., depression in eosinophils seen
immediately following exposure had normalized); however, a significant decrease
in albumin/total globulin ratio was observed in the blood of exposed rabbits
at this time. The authors contend that, because only 3 of the 41 blood-
chemistry parameters measured immediately after exposure were significantly
different (P < 0.05), and because this observation is close to that expected by
chance, further validation of these changes is warranted.
Rotkovska and Vacek (1975) reported changes in hematopoietic cell popu-
lations of mice following a single 5-min exposure to 2450-MHz (CW) radiation
2
at an intensity of 100 mW/cm (SAR estimated at 70 W/kg). The response of
microwave-exposed mice was compared with that of mice placed in a warm-air
5-40

-------
chamber at an ambient temperature of 43 °C for 5 min. Both treatments caused
a rise in rectal temperature > 2 °C. A leukocytosis occurred in mice under
both conditions; however, the time course for the leukocytosis and the response
of hematopoietic stem cells differed between the two treatments. Following
RF-radiation exposure, a decrease in the total cell volume of the bone marrow
and spleen was observed, and the number of hematopoietic stem cells in bone
marrow and spleen, as measured by the colony-forming unit (CFU) assay, in-
59
creased. Incorporation of Fe in the spleen decreased 24 h after RF-radiation
exposure. On the other hand, the exposure to heat caused a decrease in the
59
CFU's in bone marrow and spleen and an increase in the percentage of Fe
incorporation. Rotkovska and Vacek concluded that the different effects of RF
radiation and externally applied heat on the hematopoietic stem cells indicate
that biological effects caused by high intensities of RF radiation may not
necessarily be related only to increases in internal temperature. They indicated
that their results suggest a possible "direct" effect. This study is significant
because it demonstrates a marked difference in the kinetic response of the
hematopoietic system to two forms of heat stress. Consequently, these differences
must be considered in the interpretation of RF-radiation-induced changes in
the hematopoietic system.
Subsequently, Rotkovska and Vacek (1977) studied the effect of microwaves
on the recovery of hematopoietic tissue following exposure to x-irradiation.
Mice exposed to x-rays at 300 to 750 rads were then exposed to 2450-MHz (CW)
2
microwaves for 5 min at 100 iriW/cm (SAR estimated at 70 W/kg). The combined
treatment resulted in an accelerated recovery of hematopoietic tissue, a
heightened erythropoiesis and myelopoiesis, and an increased survival rate
5-41

-------
compared with x-irradiated mice. The increase in the number of endogenous
hematopoietic colonies in the spleens of the x-irradiated mice after microwave
exposure supports Rotkovska and Vacek's earlier (1975) observation of an
elevation in the number of stem cells in the spleens of intact mice after
microwave exposure alone. These investigators suggested that RF radiation may
influence the mechanisms that activate the pool of stem cells, either by
improving the repair of sublethal radiation damage or by increasing the proli-
ferative capacity of stem cells that survive x-irradiation. The authors
concluded that this acceleration of the processes of repairing radiation
damage in hematopoietic cells after thermogenic doses of RF radiation depended
on the stage of intracellular repair at the time of RF-radiation exposure. In
earlier work, Michael son et aK (1963) reported that simultaneous exposure to
2
x-rays and microwaves (2800 MHz, PW modulated, 100 mW/cm , SAR estimated at 4
W/kg) caused an accelerated recovery of the hematopoietic function in dogs.
Thomson et ajk (1965) reported that pretreatment of mice with RF radiation
(2800 MHz, PW modulated, 100 mW/cm^, SAR estimated at 70 W/kg) reduced the
mortality after x-irradiation (800 rads). The 30-day lethality was 40 to 55
percent among mice given single or multiple RF treatment prior to x-irradia-
tion, compared with 76 percent lethality in mice not pretreated with RF radi-
2
ation. Exposure of Chinese hamsters to RF radiation (2450 MHz, CW, 60 mW/cm ,
SAR estimated at 28 W/kg, for 30 min) 5 min after x-irradiation (725 to
30
950 rads) significantly increased the x-ray LD^q dose compared with exposure
of animals to x-rays only or with exposure to RF radiation before irradiation
(Lappenbusch et al_. 1973). Lappenbusch et al_. reported that the radio-
protective effect of RF radiation appears to be associated with a delayed
decrease in the number of circulating white blood cells, reduced period of
5-42

-------
decreased cell density, and complete replenishment of white blood cells within
30 days following the dual treatment. Exposure to RF radiation either alone
or combined with x-ray exposure increased the relative number of neutrophils,
reduced the relative number of lymphocytes, and slightly increased the number
of circulating red blood cells. On the other hand, animals exposed first to
RF radiation and then to x-rays demonstrated a more severe leukocyte picture
than hamsters x-irradiated only; in these animals, leukocyte counts decreased
faster, and the animals developed leukopenia.
The effect of exposure to RF radiation on circulating blood cells of de-
veloping rats has been studied by Smialowicz et aK (1979a, 1982). Rats were
exposed pre- and post-natally to 425-MHz (CW) fields at 10 mW/cm , 4 h daily
up to 41 days. Because the animals were growing, SAR's ranged from 3 to
7 W/kg. No consistent changes in blood values were observed in exposed com-
pared to sham-irradiated control rats (Smialowicz et al^. 1982). Rats exposed
under the same regimen but to 2450-MHz (CW) fields at 5 mW/cm (SAR estimated
at 1 to 5 W/kg) also showed no difference in circulating erythrocyte count,
leukocyte and differential counts, or hematocrit and hemoglobin concentration
when compared with sham-irradiated controls (Smialowicz et aj. 1979a). Rats
exposed to 100-MHz fields at 46 mW/cm^ (SAR estimated at 2 to 3 W/kg) pre- and
post-natally as above showed no change in blood parameters compared with con-
trols (Smialowicz et aj. 1981a).
Recently Pazderova-Vejlupkova and Josifko (1979) reported decreases in
the hematocrits, number of leukocytes, and absolute numbers of lymphocytes in
young rats exposed to 2736-MHz (PW, 395 Hz, 2.6-ps pulse width) microwaves
5-43

-------
2
at 24.4 mW/cm for 7 weeks (5 days/week, 4 h/day). The means of body mass of
rats at the beginning and at the end of the 7-week exposure period was 65 and
350 g, respectively (SAR estimated at 5 to 25 W/kg). These changes disappeared
within 10 weeks after termination of exposure. The activity of alkaline
phosphatase in neutrophils increased during the first week of irradiation but
decreased transiently after the irradiation. In a similar experiment by the
same authors (data not given), in which adult rats were exposed for 14 weeks
2
to 3000-MHz (PW, 300 Hz, 2.5-|js pulse width) microwaves at 1 mW/cm (SAR esti-
mated at 0.2 W/kg), no difference was observed in hematologic parameters
between exposed and control rats.
Hamrick and McRee (1975) examined the effect of RF radiation on developing
birds. Quail eggs were exposed for 24 h during the second day of incubation
to 2450-MHz (CW) fields at 30 mW/cm^ (SAR = 14 mW/g). At 24 to 36 h after
hatching, quail chicks were examined for gross deformities, changes in organ
weight, and hematologic changes. No significant effects due to RF exposure
were detected.
In another study, Smialowicz et cH. (1979b) exposed mice to 2450-MHz (CW)
fields at 30 mW/cm^ (SAR = 22 W/kg) for 30 min on 22 consecutive days. These
mice showed no significant difference in circulating-erythrocyte count, leukocyte
and differential counts, or hematocrit and hemoglobin concentration compared
with sham-irradiated controls. In this experiment, mice were maintained in an
environmental chamber where temperature, humidity, and air flow were continuously
controlled. Under the conditions of this study, the RF radiation did not
significantly elevate rectal temperatures of exposed mice. In contrast, when
5-44

-------
mice were exposed to thermogenic levels (2 to 4 °C rise in rectal temperature)
of 26-MHz (CW) radiation, 8610 mW/cm^ (SAR estimated at 13 W/kg), a decrease
in the number of circulating lymphocytes and an increase in circulating
neutrophils were observed immediately after exposure (Liburdy 1977). These
mice were held in a chamber that lacked a continuous turnover of air. Liburdy
(1977) reported that this shift reached its peak 3 h after exposure. The num-
bers of circulating lymphocytes and neutrophils were reported to return to
normal, pre-exposure levels 55 to 96 h after exposure. On the other hand,
mice exposed at high ambient temperatures (79 °C) in a vented, dry-air oven
showed an increased number of circulating lymphocytes and neutrophils for a
12-h period after exposure. It appears, therefore, that the response of
circulating leukocytes to exogenous thermal loading depends on the means by
which the body is heated. These results are similar to those reported by
Rotkovska and Vacek (1977) and indicate that the heating properties of RF
radiation differ from those of conventional modes of tissue heating.
Recently Huang and Mold (1980) reported that bone marrow (cultured in
vitro) from mice exposed to 2450-MHz (CW) fields at 15 mW/cm^ (SAR = 10 W/kg)
for 30 min on 9 consecutive days had significantly fewer (P < 0.05) CFU's of
both the erythroid and granulocyte-macrophage series. No data were presented
on the peripheral blood counts of any of these blood cells that would confirm
or expand the information gathered in the CFU assay.
5.2.1.2 In Vitro Studies—
Lin et ah (1979b) reported a reduction in the number of CFU's (granulocyte
and macrophage precursor cells) formed by bone marrow cells exposed j_n vitro
5-45

-------
to 2450-MHz fields at 60 to 1000 mW/cm2 (SAR's at 120 to 2000 W/kg). This
reduction in colony formation was reported to be dose dependent and occurred
without a significant rise in the temperature of the cell suspension. The
authors indicate that their results point to a direct effect of microwave
radiation on these hematopoietic precursor cells. Although these results are
interesting, the in situ application of fields as intense as those required to
produce the observed effects would certainly cause gross thermal injury to the
tissue.
In summary, levels of RF radiation that cause an increase in body tempera-
ture elicit changes in the hematopoietic system that can for the most part be
ascribed to a thermal stress response. Changes in the blood of animals exposed
to RF radiation at intensities below those that cause an increase in core
temperature suggest a similar stress-response mechanism. The failure to
record an increase in core temperature does not preclude that the animal is
compensating for the added thermal energy by thermoregulatory mechanisms.
Indeed, lack of a temperature change indicates that thermoregulation is operat-
ing. The response elicited by RF-radiation-induced heating, however, appears
to differ from that of conventional heating because of the differing internal
heating patterns of this form of radiation.
5.2.2 Immunology
The immune system is comprised of myriad mechanical, cellular, and humoral
components that act as the body's defense against various pathogenic microor-
ganisms, viruses, and neoplasias. The immune system is divided into the humoral
5-46

-------
element, such as antibodies and complement, and the cellular elements, which
are composed of the lymphoid and phagocytic cells. The cellular elements of
the immune system are also part of the hematologic system. The phagocytic
cells responsible for engulfing and digesting certain microorganisms are the
neutrophils or polymorphonuclear leukocytes (PMN's) and the monocytes or macro-
phages. In the presence of antibodies and complement, neutrophils are aided
in engulfing and digesting invading organisms. The monocyte is also a phago-
cytic cell. Monocytes move into an area in which an infection has begun and
then differentiate into macrophages. Macrophages can be "activated" to kill
certain microorganisms (e.g., intracellular, facultative bacteria such as
Mycobacteria tuberculosis and Listeria monocytogenes, viruses, and fungi)
through the interaction of certain subpopulations of lymphocytes, such as the
T lymphocytes.
The other cellular components of the immune system are the lymphocytes.
These cells are broadly divided into two groups, the B lymphocytes and the T
lymphocytes. Although these cells are similar morphologically, they are
different functionally; B and T lymphocytes can be distinguished by the pres-
ence of unique antigens or receptors on their membrane surface. Both T and B
lymphocytes are believed to originate in the bone, marrow and then to proceed
through various stages of development and differentiation, maturing into
functional cells of the immune system.
The B lymphocyte, or bursa-equivalent lymphocyte, is responsible for
humoral immune responses. The B lymphocytes, after appropriate stimulation by
5-47

-------
antigens, proliferate and undergo morphological changes and develop into
plasma cells that actively synthesize and secrete antibodies.
The T lymphocyte, or thymus lymphocyte, is processed through the thymus
after leaving the bone marrow. Classically cell-mediated or T-lymphocyte
responses include protection against viruses, fungi, and several bacteria. T
lymphocytes are also involved in reactions such as delayed hypersensitivity or
contact hypersensitivity and rejection of tumors and foreign tissues such as
transplants (allografts). Cell-mediated reactions are so named because these
reactions, which operate by specifically sensitized T lymphocytes, can be
transferred by these cells to normal animals. B-lymphocyte-mediated humoral
responses, in contrast, are transferable by serum.
Each element of the immune system—the T and B lymphocytes and
macrophages—plays a cooperative role in defending the host against infection
and disease. A delicate balance exists wherein the immune system is prevented
from reacting to its own tissues to avoid autoimmune reactions. The alteration
or dysfunction of any of these elements may lessen the host's ability to
combat infection or may lead to autoimmune disease. However, because of
adaptability and redundancy in the immune system, the host can survive subtle
perturbations. Consequently, although subtle effects on the immune system may
be generated by physical or chemical agents, all such effects may not lead to
clinically significant immune dysfunctions.
5-48

-------
5.2.2.1 In Vivo Studies--
A summary of in vivo studies concerning immunologic effects of RF-radiation
exposure is presented in Table 5-6.
Effects on adult animals—One of the most consistently found RF-radiation-
induced changes in the hematopoietic system is the increase in lymphocyte
formation and activity following exposure of animals of several species to RF
radiation at various frequencies (Baranski 1971, 1972a,b; Czerski 1975). Con-
sequently, there have been several studies of the effects of RF radiation on
lymphocytes and the immune system. In a study reported by Czerski (1975),
mice were exposed 2 h daily to 2950-MHz (PW modulated) microwaves at 0.5 mW/cm
(SAR estimated at 0.5 W/kg) for 6 to 12 weeks. After 6 weeks, there was a
large increase in the relative number of lymphoblasts in the lymph nodes of
exposed mice. In another series of experiments (Czerski 1975), rabbits were
exposed 2 h/day, 6 days/week for 6 months to 2950-MHz (PW) microwaves at
2
5 mW/cm (SAR estimated at 0.8 W/kg). After culturing for 7 days ifi vitro,
peripheral blood lymphocytes from these animals were found to undergo an
increase in "spontaneous lymphoblastoid transformation." Maximal increases
occurred after 1 to 2 months of exposure; the transformation rate then returned
to base line and rose again 1 month after an irradiation had been terminated.
Miro et aJL (1974) continuously exposed mice to 3105-MHz (PW) microwaves over
a 145-h period at an average power density of 2 mW/cm (SAR estimated at
2 W/kg. No description was given of how animals were fed or watered. An in-
crease in lymphoblastic cells in the spleen and lymphoid areas of exposed mice
was observed. A somewhat similar response was observed (Huang et al_. 1977)
in lymphocytes cultured from Chinese hamsters that were exposed to 2450-MHz
5-49

-------
cn
i
CJ1
o
TABLE 5-6. SUMMARY OF STUDIES CONCERNING IMMUNOLOGIC EFFECTS (IN VIVO) OF RF-RADIATION EXPOSURE
Exposure Conditions
Effects*
Species
Frequency
(mi)
Intensity
(mW/cB*)
Duration
(days x nin)
Increase in lytnphoblast: in	Mouse	2,950 (PW)
lyaph nodes and increased
response to SRBC
Increase in "spontaneous"	Rabbit	2,950 (PW)
lymphoblast transformation
of cultured lymphocytes
Increase in lymphoblasts in	House	3,105 (PW)
spleen and lymphoid tissue
Increased transformation of	Chinese	2,450 (CW)
unstimulated cultured lymphocyte hanster
and decreased altosis In PHA-
stimulated lymphocyte cultures
Transient decrease and Increased Mouse	2,450 (CW)
response of cultured lynphocytes
to PHA, Con A, and LPS
Increased mitosis of PHA-	Rhesus	10-27 (PW)
stimulated lymphocytes	Donkey
Increase in CR+, Fc+, and lg+	Mouse	2,450 (CW)
spleen cells Increased response
to B-cell mitogens Decrease
in primary response to SRBC
0 5
5 or 15
1320
42 x 120
6-8700
5, 15, 30 5 x 15
or 45
1-17 x 30
1 x 30
1 or 3 x 30
SAR
(W/kg)
0 5*
References
24-48 x 120	0 8'
Czerskl (1975)
Czerskl (1975)
Mlro et al (1974)
2 3, 6 9, Huang et al (1977)
13 8, or
20.7
3.6 or 10 Huang and Hold (1980)
0 4-2 0t Prince et aj (1972)
14	Wiktor-Jedrzejczak et al
(1977a,b,c)
(continued)

-------
TABLE 5-6. (continued)
Exposure Conditions
Effects*
Species
Frequency
(mi)
Intensity
(oM/ca2)
Ouratlon
(days x otn)
SAR
(W/kg)
References
Increase In Cr* and fc*. spleen
eel Is
Mouse
2,450 (CW)
—
1 x 15
1 x 30
11 8
5
Sulek et al (1980)
Increase In response of
cultured lyophocytes to T-
and B-cell mitogens
Rat
2,450 (CW)
425 (CW)
5
10
57 x 240
47 x 240
(Perinatal exp)
1-5
3-7
Saialowicz et a^ (1979a,
No change
Mouse
2,450 (CW)
5-35
1-22 x 15 or 30
4-25
Smalowicz et al (1979b)
No change
Rat
100 (CW)
46
57 x 240
2-3
Saialowicz et al. (1981a)
Increase in T and B
lynphocytes in spleen
Decrease in DTH
Mouse
26 (CW)
800
1 x 15 or
10 x 15
5.6*
Liburdy (1979)
Reduction of lymphocyte
traffic froo lung to spleen
House
2,600 (CW)
5 or 25
1 x 60
3.8 or 19
Liburdy (1980)
Decreased response to PWM
Rabbit
2.450 (CW)
10
1B0 x 1380
1 5
McRee et ak (1980a)
No change
Quail
2,450 (CW)
5
12 x 1440
4 03
Hamrick et a^ (1977)
Decrease in tumor developoent
Mouse
2,450 (CW)
(near-field
application)
11-14 day of
gestation or
11-14 & 19-45
x 20
35
Preskorn et al. (1978)
Decreased granulocytic response
Rabbit
3,000 (CW)
3
(continued)
42 to 84 x 360
0 5
Szmigielski et al (1975)

-------
TABLE 5-6. (continued)
Exposure Conditions
Effects*
Species
Frequency
(HHz)
Intensity
(sM/cm2)
Duration
(days x tain)
SAR References
(W/kg)
Timor regression and increase in
antitumor antibodies and anti-BSA
Rabbit
13.56
(near-field
application)
1 x 10-15
(local Shah and Dickson (1978b)
hyperthermia)
Tumor inhibition and Immune
stimulation
Rat
2,450 (CW)
200W
3 or 6 x 45
(local Szmlgielski et aj (1978)
hyperthermia)
Increased tumorlcidal activity
in lymphocytes and macrophages
Mouse
1,356
600-900
1x5
(local Maroor et al (1977)
hyperthermia)
Tumor regression
House
3,000 (CW)
40
1-14 x 120
28* Szmlgielski et a^ (1977)
Increase in lung cancer
colonies and inhibition of
contact sensitivity to
oxazolone
House
2,450 (CW)
50
4, 7, 10 or
14 x 120
36* Roszkowski et a^ (1980)
Decrease 1n response to BSA
Rabbit
1,356
(near-field
applicatlon)
3 x 60
(local Shah and Dickson (1978a)
hyperthermia)
*SRgC = sheep red blood cells, PHA+= phytoheoagglutinin, Con A = concanavalin A, LPS = lipopolysaccharide, CR = complement-receptor positive,
Ig = immunoglobulin positive, Fc = Fc portion of ionunoglobulln, DTH = delayed-type hypersensitivity, PWH = pokeweed mitogen, BSA = bovine
serum albumin
*SAR estimated

-------
2
(CW) microwaves for 15 min on 5 consecutive days at 5 mW/cm (SAR = 2.3 W/kg)
without a detectable rise of rectal temperature. Increased transformation to
lymphoblastoid forms in the absence of mitogens was maximal in cultures from
hamsters exposed at 30 mW/cm (SAR = 13.8 W/kg). Irradiation at this power
density caused a 0.9 °C rise in rectal temperature of exposed hamsters.
Mitosis of lymphocytes cultured in the presence of the mitogen phytohemag-
glutinin (PHA), however, was depressed in cells obtained from hamsters exposed
2
at 5, 15, 30, or 45 mW/cm . These effects were reported to be transient and
reversible; control levels were again observed after 5 to 10 days (Huang et aHL
1977).
More recently, Huang and Mold (1980) reported an oscillating response of
spleen cells to PHA, concanavalin A (Con A), and lipopolysaccharide (LPS) from
mice exposed to 2450-MHz fields at 15 mW/cm^ (SAR = 10 W/kg) for 30 min.
Mitogen responsiveness decreased significantly after 2 days of exposure,
returned to normal after 4 days of exposure, and was significantly increased
for all mitogens after 9 days of irradiation. Responsiveness to mitogens
tended to return to normal or to fall to subnormal levels after 17 days of
2
exposure. In contrast, when mice were exposed at 5 mW/cm for 30 min on
5 consecutive days, a significant increase in the response to LPS was observed,
2
whereas no change was observed in LPS responsiveness to exposure at 15 iriW/cm
for 5 days. When macrophages were removed from spleen-cell suspensions of
mice irradiated at 15 mW/cm for 9 days, the responsiveness to LPS was greater
than the already increased spleen-cell responsiveness without macrophage
removal. However, addition of macrophages from the RF-irradiated mice to
spleen-cell cultures from nonirradiated mice caused a significant decrease in
5-53

-------
responsiveness to LPS. The authors suggest that macrophage activation (macrophages
from irradiated mice displayed increased spreading and increased phagocytosis
of latex particles iji vitro) by microwaves may be responsible for inhibiting
LPS responsiveness.
Prince et jH. (1972) reported the opposite effect in rhesus monkeys.
These investigators found an enhanced mitotic response of peripheral blood
lymphocytes stimulated i_n vitro with PHA from monkeys 3 days following a
2
30-min exposure to 10.5-MHz (PW) radiation at 1320 mW/cm (SAR estimated at
0.4 W/kg). Enhancement of mitosis of cultured lymphocytes from monkeys
similarly exposed to 19.27- and 26.6-MHz were also reported. Also reported
were increases in circulating lymphocytes that ranged from 4 to 47 percent
above pre-exposure levels. At a frequency of 26.6 MHz (SAR estimated at 2
W/kg), the rectal temperature of monkeys following exposure was reported to
increase by 2.5 °C above pre-exposure levels.
The particular susceptibilities of lymphocytes to RF radiation has led to
an examination of the effects of this form of radiation on the immune
response. For example, Czerski (1975) reported that mice exposed for 6 weeks
to 2950-MHz (PW) microwaves at 0.5 mW/cm^ (SAR estimated at 0.5 W/kg) had
i
significantly greater numbers of antibody-producing cells and higher serum
antibody titers following immunization with sheep red blood cells (SRBC's).
Interestingly, mice that were exposed for 12 weeks did not show this increased
responsiveness. More recently, Wiktor-Jedrzejczak et aJL (1977a,b,c) exposed
mice in a rectangular waveguide to 2450-MHz radiation for 30 min at an average
dose rate near 14 W/kg. At 3, 6, 9, and 12 days after a single or multiple
5-54

-------
exposures, mice were tested for (1) the relative frequency of T and B splenic
lymphocytes, (2) the functional capacity of spleen cells to respond to T- and
B-cell-specific mitogens, and (3) the inability to respond to SRBC's (a
T-dependent antigen) or dinitrophenyl-lysine-Ficoll (DNP-lys-Ficoll, a
T-independent antigen). A single 30-min exposure induced a significant
increase in the proportion of complement-receptor-positive (CR+) lymphocytes
in mouse spleens that peaked 6 days after exposure. This effect was further
enhanced by repeated (three) exposures, which also produced a significant
increase in the proportion of immunoglobulin-positive (Ig+) spleen cells
(Wiktor-Jedrzejczak 1977a). A significant increase in the proportion of
Fc-receptor-positive (FcR+) cells in the spleens was also observed 7 days
after a single 30-min exposure (SAR - 13.7 W/kg). However, no change in the
number of Ig+ cells in spleens of these mice was observed (Wiktor-Jedrzejczak
1977c). The type and combination of surface receptors (CR, Ig, Fc) expressed
on splenic B cells represent different maturational stages in B-cell develop-
ment. Wiktor-Jedrzejczak et al. (1977a,b,c) were unable to demonstrate any
change in the total number of theta-antigen-positive (0+) T cells in the
spleens of mice following a single or multiple exposure to 2450-MHz microwaves.
No change was detected in the iji vitro spleen-cell response to stimulation by
the T-cell-specific mitogens PHA and Con A or by pokeweed mitogen (PWM), which
stimulates both T and B cells (Wiktor-Jedrzejczak et _aL 1977a). The response
to the B-cell-specific mitogens lipopolysaccharide (LPS), polyinosinic polycyti-
dylic acid (Poly I C), and purified protein derivative (PPD) of tuberculin,
however, was significantly increased over controls following a single exposure.
These results clearly correlate with the observed changes in the proportion of
5-55

-------
cells bearing different surface markers. These authors (1977a) noted that RF
irradiation did not stimulate lymphoid-cel1 proliferation j>er se but rather
appeared to act as a polyclonal B-cell activator, which led to an early
maturation of noncommitted B cells. They also found a significant decrease in
the primary immune response to SRBC's, a thymus-dependent antigen, in mice
that had been immunized just prior to the first exposure to RF radiation.
They suggested that this decreased response may result from the nonspecific
stimulation of some cells by RF radiation to mature before they are activated
by antigen (SRBC's), thereby increasing the proportion of unresponsive cells.
Recently, Sulek et al^. (1980) corroborated this increase in CR+ and Fc+
spleen cells in mice 6 days following a 30-min exposure to 2450-MHz fields
(average SAR = 12 W/kg). The kinetics for increased frequency of CR+ cells in
the spleens of irradiated mice showed an initial increase 3 days following
exposure that persisted for 5 to 6 days and then returned to normal within
9 to 10 days. The threshold for this effect was determined by varying the
time of exposure with a constant forward power (0.6 W), or by maintaining a
constant time of exposure while varying the forward power (0.1 to 0.78 W). It
was shown that a minimum of a single 15-min exposure (11.8 W/kg) or a single
30-min exposure (5 W/kg) caused significant increases in CR cells 3 or 6 days
post-exposure. The effect of absorption of multiple subthreshold quantities
of microwaves was found to be cumulative only if the exposures occurred within
1 h of one another. The increase in CR+ cells was found at dose rates ranging
from 10 to 18 W/kg for mice within a weight range of 18 to 25 g. The rectal
temperature of exposed mice was found to be elevated no more than 0.6 °C above
that of sham-irradiated mice. No change in 6+ (T lymphocytes) cells was
5-56

-------
observed. In contrast, Huang and Mold (1980) reported a significant increase
in 0+ cells but no change in B cells (Ig+) of mice exposed to 2450-MHz fields
2
at 15 mW/cm (SAR at 10 W/kg) for 30 min on 9 consecutive days.
Smialowicz et al_. (1979b) exposed mice to 2450-MHz (CW) fields under
far-field conditions for 15 or 30 min daily for periods to, 22 consecutive days
at power densities from 5 to 35 mW/cm (SAR's at 4 to 25 W/kg). Splenic
lymphocyte function was assayed by the jjn vitro mitogen-stimulated response as
3
measured by H-thymidine incorporation after culture in the presence of T
(PHA, Con A, PWM) or B (LPS, PWM, PPD) mitogens. The proportions of T (0+)
and B (CR+) splenic cells and the primary immune response of mice to SRBC's
were also studied. No difference in the response to mitogens or to SRBC's or
in the frequency of T or B cells in spleens was observed in RF-irradiated
compared with sham-irradiated mice.
Liburdy (1979) reported that changes in splenic lymphocyte populations
similar to those observed by Witkor-Jedrzejczak et al_. (1977a,b,c) can be
produced by exposure of mice to thermogenic levels of 26-MHz radiation. When
2
mice were exposed to 26 MHz at an intensity (800 mW/cm , SAR at 5.6 W/kg) that
produced a 2 to 3 °C rise in rectal temperature, a relative increase in splenic
T and B lymphocytes was observed. Similar responses (i.e., increase in T and
B cells) were induced following administration of the synthetic glucocorticoid
methyl prednisolone sodium succinate. These results indicate that these RF-
radiation-induced changes might represent some form of stress. It is difficult
to understand how a 2 to 3 °C rise in rectal temperature would occur in mice
irradiated for 15 min (SAR = 5.6 W/kg). A possible explanation is that these
5-57

-------
mice were restrained in perforated acrylic cages and held in an exposure
chamber in which the air was not circulated during this period of irradiation.
In examining further possible mechanism(s) for the shift in lymphocyte
populations in the spleens of mice exposed to RF radiation, Liburdy (1980) ex-
amined the circulation of. lymphocytes in microwave-exposed mice. Mice injected
51
with Cr-labeled syngeneic spleen cells were exposed for 1 h at either 5 or
25 mW/cm2 to 2600-MHz fields (SAR at 3.8 and 19 W/kg, respectively). Controls
included sham-irradiated mice, mice held in a 63 °C warm-air oven for 1 h, and
mice injected with methyl prednisolone sodium succinate. The distribution of
injected cells was determined for the lung, liver, spleen, and bone marrow at
o
1, 6, and 24 h post-exposure. Exposures at 25 mW/cm caused a 2.0 °C increase
in core temperature. This regimen led to a 37-percent reduction in lymphocytes
leaving the lung and migrating to the spleen. Also, a threefold increase in
spleen lymphocytes entering the bone marrow occurred in this group of mice.
A similar pattern of lymphocyte circulation was observed in the steroid-treated
group. No change in lymphocyte traffic was observed in mice of the 5-mW/cm
or warm-air groups. Liburdy (1980) concluded that these results support the
idea that steroid release associated with thermal stress and attempts by the
animal to thermoregulate during exposure to RF radiation are responsible for
effects on the immune system.
Effects on young animals—RF-radiation effects on the development of the
immune response have been studied. Smialowicz et al.. (1979a) exposed rats on
day 6 of gestation through day 41 to 2450-MHz (CW) fields at 5 mW/cm2 (SAR at
1 to 5 W/kg). The young animals absorbed microwaves at a higher rate (5 W/kg)
5-58

-------
than the adults (1 W/kg). In this study the exposed rats had lymphocytes that
responded to a significantly greater extent than those from control animals
following in vitro stimulation by T- or B-cell mitogens. A similar increase
in lymphocyte responsiveness was seen in another study in which rats were
exposed pre- and postnatally to 425-MHz radiation (SAR at 3 to 7 W/kg, with
the neonates absorbing at the latter SAR) for periods to 41 days postpartum
(Smialowicz et ah 1982). However, lymphocytes from rats exposed peri natal ly
to 100-MHz fields at 46 iriW/cm^ (SAR at 2 to 3 W/kg) showed no change in mito-
genic responsiveness (Smialowicz et ah 1981a). The results of the two former
studies indicate that long-term exposure of developing (especially neonatal)
rats to RF radiation at absorption levels higher than those achieved in the
latter study may give rise to increased responsiveness of cultured lymphocytes.
These results are similar to other reported changes in mammalian lymphocyte
responsiveness following RF-radiation exposure (Czerski 1975; Prince et ah
1972; Wiktor-Jedrzejczak et ah 1977a,b,c). Although the benefits are not
known concerning this increased responsiveness to mitogens by lymphocytes from
animals exposed peri natally to RF radiation, a recent report by Preskorn
et al_. (1978) indicates that irradiation at this time during development may be
beneficial. These investigators exposed mice to 2450-MHz fields (SAR = 35 W/kg)
for 20 min either on days 11 through 14 of gestation, or on days 19 through 45
postpartum, or during both periods. On the 16th day postpartum, all mice were
implanted with a lymphoreticular cell sarcoma. Mice irradiated i_n utero only
(colonic temperature increase of 2.2 °C in dams) showed a lower incidence of
tumors (13 percent vs. 46 percent for sham-irradiated mice) 93 days postpartum.
In mice irradiated _i_n utero and postnatal ly, tumors initially developed at a
lower rate compared with controls; however, after 2.5 months, no difference
5-59

-------
was observed in tumor incidence between groups. At the end of 4 months, the
tumor incidence in irradiated mice was slightly greater than controls (46 vs.
40 percent, respectively). An interesting finding, however, was that both
tumor-bearing and tumor-free animals that had been irradiated only j_n utero
lived longer on the average than their respective controls. This result is
somewhat similar to that reported by Prausnitz and Susskind (1962), who found
that mice briefly irradiated hundreds of times by a highly thermogenic level
of RF radiation survived longer than controls.
McRee et al^. (1980a) reported that 30 days after termination of a 6-month
23-h daily irradiation to 2450-MHz fields (SAR = 1.5 W/kg) spleen cells from
rabbits showed a decreased responsiveness to PWM. Decreased responsiveness to
PHA and Con A by these spleen cells was also reported, but responses to PHA
and Con A were not statistically different from those of controls. Although
these results are interesting, they are not conclusive, because only four
exposed and four sham-irradiated rabbits were employed. It should also
be noted that both irradiated and sham-irradiated rabbits were transported
from one laboratory to another between the termination of RF exposure and
spleen-cell assay.
Hamrick et al^. (1977) examined the avian humoral-immune response in
Japanese quail exposed to RF radiation during embryogenesis. Fertile quail
eggs were continuously exposed to 2450-MHz (CW) radiation at 5 mW/cm (SAR =
4.03 W/kg) throughout the first 12 days of development. At 5 weeks of age,
quail were immunized with SRBC's, and the levels of anti-SRBC antibodies were
determined. No difference was observed in the antibody titers of exposed and
5-60

-------
sham-exposed quails. The masses of the bursa of Fabricius (site of B-cell
production in birds) and spleen were not altered significantly by exposure to
RF radiation.
Effects on phagocytosis—RF-radiation-induced effects on phagocytic
leukocytes of animals have been reported by Szmigielski et ah (1975).
Rabbits were exposed to 3000-MHz fields for 6 h daily for 6 to 12 weeks at
2
3 mW/cm (SAR estimated at 0.5 W/kg). After the last exposure to RF radiation,
rabbits were infected with an intravenous injection of virulent Staphylococcus
aureus. At periods before and after infection, functional tests of granulo-
poiesis were performed. The investigators reported a decreased production of
mature granulocytes in infected, RF-radiation-exposed rabbits, which was mani-
fested as a more serious illness in these animals.
Summary--Exposure of laboratory animals to RF radiation can lead to
changes in the functional integrity of lymphocytes. These cells play an
important role in the immune-defense system of man and animals. The
significance of the changes caused by RF radiation is difficult to interpret.
Although some studies indicate that RF radiation causes an increased
responsiveness of lymphocytes (Czerski 1975; Smialowicz et ah 1979a, 1982;
Prince et ah 1972; Wiktor-Jedrzejczak et ah 1977a,b,c) and a potentiation of
the immune response to antigen (Czerski 1975), others indicate a depression in
responsiveness (Huang et ah 1977; Wiktor-Jedrzejczak et a2- 1977a; Liburdy
1979; Szmigielski et ah 1975). In most cases these alterations can be
attributed to a stress response, since qualitatively similar but quantitatively
more pronounced changes are observed at obviously stressful thermogenic levels
5-61

-------
of RF radiation (Huang et aK 1977; Huang and Mold 1980; Prince et ah 1972;
Liburdy 1979) or after administration of glucocorticoids (Liburdy 1979). It
is well known that stress alters physiological systems that regulate immunolog-
ical function. Both immunosuppression and immunoenhancement have been observed
to result from stress (Monjan 1981; Blecha et aL 1982), which is not inconsis-
tent with the reported effects of RF radiation on the immune system. Many
questions remain about the role of stress in the modulation of the immune sys-
tem, not the least of which, in the present discussion, deal with RF-radiation-
induced stress.
Effects caused by RF-induced hyperthermia--A1terations in the immune
system can be produced by RF-radiation-induced hyperthermia. Whole-body
microwave-induced hyperthermia has been reported to serve a therapeutic role
either alone (LeVeen et ah 1976) or in combination with ionizing radiation
(Nelson and Holt 1978). In many cases, the direct destruction of malignant
tissues by RF-radiation-induced heating is the ultimate goal. However, in
some cases, hyperthermia has led to changes in the immune response. For
example, Shah and Dickson (1978b) reported that following local heating of VX2
(carcinoma) tumor-bearing rabbits by a 13.56-MHz field, tumor regression and
host cure were observed in 70 percent of the rabbits. Intratumoral tempera-
tures of 47 to 50 °C were achieved within 30 min. Along with tumor regression,
cell-mediated immunity—as measured by skin reactivity to tumor extract and
dinitrochlorobenzene—markedly increased. A hundredfold increase in serum
levels of antitumor antibody and increased response to the antigen bovine
serum albumin (BSA) were also observed. In contrast, whole-body hyperthermia
led to temporary reduction of tumor growth, followed by a return to an exponential
5-62

-------
increase in tumor volume and rapid death of the rabbit. This course of events
following whole-body hyperthermia was accompanied by abrogation of the enhanced
cellular and humoral immune responsiveness, observed following local RF-induced
heating.
Szmigielski et al^. (1978) reported that local heating (43 °C) of the
Guerin epithelioma in Wistar rats by 2450-MHz (CW) radiation inhibited tumors
and stimulated the immune reaction against the tumor. Other immune reactions
stimulated by this treatment were the antibody response to BSA, high reactivity
of spleen lymphocytes to the mitogen PHA, and increased serum lysozyme levels
as a measure of macrophage activity. Tumor-specific reactions observed were
increased cytotoxicity of spleen cells and peritoneal macrophages to cultured
tumor cells. Similar results were reported by Marmor et ah (1977), who
exposed tumors in mice to focal 1356-MHz radiation. The EMT-6 tumor was found
to be extremely sensitive to RF heating. The cure rate was a function of
temperature and duration of exposure. A 5-min exposure at 44 °C reduced the
tumors by almost 50 percent. To determine the effectiveness of RF-induced
heating on tumor regression, tumor-cell survival was studied by treating EMT-6
tumors i_n situ. Cell inactivation by RF-radiation-induced heating was similar
to that for heating by a hot water bath. The results indicated that direct
cell killing could not account for the observed cures, and these investigators
suggested that hyperthermia (RF or convection-induced) may stimulate a tumor-
directed immune response.
Szmigielski et al_. (1977) exposed mice bearing transplanted sarcoma-180
tumors to 3000-MHz radiation, 2 h daily on the 1st through 14th day after
5-63

-------
2
transplantation, whole-body at 40 mW/cm (SAR estimated at 28 W/kg). This
exposure led to a 3 to 4 °C increase in rectal temperature and resulted in a
reduction of tumor mass by ~ 40 percent, a reduction enhanced when microwave
hyperthermia was combined with Colcemide, Streptolysin S, or both. Colcemide
enhances the inhibiting effect of hyperthermia on proliferation of cells _in
vitro (Szmigielski et al^ 1976), and Streptolysin S is an antineoplastic
agent. Szmigielski et al_. (1978) suggested that immunostimulation is important
in the complex inhibition of tumor growth by increased temperature.
While many have heralded local and systemic hyperthermia as a possible
cancer treatment, either alone or in combination with drugs or ionizing
radiation, there is evidence that hyperthermia may enhance the dissemination
of certain cancers and abrogate the immune response. For example, Roszkowski
2
et aK (1980) reported that exposure of mice to 2450-MHz radiation at 50 mW/cm
for 4, 7, 10, or 14 days, 2 h daily (SAR estimated at 36 W/kg) caused an in-
creased number of lung-cancer colonies and an inhibition of contact sensitivity
to oxazolone (a measure of T-lymphocyte activity) with increased duration of
hyperthermia. Shah and Dickson (1978a) exposed normal rabbits either to
RF-radiation-induced (13.56 MHz) or to watercuff-local hyperthermia of thigh
muscles, which were maintained at 42 °C for 1 h on 3 consecutive days. No
alteration in the response to dinitrochlorobenzene challenge was observed.
However, the humoral immune response to BSA was significantly depressed. This
response was independent of the method and degree of heating. The results
indicate that B lymphocytes might be more susceptible to hyperthermic damage
than are T lymphocytes.
5-64

-------
The above results indicate that if the applied microwaves are of suffi-
cient intensity to cause heating of tissue or of the whole animal, changes in
the immune system will follow. And, with heating to any extent, the hypotha-
lamic- hypophyseal -adrenal axis plays a major role in the responses elicited.
It is well known that endogenous or exogenous adrenal glucocorticoid hormones
affect the immune response. But what of the RF-induced responses reported in
the absence of measurable temperature increases or at rates of energy absorption
well below that of the resting or basal metabolic rate? Observed changes in
the immune response under these conditions are more difficult to explain on a
thermal-stress basis, primarily because of a lack of sensitive techniques to
detect subtle stress responses. The fact that no increase of rectal tempera-
ture occurs after exposure to RF radiation, however, does not preclude that the
animal's capablility to compensate for added thermal energy by thermoregulatory
mechanisms. RF radiation may also cause focal heating (thermal "hot spots")
in organs critical to the immune response.
5.2.2.2 J_n Vitro Studies—
Among the studies in this area (Table 5-7), several have involved attempts
to determine if in vitro exposure of lymphocytes to RF radiation leads to
"direct" changes in the metabolic or functional state of these cells. In an
early study, Stodolnik-Baranska (1967) exposed human lymphocytes in culture to
2
3000-MHz (PW) microwaves at 7 or 14 mW/cm . Some lymphocytes were irradiated
2	2
4 h/day at 7 mW/cm for 3 to 5 days, while those exposed at 14 mW/cm were
irradiated 15 min daily for 3 to 5 days. After 5 days in culture, the micro-
wave-exposed cells were found to have undergone a fivefold increase in
lymphoblastoid transformation compared with controls. Czerski (1975) attempted
5-65

-------
TABLE 5-7. SUMMARY OF STUDIES CONCERNING IMMUNOLOGICAL EFFECTS (IN VITRO) OF RF-RADIATION EXPOSURE*
Exposure Conditions
Effects
Species
Frequency
(HHz)
Intensity Ouration
(iiM/cb2) (days x min)
SAR
(W/kg)
References
tn
l
Increased blastogenesis of
exposed lymphocytes jn vitro
Increased blastogenesis
No change in Mitogen
response to PHA, Con A
or LPS
No change in mitogen
response to PHA
No change in viability
or growth
Decreased macrophage
phagocytosis
Liberation of intracellular
hydrolytic enzymes and
Increased death
Human	3,000 (PW)
lymphocytes
Hunan	10,000
lymphocytes
House spleen 2,450 (CV)
cells
Rat blood 2,450 (CW)
lymphocytes
Human	2,450 (CW)
lyophoblast
cell lines
(Daudi and
hsb2)
House	2,450 (CW)
macrophage
Rabbit	3,000 (CW)
granulocytes
7
14
5-15
10
10-500
50
1 or 5
3-5 x 240	t
3-5 x 15
Observed effect t
only when culture
temperature ap-
proached 38 °C
1 x 60, 120, or 19
240
5, 10 or 20 1 x 240, 1440,
or 2640
1 x 15
1 x 30
1 x 15, 30,
or 60
0 7, 1 4
or 2 8
25-1200
Stodolnlk-Baranska (1967)
Baranskl and Czerski (1976)
Saialowicz (1976)
Hamrick and Fox (1977)
Lin and Peterson (1977)
15 J/min Mayers and Habeshaw (1973)
t	Szmigielski (1975)
*PHA = phytoheoagglutinin, Con A = concanavalin A, and LPS = lipopolysaccharlde
^Unable to calculate SAR

-------
without success to repeat this experiment. But, in a more recent study,
Baranski and Czerski (1976) reported that exposure of human lymphocytes to
10,000-MHz fields at power densities between 5 and 15 mW/cm could induce
lymphoblastoid transformation (SAR not given). At power densities below
2	2
5 mW/cm , this effect was not observed, whereas at power levels above 20 mW/cm ,
cell viability decreased. The induction of blastic transformation depended on
2
termination of irradiation (5 to 15 mW/cm ) at the moment when the temperature
of the medium reached 38 °C. These results indicate that the microwave-induced
blastic transformation might be caused by a thermal mechanism.
Similar increases in the lymphoproliferative response of cells exposed to
temperatures > 37 °C have been reported. Ashman and Nahmias (1978) reported
that human lymphocytes, when cultured at 39 °C with the mitogens PHA or Con A,
3
showed an enhancement and earlier onset of H-thymidine incorporation compared
with cultures incubated at 37 °C. In a similar study, Roberts and Steigbigel
(1977) reported that the iji vitro human lymphocyte response to PHA and the
common antigen streptokinase-streptodornase was enhanced at 38.5 °C relative
to 37 °C. Smith et al^ (1978) reported that the in vitro response of human
lymphocytes to PHA, Con A, PWM, and allogeneic lymphocytes was markedly
enhanced by culture at 40 °C compared with 37 °C. These studies demonstrate
the need to monitor and to control the temperature of cultures exposed to RF
radiation. Without adequate temperature data, it is virtually impossible to
accept _in vitro effects as due to RF radiation itself.
The proliferative response of lymphocytes exposed iji vitro to RF radiation
appears to be related to culture temperature. Smialowicz (1976) exposed
5-67

-------
murine splenic lymphocytes to 2450-MHz (CW) radiation for 1, 3, or 4 h at
10 mW/cm (SAR = 19 W/kg). Immediately after irradiation the temperature of
the exposed cultures did not differ significantly from that of controls, and
cell viability was unchanged. These cells were then cultured for 72 h in the
presence of T- or B-cell mitogens, and the proliferative response was measured
by H-thymidine incorporation. No difference was found in the blastogemc
response of microwave-exposed and sham-exposed spleen cells to any of the
mitogens employed. In a similar experiment, Hamrick and Fox (1977) exposed
rat lymphocytes to 2450-MHz (CW) radiation for 4, 24, or 44 h at 5, 10, or
O
20 mW/cm (SAR's at 0.7, 1.4, or 2.8 W/kg, respectively). The transformation
3
of unstimulated or PHA-stimulated lymphocytes was measured with H-thymidine.
No significant differences were found in the proliferative capacity of lympho-
cytes from exposed and control cultures. The effects of RF radiation on the
growth and viability of cultured human lymphoblasts was studied by Lin and
Peterson (1977). Human lymphoblasts (cell lines Daudi and HSf^) were exposed
to 2450-MHz (CW) radiation in a waveguide for 15 min at incident power densities
of 10 to 500 mW/cm (corresponding SAR's were 25 to 1200 W/kg). No temperature
increase was observed, even at the highest power density in the capillary tube
that held the cell suspension in the waveguide. No change was observed in the
viability or growth of microwave-exposed lymphoblasts compared with controls.
These studies provide further evidence that in the absence of heating, no
change in lymphocyte activity occurs following RF- radiation exposure in
vitro.
In vitro exposure of macrophages to 2450-MHz fields has been reported by
Mayers and Habeshaw (1973) to depress phagocytosis. Monolayer cultures of
5-68

-------
mouse peritoneal macrophages were perfused with suspensions of human erythro-
cytes while being simultaneously exposed to 2450-MHz radiation at 50 mW/cm .
The rate of energy absorption in the sample was 15 J/min. The phagocytic index
of exposed cultures was significantly lower than that of the control after a
30-min exposure. Macrophage phagocytic activity returned to normal levels if
RF irradiation was discontinued. During irradiation, a 2.5 °C temperature
increase was observed; however, the final temperature in the culture vessel in
any given experiment reportedly did not exceed 36.2 °C. The investigators
concluded that the observed depression of phagocytosis in the irradiated
cultures was not thermally induced. The 2.5 °C rise in temperature during
irradiation, according to the authors, would have been expected to enhance
rather than depress phagocytosis, since optimal phagocytosis occurs at
38.5 °C. The mechanism by which the observed effect occurs is not known;
however, heating effects are difficult to dismiss because of an observed 2.5 °C
rise in culture temperature. Although the temperature of the suspension medium
did not exceed 36.2 °C, thermal gradients of much higher temperature would be
expected at the macrophage-glass interface.
An RF-induced effect on granulocyte integrity and viability was reported
by Szmigielski (1975). Rabbit granulocytes were exposed in vitro to 3000-MHz
2
(CW) radiation at 1 or 5 mW/cm (SAR not given) for 15, 30, or 60 min. Cultures
2
exposed at 5 mW/cm for 30 or 60 min showed increased numbers of dead cells as
demonstrated by an increase in nigrosine staining and an enhanced liberation
of lysosomal enzymes. Exposure at 1 mW/cm did not cause increased cell death
but did lead to a partial liberation of hydrolase enzymes. No change was
observed in the temperature of the irradiated cultures. The liberation of
5-69

-------
acid phosphatase and lysozyme from granulocytes was observed in cell suspensions
2
exposed at 1 or 5 mW/cm , both suspensions exhibiting a time- and dose-dependent
response.
In summary, exposure of laboratory animals to RF radiation may lead to
changes in the functional integrity of leukocytes, which play important roles
in the immune-defense system. The significance of the changes caused by RF
radiation is difficult to interpret, since many observed effects are transient
and reversible. Furthermore, some studies indicate that RF radiation causes
immunopotentiation, whereas others indicate immunosuppression. In many cases
the observed alterations in the immune system can be attributed to thermal
stress, because qualitatively similar but quantitatively greater changes are
observed at obviously stressful (highly thermalizing) levels of RF radiation
or following the administration of glucocorticoids. A possible explanation for
the immunomodulating effects of RF radiation arises from the timing of the
measurements of immune responsiveness after subjecting an animal to stress.
For example, corticoid-induced impairment of immune responsiveness is commonly
followed by homeostatic recovery, then subsequent overcompensation, which may
be associated with immunoenhancement.
As for those reports in which measurable elevations of temperature from
RF-radiation-induced heating are not detected, a possible role by RF-radiation-
induced thermogenesis cannot be dismissed. The failure to detect a measurable
increase in tissue or core temperature in RF-irradiated experimental animals
through the use of conventional techniques indicates that the animal was able
to compensate for the added energy. The role that thermoregulating mechanisms
5-70

-------
play in affecting the immune response needs further study. There is presently
no convincing evidence for a direct effect of RF radiation on the immune
system in the absence of a thermal (heating) effect.
5.2.3 Unresolved Questions
Several questions remain unanswered regarding the effects of RF radiation
on the hematologic and immunologic systems. Perhaps the most perplexing
question is what to make of the many Soviet reports on immunology. In most
cases these reports lack sufficient technical detail for adequate critical
assessment of reported results. Alterations in the hematopoietic and
immunologic systems have been reported in animals exposed to RF radiation at
and below 10 mW/cm over periods of weeks to months. Nevertheless, no con-
vincing evidence has been presented to demonstrate a direct effect of RF
radiation in the absence of thermal involvement; well defined and planned,
2
chronic (months-to-years), low-level (< 1 mW/cm ) studies have not been carried
out to investigate Soviet claims of possible induction of autoimmune reactions
following chronic exposure. Investigations of the possible hematologic and
immunologic effects of PW vs. CW irradiation have not been undertaken in re-
sponse to the claims by Eastern European investigators that PW modulation is
more effective than CW irradiation in causing alterations in these systems.
Investigations are lacking that would define possible synergistic effects
of other agents or drugs with RF radiation on the hematopoietic and/or immune
systems. RF radiation (at hyperthermic doses) has been shown to provide a
protective effect against damage by ionizing radiation to the hematologic
5-71

-------
system. Although this is a beneficial effect of RF radiation, it is not known
if the combination of drugs or other physical agents with RF radiation is
detrimental.
Another question has been raised that is based on the recent work of
Szmigielski et ah (1980), who described the increased incidence of cancer
development in mice exposed chronically to 2450-MHz (CW) fields at 5 and
2
15 mW/cm . This is an extremely important piece of work that merits much
additional consideration.
5-72

-------
5.3 REPRODUCTIVE EFFECTS
Ezra Berman
This section on the reproductive effects of RF radiation is organized
into three categories: Teratology (5.3.1), where the treatment is admin-
istered to the pregnant dam and observations are then made on the embryo,
fetus, neonate, or older offspring; Reproductive Efficiency (5.3.2), where the
end point of the experiment occurs in the primary and secondary sex organs of
the parent; and Testes (5.3.3), where testicular morphology and function of
testes are examined for alterations.
RF radiation has been examined intensively for its potential reproductive
effects for a variety of reasons: 1) many reproductive toxicologic tests can
be carried out in a simple and inexpensive experimental design after insult by
RF radiation, 2) the laboratory techniques used are conventionally acceptable
as toxicologic assays, and, especially, 3) detrimental reproductive changes
carry an emotional impact.
5.3.1 Teratology
While making judgments on the potential of RF radiation for teratologic
manifestations, it is important that the science of teratology and its under-
lying principles be kept in mind. Wilson (1973) has derived general
principles of teratogenesis. These principles are rephrased here to famil-
iarize the reader with the guidelines one should keep in mind when evaluating
the available literature, and so that their relevance to RF radiation teratologic
investigations is more easily understood.
5-73

-------
(1)	Susceptibility to RF radiation teratogenesis depends on species,
strain, and stage of development at the time of exposure.
(2)	There are four indications of abnormal development: death,
malformation, retarded growth, and deficient function.
(3)	These indications increase in incidence and degree with in-
creasing dosage.
Other aspects of teratology, as a branch of toxicology, are presented in
discussing effective experimental design and interpretation. The interpre-
tation of teratologic phenomena is still an art, and there is not full agreement
on the significance of certain fetal manifestations after potentially toxic
agents are administered. The discussion posits that the list of possible
deviations in the conceptus is an order of decreasing degree of severity
(death > malformation > growth retardation > functional deficit). Further,
the significance that any laboratory animal or system may have as a model of
human exposure is not universally accepted among scientists.
An attempt is made to derive from available data three aspects of terato-
logic toxicology of RF radiation: the presence of a teratologic effect, the
generalization of that effect across species, and the dose-response character
of that effect. From the evidence presented, we think that the reader will
agree with the following statements:
(1) RF radiation can cause teratogenesis in all the mammalian
species studied adequately so far if sufficiently high power
densities or SAR's are obtained.
5-74

-------
(2)	Reduced fetal weight seems to occur consistently in rodents
exposed gestational!}? to teratogenic or subteratogenic doses
of RF radiation.
(3)	Gestational exposure to RF radiation can cause a functional
deficit in later life if sufficiently high SAR's are obtained.
5.3.1.1 Nonmammalian Models--
The potential for human teratology is usually sought in mammalian models.
But other models that are lower on the phylogenetic scale or cell-culture
models are also often elucidative. Workers at the Bureau of Radiological
Health of the Food and Drug Administration have used an insect to demonstrate
the reproductive alterations due to RF radiation. Pay et a2. (1978) examined
the egg production of female fruit flies (Drosophila melanogaster) in response
to RF radiation. Using a 2450-MHz waveguide exposure system housed in an
environmentally controlled chamber (24 °C, 50 percent relative humidity),
these workers determined the survival rate of IL melanogaster pupae at SAR's
ranging from 400 to 800 W/kg. Approximately 70 percent of the pupae did not
survive a 10-min exposure when the rate of energy absorption was 640 W/kg; the
temperature of the agar surface on which the pupae rested was 45 °C. If pupae
were incubated instead in a 44 °C environment without RF radiation, the death
rate was not as severe (50 percent).
The potential for RF-radiation-induced teratology in birds has been
examined using the fetal form of birds, the egg, as a model. This is a
reasonably popular model since using bird eggs is more economical than using
mammals. Studying the bird egg can provide insight into fetal effects
independent of maternal parent influence. The egg is an object that can be
5-75

-------
placed in almost any desired position by the experimenter; it will remain
there, a distinct advantage over free-ranging mammals. The shape of the egg
represents a technical advantage over the mammal in that it is symmetrical
externally, lending itself more easily to estimates of local and total
absorption.
Theoretically, the bird egg gives no advantage to mammalian models in the
study of teratogenic potential of RF radiation because concepts of organo-
genesis apply equally to avian fetuses and mammalian fetuses. To dismiss the
considerable research on the teratologic effects of RF radiation in birds'
eggs on the basis that the egg is poikilothermic whereas the mammal's egg is
homeothermic would be an error. The mammalian conceptus is also poikilo-
thermic; it has little, if any, control of its own temperature since it is
entirely surrounded by placental fluids. It can remove itself of thermal
energy only by radiating into surrounding maternal tissues because its
capacity to rid itself of thermal energy is totally dependent on the gradient
between itself and surrrounding tissues. If the mammalian conceptus' tissue
temperature is increased because of RF-radiation absorption, it also must
radiate that absorbed energy into its surroundings (dam), just as the egg in
similar exposure conditions must radiate energy into the air of its incubator.
If air or maternal temperatures are detrimentally high, then the loss of
thermal energy from the egg or mammalian fetus is affected similarly.
Carpenter et al_. (1960a) have studied RF radiation as an inducer of
cataracts. These investigators have also likened the growth in the lens to
the process of embryonic development and growth, where proliferation and
5-76

-------
differentiation take place concurrently. To further understand the RF-
radiation-induced effects that appear in the lens, Carpenter et aK performed
experiments using the egg of the domestic chicken. They irradiated almost 500
chick embryos at the 48-h stage and examined the eggs 48 h after irradiation,
scoring incidences of survival and structural abnormalities. The irradiation
was conducted in an anechoic chamber at 2450 MHz and at power densities of
2
200, 280, and 400 mW/cm . Estimates of SAR's for these power densities are
70, 98, and 140 W/kg, respectively (Durney et aK 1978). Power densities were
determined by calorimetry using a saline-filled egg. The exposure durations
ranged from 1 to 15 min. The eggs were incubated at 39 °C air temperature
before, during, and after irradiation. At the 96th hour of development, all
eggs were opened, and the embryos were removed, fixed, stained, and examined
as whole-mounts.
Experimental results relevant to this section are presented in Table 3 of
the Carpenter et aK 1960a publication. We have used the data in that table
to calculate the relative summed incidences of dead and abnormal 96-h egg
embryos for each experimental group (power density:time) and present the
incidences in Figure 5-2.
The teratologic end points observed by Carpenter et aK (1960a) were death
and abnormal morphology, which can be viewed as a continuum of one effect,
interference with the normal processes of development. When abnormal
structure is so severe as to interfere greatly with the continuation of
development, the embryo cannot develop or maintain sufficiently normal
5-77

-------
structural and physiologic systems. The relationships between these systems
collapse and the embryo dies. As stated earlier, indications of abnormal
development increase in incidence and degree as the dosage increases. One
mechanism for the increase in death is that, as the dosage increases, more
embryos affect more changes in more systems, resulting in more deaths.
2
o
"o
£
1/3
a
o
CO
<
>
z
o
z
a
z
<
100-
90-
80-
70-
60-
50-
40-
<
S
tr
i 20
oa
<
30-
10-
0-
~r
10
~r
14
"T"
15
11
12
13
DURATION OF EXPOSURE (min)
Figure 5-2. Summed incidence of abnormal and nonviable chick embryo
eggs (percent of total) exposed to 2450-MHz radiation
for varying durations at 70, 98, or 140 W/kg (Carpenter
et al. 1960a).
Figure 5-2 contains three curves, one for each group of eggs exposed at
dose rates of 140, 98, or 70 W/kg for varying durations. The data are plotted
5-78

-------
against the summed relative incidence of affected (dead or morphologically
abnormal) eggs. The slopes of the curves are steep, indicating that small
increases in the exposure duration caused disproportionate increases of effect.
For example, the incidence in the group exposed for 10 min at 140 W/kg was 9
percent. An increase of only 30 s in exposure time caused a ninefold increase
(80 percent) in incidence. An additional characteristic of these slopes is the
disproportionate spacing along the exposure axis of the three curves.
Studies of RF radiation effects using the Japanese quail have been
conducted exclusively by workers at the National Institute of Environmental
Health Sciences. In their first attempt to induce terata in the quail egg,
Hamrick and McRee (1975) exposed fertilized eggs in an anechoic chamber to
2450-MHz (CW) fields producing an SAR of 14 W/kg in the egg. The eggs were
exposed for 24 h beginning 2 days after laying, then were returned to their
incubators until 24 to 36 h after hatching. Several experimental replications
occurred so that the sample numbers were 102 in the control group and 110
in the RF-irradiated group. After hatching, observations were made of
cellular elements of blood and of selected organ weights. No significant
differences between the two groups were seen in hatchability, blood cell
parameters, body weights, or organ weights.
These authors also reported a similarly designed experiment on Japanese
quail eggs, in which the eggs were exposed during the first 12 days of
development at SAR's of 4.03 mW/g (McRee and Hamrick 1977). As before, the
eggs were allowed to hatch and were then examined for teratologic alterations
and blood-cell parameters. When the experiment was conducted at an
5-79

-------
environmental temperature of 35.5 °C, there were no significant changes in
body weights, organ weights, or blood-cell parameters.
The authors also demonstrated the importance ambient temperature can play
in experiments in which RF radiation is the toxic agent under investigation.
Absorption of RF radiation energy produces increased temperature. In most
studies of reproductive effects reviewed in this document, fields were at
sufficiently high power densities to cause significant depositions of energy.
Temperatures in the animal subjects were usually increased by 0.5 °C or more.
Therefore, the ambient temperature in which the subject is maintained or
treated can be a significant determinant of a resulting trend.
Levels of ambient temperature can be especially relevant to experimental
subjects that do not have a significant thermoregulatory capability. For example,
the report by McRee and Hamrick (1977) included the results of an experiment
where Japanese quail eggs were in an environment at a temperature only slightly
higher (+1.5 °C) than in the experiment discussed above (35.5 °C). The authors
began their study using an ambient temperature of 37 °C because that is the
temperature at which quail eggs are conventionally incubated. But this ambient
temperature, with the addition of a 4 W/kg SAR, was sufficient to cause a 93
percent death rate of the fertilized eggs. The eggs that were irradiated at
the same SAR, but at an ambient temperature lower by 1.5 °C, sustained only a
42 percent death rate. The results of this seemingly slight (1.5 °C) increase
in incubation temperature during RF irradiation shows that for this poikilo-
thermic model, the quail egg, the temperature at which the egg is maintained
during experimentation is critical.
5-80

-------
5.3.1.2 Mammalian Models--
Morphologic teratology—The mouse is commonly used in the laboratory for
determining the teratogenic potential of toxic agents and also appears to be
a useful model for determining the teratogenic potential of RF radiation. The
mouse's size is conducive to this kind of experimentation because the apparent
resonance of mice is 2450 MHz, a frequency at which irradiation equipment is
commonly and economically available.
Rugh et aK (1974, 1975) conducted several experiments on the mouse at
2450 MHz using the experimental techniques from work on the teratology of
ionizing radiation to explore teratologic effects of microwaves. The in-
vestigators used a waveguide exposure system in exposing individual pregnant
CF1 mice. All mice were exposed once on the 8th day of pregnancy to CW fields
at 2450 MHz for 2 to 5 min at a forward power that delivered a measured
average energy dose ranging from 2.45 to 8 cal/g; there were no control or
sham-irradiated mice in this study.
These and a few other authors have reported dose parameters in calories
or Joules per gram, which we have converted to the SAR unit (W/kg) used in
this document. Each of the texts of the two papers by Rugh et a^. (1974,
1975) refers the reader to the other for more detailed methodology, but
neither adequately describes exposure factors to convert accurately the dose
values to dose rate (SAR). Our best solution was to convert to a range. The
range of doses given by Rugh et a2* is 10.3 to 33.5 J/g delivered over 2 to
5 min. The four possible combinations (Table 5-8) of two extremes of dose
5-81

-------
TABLE 5-8. CONVERSION OF J/g TO W/kg*
Average Dose	Exposure Duration	SAR
(J/g)	(min)	(W/kg)
10.3	5	34
10.3	2	85
33.5	5	112
33.5	2	280
*W/kg = J/g x 1000 t seconds (from Rugh et aT.. 1974, 1975).
(10.3 and 33.5 J/g) and two extremes for exposure duration (2 and 5 min)
result in SAR's ranging from 34 to 280 W/kg for this study. The most probable
range of values is 85 W/kg (10.3 J/g for 2 min) to 112 W/kg (33.5 J/g for
5 min).
The authors clearly demonstrated the teratogenicity of RF radiation in
mice, especially the potential for death and generation of anomalies. By
determining the energy dose in each pregnant mouse, and because varying doses
were absorbed during the 2 to 5 min of exposure, the effect observed in each
litter could be assigned to a definite whole-body average dose. This is the
strong point in the Rugh et al^. studies.
Rugh et al. (1974, 1975) showed that resorption of fetuses is a character-
istic of the levels of RF-radiation exposure. Resorptions appeared at a dose
rate of approximately 85 W/kg. Both the incidence of resorptions within
litters and the incidence of litters with resorptions increased as the average
5-82

-------
dose increased. Near the maximum dose (33.5 J/g, approximately 110 W/kg)
almost all the litters were affected. Even at 25 J/g, there were litters with
100 percent resorption rates.
Anomalies "... were various, but the one used for this study, and which
occurred most frequently, was exencephaly (brain hernia) 	" The report
also included data on resorptions, dead, and stunted fetuses. In lieu of
numerical data, the incidences of brain hernia (also termed encephalocele,
exencephaly, or cranioschisis) are given by the authors in one figure. Data,
as percentage of affected litters, were grouped into cells 8.4 J/g in width
that ranged from 10.9 to 35.9 J/g. Approximately 30 percent of the litters
given an average dose of 15 J/g (approximately 90 W/kg) had fetuses with
brain hernias. Of those litters given an average dose of 31.8 J/g (estimated
as 109 W/kg) almost all contained fetuses with some form of brain herniation.
Though sham, cage-control, or historical control value for the incidence
of brain hernia in this mouse strain (CF1) is not given, the authors state
"... since this aberration rarely if ever occurs without trauma, it is signi-
ficant when even a single exencephaly occurs.". However, one report of the
incidence of spontaneous brain hernia in CF1 mice showed an occurrence of 11
exencephalic litters in a total of 90, approximately 12 percent of litters
(Flynn 1968). Apparently, spontaneous incidences of anomalies are quite
variable.
A report by Rugh and McManaway (1977) included data on mice exposed once
for 4 min to 2450-MHz RF radiation (SAR = 100 to 114 W/kg) on days 0 to 11 (10
5-83

-------
mice per day). In this group of mice (we cannot differentiate these data from
those reported in Rugh et ah 1975 and Rugh 1976a,b) are uneven incidences of
stunted fetuses (irradiated on day 4 of gestation) and death (irradiated on
day 3, 8, or 10 of gestation). Rugh considers small fetuses as stunted and
believes that this is a teratologic symptom.
Rugh's experiments included using a single, but high field intensity for
a very short time (i.e., 4 min on the 8th day of gestation). A longer period
(100 min daily during entire organogenesis) was used by Berman et ah (1978)
in the exposure of pregnant CD1 mice to 2450-MHz CW fields at power densities
of 0 to 28 mW/cm in an anechoic chamber. Between 15 and 25 mice were
irradiated simultaneously, each in a small ventilated container. Mean rectal
temperature at the end of exposure to the highest power density (28 mW/cm )
was 38.2 °C. SAR's were determined by twin-well calorimetry and averaged 2.0
2	2
W/kg at 3.4 mW/cm , 8.1 W/kg at approximately 14 mW/cm , and 22 W/kg at
28 mW/cm^.
This study incorporated the examination of several variables of fetal
response to toxic agents. Values were tabulated by power density for
pregnancy rates, incidences of live and dead fetuses, live fetal weights, and
incidences of anomalies. Values were expressed using the litter as the
experimental unit. This is a conservative approach in experiments in which
the potential fetal toxicity of an agent is determined (Atkinson 1975).
Berman et ah (1978) also "corrected" the mean fetal weight of each litter by
a factor representing the potential influence that litter size had upon fetal
5-84

-------
weight. This technique is also endorsed by Atkinson, because it allows
greater confidence in the interpretation of data.
The study by Berman et ah (1978) demonstrated that daily exposure to
2450-MHz fields at intensities producing an SAR of 22 W/kg during the entire
period of organogenesis caused a 10-percent decrease in body weight of mouse
fetuses. The practical significance of this result is difficult to determine
without experimental confirmation of the permanency or lack of reversibility
of the smaller fetal weight. Until it can be shown that the effect is
permanent (stunting) or temporary (delayed growth), the effect remains
hygienically unresolved. The effect will have more practical significance if
it can be shown that stunting is the result.
Berman et ah (1978) also reported on the incidences of anomalies. There
did not appear to be a significant increase of individual or total anomalies
at any SAR.
The C3H-HeJ strain of mouse was used by Chernovetz et ah (1975) in an
attempt to determine the effects of fetal irradiation with 2450-MHz RF
radiation. Pregnant mice were irradiated in a multimodal cavity for 10 min on
the 11th, 12th, 13th, or 14th day of gestation at an SAR (determined by
calorimetry) of 38 W/kg. The fetuses were examined near term (near the end of
the gestation period) without any difference seen between the RF-irradiated
and the sham-irradiated groups of fetuses. The variables observed in these
fetuses related to the morphologic and the lethal aspects of the RF
irradiation. Fetal body weights were not observed in these animals.
5-85

-------
The scientific reports on the teratologic potential of RF radiation in
the rat represent the major fraction of rodent work. The main reason for
this, perhaps, is that many of the studies that have looked for morphologic
changes have also included functional teratology as one of the end points.
For that kind of study, especially behavioral studies, the rat appears to be
preferred over the mouse.
Chernovetz is perhaps the most published author on the teratologic
potential of microwaves in the rat. She reported (Chernovetz et ah 1977) the
results of exposures of pregnant rats to 2450-MHz fields in a multimodal
cavity (most likely the same device used to irradiate mice in experiments
discussed above in Chernovetz et ah 1975) at an SAR calculated by calorimetry
of approximately 31 W/kg. Exposures were conducted during 1 of 7 days, from
the 10th through 16th day of gestation, and the fetuses were examined on the
19th day of gestation. Each single exposure lasted 20 min. The experiment
included a total of 74 time-bred female rats, each assigned to 1 of 7 days of
gestation and to 1 of 3 treatment groups (RF-irradiated, infrared-irradiated,
or sham-irradiated). The result is an experimental design with 21 different
cells into which 74 rats were distributed.
By the end of the 20-min exposure period, the mean rectal temperature in
the surviving pregnant rats was 42 °C. Of the original group of 30 bred
females that were used in the RF-irradiation group, 7 (or 23 percent) died,
while none died in the sham-irradiated group. Therefore, the application of
2450-MHz RF radiation to bred rats for this duration at this SAR is lethal for
a significant portion of the animals involved.
5-86

-------
No numerical data on the gross teratologic consequences of this RF
irradiation were given in the paper. However, comments were made in the text
regarding the lack of any structural abnormalities. The number of resorptions
was significantly higher (approximately 6 times) in the RF-irradiated than in
the sham-irradiated fetuses, especially in the dams that had been irradiated
on the 11th day of gestation. Fetal weights were also altered by the irradi-
ation regimen; there was a small but statistically significant decrease in
fetal weight. The alterations seen in the fetuses, in this case decreased
fetal weights and death (in the form of resorptions), reflect the teratogenic
potential of RF radiation exposure in the rat. It is important to remember
that the fetal alterations were found from RF-radiation exposure regimens
where rectal temperatures rose above 40 °C.
Chernovetz et al. (1979) reported on the effect of 2450-MHz RF radiation
in pregnant rats, this time at SAR's of 0, 14, or 28 W/kg. The same multi-
modal cavity was used, through which air at 22 ± 1.5 °C flowed at 0.75 m/s.
Pairs of bred female rats were exposed for a 20-min period on the 8th, 10th,
12th, or 14th day of gestation and then were examined on the 18th day of
gestation. With cage controls added to each of these cells, the complete
experimental design used 4 different days of treatment and 4 treatments (cage
control, 0 W/kg [shams], 14 W/kg, and 28 W/kg), and 6 bred females per cell,
for a total of 72 bred females. Those rats irradiated at 28 W/kg developed
rectal temperatures of approximately 42 °C, temperatures that were similar to
the author's previous experiments in which bred female rats were dosed at
31 W/kg. The rats irradiated at 14 W/kg had a mean peak rectal temperature
near 40 °C.
5-87

-------
The values used by the authors in the statistical analyses were individual
fetal values, not litter mean values. At the higher SAR level (28 W/kg), only
~ 10 percent below the sublethal level used in their previous report, the
authors did not demonstrate lethality in the dams due to RF-radiation exposure.
The 20-min exposure period at other levels of SAR produced no gross morphologic
alterations in the fetuses or any severely edematous fetuses; there were also
no statistically significant changes in resorption rates.
The results of fetal weights appear to be most interesting. In the
previous study (Chernovetz et aK 1977), the authors reported that exposure
for 20 min at approximately 31 W/kg on odd-numbered days of gestation produced
a significant decrease in fetal weight. In the 1979 study, a similar result
was also obtained from exposure at 14 W/kg on even-numbered days (12 and 14),
or 28 W/kg on the 8th day of gestation. Unexpectedly, exposure at 28 W/kg on
the 12th or 14th day produced an increased body weight. The unexpected in-
creased body weight is an interesting observation, especially when it is
viewed against the decreased fetal body weight documented in the same authors'
previous paper and other reports of RF-radiation-induced fetal alterations in
rats and mice.
Berman et al^. (1981) were not able to elicit any fetotoxic or teratologic
responses (body weight; numbers alive and dead; external, visceral, or skeletal
morphology) in fetal rats irradiated daily (100 min/day) during gestation,
even with a large number of litters. The dams were exposed to 2450-MHz
radiation at power densitites of 0 or 28 mW/cm (4.2 W/kg). These conditions
5-88

-------
caused a mean rectal temperature of 40.3 °C by the end of the 100-min
exposure.
Kaplan (1982) did not find differences in the survival of squirrel
monkeys up to 10 months of age after daily i_n utero and postnatal exposures to
2450-MHz in multimodal cavities. Pregnant squirrel monkeys were irradiated
for 3 h daily, beginning in the first trimester; after birth, the daily
irradiation of the dams and their young were continued, and from 6 to 10
months of age only the young monkey was irradiated. The SAR was determined by
calorimetry to be 3.4 W/kg. In 21 irradiated and 22 sham live-born, dif-
ferences were not seen in sex ratio, the number dead by 10 months of age, or
in the mean age at death. Postmortem examinations did not reveal any tendency
to specific causes for death. The results of this study did not confirm the
results of suspected increased lethality in squirrel monkeys from a previous
study.
Functional teratology—Functional teratology is a field of reproductive
toxicology that has allowed increasingly greater insight into the toxicology
of environmental agents. The field contains a wide variety of specialists who
examine the functional capacity of neonatal or growing animals after J_n utero
exposure.
The survival rate of fetuses can be used as an indicator of immediate
effects on the fetus. The rate of survival of neonatal, infant, or older off-
spring can be also used as an indicator of delayed alterations in functional
5-89

-------
capacity long after insult during the fetal stage. For example, Kaplan et al_.
(1982) reported a study of the postnatal effects of jrn utero exposure (at dose
rates to 3.4 W/kg) to 2.45-GHz RF radiation in squirrel monkeys. One effect
the authors observed was an increase in neonatal deaths. However, a more
intensive examination (Kaplan 1981) did not demonstrate any change in death
rate and did not confirm the earlier study.
A simple test to determine whether RF irradiation iji utero would alter
the survival of mice after birth was conducted by Rugh (1976a). In this
experiment, pregnant CF1 mice were exposed on the 9th, 12th, or 16th day of
gestation to a regimen of approximately 4 min of irradiation at 2450 MHz,
which resulted in a mean dose of ~ 25 J/g (104 W/kg). The survivors of this
sublethal exposure were allowed to give birth and their young to mature to 2
months of age. At that time, the offspring were again irradiated in the same
device until each one was dead. The time-to-kill and the mean dose-to-kill
was determined as a measure of the radiosensitivity of the offspring irrad-
iated originally as fetuses. At the time of death, the rectal temperatures of
these mice ranged from 40 to 51 °C.
Because of body weight differences in the sexes at 2 months of age, the
analyses were carried out by sex. Time-to-kill was shortened in males irrad-
iated originally during the 12th or 16th day of gestation. The males
irradiated originally on the 12th day of gestation had a lower mean dose-
to- kill. Therefore, Rugh (1976a) has shown that irradiation by RF radiation
on select days iji utero can alter subsequent (and far removed in time) effects
of re-irradiation with microwaves.
5-90

-------
As noted, one effect observed commonly in fetuses exposed to RF radiation
is decreased body weight. There are difficulties associated with interpreting
altered fetal weight as an indicator of toxicity, especially when the
permanency of decreased fetal weight (delayed growth vs. stunting) is not yet
determined. But Rugh (1976a) gives evidence that mice irradiated as fetuses
weigh less at 2 months of age. In that study, the mean weights of 2-month-old
male offspring of dams irradiated during the 9th, 12th, or 16th day of
gestation were all lower than concurrent sham-irradiated males (P < 0.05).
Females irradiated during the 16th day of gestation were also smaller than
their controls (P < 0.05), but not females from the 8th or 12th day groups.
These results are evidence of a permanent change (stunting) in mice caused by
RF irradiation.
Chernovetz et a_L (1975) carried out an experiment to observe alterations
in functional capabilities after _in utero irradiation. Pregnant mice were
exposed once during the 14th day of gestation at an SAR of 38 W/kg, and the
offspring were examined until weaning for survival. There were 15 dams each
in the RF-irradiated and sham-irradiated groups. The statistical analyses,
done by litters, show that there was a slight (approximately 12 percent) in-
crease in the survival rate of the RF-irradiated litters as compared to the
sham-irradiated controls.
Some experimental designs continue irradiation past birth. One example
of chronic administration at 2450 MHz is reported by Smialowicz et al. (1979a),
where pregnant rats were irradiated daily for 4 h/day at a power density of
5-91

-------
2
4 mW/cm from the 6th day of gestation to term; irradiation of the offspring
continued through 40 days of age. The SAR's were determined by twin-well
calorimetry for several ages of the animals. Pregnant rats weighing 300 to
350 g had a mean SAR of 0.7 W/kg; offspring 1 to 5 days of age and 6 to 10 g
in weight absorbed approximately 4.7 W/kg. There was no significant dif-
ference between the mean body weights of the males (female offspring were not
used in this experiment) in the 12 sham-irradiated litters when compared to
the mean body weights of the males in the 12 RF-irradiated litters.
Shore et aL (1977) reported decreased body weight and decreased brain
weight in postnatal rats exposed in utero to 2450-MHz RF radiation. In this
experiment, pregnant rats were exposed at an average power density of
10 mW/cm^ (SAR estimated, from Durney et aL 1978, to be 2.2 W/kg) for 5 h per
day repeatedly on days 3 through 19 of gestation. Rats were allowed to de-
liver naturally and were observed frequently thereafter. Rats were grouped by
E- or H-field orientation during exposures, and only the 3-day-old offspring
that had been exposed parallel to the E-field were significantly different,
having lowered body and brain weights. We cannot easily explain this prefer-
ential decrement on the basis of either age or orientation during exposure.
Our expectation is that there is no significant difference due to orientation
in whole-body SAR in rats at this frequency (Gage et aL 1979).
In experiments conducted by Michaelson et aK (1978), some aspects of
functional teratology of RF radiation were explored in Long-Evans rats. In
these experiments, rats were exposed to 2450-MHz RF radiation at power
5-92

-------
2
densities of 10 mW/cm for 1 h on the 9th and 16th day of gestation, or
2
40 mW/cm for 2 h on the 9th, 13th, 16th, or 20th day of gestation. According
2
to data based on water-calorimetry, an exposure at 40 mW/cm produced an SAR
2
of approximately 10 W/kg and approximately 2.5 W/kg at 10 mW/cm . Exposure at
2
40 mW/cm caused an increase in the mean of rectal temperature of approximately
2
1 to 2 °C over that of the sham-exposed animals. The exposure at 10 mW/cm
(2.5 W/kg) on the 9th or 16th day of gestation caused a significant increase
(approximately 0.5 to 1 °C) in the temperature of dams at the 16th day of
gestation. There were no statistical differences in the sizes of born litters
as a consequence of exposure during gestation. There also appeared to be no
difference in the growth rates up to 21 days of age of the rat pups that were
exposed to RF radiation as compared with shams, nor in their relative brain
weights.
At the symposium where Michaelson described his work, Johnson et aK
(1978) reported on a study, then in progress, on the functional teratologic
effects in rats exposed to 918 MHz for 20 h per day for 19 days of gestation
at a power density of 5 mW/cm . The experiment is especially interesting
because it was conducted at a frequency close to the resonance of the experi-
mental subject (rat), and it is the only reported study conducted with almost
continuous exposure. The SAR in this study is approximately 2.5 W/kg,
determined by calculations from thermographic scans. The eight RF-irradiated
and eight sham*irradiated bred dams were maintained in the waveguide exposure
apparatus under environmental conditions of 22 °C, 45 percent relative
humidity, and ad libitum access to rat chow and water. The exposure was begun
5-93

-------
on the first day of pregnancy, the day on which the copulatory plug was first
seen. In this experiment, the dams remained in individual waveguides or sham
condition until day 20 of gestation, at which time they were removed and
placed in delivery cages. At 4 days of age, the litters were culled to four
males and four females per litter, which were then observed through 91 days of
age.
There appeared to be no differences in the litter means for the number of
pups born or for number dead during the first day after birth. There were no
pups in any of these litters (including RF-irradiated, sham-irradiated, and
cage-controls) with any visible physical defects. Body weights at birth, at
28 days of age, and at 91 days of age were analyzed for significant dif-
ferences on an individual pup basis. The only statistically significant
differences in body weights were found between the RF-irradiated and cage-
control females at 28 days of age. A difference also existed in males at 91
days of age, when the RF-irradiated and the sham-irradiated males weighed less
than the cage-control males.
Another parameter of development used as an assay of functional tera-
tology in this study was the age at eye-opening. The phenomenon of eye-
opening is a little understood but frequently used indicator of developmental
maturation. There appeared to be a significant shift to earlier eye-opening
in animals exposed to RF radiation during gestation, and this maturation was
earlier by approximately 1 day.
5-94

-------
Jensh reported a series of experiments designed to examine pre- and
postnatally the effects of 915-MHz, 2450-MHz, and 6-GHz irradiation of rat
fetuses (Jensh et al_. 1979; Jensh 1979, 1980). These studies included irra-
diation up to 8 h daily during most of gestation with power densities not
2
sufficient to cause increased core temperatures (915 MHz, 10 mW/cm , 2 W/kg;
2450 MHz, 20 mW/cm^, 3 W/kg; 6 GHz, 35 mW/cm^, 3.5 W/kg; SAR estimates from
Durney et al^. 1978, p. 96). Offspring were examined pre- and postpartum in a
complex series of observations meant to locate behavioral or morphological
alterations. None were noted.
In lieu of any other workable hypothesis supportable by experimental
data, the most plausible available theory that can account for RF radiation
effects is that, upon absorption, microwaves deposit energy that is converted
to thermal energy. There remain differences between theories that attribute
individual effects to a general input of energy, or the heating of local
(hot-spot) areas, or some more subtle contribution of microwaves not charac-
teristic of conventional methods of heating animals.
Comparisons between RF heating and infrared or convective heating have
been conducted in the study of hyperthermia as a teratogenic agent. Edwards is
a frequently cited investigator of the effects of hyperthermia on the develop-
ment of mammals. His extensive work on this agent as a cause of congenital
malformations is perhaps best summarized in one of his reviews (Edwards 1974),
which describes his experiments on the teratologic manifestations of hyper-
thermia in the guinea pig. It should be noted that the guinea pig is an
unusual laboratory animal in its long gestational development, approximately
5-95

-------
65 days. During the latter two thirds of gestation in the guinea pig, the un-
born offspring is in a stage of development similar to that which occurs after
birth in the mouse and the rat. In itself, a long gestation period does not
prevent use of the guinea pig as a model for teratogenesis. But the reader is
forewarned that, when reading about teratologic experimental methodology, the
administration" of the agent at 40 or 50 days of gestational age in the guinea
pig has no comparison in rodent fetuses and that this stage of development
occurs postpartum in mice and rats.
Studies of the teratogenic potential of conventially induced hyperthermia
in almost all the species used show some general changes that can be seen in
RF-radiation-treated animals as well. The symptom of hyperthermia, whether
induced by RF radiation or otherwise, causes varied anomalies (such as kinked
tail, microphthalmia, exencephalia, cleft palate, general edema). This is
evidence that hyperthermia is a general teratogen. Perhaps the only case
where RF-radiation-induced and conventionally induced teratogenesis is not
equal is in the development of smaller brain weights in fetuses heated by
non-RF-radiation-heating agents. Except for Shore et aK (1977), perhaps the
literature in RF-radiation-induced teratogenesis has not been sufficiently
developed to show this effect of decreased brain weight.
Conventionally induced hyperthermia, like RF-radiation hyperthermia,
appears to affect most species if sufficient energy is applied and proper
timing of the agent is obtained. A sufficiency is usually manifested by a
rise in maternal rectal temperature, usually over 40 °C, but often in a range
of 41 to 43 or even 44 °C. At these temperatures, the teratogenic potential
5-96

-------
of heat applied by conventional means or by RF radiation does not appear to
show any significant differences. The authors of one study attempted to
differentiate teratogenic effects of microwaves from the more conventional
methods of heating (in this case, infrared irradiation). Chernovetz et al^.
(1977) exposed pregnant rats to 2450-MHz fields at an SAR of approximately
31 W/kg, or to infrared radiation. The rectal temperatures of the RF-
irradiated and the infrared-irradiated groups were similar (approximately 42
°C). The results of the exposures showed little difference in effects due to
infrared heating or RF-radiation heating.
The symptom of decreased body weight of fetuses is a general develop-
mental effect. This response to microwaves is commonly seen in experiments,
even those experiments that have not otherwise demonstrated specific
morphologic changes (e.g., O'Connor 1980). Body weight changes alone have
also been seen in mice fetuses at temperatures < 40 °C (Berman et al_. 1978).
Body temperatures reported in the above studies were those of the dam,
usually measured in the dam's large intestine. Measurements of fetal tempera-
tures after RF-irradiation have not been reported. A report by Morishima et
al_. (1975) provides some understanding of the physiologic changes that occur
in the fetus as a result of conventionally induced hyperthermia. Morishima et
al^. used pregnant baboons and simultaneously monitored temperature and
physiologic changes in the dam and the fetus during hyperthermia. The
temperatures in these dams were raised to 41 to 42 °C by infrared lamps and
warming pads. While the pregnant dam was maintained in an analgesic state
using nitrous oxide inhalation anesthesia and intravenous succinylcholine
5-97

-------
chloride infusions, thermistors and catheters were placed at comparable
positions in the dam and the fetus. Two groups of animals were used; one
group was kept at 38 °C, and the temperature of the other was allowed to
elevate gradually to 41 or 42 °C. At the more normal temperature, 38 °C,
there was a slight but steady increase (0.5 °C) in the temperature of the
fetus over that of the dam. When the temperature was increased to 40 °C in
the dam, the fetal temperature rose accordingly and remained 0.75 °C higher
than the dam's temperature. The hyperthermia produced increased uterine
activity and some acidosis in the dam, and a profound acidosis and hypoxia in
the fetus. If conditions comparable to those seen in the baboon are also seen
in the mouse, then we might expect the mouse fetus to have temperatures
greater than that of its dam.
That the teratogenic potential of RF radiation might be dependent only on
the deposition of energy as heat was shown experimentally by Rugh and McManaway
(1976). They exposed pregnant mice to highly teratogenic (high incidence of
fetal death) levels of RF radiation with and without pentobarbital anesthetic.
The mice exposed during anesthesia had normal rectal temperatures and normal
incidences of fetal death. Therefore, the anesthetic protected against the
primary temperature effect of the radiation by reducing "... the body tempera-
ture to a degree equivalent to the rise in temperature expected from the
conditions used." This is clear evidence that the increased fetal
t
abnormalities are strongly, if not solely, associated with increased maternal
temperature.
5-98

-------
Further evidence for a direct relationship between fetal effects and
hyperthermia induced by RF radiation is shown in several reports (Dietzel and
Kern 1970; Dietzel et a^. 1972; Dietzel 1975). In these experiments, groups
of pregnant rats were subjected to 27-MHz radiation to raise rectal
temperature to 42 °C. From these data, the thresholds for malformation rates
appear to be between core temperatures of 39.0 °C maintained for 5 min and
40.5 °C for 10 min.
Observed teratological effects are summarized in Table 5-9.
5.3.2 Reproductive Efficiency
This section discusses aspects of reproduction where the conceptus is not
"insulted" directly by the agent, although the fetus may be involved in demon-
strating the effect. Generally, reproductive efficiency is the capacity of
the dam or sire to effect conception and to bear and rear offspring. Changes
in this capacity might be due to alterations in behavior, physiology, or
morphology. For example, reproduction efficiency might be affected by changes
in maternal cyclical hormone secretions. Tests for reproductive efficiency
are not conducted as frequently as those for teratologic effects, usually
because the male and female reproductive systems require considerable altera-
tion by a toxic agent to cause significant fetal wastage or increased
reproductive efficiency. The testes as factors in reproductive efficiency are
discussed separately in the next section.
5-99

-------
TABLE 5-9. SUMMARY OF STUDIES CONCERNING TERATOLOGIC EFFECTS OF
RF-RADIATION EXPOSURE
Exposure Conditions*
Effects
Species

Intensity
(mW/co2)
Duration
(days x «in)
SAR
(W/kg)
30% survival of pupae
D. aelanoaaster

1 X
10
640
Eobryonic LD^q
Chicken

200
1 X
12
70


280
1 X
7
98



400
1 X
4
140
Decreased postnatal survival
House


1 X
4
104
Teratogenesis
House


1 X
2-5
85-112
No change in teratogenesis
House


1 X
10
38
Increased postnatal survival
House


1 X
10
38
Maternal lethality, resorptions,
Rat


1 X
20
31
decreased fetal weight






Decreased fetal weight
House

28
12 x 100
22
No change post-hatching, hatchability,
Japanese
quail

1 X
1440
14
heoograo, body or organ weights






No change
Rat


1 X
20
14
No change
Rat

40
1 X
120
10
No change
House

3.4-14
12 x 100
2-8
Teratogenesis
House


12

4
No change
Rat

5
Hany x 240
0 7-4 7
No change
Rat

28
12 x 100
4 2
No change
Squirrel
¦onkey

Hany x 180
3 4
No change
Rat

10-35
Hany
1-3.5
No change
Rat

5
19 x 1200
2.5
Decreased body and brain weight
Rat

10
16 x 300
2 2
" Frequency used was 2450 MHz, except for Jensh (914, 2450, and 6000 Wz), and Johnson 1978 (918 MrizJ

-------
We have previously discussed the effects of 2450-MHz RF radiation on D.
melanogaster reported by Pay et al.. (1978). The fruit flies that were the
subjects of that study were irradiated before production of ova. The SAR used
in the study ranged from 400 to 800 W/kg. Reproductive efficiency, in this
case the production of eggs, was significantly reduced in both conventionally
heated and RF-irradiated females as compared with the shams. But, there
appeared to be no significant difference in the production of ova from females
exposed to RF radiation compared to those exposed to the conventional source
of thermal energy.
The egg production of chickens can also be used to provide an index of
reproductive efficiency or reproductive wastage. A single, intense exposure
of day-old chicks was made in the experiment reported by Davidson et ajL
(1976). Described in one of four experiments is the reproductive efficiency
of 28-week-old hens that had been irradiated as chicks on day 1 of age for
2
4.5 s at a power density of 800 W/crrl in a multimodal cavity. The estimated
SAR was 2770 W/kg. No differences appeared between the control group and the
irradiated group in the production of eggs during 100 days of laying. The
authors also stated that a dose rate of 2500 W/kg for 9 s at 2450 MHz is a
lethal dose in day-old chicks and that 42 percent of the chicks dosed for 6 s
at 2810 W/kg died. The animals that survived were unconscious for up to
5 min. The day-old chicks had approximately the same sublethal SAR (2770 W/kg
for 4.5 s) used in the examination for latent reproductive effects, but there
were no deaths that could be directly attributed to the RF-radiation exposure.
5-101

-------
From these experiments, it is concluded that the massive doses of this experi-
mental regimen left no latent alterations in reproductive efficiency.
Rugh et a^. (1975) reported an experiment in which they examined dif-
ferences in average lethal dose of RF radiation (2450 MHz) during the estrous
cycle of CF1 mice. The weights of the mice ranged from 29 to 31 g. There was
a prior 20-min acclimation to the waveguide exposure chamber, and environ-
mental conditions were 23.5 °C, 50 percent relative humidity, and an airflow
of 38 L/min (0.38 km/h). There was a significant decrease in the average
lethal dose for females in estrus as compared to the average lethal dose for
those in diestrus (P < 0.01). The forward power was 8.24 W. This experiment
shows that changes brought about during the reproductive cycle can affect
radiosensitivity.
In summary, it appears that the efficiency of the female reproductive
system is not easily altered. Only irradiation at extremely high dose rates
has made changes in reproductive patterns. Small alterations from normal
values, though detectable by modern scientific and statistical methodology, do
not seem able to sway the general outcome of the reproductive cycle. Observed
reproductive effects are summarized in Table 5-10.
5-102

-------
TABLE 5-10. SUMMARY OF STUDIES CONCERNING
REPRODUCTIVE EFFECTS OF RF-RADIATION EXPOSURE


Exposure Conditions*

Effects
Species
Duration
(days x min)
SAR
(W/kg)
Reference
Decreased ova
production
L.
melanogaster

400-800
Pay et al.. (1978)
No change in egg
production
Gallus
1 x 0.08
2770
Davidson et al.
(1976)
Lethality changes
with estrous cycle
Mouse


Rugh et aK (1975)
*Frequency was 2450 MHz in all cases.
5.3.3 Testes
Considerable scientific work has been done to determine effects of RF
radiation on the testes. The testes are so placed anatomically that they can
be conveniently irradiated without irradiating the remainder of the body.
Tests for testicular function are also conducted easily. Quantitative changes
in sperm concentration can be conveniently assessed using repeatable laboratory
techniques. It is known that when the mammalian testes, which have a normal
temperature of 33 to 35 °C, are heated to temperatures approaching abdominal
temperature (37-38 °C) sterility can occur. Sterility consequent to high testi-
cular temperatures can be viewed as a purposeful contraceptive agent or as an
unintended toxic agent.
5-103

-------
A report by Muraca et al. (1976) describes the results of irradiation of
rat testicles with 2450-MHz radiation. Each animal was anesthetized, placed
in an anechoic chamber, then irradiated in a free field at a power density of
80 mW/cm^. The SAR for adult rats was 16 W/kg (Durney et aK 1978). But, the
philosophical aim of this experiment was to examine the effects of RF radiation
on the testes, and it may be appropriate to use the SAR of the testes alone as
a more adequate expression of absorption. For that reason, we have also esti-
mated (Durney et aK 1978, Figure 41, model of a quail egg) that the average
SAR to the testes might have been 60 W/kg.
In the experiment, the authors irradiated rat testes to produce increases
of intratesticular temperature to 36, 38, 40, and 42 °C; the testes were then
maintained at these temperatures. The technique of irradiation included
implantation of a thermistor into one testis of the anesthetized male, with the
temperature of that testis acting as the control of the on-off-on sequence of
the RF irradiator. The authors irradiated both testicles, so that temperatures
to which the testes were brought were within ±0.5 °C of the levels mentioned
above. The duration of each irradiation at a power density of 80 mW/cm varied
from 10 to > 70 min. The animals were irradiated once or repeatedly
during 5 consecutive days. Five days after the treatment ended, the animals
were killed and their bodies were infused with solutions to preserve the testis
that was not "invaded" by the thermistor. The microscopic appearance of the
testicular tissue was categorized as follows: apparently normal; appearance of
early inflammatory or degenerative changes in the spermatogenic epithelium
without well developed necrosis; or severe degeneration of the majority of
5-104

-------
seminiferous tubules with coagulated cellular elements. This study also
included a report comparing the effects of immersing the scrotum in a water
bath.
2
When male rats were exposed once (80 mW/cm ; 2450 MHz; 10 to 73 min;
temperature of the testis maintained at 40 °C), there appeared to be no signi-
ficant change in the incidence of apparently abnormal testicular tissue.
After a single exposure of 10 min, during which the temperature of the testis
was allowed to reach 42 °C, the number of animals that had some abnormal
changes tripled. Multiple (five) exposures, even where the temperature
reached only 36 °C for 60 min, caused all the testes to have some changes in
the spermatogenic epithelium. When temperatures were allowed to reach 40 °C
in the testes for repeated short periods (10 to 27 min, five times), severe
degeneration in the spermatic tubules was seen in all the testicular samples.
This study points to two important factors of irradiation of the testes
with 2450-MHz RF radiation: that a minimum temperature (> 40 °C) must be
reached in an acute exposure, and that repetitive treatments are much more
effective. These two factors interplay, as well, so that an acute temperature
excursion, even as high as 40 °C for over 70 min, is not as deleterious as a
lower temperature (36 °C) reached repeatedly (five times).
Fahim et aK (1975) conducted an experiment to compare the contraceptive
capability of microwaves with that of other methods of heating. The authors
used male Sprague-Dawley rats (Holtzman strain) and irradiated them with
2450-MHz radiation (from a diathermy unit of maximum output of 100 W) during
5-105

-------
pentobarbital anesthesia. The applicator of the diathermy unit was 7.5 cm
from the testicles of the rat and provided near-field exposures that do not
allow a good estimate of the associated SAR. By varying the power output of
the unit and by varying the exposure time from 1 to 15 min, the authors
developed four subgroups of animals: those in which testicular temperature
reached 65 °C for 5 min, 45 °C for 15 min, or 39 °C for 1 or 5 min. The males
were allowed to mate with normal females 24 h after treatment, and every 5
days thereafter until pregnancy was observed in the females. "The endpoint
for fertility was the amount of time required for every surviving male in the
treatment group to impregnate a female." Later, the sexual organs were
weighed, and histological examinations were made of the testes and secondary
sex organs.
Raising the temperature of the testes to 65 °C for 5 min or to 45 °C for
15 min caused complete infertility in the males for 10 months. When these
testes were examined histologically, there was no observable spermatogenesis.
When the temperature was raised to only 39 °C, 70 percent of the males
retained normal breeding capability, whereas the remaining 30 percent recov-
ered their fertility within 2 wk. Histological sections indicated normal
spermatogenesis. There appeared to be no differences in testicular weights or
secondary sex-organ weights, even in the group (45 °C, 15 min) that was
sterile 10 months after the RF-radiation exposure had ended.
This experiment demonstrates that a temperature of 45 °C caused by RF-
radiation exposure must be attained in the rat testes to produce permanent
5-106

-------
sterility. Temperatures that ranged to 39 °C, somewhat above rectal temper-
ature, for only a short period (5 min) were not effective and produced only
temporary, if any, infertility. Along with the permanent sterility induced by
the high temperatures (minimum of 45 °C), there was damage to the tubules and
a lack of spermatogenesis.
Sterility can occur in rats following RF-radiation exposure. At high
temperatures, epithelial and tubular structures of the testes can be damaged
permanently. At lower temperatures (approximately normal body temperature)
for short durations, the temperature in the testicles due to RF-radiation
exposure resulted in temporary sterility. This last condition implies that
the spermatogenic cells themselves are not damaged permanently. The rat's
spermatogenic cycle has a duration of about 10 wk, i.e., it takes about 70
days from the first meiotic divisions of sperm generation in the germinal
epithelium until the time when the sperm are fully matured, in place, and
ready for ejaculation. During the last 2 wk of spermatogenesis, sperm are
located in the tubules of the testes and in the epididymis and require very
little additional maturation. There is a constant stream of maturing sperm,
because various seminiferous tubules are in different stages of this long
cycle. The most mature stages of the spermatogenic cycle are most sensitive
to heat (Van Demark and Free 1970). The permanent (10-month) sterility seen
in the experiment above is due not only to loss of the most mature and most
heat-sensitive stages, but also to the loss of sperm at all stages, including
those stages involving the germinal epithelium. The 2-wk temporary infer-
tility seen in the less affected males was caused by the loss of only the most
mature and most heat-sensitive sperm.
5-107

-------
A study to demonstrate an alteration in functional fertility was conducted
by Berman et al^. (1980). The authors' report concentrated on chronic exposures
of rats to 2450-MHz radiation in an anechoic chamber. There were three separate
experiments:
2
(1)	Exposure at a power density of 5 mW/cm for 4 h/day, daily,
from the 6th day of gestation through 90 days of age post-
partum;
p
(2)	Exposure at a power density of 10 mW/cm for 5 h/day for 5 days,
beginning on the 90th day of age postpartum; and
2
(3)	Exposure at 28 mW/cm for 4 h/day, 5 days a week, for
4 continuous weeks, beginning on the 90th day of age
postpartum.
The main purpose of the experiments was to evaluate potential mutagenic
effects of RF radiation on the germ cells of the male rat. Male rats were
bred to untreated female rats shortly after the end of the treatment period.
No mutagenic effect was demonstrated. The breeding data of the assays used in
these experiments were used to evaluate the effects on the spermatogenic func-
tion in rats.
In the first experiment, the animals were exposed daily from the 6th day
2
of gestation to 90 days of age, at a power density of 5 mW/cm . In this
experimental group, the SAR varied inversely with the growth of the animals,
i.e., from approximately 4.5 W/kg in neonates to approximately 0.9 W/kg at 90
days of age. Twin-well calorimetry was not used in the other two experiments,
p
and SAR's are estimated at 2 W/kg at a power density of 10 mW/cm , and
5.5 W/kg at 28 mW/cm^.
5-108

-------
In this study, temperatures of the testes of rats exposed up to 90 min to
2
2450-MHz fields at 28 mW/cm showed an increase from a pre-exposure level of 34
°C to almost 38 °C after exposure. Simultaneously recorded rectal tempera-
tures were ~ 3 °C higher than temperatures of the testes during the entire
period. Sham-irradiated animals had testicular temperatures ranging from
almost 32 °C to approximately 35 °C. The exposure at 28 mW/cm during the
90-min period produced a temperature in the testis equivalent to the normal
rectal temperature.
2
Chronic exposure at 5 mW/cm from the 6th day of gestation through
90 days of age appeared to have no effect on the reproductive efficiency of the
male rats when bred with normal females. Also, exposure at a power density of
10 mW/cm (SAR ~ 2 W/kg) at 90 days of age for a period of 5 days had no
effect on the reproductive efficiency of the males. It was only the most
severe regimen, the exposure of adult male rats at 28 mW/cm for ~ 4 wk
(estimated SAR ^5.5 W/kg), that caused any alteration in reproductive
function. This exposure produced a severe decrease in the reproductive
ability of the males; only 50 percent of the females that were
available to the males for breeding became pregnant during the week
immediately following the exposure period. The breeding returned to normal
beginning at the 3rd week after irradiation (the next period of test breeding).
The animals bred normally thereafter. No examination was made of the testes
2
of animals exposed at 28 mW/cm . The temperatures reached in the testes at
2
the highest power density (28 mW/cm ) were similar to those reported by Fahim
et al_. (1975) where histological changes were seen.
5-109

-------
Cairnie et aL (1980) studied the effects in mouse testes after a range
of whole-body exposures of up to 30 days for 16 h/day at 2450-MHz and power
2
densities up to 36 mW/cm (average whole-body SAR = 7 W/kg; average testicular
SAR = 14 W/kg). These exposures caused no measurable increase in testicular
temperature. No changes were seen in the number of dead testicular cells, the
number of epididymal sperm, or in the percentage of abnormal sperm after
exposure, even up to 8 wk after exposure. Four strains of mice were used, and
no strain susceptibility was observed.
Varma and Traboulay (1975) were able to cause severe scrotal and testi-
cular burns in anesthetized mice after exposure for 20 min to 1.7 GHz and 200
2
mW/cm (whole-body SAR estimated at 80 W/kg). Exposures of < 100 min to 10
2
mW/cm at this frequency caused little or no damage. But a 100-min exposure
caused general testicular damage without changing Sertoli and interstitial
2
cells. No severe changes occurred at 3 GHz, 50 mW/cm , for 20 min (SAR = 20
W/kg).
In the studies reported by Ely et al_. (1964), 2880-MHz radiation was used
to irradiate only the testicular area in groups of dogs. The animals were
anesthetized during exposures while temperatures were measured in the testis.
Irradiation of the testicular area continued until a peak temperature was
reached; the RF radiation was then turned off manually so that the testes
could cool, at which point the RF radiation was turned on once again. This
cycling of exposure was completed many times so that the animal's testes could
be kept at a "steady" temperature for a considerable length of time. The
5-110

-------
normal temperature of the testes of the dogs was ~ 33 °C, 5 °C lower than
that of the rectum.
2
According to this report, exposure of the testes at 20 mW/cm caused an
2	2
elevated testicular temperature of 36 °C at 33 mW/cm , 38 °C at 45 mW/cm and
up to 40 °C. As 38 °C is the approximate body temperature in the dog, bringing
the testes to body temperature can expect to produce sterility if the elevation
2
is sufficiently chronic. It appears then, that 45 mW/cm is required to
produce this type of effect.
There is a large body of literature on the fertility effects caused by
temperature increases, such as when the testicles are clothed to prevent the
normal thermal radiation. Even high environmental temperatures, when sustained,
can produce infertility in male rats. One example of this is a report (Pucak
et al_. 1977) describing deaths in a large rat-production colony in which room
temperatures accidentally reached as high as 31.6 °C for 2 days and as high as
37.7 °C in individual cages. Under these conditions, approximately 3,000 of
14,000 Sprague-Dawley rats died from heat prostration. When examined 18 days
after the incident, 25 percent of the surviving males showed bilateral atrophy
of the testicles, which were approximately half the normal size. Histological
examination showed atrophy of the spermatic tubules and failure of spermato-
genesis. The proportion of affected testicles ranged from 50 to 75 percent.
Five weeks after the incident the animals with small testes were still sterile.
In comparison with the temperatures seen in this study, raising the tempera-
tures of testes to 42 °C by RF-radiation exposure appears to be an extreme
/
experimental regimen. Observed testes effects are summarized in Table 5-11.
5-111

-------
TABLE 5-11. SUMMARY OF STUDIES CONCERNING REPRODUCTIVE
EFFECTS OF RF-RADIATION EXPOSURE IN THE RAT
Exposure Conditions*
Effects	Intensity Duration	SAR	Reference
(days x min) (W/kg)
Spermatogenic tissue	80	1 x 10-73	16	Muraca et al_. (1976)
abnormal	5 x 10-73	16
No change	5	many x 240	0.9-4.5 Berman et aK (1980)
No change	10	5 x 360	2
Temporary sterility	28	20 x 240	5.5
^Frequency was 2450 MHz in all cases.
5.3.4 Unresolved Questions
5.3.4.1 Teratology--
A threshold of teratogenic effects due to RF radiation has not been
determined. The measurement of threshold will have to include the entire
range of teratogenic effects: lethality, anomaly production, decreased body
weight in fetuses, and alteration of postnatal function. Even at this stage
of development of the data relating to RF-radiation-induced teratogenesis, a
picture is appearing in which the degree of response is dependent on the level
of whole-body SAR, the duration of exposure, the timeliness of the exposure,
and the species, all of which complicate the threshold determination. It is
not yet clear if threshold can be based on a finite whole-body SAR for all
species or is some adjustment or proportion will be necessary for species
size, metabolic rate, thermoregulatory capacity, etc. So far, the only
5-112

-------
physiologic variable that can be supportably associated with teratogenic
effects is the colonic temperature of the dam during or at the end of
exposure. The mouse and rat, two extensively studied species, show a minimum
temperature of approximately 40 °C in the dam is associated with teratologic
symptoms. However, whole-body SAR's required to cause temperature excursions
like this are much higher in the mouse than in the rat. Therefore, an adequate
relationship of teratogenesis to SAR alone is not apparent. The maternal
colonic temperature, then, is the only available indicator of a threshold for
teratogenesis in mammals.,
Published reports that meet the criteria for consideration in this
document have limited their examination of the fetal results of gestational
exposure to a gross morphological change or one that might be seen under low
magnification (15X). There has been no organized attempt to examine in
histologic detail fetuses that have been irradiated jfrv utero. Authors of one
study (McRee et a^. 1980b), however, made a detailed examination of embryonic
hearts but could not demonstrate changes in morphologic, ultrastructural, or
enzymatic activity. The subjects were Japanese quail which had been exposed
daily for 8 days to 2450-MHz radiation at SAR's of 4 and 16 W/kg.
There are classifications of fetal changes that represent no gross
structural deficit, but nevertheless represent some variation of structure.
An example of a variation that may not be considered by all teratologists as a
"deficit" is a small but normal fetus. However, if the incidence of this
variation is consistently increased because of the application of a toxic
agent, it could be considered an expression of embryotoxicity. The decreased
5-113

-------
body weight so often seen in offspring exposed to RF radiation (Berman et al.
1978; Chernovetz et aj[. 1977, 1979; Rugh 1976a) might otherwise be considered
as a lesser category of "structural variation without deficit" were it not
that decreased weight is so consistent in these experiments. Whether this
decreased fetal weight is temporary (i.e., only a delayed growth that will
disappear in the neonatal stage) or permanent (i.e., a stunting of the fetus
that will persist) has not yet been scientifically resolved.
There is one aspect of the literature on the teratogenic potential of RF
radiation that deserves further discussion. More than half the papers in this
document on teratogenesis report experiments with the rat. Chernovetz et al.
(1977) raise an interesting point about using the rat in determining the
teratogenic potential of RF radiation. The authors argue that because levels
of microwave exposure which are associated with high mortality rates of dams
do not also produce fetal structural abnormalities in the rat, "... that the
teratogenic threshold of microwave radiation is higher than the dam's
threshold of mortality," and hence "... that maternal mortality is more
probable than malformation of the fetus, irrespective of the dose."
The problem of using animal models in determining teratogenic potential
in humans is that there is no assurance that any of these models is a real
estimate of effects in humans. The concepts supporting the use of animals as
models of human beings require that the response be demonstrated in a number
of species, so that the resultant generality can be more confidently extended
to humans. Therefore, we seek among species some generality of effects of
5-114

-------
microwaves on the fetus. Mice and rats are the two laboratory animals upon
which rest almost all of the data of RF-radiation-induced teratogenesis. Any
difference between the two species in their teratogenic response to RF radiation,
therefore, becomes important.
5.3.4.2 Reproductive Efficiency and Testes--
The testes contain, besides sperm-producing tissues, interstitial cells
that secrete testosterone, the male hormone. Gunn et cH. (1961) described
the effect of 24-GHz radiation on the morphology and function of the interstitial
cells in rats. They exposed rats once at this frequency for a period of 5 min
o
at a power density of 250 mW/cm , which caused the temperature to rise to 41 °C
in the testes. As a result, there were scrotal burns, severe edema, and
spermatic tubular degeneration, but no interstitial-cell pathology.
The testosterone secreted by the interstitial cells of the testes
regulates zinc uptake by the dorsolateral prostate. When Gunn examined zinc
uptake in RF-irradiated rats that had no interstitial cell pathology, he found
a decreased uptake. Gunn related the decreased function of this secondary
sex organ to decreased testosterone secretion.
The effects seen by Gunn were produced at a frequency of 24 GHz. This
is such a short wavelength that, theoretically, there would not be any sig-
nificant penetration to affect the dorsolateral prostate directly. The tests
using the dorsolateral prostate as an indicator of sexual function in the rat
have been commonly used by Gunn and others in other types of experimental
situations.
5-115

-------
It is not known why and how the dorsolateral prostate lost its capacity
to function normally when there was no observable change in the interstitial
tissues, the other testicular damage appeared to be minimal, and the energy
could not reach the prostate. This appears to deserve additional scientific
attention and exploration.
The literature we have cited does not lend itself to extrapolation of
effects of RF radiation that cause only small increases in the temperature of
the testes. The lack of data on RF-radiation effects at lower power densities
(which cause lower testicular elevations of temperature than have been cited
in the articles above) is especially important in light of suggestions that
thermal energy, in the form of absorbed RF radiation, can be used as a contra-
ceptive in men (Medical World News 1974; Arehart-Treichel 1974).
One report by Rugh (1976a) contains data on survival in RF-radiation
fields that are different in males and females. In this study, mice of both
sexes and of three ages (weanlings, young mature, and aged) were irradiated
with 2450-MHz RF radiation until dead. Of the three variables (age, sex, body
weight) examined for their contribution to RF-radiation lethality, Rugh found
that "The overall conclusion would be that no matter at what age ... absorbed
dose to death ... is different for the two sexes. Females are slightly more
radiosensitive ...." These results are not explained easily. Although they
are not apparently related to sexual function, they were included here on the
basis of sexual differences.
5-116

-------
5.4 NERVOUS SYSTEM
Ernest N. Albert
The nervous system in higher forms of life is the result of an evolution-
ary development in response to the environment, and has reached a very high
stage of development in man. The body experiences many environmental stimuli
and responds to these conditions either voluntarily or involuntarily. These
responses are mediated by the nervous system.
Briefly, specialized receptors enable the human being to perceive visual,
auditory, thermal, mechanical, chemical, electrical, pain, and pressure stimuli,
and send the sensory information to the central nervous system (CNS, i.e.,
brain and spinal cord) via the peripheral nervous system (PNS). In addition,
the brain is the site of mental operations such as mathematical integration,
thought, memory, will, anger, and love. The brain integrates all of the sen-
sory information and provides a response appropriate to maintaining homeostasis
within the entire organism.
These functions of the nervous system are carried out by its two major
divisions: (1) the somatic or voluntary, and (2) the autonomic or involuntary.
The somatic or voluntary division is primarily involved in controlling con-
scious activities. On the other hand, the autonomic or involuntary division
of the nervous system is more involved with those functions of the body that
an individual cannot control (e.g., regulation of blood pressure and heart
rate). When the body is suddenly stressed, the heart begins to beat faster;
respiration rate and blood pressure increase involuntarily.
5-117

-------
The basic structural units of all parts of the nervous system consist of
cells (neurons and glia) and their processes (axons, dendrites, and termi-
nals). There are many types of neurons and glial cells as well as many
modifications of their axons and dendrites. These cells and their processes
communicate with each other, as well as with muscles and glands of the body,
via chemical and electrical means in integrating information and maintaining
homeostasis.
When an environmental condition fluctuates within acceptable limits, the
nervous system responds by compensating within the body's physiological limits
with no apparent harm. However, extreme environmental conditions stress the
nervous system beyond these limits, resulting in abnormal physiological func-
tions. Such abnormal functioning could result from exposure to RF radiation.
Therefore, it is important to investigate and evaluate the interactions and
thresholds of interactions between RF radiation and the nervous system.
A critical review of the literature in this area permits the following
general statements to be made about the effects of RF radiation:
(a)	Morphology. There is ample evidence that acute and chronic
exposure of animals at sufficiently high intensities of RF
radiation (CW or PW) produces morphological alterations in the
nervous system. The morphological changes are qualitatively
similar after acute and chronic exposure (at high and low
densities), but quantitatively there are greater alterations in
neuronal structure at higher (> 10 mW/cm2) power densities and
after chronic exposures. In general, PW radiation produces
greater changes than CW radiation. Almost all reported changes
due to RF-radiation exposure are reversible.
(b)	Blood-brain barrier (BBB). There is an ongoing controversy
whether RF radiation at low power densities alters the BBB of
experimental animals. At present, no definitive conclusions
5-118

-------
are possible concerning RF-radiation effects on the BBB at low
power densities (< 10 mW/cm2).
(c)	Pharmacologic effects on CNS. PW radiation appears to have a
potentiating effect on drugs that affect nervous system
function. Whether RF radiation also has an inhibiting effect
on neuropharmacologic drugs is not certain.
(d)	Effect on neurotransmitters. There are no data on this subject
at low power densities (< 10 mW/cm2). However, at high power
densities and SAR's (> 24 W/kg) there appears to be an effect
on the release of neurotransmitters after exposure. Whether
there is an increased or a decreased release depends on the
specific neurotransmitter.
5.4.1 Morphologic Observations
The nature of morphologic changes in the nervous system of exposed
animals depends on the frequency, power density, duration, and modulation
characteristics (e.g., PW or CW) of the radiation. Gordon (1970) and
Tolgskaya and Gordon (1973) report that severe damage to the brain occurs when
2
RF radiation of various frequencies at high power densities (> 40 mW/cm ),
producing measurable AT's, is used. These changes consist of hemorrhages,
edema, and vacuolation of neurons after a 40-min exposure to 3000- or
10,000-MHz PW or CW radiation at 40 to 100 mW/cm^. At 20 mW/cm^ similar but
less severe effects were observed. Further, 3000-MHz radiation produced more
marked changes than 10,000 MHz of equal power density. However, Austin and
Horvath (1954) did not observe similar changes in brains of rats that became
convulsive and hyperthermic (brain temperature ~ 43 °C) due to a single
exposure to 2450-MHz radiation. These authors observed only mild pyknosis and
hyperemia in some areas of the brain. Albert and DeSantis (1975, 1976) did
not observe hemorrhage, gliosis, or focal necrosis in Chinese hamsters exposed
5-119

-------
to 2450-MHz (CW) radiation at 50 mW/cm^ (SAR estimated at 17.5 W/kg) for 30 to
120 min, but they did observe swollen neurons with frothy cytoplasm in the
hypothalamic and thalamic regions of the brain. Such observations were not
seen in the cerebellum, pons, or spinal cord. Histologic change in the brains
of rats have also been reported at low power densities (< 10 mW/cm ) of
3000-MHz microwaves after multiple 30-min exposures. The histological alter-
ations include cytoplasmic vacuolation of neurons, axonal swelling and
beading, and swelling in and decreased numbers of dendritic spines (Tolgskaya
and Gordon 1973; Albert and DeSantis 1975; Gordon 1970). All of the above-
mentioned changes were reversible after 3 to 4 weeks. Baranski (1972b) noted
that exposure of guinea pigs and rabbits to 3000-MHz (PW and CW) radiation at
power densities of 3.5 to 25 mW/cm^ (SAR estimated at 0.4 to 2.5 W/kg)
resulted in myelin degeneration and metabolic alterations in glial cells.
Qualitatively, the morphological effects on the CNS are similar at 10 to
50 mW/cm power densities, but quantitatively they are greater at higher power
densities. Most scientists agree that irradiation at these power densities
can raise the body temperature. Soviet scientists have reported similar
2
morphological changes at 1 mW/cm after chronic irradiation, and they do not
2
consider these alterations at 1 mW/cm of thermogenic origin. Tolgskaya and
Gordon (1973), Baranski (1972b), and Baranski and Edelwejn (1968) further
state that morphological effects are more marked after PW or chronic exposure
than after CW or acute exposure. Most Eastern European studies claim full
recovery by irradiated animals in 1 to 3 weeks after exposure at less than 10
mW/cm . Albert and DeSantis (1975, 1976) found continued existence of
5-120

-------
neuronal cytopathology in animals 2 weeks after exposure. Perhaps a longer
recovery period in the latter study would have shown complete reversibility.
Switzer and Mitchell (1977) described the existence of myelin figures in
brain dendrites of rats exposed to 2450-MHz (CW) fields for 550 h over a
110-day period (average SAR =2.3 W/kg) and allowed 6 weeks for recovery.
These authors did not observe any gliosis, perivascular edema, or synaptic
pathology. However, irradiated rats in this study exhibited marked disruption
of discriminative performances during exposure.
Takashima et al_. (1979) exposed male rabbits to 1- to 30-MHz fields
amplitude modulated at ~ 15 Hz or 60 Hz during 2- to 3-h single exposures or
during 4 to 6 weeks of chronic exposure. The electric field strengths ranged
from 60 to 500 V /m. Acute and chronic EE6 recordings were obtained from
rms
animals under sodium pentobarbital anesthesia. There was no temperature rise
in the exposed animals. The EEG recordings after acute exposures at field
strengths of 60 to 500 Vrms/m or after chronic exposures at strengths up to
70 V /m showed no differences between control and experimental animals,
rms
However, chronic irradiation at higher field strengths was associated with
abnormal patterns. These consisted of bursts of high amplitude spindles at
90 \L /m as well as suppression of activity at 500 V„ /m. All brain activi-
rms	rms
ties returned to normal a few hours after irradiation. The results in this
study appear to be free of electrode artifacts because chronic exposures were
made without electrodes, and all recordings were made when the fields were
switched off. The effects of anesthesia are not clear, and the natural
fluctuations of brain activity also complicate interpretation of the results.
5-121

-------
Although the occurrence of high-amplitude spindles in irradiated animals in
this study is similar to that described by Bawin et afL (1973), it is difficult
to compare the two results because, in the latter study, chronically implanted
electrodes may have interfered with the imposed fields. Takashima et a^. (1979)
confirmed the existence of artifacts during irradiation when implanted electrodes
were used.
We can conclude that RF radiation of centimeter wavelengths causes
morphological changes in the CNS of experimental animals following acute or
2
chronic exposures at low power densities (< 10 mW/cm ). Note that, in most of
the low-exposure studies, total recovery was observed after 3 to 4 weeks.
5.4.2 Blood-Brain Barrier Studies
In the past few years, contradictory reports have been published concern-
ing the effects of RF radiation on the permeability of the BBB. (For reviews,
see Albert 1979a and Justesen 1980.) Some of these reports indicate that direct
RF-radiation effects in experimental animals might result in increased perme-
ability of the BBB (Frey et aL 1975; Oscar and Hawkins 1977; Albert and
DeSantis 1976; Albert 1979b). Other reports indicate that increased perme-
ability might be mediated by hyperthermia induced by intense RF fields (Sutton
and Carroll 1979; Merritt et aL 1978; Lin and Lin 1980). Preston et al.
(1979) and Preston and Prefontaine (1980) reported negative findings at lower
power densities. Some of these discrepancies may be attributed to differences
in techniques employed to assess changes of permeability. These methodologies
consist of gross examination of brain slices, single-passage isotopic tracers,
5-122

-------
fluorometric measurements, arid electron microscopic tracers. All of these
techniques have some inherent shortcomings, either in quantitation or sensi-
tivity. Thus, unless the changes of permeability are diffuse and significant,
positive results may not be readily apparent. Further complications are
reports that average power density, peak power, and pulse width may be
important parameters affecting the BBB permeability (Frey et al_. 1975, Oscar
and Hawkins 1977). Therefore, one must consider the limitations of the tech-
niques and the exposure parameters before reaching conclusions regarding
effects of RF-radiation on the BBB.
Frey et aK (1975) were the first authors in the United States to report
a permeability increase in the BBB of the rat after RF-radiation exposure.
2
They observed that a 30-min 1200-MHz exposure (CW) at 2.4 mW/cm (SAR esti-
mated at 1.0 W/kg) resulted in a statistically significant increase in fluo-
rescein in brain slices of experimental animals over controls. Most of the
fluorescein appeared to be concentrated in the vicinity of the lateral and
third ventricles. Some dye also was detected in the metencephalon. These
authors also reported similar but heightened alterations in the permeability
2
of the BBB when rats were irradiated with PW radiation at 2.1 mW/cm peak
2
and 0.2 mW/cm average power density (SAR estimated at 0.8 W/kg). Their
results also indicated that PW radiation was more effective in altering brain
permeability than CW—once again, the theme of PW radiation producing greater
biological effects than CW surfaces in these studies.
Merritt et jfL (1978) were unable to replicate the fluorescein studies of
Frey et al^. (1975). However, increased brain permeation of fluorescein-albumin
5-123

-------
(mol. wt. 60,000) was produced in rats heated to 40 °C by hot air or by RF
radiation. They concluded that hyperthermia per se, and not field-specific
effects of RF radiation, is the essential determinant of increased perme-
ability. Using sodium fluorescein and Evan's blue, Lin and Lin (1980) also
found no change in BBB permeation after a single 20-min focal exposure within
the rat head at 0.5 to 1000 mW/cm^ (local SAR's ranged from 0.04 to 80 W/kg)
at 2450 MHz (PW). Albert (1977, 1979b) and Albert and Kerns (1981), using
electron microscopic tracer methodology, followed the movement of horseradish
peroxidase (mol. wt. 40,000) in rat and Chinese hamster brains after 2450-MHz
(CW) irradiation in the far field at 10 mW/cm^ (SAR estimated at 0.9 to 2.0 W/kg).
The authors reported focal areas of increased permeability in brains of 35 per-
cent of the irradiated animals in contrast with 10 percent in controls. Areas
with increased permeability were seen with greater frequency in the thalamus,
hypothalamus, medulla, and cerebellum than in the cortex or hippocampus. Using
the same parameters of exposure and species as above, Albert (1979b) and Albert
and Kerns (1981) showed that within 2 to 4 h after irradiation, there was no
evidence of increased BBB permeability, thus demonstrating complete reversi-
bility of the RF-radiation effect. It was also shown that the increased perme-
ability of the BBB appeared to be due to increased pinocytotic transport of
the tracer rather than to opening of the endothelial tight junctions (Albert
and Kerns 1981).
Sutton and Carroll (1979) observed increased permeability of the BBB to
horseradish peroxidase when brain temperatures of rats were elevated to 40
to 45 °C for a few minutes by 2450-MHz (CW) microwaves. A significant
observation was that the integrity of the BBB and survival times at 45 °C
5-124

-------
(cerebral temperature) could be prolonged if core temperature was maintained
at 30 °C. They hypothesized that perfusion of vessels with cool blood had a
protective effect. This study indicates that severe hyperthermia induced by
RF radiation might result in increased permeability of tracer proteins in
rat brains.
In a physiological study using radioisotope tracers, Oscar and Hawkins
(1977) exposed rats to 1.3-GHz (CW or PW) radiation for 20 min. Using the
2
technique of Oldendorf (1970), they found after CW irradiation at 1 mW/cm
(SAR estimated at 0.4 W/kg) that there was a significantly greater uptake of
mannitol (mol. wt. 182) and inulin (mol. wt. 5000), but not dextran (mol. wt.
60,000), in brains of exposed animals. Similar, but greater, uptake of these
2
compounds was observed after PW irradiation (average power density 0.3 mW/cm ;
SAR estimated at 0.1 W/kg) than after CW irradiation. Uptake of mannitol by
the brain was quite dependent upon power density, pulse width, and number of
pulses per second. Merritt et aK (1978), who also used the Oldendorf tech-
nique, reported that there was no significant change in uptake of mannitol or
inulin in rats exposed to RF radiation under conditions similar to those used
by Oscar and Hawkins (1977).
Preston et al_. (1979), who also used the Oldendorf technique, exposed
rats to 2450-MHz (CW) fields for 30 min at 0.1 to 30 mW/cm^ (SAR estimated at
0.02 to 6 W/kg); no change in uptake of mannitol in the brain was found. They
also speculated that the changes reported by Oscar and Hawkins (1977) may have
been due to changes in blood flow. Later, Oscar et a^. (1981) measured the
blood flow in several brain regions during exposure to 2800-MHz (PW) fields
5-125

-------
o
at 15 mW/cm (average) for 5 to 60 min and found increased local blood flow.
They then suggested that previously reported BBB-permeability changes (Oscar
and Hawkins 1977) may be less in magnitude than originally indicated.
A recent report by Preston and Prefontaine (1980) described studies on
BBB permeability in rats exposed to 2450-MHz (CW) radiation both in the near
and the far field. The far-field exposure took place in an anechoic chamber
at 1 or 10 mW/cm^ (SAR estimated at 0.2 to 2.0 W/kg) for 30 min. In the second
study, a microwave applicator was placed on the rat head for a near-field ex-
posure. The exposure consisted of a single 25-min irradiation (SAR's at 0.08
to 1.6 W/kg). In this second study the exposure took place after the tracer
^C-sucrose was injected into the animal so that BBB function during irradia-
tion could be examined. No change in permeation was found in either study.
5.4.3 Pharmacological Effects
RF radiation has been reported to alter effects of drugs that influence
CNS functions. Baranski and Edelwejn (1968) noted altered EEG tracings and
decreased tolerance to Cardiasol, a CNS stimulant, in persons working in micro-
wave fields. To better understand their observations of humans, they conducted
animal experiments. In their investigations on rabbits, these authors found
that administration of Phenactil, a depressant of cortical activity, followed
by 3000-MHz (PW) irradiation at 20 mW/cm^ (SAR estimated at 3.0 W/kg) for
20 min, resulted in desynchronization of the EEG; i.e., the synchronizing of
the EEG by Phenactil was reversed by irradiation, which indicates a stimu-
latory effect by microwaves on the brain-stem reticular formation. Under the
5-126

-------
same exposure conditions, these authors demonstrated that half the dosage of
Cardiasol was required to obtain the same EEG reaction that was observed in
control animals without previous irradiation; the authors concluded that
microwaves potentiate the effects of Cardiasol. Chronic exposure of rabbits
to 3000-MHz (PW) radiation at 7 mW/cm^ (SAR estimated at 1.0 W/kg), 3 h/day
for a total of 70 to 80 h resulted in violent convulsions after injection of
the same dose of Cardiasol (3 mg/kg) as in controls, thus indicating that
tolerance to Cardiasol was decidedly lower in chronically exposed animals than
in controls. Chronic exposure also resulted in desynchronization and high-
amplitude recording potentials in the EEG. In this study, thermal effects of
microwaves were considered unlikely, and the authors suggested that microwaves
may exert a stimulating effect on specific areas of the CNS. It should be
noted that the EEG records were obtained with screw electrodes implanted into
the skull of the rabbit and that the pulse modulation characteristics were not
specified.
Servantie et al_. (1973) investigated the combined effects of microwave
radiation and Pentetrazol, a convulsant drug. Mice were subjected to 3000-MHz
(PW) radiation (maximum peak power = 600 kW and maximum mean power = 350 W),
presumably for a few minutes per day for 8, 15, 20, 27, and 36 days. The mean
2
power density measured in the absence of the animal was 5 mW/cm . At the end
of exposure, 50 mg/kg of Pentetrazol was administered intraperitoneally. They
observed that RF radiation affected the convulsive time and mortality rate.
Specifically, they noted that there was no difference between controls and ex-
perimental animals through 8 days of exposure. After 15 days of irradiation,
the animals exhibited delayed appearance of convulsions and were less
5-127

-------
susceptible to the epileptogenic action of Pentetrazol. However, after 27 to
36 days of exposure, the effect was reversed. The results indicate a biphasic
response of mice over 36 days. Decreased incidence of mortality was observed
in drug-injected animals if they were irradiated more than 8 days. Servantie
et al. (1973) also investigated the effect of curare-like drugs on iji vivo and
in vitro preparations of the rat's sciatic nerve. They found that irradiated
rats were less susceptible to paralyzing drugs. Similar findings were noted
in sciatic-nerve preparations from rats. In addition, sciatic nerves isolated
from irradiated rats were paralyzed to a lesser extent and recovered sooner
than those of control rats. Servantie et aL (1973) suggested that, based on
their observations, the effect of microwaves is more likely at the neuro-
muscular junction (synapse) and less likely due to secondary effects on the
metabolism or binding of the drugs.
Goldstein and Sisko (1974) investigated the effects of pentobarbital and
o
9.3-GHz (CW) microwaves at a power density of 0.7 to 2.8 mW/cm (SAR estimated
at 0.1 to 0.3 W/kg) on the behavior and EEG patterns of rabbits. They reported
that there was no difference in EEG patterns between control (pentobarbital-
injected) rabbits and rabbits treated with pentobarbital plus RF irradiation
(0.7 mW/cm ) during the the first 5 min. However, after a latent period,
cycles of intense arousal and sedation accompanied by other behavioral changes
occurred in animals exposed to RF radiation after the barbiturate injection,
but not in controls. Similar but more pronounced effects were seen at 2
mW/cm (SAR estimated at 0.2 W/kg). In this study, the EEG was recorded with
the aid of implanted stainless steel electrodes.
5-128

-------
2
Recently, Thomas et a_h (1979) reported that acute, low level (av 1 mW/cm )
2450-MHz (PW) radiation in the near field (SAR estimated at 0.2 W/kg) poten-
tiates the behavioral response to chlordiazepoxide (a tranquilizer) in rats.
This was primarily a behavioral study and is discussed in § 5.5, Behavior.
It can be concluded from such pharmacologic studies that low level
microwaves may interact in a specific manner on the nervous system in con-
junction with drugs. Such interactions may prove potentially to be both
useful and harmful as more information becomes available.
5.4.4 Effects on Neurotransmitters
Specific neural systems that contain various neurotransmitters are known
to affect the inhibitory or excitatory states of the brain. The relative
firing rates of these neuronal systems is reflected in the turnover of their
neurotransmitters. Since RF radiation has been reported to stimulate and
depress the CNS, several scientists have investigated the effects of RF
radiation on CNS neurotransmitters.
Snyder (1971) made neurochemical measurements in rats exposed to 3000-MHz
(CW) radiation. He observed that a 1-h exposure at 40 mW/cm (SAR estimated
at 8 W/kg) resulted in a significant increase of 5-hydroxyindolacetic acid
(5-HIAA) and 5-HT (serotonin) in discrete nuclei of the brain. The opposite
effect, that is, reduced levels of 5-HIAA and 5-HT, was found in rats exposed
?
at 10 mW/cm (SAR estimated at 2 W/kg) 8 h/day for 7 days. The body tempera-
2
ture in the rats exposed at 10 mW/cm rose by 1 to 2 °C during irradiation,
5-129

-------
and the animals showed signs of moderate heat stress. To compare the effects
p
of microwave exposure (10 mW/cm ) to those of conventional radiant heat loads,
rats were placed in an incubator maintained at 34 °C by thermostatically con-
trolled incandescent lights for 8 h/day for 7 days. Body temperature was
2
raised to the equivalent of that observed after irradiation at 10 mW/cm . No
difference was found in the turnover rate of norepinephrine or 5-HT or steady
state levels of 5-HIAA between heated and control animals. Snyder concluded
that exposure to RF-radiation produced distinctly different effects on 5-HT
and 5-HIAA in rat brains than conventional heating.
Zeman et aK (1973) investigated the effects of acute and chronic ex-
posure of rats to 2860-MHz (PW, average power 300 W) radiation on rat brain
gamma-aminobutyric acid (GABA). Chronic exposures were conducted for 4 to
2
6 weeks, 4 to 8 h/day, at 10 mW/cm (SAR estimated at 2 W/kg). Acute ex-
2
posures were conducted at 40 to 80 mW/cm power density (SAR estimated at
8 to 16 W/kg) for 5 or 20 min. No significant change in body temperature was
2
noted during chronic exposures at 10 mW/cm , while rectal temperature in-
2
creased by 3 °C after acute exposures at 40 to 80 mW/cm . There was no
significant difference in whole-brain GABA levels between control and
irradiated animals in acute or chronic experiments. These results would be
more meaningful if the investigation had focused on discrete areas of the
brain rich in GABA, rather than whole brains; however, others have used the
procedure of whole-brain assays successfully.
Merritt et al^. (1976) reported a decrease in norepinephrine, dopamine,
and serotonin in discrete areas of rat brains after whole-body exposure for
5-130

-------
2
10 min to 1.6-GHz microwaves at 80 mW/cm (SAR estimated at 24 W/kg). These
exposures raised rectal temperatures by 3.7 °C. Hypothalamic epinephrine
decreased significantly in irradiated and hyperthermic controls, although
serotonin levels decreased in hippocampi of irradiated rats but not in those
of control hyperthermic rats. There were no differences in the serotonin
contents in the hypothalamus and striatum between hyperthermic controls and
irradiated animals. Merritt et ah concluded that hyperthermia was responsible
for the effects on neurotransmitters.
In a separate study, Merritt et al_. (1977) observed a decrease in rat
hypothalamic norepinephrine and dopamine after a 10-min exposure to 1.6-GHz
2	2
microwaves at 20 mW/cm (SAR estimated at 6.0 W/kg), but not at 10 mW/cm .
The higher power density was associated with increased temperature, whereas
the lower was not. Serotonin in the hypothalamus was unaffected by microwaves
even at power densities as high as 80 mW/cm . No comment was made on hippo-
campal levels of serotonin in this study.
The effect of RF radiation on calcium ion efflux from brain and other
tissues is discussed in § 5.7.5, Calcium Ion Efflux.
5.4.5 Unresolved Questions
The effects of low-level exposure to RF radiation on the development of
the CNS need much greater attention. Thus far, only gross brain teratologic
and brain-weight studies have appeared in the literature. Investigations on
subtle effects of RF radiation on CNS development are lacking. The effects
5-131

-------
on the growth and maturation of the nervous system at the cellular and sub-
cellular levels need to be examined. Albert et al_. (1981) reported permanent
loss of cerebellar Purkinje cells in rat pups after exposure of pregnant dams
to 100-MHz (CW, SAR = 2.77 W/kg) and 2450-MHz (CW, SAR = 2 W/kg) radiation.
These authors reported a reversible decrease in cerebellar Purkinje cells if
newborn rat pups were exposed to 2450-MHz radiation.
Synergistic and antagonistic effects of RF radiation with alcohol,
x-rays, and other agents should be investigated, since some findings already
indicate that this area should not be ignored.
Nervous system effects are summarized in Table 5-12. There appear to be
ample data that suggest effects of RF radiation on the nervous system of
animals and humans. On the other hand, data on the BBB are controversial.
Long-term low level studies on the adult nervous system are conspicuously
lacking in the U.S. In addition, information on the effects of chronic, low
level RF-radiation exposure on neurotransmitters is lacking.
In summary, the state of the art does not permit one to assume that ex-
posure to low level RF radiation produces a significant effect on nervous
system morphology, the blood-brain barrier, CNS-active drugs, or neurotrans-
mitters. Since chronic low level studies are lacking in this country, great
care must be taken when making judgments until long-term studies are
completed.
5-132

-------
TABLE 5-12. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON THE NERVOUS SYSTEM*
Effects
Species
Frequency
(MHz)
Exposure Conditions
Intensity
(bM/co2)
Duration
(days x oin)
SAR
(W/kg)
Reference
Desynchronized EEG -	Rabbit 3000 (PW)
Lowered tolerance to CMS	Rabbit 3000 (PW)
stimulating drug
Biphasic effect on the potency Mice	3000 (PW)
of a convulsive drug
Changes in EEG patterns of	Rabbits 9300 (CW)
anesthetized aninals
Potentiation of drug response	Rats	2450 (PW)
Decreased brain NE, DA,	Rats	1600 (CW)
serotonin in hyperthermic
aninals
Decreased hypothalamic NE,	Rats	1600 (CW)
DA, serotonin
No effect on neurotransaltter	Rats	1600 (CW)
levels
No effect on GABA content	Rats	2860 (CW)
20 (av)
7 (av)
0 7-2 8
1.0 (av)
80
20, 80
10
80
40
10
1 x 20
24-26 x 180
8-36 x
Unknown
1x5
1 x 30
1 x 10
1 x 10
1 x 10
1 x 50
1 x 20
24-56 x
240-480
3 0 (est)
1 0 (est)
0 1-0 3 (est)
0 2 (est)
24 (est)
6-24 (est)
3 0 (est)
16 0 (est)
8 0 (est)
2 0 (est)
Edelwejn (1968)
Edelwejn (1968)
Servantle et al. (1973)
Goldstein and Sisko (1974)
Thosas et al (1979)
Merritt et al (1976)
Herritt et al (1977)
Merritt et al (1977)
Zeoan et al. (1973)
(continued)

-------
TABLE 5-12. (continued)
Exposure Conditions
Effects
Species
F requency
(HHz)
Intensity
(¦W/co2)
Duration
(days x mm)
SAR
(V/kg)
Reference
Swollen neurons in specific
brain regions
Chinese
hamsters
2450 (CV)
25-50
25
1 x 30-1440
22 x 840
3 5-17 5 (est)
8 8
Albert and DeSantis (1975)
Swollen neurons in hypo-
thalamus and subthalamus
Chinese
hamsters
1700 (CV)
10, 25
1 x 30-120
5, 12.5 (est)
Albert and DeSantis (1976)
Myelin figures in dendrites
6 weeks post-exposure
Rats
2450 (CV)
23
110 x 300
2 3
Switzer and Mitchell (1977)
Increased permeability of
BOB to fluorescein
Rats
1200 (CV)
1200 (PV)
2.4
0.2 (av)
1 x 30
1 x 30
1 0 (est)
0 08 (est)
Frey et a^ (1975)
Myelin degeneration and
metabolic alterations in
glial cells
Guinea
pigs
Rabbits
3000 (CV)
3000 (CV)
3.5-25
5

0 5-3 5 (est)
0 4-2 8 (est)
Baranski (1972b)
Focal areas of increased
BBB permeability to
peroxidase
Chinese
hamsters
Rats
2450 (CV)
2450 (CV)
10
10
1 x 120-480
1 x 120-480
2.0
0 9
Albert (1979b)
Brain temperature elevation
(40-45 °C); Increased perme-
ability of BBB
Rats
2450 (CV)
80 V
1 x 10-30
"
Sutton and Carroll (1979)
Increased permeability of BBB
(mannitol and inulln)
Rats
1300 (CV)
1300 (PV)
1.0
0 3 (av)
1 x 20
1 x 20
0 4
0 1
Oscar and Hawkins (1977)
en
i
M
co
(continued)

-------
TABLE 5-12. (continued)
Effects
Species
Frequency
(MHz)
Exposure Conditions
Intensity
(nW/ca*)
Duration
(days x Bin)
SAR
(V/kg)
Reference
in
i
Co
-------
5.5 BEHAVIOR
Michael I. Gage
5.5.1 Introduction
Behavior has been defined as "anything an organism does" (Catania 1968),
or as "the actions or reactions of persons or things under specified circum-
stances" (Morris 1976). Behavior may also be defined as the actions of an
organism in relation to its environmental stimuli.
Behavioral effects of RF radiation have been extensively studied for
several reasons. The first is the reports of experiments (not fully docu-
mented with methods and detailed data analysis) from the Soviet Union and
other European countries that behavioral effects of microwaves are seen at
2
relatively low levels, i.e., 500 pW/cm and below at 2375 MHz in rats
(Dumansky and Shandala 1974; Shandala et aK 1977). The second is that
behavior can serve as an index of total organism functioning, especially by
elucidating the integrative status of the nervous system as well as that of
other organ systems of the body. Moreover, behavior can be analyzed in a non
terminal fashion, without resort to surgically or biochemically invasive pre-
paratory techniques.
Behavior may be separated into two major categories: naturalistic and
acquired. Spontaneous or naturally occurring behavior may be innate and is
often species-specific in frequency of occurrence. Examples of naturally
occurring behavior include locomotor activity, eating, drinking, and mating.
5-137
-------
There are two general categories of acquired or learned behavior—respon-
dent and operant, which are distinguished by the procedures used in the
acquisition or conditioning of the behavior. Respondent conditioning occurs
as a consequence of the temporal contiguity between stimuli. Stimuli paired
in time with another stimulus, which reflexively elicits a response, gradually
elicit the response. Examples of acquired respondent behavior are salivation
and hunger pangs when passing a restaurant or eye blinks to acoustic stimuli.
Responses conditioned by respondent procedures are usually measured by their
occurrence and their magnitude.
Much complex behavior of human beings and other higher animals in the
course of daily activities can be viewed as emitted or operant behavior.
Operant conditioning occurs as a consequence of reinforcement that follows
the emission of a response. Reinforcers may be positive (such as food or
water), or negative (such as termination of electric shock or intense radiant
energy). All reinforcers maintain or increase the frequency of response.
By definition, a positive reinforcer is a stimulus that increases or maintains
the probability of emission of an operant by its presentation. A negative
reinforcer is a stimulus that increases or maintains this probability by its
removal after emission of the operant. Reinforcement need not follow the
occurrence of each response. It may be intermittent according to a schedule
(i.e., a schedule of reinforcement). Responses are said to be conditioned
when they have a highly predictable probability of occurrence. This prob-
ability is often expressed by the average occurrence within a period of time,
or the response rate. Examples of operant behavior include eating with a
5-138
-------
knife and fork, driving a car, or writing a review on behavioral effects of
microwaves.
Operant conditioning may be used to answer specific questions about be-
havior that may be similar across species. For example, by reinforcing the
responses in the presence of light of one wavelength but not reinforcing
responses in the presence of light of any other wavelength, animals can learn
to respond selectively in the presence of light of the first wavelength. By
presenting light at wavelengths close to the one that is reinforced, the
threshold for discriminability of colors can be obtained.
Specific types of behaviors investigated in behavioral research are so
numerous that no attempt can be made to describe all of them here. Descrip-
tions will be given of behaviors that have been studied as a function of
microwave exposures. For a more complete description of behavior as studied
by ethologists and psychologists, the reader is referred to several standard
books (Brown 1975; Hinde 1970; Honig and Staddon 1977; Kling and Riggs 1971;
Konorski 1967; and Pavlov 1960).
5.5.2 Summary
Some general statements can be made regarding the effects of RF-radiation
exposure on behavior:
o Some microwave effects have been reported on various kinds of
animal behavior. Although most of the studies have used rats
as subjects, a few have used mice, squirrel monkeys, and rhesus
monkeys.
5-139
-------
• Changes in locomotor behavior have occurred after CW exposures
at an SAR as low as 1.2 W/kg in rats (D1 Andrea et a^. 1979).
Changes in food and water intake or body mass have not been
consistently reported at such levels.
• Decreases in rates of ongoing operant behavior have been seen
during exposures at SAR =2.5 W/kg in rats (de Lorge and Ezell
1980), and cessation of operant behavior has been seen at an
SAR of 10 W/kg in rats (D1 Andrea et al_. 1976).
• Alterations in operant performance measured after exposure was
terminated also occurred with SAR's of 2.5 W/kg or more in rats
(Gage 1979a).
• The threshold for detection of microwaves may be as low as
0.6 W/kg in rats (King et al. 1971). However, it is not cer-
tain that animals avoid or attempt to escape from CW microwaves
except at very high power levels.
• Drug effects on behavior in rats have been augmented after
PW-radiation exposures lasting 0.5 h at average SAR = 0.2 W/kg
(Thomas et aK 1979). Behavioral thermoregulation has been
altered after only several minutes of exposure at SAR =1.0
W/kg in the rat (Stern et al. 1979) and at SAR =1.0 W/kg in
the squirrel monkey (AdaTr and Adams 1980b).
• Although the same behavioral effects during or after microwave
exposure of the same magnitude may not be consistently predict-
able, enough behavioral changes have been reported after simi-
lar exposures to warrant the conclusion that behavior is
disrupted by microwaves with an energy input that approximates
one quarter to one half of the resting metabolic rate of many
animals.
• These behavioral alterations are reversible with time.
5.5.3 Naturalistic Behavior
Spontaneous locomotor behavior has been studied with both acute- and
chronic-exposure regimens in the rat. Hunt et al. (1975) found decreased
initial exploratory activity by male Wistar rats after a 30-min exposure to
2450-MHz fields in a multimodal cavity with power adjusted to produce an SAR
of 6.3 W/kg. The activity of exposed animals returned to control values
5-140
-------
within a 2-h period. The decrease in initial exploratory activity was the
same whether the rat was placed in the apparatus immediately after or 1 h
after exposure was terminated. Core body temperature was increased to 40.3 °C
at termination of exposure but dropped to 37.8 °C, within the normal range,
1 h after exposure. Decreased swimming speed in water at 24 °C was also seen
in rats that were practiced swimmers after 30-min exposures to 2450-MHz fields
at SAR's of either 6.3 or 11 W/kg. The effects were seen only after 1.2 km of
swimming following a 6.3-W/kg exposure; after an 11-W/kg exposure, effects
were seen immediately for the first few meters and again after 0.6 km of
swimming despite the fact that water at 24 °C would reduce persistence of a
microwave-related hyperthermia.
Roberti et ah (1975), on the other hand, did not see changes in spon-
taneous motor activity, as measured in a glass cage, in male Wistar rats after
four different exposure conditions in the far field of an anechoic chamber.
The locomotor activity measured by Roberti et ah may not have required as
much physical effort as continuous swimming. Exposures (estimates of SAR
values based on single animal exposures) were either to 10,700-MHz (CW) fields
at 0.6 to 0.9 mW/cm2 (SAR estimated at 0.15 W/kg) for 185 continuous hours
(24 h daily for 7 2/3 days); to 3000-MHz (CW) fields at 0.5 to 1.0 mW/cm2 (SAR
can be estimated at 0.3 W/kg) for 185 continuous hours; to 3000-MHz (PW)
fields, 769 pulses/s, at 1.5 to 2.0 mW/cm2 (SAR estimated at 0.6 W/kg) for 185
2
continuous hours; or to 3000-MHz (PW) fields, 769 pulses/s at 24 to 26 mW/cm
(SAR estimated at 8.3 W/kg) for 408 continuous hours (24 h daily for 17 days).
After this last exposure condition, no change was also observed in running
speed during forced runway running by the rats.
5-141
-------
Locomotor activity on a small platform was increased (as compared with
five controls) in five female Sprague-Dawley rats during the course of
repeated exposures to 2450-MHz (CW) fields in a multimodal cavity (Mitchell et
al. 1977). Exposures of 5-h duration occurred 5 days weekly for 22 weeks.
The average SAR was determined to be 2.3 W/kg, which is similar to the SAR at
p
power density of 10 mW/cm in a plane-wave environment. The activity in-
creased within the first week of exposure and remained high throughout the
course of the exposure period.
Decreases in activity, measured by visual observation, were reported
during repeated exposures of eight Wistar male rats to 918-MHz (CW) fields in
a circular waveguide (Moe et ah 1976). Each exposure lasted 10 h (from 2200
to 0800) and occurred nightly for 3 weeks (total of 210 h), and the range of
p
whole-body average SAR's was measured as 3.6 to 4.2 W/kg (10 mW/cm average
power density). The decreased activity predominantly occurred shortly after
the microwave field was activated. The exposed rats were stretched out in a
prone position more frequently than control rats in the early morning hours.
In addition, exposed rats were reported to consume less food than did
the controls over the course of the exposure, even though their body mass was
not different from that of controls. A repeat of this experiment where eight
male Wistar rats received an average SAR of 0.9 to 1.0 W/kg (average power
density of 2.5 mW/cm ; daily 10-h exposures lasting for 13 weeks) resulted in
no differences in food consumption or in activity measured during the
eleventh week (Lovely et al. 1977). The two studies indicate that a dose-
rate-related threshold for these effects might be somewhere between 1.0 and
3.6 W/kg. Unfortunately, in the circular waveguide, the power density is
5-142
-------
twice the average on axis and falls off rapidly toward the wall, resulting in
the possibility of fluctuating and uncertain quantitative description of SAR
at specific times during exposure (Guy and Chou 1976).
Effects on spontaneous behavior of rats were reported in two other
chronic exposure experiments. Both experiments used 15 exposed and 15 control
male Long-Evans rats irradiated from 0900 to 1700 (8-h), 5 days weekly for 16
weeks in an anechoic chamber with a central monopole antenna and ground plane.
The rats were adapted to the exposure and testing apparatus for 4 or 8 weeks
before irradiation. In one experiment, exposures were to 2450-MHz (CW) fields
2
at an average SAR of 1.2 W/kg (power density, 5 mW/cm ) (D'Andrea et a2-
1979); and in the other, exposures were to 915-MHz (CW) fields at an average
SAR of 2.5 W/kg (power density, 5 mW/cm^) (D1Andrea et aK 1980).
In the study at 2450 MHz, rats showed decreases in activity as measured
on a stabilimetric platform throughout the course of exposure but increased
running-wheel activity overnight through the course of exposure (this latter
effect was not significant). No significant differences were seen in food and
water intake and in body mass. In the study at 915 MHz, exposed rats showed
increased activity as measured both in the running wheels and on the
stabi1imeter. Again, no significant changes in food and water intake or body
mass were observed.
Rudnev et a^. (1978) reported effects of exposure of 25 male albino rats
to 2375-MHz (CW) fields at 0.5 mW/cm^ (SAR estimated at 0.1 W/kg for indi-
vidual animal exposure) for 7 h daily for 1 month. Open field activity was
5-143
-------
measured in irradiated rats and in 25 controls as the number of squares
crossed in 3 min on 2 successive days. The count on the first day was defined
as exploratory activity, whereas that on the following day was defined as
motor activity. Shock-induced aggression was measured by observing battles
between an exposed and control rat. Maintenance of balance on a rotating
treadmill (dynamic-load endurance) and on an inclined rod (static-load
endurance) was measured, as was the amount of food consumed in 20 min after
23 h of deprivation. Electrodermal skin sensitivity was measured as the
voltage of 100-Hz square-wave electrical stimulus that was needed to elicit
paw withdrawal from the metal bars on the cage floor. Measurements were made
prior to irradiation on the 10th, 20th, and 30th days of exposure, and every
15 days for 3 months after exposure.
A significant reduction in food intake, time on the treadmill, and in
time on the inclined rod was seen by the 10th day of exposure. Exploratory
activity was significantly decreased, and shock sensitivity was increased
after 20 days of exposure. At the end of exposure, exploratory and motor
activity, time on the inclined bar, and shock sensitivity were significantly
decreased. The latency to start eating was increased for 30 days after
exposure ended. Time on the treadmill was reduced for 15 days after the
termination exposure. Exploratory activity was increased throughout a 3-month
postexposure period. Sensitivity to electric shock was reduced significantly
on the 30th and 60th day after exposure ended and was still below control
levels on the 90th day.
5-144
-------
5.5.4 Learned Behavior
5.5.4.1 Respondent Conditioning—
Currently, there appear to be no reports by U.S. investigators of micro-
wave exposure altering respondent behavior, although an experiment using
microwaves as an unconditional stimulus has been published. Some studies
using respondent conditioning techniques that report changes in behavior as a
consequence of microwave exposure appear in the Soviet literature (e.g.,
Dumansky and Shandala 1974; Lobanova 1974). Unfortunately, details regarding
exposure conditions or behavioral methodology and results are too sketchy to
permit inclusion in this review. An attempt was made to use microwave ex-
posure as an unconditional stimulus in one experiment (Bermant et al_. 1979).
A 30-s presentation of a 525-Hz tone that preceded 2450-MHz sinusoidally
modulated microwaves that were either presented for 10 s (SAR = 420 W/kg) or
30 s (SAR = 220 W/kg) or that preceded an electric shock to the tail produced
rises in rectal temperature of 0.7 C, 0.5 C, or 0.37 °C, respectively, during
a base-line period over the course of 10 conditioning sessions. Control
female Sprague-Dawley rats presented with the tone alone showed a decrease of
0.47 °C in rectal temperature during this period. The microwave exposure
itself was designed to produce an increase in rectal temperature of 1.5 °C.
Aside from this study there are no reports of microwaves altering respondently
conditioned behavior in which exposure parameters allow clear determinations
of the SAR.
5-145
-------
5.5.4.2 Operant Conditioning—
There is a large body of literature examining alterations of operant
behavior produced by microwaves for which SAR values have been, or can be,
determined.
Reduction in rates of responding on an operantly conditioned task has
been observed during the course of microwave irradiation. Rats were trained
to lever-press for food pellets on a random-interval, 30-s schedule. After
behavior on this task stabilized, the rats showed a response rate that was in
general linearly uniform and typical of random- or variable-interval schedule-
controlled performance. The rats were then exposed to microwaves under each
of several conditions in sessions lasting 25 min, or until the rat's response
rate fell below one third of its base-line control rate. In an initial experi-
ment (D'Andrea et al_. 1976), six male Long-Evans rats were exposed to 360-,
480-, or 500-MHz (CW) fields in a parallei-piate apparatus at an incident
o
power density of 25 mW/cm (the long axis of the rat was parallel either to
the electric-field vector or to the vector of wave propagation). Responding
was reduced only during exposures to 500-MHz fields when the long axis of the
body was parallel to the electric-field vector. Behavior stopped abruptly
after ~ 11 min of exposure, and upon removal from the apparatus the animals
appeared flaccid, wet, and, according to the authors, heat stressed. At this
exposure the SAR was computed from the measured 0.16 °C/min rise in rectal
temperature to be ~ 10 W/kg. Exposures that produced no change in behavior
had SAR's ranging from 5 to 6 W/kg.
In a second experiment (D1 Andrea et al_. 1977), these results were con-
firmed and extended. Exposures were conducted in a monopole-above-ground
5-146
-------
radiation chamber and lasted for up to 55 min, or until responding on the
random-interval schedule fell below one third of the base-line rate.
Exposures of five male Long-Evans rats for periods up to 55 min to 400-, 500-,
600-, and 700-MHz (CW) fields at 20 mW/cm2 power density and 22 ± 1 °C ambient
temperature yielded a U-shaped function of time to the criterion of reduction
in rate, with the minimum of ~ 23 min at 600 MHz when rectal temperature
increased 0.09 °C/min (SAR estimated at 16.4 W/kg). At the time they stopped
responding all rats appeared heat stressed and were engaged in spreading
saliva on their fur. Six additional rats, exposed at 600-MHz (CW) fields at
2
power densities of 5, 7.5, 10, and 20 mW/cm , showed decreased times to stop
2
their responding, at power densities above 7.5 mW/cm when rectal temperature
2
increased 0.04 °C/min or more. At 10 mW/cm (SAR estimated at 7.5 W/kg) the
rats stopped responding after ~ 45 min, and rectal temperature increased
2
0.04 °C/min. Three rats exposed to PW microwaves with 170 mW/cm peak and
2
5.10 mW/cm average power density showed no change in performance. As in
the earlier experiment of D'Andrea et al_. (1976), response reduction was
abrupt and was correlated with the rectal temperature increase.
A series of experiments by de Lorge (1976, 1979) also examined altera-
tions in operant performance during exposure to microwaves. In most of these
experiments, the schedule of reinforcement was used to measure observing and
detection responses in a vigilance task. Two levers were present in a testing
chamber or in front of a primate restraint chair made of Styrofoam and sheet
plastic for optimal transparency to microwaves. Responses on the right lever,
called observing responses, produced either a low- or a high-frequency acoustic
signal. The high-frequency signal was scheduled to occur on a variable inter-
5-147
-------
val of either 30 or 60 s. If the animal made a response on the left lever
when the high-frequency signal occurred—a detection response—it received a
food pellet as a reinforcer. Observing responses occurred at fairly linear
response rates, a pattern similar to that seen on variable interval schedules
of reinforcement.
In the first experiment (de Lorge 1976) five male rhesus monkeys (Macaca
mulatta) showed no change in behavior on this schedule during 1 or 2 h of
exposure to 2450 MHz in the far field of an anechoic chamber when power
2
densities ranged to 16 mW/cm (SAR estimated at 1.2 W/kg). The field was 100
percent amplitude modulated at 120 Hz. Three of these monkeys showed reduced
2
rates of observing response during 1-h exposure to this field at 72 mW/cm
(SAR estimated at 5.0 W/kg) but not at 32, 42, 52, or (in two monkeys) 62
2 2
mW/cm . At 72 mW/cm , rectal temperatures rose ~ 2 °C, and were still
increasing at the end of the hour-long session. The animals moved more in
their chairs after 20 min, and were observed to take short naps after ~ 30 min
at this power density. When observing responses decreased, reaction time to
respond on the left lever increased.
In the second experiment (de Lorge 1979) four male squirrel monkeys
(Saimeri sciureus) were tested and exposed to 2450-MHz 120-Hz modulated
microwaves under conditions similar to those described above at 10, 20, 30,
40, 50, 60, and 70 mW/cm . No reduction in the mean rate of right-lever
observing responses > 1 standard deviation below the mean of the control rate
was seen at any power density during the 30-min exposures. But brief pauses
in responding were seen at microwave onset and offset at 50 mW/cm and above
5-148
-------
(SAR estimated at 2.75 W/kg). At this power density rectal temperature rose >
1 °C. Three of the monkeys were also given 1-h exposures. Observing response
rate seemed to decrease more, and, according to an impression gained from
performance of one monkey (de Lorge 1979, Figure 6), this decrease began
earlier in sessions at higher power densities. One of the three monkeys
exhibited an increase of response rate as power densities increased. After
microwave exposure, all monkeys also showed decreased responding that was
directly related to the power density. All monkeys showed increase in
frequency of incorrect responding on the left lever for food, which was a
direct function of power density. Reliable changes in behavior were
considered to occur only at power densities of 40 to 50 mW/cm at a time when
rectal temperatures had risen by > 1 °C.
In a third experiment (de Lorge and Ezell 1980) eight male Long-Evans
rats were exposed to 5620-MHz (PW) fields at 662 pulses/s, and then to 1280-MHz
(PW) fields at 370 pulses/s. They were tested on the behavioral task of
vigilance during exposures (SAR's were reported at 0.19 W/kg per mW/cm at
5620 MHz, and 0.25 W/kg per mW/cm^ at 1280 MHz). Rates of observing responses
2
during exposures to 1280-MHz fields decreased somewhat at 10 mW/cm (average
2
power density), and decreased markedly after 15 to 20 min during 15 mW/cm .
At 5620 MHz, rates of observing responses decreased only during exposures at
o
26 mW/cm and above. Behavior decrements occurred at SAR =2.5 W/kg at
1280 MHz but required SAR =4.9 W/kg at 5620 MHz.
In a related experiment (Sanza and deLorge 1977), four male Sprague-
Dawley rats were trained to respond on a fixed-interval 50-s schedule for
5-149
-------
food pellets as reinforcers. With this schedule, the first response emitted
50 or more seconds after arrival of the last food pellet produced another food
pellet. Exposures for 60 min to 2450-MHz, 120-Hz modulated at fields at
2
37.5 mW/cm in the far field of an anechoic chamber produced decreases in
response rates that had a fairly abrupt onset. The response decrements were
seen only in two rats that had high base-line rates and not in two rats with
low response rates. Exposures at 8.8 and 18.4 mW/cm produced no decrements
in performance. All rats spent more time at the wall opposite the food cup
2 2
during 18.4- and 37.5-mW/cm exposures than during sham or 8.8-mW/cm ex-
2
posures. The SAR was not given but is estimated at 3.7 W/kg at 18.4 mW/cm
and 7.5 W/kg at 37.5 mW/cm^.
Responding of three male rhesus monkeys trained to a high level of pro-
ficiency on a visual tracking task was not disrupted by exposures at 10 or
20 mW/cm^ to 1200-MHz (CW) fields (reported SAR estimated at 0.8 and 1.6 W/kg)
during behavior sessions lasting ~ 2 h (Scholl and Allen 1979).
Some other studies have looked at changes in previously learned operant
behavior at the termination of single or multiple exposure periods. Thomas
et al. (1975) trained four male Sprague-Dawley rats to respond on a multiple
schedule. One component was fixed ratio 20 (FR20): every 20th response on
the right lever was reinforced by a food pellet. The other component was a
differential-reinforcement-of-low-rate of 18 s with a limited hold of 6 s
(DRL 18 LH 6): Responses on the left lever separated by at least 18 s, but
by no more than 24 s, were reinforced. Components alternated on an irregular
basis. Only one of the schedules was in effect during a given time period.
5-150
-------
Exposures to microwaves lasted for 30 min, and behavioral testing began 5 to
10 min after exposure. Exposure parameters were as follows: 5, 7, 15, and
20 mW/cm2 at 2450-MHz (CW) fields; 5, 10, 15, and 20 mW/cm2 at 2860-MHz (PW)
2
fields at 500 pulses/s and 1-ps pulse width; and 1, 5, 10, and 15 mW/cm at
9600-MHz (PW) fields at 500 pulses/s and 1-ys pulse width. All rats were
exposed to all parameters while restrained in the far field of an anechoic
chamber. In general, response rates increased on the DRL schedule and de-
creased on the FR schedule. Increased rates on the 0RL schedule were seen
2
following exposures to 5 mW/cm and above at 9600-MHz fields (SAR estimated
at 1.5 W/kg), 7.5 mW/cm2 at 2450-MHz (CW) fields (SAR estimated at 2.0 W/kg),
and 10 mW/cm2 and above at 2860-MHz fields (SAR estimated at 2.7 W/kg). De-
creased response rates on the FR schedule were measured following exposures
2
to 5 mW/cm and above at all frequencies (SAR estimated at 1.5 W/kg for
9600 MHz, and 1.4 W/kg for 2450 and 2860 MHz). Increased response rates
2
during time-out periods between components were seen following 5 mW/cm at
all three frequencies. Time-out responses peaked and then dropped after ex-
posures at higher power densities.
In a second study, Thomas et al_. (1976) trained four male Sprague-Dawley
food-deprived rats on a fixed-consecutive-number-eight (FCN 8) schedule. With this
schedule, at least eight presses had to be made on the right lever before
depression of the left lever would yield a food pellet reinforcer. If the rat
made fewer than eight consecutive responses on the right lever before switch-
ing, the count was restarted at zero. Well trained rats were tested after
2
30-min exposures in the near field at 5, 10, and 15 mW/cm to 2450-MHz (PW)
fields at 500 pulses/s, and 1-ps pulse width. Because exposures were in the
5-151
-------
near field, SAR cannot be precisely estimated but may be assumed at 0.4 W/kg
2
for each mW/cm (Durney et aK 1980). The percentage of eight or more con-
secutive responses on the right lever (reinforced runs) decreased, and the
length of these runs decreased after all exposures. Those decreases were
direct functions of power density. However, the overall rate of responding
and the running rate (i.e., the rate after the first response following rein-
forcement) hardly decreased during these postexposure sessions.
Gage (1979a) reported decreases in response rates of eight adult male
Sprague-Dawley rats trained to alternate between levers either 11 or 33 times
for a food-pellet reinforcer. The decrements occurred in sessions following
o
overnight, 15-h exposures to 2450-MHz (CW) fields at 10, 15, and 20 mW/cm ,
2
but not after exposures at 0.5 and 1.0 mW/cm . Only very small decrements
2
were seen after 55-min exposures at power densities to 30 mW/cm (the SAR
2
measured in rats under similar exposure conditions was 0.3 W/kg per mW/cm ).
Vigilance discrimination was tested immediately after 30 min of irradiation
in 10 well trained, young adult male Wistar rats (Hunt et 1975). Exposures
produced SAR's of 0.0, 6.5, and 11.0 W/kg at 2450 MHz (modulated in a quasi -
sinusoidal fashion at 120 Hz in a multimodal cavity). All rats were exposed at
both SAR values, but the sequence of exposures was varied. The task required
the rats to obtain saccharin-flavored water reinforcers by pressing a bar when
a light flashed, but by not pressing a bar when a sonic stimulus was presented.
One of the two stimuli was presented every 5 s, and the light was presented
12.5 percent of the time. Rats at both SAR's had an increased number of omis-
sion errors (i.e., failures to respond to the light after exposure), but there
5-152
-------
were no increases in errors of comission (i.e., responding wrongly when the noise
was presented). The failure to correctly respond was more frequent at the
start of the session, as well as after exposure at the higher SAR.
Schrot et aK (1980) experimented with three male albino rats trained to
learn a new sequence of pressing three levers for food reinforcers daily in
repeated-acquisition procedure. The rats increased their number of errors and
decreased their rate of sequence completions when tested immediately after
30-min exposures to 2800-MHz (PW) fields (500 pulses/s, 2-(js pulse width)
at 5 and 10 mW/cm average power density. No effects were seen at lower
o
average power densities of 0.25, 0.5, and 1 mW/cm . Peak powers were 0.25,
0.5, 1, 5, and 10 W/cm in these exposures. The rats were exposed in a sleeve
holder with the electric-field vector perpendicular to the long axis of the
animal's body. The reported SAR's based on temperature measurements were 0.7
o
and 1.7 W/kg at 5 and 10 mW/cm , respectively.
Operant behavior has also been examined during or following the course of
chronic microwave exposures in some reports also described in §5.5.3, Natural-
istic Behavior. Mitchell et a^. (1977) measured performance on two schedules
in separate groups of rats pretrained before the start of 22 weeks of 5-h
exposures, 5 days/week, to 2450-MHz fields (average SAR at 2.3 W/kg) in a
multimodal cavity. Five exposed and five control female Sprague-Dawley rats
were tested for 30-min sessions on a schedule of multiple FR5 extinction for
15 s (MULT FR5 EXT 15 s). When a white lamp was on, every fifth response was
reinforced by a food pellet, but during periods when the lamp was off, no
responses were reinforced. Although the control rats had higher response
5-153
-------
rates during the FR component, this difference was not significant. Exposed
rats had higher response rates than controls during the extinction component,
and this difference was statistically significant. The ratio of response rate
during the MULT-FR5 component to response rate during the EXT-15-s component
was significantly different in exposed as compared with control rats: Exposed
rats exhibited a higher ratio, indicating poorer discrimination over the
course of exposure, although they had a value like that of the control rats
before irradiations began. Another schedule, Sidman avoidance, was used to
test escape and avoidance response. Five exposed and four control rats were
trained to postpone an unsignaled 2.0-mA foot shock for 15 s (response-shock
interval) by pressing a lever. During 30-min sessions, the animals received
a shock every 0.5 s (shock-shock interval) until they pressed the lever.
No significant effects of microwave exposures were seen with this schedule,
although improvement in avoidance was seen in all rats both within and over
the course of testing sessions.
Conditioned taste aversion was studied in experiments reporting chronic
2
exposures of rats to 918-MHz fields in circular waveguides at 10 mW/cm
(Moe et a2. 1976) and 2.5 mW/cm^ (Lovely et aH. 1977). Rats were given a
saccharin solution to drink in place of water during the microwave exposure
period. Presumably, if the saccharin were drunk in conjunction with an agent,
such as microwave radiation, which made the rat sick or produced some undesirable
sensations, this connection would be learned and the solution would be avoided
in the future. Preference was measured after exposure by allowing a water-
deprived rat to choose between drinking the saccharin solution and water for
p
20 min. In the experiment at 2.5 mW/cm (Lovely et aK 1977), saccharin
5-154
-------
preference was tested only from the 9th to the 13th week of exposure. No
difference between exposed and control rats was seen in amount of saccharin
2
solution consumed either at 2.5 mW/cm (average measured SAR at 1.0 W/kg) or
o
at 10 mW/cm (average measured SAR at 3.9 W/kg).
Several studies have investigated the ability of animals to detect or to
take behavioral action to minimize, avoid, or escape from microwaves. In an
early paper, King et al_. (1971) showed that three irradiated and three control
male albino rats could respond to 2450-MHz microwaves doubly modulated at 60
and 12 Hz as the conditioned stimulus in a measure of conditional suppression.
In this experiment, rats were reinforced with sugar water on a random interval
schedule for licking at a water tube. At various times during each 2-h
session either a 525-Hz tone or the microwaves were presented for 1 min, and
an unavoidable 0.5-s electrical foot-shock followed. Conditioned suppression
to the tone was reliably indicated by no licks being emitted during the tone.
Microwave dose rates of 0.6, 1.2, 2.4, 4.8, and 6.4 W/kg were substituted for
the tone in some sessions. During irradiation at SAR = 0.6 W/kg one of three
rats suppressed responding, at SAR =1.2 W/kg two rats suppressed, and at SAR
=2.4 W/kg all three suppressed reliably.
Johnson et aK (1976) trained two male Wistar-derived rats to nose poke
in a restraint for food pellets on an FR5 schedule in the presence of an
acoustic-pulse stimulus of 7.5 kHz, 10 pulses/s, 3-ps pulse duration for 3-min
periods, which alternated with 3-min periods of no stimulation during which
nose pokes were not reinforced (extinction). When microwaves at 918 MHz
(10 pulses/s, 10-ps pulse durations) were presented for 30-s intervals during
5-155
-------
an extinction period, or when microwaves were substituted for the auditory
stimulus during reinforced periods, response rates were observed that were
similar to those seen in periods when the acoustic stimulus was present. The
2
energy density per pulse of microwaves was 150 pJ/cm , and the average power
o
density was 15 mW/cm . (The SAR would be near 7.5 W/kg if the rats were in
the far field.)
Detection of microwaves does not imply that there are affective proper-
ties of this stimulus, i.e., that they hurt or feel good. In fact, such
detection may be further evidence of the RF hearing phenomenon (Frey and
Messenger 1973). Indeed, in several experiments in which PW microwaves are
presented during exposure, the alteration in behavior of the exposed animal
might be due to effects of acoustic stimulation by the microwave pulses. Such
experiments should have as a control the presentation of pulsed auditory
stimuli.
Frey et al^. (1975) experimented with female Sprague-Dawley rats exposed
to 1200-MHz (PW) fields at 1000 pulses/s (0.5-s pulse duration) at an average
2
power density of 0.2 mW/cm (SAR estimated at 0.2 W/kg) and a peak power
2
density of 2.1 mW/cm . Six rats spent only 30 percent of a 30-min period of
exposure in an unshielded half of a Styrofoam shuttle box during the last 2 of
4 successive daily exposures. Six other rats exposed to 1200-MHz (CW) fields
2
at 2.4 mW/cm (SAR estimated at 2.2 W/kg) spent 52 percent of the time in the
unshielded half of the box on these days. Six other control rats spent 64
percent of the 30-min period in the unshielded half. Only the rats exposed to
PW microwaves could be said to escape from the stimulus or exhibit a modest
5-156
-------
preference, which would decrease their exposure. A similar finding was also
reported to occur in male rats during exposures to a 1200-MHz (PW) field at
2
100 pulses/s at 0.4 or 0.9 iriW/cm average power densities (SAR estimated at
2
0.4 and 0.81 W/kg, respectively) and at 133 or 300 mW/cm peak power densities
(Frey and Feld 1975). These rats spent an average of only 29 percent of their
time in the unshielded half in seven 90-min sessions, and this side preference
was maintained throughout all 7 days. Sham-irradiated rats spent an average
of 57 percent of their time in the unshielded half of the box.
Two groups of eight male Wistar rats spent more than half of each of nine
weekly hour periods in the side of a shuttlebox when occupancy of that side
kept a PW microwave field turned off (Hjeresen et ah 1979). The 2880-MHz
2
field was pulsed at 100 pulses/s (3.0-|js pulse width) with a 9.5-mW/cm
o
average and 33-mW/cm peak power density (SAR calculated at 2.1 W/kg). The
exposure was in the far field of an anechoic chamber. A group of eight rats
that could not extinguish the field and another group that received no micro-
waves showed no side preferences. Preference for occupancy of the side that
extinguished the field increased across each weekly session. Substitution of
a 37.5-MHz (PW) acoustic stimulus for the microwaves in one session resulted
in the rats' spending most of the session in the side that kept the acoustic
stimulus off. A continuously occurring broadband "pink" noise in the anechoic
chamber prevented appearance of side preference in two other groups. In
addition to confirming that rats avoid or escape from pulsed microwaves, this
experiment suggests that microwaves may be detected as an auditory stimulus.
5-157
-------
Monahan and Ho (1976) showed that male CF1 mice irradiated for 10 or
15 min at 2450-MHz (CW) fields in a waveguide when forward-power levels were
stabilized at 0.4, 0.8, 1.6, 2.4, 3.2, 4.0, or 4.8 W exhibited a decline in
energy absorption rate at 2.4 W and above. This decline usually occurred
within the first 5 min of exposure and stabilized toward the latter part of
the period. Monahan and Ho interpret the results as a behavioral action to
minimize exposure to the microwaves. However, they did not directly measure
any behavior. During the first 5 min of exposure the SAR's measured in this
apparatus ranged from 7.7 to 65.7 W/kg. The lowest SAR at which the mice
clearly altered their rates of energy absorption was 28 W/kg. In a second
study Monahan and Ho (1977) showed that the SAR associated with reduction of
absorption during 20 min of irradiation decreased reliably from 43.6 W/kg when
ambient temperature was 20 6C, to 0.6 W/kg, when ambient temperature was
35 °C.
Videotape observations of rats and mice exposed to 2450-MHz (CW) fields
2
at 15 mW/cm for 1 h in the far field below a radiating horn in an anechoic
chamber did not reveal any preferential behavior that minimized whole-body
absorption rates through parallel orientation to the magnetic- as opposed to
the electric-field vector (Gage et aj. 1979). These observations occurred at
ambient temperatures of 22 or 28 °C while the animal was held in a cuboid or
cylindrical enclosure. Usually, the animals assumed a curled, sleep-like
posture in the hour before the microwaves were turned on and maintained that
posture throughout most of the exposure period; an exception occurred in mice
exposed at 28 °C, in which case they more often assumed positions that were
oriented parallel to both field vectors. The SAR of the rats without any
5-158
-------
2
enclosure was 3.3 W/kg at 15 mW/cm , independent of their orientation. The
SAR of the mice when parallel to the electric-field vector was 12.3 W/kg, and
the SAR when the mice were parallel to the magnetic-field vector was 6.2 W/kg,
without any enclosure. The animals changed positions more frequently when
some difference in SAR existed relative to the position assumed.
In a recent article, Carroll et aK (1980) reported that 20 female
Long-Evans rats did not go to a marked-off floor area in a multimodal cavity
to reduce the intensity of 918-MHz microwaves (modulated at 60 Hz with 3-Hz
mode stirrer modulation) from 60 W/kg to either 40, 30, 20, or 2 W/kg. The
mean number of entries into the marked-off "safe" area and the percentage of
time spent there during 22-min sessions were not different when the microwaves
were on (for five 2-min periods) or off. SAR's > 60 W/kg were associated with
lethality within 8 min of continuous exposure in tests of a separate group of
rats. However, in the same apparatus 10 rats quickly learned to escape from
an 800-pA electric foot shock by going to this marked-off area within a 22-min
session. Shocked rats would remain in this area over 90 percent of the
session time.
Sanza and de Lorge (1977) noted the fact that their rats first tried to
jump out of a testing chamber and then assumed a stationary position after
2
failing to escape during exposures to 2450 MHz at 37.5 mW/cm (SAR =
7.5 W/kg).
5-159
-------
5.5.5 Interactions With Other Environmental Stimuli
Interactions of microwaves with two types of environmental stimuli have
been reported to affect behavior. One of these stimulus types is chemical,
specifically, several commonly used psychoactive drugs; the other is physical,
e.g., ambient temperature during exposure. Response rates of rats performing
on a fixed-interval 1-min schedule of reinforcement beginning 0.5 h after 2.5-
to 20-mg/kg dosages of chlordiazepoxide (Librium) (Thomas et al^ 1979) were
increased over control values. These increases were further augmented if an
2
exposure at an average power density of 1 mW/cm of 2450-MHz (PW) fields (500
2
pulses/s, 2-|js pulse width, 100-mW/cm peak power density) occurred in the
half hour between the injection and behavioral testing. In a second study
(Thomas and Maitland 1979), the dose-response function of rats given
d-amphetamine sulfate was shifted to the left after exposure to microwaves
having the same parameters and conditions of exposure as those mentioned
above. A shift to the left of a dose-response function indicated an increased
potency of the drug. Rats in this second study were reinforced for responses
separated by more than 18 s (DRL 18 s). Drug doses producing increases and
decreases in response rates were lower after microwave exposures than those
after no microwave exposures. Although the rat was in the near field in these
studies, the dose rate measured in a Styrofoam-insulated water model was
0.2 W/kg. Thomas and Maitland (1979) also showed that 0.5-h exposure to
microwaves, at the parameters indicated above, 4 days/week when amphetamine
was not administered, shifted the dose-response function of this drug to the
left on the fifth day when a microwave exposure did not occur. In both of
these studies, exposure to microwaves alone or after saline injection had no
5-160
-------
effect on behavior. The shift to the left of a dose-response function indi-
cated that an exposure to microwaves, which by itself did not affect
behavior, acted synergistically with chlordiazepoxide or amphetamine to
increase sensitivity of the organism to the drug. Similar results have not
been seen with chlorpromazine or with diazepam (Valium), an analog of
chlordiazepoxide, in experiments in the same laboratory (Thomas et al_. 1980).
In a related experiment, Monahan and Henton (1979) showed that chlordiaz-
epoxide and, with less consistency, chlorpromazine and d-amphetamine altered
response rates of mice trained in an operant-conditioning procedure to escape
from or to avoid 2450-MHz (CW) fields at an average dose rate of 46 W/kg.
While this experiment showed that drug effects can interact to alter microwave
exposure effects, the design of the experiment does not allow any conclusion
to be drawn about interactions between drugs and microwaves on performance.
Interactions with ambient temperature and humidity were predicted by
Mumford (1969). Monahan and Ho (1977) showed that reduction in the rate of
energy absorption by mice exposed in a waveguide to 2450-MHz (CW) fields
occurred at a low dose rate of 0.6 W/kg when ambient temperature was 35 °C but
required higher doses at lower ambient temperatures. Although the mice were
not directly observed, the authors presumed the mice reoriented in the micro-
wave field to reduce the amount of absorbed energy.
2
Gage (1979b) showed that overnight exposures at 5, 10, or 15 mW/cm to
p
2450-MHz (CW) fields (SAR =0.2 mW/g per mW/cm ) at an ambient temperature of
28 °C reduced operant response rates of rats measured the morning after ex-
posure was terminated. Similar exposures when ambient temperature was 22 °C
5-161
-------
did not result in reduced response rates except after exposure at the highest
power density.
Two reports have indicated mammals will alter thermoregulatory behavior
2
in the presence of as little as 5 mW/cm of 2450-MHz (CW) microwaves. In one
(Stern et aK 1979), six male rats were trained to press a lever to switch on
an infrared (IR) heat lamp for 2s in a chamber at an ambient temperature of
3.9 to 5.3 °C. Microwaves at power densities as low as 5 mW/cm reduced the
rate of responding for the heat lamp. The reduction in response rate that
occurred shortly after microwave onset was a direct function of power density
2
in the range of 5 to 20 mW/cm and returned to base-line values when micro-
waves were switched off. The rats were exposed in the near field (SAR =0.2
W/kg per mW/cm ).
Adair and Adams (1980b) showed that thermoregulatory behavior of the
squirrel monkey (a new world primate found in equatorial jungle areas) was
significantly altered at power densities as low as 6 mW/cm in the far zone of
a 2450-MHz (CW) field within 10 min of onset of exposure. In this study, the
monkeys were trained to select a preferred ambient air temperature by making
an operant response to obtain 15 s of 55 °C air when the air was otherwise 15
°C or to obtain 15 s of 15 °C air when the air was otherwise 55 °C. Preferred
air temperatures ranged between 35 and 36 °C without microwaves and decreased
as a direct function of the microwave power density during exposure. (The SAR
in this experiment was determined calorimetrically on saline models to be
0.2 W/kg per mW/cm .) As in the experiment by Stern et aJL (1979) with rats,
preferred temperatures returned to base-line levels when the microwave field
was extinguished.
5-162
-------
The above-mentioned studies of behavioral effects caused by microwave
exposure are summarized in Table 5-13.
In conclusion, there is ample evidence to suggest that microwaves alter a
variety of unlearned and learned behaviors occurring during and after ex-
posures. In most cases the behavior change can be described as a reduction in
the level of ongoing activity. However, there are some situations in which
increases in activity have been seen. When measured, the magnitude of
behavioral change seems to be related to the power density or SAR of the ex-
posure. Behavioral changes usually revert to base-line levels after removal
of the microwave field.
5.5.6 Unresolved Questions
Most unresolved questions regarding behavioral effects of microwaves
arise because observed findings have not been verified within the same laboratory
or in other laboratories. Repeating a study involves high cost and the
risk of failure to confirm a finding due to small unnoticed differences, e.g.,
between standard procedures in two laboratories. Possibly for these reasons
verification has often not been attempted. However, verification and sys-
tematic replication would allow determination of the limits of conditions
within which a behavioral effect may be expected, as well as definition
of the range of conditions adequate to observe a threshold of effect.
5-163
-------
TABLE 5-13. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON BEHAVIOR
Exposure Conditions
Effects Species Frequency Intensity Duration SAR* Reference
(Weight, g) (PWz) (riW/ca2) (days x «in) (W/kg)
Decreased exploratory activity Young sale
and swinging speed, AT increase rat
2.5 °C
No effect on spontaneous activity Hale rat
or activity and forced running (160-180)
Increased locosotor activity Feoale
rat (307)
Decreased spontaneous activity and Male rat
food intake (360-410)
No effect on spontaneous activity Male rat
or food intake (316-388)
Decreased activity on stabillaetric Hale rat
platform, no significant increase (350-375)
In wheel running
Increased activity on stabllinetric Hale rat
platfonn and 1n wheel running (350-375)
2450 (PW, 7
¦ultlaodal cavity
120 Hz, AH)
10700 (CW)
3000 (CW)
3000 (PW)
3000 (PW)
2450 (CW, aulti-
¦odal cavity)
918 (CW)
918 (CW)
2450 (CW)
915 (CW)
0.6-0.9
0 5-1.0
1.5-2 0
24-26 (av)
10
10
2.5
5
1 x 30
7.7 x 1440
7 7 x 1440
7 7 x 1440
17 x 1440
110 x 300
21 x 600
91 x 600
80 x 480
80 x 480
6.3
0.2
0.3
0.6
8.3
2 3
3.6
1 0
1 2
2 5
Hunt et al (1975)
Robertl et aK (1975)
Hitchelletal (1977)
Hoe et al (1976)
Lovely et al (1977)
D'Andrea et al^ (1979)
D1Andrea et al (1980)
(continued) -
-------
TABLE 5-13. (continued)
Exposure Conditions
Effects Species Frequency Intensity Duration SAR* Reference
(Weight, g) (MHz) (wU/cw2) (days x mm) (W/kg)
Decreased time on treadmill and Hale rat
inclined rod, decreased exploratory
activity, increased then decreased
shock sensitivity Decreased
activity and shock sensitivity
persisted 90 days after exposure
Rectal temperature rise = 0 37 °C Female
before start of test, AT = 1 5 °C rat
with microwaves
Response decreased during exposure
on random interval schedule (lowest
intensity for effect, AT = 1 8 °C)
Response decreased during exposure
(maximum effect) on random
interval schedule, AT = 1 8 °C
Decreased observing responses
on vigilance task, AT = 2 °C
No effect on observing responses
Decreased observing responses
on vigilance task
No effect on observing responses
Hale rat
(350-380)
Hale rat
(357-382)
Hale rhesus
monkey
(4 kg)
2375 (CW)
2450 (PW, multi-
modal cavity,
60 and 12 Hi AH)
0 5
30 x 420
0 1
Rudnev et al (1978)
500 (CW)
600 (CW)
2450
(120 Hz, AH)
Hale squirrel 2450 (120 Hz,
monkeys AH)
(850-950)
25
10
72
16
50
10 x 0 17
10 x 0 5
1 x 11
1 x 55
1 x 60
1 x 20
1 x 30
1 x 60
1 x 60
420 Bermant et aj^ (1979)
220
10
7 5
5 0
1 1
2 8
0 6-17
0'Andrea et al (1976)
D'Andrea et al (1977)
de Lorge (1976)
de Lorge (1979)
(continued)
-------
TABLE 5-13. (continued)
Exposure Conditions

Effects
Species
(Weight, g)
Frequency
(MHz)
Intensity
(mW/cm2)
Duration
(days x mm)
SAR*
(W/kg)
Reference

Decreased observing responses
on vigilance task
Hale rat
(362-400)
1280
5620
(PW)
(PW)
10
26
1 x 40
1 x 40
2 5
4 9
de Lorge and Ezell
(1980)

Response rate decreased on fixed
interval schedule in rats with
high baseline rates, spending time
away from lever
No effect on response rate
Hale rat
(290-340)
2450
AH)
(120 Hz,
37 5
8 8-18 4
1 x 60
1 x 60
7 5
1 8-3 7
Sanza and de Lorqe
(1977)
VJ1
1
o
No effect on visual tracking task
Response rate decreased on FR and
increased on DRL schedules
Hale rhesus
monkey
(6 2-7 9 kg)
Male rat
(120 days,
150')
1200
2450
2860
9600
(CW)
-------
TABLE 5-13. (continued)
Exposure Conditions
Effects
Species
(Weight, g)
Frequency
(MHz)
Intensity Duration
(mW/cm2) (days x mm)
SAR*
(W/kg)
Reference
Increased response rates in
extinction, decreased stimulus
control, no effect on Sidman
avoidance
No effect on flavor aversion test
No effect on flavor aversion test
Microwaves detected as stimulus
Microwaves detected as stimulus
Spending more time in shielded vs
unshielded compartment
Spending equal time in shielded vs
unshielded compartment
Spending more time in shielded vs
unshielded compartment (occurred
in first of 7 sessions)
Spending more time in unirradiated
compartment
Female
rat
Male rat
(360-410)
Male rat
(316-388)
Male rat
(409-427)
Male rat
(300-350)
Female
rat
Female
rat
Male rat
(250)
Male rat
2450 (CW,
multimodal
cavity)
918 (CW)
918 (CW)
2450 (PW,
120 Hz, AM,
multimodal
cavity)
918 (PW)
1200 (PW)
1200 (CW)
1200 (PW)
2880 (PW)
10
10
2 5
15
0 2
2 4
0 4
9 5
(continued)
110 x 300
21 x 600
91 x 600
1 x 1
1x05
4 x 30
3 x 30
1 x 90
9 x 60
2 3
3 9
1 0
0 6-2 4
7 5
0 2
2 2
0 4
2 1
Mitchell et al^ ( 1977)
Moe et a^ (1976)
Lovely et aj (1977)
King et a| (1971)
Johnson et a^ (1976)
Frey et a^ (1975)
Frey et aj[ (1975)
Frey and Feld (1975)
Hjeresen et a| (1979)
-------
TABLE 5-13. (continued)
Effects
Species
(Weight, g)
Frequency
(MHz)
Exposure Conditions
Intensity
(mW/cm2)
Ouration
(days x mm)
SAR*
(W/kg)
Reference
Decrease in SAR at 24 °C
Decrease in SAR when ambient
temperature increased from 20 °C
to 35 °C
No preferential orientation of rats
or mice in far field of plane wave
Cannot take specific action to
reduce intensity of irradiation
Augmentation of increased
response rates produced by
chlordiazepoxide
Shift to left of dose-response
curve for d-amphetamine in DRL
schedule
No effect on dose response curve
for chlorpromazine or diazepam
Hale mouse
(30-34)
Hale mouse
(30-34)
Hale rat
(200-360)
Hale mouse
(25-33)
Female
rat
(290)
Hale rat
(325-375)
Hale rat
(250-300)
Hale rat
(360-380)
2450 (CW)
2450 (CW)
2450 (CW)
2450 (CW)
918 (PW,
60 Hz, AM,
multimodal
cavity)
2450 (PW)
2450 (PW)
3800 (PW)
15
15
1 x 15
1 x 20
1 x 50
1 x 60
5 x 2
1 x 30
1 x 30
4 x 30
1 x 30
28 Monahan and Ho (1976)
43 6-0 6 Monahan and Ho (1977)
3 3
Gage et a^ (1979)
6 2-12 3
(depending on
orientation)
60 Carroll et al (1980)
0 2
0 2
0 2
0 2
Thomas et al (1979)
Thomas and Maitland
(1979)
Thomas et al (1980)
(continued)
-------
TABLE 5-13. (continued)
Exposure Conditions
Effects
Species
(Weight, g)
Frequency
(MHz)
Intensity
(mW/cm2)
Duration
(days x mm)
SAR*
(W/kg)
Reference
Chlordiazepoxide reduced responses,
decreased avoidance responses, and
increased escape responses to
microwaves
Male mouse
(35-44)
2450 (CW)

1 x 30
46
Monahan and Henton
(1979)
Increased temperature during
exposure caused greater response
rate decreases after exposure
Hale rat
(315-365)
2450 (CW)
10
1 x 930
2 0
Gage (1979b)
Reduced responding for heatlamp
in a cold room
Hale rat
325-450)
2450 (CW)
5
1 x 15
1 0
Stern et a! (1979)
Selection of a lower ambient
air temperature
Squirrel
monkey
(750-1100)
2450 (CW)
6
1 x 10
1 0
Adair and Adams (1980b)
*If measured SAR was not reported, SAR was estimated when possible
-------
There is no unifying hypothesis to explain all the observed behavioral
changes. The research on interaction between microwaves and chemicals has not
been verified in independent laboratories, and it has not been extended to see
if the interaction is limited to particular classes of compounds. Delineation
of differences between effects of single and multiple exposures would help
determine if effects of chronic exposure are qualitatively different from
repeated measurement of effects of each single exposure.
Behavioral experiments have used only a limited sample of microwave
exposure conditions. Exploration of frequency spectrum, modulations, wave-
forms, and interactions between waves of different parameters has hardly
begun. Most behavioral work has used rats as subjects. Animals more like
humans in physical size and shape have only infrequently been studied to help
extrapolate findings to humans. Specifications of conditions of exposure
other than the microwave stimulus such as ambient temperature and humidity
have not been consistently controlled and reported so that their influence on
behavior may be determined. There is no information on exposure effects of
specific or limited areas of the body in comparison to total body exposure to
evaluate the effects of localized energy absorption.
5-170
-------
5.6 SPECIAL SENSES
Joe A. Elder
5.6.1 Cataractogenic Effects
A cataract is an opacity in the crystalline lens of the eye. These lens
defects may be clinically insignificant or may cause partial or total blind-
ness. The following conclusions may be drawn from animal experiments on the
cataractogenic potential of RF radiation (Tables 5-14 and 5-15):
1. High-intensity RF radiation is cataractogenic.
2. For single acute exposures, the threshold intensity for cataract
production exceeds 100 mW/cm2. Multiple exposures at intensities
near threshold values for single acute exposures will result in
lens opacities.
3. The cataractogenic potential of RF radiation varies with frequency;
the most effective frequencies appear to be microwave frequences
in the 1-to 10-GHz range.
4. Similar ocular effects are produced by CW and PW radiation of
the same average intensity.
5. In contrast with the above conclusions, which are based on
acute, near-field exposures to the eye or head, no cataracts
have been reported in unrestrained animals after far-field
exposure at power densities near lethal values.
5.6.1.1 Microwave Radiation in Experimental Cataractogenesis--
The rabbit eye has been the experimental model most often used in animal
studies because of its similarity in size and anatomy to the human eye
(Figure 5-3). Its diameter is ~ 75 percent that of the human eye; the cornea
is as large, and although the lens is thicker, its diameter is the same
5-171
-------
TABLE 5-14. SUMMARY OF STUDIES CONCERNING OCULAR EFFECTS OF NEAR-FIELD EXPOSURES
Exposure Conditions
Effects
Species
Frequency

Intensity
Duration
SAR
Reference

(MHz)

(raW/cm2)
(days x mm)
(W/kg)

Cataract
Rabbit
5500
(CW and
PW)
470-785*
1 x 2-100
300-500t
Birenbaum et al (1969a)
Cataract
Rabbit
800 (CW)

785*
1 x 25
50ot
Birenbaum et a_[ (1969b)

4200
(PW)

785*
1 x 17
5oo;

4600
(PW)

785*
1 x 15
500 +

5200
(PW)

500-785*
1 x 5-12
350-5001

5400
(CW and
PW)
500-785*
1 x 3-4
300-500J

5500
(CW and
PW)
500-785*
1 x 2-3
300i500

6300
(PW)

785*
1x5
500

Cataract
Rabbit
2450
(CW)

180
1 x 240

Carpenter (1979)

120-180
20 x 60

No cataract
Rabbit
2450
(CW)

75
20 x 60

Carpenter (1979)
Cataract
Rabbit
2450
(CW)

150
1 x 100
138*
Guy et al (1975a)
Cataract
Rabbit
2450
(CW)

295
1 x 30

Hagan and Carpenter (1976)

10,000 (CW)

375
1 x 30

No cataract, keratitis
Rabbit
35,000

~ 40
1 x 60
s 175#
Rosenthal et al (1976)
(inflammation of cornea)

107,000

- 40
1 x 60
> 238

iEstimate calculated by dividing the microwave power by the irradiated area (d = 1 27 cm) of the eye
lEstimate based on the assumption that all the incident power was absorbed by the eye (2 g)
^Maximum SAR in the eye
Estimated SAR values for the cornea (See text for discussion of Rosenthal et ^ studies of frequency specificity )
-------
TABLE 5-15. SUMMARY OF STUDIES CONCERNING OCULAR EFFECTS OF FAR-FIELD EXPOSURES
Exposure Conditions

Effects
Species
Frequency
(MHz)
Intensity
(mW/cm2)
Duration
(days x mm)
SAR
(W/kg)
Reference

No ocular effects, including no
lenticular changes
Rabbit
3000 (CW)
100, 200
1 x 15, 30
14, 28*
Appleton et aj (1975)
un
I
Acute ocular changes, e g ,
hyperemia of lids and conjunctiva,
meiosis, anterior chamber flare,
engorgement of iris vessels, and
periorbital cutaneous burns, no
lenticular changes

300, 400,
500
1 x 15
42, 56*
70*

Death

300
500
1 x 30
1 x 15
42*
70*

No cataracts
Rabbit
385 (CW)
385 (CW).
468 (CW)T
60
30
60
10 x 15
10 x 90
10 x 20
48*
24*
8 1
Cogan et a| (1958)

No cataracts
Rabbit
2450 (CW)
10
5 x 480
(x 8-17 weeks)
1 5*
Ferri and Hagan (1976)

No ocular effects
Monkey
(H mulatta)
9310 (PW)
150
30-40 x 294-665**

McAfee et al (1979)
JEstimated average whole-body SAR values (Ourney et 1978, figure 31)
Iwaveguide average whole-body exposure
Total exposure time in minutes for the entire 30- to 40-day experimental period
-------
CILIARY BODY
RETINA
CHOROID
CILIARY BODY
SUSPENSORY
LIGAMENT
OF LENS
SCLERA
CORNEA
LENS
IRIS
TOVEA
OP1 ic
NtRVE
LENS
CORNEA
RIS
ANTERIOR
CHAMBER
POSTERIOR
CHAMBER
OPTIC
NERVE
CENTRAL
ARTERY AND
VEIN OF RETINA
POSTERIOR
CHAMBER
VITREOUS
BODY
VITREOUS BODY CONJUNCTIVE
Figure 5-3. Cross-sectional sketch of the human (left) and the rabbit (right)
eye (from Birenbaum et a^. 1969a, Figure 1).
as that of the human eye. In a typical study (Carpenter et ak 1960b), one eye
of an anesthetized rabbit was exposed to microwave radiation in the near
field, and the nonirradiated eye served as the control. The eyes were then
examined at various intervals using an ophthalmoscope, slit-lamp biomicroscope,
or both instruments. The earliest positive reaction in this type of study,
occurring within 24 to 48 h after a cataractogenic exposure, is the appearance
of one or two narrow translucent or milky bands in the posterior cortex of the
lens, just under the capsule, which extends no further than the lens equator.
These bands can be seen only by slit-lamp examination with an angled beam.
If the ocular injury is minimal, no further change occurs, and the cortical
5-174
-------
banding disappears within a few days. Otherwise, in 2 to 4 days after exposure
small granules, appear in the region of the suture of the posterior lens. If a
more intense reaction occurs, larger numbers of granules appear over a larger
area within the next few days, and small vesicles may develop. These early
changes may develop further and become either well-defined circumscribed or
diffuse cataracts. These lens changes remain as permanent ocular defects. In
general, it has been found that microwave cataracts in rabbits involve only
the posterior cortex of the lens, unless the exposure is so intense that the
opacity extends throughout the lens.
At exposure levels that cause cataracts, other ocular reactions also
occur, but these are transient and differ in severity with the intensity and
duration of the exposure. Examples include swelling and chemosis of bulbar
and palpebral conjuctivae, pupillary constriction, hyperemia of iris and
limbal vessels, and vitreous floaters and filaments (Carpenter 1979).
5.6.1.2 Exposure Threshold Values for Cataracts--
Following the publication of reports demonstrating that microwaves can
cause cataracts in experimental animals (Richardson et al. 1948; Daily et aj.
1950a,b), three laboratories (Williams et a^. 1955; Carpenter et al^. 1960b;
Carpenter and Van Ummersen 1968; and Guy et cH. 1975a) published time vs.
power-density threshold curves for cataract induction in rabbits by a single
exposure to near-field 2.45-GHz radiation. The time vs. power-density thres-
hold curves originally published by the three laboratories are similar in
shape but are quantitatively different. The more recent studies found lower
threshold values than those reported by Williams et al. (1955). This probably
5-175
-------
reflects differences in the irradiation method and in techniques used to
measure power density. In fact, Carpenter (1979) has determined that his
power densities were 50 percent higher than originally published. His
corrected data are plotted in Figure 5-4 along with the results of Williams et
al. (1955) for comparison. Guy et ah (1975a) replicated Carpenter's work for
single acute exposures with essentially the same results, and also quantified
the threshold of cataractogenesis in terms of SAR (see Figure 5-4). For
example, these workers found the cataractogenic threshold for a 100-min exposure,
the longest period of irradiation, to be 150 mW/cm (138 W/kg) peak absorption
(cf. Figure 5-5).
The cumulative effect of microwave radiation on cataractogenesis in the
rabbit has been examined by repeatedly exposing the eye at power densities
below the threshold for single acute exposures (Carpenter 1979). For example,
2
daily 1-h exposures at 180 mW/cm for 13 to 20 days were found to be catar-
actogenic in 8 of 10 animals, whereas single exposures at this power density
2
were not effective. At 150 mW/cm , 4 of 10 rabbits gave a positive response
after 18 to 32 daily exposures. No cataracts were observed after 20 daily 1-h
2
exposures at 75 mW/cm .
If the entire body of an unanesthesized rabbit is exposed at power
densities similar to those that cause cataracts under near-field conditions,
the animal exhibits acute stress; Appleton et ah (1975) reported that rabbits
became heat stressed and struggled out of the field during a 15-min exposure
2
at 100 mW/cm . This report is described in more detail in § 5.6.1.5.
5-176
-------
700
-600
600 -
-500
500-
-400
400-
-300
WILLIAMS ef a/ (1955)
z 300 -
CARPENTER et a! (1960b 1979)
-200
200-
GUY eial (1975a)
-100
100 -
100
80
60
40
20
0
EXPOSURE TIME (min)
Figure 5-4. Time and power-density threshold for cataractogenesis in rabbits
exposed to near-field 2450-MHz radiation; values of maximal SAR
are also given (from Guy et a^. 1975a, Figure 7).
5.6.1.3 Frequency Specificity—
As indicated above, most studies of experimental cataractogenesis were
conducted at 2.45 GHz, but opacities in rabbit eyes have been reported after
near-field exposures at 0.8, 4.2, 4.6, 5.2, 5.4, 5.5, 6.3, and 10 GHz (Birenbaum
et al. 1969a,b; Hagan and Carpenter 1976). In several studies, the cataractogenic
potential of different frequencies was addressed. For example, after Hagan
and Carpenter (1976) determined the relative effects of 2.45- and 10-GHz (CW)
radiation on the rabbit eye, they concluded that the cataractogenic potential
5-177
-------
1 25-i
E 100-
5
E
E 0 75 -
cn 0 50-
<
>
u
cc
RETINA
0 25-
0 00
POSTERIOR
POLE
— ANTERIOR
POLE
CORNEA
DEPTH FROM CORNEAL SURFACE
Figure 5-5.
Distribution of energy absorption rate (W/kg) per mW/cm incident
power density in the rabbit's eye and head exposed to 2450-MHz
radiation (from Guy et al. 1974, Figure 3).
for single acute exposures is greater at the lower frequency. At both
frequencies, the opacities were characteristically located in the posterior
subcapsular cortex of the lens, although the initial appearance and subsequent
development differed. These differences probably reflect differences in the
pattern of absorbed microwave energy in the eye due to the different depths of
penetration of the radiation at these two frequencies.
Guy et ak (1974, 1975a) measured the distribution of absorbed energy in
rabbit eyes exposed to 918- and 2450-MHz radiation and found the patterns to
be significantly different. At 2450 MHz, energy absorption was maximal in the
vitreous body at a point midway between the posterior surface of the lens and
5-178
-------
the retinal surface (Figure 5-5). The locus of peak absorption thus cor-
relates well with the observation of irreversible changes in the posterior
cortical lens. Exposure to 918-MHz fields in a specially devised cavity
resulted in relatively uniform absorption in the eye, but maximal absorption
was only ~ 25 percent of the peak absorption at 2450 MHz (Figure 5-6). There-
fore, one would expect the threshold for cataractogenesis in rabbits exposed
to 918 MHz to be considerably higher than the threshold at 2450 MHz. But more
importantly, at 918 MHz, peak absorption in the rabbit brain was 36 percent
higher than in the eye. It is possible that lens effects or ocular changes
may not occur before more severe damage occurs in other sensitive tissues,
such as the brain, during exposure to 918-MHz fields.
0 30 -i
5
E 0 25*
cc
o
0 70 -
~c
O
>
o
CC
0 15 -
n
HEAD
RETINA
LENS
2 3
DEPTH FROM CORNEAL SURFACE (cm)
Figure 5-6. Distribution of energy absorption rate (W/kg) per mW/cm inci-
dent power density in rabbit's head and eye exposed to 918-MHz
radiation (from Guy et a]_. 1974, Figure 27).
5-179
-------
Both Hagan and Carpenter (1976) and Guy et a^. (1975a) used exposure sys-
tems that applied microwave energy across an air space to the eye, and both
reported lenticular cataracts in the posterior subcapsular cortex. Birenbaum
et al. (1969b) produced cataracts in the anterior cortex of rabbit lens with
an exposure system that applied PW microwaves to the corneal surface. Further-
more, as the frequency decreased from 6.3 to 5.2 to 4.6 to 4.2 GHz, longer
exposure times at a constant field strength were required to produce lens
defects. Under similar experimental conditions, even longer exposure times
were required to induce cataracts at a lower frequency, i.e., 0.8 GHz (CW)
radiation. Furthermore, Birenbaum et £l. (1969a) found no substantial differ-
ence in the cataractogenic threshold values for CW and PW 5.5-GHz radiation
and concluded that the average, not the peak, rate of energy absorption deter-
mines whether lens injury will occur.
Rosenthal et a_h (1976) examined the effects of 35- and 107-GHz radiation
on the rabbit eye. At both frequencies keratitis (inflammation of the cornea)
occurred at lower intensities than required to produce any other demonstrable
ocular effect, such as lens injury (cataract) or iritis. Irradiation at
107 GHz was more effective in producing immediate corneal damage, but this change
was generally gone by the next day. Effects at 35 GHz were persistent, were
almost always present the next day, and were associated with marked injury to
the corneal epithelium. Effects on the cornea correlate well with the pattern
of microwave-energy absorption because most of the energy at these high
frequencies is absorbed in the outer regions of the eye. The earliest stage
of keratitis or minimal corneal stromal injury occurred after a 30-min exposure
at an incident power of 50 mW or after a 60-min exposure at 25 mW. Estimates
5-180
-------
of the rate of energy absorption in the eye at the lower incident power (25
mW) are 35 mW at 35 GHz and 47.5 mW at 107 GHz (Rosenthal et ah 1976); there-
fore, the average SAR's of a rabbit eye weighing 2 g are 17.5 and 23.8 W/kg,
respectively. Since maximal absorption occurs in the outer structures of the
eye, the SAR in the cornea is estimated to exceed the average SAR of the eye
by more than one order of magnitude.
Although the above data cannot be compared directly because of widely
varying experimental procedures, these results indicate that the potential for
cataract induction in rabbits is higher in the frequency range between 1 and
10 GHz than at either lower or higher microwave frequencies. Power densities
that would cause cataracts at frequencies below 1 GHz and above 10 GHz give
rise to insult in other ocular sites and extra-ocular tissues. Rosenthal et
al. (1976) found that 35- and 107-GHz radiation primarily affected the outer
structure of the eye, for example, the cornea; and at 918 MHz, Guy et ah
(1974) showed that maximal energy absorption occurred in the brain, not in the
eye.
The effects of near-field microwave exposures on the rabbit eye have been
summarized by Cleary (1980) as follows:
"The induction of cataracts in experimental animals, principally New
Zealand white rabbits, has been described by a number of investi-
gators using a variety of microwave exposure modalities. Generally
microwave field intensities necessary for cataract induction in the
rabbit are such that acute whole body exposures would be lethal due
to hyperthermia. Most cataract studies have, therefore, employed
focused or near-zone fields which limit exposures to the head or
eyes. Localized thermal trauma is still of such a magnitude to
necessitate the use of general or local anesthesia. Corneal irri-
gation with physiological saline solutions have also been used to
5-181
-------
prevent corneal damage due to tissue dehydration during microwave
exposure Anesthesia and corneal irrigation, as well as air temper-
ature and humidity, may significantly affect the temperature of
ocular structures."
The experimental results strongly suggest that radiation-induced tempera-
ture elevation may be essential for the cataractogenic effect of microwaves.
Additional evidence for this position has been provided by Kramar et ak
(1975), who reported that rabbits kept under general hypothermia during
irradiation at known cataractogenic levels of 2.45-GHz (near-field) radiation
did not develop cataracts. This study is described more fully in § 5.6.2.
5.6.1.4 Far-field Exposure Studies--
In contrast with the acute, near-field exposures that can cause cataracts
and other ocular effects, cataracts have not been produced in rabbits whose
entire bodies were exposed to radiation in the far field (Table 5-15). Appleton
et al. (1975) exposed anesthetized rabbits to far-field radiation at 3000 MHz
2
for 15 or 30 min at 100 or 200 mW/cm . No ocular changes were observed during
or immediately after exposure. Fourteen daily examinations and four weekly
examinations, followed by monthly examinations for 1 year revealed no
2
lenticular changes. During exposure at higher levels (300, 400 or 500 mW/cm
for 15 min) animals exhibited acute ocular changes consisting of hyperemia of
lids and conjunctiva, meiosis, anterior chamber flare, engorgement of iris
vessels, and periorbital cutaneous burns. Subsequent examinations revealed no
morphologic lenticular abnormalities. The authors concluded that "It is
noteworthy that one year after a single microwave exposure, sufficient in
intensity to cause both thermal cutaneous and acute gross ocular effects, no
lens changes or cataracts were observed." It is also noteworthy that these
5-182
-------
power levels and durations were well above the cutaneous sensation level,
because unanesthetized animals became heat stressed and struggled out of the
2 2
field during a 15-min exposure at 100 mW/cm . Exposures at 300 mW/cm for
2
30 min or 500 mW/cm for 15 min were lethal to some of the rabbits.
Ferri and Hagan (1976) exposed unanesthetized rabbits to 2450-MHz (CW)
2
radiation in the far field at 10 mW/cm , 8 h/day, 5 days/week for 8 to 17
weeks. Weekly examinations of the eye showed no abnormal changes during the
study, and no post-irradiation changes were observed during the following 3
months.
Cogan et aK (1958) found no cataracts in rabbits 4 weeks after exposure
2
at 385 MHz; the rabbits were irradiated twice weekly for 5 weeks at 60 mW/cm
2
for 15 min or 30 mW/cm for 90 min. Six weeks after exposure in a waveguide,
2
no cataracts were observed in rabbits irradiated at 468 MHz, 60 mW/cm (SAR =
8.1 W/kg), for 10 days (20 min daily). Although the authors reported that the
exposure levels at both frequencies were near lethal values, the 8.1 W/kg SAR is
considerably lower than the value other investigators have found to be sublethal
to rabbits (see Table 5-15). The environmental conditions (temperature,
airflow, etc.) within the waveguide were not given; therefore, one must assume
that the ambient conditions were significantly different from normal values or
that the measured SAR is in error.
McAfee et aL (1979) trained unfettered monkeys (Macaca mulatta) to
2
expose their faces and eyes to 9.31-GHz (PW) radiation at 150 mW/cm average
power density. Over a period of about three months, the animals were irradiated
5-183
-------
for a total of 294 to 665 min during 30 to 40 daily sessions. No cataracts or
corneal lesions were observed in these monkeys during a 1-year period follow-
ing irradiation.
5.6.2 Unresolved Questions
Exposure of the rabbit eye to microwave radiation at sufficient power
densities and durations causes an immediate increase in intra-ocular temper-
ature, and, after a latent period of a few days, opacities develop in the
posterior subcapsular cortex of the lens (Kramar et a^. 1975). This sequence
of events has led to the assumption that microwave-induced lens opacities are
thermally caused. Several experiments have been designed to test directly the
cause-and-effect relationship between temperature increase and cataract formation.
Kramar et aK (1975) exposed rabbit eyes to cataractogenic levels of microwave
radiation while the animal's body was submerged in cold water. By this means,
microwave-induced intra-ocular temperature was limited to < 41 °C, and no lens
opacities developed. In a later experiment, Kramar et aK (1976) used heated
water to produce ocular and rectal temperatures characteristic of those in
rabbits exposed to a cataractogenic level of microwaves. Although the vicinity
of the lens was heated to temperatures above those known to be associated with
microwave cataracts, no lens opacities were observed; however, the rate of
ocular temperature rise was about one-tenth the rate of increase with micro-
waves. Kramar et aK (1976) concluded that a combination of a sharp temper-
ature gradient and rapid rise in temperature following irradiation may be more
traumatic to the lens than a critical temperature per se.
5-184
-------
Carpenter et ak (1977) reported no cataracts of the posterior cortex of
the lens in rabbits after 6 weeks of treatment in which "the eye was heated at
the same rate, to the same extent, and for an equal period of time as it would
experience during a cataractogenic microwave exposure, the difference being
that the equal heating was provided by other means, namely, direct application
of heat to the surface of the eye." In addition, elevating retrolental and
rectal temperatures to values characteristic of a cataractogenic microwave
exposure, through a combination of restricted body heat loss and irradiation
of one eye to power densities slightly below the cataractogenic threshold,
produced cataracts in only 3 of 10 rabbits. According to their hypothesis, if
cataracts were solely of thermal origin, all animals given the two treatments
should have developed cataracts. Carpenter et a_h (1977), therefore, concluded
that the increase in intra-ocular temperature occurring during microwave
irradiation is not the sole causative factor in microwave cataractogenesis.
The reason for the apparent disagreement between the conclusions of
Carpenter et aK (1977) and Kramar et cH. (1976) probably rests with the
difficulty of duplicating by nonmicrowave heating techniques the temporal and
spatial temperature profiles induced by microwave irradiation of the eye.
Note that the temperature at a single site in the eye was the basis for evidence
of duplicating microwave heating of the eye at the same rate and to the same
extent. Furthermore, even though retrolental and rectal temperatures char-
acteristic of cataractogenic microwave exposures were produced, the overall
effects on the rabbits were more traumatic than a near-field microwave exposure
to one eye. Within 24 h after heated water was applied directly to the eye,
almost all the eyes exhibited hemorrhage into the anterior cavity, so that
5-185
-------
observation of the lens became impossible. Only 21 of 32 rabbits survived the
heat treatment. Of these, five developed lenticular cataracts of the anterior
cortex, and only three developed cataracts of the posterior cortex of the lens
that are characteristic of microwave-induced opacities (Carpenter et al.
1977).
As mentioned above, the difficulty of duplicating by nonmicrowave tech-
niques the time-temperature profile of microwave energy absorption in the eye
is probably responsible for the unsuccessful attempts to prove that an elevation
of temperature is responsible for microwave cataracts (cf. Kramar et ah 1976;
Carpenter et ah 1977). On the other hand, strong evidence for thermalization
being the causative factor in microwave cataracts is provided by the experiment
of Kramar et ah (1975), which showed that cataracts were not produced in
hypothermic rabbits receiving a cataractogenic microwave exposure. At present,
it is generally understood that intense localized exposure of the eye for
p
substantial durations (i.e., 150 mW/cm for 100 min) is necessary to induce
cataracts in laboratory animals, and that such acute exposures cause death by
hyperthermia if the entire animal is irradiated.
5.6.3 Auditory Effects
When the human head is exposed to PW RF radiation, an audible sound
described as a click, buzz, chirp, or a knocking sensation is perceived by some
individuals; the sound appears to originate from within or behind the head.
This auditory phenomenon is called "RF sound" or "RF hearing." Our present
knowledge of RF hearing is summarized below.
5-186
-------
1. The RF sound occurs only upon exposure to PW sources; it appears
to originate from within or near the back of the head regardless
of orientation of the head in the RF field, and the sound varies
with pulse width and pulse repetition rate.
2. The effective frequency range is 216 to 6500 MHz. The effect
has been found to occur at average power densities as low as
0.001 mW/cm2 with peak power densities in the range of 100 to
300 mW/cm2. Effective pulse widths vary from 1 to 1000 ps.
(See Table 5-16.)
3. The ability to perceive RF pulses has been shown to be related
to bone-conduction hearing and to the ability to hear high fre-
quency acoustic waves above 5 to 8 kHz.
4. The available data support the conclusion that the RF auditory
effect is evoked by a mechanism similar to that for conventional
acoustic stimuli and that the primary site of interaction is
peripheral to the cochlea.
5. The most generally accepted mechanism responsible for the RF
auditory sensation is thermoelastic expansion. That is, the
absorption of the energy in a brief RF pulse causes a small
temperature rise (~ 10"6 °C) in a short time (~ 10 ps), which
results in thermoelastic expansion of matter within the head,
which then launches an acoustic pressure wave that is detected
by the hair cells in the cochlea via bone conduction.
5.6.3.1 Human Perception of Pulsed RF Radiation--
Frey (1961) was the first to study systematically the human auditory
response to pulse-modulated radiation. The subjects, who were more than 30 m
from an enclosed antenna, reported hearing a transient buzzing sound upon ex-
posure to the intermittent rotating beam. The apparent location of the sound,
which was described as a short distance behind the head, was the same no
matter how the people were oriented in the RF field. When an RF shield
(aluminum flyscreen) was placed between the subject and the RF source, the
subject did not perceive RF sounds (Frey 1973). When earplugs were used, a
reduction in the ambient noise level and an increase in the RF sound level
were reported. The sensitive area for detecting RF sounds was later described
5-187
-------
TABLE 5-16. SUMMARY OF STUDIES CONCERNING AUDITORY EFFECTS OF RF RADIATION IN HUMANS
Exposure Conditions
Effect
Comment
Number
of
Subjects
Frequency
(MHz)
Pul se
Repetition
Rate (s*1)
Pulse
Width
(MS)
Peak
Intensity
(mW/cm2)
Average
Intensity
(mW/cm2)
Energy/
Pul se
(J/cm2)
Noise Level
(dB)
References
RF hearing
"distinct"
cltcks"
Threshold
values
8
3000
0 5
5
10
15
2500
225-2000
300-1000
0 006
0 001-0 01
0 002-0 007
12 5
2 3-20 0
4 5-15 0
45 (+ plastic
foam ear
muffs)
Cain and Rissmann (1978)
RF hearing
buzz heard
at PRR > 100,
individual
pulses heard
at PRR < 100

3
3000
6500
<100-1000
<100-1000
1-2
1-2
2500-50,000
2500-50,000
5
5
40

Constant (1967)
No auditory
response

9500

0 5

Constant (1967)
RF hearing
"buzzing sound"

4
1245
1245
50
50
10
70
370
90
0 19
0 32

Frey and Messenger
(1973)
RF hearing
"buzz,
clicking, hiss,
or knocking"
Threshold
values
Not
given
216
425
425
425
425
27
27
27
27
125
250
500
1000
670
263
271
229
254
4 0
1 0
1 9
3 2
7 1

70-90 (+ ear
stopples)
Frey (1962, 1963)
No auditory
response

8900
400
2 5
25,000
25

70-90 (+ ear
stopples)
Frey (1962)
SF hearing
"buzzing sound"
Threshold
values
8
7
1310
2982
244
400
6
1
267
5000
0 4
2

70-80
(+ ear plugs)
Frey (1961)
RF hearing
"clicks, chirps"
Threshold
values
1
2450
3
1-32
1250-40,000
0 1
40*
45
Guy et a! (1975)
RF hearing
Polytonal
sound
18
800
1000-1200
10-30
> 500

40 ( + ear
(stopples)
Tyazhelov et al (1979b)
Calculated peak-absorbed-energy-denslty per pulse is 16 mJ/Kg
-------
as a region over the temporal lobe of the brain, because the placement of a
small piece of metal screen (5x5 cm) over this area completely stopped the
RF sound (Frey 1962).
Guy et aK (1975b) described the effect of PW radiation on two of the
coinvestigators. Three pulses, 100 ms apart, were presented each second to
keep the average power density below 1 mW/cm . Each individual pulse was
heard as a distinct and separate click, and short pulse trains were heard as
chirps with the tone corresponding to the pulse repetition rate (PRR). The RF
sound appeared to originate from within or near the back of the head. This
report also included the note that transmitted digital codes could be
accurately interpreted by the subject when the pulse generator was keyed
manually. Guy et aK (1975b) also reported that the threshold for RF hearing
was lower when ear plugs were used.
In a study by Constant (1967), the RF sound was described as being in the
area of the ear on the side opposite to the one that was irradiated. All
three of his subjects readily detected 2-ps pulses, whereas 0.5-ps pulses
were not perceived. All three experienced a buzzing sensation at PRR's
greater than 100/s, whereas individual pulses were heard when subjects were
exposed to PRR's below 100/s.
Five of eight human subjects reported hearing distinct clicks either
inside the head or behind the head when exposed to 15-ps pulses (Cain and
Rissmann 1978). The remaining three people heard faint clicks when the pulse
width was increased to 20 ps.
5-189
-------
5.6.3.2 Effective Radiation Parameters—
In the initial report by Frey (1961), human subjects perceived PW
radiation at frequencies of 1310 and 2982 MHz. Although the peak of power-
density thresholds for RF hearing was 266 mW/cm^ for 1310-MHz and 5000 mW/cm^
2
for 2982-MHz fields, the average power density thresholds were 0.4 mW/cm and
2
2 mW/cm , respectively. When ear plugs were used to attenuate the ambient
noise level of 70 to 80 dB, the subjects reported an increase in the RF sound
levels.
In the following year, Frey (1962) reported that humans could perceive PW
2
radiation at 425 MHz with an average power density threshold of 1 mW/cm ; the
2
peak of power density was 263 mW/cm . A frequency of 8900 MHz was not effective
2
even at an average power density of 25 mW/cm ; the peak of power density was
2
25,000 mW/cm . At 216 MHz, the lowest effective frequency reported in the
2
literature, the average power density threshold was 4 mW/cm ; the peak of
2
power density was 670 mW/cm (Frey 1963). Later, in a study of four subjects
exposed to 1245-MHz PW radiation, Frey and Messenger (1973) concluded that
perception was dependent primarily upon peak power and secondarily upon pulse
width. They reported that the peak power for perception was somewhat less
2
than 80 mW/cm for pulse widths of 10 to 70 ps.
In the study by Constant (1967), three human observers were exposed to PW
2
radiation at 3, 6.5, and 9.5 GHz at an average power density of 5 mW/cm
(pulse width was 0.5 to 2.0 ps; PRR was up to 1000/s). Only two of the three
observers perceived 3- and 6.5-GHz radiation; none experienced a response to
5-190
-------
the highest frequency. Cain and Rissmann (1978) reported that all eight of
their subjects heard RF pulses at 3 GHz. In this study, plastic foam earmuffs
were worn to attenuate the ambient noise, which was 45 dB. The average
2
threshold energy density per pulse was 10.6 pJ/cm (range of values was 3.4 to
17.5 pJ/cm2).
The range of microwave pulse widths varied from 1 to 32 ps in the study
by Guy et al^. (1975b) on one human subject. The results indicate that regard-
less of the peak power of the pulse or the pulse width, the threshold for RF
2
hearing of 2450-MHz radiation was related to an energy density of 40 pJ/cm
per pulse, or energy absorption per pulse of 16 pJ/g, as calculated with the
aid of a spherical model. The background noise of the exposure chamber was
2
45 dB. When earplugs were used, the threshold level decreased to 28 pJ/cm .
The threshold for a second subject, who had a hearing deficit, was approxi-
2
mately 135 pJ/cm . Guy et ah (1975b) stated that two pulses which occurred
within several hundred microseconds of each other were perceived as a single
pulse with energy equal to the sum of the two pulses.
The human studies cited above indicate that the highest effective fre-
quency of RF hearing is between 6.5 and 8.9 GHz and that the lowest effective
frequency is 216 MHz. Also, the results describe other radiation parameters
(peak power density, energy density per pulse, and pulse width) that are
important in determining the threshold for RF hearing in humans.
5-191
-------
5.6.3.3 Dependence of RF Hearing on Acoustic Hearing--
Standard audiograms measure hearing thresholds for air conduction at
acoustic frequencies of 250 to 8000 Hz and for bone conduction to 4000 Hz.
Cain and Rissman (1978) measured the hearing ability of eight subjects over
the frequency range of 1 to 20 kHz in addition to determining their standard
audiograms. They found that, although there was no apparent correlation
between the ability to perceive pulsed microwaves at 3000 MHz and hearing
ability as measured by standard audiograms, there was a strong correlation
between microwave-hearing threshold and hearing thresholds to air-conducted
acoustic signals above 8 kHz. For example, three of the subjects who had
normal hearing below 4 kHz could not hear microwave pulses of less than 20-ps
duration under conditions in which the other subjects could perceive RF sounds.
All three had a hearing deficit at frequencies above 8 kHz.
Frey (1961) compared human acoustical hearing and RF hearing and reported
that a necessary condition for perceiving the RF sound was the ability to hear
audiofrequencies above approximately 5 kHz, although not necessarily by air
conduction. This conclusion was based on one subject who had normal air-
conduction hearing but failed to hear the microwave pulses. The person was
subsequently found to have a substantial loss in bone-conduction hearing. On
the other hand, a subject with good bone-conduction hearing but with poor
air-conduction hearing perceived the RF sound at approximately the same power
density that induced threshold perception in subjects with normal hearing.
The studies by Cain and Rissmann (1978) and by Frey (1961) show RF hearing to
be dependent on high-frequency hearing above 8 kHz and bone-conduction hearing
at lower acoustic frequencies.
5-192
-------
5.6.3.4 Similarity of Auditory Response to Microwave and Conventional Acoustic
Stimuli--
Taylor and Ashleman (1974) measured the electrical response in three
successive levels of the cat auditory nervous system (eighth cranial nerve,
medial geniculate nucleus, and primary auditory cortex) to both acoustic and
pulsed-microwave (2450-MHz) stimuli. They concluded that the microwave-
induced auditory effect on the animal is exerted similarly to that of con-
ventional acoustic stimuli. Furthermore, these authors reported that
inactivation (perforation of the round window and aspiration of perilymph) of
the cochlea, the known first stage of transduction for acoustic stimuli,
affected the central nervous system (CNS) response to acoustic and microwave
energy in the same way; i.e., the evoked electrical activities of all three
sites were abolished by cochlear destruction. These results indicated that
the locus of the initial interaction of pulse-modulated microwave energy with
the auditory system might reside peripheral to the cochlea.
In an experiment in which the thresholds of evoked electrical responses
from the medial-geniculate body in cats were determined as a function of back-
ground noise, Guy et aK (1975b) found that as the noise level (50 to 15,000-Hz
bandwidth) increased from 60 to 80 dB, there was only a negligible increase in
the threshold for the 2450-MHz microwave stimuli, a moderate increase in the
threshold for a piezoelectric bone-conduction source, and a large increase in
the threshold for loudspeaker-produced stimuli. The finding that the evoked
response to microwave stimuli did not increase in relation to background
5-193
-------
noise, which included acoustic frequencies to 15,000 Hz, indicates that
microwaves may interact with the high-frequency portion of the auditory
system.
Guy et tH. (1975b) also demonstrated that potentials evoked by microwave
stimuli could be recorded at CNS sites other than those that correspond to the
auditory nervous system. This finding indicates that elicited potentials
recorded from any CNS location could be misinterpreted as indicating a direct
microwave interaction with the particular system in which the recording is
made.
Prior to 1970, Frey (1962) had suggested that RF hearing might be a
result of direct cortical or neural stimulation. He based this suggestion on
(1) his observations that the perception of RF pulses was instantaneous and
occurred at low average-incident-power densities and on (2) the failure to
record cochlear microphonics at power densities much, higher than those
required to elicit auditory nerve responses. Cochlear microphonics are
electrical potentials that mimic the sonic waveforms of acoustic stimuli; they
are the signature of mechanical distortion of cochlear hair cells, the first
stage of sound transduction. The failure to observe microwave-induced
cochlear microphonics had led to the suggestion that pulsed microwaves, unlike
conventional acoustic stimuli, may not act on any sensor prior to acting
directly on the inner ear apparatus (Frey 1967).
5-194
-------
In 1975, Chou et a_L reported their success in overcoming the technical
problems that had prevented investigators from recording cochlear microphonics
from microwave-irradiated animals. The cochlear microphonics of guinea pigs
exposed to 918-MHz (PW) radiation were found to be similar to those evoked by
acoustic stimuli.
The results of the above studies of evoked electrical potentials in the
auditory system, including the demonstration of pulsed-microwave-evoked
cochlear microphonics, strongly indicate that the microwave-induced auditory
sensation is detected similarly to conventional sound detection and that the
site of conversion from microwave to acoustic energy resides peripheral to the
cochlea. However, it is not known what structure(s) in the head transduce(s)
the microwave energy to acoustic energy.
5.6.3.5 Mechanism of RF Hearing—
As mentioned above, Frey (1967) had suggested that RF hearing might be a
result of direct cortical or neural stimulation because of the failure to
record cochlear microphonics and because the perception of RF pulses was
instantaneous and occurred at low average-incident-power densities. The
latter points were evidence against a radiation-pressure/bone-conduction
hypothesis (see also Guy et aK 1975b). Sommer and von Gierke (1964) had
suggested that radiation pressure exerted by the RF pulse impinging on the
surface of the head could launch an acoustic signal of sufficient amplitude to
be detected by the inner ear via bone conduction. Other types of pressure
much greater than radiation pressure can be produced in tissue exposed to RF
5-195
-------
pulses. These include thermal expansion forces that are proportional to the
square of the electric field in the material. For example, Gournay (1966) has
shown that pressures greatly exceeding radiation pressure result when visible
light from a laser is converted to acoustic energy by thermal expansion due to
absorbed energy in various liquids.
Foster and Finch (1974) extended Gournay1s analysis to a physiological
solution exposed to microwave pulses similar to those that produce RF hearing
in humans. They showed both theoretically and experimentally that radiation-
induced pressure changes would result from the absorption of RF pulses and
could produce significant acoustic energy in the solution. In fact, audible
sounds were produced by rapid thermal expansion, resulting from only a
5 x 10 k °C temperature rise in the physiological solution, due to the
absorption of the energy in the RF pulse. The authors concluded that the RF
sounds involve perception, via bone conduction, of the thermally generated
sound transients caused by the absorption of microwave pulses. The pulses
must be moderately intense (typically 500 to 5000 mW/cm at the surface of the
head). However, they can be sufficiently brief (< 50 ps) such that the
maximum increase in tissue temperature after each pulse is very small
(< 10~5 °C).
The hypothesis of Foster and Finch (1974) predicts that the RF hearing
effect is related to thermoelastically induced mechanical vibrations in the
head. Vibrations of this type can be produced by other means such as a laser
5-196
-------
pulse or by a pulsed piezoelectric crystal in contact with the skull (Chou et
al. 1976). Frey and Coren (1979) used a holographic technique to test whether
the skull and the tissues of the head of an animal have the predicted vibra-
tions when exposed to a pulsed RF field. No displacements were recorded, but,
subsequent to this report, Chou et a2- (1980) demonstrated that the sensi-
tivity of the holographic technique used by Frey and Coren (1979) was three to
four orders of magnitude too low to detect displacements related to vibrations
from microwave-induced thermoelastic expansion in biological tissues.
Tyazhelov et al_. (1979b) conducted a series of psychophysical experiments
with 18 subjects to evaluate the adequacy of the thermoelastic hypothesis and
to study the perceptual qualities of RF-induced sounds. Audiofrequency sig-
nals were presented alternately to or concurrently with microwave pulses (see
Table 5-16) under conditions in which the subject could adjust the amplitude,
frequency, and phase of the audio signal. The authors concluded that the
thermoelastic hypothesis adequately explained some of their findings for
microwave pulses of high peak power and short width (< 50 ps), but other
results were interpreted as inconsistent with a thermoelastic mechanism for RF
hearing. For example, pulse widths greater than 50 ps, which increased the
mean power level, produced increases in loudness that rose more rapidly than
predicted by the thermoacoustic model. In addition, suppression of RF sounds
by the audio signal was reported to be inconsistent with the model.
5-197
-------
Lebovitz and Seaman (1977) compared the response to pulsed microwaves
(915 MHz) with the response to acoustic clicks by monitoring single auditory
neurons in the cat. A response of these neurons to pulsed microwaves was
predicted by earlier studies that demonstrated subjective auditory perception
(Frey 1962), auditory evoked potentials (Taylor and Ashleman 1974), and
cochlear microphonics (Chou et al^. 1975). Furthermore, the thermoelastic
model predicts that a mechanical wave of pressure stimulates the inner ear via
bone conduction. Thus, the response of the neurons in the auditory pathway to
pulsed microwaves should be similar to their response to transient mechanical
stimuli such as acoustic clicks. The results indicated that mechanical factors
within the cochlea are similarly involved in determining both the acoustic and
the pulsed microwave response (Lebovitz and Seaman 1977).
Other data in this report (Lebovitz and Seaman 1977) appeared to be
inconsistent with the thermoelastic model that predicts a high-frequency
component such as the microwave-induced cochlear microphonic recorded by Chou
et ak (1975). That is, Lebovitz and Seaman (1977) observed a decrease in
sensitivity of high-frequency auditory units to microwave pulses. However,
they used long pulses of 250 to 300 ps in duration to obtain maximal energy
per pulse. More recently, Tyazhelov et al^ (1979b) reported that long pulses
(~ 100 ps) result in a lower pitch of the RF sound in humans. Two of their
observers who had a high-frequency auditory limit of 10 kHz could not hear
short RF pulses but could hear long pulses. Thus, the results of single unit
recordings in cats are consistent with human perception of RF pulses when the
pulse widths are long.
5-198
-------
Lin (1977) developed a theoretical model that estimates the characteristics
of acoustic signals induced in laboratory animals and humans by microwave
pulses; his model is based on thermal expansion in spherical heads irradiated
by pulsed microwave energy. The frequency of the induced sound was found to
be a function of head size and of acoustic properties for brain tissues;
hence, the acoustic pitch perceived by a given subject will be the same
regardless of the RF frequency. The calculations of Lin show that the
fundamental frequency predicted by the model varies inversely with the radius
of the head; i.e., the larger the radius, the lower the frequency of the
perceived RF sound. He estimated the fundamental frequency of vibration in
guinea pigs, cats, and adult humans to be 45, 38, and 13 kHz, respectively.
The frequency for an infant head was estimated to be about 18 kHz. These
results appear to be in good agreement with the measurements of Chou et al.
(1975), who found cochlear microphonics in guinea pigs to be 50 kHz. In
addition, the calculations provide further evidence that a necessary condition
for auditory perception by adult humans is the ability to hear sound above 5
to 8 kHz (Frey 1961, Rissmann and Cain 1975). However, the prediction that the
pitch perceived by a given subject will be the same regardless of the RF
frequency is not in agreement with the experimental finding that long pulses
result in a lower pitch of the RF sound in humans (Tyazhelov et al_. 1979b).
Other unresolved questions are discussed below.
5.6.4 Unresolved Questions
Several investigators have tried to determine the thresholds for the
RF-induced auditory sensation in human beings and in laboratory animals
5-199
-------
(Table 5-17). Human studies and animal experiments, in general, are few and
have used small sample sizes, different frequencies, and different experimental
procedures. However, the radiation parameters that are most important in
determining the threshold for the auditory response in humans (discussed in
§ 5.6.3.2, Effective Radiation Parameters) and in laboratory animals (discussed
below) are being characterized.
The threshold for an auditory response in cats exposed to 918- and 2450-MHz
(PW) radiation was studied by recording electric potentials from the medial
geniculate body (Guy et ah 1975b). As the pulse width increased from 0.5 to
32 |js, the threshold value for the peak of power density decreased proportion-
ately for pulse widths below 10 |js. The thresholds of average power density
and energy density per pulse also increased with pulse width, but these para-
meters did not show the strong proportional relationship with pulse width as
did the peak of power density. Guy et al^. (1975b) concluded that the thresh-
old for the evoked auditory response was related to the incident energy density
per pulse, at least for pulse widths less than 10 jjs. The threshold energy
density per pulse was found to be about one-half of that which had produced a
sensation in one human subject. Later, one of the coauthors (Lin 1978) stated
that one cannot easily rule out the possible connection between the pulsed-
microwave-evoked auditory responses and other radiation parameters, including
peak incident power density, absorbed energy, and pulse width.
In a similar experiment conducted at frequencies between 8670 and 9160
MHz, Guy et sH. (1975b) found that the threshold values of power density and
of energy density per pulse, which include the auditory response of cats, were
5-200
-------
TABLE 5-17. SUMMARY OF STUDIES CONCERNING THRESHOLD VALUES FOR
AUDITORY-EVOKED POTENTIALS IN LABORATORY ANIMALS
Exposure Conditions
Effect
Species
(n)
Frequency
(MHz)
Pulse
Repetition
Rate
(s-1)
Pulse Width
(ps)
Pulse
Intensity
(raW/cio2)
Average
Intensity
(mW/craz)
Incident
Energy Density
Per Pulse
(pj/cm2)
Peak Absorbed
Energy Density
per Pulse
(pj/g)
References
Response obtained
with scalp
electrodes
Cat (2)
3000
0 5
5
10
15
2200,2800
1300
580

11,14
13
8 7

Cain and Rissmann (1978)
Response obtained
from round window
with carbon lead
Guinea"
pig (5)
918
100
1-10
A
A

20
Chou et a| (1975)
Response obtained
from medial
geniculate with
glass electrode
Cat (2)
918
2450
8670-9160
1
1
1
3-32
0 5-32
32
800-5800
600-35600
14800-38800
0 017-0 028
0 015-0 047
0 472-1 24
17 4-28 3
15 2-47 0
472-1240
12 3-20 0
8 7-26 7
Guy et a^ (1975b)
Response obtained Cat
from individual (not
auditory neurons given)
with glass electrode
915
< 10
25-250

<10

4-40
Lebovitz and Seaman
(1977)
'Direct comparison of power density in the waveguide exposure system to free-field power density is improper because the efficiency of energy coupling is
ten times higher than for free-field exposure (See Chou et al[ 1975, p 362 )
-------
an order of magnitude higher than those required at 918 and 2450 MHz
(Table 5-17). Furthermore, no auditory response was obtained at 8670 to
9160 MHz until the brain was exposed by enlarging the hole in the skull that served
as the electrode access port.
Cain and Rissmann (1978) determined the threshold for auditory responses
in animals exposed to 3000-MHz (PW) radiation. Although their results are in
general agreement with Guy et a|. (1975b) in that the threshold energy density
per pulse was relatively constant for pulse widths of 5, 10, and 15 ps, their
results are confounded by the use of scalp electrodes and only few animals of
three species (two cats, two chinchillas, and a dog).
Lin (1978) concluded that the studies on microwave-induced auditory sen-
sations in laboratory animals strongly indicate a threshold, but the exact
numerical value must await further experimentation. Furthermore, if the
available threshold data are analyzed as a function of microwave frequency, it
becomes clear, both in terms of the peak of power density and peak rate of
energy absorption, that the threshold differs for different frequencies even
in the same animals (Lin 1978).
Recently, Wilson et ah (1980) described an autoradiographic technique
14
using [ C]2-deoxy-D-glucose to map auditory activity in the brain of rats
exposed to acoustic stimuli and to PW and CW microwave radiation. With this
technique, i_n vivo determination of metabolic activity, i.e., glucose utili-
zation and associated functional activity in the brain, can be visualized.
5-202
-------
Prior to exposure to the acoustic stimuli or to microwaves, one middle ear was
ablated. First, the authors showed the expected bilateral asymmetry of radio-
active tracer uptake in the auditory system of rats exposed to acoustic clicks
or weak background noise. Second, in contrast, a symmetrical uptake of tracer
was found in animals exposed to PW radiation. Thus, the autoradiographic
results confirmed the finding that RF hearing does not involve the middle ear
(Frey 1961; Chou and Galambos 1979). Unexpectedly, Wilson et aK (1980) found
similar patterns of radioactive tracer uptake in the auditory system of rats
exposed to CW radiation (918 MHz; 2.5 and 10 mW/cm^) and to PW radiation (2450
MHz). This result, indicating an auditory response to CW microwaves, was
unexpected, because no report of a direct hearing sensation due to exposure to
CW microwaves had appeared in the literature. Since the thermoelastic
hypothesis of RF hearing is based on the properties of PW radiation, this
observation suggests that another mechanism may be involved in the interaction
of RF radiation with the auditory system of the brain.
5.6.5 Human Cutaneous Perception
Exposure of the human body to microwave radiation can cause heating that
is detectable by the temperature sensitive receptors in the skin. As shown in
Table 5-18, several investigators have experimentally determined the microwave
intensities that cause sensations of warmth and pain in human subjects.
5.6.5.1 Frequency Specificity—
2
Hendler and colleagues (1963, 1968) exposed a circular area (37 cm ) of
the forehead to 3- and 10-GHz (PW) radiation and to infrared (IR) radiation.
5-203
-------
TABLE 5-18. SUMMARY OF STUDIES CONCERNING HUMAN CUTANEOUS
PERCEPTION OF RF RADIATION
Effect
Frequency
(MHz)
Intensity
(mW/cm2)
Duration
(s)
SAR*
(W/kg)
Reference
Sensation of
warmth on
forehead
3000
(PW)
10,000
(PW)
—
1-5
1-5
20-40
25-35
Hendler (1968)
Hendler et al.
(1963)
Sensation of
warmth on
forehead
2880
74
56
15-73
50-180

Schwan et al^ (1966)
Sensation of
warmth on
inner forearm
3000
(PW)
300-2500
1 6
—
Vendrick and Vos
(1958)
Sensation of
pain on inner
forearm
3000
2500
1000
30
130
2000
Cook (1952)
*SAR value is estimated.
The forehead was selected for a study of warmth sensations because the
temperature receptors in the skin of the forehead are relatively numerous and
are evenly distributed, so that the area constitutes a low-threshold region of
uniform temperature sensitivity. The lower-frequency microwaves (3 GHz) had a
higher intensity threshold for warmth sensation. The higher-frequency IR
radiation was more effective than either microwave frequency, because the IR
energy was absorbed more effectively in the outer skin layers containing the
thermal sensors.
5-204
-------
5.6.5.2 Temperature Thresholds--
For both microwave and IR radiation at intensities producing warmth
sensations, a threshold of warmth was experienced when the temperature of a
more superficial layer of subcutaneous tissue ~ 0.2 mm below the skin's
surface was increased ~ 0.01 to 0.02 °C over the temperature of a deeper layer
in the skin lying ~ 1 mm below the surface. In this study it was also noted
that there was a persistent sensation of warmth for ~ 7 s after cessation of
the exposure, which indicated the continued existence of an effective tempera-
ture difference between the subcutaneous tissue layers (Hendler 1968).
Schwan et a2- (1966) exposed a small area of the forehead (7-cm diameter)
equivalent to the area exposed in Hendler's studies to 2.88-GHz radiation and
measured the length of time that elapsed before the person was aware of a
sensation of warmth. The times for four subjects varied from 15 to 73 s at
2 2
74 mW/cm and from 50 to 180 s at 56 mW/cm . The authors found that the
reaction times were not linearly proportional to the reciprocal of the
incident power density and concluded that subjective awareness of warmth was
not a reliable indication of microwave hazard.
2
Vendrik and Vos (1958) exposed a 13-cm area of the inner forearm to
3-GHz (PW) radiation (300 to 2500 mW/cm^) and found the threshold for tempera-
ture changes to be 0.4 to 1.0 °C. Skin temperature increases that were kept
below 1 °C were linear with microwave intensity for six exposure durations.
In contrast to the regularity of skin temperature changes induced by microwaves,
the reports of temperature sensations were variable. Sensations of warmth
5-205
-------
occurred < 0.5 to 3.5 s after rapid rises in skin temperature. The sensations
did not cease when the skin temperature began to drop. In this study, microwave
radiation (3 GHz) was found to be a factor of 10 less effective in producing a
temperature elevation than was IR radiation at a similar intensity.
Cook (1952) determined the pain threshold in six subjects who were exposed
to 3-GHz radiation at five different sites on the body's surface. The initial
skin temperature ranged from 31.5 to 33.5 °C. Pain resulted when a critical
skin temperature (~ 46 °C) was reached rather than from a critical temperature
2
rise (AT). For an inner forearm area of 9.5 cm , the power density pain
2 2
thresholds varied from 2500 mW/cm for a 30-s exposure to 1000 mW/cm for an
exposure of 130 s. The pain threshold was lower for a larger exposed area of
2
53 cm . The skin temperature corresponding to burning pain was found to be
independent of the area of exposure, radiation intensity, exposure time, and
anatomical site. At high intensities, the exposure time needed to produce
pain was an inverse function of radiation intensity. The author reported that
the sensations of warmth and pain with microwave heating differed little from
those resulting from IR radiation.
5.6.6 Unresolved Questions
The few studies on pain and warmth sensations in human beings exposed to
frequencies in the range of 3 to 10 GHz provide useful data on exposure levels
that are clearly undesirable for the general population. Cutaneous perception,
however, may be a reliable indicator of an unsafe exposure level only at RF
5-206
-------
frequencies with wavelengths small in comparison to the length of the exposed
body, i.e., wavelengths comparable to or smaller than the thickness of skin.
Under these conditions, most of the energy is absorbed in the outer tissue
layers containing thermal sensors. At lower frequencies, which have wavelengths
approximately equal to or longer than the human body, modeling studies have
shown that much of the energy is absorbed within the body below the superficial
skin layers. One should make note of the fact that, in all mammals tested,
the threshold temperature (~ 42 °C) of cellular injury for sustained elevations
(seconds to tens of seconds) is below the threshold (~ 46 °C) of pain. These
results strongly indicate that cutaneous perception of RF energy is not a
reliable sensory response that will protect against potentially harmful levels
of RF radiation over the broad frequency range of 0.5 MHz to 100 GHz.
5-207
-------
5.7 OTHER PHYSIOLOGICAL AND BIOCHEMICAL EFFECTS
Charles G. Liddle
5.7.1 Clinical Chemistry and Metabolism
An individual's response to many stresses manifests itself through
changes in some of the clinical chemistry indices. Serum calcium and phos-
phate levels normally increase initially and then decrease below normal in
response to stress; whereas serum glucose, blood urea nitrogen, and uric acid
levels increase following stress. These blood chemical responses are con-
sistent with induction of release of adrenocortical hormones in response to
stress. Physiological responses, including thermal responses, to RF-radiation
exposure may occur at exposure levels too low to produce large changes in
thermoregulatory behavior or colonic temperature. In such instances there may
be changes in various metabolic parameters. There are few clinical chemistry
and metabolism reports where exposures were sufficiently defined to relate
results to the power density or SAR. These studies are described below, and
are outlined in Table 5-19.
A study of the effects of microwaves on serum chemistry values was
reported by Lovely et ah (1977), who exposed rats in a circularly polarized
waveguide to 918-MHz (CW) radiation. The animals were irradiated 10 h/day for
2
13 weeks at a power density of 2.5 mW/cm (SAR ~ 1.04 W/kg). They reported no
change in Na+, K+, ion gap, CI , blood urea nitrogen, or glucose values
compared to sham-irradiated animals. They did report a significant difference
in the serum calcium values at the end of the 12-week exposure period.
5-209
-------
TABLE 5-19. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON
CLINICAL CHEMISTRY AND METABOLISM, ENDOCRINOLOGY, AND GROWTH AND DEVELOPMENT
Exposure Conditions
Effects
Species
Frequency
(MHz)
Intensity
(raW/cmJ)
Duration
(days x mln)
SAR
(W/kg)
At
(°C)

Reference

No effect on serum chemistry
values
Rat
918 (CW)
2 5
91 x 600
1 0
0

Lovely et a^
(1977)
Increase In serum glucose
Rabbit
2450 (CW
and PW)
5, 10,
25
1 x 120
0 8-4 0 (est)
0. 0,
0, 0,
1 7 (PW)
2 9 (CW)
Wangemann and Cleary (1976)
Increase in blood urea nitrogen
Rabbit
2450 (CW)
25
1 x 120
4 0 (est)
2 9

No increase 1n blood urea
nitrogen
Rabbit
Rabbit
2450 (CW)
2450 (PW)
5 and 10
5, 10,
25
1 x 120
1 x 120
0 8, 1 6 (est)
0 8-4 0 (est)
0
0, 0,
1 7

Increased In uric acid values
Rabbit
2450 (CW
and PW)
10, 25
Neg* 5
1 x 120
1.6, 4 0 (est)
Neg 0 8
0, 1 7 (PW)
0, 2 9 (CW)
0

No effect on other serum
chemistry values
Rabbit
2450 (CW
and PW)
5, 10,
and 25
1 x 120
0 8-4 0 (est)
0, 0,
0, 0,
1 7 (PW)
2 9 (CW)

Increased Iron and manganese
levels In brain
Rat
1600 (CW)
80
1 x 10
48 (est)
4 5

Chamness et a2
(1976)
Decrease in specific metabolic
rate (Ambient T = 24 °C)
House
2450 (CW)
...
1 x 30
10.4
(Neg 5 5)

Ho and Edwards
(1977b)
Increase in specific metabolic
rate (Ambient T = 35 °C)
Mouse
2450 (CW)
...
1 x 30
8 6
(Neg 3 6)

Ho and Edwards
(1979)
Increased NADH fluorescense
Decreased ATP
Rat
(exposed
591 (CW)
13 8
1x05
0 36-2 2
0

Sanders et al
(1980)
l
ro
Decreased CP
brain)
(continued)
-------
TABLE 5-19. (continued)
Exposure Conditions
Effects Species Frequency Intensity Duration SAR At Reference
(~Wz) (iiW/cb2) (days x mm) (W/kg) (°C)
(Ji
l
ro
Decreased ATP
Decreased CP
No effect on blood coagulation
Increased thyroxine and tri-
iodothyronine
No effect on thyroid gland or
thyroid hormone
No effect on thyroid function
Decrease in seruoi-protein-bound
iodine, thyroxine and thyroxine/
serum ratio
Increase in thyroid hormone
Decrease in serum thyroxine levels
Increase in corticosterone levels
Decrease in thyrotropin levels
Increase in corticosterone levels
Decrease in thyrotropin levels
Rat
(exposed
brain)
Human
plasma
Dog
Rat
Rat
Rat
Rabbit
Rat
Rat
591 (CV)
2450 (CW)
2450 (CW)
2450 (CW)
2450 (CW)
2450 (CW)
3000 (PW)
2450 (CW)
5
10-280
72-236
1,10. 100
1. 10
10, 20, 25
15
20
Neg 1-10
2450 (120 Hz 40-70
AM) Neg 1-20
10, 40-70
Neg 1-5, 20
1x05
1 x 30
1 x 120
1 x 10-45
56 x 480
1 x 240-960
2 5 x 1440
48 x 180
1 x 240-480
1 x 60-480
1 x 60
1 x 240
10-40
Neg 1-5
25, 40 5 2, 8 4
Neg 1-20 0 02-4 2
0 13-0 8
1 3-38 (est)
58-190
0 25-25
0 25-2 5
2.5, 5.
6 25 (est)
3 8 (est)
0 25-0 75
(est)
0 25-25
8 4-14 7
Neg 0 21-4 2
2 1-14 7
Neg 0 21-4 2
2 1-8 4
Neg 0 2-1 0
0 6-2 1
0-1
Not reported
Not reported
S 10 mW/cm2 = 0
100 mW/cm2 i 5
S 20 mW/cm2 = 0
25 mW/cm2 =17
Not reported
0-0 6
0 6-1 4
1 3-3 0
0-0 6
0-3 0
0-0 6
0 3-2 1
0
Sanders et a^ (1980)
Boggs et a| (1972)
Magin et a2 (1977a,b)
Hilroy and Hichaelson (1972)
Parker (1973)
Parker (1973)
Baranski et a^ (1972)
Lu et a! (1977)
Lu et al (1981)
(continued)
-------
Effects
Species
No effect on thyroid, pituitary, Rat
or adrenal gland weight or growth
hormone levels
No effect on thyroid, anterior Rat
pituitary gland, adrenal, prostate
or testes weights, no change in
follicle stimulating hormone or
^ gonadotropic hormone levels
N> Increase in leutinizing hormone Rat
Increased adrenal weights and Infant
significant adrenal responses rat
Increased plasma corticosterone Rat
levels
No effect on serum corticosterone Rat
levels
No effect on weight gain Rat
TABLE 5-19. (continued)
Exposure Conditions
Frequency Intensity Duration SAR
(MHz) (mW/cm2) (days x mm) (W/kg)
2450 (CV) 1-20 1 x 60-480 0 25-25
2860-2880 (CW) 10 36 x 360
At Reference
(°C)
0-1 4 Lu et al (1977)
Not reported Mikolajczyk (1976)
1-2 (est)
2860-2880 (CW) 10 36 x 360 1-2 (est) Not reported Mikolajczyk (1976)
2450 (CW) 40 6x5 20-60 (est) 1 5-2 5 Guillet and Michaelson (1977)
2450 (CW) 50, 60 1 x 30-60 11 5-13 8 (est) 13 mW/cm2 =05 Lotz and Michaelson (1978)
Neg 13-40 Neg 3 0-9 2 20 mW/cm2 =07
(est) 30 raW/cm2 =09-14
20-40 1 x 120 4 6-9 2 (est) 40 mH/cm2 =13-14
Neg 13 Neg 13 50 mW/cm2 =16-2 4
60 mW/cm2 =25-29
918 (CW) 2 5 91 x 600 10 0 Lovely et al (1977)
2450 (CW) 5 55 x 240 0 7-4 7 Not reported Smialowicz et al (1979a)
(continued)
-------
TABLE 5-19. (continued)
Exposure Conditions
Effects
Species
Frequency
(MHz)
Intensity
(oW/cmJ)
Duration
(days x mm)
SAR
(W/kg)
at
(°C)
Reference
No effect on growth, neuro-
logical or lmauoological develop-
ment or mutagenicity
Rat
100 KHz (CW)
46
112 x 240
2 8
Not reported
Smialowicz et a^ (1981a)
Possible decrease in brain
acetylcholinesterase activity

37, 55
x 240

No effect on growth
House
2450 (CW)
10
24 x 48
6-8 (est)
Not reported
HcAfee et al (1973)
No effect on body weights
Infant
rat
2450 (CW)
40
6x5
20-60 (est)
1 5-2 5
Guillet and Hichaelson (1977)
No effect on growth
House
10 5, 19 27,
26 6 (CW)
8900
1 x 20
0 9, 1 8,
3.6 (est)
Not reported
Stavinoha et a^ (1975)

House
19
17,000-
114,000
5 x 40
6 3 (est)

No effect on weight gain
House
148 (CW)
0 5
50 x 60
0 013
Not reported
Lin et §2 (1979a)
*Neg = Effect not found at value indicated
-------
However, this is most probably a spurious result because the calcium
levels in the sham-irradiated animals were decreased from previous values,
while the levels in the irradiated animals remained unchanged from earlier
values. These results, therefore, mean that 918-microwaves at 2.5 mW/cm do
not alter the measured clinical chemistry values. Colonic temperature
measurements were made on the animals, and no detectable changes were
observed.
Somewhat differing from these results, Wangemann and Cleary (1976)
reported an increase in serum glucose levels in rabbits exposed for 2 h to
2
2450-MHz microwaves (CW and PW) at power densities of 5, 10 and 25 mW/cm
(SAR's estimated at 0.8, 1.6, and 4.0 W/kg, respectively), and an increased
2
blood urea nitrogen value in animals exposed to 25-mW/cm microwaves only.
2 2
Increased uric acid levels were found at 10 and 25 mW/cm but not at 5 mW/cm
(both CW and PW). Levels of calcium, phosphorus, cholesterol, total protein,
alkaline phosphatase, lactic dehydrogenase, and serum glutamic oxalic trans-
aminase were unaffected at the three power densities. The authors stated that
the PW and CW results could not be compared directly because the exposure con-
ditions were different. The results were those that would be expected from
2
heat stress. Animals exposed at 25 mW/cm for 2 h showed a significant rectal
temperature increase of 1.7 and 2.9 °C for PW and CW exposures, respectively.
2
Those exposed at 10 mW/cm showed evidence of mild heat stress, such as
peripheral vasodilation, but no significant increase in rectal temperature.
One possible explanation for the difference between these results and those of
Lovely et al_. is the difference in absorbed energy patterns in rats and
rabbits at 2450 MHz.
5-214
-------
Brains from rats exposed at 1600 MHz for 10 min at a power density of
80 mW/cm (SAR estimated at 48 W/kg) were analyzed for selected heavy metals
by Chamness et aK (1976). Iron levels were increased in all the areas of
the brain (hypothalamus, corpus striatum, midbrain, hippocampus, cerebellum,
medulla, cortex). Manganese was increased in the cortex and medulla, and
copper was increased in the cortex; while calcium, zinc, sodium, and potassium
levels were unchanged. The observed changes were probably a result of hyper-
thermia since most alterations seen were also observed in rats subjected to a
hyperthermal environment, and the irradiated animals showed a rectal tempera-
ture increase of 4.5 °C.
Platelet-rich human plasma was irradiated i_n vitro with 2450-MHz micro-
waves at power densities of 10 to 280 mW/cm^ (SAR's of 1.3 to 38 W/kg) for
0.5 to 24 h by Boggs et aK (1972), and effects on blood coagulation were
analyzed. They reported no significant changes in platelet count, coagulation
time, or clot strength at these power densities. The plasma temperature was
maintained below 37 °C during the exposures. They also conducted studies on
the effect of heating on coagulation time and clot strength. Samples were
heated either by microwaves or by radiant heating, and the results were com-
pared to unheated control samples. The relative coagulation time for the
microwave-heated samples remained unchanged throughout the temperature range
studied (34, 37, 39 and 42 °C), while the samples heated by radiant energy
showed increases in the relative coagulation time (1.58 and 2.03 times the
non-heated samples) at 39 and 42°C, respectively. A similar, but reverse
effect was seen in the relative clot strength. The samples heated to 42 °C
by microwaves showed a decrease to 0.74 times the control samples, while the
5-215
-------
radiant-heated samples showed a decrease to 0.31 times the control values. A
possible explanation for these differences is the difficulty of producing the
same heating rate and pattern within a sample using different sources of heat.
Ho and Edwards (1977b) measured the rate of oxygen consumption of mice
exposed to 2450-MHz microwaves in a waveguide for 30 min at dose rates from
1.6 to 44.3 W/kg at an environmental temperature of 24 °C. They found a
significant decrease in the specific metabolic rate (SMR) at 10.4 W/kg or
higher, but not at 1.5 or 5.5 W/kg. There was no detectable rectal tempera-
ture increase at or below 10.4 W/kg, but there was a 0.5 °C increase at
23.6 W/kg and a 1.0 °C increase at 44.3 W/kg. These results indicate that the
mouse compensates for large dose rates of microwave energy by adjusting its
metabolic rate downward to compensate for the thermal load.
In a more recent report, Ho and Edwards (1979) reported a continuation of
the previous study. Exposure conditions were identical to those of the pre-
vious study except that environmental temperatures of 20, 30, and 35 °C were
also used. At 20 °C they found a significant decrease during exposure in the
SMR at 12.1 W/kg and above, but not at 6.0 W/kg or below. A significant
decrease was found in the SMR during the 30-min post-exposure period at
45.1 W/kg, but not at 27.0 W/kg or less. For the exposures at 30 °C there was no
change in the SMR at any of the dose rates used (1.4 to 23.7 W/kg) either
during exposure or in the post-exposure period. , At 35 °C there was a signi-
ficant increase in the SMR at 8.6 W/kg and higher, but not at 3.6 W/kg or
lower during exposure, and no changes were observed during the postexposure
period. The authors re-evaluated their first report (Ho and Edwards 1977b) by
5-216
-------
comparing the results during exposure to the pre-exposure values instead of
comparing the values of the animals during exposure with those of sham-exposed
animals. With this method of evaluation, the SMR values during exposure at 24
°C were not significantly different from the pre-exposure values. They re-
ported that this method of comparison accounted for the lack of uniformity
among animals at the beginning of each experiment and therefore was a better
method of comparison.
The general trends indicated by their studies are that microwave exposure
at the lowest ambient temperature resulted in a reduction in the SMR, exposure
at the highest ambient temperature resulted in an increase in the SMR, and no
significant changes occurred at the intermediate temperatures. A possible
interpretation of these trends is that at the lower ambient temperatures the
animals are producing heat to maintain thermal neutrality; addition of the
microwave heating reduces the animals' demand for additional heat. At the
higher amibent temperatures the animals may be actively trying to dissipate
the additional heat from the microwaves, thereby increasing their SMR.
A study on the effects of microwaves on energy metabolism of the rat
brain was reported by Sanders et aj. (1980). First, a small area of the brain
of anesthetized animals was surgically exposed. Then a horn antenna was
positioned so that exposures were in the far field and only the exposed sur-
face of the brain was irradiated with the electric field parallel to the body
axis. Animals were exposed to 591-MHz (CW) radiation for 0.5, 1, 2, 3, or
2 2
5 min at 13.8 mW/cm or for 0.5 or 1 min at 5 mW/cm . (Calculated SAR for
5-217
-------
2
5 mW/cm = 0.13 W/kg using a 2-cm sphere model or 0.8 W/kg using a semi-infinite
2
plane model.) During exposures at 13.8 mW/cm they found an increase in
nicotinamide adenine dinucleotide (NADH) fluorescence to a maximum of 4.0 to
12.5 percent above pre-exposure control levels. In addition, adenosine tri-
phosphate (ATP) levels were significantly decreased at all exposure times, as
2
were creatine phosphate (CP) levels. At 5 mW/cm , ATP and CP levels were also
significantly decreased following 0.5- and 1-min exposures. The ATP and CP
2
changes at 5 mW/cm were not significantly different from those seen at 13.8
2
mW/cm . There were no changes in rectal temperature at any of the exposures
and no significant difference in brain temperature between exposed and sham-
exposed animals. The authors concluded that the results (increased NADH
fluorescence; decreased ATP and CP levels) support the hypothesis that RF
radiation inhibits mitochondrial electron transport chain function and that
the changes cannot be attributed to general tissue hyperthermia.
In summary, changes in clinical chemistry values have been reported at
dose rates as low as 0.8 W/kg in rabbits, and negative results have been
reported at exposures as high as 1 W/kg in rats. The clincial chemistry
changes that have been reported are those that would be expected from heat
stress. In other studies, effects on the rate of oxygen consumption of mice
have been reported at 9.6 W/kg but not at 5.5 W/kg, and changes in brain
energy metabolism have been found at 0.13 to 0.8 W/kg (Table 5-19).
5-218
-------
5.7.2 Endocrinology
The endocrine system in coordination with the nervous system is a major
regulatory system in the body. Alterations in the function of the
neuroendocrine system often reflect the efforts on the part of the animal to
maintain homeostasis when subjected to stressful internal and external stimuli.
Detection of changes in the endocrine system is a sensitive means of analyzing
the animal's responses either to direct stimulation of the endocrine organs
themselves or to stimulation of the CNS. The thyroid gland plays a basic role
in regulating basal metabolism as well as in generating metabolic heat in the
tissues. Changes in thyroid activity can result from changes in thyroid-
stimulating hormone from the hypothalamic-hypophyseal system and/or increased
metabolic activity of the thyroid gland due to heating. Direct interaction
with the CNS could also produce changes in thyroid activity. Moderate or
gradual heating results in a reduction of thyroid hormone, while rapid or
marked elevation of body temperature results in an increase in thyroid
activity. The results of RF-radiation-effects studies on the endocrine system
are discussed below and summarized in Table 5-19.
Thyroid function of rats following exposure to 2.45-GHz microwaves at 10,
2 2
20, and 25 mW/cm (SAR's estimated at 0.25 W/kg per mW/cm ) for 4 or 16 h was
studied by Parker (1973). No effects on thyroid gland function were found at
2
these exposures. However, at 15 mW/cm (SAR estimated at 3.75 W/kg), exposure
for a longer period (60 h) was reported to produce a significant decrease in
serum-protein-bound iodine, thyroxine and thyroid/serum iodine ratio. A
2
significant rectal temperature increase (1.7 °C) was reported at 25 mW/cm ,
but not at lower power densities.
5-219
-------
Magin et aJL (1977a, b) irradiated the surgically exposed thyroid gland
of anesthetized dogs with 2450-MHz (CW) microwaves using a waveguide appli-
2
cator at power densities from 72, 162, and 236 mW/cm (SAR's from 58 to
190 W/kg) for 2 h. One thyroid was irradiated while the other was used as
a control. Tissue temperatures of 39, 41, and 45 °C were maintained in the
thyroid at the three power densities. They reported an increased release of
thyroxine (TH) and triiodothyronine (T3) at all power densities, demonstrating
that the thyroid gland in the dog can be stimulated directly by microwave
heating.
Milroy and Michaelson (1972), who exposed rats to 2450-MHz microwaves at
1, 10, and 100 mW/cm^ (SAR = 0.25 W/kg per mW/cm^) for single exposures of 10,
2
20, 30, and 45 min and at 1 and 10 mW/cm 8 h/day for 8 weeks, reported no
effect on T3 levels, thyroxine levels, or on the uptake of radioactive iodine.
2
No rectal temperature increase was observed at 10 mW/cm or less. At 100 mW/
2
cm there was a constant rise in rectal temperature throughout exposure, up to
42 °C at the end of the exposure period. Histopathological examination of the
thyroid glands also showed no effect from the exposure.
An increased production of thyroid hormone in rabbits as measured by in-
131
creased incorporation of I and increased radioactivity per gram of thyroid
(verified by autoradiography) was reported by Baranski et al^. (1972). The
animals were exposed for 3 h/day for 4 months to 10-cm (3-GHz, PW) micro-
2
waves at an average incident power density of 5 mW/cm (SAR estimated at 0.25
to 0.75 W/kg). They reported no increase in body temperature or thyroid
temperature. (The pulse parameters were not given, so peak power density
cannot be calculated.)
5-220
-------
Lu et a^. (1977) reported serum thyroxine levels in rats exposed to
2450-MHz microwaves at 1, 5, 10, and 20 mW/cm^ (0.25, 1.25, 2.5, and 5 W/kg)
2
for 1, 2, 4, or 8 h. Decreased thyroxine levels were observed at 20 mW/cm
following 4- and 8-h exposures. The thyroxine values at shorter exposures
and lower power densities were not significantly different from the sham
2
values, except for an increase after 4 h at 1 mW/cm , which appears to be a
chance variation, since at both higher and lower power densities and at longer
and shorter periods of exposure no effect was detected. There were small but
2
statistically significant rectal temperature increases at 1 mW/cm after 4 h,
2 2
at 5 mW/cm after 1 and 2 h, and at 10 mW/cm after 2 and 4 h of exposure.
The increases were in the 0.2 to 0.56 °C range. The lack of correlation
between power density, exposure time, and rectal temperature increase, along
with the small rectal temperature change, suggests that the effects may not
have been due entirely to the microwave exposure but possibly from the stress
2
of confinement. At 20 mW/cm , however, the rectal temperature was signifi-
cantly elevated at all four exposure times and tended to increase with longer
exposures. At 1 and 2 h the changes were small (0.64 and 0.54 °C), but larger
(1.01 and 1.35 °C) increases occurred after 4 and 8 h. Serum corticosteroid
2
levels were decreased at 20 mW/cm after 8 h only, which the authors report as
a shift in the circadian periodicity. Serum growth hormone measurements
showed no change at any of the power densities or times reported. They also
measured no change in the mass of the thyroid, pituitary, and adrenal glands
following irradiation.
In another study on the effects of microwaves on acute endocrine responses
in rats, Lu et aJL (1981) exposed animals to 2.45-GHz microwaves (AM at 120 Hz)
5-221
-------
for 1 and 4 h, and measured colonic temperature, thyrotropin, and cortico-
sterone levels immediately after exposure. The 1-h exposures were conducted
2
at power densities of 1, 5, 10, 20, 40, 50, 60, or 70 mW/cm , and the 4-h
exposures at 0.1, 1, 5, 10, 20, 25, and 40 mW/cm^ (measured SAR = 0.21 W/kg
2
per mW/cm ). The results for the 1-h exposure showed an increase in colonic
temperature with increasing power density with a significant increase at
2 2
20 mW/cm and above, but not at 10 mW/cm or below. Corticosterone also
showed an increase with increased power density with evidence of a threshold
2
between 20 and 40 mW/cm . Thyrotropin values showed a decrease with increasing
2
power densities, significant results at 40 mW/cm and above, and equivocal
2 2
results in the 10 to 20 mW/cm range (10 mW/cm values were significantly
2
lower while those at 20 mW/cm were not). Similar trends were observed fol-
lowing 4-h exposures, with a significant increase in colonic temperatures at
2 2
10 mW/cm and above, and no increases at 5 mW/cm or below with the exception
2
of one small group at 1 mW/cm . The trends for the increase in corticosterone
levels and the decrease in thyrotropin levels were not significant; however,
2
the corticosterone level at 40 mW/cm was significantly increased, but not at
2 2
25 mW/cm or lower values. Also, the thyrotropin values at 25 mW/cm and
2 2
40 mW/cm were significantly decreased, whereas those at 20 mW/cm and below
were not. Comparing the effects of 1- and 4-h exposures on the three param-
eters, the change in colonic temperature was similar for both exposures;
however, the stimulatory effect on serum corticosterone levels was less for
the 4-h exposures than for the 1-h exposures, while the depression of serum
thyrotropin was more pronounced after 4 h than after 1 h. The authors con-
cluded that the hormonal changes probably represented a general nonspecific
5-222
-------
stress reaction that was related to the intensity and duration of the stress-
ing agent, rather than to the nature of the agent itself.
No change in thyroid weight was seen in a study by Mikolajczyk (1976),
where rats were exposed to 2860 to 2880-MHz microwaves at 10 mW/cm , 6 h/day,
6 days/week for 6 weeks. The SAR is estimated to be 1 to 2 W/kg for a single
animal exposure, but the animals were exposed close together in a box with
dividers every 10 cm, which should give a somewhat higher SAR value. The
author did find a significant increase in luteinizing hormone from the
anterior pituitary gland but no change in follicle-stimulating or gonadotropic
hormone levels. The weights of the anterior pituitary, adrenal, prostate, or
testes were not affected by the exposure.
Other investigators also have reported the effects of microwaves on
adrenal function. An increase in blood-corticosteroid concentration above
that which would normally occur at that time of day in the absence of a
stimulus is considered by many to be an indicator of an animal's response to
stress. This results when an internal or external stimulus, either chemical,
physical, or emotional, excites neurons of the hypothalamus to release
corticotropin-releasing hormone, which drives the pituitary to release
adrenocorticotropic hormone (ACTH). This hormone then stimulates the adrenal
cortex to secrete corticosteroids. Among the strongest stressful stimuli are
surgery, anesthesia, cold, narcosis, burning, high environmental temperature,
and rough handling or restraint. Additional studies examining the effects
of microwaves on adrenal function are described below.
5-223
-------
2
Lovely et a2. (1977) exposed rats to 918-MHz microwaves at 2.5 mW/cm
(SAR ~ 1.04 W/kg), 10 h/day for 13 weeks, and observed no changes in serum
corticosterone levels. In a study by Guillet and Michaelson (1977), rat pups
2
were exposed to 2450-MHz microwaves at 40 mW/cm (SAR estimated at 20 to
60 W/kg), 5 min/day for 6 days beginning at day 1 postpartum; no change in basal
corticosterone levels was found. There was a significant adrenal response to
the microwaves, but this was the same as seen following ACTH administration,
indicating a stress reaction. Rectal temperatures of the exposed animals
averaged 1.5 to 2.5 °C higher than those of the sham-exposed animals. Adrenal
mass of the irradiated animals was significantly greater than those of the
control animals.
In another study, Lotz and Michaelson (1978) irradiated rats with
2
2450-MHz microwaves at power densities of 13, 20, 30, and 40 mW/cm for 30,
60, or 120 min and at 50 and 60 mW/cm^ for 30 or 60 min (SAR estimated at 0.21
2
W/kg per mW/cm ). Plasma corticosterone levels were increased at power
2 2
densities at or above 50 mW/cm , but not at 40 mW/cm or less, for the 30- and
60-min exposures. At the longest exposure time (120 min), increased levels
2 2
were seen at or above 20 mW/cm but not at 13 mW/cm . Graphs were presented
of the rectal temperatures taken at the completion of exposures; estimates of
the rectal temperature increases taken from the graphs for 50 mW/cm were 1.6
O
and 2.4 °C and for 60 mW/cm were 2.5 and 2.9 °C for 30 and 60 min, respectively.
2
The temperature increases for 40 mW/cm were 1.3 and 1.4 °C for 30 and 60 min.
2
For the 120-min exposures, rectal temperature increased 0.5 °C at 13 mW/cm
and 0.7 °C at 20 mW/cm^.
5-224
-------
In summary, microwave effects on thyroid function have been reported at
SAR's as low as 2.1 W/kg, and negative results have been reported at values
as high as 25 W/kg. The duration of exposure as well as the exposure rate
appears to be important, as demonstrated by the study of Parker (1973), where
exposures at 6.25 W/kg for 16 h produced no changes, while 3.75-W/kg exposures
for 60 h resulted in a decrease in serum thyroxine levels. Changes in corti-
costerone levels have been reported from microwave exposure at SAR's as low
as 10 W/kg but not at 6.25 W/kg, and adrenal responses have been reported at
levels as low as 4.6 W/kg but not at 3 W/kg.
5.7.3 Growth and Development
Few investigators have reported the effects of a combination of pre- and
postnatal exposure or postnatal exposure only to RF radiation on the growth
of laboratory animals. Smialowicz et al^. (1979a) exposed rats for 4 h per day,
beginning on day 6 of gestation through 40 to 41 days postpartum to 2450-MHz
(CW) microwaves at 5 mW/cm (SAR's = 0.7 to 4.7 W/kg), and reported no dif-
ference between the weight gains of the exposed and sham-exposed animals.
In another study, Smialowicz et a^. (1981a) reported the growth and
development of rats exposed to 100-MHz (CW) microwaves at an incident power
density of 46 mW/cm (average SAR =2.8 W/kg) for 4 h/day from day 6 of
gestation through 97 days of age. There was no consistent difference between
the body weights of the exposed and sham-exposed animals, though the exposed
animals tended to be larger than the sham-exposed animals. Some of the animals
were tested for neurological development, and no differences were observed in
5-225
-------
the development of a startle response or righting reflex. There was a sig-
nificant difference in the age of eye opening, with the mean age of eye open-
ing in the sham-irradiated animals occurring almost one day later than the
irradiated animals. The authors stated that the change probably did not
represent an acceleration of eye opening, as the age of eye opening in the
irradiated animals was similar to that normally seen in control animals in
other experiments, but also mentioned that eye opening was delayed in the
sham-irradiated animals. Tests for development of motor activity at 35 and
84 days of age showed no difference between exposed and sham-exposed animals.
No difference was observed for complete blood counts, mitogen-stimulated
response of lymphocytes, frequency of T- and B-lymphocytes, or antibody
response to Streptococcus pneumoniae capsular polysaccharide between the ex-
posed and sham-exposed animals. No mutagenic effect was observed on the sperm
cells after 20 days using the dominant lethal assay. Seven regions of the
brain were weighed at 22, 40, and 97 days of age, and there was a significant
increase in the weight of the medulla in the irradiated animals at 40 days of
age but not at 22 or 97 days of age. There were no differences in the weights
of any other brain region. There were also no differences in the brain pro-
tein concentrations for the seven regions. Brain acetylcholinesterase (AChE)
activity, however, was reduced in the striatum and medulla at 22 days of age
and in the midbrain at 40 days of age. No effects on AChE activity were noted
in the other regions at 22 and 40 days of age, and no differences were noted
at 97 days of age. The differences were small, and there appeared to be no
pattern to the changes; the sample size was small (3 to 5 animals), and out
of 21 comparisons made at the 5 percent confidence level, one significantly
different result would be expected. Consequently, without replication with
5-226
-------
a larger sample size, it is difficult to ascribe these changes to the micro-
wave exposure.
The effect of postnatal exposure to microwaves on growth and development
was studied by McAfee et al^ (1973). Weanling mice were exposed to 2450-MHz
microwaves at 10 mW/cm , 2 min each hour for 24 days; no effect on animal
growth was found. There was no elevation of body temperature in the exposed
animals. The authors also stated that a previous study, in which a stimula-
tory effect on growth from microwaves was claimed, was probably in error because
of inaccuracies in weighing the animals (see Nieset et ah 1958).
Guillet and Michael son (1977) exposed neonatal rats to 2450-MHz micro-
waves at 40 mW/cm (SAR's at 20 to 60 W/kg), 5 min/day for 6 days beginning on
day 1 postpartum. No effect on body mass was found, though the authors did
report adrenal changes, as discussed in the previous section, § 5.7.2,
Endocri nology.
Stavinoha et a2- (1975) irradiated 4-day-old mice at a field intensity of
5800 V/m (8.92 W/cm^) for 20 min with 10.5-, 19.27-, or 26.6-MHz radiation,
and observed no change in gr.owth rate to 22 days of age compared to the
control animals (SAR's estimated at 0.9, 1.8, and 3.6 W/kg). Mice were also
exposed to 19-MHz microwaves for 40 min/day for 5 days using a near-field
2
synthesizer to deliver an E field of 800 V/m (17 W/cm ) and an H field of
2
55 A/m (114 W/cm ); no change in body mass through 120 days of age was found
(SAR estimated at 6.3 W/kg).
5-227
-------
In a chronic study, Lin et aj. (1979a) exposed 4- to 7-day-old mice to
148-MHz microwaves at 0.5 mW/cm^ (SAR = 0.013 W/kg), 1 h/day, 5 days/week for
10 weeks and reported no significant effect on weight gain.
These studies indicate that RF exposures at SAR's < 8 W/kg have no effect
on growth and development of animals exposed pre- and postnatally or post-
natal ly only. Exposures at 20 to 60 W/kg postnatally for short exposure times
also did not alter the growth rate of animals even though adrenal changes were
seen.
5.7.4 Cardiovascular System
5.7.4.1 Whole-Body Exposures—
Some of the initial studies of biological effects of microwave radiation
were made on the cardiovascular system. Presman and Levitina (1962) reported
that whole-body or ventral exposure to CW microwaves at 2400 MHz, 7 to
2
12 mW/cm , promoted a decreased heart rate (bradycardia) in the rabbit.
However, exposing only the head to microwaves at the same power densities
resulted in an increased heart rate (tachycardia).
Kaplan et al_. (1971) attempted to replicate the tachycardia reported by
Presman and Levitina by exposing the head of rabbits to 2400-MHz (CW) micro-
2
waves at an incident power density of 10 mW/cm (SAR estimated at 2 W/kg).
The exposure conditions differed in that Kaplan et aK exposed the animals in
the far field, while Presman and Levitina used near-field exposures. Kaplan
et al. reported no change in heart rate. They then exposed rabbits at
5-228
-------
2
increasing power densities (20, 40, 60, 80, and 100 mW/cm ) and measured
respiration rate, body temperature, and heart rate. Respiration increased at
2 2
40 mW/cm and greater, and body temperature was elevated at 80 and 100 mW/cm
p
(0.5 °C at both power densities). Heart rate increased at 100 mW/cm only.
Birenbaum et a\_. (1975) exposed the heads of unanesthetized rabbits to
2.4-GHz (CW) microwaves for 60 min at 20, 40, 60, and 80 mW/cm^ (SAR estimated
at 3 to 12 W/kg), and found increases in heart rate, respiration rate, and
subcutaneous temperature (lower back) at all four power densities, with greater
increases at higher exposure levels. They also compared 2.8-GHz (CW and PW,
o
1000/s, 1.3 ps) microwaves at 20 mW/cm for the same three parameters and
found no differences in the responses to CW or PW irradiation.
Phillips et aK (1975b) exposed adult rats to 2450-MHz microwaves pulsed
at 120 Hz for 30 min (SAR's = 4.5, 6.5, and 11.1 W/kg) and measured colonic
and skin temperatures and heart rate after exposure. They reported increases
in colonic and skin temperatures with increased exposure levels immediately
after irradiation. At the highest exposure, the colonic temperature dropped
below normal at 1 h after exposure and remained subnormal for 4 h. There was
no post-exposure effect on heart rate at 4.5 W/kg; however, there was a mild
bradycardia at 6.5 W/kg and a more pronounced decrease in heart rate at
11.1 W/kg. They attributed the effect to the heat stress induced by the
microwaves.
5-229
-------
Hamrick and McRee (1980) assessed the effects of body temperature on the
heart rate of embryonic quail. They exposed the embryos to 2450-MHz (CW)
microwaves for 5 to 10 min (SAR's = 3, 6, 15, and 30 W/kg) at incubation
temperatures from 35 to 38 °C, and to 2450-MHz (PW) microwaves (10-ps pulses,
varied from 10 to 50 pulses/s, SAR's at 0.3, 1.5, and 3 W/kg) at incubation
temperatures of 35 to 39 °C. There were no significant differences between
the heart rates of the exposed and control embryos in any of the groups at any
of the temperatures used. They did observe that the embryonic heart rate
increased ~ 23 beats/min for each °C rise in incubation temperature in the 36
to 39 °C range.
5.7.4.2 Isolated Heart Preparations—
Although the data supporting a microwave-induced bradycardia in the in-
tact animal are equivocal, some researchers have exposed isolated heart
preparations and reported an effect of microwaves on heart rate. Tinney et
al. (1976) attempted to determine the locus of microwave-induced bradycardia
in the isolated turtle heart using 960-MHz (CW) microwaves. Although the
heart was isolated from the CNS, it was still capable of responding to neuro-
humoral agents. They found that exposing the isolated heart to microwaves at
2
8 mW/cm caused bradycardia. When the heart was treated with atropine,
blocking the sympathetic system, tachycardia resulted. However, when the
heart was treated with propranolol, blocking the parasympathetic system, the
heart rate decrease was more significant than that following microwave
exposure alone, indicating that microwaves may have affected the para-
sympathetic reflexes. With both blocking agents added to the heart, microwave
exposure had only slight effects on the heart rate. Tinney et a^. postulated
5-230
-------
that microwave exposure of the isolated heart equally enhances the release of
acetylcholine and norepinephrine; however, since the activity of the former
usually predominates over the latter (Johansen 1963), the net effect of micro-
wave exposure is a decrease in heart rate.
A similar drug-microwave interaction has been observed in the isolated
rat heart by 01 sen et al^ (1977). They exposed the heart to 960-MHz (CW)
microwaves for 4 min at 1.3 and 2.1 W/kg and observed a decrease in heart
rate. The decrease was greater at 1.3 W/kg than at 2.1 W/kg. They also
conducted studies at 2.1 W/kg with drugs to block the sympathetic and para-
sympathetic nervous system and obtained the same results as Tinney et al.
(1976).
Paff et al_. (1963), working with the isolated embryonic chicken heart,
were unable to detect changes in heart rate during exposure to 24,000-MHz
fields. They did, however, detect effects on the electrocardiogram (ECG),
2
including abnormal P and T waves from 3-min exposures at 74 mW/cm .
Frey and Seifert (1968) showed that 10-ps pulses at a carrier frequency
of 1.425 GHz given at a synchronous period with the ECG (220 ms after the P
wave) resulted in tachycardia or heart arrhythmia in the isolated frog heart.
2 2
The peak power density was 60 mW/cm (average power density -0.6 pW/cm ).
Liu et a2. (1976) reported no effect on heart rate with isolated frog hearts
or in hearts irradiated i_n situ in a similar study. The iji situ hearts were
exposed to 100-ps pulses of either 1.42 or 10 GHz, and the isolated frog
hearts were exposed to 100-ps pulses of 1.42 GHz. The pulse was delivered
5-231
-------
on the rising phase of the R-wave from the ECG, which was somewhat similar to,
but not exactly the same as, the 200-ms delay following the P-wave used by
Frey and Seifert. (The R-wave follows the P-wave by around 200 ms.) The peak
and average power densities of 320 mW and 32 ^iW were also considerably higher
than those used by Frey and Seifert. These factors, plus differences in the
manner of preparing the isolated hearts (Liu et al_. curarized the frogs,
whereas Frey and Seifert decapitated the frogs), make it difficult to compare
the results of the two studies.
Clapman and Cain (1975), however, tried to replicate the study of Frey
and Seifert using similar pulse widths (10 ps), peak and average power densities
2 2
(60 mW/cm and 0.6 pW/cm ), carrier frequency (1.42 GHz), and method of isolating
the frog heart; they reported no change in heart rate. When they conducted
studies with a different peak power (5.5 W/cm ), frequency (3 GHz), and pulse
widths (2 and 150 ps), they also found no heart rate changes. Clapman and
Cain were able to produce an increased heart rate with 20-mA current pulses
synchronized 200 ms after the P-wave peak.
The results of microwave exposure on the cardiovascular system indicate
that whole-body exposures of sufficient intensity to produce heating will also
produce an increase in heart rate similar to that which would be expected from
heating alone. In the isolated heart there appears to be a stimulation of the
autonomic nervous system from microwave exposure at levels where little heat-
ing would be expected (1 to 2 W/kg). Low levels of synchronized PW microwaves
(0.6 to 32 mW/kg) apparently are ineffective in producing detectable alterations
5-232
-------
in heart rate. A summary of the RF-radiation effects on cardiac physiology is
given in Table 5-20.
5.7.5 Calcium Ion Efflux
Calcium ions play a prominent role in many biochemical and biophysical
processes. They are involved in maintaining cellular membrane integrity and
function, as a cofactor in many enzyme reactions, as a putative second mess-
enger for the conduction of extracellular signals to the nucleus of the cell,
and in neural tissue for excitation and for the secretion of transmitter
substances at synapses. Several investigators have reported the effects of
electromagnetic fields on the response of calcium ions in biological systems,
principally brain tissue. Bawin et aK (1975) showed that a 20-min in vitro
exposure of chicken brain tissue to a 147-MHz field, at a power density of 1
2
to 2 mW/cm , could cause enhanced efflux of calcium ions from the samples, but
only if the field was sinusoidally amplitude modulated with frequencies of 6,
9, 11, 16, or 20 Hz. Maximal efflux was measured at 16 Hz. Modulation fre-
quencies of 0, 0.5, 3, 25, or 35 Hz were ineffective. This result was corrob-
orated and extended by Blackman et al_. (1979, 1980a), who showed that the
2
response occurred only within a narrow intensity range around 0.83 mW/cm .
Sheppard et a^. (1979) confirmed this intensity response with a 450-MHz
carrier wave amplitude modulated at 16 Hz; calcium efflux was enhanced at 0.1
2 ?
and 1.0 mW/cm but not at 0.05, 2.0, or 5.0 mW/cm .
Blackman and co-workers also examined the influence of 50-MHz fields,
amplitude modulated at 16 Hz, and found that the response could be elicited
5-233
-------
TABLE 5-20. SUMMARY OF STUDIES CONCERNING RF-RADIATION EFFECTS ON
VARIOUS ASPECTS OF CARDIAC PHYSIOLOGY
Effects
Species
Frequency
(MHz)
Exposure Conditions
Intensity Duration SAR
(mW/cm2) (total mm) (W/kg)
Reference
cn
i
ro
CO
Bradycardia develops after
whole-body exposure along
with hyperthermia
Exposure to head promotes
tachycardia, exposure to
back raises respiratory rate
but not heart rate
Increased respiration
Increased heart rate from
dorsal exposure of the head
Alterations in ECG
(shortening of QT interval,
increased height of T-wave,
appearance of U-wave)
No effect on heart rate that
cannot be attributed to
microwave heating
Pulses synchronized with each
R-wave do not affect heart
rate
Synchronized pulses with QRS
complex causes increase in heart
rate with some arrhythmias
Rat
Rabbit
Rabbit
Rabbit
72-h
chick
heart
Quail
embryo
Frog
Frog
2,450 (PW)
28 & 48
2,450 (CW and 20
PW)
2,400
2,400
24,000 (PW)
2,450 (CW and
PW)
1,420-10,000
(PW)
1,425 (PW)
40-100
100
74
*
NA
NA
0 6 jjW/cm2
30
60
20
20
5 to 10
100-ps
pulses
10-MS
pulses
6 5 & 11 1 Phillips et al (1975b)
Birenbaum et al (1975)
8-20 Kaplan et al (1971)
20 Kaplan et a2 (1971)
NG Paff et al (1963)
0 3-30 Hamrick and McRee (1980)
12 or 16 pW Liu et a! (1976)
per heart
NG
Frey and Seifert (1968)
(continued)
-------
TABLE 5-20 (continued)
Exposure Conditions
un
Effects
Species
Frequency
(MHz)
Intensity
(mW/cn2)
Duration
(total mm)
SAR
(W/kg)
Reference
PO
oo
vn
Low power levels cause
bradycardia in the isolated
turtle heart
Turtle
960 (CW)
NA
60
2-10
T inney et al^ (1976)

Causes slight decrease in the
isolated heart
Rat
960 (CW)
NA
5 to 10
1 3 and 2 1
Olsen et a! (1977)

Synchronized exposures with ECG
have no effect on heart rate
Frog
1420-3000
(PV)
0 0006
2-, 10-,
150-ys
pulses
NG
Clapman and Cain (1975)
*
NA = Not appltcable
-------
2
within two intensity regions (between 1.44 and 1.67, and at 3.64 mW/cm )
2
separated by power densities of no effect including 0.72 mW/cm (Blackman et
al. 1980b). The apparent discrepancy in the calcium ion efflux data at dif-
ferent power densities at the three different carrier frequencies (50, 147,
and 450 MHz) has been resolved by the finding that the efflux is dependent
upon the electric field strength within the sample and not on the incident
power density (Joines and Blackman 1980). The empirical model used to trans-
form the incident power density to internal field strength has also been used
to predict additional values of intensity that produced both alteration and no
alteration in calcium ion efflux (Joines and Blackman 1980). It should be
noted that the 147-MHz and 50-MHz exposures caused no generalized heating of
the sample. The maximum temperature rise was calculated to be less than
0.0004 °C, and calculated SAR's at each carrier frequency were less than
0.0014 W/kg (Blackman et aK 1980b).
Although outside the range of the frequencies of interest for this review,
frequencies within the same 6- to 20-Hz region applied directly, without a
carrier wave, were demonstrated to alter the efflux of calcium ions from
chicken brain tissue (Bawin and Adey 1976). In this case, the efflux of cal-
cium ions was reduced by the exposure. At the intensities used, there is no
heating of the sample exposed to 6- to 20-Hz fields because energy absorption
is very small.
No change in calcium ion efflux from brain tissue has been reported by
Shelton and Merritt (1981), who used procedures somewhat different from those
described by Bawin et a^. (1975), Bawin and Adey (1976), and Blackman et a]_.
5-236
-------
(1979, 1980a). They irradiated brain tissue from the rat with 1-GHz radiation,
2
square-wave, pulse-modulated at 16 Hz (0.5, 1.0, 2.0, or 15 mW/cm ). In their
original study, Bawin et aK (1975) also exposed skeletal muscle to the same
exposure conditions that produced calcium ion efflux from chick brain tissue, but
no effect was observed.
An increase of calcium ion efflux from rat pancreatic tissue exposed at
?
2 mW/cm for 1 to 2.5 h at 147 MHz, sinusoidally amplitude modulated at 16 Hz
(estimated SAR < 0.075 W/kg), has been reported by Albert et fH. (1980).
However, the efflux was not accompanied by a change in protein secretion,
which is normally associated with calcium mobilization in the pancreas. The
authors attributed the lack of protein secretion to the relatively small
volume of medium available to the tissue slices for normal metabolic activity.
Although the efflux of calcium ion from chick brain tissue has been docu-
mented in two laboratories, the mechanism of the effect and physiological
significance are still highly uncertain. These points are addressed in more
detail in Unresolved Questions.
5.7.6 Unresolved Questions
No report has yet described a mechanism of action in sufficient detail to
identify all the relevant conditions necessary and sufficient to unequivocally
explain the calcium ion response in the brain or pancreas. The response to
specific intensities is unusual and presently unexplained. Any critical
5-237
-------
evaluation of this phenomenon, particularly in brain tissue, must also con-
sider corollary supporting studies with brain tissue from cats (Bawin and
Adey 1977), cat brain tissue i_n vivo (Kaczmarek and Adey 1973), and specific
brain rhythms in cats (Bawin et aj. 1973), as well as behavioral changes in
monkeys (Gavalas-Medici and Day-Magdaleno 1976). This response to modulation
of an RF-carrier wave or to sub-ELF signals alone may be a true field effect
at a very low SAR and at biologically relevant frequencies, i.e., in the range
of frequencies normally present in the electroencephalogram (EEG).
Other areas where there are unanswered questions is in comparisons of CW
vs. PW microwaves under identical exposure conditions. Such studies would
help determine if the differences seen by Wangemann and Cleary (1976) were
due to different exposure conditions or to the irradiation parameters (CW or
PW).
There is also a paucity of information on the effects of RF radiation at
different frequencies, particularly at frequencies of environmental impor-
tance. Studies at different frequencies would help to determine the reasons
for differences in effects at similar SAR's. Such studies might help explain
why Wangemann and Cleary (1976) reported serum chemistry changes in rabbits
at 0.8 W/kg (2450 MHz), and why Lovely et al^. (1977) reported no change in
serum chemistry values in rats at 1 W/kg (918 MHz).
There are also data such as those by Boggs et al^. (1972) where the
results from microwave heating to a predetermined temperature are different
from those resulting from the same temperature produced by other means of
5-238
-------
heating. Perhaps there are differences in the uniformity of heating or the
rate of heating that would account for these differences. Additionally, a
study by Deficis et aK (1979) reported elevated serum triglyceride and
P~1ipoprotein levels in mice exposed to 2450-MHz microwaves at power densities
2 2
of 1.5, 3.3, and 4 mW/cm , but not at 1 mW/cm . Because the exposures were
conducted in a multimodal cavity, SAR values were not reported and cannot be
predicted. If this study is repeated, particular attention should be given
to dosimetry. An alternative is to make or report dosimetric measurements in
the exposure system used.
The reported effects of thyroid function at 3.75 W/kg for 60 h contrasted
with no effect at 6.25 W/kg for 16 h (Parker 1973) suggests that the total
amount of energy absorbed may also be an important consideration. Additional
studies could define further the relative importance of dose rate com-
pared with total dose.
5-239
-------
5.8 GENETICS AND MUTAGENESIS
Carl F. Blackman
5.8.1 Introduction
Genetics is the branch of biology that deals with the heredity and variation
of organisms. The biochemical basis of heredity lies in the sequence of bases
found in the nucleic acids, deoxyribonucleic acid (DNA), and in a few cases,
ribonucleic acid (RNA). All living cells have the biochemical machinery to
detect the sequence of bases in the DNA and to transcribe sections of DNA in-
formation into similar sequences of bases in RNA. The RNA molecules then
move to other locations inside the cell where their information is translated
into various series of amino acids joined together in sequences that were
precisely defined in the original DNA molecule. These specifically arranged
amino acids form proteins, some of which are enzymes. Enzymes, in turn,
catalyze biochemical reactions that ultimately result in the growth and
propagation of intact organisms. Hereditary (genetic) material can be either
nucleic acids alone, as found in bacteria and viruses, or a nucleic acid in
association with proteins, which form the chromosomes found in more complex
organisms, including man.
Heat is a physical agent that can disrupt genetic material by causing
the temperature to rise above a normal physiological range. The effect of
heat, or temperature rise, has been studied in many biological systems. As
examples, heat has been shown to cause: physicochemical damage in isolated
DNA preparations (Lindahl and Nyberg 1974; Ginoza et a2_ 1964; Ginoza and
5-241
-------
Miller 1965) and in bacteria (Pellon et ah 1980); chromosome changes in
Drosophila melanogaster (Grell 1971) and in the locus (Buss and Henderson 1971),
including a change from dipolid to haploid in pollen from maize (Mathur et ah
1980); enhanced sensitivity to other agents in mammalian cell cultures (Ben-Hur
et ah 1974) and in D. melanogaster (Mittler 1979); reduced fertility in rats
(Bowler 1972; Fahim et ah 1975); and mutations in bacteriophages (Bingham et ah
1976), in bacteria (Zamenhof and Greer 1958) and in D. melanogaster (Muller and
Altenburg 1919). Because absorbed RF energy is usually dissipated as heat, all
reports of genetic and mutagenic changes caused by exposure to RF radiation must
be examined closely to determine whether temperature rise, or some other mechanism,
is the causative agent.
Mutations are relatively permanent changes in the hereditary material
involving either a physical change in chromosomal relations or a biochemical
change in the sequence of nucleic acid bases that make up the genes. These
changes can be passed on to future generations of cells. Two different types
can be affected: germ cells, which are egg or sperm or their antecedent cells;
and somatic cells, which form all other tissues in the body. Mutations in
germ cells can be passed on to future generations of the organism, whereas
mutations in somatic cells of an organism may lead to impairment of organ
function and, in the extreme case, to cancer. Most mutations are considered
harmful because they disturb the biochemical processes necessary for the sur-
vival of the organism and for the propagation of the species.
Evaluation of the mutagenic potential of RF radiation has generally fol-
lowed the procedures established for testing the mutagenic activity of chemicals
5-242
-------
(Hollaender 1971). This evaluative scheme utilizes
• Molecular systems, to detect changes in the hereditary material
• Single-cell organisms, to detect changes in structure and func-
tion that are transmissible to the next generation
• Multi-cellular systems, including plants and animals, to detect
changes in reproductive potential
• Infra-human primates, to detect changes in reproductive potential
The evaluation scheme starts with simple, well-defined genetic systems that
enable rapid analysis and identification of suspected agents meriting further
evaluation. The more complex biological systems used in further testing
require substantially larger investments of time and resources.
As discussed below, the experiments reviewed in this section have demon-
strated that (1) RF irradiation of low-to-moderate intensity does not cause
mutations in biological systems, and (2) the only exposures that are potentially
mutagenic are those at high power densities of CW or PW radiation—exposures
that result in substantial thermal loading or in extremely high electric-field
forces within cells.
5.8.2 Effects on Genetic Material of Cellular and Subcellular Systems
5.8.2.1 Physical and Chemical DNA and Chromosome Studies --
Mechanisms for the absorption of RF radiation by DNA molecules are treated
elsewhere in this document (see § 3.2, Mechanisms of RF-Field Interactions
with Biological Systems; and § 5.1, Cellular and Subcellular Effects). This
5-243
-------
section will address the experimental studies that have examined changes in
genetic material induced by RF radiation. Purified DNA or DNA extracted from
the testes of exposed mice has been examined to determine if the physical
properties of the molecule can be altered by the exposure to RF radiation.
Hamrick (1973) examined the thermal-denaturation profiles of aqueous solutions
of isolated DNA exposed iji vitro at 37 °C to 2450-MHz (CW) radiation (SAR =
67 W/kg). No changes were found. Even elevated temperatures (to 50 °C)
produced by microwaves (1 h at SAR < 160 W/kg) did not cause a difference in
the thermal denaturation profile. (See § 5.1 for details.)
Varma and Traboulay (1976) reported radiation-induced changes in thermal
denaturation profiles (i.e., shift to lower temperature in the midpoint of the
transition curve [Tm] and reduction in the maximum hyperchromicity) as well
as changes in the base composition of testicular DNA extracted from anesthetized
mice whose testes were irradiated in the near field. Ten animals were exposed
2
individually either to 1.7-GHz radiation at 50 mW/cm for 30 min (SAR estimated
2
at 2.4 W/kg for testes alone) or at 10 mW/cm for 80 min (SAR estimated at
O
0.48 W/kg for testes alone), or to 0.985-GHz radiation at 10 mW/cm for 80 min
(SAR estimated at 0.26 W/kg for testes alone). The animals exposed to 1.7-GHz
2 2
fields at 50 mW/cm or to 0.985-GHz fields at 10 mW/cm were given a 1-day
recovery period before they were subjected to euthanasia, and the DNA was ex-
2
tracted. The animals exposed to 1.7 GHz-fields at 10 mW/cm were used in another
test for 8 weeks before euthanasia, and the DNA was extracted. Identical results
p
are reported for the 1.7-GHz exposure at 50 mW/cm in another publication (Varma
and Traboulay 1977); they are assumed to describe the same experiment rather than
a duplicate experiment because it is unlikely that all the experimental measure-
s' 244
-------
ments upon repetition would have given identical values. They conclude "that
biological damage may be due to the combined effect of thermal and nonionizing
radiation."
The experimental results reported by Varma and Traboulay must be examined
with careful attention given to the control experiments and to the potential
size of the thermal insult. Although the percent adenine/thymine is higher in
the DNA extracted from exposed animals, which could be responsible for the drop
in Tm, it is highly unlikely that identical values would be obtained for each
DNA extract from the three groups of control animals. Based on the variability
normally inherent in these biochemical measurements, a more probable explanation
is that the control values were a pool of all the control groups or represent
just one of those groups. Since the normal experimental variability in the
measurements is not described—either for repeated measures on one DNA-extract
preparation or between extract preparations—it is impossible to conclude that
the differences cited between control and exposed samples are significant; nor
can it be concluded that the differences result from the radiation exposure
directly, rather than from the extraction procedures used. Similarly, the crude
SAR values estimated for testicular exposures appear to be quite low to produce
the types of damage cited by these authors. A more likely explanation is that
the assumptions underlying either the SAR estimates or the exposure conditions
are in error, and that elevated temperatures produced by the radiation exposures
were the causative agent. (For example, it is possible that the temperature of
the absorbing material used to shield the animals was highly elevated, and it
may have elevated the testicular temperature.) In an anesthetized animal ex-
posed in the near field to 1.7-GHz radiation at 50 mW/cm for 30 min and shielded
5-245
-------
with urethane foam except for the testes (SAR estimated at 2.4 W/kg), a 1 to
2 °C rectal temperature rise was recorded following exposure (Varma and Traboulay
1977). In another report (Varma and Traboulay 1975), animals were exposed under
the same conditions, except that the exposure time varied between 30 and 40 min,
and the testes were examined histologically. The authors report, "the lumens
were empty with complete disintegration of spermatids, Sertoli cells and the
delicate connective tissue which surrounds the seminiferous tubules." In the
same report, with respect to animals exposed to 1.7-GHz fields at 10 mW/cm ,
the authors state: "there was little or no damage to testes, except when the
time of exposure was increased to 100 min, then severe changes in morphology
were observed." There was no explanation why little or no damage suddenly
became severe damage at 100 min of exposure. Anesthetized animals have been
shown to lack homeothermic capacity (Cairnie et aK 1980). It is probable that
2
even exposures to 10 mW/cm deposited sufficient energy in these anesthetized
animals to cause the temperature elevation responsible for many of the changes
that are attributed to microwave-specific effects.
Thus, an evaluation of the existing evidence from physical studies on DNA
indicates that RF irradiation at low-to-moderate intensities not accompanied by
temperature rise causes no changes in DNA bases, the fundamental unit of the
genetic code. However, if substantial elevations of temperature occur during
exposure, they may cause disruptions in the pairing of the two complementary
strands as well as other damage.
Other researchers have used cytogenetic techniques to examine some physi-
cal and chemical properties of chromosomes in intact cells to determine if the
5-246
-------
relationship of various parts of the genetic material is altered by RF-radiation
exposure. Huang et ah (1977) reported no RF-induced chromosomal aberrations
in white blood cells from Chinese hamsters exposed to 2450-MHz radiation at
power densities up to 45 mW/cm (SAR = 20.7 W/kg) for 15 min a day on 5 con-
secutive days. McRee et ah (1978) in a preliminary report found no sister
chromatid exchanges in bone-marrow cells of mice exposed to 2450-MHz fields at
20 mW/cm^ (SAR = 15.4 W/kg), for 8 h daily, 28 days total. Alam et ah (1978)
showed, in great detail, that chromosomal aberrations occur in a Chinese-
hamster-cell line (CH0-K1) exposed 30 min to 2450-MHz radiation from a diathermy
applicator, but only if the temperature of the culture is allowed to rise to
49 °C during exposure. These authors demonstrated that irradiation of cell
lines at high (> 200 mW/cm ) power densities (SAR estimated at 360 W/kg) would
cause no detectable cytogenetic effects, provided proper temperature control
was maintained. Thus, heating seems to account for the observed cytological
changes.
Authors of one detailed study used a frequency well below the 1.7- to
2.45-GHz range. McLees et ah (1972) exposed rats treated to undergo liver
regeneration either to 13.12-MHz CW (4.45 kVp_p/m) fields or to PW [44.1
kVp_p/m, 200-jjs pulse width, 50-Hz pulse repetition rate (PRR)] radiation for
28 to 44 h. They examined the effects of radiation on liver cell mitotic
activity, that is, the percentage of cells in mitosis, and the number of
chromosomal aberrations. Rat liver cells may be very sensitive to exogenous
stresses during this regenerative process because they are normally non-
dividing cells in an intact animal that have been challenged to divide in
situ. However, the authors found no RF-induced alterations in chromosomal
5-247
-------
morphology (SAR estimated at 1.2 to 1.3 W/kg). This result indicates that no
cytogenetic changes would be expected in this range of frequencies for low-
intensity exposures. Thus, although there are reports indicating that
exposure to RF radiation can cause cytogenetic changes, these changes probably
result from radiation-induced elevations of temperature.
5.8.2.2 Biological Studies of DNA and Chromosomes-
Bacteria have been used to study the mutagenic potential of RF radiation,
because the single-cell system is simple, easy to culture, quick to test, and
relatively sensitive to the action of mutagenic agents. By using the
bacterial system and the biological amplification it provides for any DNA
change, molecular biologists have been able to decipher the genetic code and
to identify a change in as few as one DNA subunit out of 10^ to 10"^
subunits. In contrast, biologists using standard physical techniques such as
DNA melting curves usually can detect changes in no better than 0.1 percent of
the DNA. Blackman et al_. (1976) exposed growing cultures of the bacterium
Escherichia coli either to 1.7- or to 2.45-GHz (CW) radiation for 3 to 4 h.
Exposure at 1.7 GHz was in the near field at 88 V /m or ~ 250 V„ /m (SAR =
rms p-p
3 W/kg). The 2.45-GHz exposures were in the far field at either 10 or 50
2
mW/cm (SAR = 15 or 70 W/kg, respectively). Although exposure of growing
cells provided enhanced sensitivity to mutagenic agents, no mutagenic activity
was detected. A positive control, ultraviolet (UV) light, caused mutations
and was used to demonstrate the sensitivity of the assay method. Dutta et aH.
(1979a) exposed growing cultures of various bacterial strains of Salmonella
typhimurium, commonly used in the Ames testing procedures to detect chemical
mutagens (Ames et aK 1975), to 2.45-GHz (CW) radiation for 90 min at
5-248
-------
20 mW/cm^ (SAR = 40 W/kg) and to 8.6-, 8.8-, 9.0-, 9.2-, 9.4-, and 9.6-GHz
2
(PW) radiation (1-ps pulse width, 1-kHz PRR) at 10 and 45 mW/cm average power
2
densities, and 10,000 and 45,000 mW/cm peak power densities. (The SAR at 45
2
mW/cm is estimated at 80 W/kg.) No mutagenic activity was observed under any
of these exposure conditions.
Another approach with bacterial systems is to test for radiation-induced
alterations in genetic processes, including cell death. Since most mutations
are detrimental, they might be detected indirectly by this method. Corelli et
al. (1977) exposed cultures of E. coli to microwaves at frequencies swept
between 2.6 and 4.0 GHz for 8 h (SAR = 19 W/kg). Although at 26 °C these
cultures were probably growing slowly, no change was noted in the number of
colony-forming units (CFU) in the cultures following irradiation, indicating
no detectable lethal events because of the exposure. These workers also
examined the infrared (IR) spectrum of these cells when exposed to 3.2-GHz
radiation for 11 to 12 h (SAR is either 21 or 16 W/kg). There was no
observable effect on the molecular or conformational structure of these cells,
in contrast to results obtained using a positive control, ionizing radiation.
Two strains of E. coli, one deficient in an enzyme needed to repair damaged
DNA, were tested (Dutta et a2- 1979b) for survival following microwave exposure.
The exposure conditions were,8.6-GHz (PW) radiation ( 1-jjs pulse width, 1-kHz
PRR, SAR = 12 W/kg) for 1, 2, 4, or 7 h. There was no significant change in
the relative growth patterns of these strains that could be attributed to
microwave-induced DNA damage repairable in one strain, but not in the other.
Blackman et ak (1975) exposed a different strain of E. coli in log phase
(actively dividing) and in lag phase (undergoing metabolic activities pre-
5-249
-------
paratory to division) at 32 °C for 4 h to 2.45-GHz radiation at 0.005, 0.5,
5.0, or 50 mW/cm^ (at 50 mW/cm^ power density the SAR = 75 W/kg). Additional
experiments were conducted at 5 mW/cm and 25 °C to test for the influence of
cold stress, and in two-culture media at 30 and 35 °C to compare the relative
influence of a rich medium with a minimal medium, the latter requiring greater
utilization of the genetic apparatus of the cell for growth to occur. They
found no change in the colony-forming ability of the cultures due to the
2
exposure except for enhanced growth at 50 mW/cm , which was attributed to
slight temperature rises in the exposed cultures.
Few researchers have used single-cell systems more complex than bacteria
specifically to look for RF-radiation-induced mutagenesis. Dutta et a_[.
(1979a) exposed a diploid strain of the yeast Saccharomyces cerevisiae, a
2
primitive eukaryote, to 2.45-GHz (CW) radiation for 2 h at 20 mW/cm (SAR =
40 W/kg). They found essentially no change in the number of mutations at
either of two loci affecting the nutritional requirements for adenine or
tryptophan. These investigators conducted additional tests at 8.5-, 8.6-,
8.8-, 9.0-, 9.2-, 9.4-, and 9.6-GHz (PW) radiation (1-ps pulse width, 1-kHz
PRR) for 2 h at average power densities of 1, 5, 8.9, 10, 15, 30, 35, 40, or
2
45 mW/cm . Although no measurements of SAR are cited, thereby making com-
parisons with CW exposures difficult, the highest power density was reported
to raise the temperature of the culture by 12 °C, indicating substantial
2
absorption of the radiation. (At 45 mW/cm power density the estimated SAR =
80 W/kg). In no case did the exposures cause a change in the frequency of
genetic events, altering the requirements for either adenine or tryptophan, in
the treated population as compared with the control population. Dardalhon et
5-250
-------
al. (1980) exposed a diploid strain of S. cerevisiae at 20 °C in the near
field of 9.4-GHz (CW) radiation (SAR estimated at < 2) for 1 to 5 h, or in the
near field of 17-GHz (CW) radiation at either of two power densities (SAR's
estimated at 28 or < 6 W/kg, respectively) for various times to 24 h. No
significant changes attributable to the irradiation were observed in the
percent survival, in the induction of cytoplasmic "petite" mutations, in the
induction of mitotic recombinations, or in sporulation. Thus, no changes were
detected in complex single-cell systems used to examine directly the mutagenic
potential of microwaves.
Some work has been done with bacteria and yeast cultures that compares
the lethal and mutagenic effects of microwaves with those induced by con-
ventional heating. Dutta et aK (1980) examined the responses of various
strains of the bacteria S. typhimurium and E. coli and those of a diploid
strain of the yeast S. cerevisiae, exposed to 8.6-, 8.8-, or 9.0-GHz (PW)
radiation (1-kHz PRR, l-|js pulse width) at average power densities to
45 mW/cm (SAR estimated maximum at 80 W/kg). The bacteria were exposed
90 min at an ambient temperature of 37 °C, whereas the yeast was exposed 2 h
at 30 °C. The results of these irradiation treatments were compared with
those obtained from treatments at elevated temperatures produced by conven-
tional heating. The comparison indicated that conventional heating could
produce cellular damage Reading to reduced survival in a manner similar to the
changes caused by microwave-induced elevations of temperature (to 10 °C above
normal growth temperatures). No mutational events occurred in bacteria to
42 °C, or in the yeast to 40 °C; however, at 45 °C, a slight increase in
mutational events occurred in the yeast, and at 47 °C, in S. typhimurium.
5-251
-------
Dardalhon et

-------
pectively) and mated each of the surviving males individually with two virgin
females during 15 consecutive days. No changes were observed in the genera-
tion time or sex-ratio pattern of the offspring. These offspring were
then mated and observed for possible sex-linked lethal mutations in their
offspring. No such mutations were found at frequencies > 1 percent, a
detection limit based on the small number of chromosomes (< 800) actually
evaluated. An additional caveat by the authors was that the most sensitive
stages to detect recessive sex-linked lethal mutations, late spermatocytes and
early meiosis, were not tested in their study, because few sperm were still in
meiosis at the time of exposure. Mittler (1976), who exposed adult males from
various strains of D. melanogaster to 29-MHz (CW) fields at 600 Vrms/m (SAR
roughly estimated at 0.024 W/kg) and to 146-MHz (CW) fields at 62.5 Vrms/m
(SAR roughly estimated at 0.015 W/kg) for 12 h, mated them with virgin
females for 12 h every 2 days in production of 4 or 5 broods. In these
experiments, brood 4 was produced from sperm irradiated "in or about meiosis."
There were no mutations induced by these treatments as evidenced by the lack
of chromosome loss, nondisjunction, or sex-linked recessive lethals. In
addition, Mittler (1977) observed no mutagenic effects (recessive lethals)
when exposing adult females to 98.5-MHz frequency-modulated fields (composed
of standard commercial broadcast audio frequencies) at 0.3 V /m (SAR
^	rms
estimated at 0.0004 W/kg), 134 h per week for 32 weeks. Although the results
with D. melanogaster are difficult to extrapolate to the human condition, this
test system as a qualitative index was not able to detect any mutagenic
alterations by RF radiation over a wide range of frequencies.
5-253

-------
5.8.3.3 Vertebrates—
Experiments with vertebrates address the impact of the genetic effects of
RF radiation most directly because of the general biological similarities
between man and other vertebrates. However, there are very few such experi-
ments directed at this subject. Three studies have been reported that are
directly focused on the mutagenic potential of microwaves in mice and rats.
Varma et a^. (1976) used the dominant-lethal test to investigate the effects
of 2.45-GHz (CW) radiation. The testes of anesthetized mice were treated
either by a single 10-min exposure at 100 mW/cm (SAR estimated at 11.4 W/kg),
or by three exposures of 10 min each at 2-h intervals on the same day at
2
50 mW/cm (SAR estimated at 5.7 W/kg), or by four exposures of 10 min each at
o
3-day intervals over 2 weeks at 50 mW/cm (SAR estimated at 5.7 W/kg). Varma
and Traboulay (1976) in a similar study exposed the testes of anesthetized
2
mice to a single dose of 1.7-GHz radiation, at 10 mW/cm for 80 min (SAR
estimated at 0.48 W/kg) or at 50 mW/cm^ for 30 min (SAR estimated at 2.4 W/kg).
Following a 24-h recovery period, the males were bred with virgin females—one
group of females per week for 6 to 8 weeks. In the 2.45-GHz study, the authors
used the week-6-group results for comparison; they found a higher mutagenicity
2
index in the results of the 100-mW/cm group bred at week 1, and in the results
2
of the 50 mW/cm group exposed three times in one day and bred at week 4. The
authors concluded that a single, intense exposure or multiple exposures during
one day induced a number of significant mutations in the mice. The results of
the 1.7-GHz studies at 50 mW/cm for 30 min indicated a radiation-induced
increase in infertility, in pre-implantation losses, and in the mutagenicity
index of the groups bred at 3, 4, 5, and 6 weeks. In addition, for the group
2
exposed to 1.7-GHz fields at 10 mW/cm for 80 min, an increase was seen in the
5-254

-------
mutagenicity index of the groups bred at 1, 2, 3, and 6 weeks. The authors
concluded that the "biological damage may be due to the combined effect of
thermal and nonionizing radiation." Both of these reports should be evaluated
on the basis described in § 5.8.2.1, Physical and Chemical DNA and Chromosome
Studies, for the physical and chemical studies reported by these authors. It
is not possible to determine the extent of the biological damage caused by
elevated temperature in either of these reports, especially because anesthetized
mice lack an effective temperature-regulation capability (Cairnie et al^
1980). However, if the histological damage found in the testes of animals
exposed under similar conditions (Varma and Traboulay 1975) is reflected in
damage to sperm, it could account for many if not all of the changes in the
mutagenic index, fertility, and pre-implantation loss reported by those authors
in later studies.
The conclusion that radiation in this frequency range could cause mutagenic
changes in mice was challenged in a detailed study by Berman et aK (1980),
who exposed unanesthetized male rats to 2.45-GHz (CW) radiation under three
treatment regimens: 4 h/day from day 6 of gestation to 90 days of age at
2
5 mW/cm (SAR varied from 4.7 W/kg to somewhat less than 0.9 W/kg at day 90,
because growth of the animals changed their energy absorption efficiencies),
2
5 h/day for 5 days beginning on day 90 at 10 mW/cm (SAR estimated at 2 W/kg),
2
and 4 h/day, 5 days/week, for 4 weeks beginning on Day 90 at 28 mW/cm (SAR
estimated at 5.6 W/kg). During selected weekly periods after treatment, the
exposed males were bred to untreated females that were examined in late pregnancy
by the dominant-lethal assay. No significant germ-cell mutagenesis was detected
under any treatment condition, even though significant increases in rectal and
5-255

-------
2
testicular temperature were observed during the 28-mW/cm exposure and were
associated with a concomitant decrease in incidence of pregnancy during some
of the breeding periods, which indicates temporary sterility. In addition,
these authors reexamined the data of Varma and concluded they had been
interpreted incorrectly because the effects of litter size on fetal mortality
had been ignored and the differences between treated and control groups had
been overemphasized when the control values were not representative of normal
values. Berman et aK (1980) concluded "it still remains to be demonstrated
that microwaves, even at near-lethal doses, can cause a dominant lethal mutagenic
effect in the mouse or the rat." Thus, well designed experiments with biological
systems having complex genomes similar to man's have demonstrated no mutagenic
effects from low- to moderate-intensity RF radiation when the temperature of
the biological system is maintained in the physiological range. Most of the
reported changes are consistent with the effects expected from elevations in
temperature, caused by CW exposure to high-intensity RF radiation. The other
changes cited in this section may result from incomplete experimental design.
5.8.4 Unresolved Questions
\
Heating, which raises the temperature of biological samples above normal
physiological ranges and may result in genetic and mutagenic changes, is a
well known result of exposure to high-intensity RF radiation. The choice of
exposure conditions to study the genetic and mutagenic effects of moderate- to low
intensity radiation is arbitrary because no other well defined mechanism of RF
interaction has been developed to address this biological problem. The
experimenter must select the frequency, intensity, and duration of exposure, as
5-256

-------
well as the type and characteristics of modulation. To compound the problem
further, there are an almost limitless number of biological systems that could
be selected for study. There are several guides that can be developed from
recent experimental findings to help delineate a future experimental plan.
Although the conclusion has been stated above that no strong evidence
exists in the cited reports to demonstrate mutagenic activity for low intensity
RF radiation per se, it should be qualified by the following observations:
1.	These experiments are limited to small portions of the RF
region of interest, i.e., 0.5 MHz to 100 GHz. This 106-Hz
frequency range is so large that if detrimental effects occurred
over very narrow frequency ranges, they might have gone undetected.
2.	Many of the experiments were conducted with CW radiation. If
effects exist, they may occur with amplitude-, frequency-, or
PW-modulated radiation at frequencies associated with ongoing
biological or biochemical processes. An exposure to a
short-duration, high-intensity PW radiation (radar-like) may
not produce the same biological response as a CW exposure of
the same average power density. Thus, the concept of averaging
the intensity of PW radiation over time to determine an effective
intensity may not be generally valid.
3.	The general approach in most experimental studies has been to
assume that the higher the radiation intensity, the more pro-
nounced the biological effect. It is possible that several
mechanisms are operative during radiation exposure and that
mechanisms with thresholds at relatively high intensities
(e.g., heating) mask the effects being caused by the more
subtle mechanisms.
4.	RF radiation, as a co-stressor, may enhance the effects of known
mutagenic agents. In environmentally relevant situations,
biological systems are exposed simultaneously to many agents,
both chemical and physical.
5.	Since no mechanism has been developed to describe the interaction
of RF radiation with complex biological structures and genetic
material, it could be argued that standard procedures that have
been optimized to identify ionizing and chemical mutagens are
not necessarily applicable to this form of radiation. Thus, in
view of this lack of knowledge, no test system should be overlooked
5-257

-------
as potentially sensitive for detection of mutagenic activity.
For example, plant systems have been shown to be extremely
sensitive indicators of environmental pollution.
Recent experimental developments support some of the suggestions made
above. Possible frequency-specific responses were reported by Smolyanskaya
and Vilenskaya (1973), who described the induction of colicin in E. coli by
extremely high-frequency radiation, near 37 GHz, at power densities of 0.001,
2
0.01, 0.1, or 1.0 mW/cm for 0.5, 1, and 2 h. Colicin is a protein usually
not made by the cell. Its induction, while not well understood, requires the
transcription of a new portion of the DNA molecule. This induction demonstrates
that specific microwave frequencies can induce changes in the genetic processes
of cells. More recently, Grundler et al_. (1977) described frequency-dependent
growth responses in the yeast S. cerevisiae exposed for periods to 11 h to
2
microwave frequencies between 41 and 42 GHz at 1 to 3 mW/cm (average SAR
estimated at 4 to 11 W/kg). (See § 5.1, Cellular and Subcellular Effects, for
details.) Although this work is controversial, it supports the frequency-
specific-response concept.
The importance of examining the effects of RF radiation as a co-stressor
with other agents is emphasized in the report by Dardalhon et al. (1980) that
was discussed earlier with respect to exposure either to heat or to microwave
radiation. These authors exposed haploid and diploid strains of S. cerevisiae
2
for 1 h to 17-GHz (CW) radiation, near field, at 50 mW/cm (SAR estimated at
< 6 W/kg) and 20 °C, followed by exposure to various doses of UV light
(254 nm), which is known to cause damage directly to the DNA. Both haploid and
diploid strains were affected. The exposure combination caused a tendency—
5-258

-------
but apparently not a statistically significant change—toward diminished
survival, increased mitotic recombination, and increased cytoplasmic "petite"
mutations, when compared with the effects caused by exposure to the same
doses of UV light alone. This microwave-radiation enhancement of effects
due to UV-light exposure is unusual, because enhancement was observed only at
high doses of UV light, where, presumably, additional repair systems are
operative. Similar trends were obtained by replacing microwave-radiation
exposure with temperature elevation to 46 °C before treatment with UV light.
No influence from exposure to microwave radiation, followed by UV light, was
7	2
observed for 17-GHz fields at 2.5 mW/cm , or for 9.4-GHz fields at 5 mW/cm .
Because the microwave-radiation treatment caused only a 1 °C global temperature
rise in the sample that was 10 °C below its normal growth temperature, the
authors reasoned that if temperature rise is the basic agent enhancing the
UV-1ight-induced response, it must be occurring selectively at specific sites
within the biological system other than on the DNA itself, causing change
perhaps in metabolic processes or in the structural integrity of cytoplasm, or
of membranes. In this case, there may be a differential sensitivity among the
various repair systems induced by the UV light. This type of investigation
potentially can reveal biological processes, including repair of damage or
detoxification of chemicals, that may be particularly sensitive to microwave-
radiation exposure.
In another area, radiation-induced efflux of calcium ions from jm vitro
brain tissues has been shown to depend on the modulation frequency of a carrier
wave (Bawin et aK 1975; Blackman et a^. 1979). The effective modulation
frequencies were in the same frequency range as the natural biological rhythms
5-259

-------
associated with the EEG in the intact animal, and thus the radiation may have
coupled with an existing oscillatory system. Further studies by Blackman et
al. (1979; 1980a,b) and Sheppard et aK (1979) indicated that higher and lower
intensities of radiation can lead to the disappearance of the effect. This
suggests that examining biological processes only at a single power density
may provide misleading information.
Three additional, incomplete reports have been included in this section.
These studies are included either because they present positive findings at
frequency ranges or intensities seldom studied or because of a strong claim of
the authors. Heller (1970) and Mickey and Koerting (1970) reported chromosomal
aberrations in cultured Chinese-hamster-1ung cells exposed for 30 min to 19-
or 21-MHz fields, although no changes were observed after exposure to 15- or
25-MHz (PW) fields (100-Hz PRR, 50-ps pulse width) at field intensities to
300 kVp_p/m. It is very likely either that the extremely high peak voltages
(up to 300 kVp_p/m) were producing intense, rapid heating within the cells, or
that the field per se was causing major stresses within the system. These
reports provide insufficient descriptions to allow an estimate of the SAR
values, and, thus, these results cannot be compared with those found by others.
It is possible that fields at extremely high intensities can account for some
cytogenetic changes. Manikowska et a^. (1979) reported a non-dose-dependent
increase in chromosomal translocations and in chromosomal pairs remaining as
univalents at Metaphase I in the sperm cells of mice exposed 1 h/day, 5 days/
week, for 2 weeks to 9.4-GHz (PW) radiation at 200-, 1000-, 2000-, or
20,000 mW/cm^ peak power densities (0.5-ps pulse width; 1-kHz PRR; 0.1-, 0.5-,
1-, and 10-mW/cm average power densities, respectively). (Average SAR is
5-260

-------
roughly estimated at 5 W/kg, and peak SAR roughly at 9000 W/kg for the highest
power density used.) The authors do not describe the relation of the animals'
testes to the incident field, thereby raising the question of the actual
quantities of energy coupled to the target cells at 9.4 GHz, where tissues
exhibit large attenuation coefficients. A description of the environmental
conditions during exposure is also absent. These omissions essentially
prevent any critical assessment of the results. Furthermore, because of the
small number of animals used in the study, the authors state that "the
findings ... obviously need confirmation on larger numbers of animals."
In another study, based on bacteria exposed to extremely high temperatures,
an incomplete analysis of the distribution of temperatures within the samples
led to an unjustified conclusion. Blevins et al^. (1980), who exposed several
strains of S. typhimurium commonly used in the Ames testing procedures (Ames
et al. 1975) to 2.45-GHz radiation in a microwave oven at a calculated power
density of 5100 mW/cm for periods of 2 to 23 s, examined the cultures
for lethality and for mutation induction. These exposures caused extremely
large elevations of temperature. Corollary heating experiments were performed
in high-temperature water baths to determine the extent to which temperature
change alone contributed to lethality and mutations. Their conclusion that
microwave radiation is a potent mutagen was not supported by their data,
because the authors failed to demonstrate that the uniformity of microwave
heating was duplicated by the water bath experiments. The authors acknowledge
that "differences in the kinetics of water bath and microwave heating are
possible" but fail to evaluate this possibility further. Because of the
large temperature increases in their culture systems—apparently as large as
5-261

-------
46 °C in 14 s, more definitive temperature distribution work must be done
to establish accurately the contribution that temperature change makes to
induction of mutations before additional mutagenic properties are assigned to
microwaves. Blevins et aK (1980) reported using a power density 500 times
greater than the current U.S. occupational guidelines, and, because the apparent
mechanism is based on such high, nonphysiologic temperatures, the authors'
conclusions are not applicable to the other cited studies, which have used
2.45-GHz radiation. However, the study may support a reevaluation of the con-
cept that brief, high-intensity exposures can be averaged over a longer time
period to define the average intensity. For example, an exposure to 5100
2	2
mW/cm for 2 s is 28 mW/cm if averaged over 6 min, and an exposure for 14 s
2
becomes 200 mW/cm if averaged over 6 min. The Blevins et cfL study would
also support the concept that high-intensity pulses of microwaves may affect
biological systems differently from convection or conduction heating.
Although no solid evidence exists (i.e., there are no independently
verified reports) to indicate that low-intensity RF radiation is mutagenic,
sophisticated concepts and designs are just beginning to appear in experiments
concerning the biological effects of RF radiation. Thus, although it is pre-
mature to proclaim unequivocally that RF radiation is not mutagenic, the vast
majority of evidence presently indicates that in the absence of temperature
elevation this statement is true. The above studies of the biological effects
of RF radiation that relate to genetics and mutagenesis are summarized in
Table 5-21.
5-262

-------
TABLE 5-21. SUMMARY OF STUDIES CONCERNING GENETIC AND MUTAGENIC EFFECTS OF RF-RADIATION EXPOSURE
Exposure Conditions
Effects
Species
F requency
(MHz)
Intensity Duration
(mW/cm2)	(days x rain)
SAR
(W/kg)
Reference
tn
i
NJ
CTi
U>
No change in thermal denaturation profile,
except at elevated temperature
Change in thermal denaturation profile and
hyperchromicity of ONA extracted from testes
following exposure
No chromosome aberrations in white blood
eel Is
No sister chromatid exchange in bone marrow
cells
No chromosome aberrations in CH0-K1 cell
line if temperature maintained
No chromosome aberrations or change in
mitotic activity in regenerating liver
cells in rat
No mutation induction
No mutation induction observed in Ames
tester strains
DNA
House
2450 (CW)
1700 (CW)
98b (CW)
Chinese hamster 2450 (CW)
House	2450 (CW)
Chinese hamster 2450 (CW)
Rat
13 12 (CW)
13 12 (PW)
£ coll	2450 (CW)
i coTT	1700 (CW)
S t.yphimurmm 2450 (CW)
8600-9600 (PW)
134
< 50
10
5-45
20
< 200
1 x 960
1 x 80
5 x 15
28 x 480
1 x 30
4 45 kV /mix 1680-2640
44 1 kVP p/ra
p-p
10 or 50	1 x	180-240
250 V /m	1 x	210
p-p
20	1 x	90
10, 45	1 x	90
67 (est)	Hamrick (1973)
<24 (est) Varma and Traboulay
< 0 26	(1976, 1977)
21	Huanq et (1977)
15	McRee et a^ (1978)
< 360 (est)	Alara et a^ (1978)
1 3 (est)	McLees et a] (1972)
15 or 70
3
40
18, 80
(est)
81ackman et a^
(1976)
Dutta et al (1979a)
(continued)

-------
TABLE 5-21. (continued)
Effects
Species
Frequency
(MHz)
Exposure Conditions
Intensity
(mW/cm2)
Duration
(days * nin)
SAR
(W/kg)
Reference
Reduction in survival concomitant with rise
in sample temperature
E col i
8600-9000 (PW)
typhi murium 8600-9000 (PW)
cerevisiae 8600-9000 (PW)
1-20
<	45
<	45
1 x 90
1 x 90
1 x 120
> 50
<	80
<	80
Outta et al (1980)
No reduction in survival or mutational events S cereytsiae
Ui
ro
CTi
No detectable lethal	events due to no change
in CFU's
No observable change in molecular structures
because no change in IR spectrum
No repairable DNA damage
No change in growth pattern,
enhanced colony-forming activity
No change in mutation frequencies at either
of two loci controlling requirements for
adenine and tryptophan
No mutagenic effects in exposed embryos
No changes in generation time, sex ratio,
or sex-linked lethal nutations in offspring
No mutations in adult males as evidenced by
chromosome loss, nondisjunction, or sex-
1 inked recessive lethals
coll
coll
co) 1
co) i
cerevisiae
9 4 (CW)
17 (CW)
2600-4000 (CW)
3200 (CW)
8600 (PW)
2450 (CW)
2450 (CW)
8500-9600 (PW)
0 melanoqaster 2450 (CW)
D melanoqaster 2450 (CW)
0 melanoqaster 29 (CW)
146 (CW)
0 005-50
20
1-45
4600-6500
1 x 300
1 x 1440
1 x 480
1 x 660-720
1 x 60-420
1 x 240
1 x 120
1 x 120
1 x 360
1 x 45
< 28 (est)
19
21 or 16
12
0 008-75
40
< 80 (est)
100
Oardalhon et al
(1980)
Corel 1 i et aj[
(1977)
Corelli et al
(1977)
Outta et al^ (1979b)
Blackman et a^
(1975)
Outta et al (1979a)
Hamnerius et aj
(1979)
600 V /m 1 x 720
62 5 /m 1 x 720
ruts
150-210 (est) Pay et al (1972)
0 024 (est) Mittler (1976)
0 015 (est)
(continued)

-------
TABLE 5-21. (continued)
Exposure Conditions
Effects
Species
Frequency
(KHz)
Intensity Duration
(rtf/cra2)	(days x mm)
SAR
(W/kg)
Reference
cn
i
rvj

CP
No mutagenic changes (recessive lethals) in
adult females
No significant germ-cell mutagenesis in
weekly breedings
No significant germ-cell mutagenesis in
weekly breedings
Same, except decrease in pregnancies,
indicating temporary sterility caused by
elevated testicular temperatures
Induction of a repressed protein, colicin,
indicating a change in the genetic processes
Change in growth rate that was very fre-
quency specific, indicating an alteration in
the processes of the cell
Chromosome aberrations in lung cells hi
vitro at two frequencies but not at two
closely related frequencies, 15 or 25 MHz
Increase in chromosome translocations in
sperm eel Is
Increased mutations and lethality
0 aelanoqaster 98 5 (CW)
Rat	2,450 (CW)
Rat	2,450 (CW)
Rat	2,450 (CW)
E col i
Chinese
hamster
37,000 (CW)
S cerevisiae 41,000-42,000
(CW)
19 (PW)
21 (PW)
Mouse	9,400 (PW)
S typhimurium 2,450 (CW)
0 3V /m, 224 x 1140
(FM) £?$audio
5	106 x 240
10
28
0 001-1
1-3
up to
300 kV /m
p-p
0 1-10
5100
5 x 300
20 x 240
1 x 30-120
1 x 660
1 x 30
10 x 60
1 x 0 03-0 48
0 0004 (est)	Mittler (1977)
4	7-0 9	Herman et a]^ (1980)
2	Berman et a| (1980)
5	6	Berman et al (1980)
Smolyanskaya &
Vilenskaya (1973)
4-11	Grundler et a^
(est, av)	(1977)
Heller (1970)
Mickey and
Koerting (1970)
0 05-5 (est) Manikowska et aj
(1979)
Blevins et aj
(1980)
*est = estimated, av = average

-------
5.9 LIFE SPAN AND CARCINOGENESIS
William P. Kirk
5.9.1 Life Span
There are few data (Table 5-22) and no definitive studies on which to judge
the effects of RF-radiation exposure on human longevity. Pazderova-Vejlupkova
and Josifko (1979) mention some previous studies in which various clinical para-
meters were examined in 82 employees of radio transmitting stations (0.3 to 30
MHz; 14.4-year mean exposure; mean intensity = 80 V/m) and 58 employees of TV
transmitting stations (30 to 300 MHz; 7.2-year mean exposure; mean intensity =
2.9 V/m). No signs of damage were found in these personnel; however, the
study did not directly address mortality. More recently, information has been
developed on the mortality experience of U.S. Government employees assigned to
the Moscow Embassy during the period 1953 to 1976, when Soviet microwave
irradiation of the U.S. Embassy was taking place. Comparisons were made with
a group of employees who had been stationed at other U.S. Embassies in Soviet
Bloc cities (Budapest, Leningrad, Prague, Warsaw, Belgrade, Sophia, and Zagreb).
The comparison group was chosen to be as similar as possible to the 1800
employees in the Moscow group for selection (i.e., posting) criteria and
environmental influences, except that the posts were not subject to microwave
exposure (Lilienfeld et al^ 1978). Exposures to 2.56- to 4.1-GHz radiation at
2
a maximum power density of 5 pW/cm were documented from August 1963 to May
1975; the frequency range was expanded to 0.6 to 9.5 GHz in May 1975, and the
2
maximum power density was increased to 18 pW/cm from August 1975 to February
1976. The duration of exposure ranged from 0.5 h/day to 20 h/day (United
5-267

-------
TABLE 5-22. SUMMARY OF STUDIES CONCERNING RF-RADIATION EXPOSURE EFFECTS ON LIFE SPAN/CARCINOGENESIS
Exposure Conditions
Effects
Species
Frequency
(MHz)
- Intensity
(mW/cm2)
Duration
(days x nin)
SAR
(W/kg)
References
No effect on life span or cause
of death
Human adult
male & female
2560-4100
600-9500
0 005 (max)
0 01B (max)
8030* x 600
180 X 1200
-4
2 x 10 . (max)
7 x 10 (max)
Lilienfeld el al^ (1978)
No effect on mortality in a
military population followed
for 20 years
Human adult
male
200-5000
(est) (PW)
~ 1
(routine)
100
(occasional)
730 x 480
(est)
<0 05 (est)
<5 (est)
Robinette et (1980)
Slight increase in mean life span
Adult mouse
800
43
175 x 120
12 9 (est)
Spalding et al^ (1971)
Increased mean and maximum life
span for "irradiated mice with
tumors " increased mean life span
but no change in maximum life span
of non-tumor-bearing mice Delay
in development of tumors in
irradiated mice but no change in
ultimate number of tumors
Infant mouse
2450

4 x 20
(In utero)
35
Preskorn et aj[ (1978)
Increased mean life span in
irradiated mine (concurrent
infection-pneumonia)
Adult mouse
9270
(PW)
100
4	4 mm/day
5	days/week
59 weeks
40 (est)
Prausnitz and Susskind
(1962)
*The duration of 8030 days equals the number of years (22) of irradiation of the embassy and length of the study period, but the
average exposure of individuals is estimated to be 2 to 4 years

-------
States Senate Committee on Commerce, Science, and Transportation 1979). No
evidence was found that the Moscow group had experienced any higher mortality
or any differences in specific causes of death up to the time of the report.
However, investigators noted that, because the population was relatively
young, it is too early to detect long-term mortality effects except for
those serving in the earliest period of the study. This important study is
covered in more detail in § 5.10 Human Studies.
Robinette et al_. (1980) examined the records of a group of 40,000 U.S.
Navy personnel who enlisted during the period 1950 to 1954. Approximately
20,000 had job classifications with maximum potential for exposure to radar
(i.e., electronics technicians, fire control technicians, and aviation elec-
tronics technicians). A similar number of radio and radar equipment operators
believed to have a minimal potential for exposure were used as a comparison
group. No specific data on exposure or frequencies were provided; however,
the authors state that the low exposure group of radio and radar operators
2
were exposed to levels well below 1 mW/cm , whereas the high exposure occu-
2
pations involved average exposure levels below 1 mW/cm during duty hours
combined with infrequent higher exposures, which on occasion may have exceeded
2
100 mW/cm . The authors found no apparent difference in the long-term
mortality patterns between the two exposure groups more than 20 years post
exposure. This study is also covered in more detail in § 5.10.
Although there have been many experiments to determine the lethal effects
in animals exposed to high levels of RF radiation, data on life span effects
of experimental animals exposed to low power levels are scarce. To our knowledge
5-269

-------
there is only one report of a study in which animals were exposed to microwave
radiation as the sole stressor and observed throughout their life span.
Spalding et al_. (1971) exposed 24 adult female mice to 800-MHz fields at a
2
power density of 43 mW/cm , for 2 h/day, 5 days/week for 35 weeks. A waveguide
apparatus was used to irradiate the animals, which were cooled by forced air.
No temperature or relative humidity data are reported. The whole-body-averaged
SAR is estimated to be 12.9 W/kg. Four exposed mice died of "thermal effects"
during the experiment, and a fifth is reported to have died during exposure
because it became too obese for the exposure chamber. The mean life span of
the remaining 19 exposed mice was 664 ± 32.2 (SEM) days, while that of the 24
sham-irradiated mice was 645 ± 32.2 (SEM) days.
Prausnitz and SUsskind (1962) studied pathological and longevity effects
on male Swiss mice exposed to 9270-MHz (PW) radiation (duty cycle = 0.001) at
an average power density of 100 mW/cm for 4.5 min daily, 5 days/week for
59 weeks. The daily exposure produced an average body temperature rise of 3.3 °C.
This daily dose is stated to be half the LD^q for the mice. Originally there
were 200 irradiated mice and 100 control mice. Five percent of each group
was killed for pathological and hematological examination 7 months after the
beginning of irradiation, and an additional 10 percent of each group was
killed at 16 months. The latter series was done within a month of the final
irradiation. Because of a partial contradiction in pathological findings in
the animals killed at 7 and 16 months, all surviving mice (19 controls and 67
irradiates) were killed 19 months after the beginning of irradiation (4 months
after final irradiation). The data from this experiment are summarized in
Table 5-23. The authors state that the longevity of the mice did not appear to
5-270

-------
TABLE 5-23. SUMMARY OF PRAUSNITZ AND SUSSKIND DATA (1972)
Controls (100 total)	Irradiates (200 total)
When		
Sacrificed	No X of No with % This Series No	% of No with % This Series
Series	(months)	Animals Total "Leucosis" with "Leucosis" Animals Total "Leucosis" with "Leucosis" Comment
tn
i
(V)
--J
Sacrifice Series 01 7
Sacrifice Series #2 16	10
(1 month post-
irradiation)
Sacrifice Series #3
(all animals
surviving at 19
months)
Longevity Series
19	19
(4 months post-
lrradiation)
As	40
occurred
Spontaneous Deaths No data
given
26
10
19
40
26
None
apparently
given
10
21
10
10
20	10
None apparently given
30
67	33 5 12
60	30	21
43	22
18
35
Reported as
negative for
"leucosis"
Reported as
positive for
"leucosis"
Not statistically
significant
X2 (1 df) = 1 49
Reported as
negative for
"leucosis"
See text
Autolysis too
advanced for diag-
nosis of cause
of death

-------
be affected under the prevailing conditions. The data appear to support a
stronger statement, i.e., significantly more irradiated mice than controls
survived until the termination of the experiment 19 months after the beginning
of irradiation.
The difference in survival is not statistically significant at 14 months
2
into the experiment (x ~ 2.2, using data from Fig. 3, Prausmtz and Susskind
1962) but is significant at 19 months (x^ =4.9, 0.05 > P > 0.025). The data pro-
vided are not sufficiently detailed to apply more refined methods of analyzing
survival (Peto et al_. 1976, 1977). The authors suggested that, in view of problems
with pneumonia experienced in both control and irradiated mice in the last few
months of irradiation, "conceivably microwave irradiation in this modality,
with periodically induced slight artifical fevers, is of some benefit to the animal
in combatting disease."
There have been some reports of altered and possibly enhanced immunological
competence after microwave exposure. These are discussed in detail in § 5.2,
Hematologic and Immunologic Effects. In a relevant study, Preskorn et aK
(1978) compared tumor development and longevity in female mice prenatally
exposed to 2450-MHz microwaves with sham-irradiated controls. The irradiated
mice were offspring of dams exposed for 20 min/day on days 11, 12, 13, and 14
of gestation in a multimodal cavity (SAR = 35 W/kg). Both groups were injected
with sarcoma cells at age 16 days. Both the mean and maximum survival time of
"irradiated mice with tumors" exceeded those of "non-irradiated mice with
tumors" (P < 0.05). The mean survival time of "irradiates without tumors" was
also greater than "non-irradiated mice without tumors." When 50 percent of
5-272

-------
the controls had died, 67 percent of the irradiates were alive. The difference
was not significant (0.05 < P > 0.1), however, and maximum life span did not
di ffer.
Rotkovska and Vacek (1972, 1977) reported increased survival and increased
LD^o_3q in m1ce for sublethal (600 to 750 R) acute exposures to 200 kV x-rays
if the x-ray exposures were preceded (1, 3, 14 days) or followed (30 min) by a
2
5-min exposure to 2450-MHz microwaves at 100 mW/cm . In their initial experi-
ment, the 30-day survival after a 600-R x-ray exposure was increased from
14 percent (controls) to 53, 88, or 90 percent when such exposure followed the
microwave irradiation by 1, 3, or 14 days, respectively. In their second
experiment, the LD^q_^q was increased by 100 R with all microwave-irradiated
groups having significantly higher survival at 30 days than their control
counterparts. These data illustrate the plethora of potentially confounding
variables that may have to be considered in determining RF-radiation exposure
effects.
5.9.2 Carcinogenesis
Carcinogenesis is the process of inducing cancer or malignant neoplasia.
Neoplasia is uncontrolled growth or cell division in a tissue and is considered
malignant if cells from the affected tissue (tumor) enter the blood and travel
to other parts of the body to colonize and form new tumors (metastases). A
tumor that does not metastasize is considered benign, although it may well
grow to sufficient size to be 1ife-threatening. Consideration of the various
theories of carcinogenesis is beyond the scope of this discussion, except to
5-273

-------
note that most authorities believe that the initiating event on the cellular
level involves physical or chemical alterations of a cell such that its descendent
cells are abnormal and may override the body's cellular proliferation control
processes. The role that the immune system plays in primary carcinogenesis has
not been well documented and is somewhat controversial. However, it is thought
that neoplasia may arise through an epigenetic mechanism in which immunosuppression
allows the development of tumors initiated by a genetic event.
The specter of potential carcinogenicity has been periodically raised in
connection with RF-radiation exposure since 1953, when J. R. McLaughlin, a
medical consultant to the Hughes Aircraft Corporation, submitted a report to
the military, listing leukemia as one of the possible effects of radar exposure
(McLaughlin 1953). The relevant literature, which is sparse, has been re-
viewed by Baranski and Czerski (1976), Justesen et cH. (1978), and Dwyer and
Leeper (1978) with little supportable evidence that RF exposures are likely to
be carcinogenic. However, the subject remains controversial for many reasons,
including
(1)	Widespread fear of cancer in large segments of the population
(2)	The proven carcinogenicity of ionizing radiation
(3)	The failure of the public and the media to distinguish between
ionizing and nonionizing (i.e., RF) radiation
(4)	The existence of several anecdotal reports of association of
cancer with RF-radiation exposure in humans
(5)	The publication of an early animal study (Susskind 1962,
Prausnitz and Susskind 1962) reported by the authors to show
leukemogenesis in mice chronically irradiated with microwaves
(6)	The unsupported reports by Zaret (1976) of an increased incidence
of cancer in North Karelia followed by the misinterpretation of
5-274

-------
the Zaret article by Dwyer and Leeper (1978), coupled with their
editorial transformation of the North Karelia Project from a
study of cardiovascular disease to a study of cancer
(7)	The recent reports by Szmigielski et a^. (1980) that exposure
to 2450-MHz microwaves accelerated the development in mice of
spontaneous breast cancer and 3,4-benzopyrene-induced skin
cancer
(8)	The postulation that, since heat is a known mutagen and there
is a strong correlation between mutagenesis and carcinogenesis,
under certain conditions of relatively high-intensity exposure,
RF radiation that causes significant tissue heating could theoreti-
cally induce cancer
(9)	The lack of well designed human or animal studies that have sufficient
exposure data and are statistically adequate to provide credible
data that would support a reliable conclusion.
Reviewed in depth herein are four published reports that have received
extensive publicity, including (1) the Prausnitz and Siisskind (1962) report of
leukemogenesis in mice chronically irradiated with microwaves, (2) the "North
Karelia Connection" suggested by Zaret (1976), (3) the irradiation of U.S.
Government employees stationed at the U.S. Embassy in Moscow (Lilienfeld et
aK 1978), and (4) the health status of U.S. Navy personnel exposed to radar
during the Korean War (Robinette et aJL 1980). Brodeur (1977) gives anecdotal
accounts of increased incidence of cancer in several groups of defense con-
tract personnel working in RF-radiation research and development, but no
reliable reports on these incidents have appeared in the scientific or medical
literature. Since the cutoff date for general literature review, two letters
to the editor have appeared in the New England Journal of Medicine suggesting
an association of polycythemia vera with occupational microwave exposure
(Friedman 1981) and leukemia with occupational exposure to a variety of elec-
tric and magnetic fields (Milham 1982). Neither of these letters would meet
5-275

-------
the criteria established for this review but should lead to more elaborate
studies to resolve the questions raised.
5.9.3 Prausnitz and Susskind Study
In 1962, Prausnitz and Susskind published reports on an experiment con-
ducted to determine the pathological and life-shortening effects of chronic
microwave exposure of mice. In 200 male Swiss mice irradiated with
2
9270 MHz (PW) at an average power density of 100 mW/cm for 4.5 min daily, an
average body temperature increase of 3.3 °C was produced. Exposures were
conducted 5 days/week for 59 weeks, and survival, body mass, blood parameters,
and certain postmortem pathological findings were compared with those of a con-
currently sham-irradiated group of 100 animals. Animals were sacrificed at
7, 17, and 19 months after the beginning of irradiation. In addition,
23 percent of the animals were lost to the study by spontaneous death and
autolysis prior to necropsy. The remaining 100 animals (40 controls,
60 irradiates) formed a "longevity" study group. One of the findings reported
in this group was "cancer of the white cells," defined as monocytic or
lymphocytic leucosis, or lymphatic or myeloid leukemia. Leucosis was defined
as a noncirculating neoplasm of the white cells, whereas leukemia was defined
as a circulating leucosis. Data were grouped and reported as "leucosis."
Although the liver and spleen were removed for examination at necropsy, they
were not considered in determining lymphoid infiltration, even though these
organs are usually involved in this kind of reaction. Data describing the
differential composition of the blood cell types are not given. A summary of
the fate of the 200 irradiates and 100 controls is shown in Table 5-23.
5-276

-------
The results as reported are confusing. The authors performed a third
sacrifice series at 19 months to resolve a perceived contradiction in the
"leucosis" findings from the first and second sacrifice series. Yet, had they
used even a crude 2x2 contingency table analysis to test the prevalence of
leucosis in the control and irradiated animals sacrificed in the second series,
2
they would have arrived at a x of only 1.49 (uncorrected for continuity) with
one degree of freedom (P > 0.20). Thus, the prevalence of leucosis in irradiated
animals was not significantly different from that in the control animals for
any of the "sacrifice" groups.
While leucosis in the "longevity" series is highly significant (P <
2
0.005, x >8.0), these data are severely compromised by (1) the marked loss
of animals that would have fallen in this group due to autolysis, (2) the
differential loss of control animals due to infection, and (3) the premature
sacrifice of all remaining animals to resolve a nonexistent paradox. Com-
paring data from two groups of animals that selected themselves (by dying
prior to 19 months and being found in time for necropsy) is not an acceptable
statistical analysis of experimental data. An acceptable method would be to
compare the prevalence rates of leucosis in the controls and irradiates in
combined groups that include all the animals at risk in the original longevity
group (longevity and third sacrifice series). If this is done, the following
results are obtained:
Group	With Leucosis Without Leucosis Total
Control	8 (a)	51 (b)	59 (a+b)
Irradiated 33 (c)	94 (d)	127 (c+d)
Total	41 (a+c)	145 (b+d)	186 (n)
5-277

-------
To test whether the leucosis rates are different in controls and irradiates,
the following statistic may be used:
S = n(ad - bc)^/[(a+b) (c+d) (a+c) (b+d)]
If the animals comprising the table represent a random sample from a larger
2
population, S approximately follows a x distribution with one degree of
2
freedom. Upon substitution, the value of 3.72 is obtained. Since x >3.62
(P - 0.06), the rates would not be declared different at the 5 percent significance
level. Furthermore, if a continuity correction were used in computating this
2
statistic, the resultant x = 2.93, which is significant only at the 9 percent
level.
There are now better statistical methods available to analyze survival
data (Peto et al^. 1976, 1977), but they require detailed documentation of the
fate of each animal during the course of the experiment. Given (a) the loss of
histopathological data on animals because of autolysis and infection, (b) the
further complication that different pathologists were used for different
sacrifice series, and (c) the absence of historical data on the incidence of
leucosis in the mouse strain used, any attempt to apply these methods retro-
actively would not be beneficial.
In summary, because of the previously described problems with the bio-
logical protocol, the lack of sound statistical methodology in experimental
design and data analyses, and the questionable significance of what was reported,
5-278

-------
the study by Prausnitz and Susskind is of limited value in defining the
pathological effects of chronic whole-body microwave radiation. As noted
previously, the data on survival times are of considerable interest due to
similar findings by Spalding et a^. (1971) and Preskorn et al^. (1978).
5.9.4 North Karelia Connection
In 1971, the Finnish government, in conjunction with the World Health
Organization, began a program known as the "North Karelia Project" (Puska et
aK 1978). Finland has one of the highest rates of mortality from cardio-
vascular disease (CVD) in the world. Within the country, the eastern provinces
had higher rates of CVD than those in the west. North Karelia, an eastern
province that borders the Soviet Union, was identified in the 1950's as
having a particularly high incidence of CVD. The objectives of the project
were to decrease CVD morbidity and mortality by identifying the causative
factors, devising means for primary prevention, and strengthening treatment
and secondary prevention. Cigarette smoking, elevated blood pressure, and
serum cholesterol levels were postulated as three major risk factors. The
proximity of North Karelia to the Soviet border prompted an American physician,
Dr. Milton Zaret, to speculate that RF radiation from Soviet communications or
radar might be contributing to the incidence of CVD. In a presentation
at the symposium Biologic Effects and Health Hazards of Microwave Radiation
(Warsaw, October 15 to 18, 1973), he suggested that microwave radiation from
the Soviet Union might be a factor. His remarks were later withdrawn and do
not appear in the record of the Symposium. However, he restated this hypo-
thesis in a published article (Zaret 1976), in which he additionally stated,
5-279

-------
without data or reference, that "a new finding, an increased incidence of
cancer, also appears to be emerging in North Karelia." The major thrust of
the article was directed to the possibility of exploiting the North Karelia
Project to investigate the role of RF radiation in the evolution of CVD. He
also noted that the potential role of RF radiation in carcinogenesis should be
studied.
In their literature review on carcinogenic properties of microwave and RF
radiation, Dwyer and Leeper (1978) misinterpreted the 1976 Zaret article. They
editorially transformed the North Karelia Project into an epidemiological
project designed to test the hypothesis that RF radiation caused or contributed
to heart attacks or cancer.
The Finnish Project (1972 to 1977) caused significant changes in diet and
lifestyle, which correlated with a significant reduction in CVD mortality.
Further, information provided by the Finnish Institute of Radiation Protection
(K. Jokela, personal communication to M. Hattunen, Scientific Counselor of
Embassy of Finland, 2133 Wisconsin Avenue, N.W., Washington, D.C. 20007,
November 22, 1978) specifically denies the existence of an abnormal cancer
incidence in eastern Finland and knowledge of any possible linkage to microwave
fields.
5.9.5 Moscow Embassy Study
The long-term microwave irradiation of the American Embassy buildings in
Moscow was highly publicized in 1976. This led to a comprehensive study of
5-280

-------
the morbidity and mortality of the Moscow Embassy employees and dependents as
compared with employees and dependents of U.S embassies in other Eastern
European locations (Lilienfeld et a^. 1978).
The Moscow group experienced less overall mortality for all causes of
death than did the comparison group. However, the death rate in females from
malignant neoplasms was slightly, but not significantly, higher than expected
in the Moscow group; and the incidence of malignant neoplasms, other than of the
skin, was significantly higher in the Moscow female group. In both cases, the
numbers were very small. In the former case, there were 7 different cancer
sites involved in 8 cases, which, according to the authors, virtually eliminates
a single causal factor; in the latter case, there were 10 cases involving 7
different sites, and, when exposure to microwaves was considered, it was found
that the death rate was highest for those who had the least exposure to micro-
waves.
5.9.6 U.S. Navy Study
A study of the effects of occupational exposure to radar during the
Korean War on U.S. Navy enlisted personnel was recently published (Robinette
et al. 1980). Approximately 20,000 men whose occupations possibly involved
relatively more frequent exposures to higher intensity fields were compared
with 20,000 men in other occupational classifications. One of the end points
evaluated was the effect of exposure from 1950 to 1954 on the relative inci-
dence of cancer during 1950 to 1976. No statistically significant differences
5-281

-------
between the two groups were evident for malignant neoplasms as a cause of
death from 1950 to 1974, or as a cause of hospitalization in Navy or VA hospitals.
Within the high exposure group, subgroups were created, depending on Hazard
Number. The Hazard Numbers were measures of potential individual exposures,
dependent upon the type and power rating of the radar and upon the ship and the
time spent in service on the ship. Hazard Number categories were 0, 1 to
5000, and > 5001. In considering the causes of death within the highest
exposure group, there are significantly more malignant neoplasms of the
respiratory tract in the subgroup with Hazard Number > 5001 than in the
subgroup with the Hazard Number < 5000. A significant trend for increased
death from all diseases in the > 5001 subgroup was also noted. However, informa-
tion relevant to respiratory neoplasms such as smoking histories was not available
because the study was an analysis of cause of death from death certificate data.
Also, in a study where so many statistical comparisons are done using a P = 0.05
criterion, one or two positive findings would be expected by chance.
5.9.7 Unresolved Questions
Because few RF-radiation studies in man or animals have employed life span
or cancer as end points, and none has had sufficient statistical power and
adequate quality control to place an upper limit of risk at less than two times
control incidences, the questions of microwave carcinogenesis or life shortening
are still open. Conversely, none of the complete reports in the literature
presents a convincing case for the existence of a significantly increased risk
of cancer or life shortening in exposed populations. A critical issue is the
difficulty of developing exposure data or information in human population studies.
5-282

-------
The interaction of RF radiation with other moieties (such as ionizing
radiation) in the induction, prevention, or treatment of cancer has not been
fully characterized. Microwave-induced hyperthermia has been demonstrated to
increase the effectiveness of x-ray treatment of various cancers. In nonthera-
peutic situations, however, both inhibition and enhancement of tumor development
in microwave-irradiated mice have been reported.
In their paper, discussed in the section on longevity (§ 5.9.1, Life Span),
Preskorn et a_h (1978) found that the development of tumors following injection
of sarcoma cells into 16-day-old CFW mice was significantly delayed if the mice
had been exposed to 2450-MHz microwaves jn utero on day 11, 12, 13, and 14 of
gestation (20 min/day, SAR = 35 W/kg). There was no difference, however, in the
ultimate number of tumors.
Conversely, Szmigielski et al^. (1980) found that repeated exposure to
2450-MHz microwaves 5 or 15 mW/cm^ (SAR 2 to 3 or 6 to 8 W/kg), 2 h/day,
6 days/week for up to 10 months significantly accelerated the appearance of
spontaneous breast tumors in C^H/HeA mice, and the appearance of skin cancer
in BALB/c mice treated with 3,4-benzopyrene during or after the microwave
exposure. It is noted that chronic stress due to deliberate overcrowding
2
produced essentially the same effect as the 5 mW/cm exposure. In both
?
experiments, 15 mW/cm produced greater acceleration in tumor production than
2
did 5 mW/cm or overcrowding. Workers in this field have only begun to discover
the effects of RF-radiation on animals under different immunologic and
hormonal equilibrium levels, e.g. animals subjected to prolonged stress.
5-283

-------
5.10 HUMAN STUDIES
Doreen Hill
The general approach used in this document for the evaluation of the
health effects literature is stated in the Introduction. Rigorous criteria
were established and applied to the reports of experimental results in order
to establish what is believed to be a credible data base. It is, however,
difficult to apply these selection and review criteria uncompromisingly to the
human studies or epidemiological literature, largely because the research
method is population-based and observational rather than experimental. The
papers included here were selected because they were judged to present rela-
tively more information on exposure parameters and/or more rigorous or ana-
lytical study designs. The amount of detail in reporting and the degree of
specification of study methods and procedures (use of controls, statistics,
control of confounding variables, and so forth) were considered important.
Case reports are not included.
5.10.1 Occupational Surveys/Clinical Studies
The majority of reports in the literature concern people occupationally
exposed in military or industrial settings. A wide variety of conditions,
symptoms, and clinical measurements are usually evaluated. The health condit-
ions investigated usually are pre- or subclinical instead of overt or diag-
f
nosed disease. Studies of this type that focused on single rather than multi-
ple end points are described later.
5-285

-------
It is in occupational surveys and industrial health surveillance programs
that the potential for neurological and behavioral changes has been investi-
gated. The Soviet and Eastern European literature frequently describes a
collection of symptoms reported to occur in personnel industrially exposed to
microwaves. These collective symptoms, which have been variously called the
"neurasthenic syndrome," the "chronic overexposure syndrome," or "microwave
sickness," are based on subjective complaints that include headaches, sleep
disturbances, weakness, decrease of sexual activity (lessened libido),
impotence, pains in the chest, and general poorly defined feelings of non-
well-being (Baranski and Czerski 1976, p. 157). Also described are a set of
labile functional cardiovascular changes including bradycardia (or occasional
tachycardia), arterial hypertension (or hypotension), and changes in cardiac
conduction, and this form of neurocirculatory asthenia is also attributed to
nervous system influence (Silverman 1980).
Barron et a^. (1955) presented results of a study conducted to evaluate
changes in various physical and functional characteristics of radar personnel
employed by an airframe manufacturer. A total of 226 exposed workers were
initially included in the medical surveillance program. The radar workers
were characterized by their duration of exposure as seen in Table 5-24.
Controls totaling 88 subjects, stated to have had no industrial radar
exposure, were also examined. Methods of selection of cases or controls were
not specified. The age distribution of all subjects ranged from 20 to more
than 50 years, with the majority under 40 years of age. The controls were,
however, substantially older, as can be seen in Table 5-25.
5-286

-------
TABLE 5-24. DISTRIBUTION OF YEARS OF EXPOSURE
FOR 226 RADAR WORKERS*
Years of Exposure	No. of Workers	Percent of 226
0 to 2	106	47
2 to 5	83	37
5 to 13	37	16
*Data from Barron et al.	1955.
TABLE 5-25. AGE DISTRIBUTION OF 226 MICROWAVE WORKERS
AND 88 CONTROLS*
Age


Radar

Controls
% of 226
Cumulative %
% of
88 Cumulative %
20 to
29
34
34
14
14
30 to
39
49
83
40
54
40 to
49
13
96
27
81
50+

4
100
19
100
*Data from Barron et al. 1955.
A decrease in polymorphonuclear cells was reported in 25 percent of the
radar workers vs. 12 percent of the controls. An increase in monocytes and
eosinophils was also observed for the exposed group, but in a later report
(Barron and Baraff 1958) these effects were attributed to a technical error.
Platelet counts and urinalyses were similar in the two groups. Ophthalmo-
logical examinations revealed ocular anomalies of several diverse types among
5-287

-------
12 exposed persons vs. 1 case in the control group. The medical surveillance
program was extended to permit periodic reexaminations (Barron and Baraff
1958). No significant changes in health status were noted.
The radar bands of exposure include S-band (2880 MHz) and X-band
(9375 MHz). Exposure times and power densities for individuals could not be
developed, but zones at various distances from the antenna were specified and
used to estimate three ranges of power densities. The minimal average power
2	2
density in Zone A was 13.1 mW/cm . Zone B ranged from 3.9 to 13.1 mW/cm .
2
Zone C was <3.9 mW/cm . The authors stated that because of the relatively
low power densities, personnel working in Zone C were eliminated from the
study. The majority of personnel worked with or around APS-45 and AN/APS-20-B
and -E radars. By establishing these zones of potential exposure, steps were
taken to limit entry of personnel to Zone A. Zone B was judged to be safe for
occasional but not constant exposure. As a result, most subjects continuing
in or added to the examination program are believed to have had only incidental
exposures to power densities less than 13.1 mW/cm .
The strengths of this study are the attempts to estimate potential
exposures and to periodically reexamine the workers. But, there are also some
major problems with study methods and analyses. For example, there are dis-
proportionate numbers of exposed vs. control subjects as well as questionable
comparability because of age differences (Table 5-25). None of the observed
results were tested statistically.
5-288

-------
Czerski, Siekierzynski, and co-workers in Poland evaluated 841 men
occupationally exposed to pulse-modulated microwave radiation (Czerski et al_.
1974; Siekierzynski et a^. 1974a, 1974b). The radiation frequencies were not
explicitly stated, but one can infer from the references to pulse-modulation
and radio-location that the working environment dealt with radar frequencies.
The age distribution of the men ranged from 20 to 45 years. The men had
worked varied periods of time, with some employed over 10 years. An unexposed
control population comparable in general working conditions and socio-economic
status could not be established; therefore, the study group was subdivided
into two groups on the basis of level of microwave exposure. One group con-
2
sisted of 507 men exposed to mean power densities greater than 0.2 mW/cm ,
2
with short-term exposures estimated to reach 6 mW/cm . The other group was
2
334 men exposed to mean power density levels less than 0.2 mW/cm .
The health end points evaluated covered three major categories, e.g.,
neurotic syndrome, digestive tract functional disturbances, and cardio-
circulatory disturbances with abnormal electrocardiogram (ECG) findings.
According to Polish occupational exposure criteria, these conditions are
considered contraindications for work in a microwave environment. The
neurotic syndrome was defined by a variety of symptoms such as fatigue,
headaches, sleep disturbances, and difficulties in memorizing and con-
centrating. Psychologic examinations were given. Comparisons were made
between and within exposure groups according to age and duration of
occupational exposure. The two groups were found to be similar with respect
to the distribution of these symptoms and conditions. There was no dependence
on the duration of exposure.
5-289

-------
In 1979, Djordjevic et a^. reported on medical evaluations of radar
workers, aged 25 to 40 years, with a work history of 5 to 10 years. Specific
frequencies of exposure were not cited but were stated to be within the whole
range used in radar operations. Evaluation of the working environment was
undertaken, including power density measurements. While the environmental
analyses are not given in the paper, it was concluded that the workers were
exposed to pulsed microwaves within a wide range of intensities but generally
?
at levels less than 5 mW/cm . The lower limit of this exposure may have been
2
1 mW/cm , but it is not clear from the discussion whether this estimate refers
to the workers included in this study or to radar station personnel in
general. The control group consisted of 220 persons reported to be similar in
age, working environment, and socioeconomic status. The controls did not have
work experience with microwave sources. Selection criteria or further
descriptive information was not given for either the cases or controls.
Ten major end points or diagnoses were covered in the clinical eval
uation, including ophthalmologic examinations. These were conducted by
clinical specialists following the same scheme for classification of
abnormalities. The two groups did not differ with respect to the 10 factors.
The type of procedure was not reported. The groups also did not differ
statistically in terms of electrocardiogram results or on multiple biochemical
and hematologic indicators. Radar workers did demonstrate more subjective
complaints, including headache, fatigue, irritability, sleep disturbance,
inhibition of sexual activity, and impairment of memory. Based on their
survey of working conditions, the authors attribute the latter result to
specific problems such as lighting or poor ventilation in the environment
5-290

-------
of the radar workers. If true, it appears that either the subject and control
groups were not as comparable with respect to working environment as
originally stated, or some other factor(s) is (are) operating to produce more
subjective complaints. One possibility is a selective or reporting bias on
the part of the radar workers, e.g., enhanced awareness of the possible
"microwave sickness1' syndrome. Some other environmental factor may have been
present that may not be measurable in the type of environmental survey
conducted or detectable through these clinical evaluations but microwave
exposure cannot be clearly excluded.
5.10.2 Mortality Studies
Lilienfeld and assodiates (1978) completed a broad survey of the
mortality and morbidity experience of Foreign Service employees and their
dependents to assess the potential health consequences of microwave
irradiation of the American Embassy in Moscow. Foreign Service employees and
those from other agencies who had served in the Moscow Embassy during 1953 to
1976 were compared with employees at eight other embassies or consulates in
Eastern Europe over the same time period.
The microwave irradiation of the Moscow Embassy was first detected in
1953 and subsequently varied in intensity, direction, and frequency over time.
The frequencies ranged from 0.6 to 9.5 GHz (United States Senate, Committee on
Commerce, Science, and Transportation 1979; Pollock 1979). The measured
average power densities over time are given in Table 5-26.
5-291

-------
TABLE 5-26. MICROWAVE EXPOSURE LEVELS AT THE U.S. EMBASSY IN MOSCOW
Time Period
Exposed
Area of Chancery
Power Density and
Exposure Duration
1953 to May 1975
West Facade
2
Max of 5 pW/cm

9 h/day
June 1975 to Feb. 1976
South and East Facade
18 pW/cm2


18 h/day
Since Feb. 7, 1976
South and East Facade
2
Fractions of a pW/cm

18 h/day
*Data from Lilienfeld et al. 1978.
Extensive efforts were launched to identify and trace the populations.
Information on illnesses, conditions, or symptoms were sought from two major
sources: (1) employment medical records, which were fairly extensive, given
examination requirements for foreign duty and (2) a self-administered health
history questionnaire. Questionnaire responses were validated for a
stratified sample by review of hospital, physician, and clinic records. Death
certificates were also sought, although other sources also were used to
ascertain mortality status.
Standardized mortality ratios for various subgroups were developed for
each cause of death, were standardized for age and calendar period, and were
specific for sex. Similar procedures were applied to develop summary indices
of morbidity.
5-292

-------
A total of 4388 employees and 8283 dependents was studied. More than
1800 with 3000 dependents were employed at the Moscow Embassy and 2500 with
more than 5000 dependents worked at the comparison posts. Ninety-five percent
of the employees were traced. Receipt of completed questionnaires was less
successful, with an overall response rate of 52 percent for State Department
personnel.
Hundreds of comparisons between the Moscow and control posts were made
from information in the medical records. Various health problems were
generally similar with two exceptions. Moscow employees had a threefold
greater risk of acquiring protozoal infections than comparison-post employees.
In general, both sexes in the Moscow group had somewhat higher frequencies of
most of the common kinds of health conditions reported. The authors stated,
"However, these most common conditions represented a very heterogeneous
collection and it is difficult to conclude that they could have been related
to exposure to microwave radiation since no consistent pattern of increased
frequency in the exposed group could be found."
Some excesses were reported by Moscow employees in the health history
questionnaire. Both sexes reported more eye problems due to correctable
refractive errors. More psoriasis was reported by men and anemia by women.
The Moscow employees, especially males, reported more symptoms such as
irritability, depression, difficulties in concentration, and loss of memory.
It is possible, however, that a bias due to awareness of potential adverse
effects is operating, since the strongest differences were present in the
subgroup with the least exposure.
5-293

-------
The observed mortality was less in both male and female employees than
expected, based on U.S. mortality rates; the male employees had more favorable
experience than female employees. In both sexes, cancer was the predominant
cause of death. The Moscow and comparison groups did not differ appreciably
in overall and specific mortality. The authors noted, however, that the
population was relatively young, and it may be too early to detect long-term
mortality effects.
The authors concluded that no convincing evidence was discovered to
implicate microwaves in the development of adverse health effects at the time
of the analysis. But they also carefully discussed the limitations inherent
in the study. These included uncertainties associated with the reconstruction
of the employee populations and dependents, difficulties of obtaining death
certificates, the low percentage of responses for the questionnaire, and the
statistical power of the study. The limitation most critical for consid-
eration in a document such as this relates to ascertainment of exposure.
Problems relative to individual mobility within the embassy and variation of
field intensities within the building are present in this study as in any
other. No records were available on where employees lived or worked, so one
had to rely on questionnaire responses to estimate an individual's potential
for exposure. The highest exposure levels (18 pW/cm ) were recorded for only
6 months in 1975-76, thus the group exposed to the most intense fields had the
shortest cumulative time of exposure and of observation in the study.
Robinette and Silverman (1977) and Robinette et al^- (1980) examined
mortality and morbidity among U.S. naval personnel occupationally exposed to
5-294

-------
radar. Records of service technical schools were used to select subjects for
the study. Exposure categorizations were made on the basis of occupational
specialty. The exposure group (probably highly exposed) consisted of
technicians involved in repair and maintenance of radar equipment. The
controls (probably minimally exposed) were involved in the operation of radar
or radio equipment. It was estimated from shipboard monitoring that radiomen
and radar operators (in the low-exposure group) were generally exposed at less
2
than 1 mW/cm , while gunfire control and electronics technicians (in the high
exposure group) were exposed to higher levels during their duties. Over
40,000 veterans were included in the study, with about equal numbers in these
two major exposure classifications. In conjunction with naval personnel, an
effort was also undertaken to develop an index of potential exposure, termed
Hazard Number, for a limited portion of the population. This number was based
on the duty months multiplied by the sum of the power ratings (equipment
output power) of gunfire-control radars (ship) or search radars (aircraft)
where technicians were assigned.
Medical information was obtained through Navy and Veterans Administration
records. Records were searched for information on four major end points:
(1) mortality; (2) morbidity via in-service hospitalizations; (3) morbidity
via VA hospitalizations; and (4) disability compensation.
Mortality was ascertained through the VA beneficiary system. Strokes,
cancers of the digestive tract and respiratory system, and leukemias were
elevated for the high exposure group, but none of the increases was statis-
tically significant. The authors note that the differences in mortality from
5-295

-------
malignant neoplasms of the lymphatic and hematopoietic system are not
statistically significant. As seen in Table 5-27, comparisons were also made
within the high exposure group across Hazard Number categories. In this case,
only two comparisons were statistically significant: (1) the difference in
TABLE 5-27. NUMBER OF DEATHS FROM DISEASE AND MORTALITY RATIOS3 BY HAZARD NUMBER:
U.S. ENLISTED NAVAL PERSONNEL EXPOSED TO MICROWAVE RADIATION
DURING THE KOREAN WAR PERIOD
Number of Deaths
International		
Classification
^	High Exposure, by Hazard No.
Diseases Low		
Cause of Death	(8th Rev.) Exposure Total 0	1-5000 5000+
All diseases	000-796 325	309	63 160	86
(1.04)	(0.96)	(0.82) (0.91)	(1.23)
Malignant neoplasms	140-209 87	96	22 45	29
(0.96)	(1.04)	(0.99) (0.90)	(1.44)
Digestive organs	150-159	14	20	6	11	3
(0.85)	(1.14)	(1.49) (1.14)	(0.78)
Respiratory tract	160-163	16	24	4	10	10
(0.85)	(1.14)	(0.82) (0.86)	(2.20)
Lymphatic and	200-209	20	26	6	12	8
hematopoietic system	(0.83)	(1.18)	(1-09) (1.04)	(1.64)
Other malignant	Residue	37	26	6	12	8
neoplasms	(1.19)	(0.82)	(0.78) (0.70)	(1.17)
Diseases of circulatory 390-458	167	150	36 73	41
system	(1.07)	(0.93)	(0.94) (0.83)	(1.17)
Other diseases	Residue	71	63	5	42	16
(1.08)	(0.92)	(0.30) (1.13)	(1.08)
aMortality ratio (in parentheses) standardized for year of birth; the combined
experience of the low and high exposure groups is taken as the standard.
^Data from Robinette et al. 1980.
5-296

-------
respiratory tract cancer between those with a Hazard Number less than 5000 vs.
more than 5000 and (2) the test for trend for all diseases combined. However,
information relevant to respiratory neoplasms such as smoking histories was
not available because the study was an analysis of cause of death from death
certificate data. Differential health risks with respect to hospitalized
illness around the period of exposure were not apparent. Subsequent VA
hospitalizations and disability awards provided incomplete information.
Since the study focused largely on the use of automated VA record
systems, it was not possible to determine non-Navy or non-VA hospitalizations,
nonhospitalized conditions, reproductive histories, or subsequent employment
histories. Actual individual exposure could not be reconstructed retro-
spectively, thus the information represents the potential exposure of the men.
5.10.3 Ocular Effects
Many surveys on ocular effects have been conducted. The potential of RF
radiation to induce cataracts and the production of lens defects and opacities
have been cited in both American and foreign literatures. Ocular effects
have, in fact, been the major end point of study in U.S. research, primarily
in military populations. For this health end point more than others, much
attention has been devoted to careful, detailed clinical examinations and to
use of standard procedures or protocols. Population selection criteria,
however, is not well elaborated; therefore, the possibility of selection bias
cannot be excluded for most of these surveys. Ocular studies have largely
taken the form of cross-sectional clinical surveys in actively working popula-
5-297

-------
tions and, as such, have focused on minor lens changes. Longitudinal studies
of populations with known exposure, with followup of either a prospective or
retrospective nature, are nonexistent despite the identification of potential
cohorts from the cross-sectional surveys. The research to date has not yet
fully examined the question of whether lens changes, if observed, lead to
overt clinical eye disease or conditions, e.g., the functional and clinical
significance of minor lens changes in general and over time. Thus, while it
appears reasonable that RF radiation at high levels could influence cataract-
ogenesis, it cannot be confirmed without more rigorous followup studies in
already identified exposed populations. This would be a reasonable line of
future research if such epidemiological studies are feasible.
Utilizing Veterans Administration hospital records and military personnel
records, Cleary et aj. (1965) conducted a case-control study to examine
cataract formation among Army and Air Force veterans. This is the only case-
control study focused on clinically diagnosed cataracts. Cases were defined
as white males born after 1910 who were treated for cataracts between 1950 and
1962, based on diagnoses given in VA hospital records. Controls were drawn
from the same sources by selecting men with adjacent hospital register
numbers; therefore, the control group had random diagnoses made in the same
hospitals at the same time the cataract cases were diagnosed, and the controls
were also born after 1910. Military occupational specialties, as abstracted
from service records, were used to denote radar or nonradar workers. Job
classification was then used as an indicator of potential exposure. The
distribution of cases and controls according to radar exposure is seen in
Table 5-28. The frequency of microwave exposure as denoted by radar-work
5-298

-------
history was similar between cases and controls; the odds ratio was less than
one (0.67) and was not statistically significant. The distribution over all
age groups is seen in Table 5-29. Excess risk was noted for Air Force
veterans, but the numbers were small. Only 40 radar workers were found in
over 5000 cases and controls.
TABLE 5-28. CLASSIFICATION BY MILITARY OCCUPATION OF WORLD WAR II
AND KOREAN WAR VETERANS WITH AND WITHOUT CATARACTS,
BASED ON DISCHARGES FROM VA HOSPITALS*
Veterans' Cataract Status
Occupation	Yes	No	Total
Radar Workers	19	21	40
Nonradar Workers	2625	1935	4560
TOTAL	2644	1956	4600
Odds Ratio = OR = [21] (2625) = °'67
*Data are from Cleary et aK 1965.
Cleary and Pasternack (1966) examined and scored subclinical or minor
lens changes in 736 microwave workers and 559 controls. A relative exposure
index or score was developed by weighing various parameters such as power
output and distance from the microwave generating source. The work environ-
ment in terms of microwave parameters was not specified. Utilizing regressi
techniques, radar workers were found to have more lens changes than controls
and the number of defects was correlated with duration of exposure and expo-
sure score among the workers. The authors suggested that the subclinical
changes may indicate accelerated aging of lens tissues.
5-299

-------
TABLE 5-29. ESTIMATED RELATIVE RISK OF CATARACTS
AMONG ARMY AND AIR FORCE VETERANS*
Cataracts

Radar
Workers
Nonradar
Workers
Odds
Ratio
2
X

Total
Yes
No
19
21
2625
1935
0.67
1.26
NS
20-39
Yes
No
3
4
418
522
0.94
0.08
NS
40-49
Yes
No
7
13
1517
1101
0.39
3.39
NS
50-59
Yes
No
9
4
699
316
1.02
0.09
NS
*Data are from Cleary et aK 1965.
Majewska (1968) studied the eyes of 200 Polish workers employed from 6
months to 12 years at installations with microwave-generating equipment that
operated from 600 MHz to 10.7 GHz (2.8 to 50 cm). Although cited as "high
intensity" by the author, intensity levels were not specified. Two hundred
comparably aged unexposed controls were also examined. No methods of subject
selection were reported, nor was the sex of the participants reported. After
dilation of the subjects' pupils, lenses were examined with an ophthalmoscope
and a slit lamp. It was not stated whether procedures ("blinding") were
applied to mask the group assignment of the subjects for the examiners. Lens
changes were noted in 168 of the subjects vs. 148 controls. This difference
was calculated as statistically significant. The result was presented as a
summary measure over all ages; age-specific differences were not presented.
The authors stated that the controls were age-matched.
5-300

-------
In the same study, the effects of longer-term exposure were evaluated by
comparing 100 controls with 102 employees, drawn from the original group, who
had worked with high-frequency-electromagnetic-wave generators for more than
4 years. Subjects were graded on a five-point scale for degree of lens
opacity. The mean grade of opacities in the exposed group was greater than in
controls in each 5-year age-group interval for ages ranging from 20 to 50
years. Among microwave workers, the mean grade of lens changes, uncontrolled
for age, also showed an increase with length of employment. Although this
part of the study was stated to be focused on employees with four or more
years of work experience, the data on lens scores and duration of employment
list results for persons with less than four years of employment. This
evaluation also suffers from a lack of quantitative measures of exposure.
In 1972, Appleton and McCrossan reported a clinical survey conducted
among 226 military personnel stationed at Fort Monmouth, New Jersey.
Microwave-exposed workers were defined as those who worked with Signal Corps
electronic communication, detection, guidance, and weather equipment.
Controls did not report such a work history. Likely intensities or
frequencies were not specified. There were 135 controls and 91 exposed
personnel. Personnel were drawn from the post's Occupational Vision Program,
but sampling or selection procedures were not documented. Ophthalmologists
based diagnoses on si it-lamp examination of dilated pupils. The presence or
absence of opacities, vacuoles, and posterior subcapsular iridescence was
noted. The examination results were similar in the two groups over all age
groupings, although no statistical tests were applied.
5-301

-------
The survey was later extended to six other installations with results
reported by Appleton in 1973. The exposure of military personnel was
classified as "likely" vs. "unlikely." The latter group served as controls.
Blind procedures were utilized for the examining ophthalmologists; that is,
they were not aware of the exposure classification of subjects. The same team
of examiners performed all tests, except at one location, in an attempt to
minimize inter-observer variation. The same three end points used in the 1972
study were also used in the 1973 study. Older age groups demonstrated a trend
of greater opacities among exposed personnel, but, since the numbers in some
age groups were small, the validity of this result is questionable.
Odland (1973) reported results of a survey of ocular anomalies in
personnel from eight military installations. The population consisted of 377
exposed individuals and 320 controls. Exposure conditions were not specified.
Exposed personnel were defined as individuals whose primary duties involved
the operation or maintenance of radar equipment. The selection criterion for
controls was duty that did not permit actual or potential exposure to radar.
The actual work assignments of control persons were not stated. Medical
histories were taken, and ophthalmic examinations were performed in a double-
blind manner. The occurrence of lens anomalies was similar in the two groups;
however, the frequency of anomalies between control and exposed groups was
different for individuals who had a family history of diabetes, nontraumatic
cataract, glaucoma, or defective vision. Lens changes were noted in 29
percent of the exposed individuals with such a history vs. 17 percent in the
controls with a family history of eye problems. No statistical tests were
applied to any of the reported frequency distributions.
5-302

-------
Shack!ett et aK (1975) reported eye examinations of 817 military and
civilian personnel. There were 477 persons with a history of microwave
exposure and 340 controls without exposure drawn from eight Air Force bases
between November 1971 and December 1974. The authors stated that detailed
work histories were recorded (including time with different types of equip-
ment), but information on typical exposure settings is not given. Local unit
commanders selected the subjects by using criteria established by the
examining team. Standard diagnostic criteria were established. The same
ophthalmologists performed all examinations and were not aware whether a
subject was considered as exposed or a control. No differences were noted
between the two groups in the frequency of opacities, vacuoles, and posterior
subcapsular iridescence. Differences in results were not statistically
significant, but the type of test was not stated. An age-dependent increase
in lens changes was noted in both groups.
The study by Siekierzynski et aT. (1974b) discussed earlier also compared
2
lens opacities in the two exposure groups (less than 0.2 mW/cm and 0.2 to 6
2
mW/cm ). Ophthalmologic examinations were performed. Lens translucency was
assessed with a slit lamp after pupil dilatation and according to criteria
established for five grades. No differences were reported between the
exposure groups nor within the groups for duration of exposure.
5.10.4 Reproductive Effects
In 1965, Sigler et aK reported a history of occupational exposure to
radar and more military service among fathers of children with Down's Syndrome.
5-303

-------
The association with radar exposure was an ancillary observation; this study
was specifically directed toward examination of the relationship between
ionizing radiation exposure and Down's Syndrome. A case-control approach was
used, and 288 children born with Down's Syndrome in Baltimore between January
1946 and October 1962 were identified for inclusion in the study. Controls
were selected by matching each case of Down's Syndrome with another normal
birth for (1) hospital of birth, (2) sex, (3) date of birth, and (4) maternal
age at birth of child. Of the original 288, 216 matched pairs were available
for final analysis. Eliminations occurred for various reasons such as non-
cooperation or equivocal diagnoses. Occupational histories and other data
were obtained by interview of the parents. The fathers of the children with
Down's Syndrome had more military service experience than control fathers, but
the difference was not statistically significant. A greater history of radar
exposure, as radar technicians or operators, was reported by case fathers.
This difference was statistically significant. It should be noted that radar
operators are not necessarily "exposed," since the place of operation may be
away from the power-generating equipment or the microwave source.
Cohen et aK (1977) extended the study. The case series was expanded to
include 128 additional verified cases and their matched pairs born through
1968. To serve as an independent replication, essentially the same procedures
were applied along with the following expansions: (1) more extensive questions
on microwave/radar exposure and military service, (2) validation of exposure
histories by searching armed service records, and (3) chromosomal studies.
The previously noted differences disappeared in the extended analysis as seen
in Table 5-30. Results of the cytogenetic analyses are not yet available.
5-304

-------
TABLE 5-30. PATERNAL RADAR EXPOSURE BEFORE CONCEPTION OF INDEX CHILD
(FROM INTERVIEW AND/OR NAS)a



Down1
s Cases
Controls
Case Series'3


No.
%
No.
%
1. Current
Exposed

20
15.7
27
21.3

Exposure
Known
127
100.0
127
100.1
2. Original
Exposed

36
18.6
30
15.2

Exposure
Known
194
100.1
198
100.0
3. Combined
Exposed

56
17.4
57
17.5

Exposure Known
321
99.9
325
100.0
aData are from Cohen et aK 1977.
^1: 128 pairs; 2: 216 pairs; 3: 344 total pairs.
The authors stated that although the extended study did not confirm
excess radar exposure among Down's case fathers, the possible relationship of
such exposure to increased risk of Down's offspring cannot completely be ruled
out. They added further that the most challenging aspect of the investigation
was the definition of radar "exposure." In discussing explanations for lack
of confirmation of a radar exposure factor (if one does in fact exist) the
authors offer several possibilities. As mentioned above, inaccurate exposure
estimates could distort results. The role of maternal factors in Down's
Syndrome is so important that paternal factors could be masked. It was also
suggested that the most severely impacted males WitK the highest risk of
abnormal offspring also experienced germinal tissue damage, eliminating them
from reproductive experience or ascertainment through live birth cases. It
5-305

-------
was further suggested that a prospective approach might, then, prove more
fruitful. Although not discussed by the authors, radar equipment and military
occupational specialties could have changed over time in such a way as to
lower risk. It has also been suggested that since microwave-generating
equipment, especially of an older vintage, may have emitted more ionizing
radiation than modern equipment, the ionizing radiation presents the actual
risk factor, with microwaves possibly operating as a co-variable.
Lancranjan et cH. (1975) studied 31 adult males with a mean age of 33
years and a mean exposure of 8 years (a range of 1 to 17 years) to electro-
magnetic fields that "frequently were in the range of tens to hundreds of
2
pW/cm ." The frequencies were defined as microwaves of wavelengths between 3
and 12 cm and frequencies between 10,000 and 3,600 MHz. No details on the RF
source(s) were provided. A group of 30 men of similar mean age and no known
exposure to microwaves served as a control for the analysis of spermatic
fluids and hormones. Statistical analysis of the results showed no
differences in urinary content of 17 keto-steroids (as an indirect measure of
Leydig cell function) or total gonadotropin between the exposed and controls.
Statistically significant decreases were reported for exposed personnel in the
number of sperm per milliliter of semen, percent of motile sperm in the
ejaculate, and the percent of normal sperm.
The goal of this study was to try to apply objective measures to assess
the subjective reports of decreases in libido or of other sexual disturbances.
This was accomplished, but the exposures are poorly defined, and the number of
5-306

-------
men evaluated was very small. Spermatogenesis improved in two-thirds of the
subjects after cessation of exposure. The investigators felt this supported
an argument for microwave influence on the observed alterations. It is
interesting to note that the values obtained by the semen analyses for both
exposed and control groups might be considered to be low-normal or below
normal values.
Table 5-31 summarizes those studies on human beings for which SAR values
could be estimated.
5.10.5 Unresolved Questions
There are some general issues surrounding the use and applicability of
epidemiological research methods to study the effects of environmental agents,
including radiofrequency radiation. Examples of such issues include the
ability of epidemiological studies to detect low-level risks, to separate the
effects of multiple causes, and to identify and control confounding factors.
But, in addition, there are specific problems seen in the literature on human
beings exposed to RF radiation. The reports in the literature have some
important and common problems that limit interpretation and use of the data in
determining population exposure levels. These problems are briefly discussed
i
bel ow.
5.10.5.1 Exposure Assessment--
This is perhaps the largest single problem in conducting epidemiological
studies on RF radiation. It is difficult to determine the actual exposure of
individuals, and even of groups, for a variety of reasons such as difficulties
5-307

-------
in developing methods of personal dosimetry for this type of radiation. There
may be no continuous surveillance programs in the workplace, for example, that
could yield data for use in epidemiological studies. In many cases, health
studies, especially of chronic conditions, are conducted at a point in time
when it is hard, if not impossible, to reconstruct exposure patterns. The
study by Robinette and Silverman (1980) is a good example of attempts to deal
with the difficulties of retrospective exposure assessment, e.g., two
approaches (job type vs. Hazard Number) were used and an exposure gradient was
obtained with the Hazard Number which simultaneously considered the RF sources
and the length of service assignment. Despite these efforts, the results are
still estimates of potential rather than actual exposure. On the other hand,
the levels and frequencies of exposure are either not known, not estimated, or
not reported in many other studies. If well-developed exposure data is not
available, it is difficult to develop dose-response relationships, to inter-
pret the significance of human health studies, and to use the data in
establishing protective exposure limits.
5.10.5.2 Documentation and Methods--
Another problem in the RF radiation literature on human beings relates to
documentation of methods and procedures. Degree of detail in reporting seems
to be a major difference between studies done in the U.S. and those conducted
in other countries. The paper "Guidelines for Documentation of Epidemiologic
Studies" (Epidemiology Work Group 1981) suggests the types of topics that are
useful to document when reporting an epidemiologic study, especially one used
to support regulatory decisions. These major elements include background and
objectives, study and comparison subjects, data collection procedures, and
5-308

-------
analysis. Few reports on human beings exposed to RF radiation adequately
address these topics. It is frequently hard to tell whether certain research
methods were applied or simply not reported. The selection and use of control
groups is an example. The criteria for selecting controls are frequently not
stated. Controls are often said to be comparable in all respects except
exposure but analyses or data to support such statements may not be supplied.
Also, practices such as development of standardized rates or use of procedures
to control confounding variables, e.g., age adjustment, are not common in the
literature. Statistical power is rarely evaluated or discussed; however, it
is hard to estimate this if the underlying prevalence or incidence of the
disease under study is not well known. This is especially true for some of
the conditions and symptoms, e.g., functional disturbances, studied in
relation to RF radiation exposure.
5.10.5.3 Health End Points: Design and Populations—
Another issue is the medical significance of any changes that may be
induced by exposure to RF radiation. For example, the studies on ocular
effects usually have examined a subclinical end point, e.g., lens opacities,
which may not necessarily be an early marker or risk factor for cataracto-
genesis and visual problems (Silverman 1979). Further, the study populations
have not been followed to develop incidence or longitudinal data; thus, most
studies present only single measures at one point in time. Future studies
should try to determine whether symptoms and subtle changes lead to any sub-
sequent disability or disease; longitudinal data needs to be developed. A
similar concern surrounds the problematic information on functional changes
and nervous system effects reported in some of the Eastern European
5-309

-------
literature. Not yet substantiated, the potential for neurological or
behavioral effects needs to be more thoroughly and rigorously evaluated. In
addition, standardized questionnaires and more objective medical measurements
need to be applied in such studies.
Most studies concern occupational groups of relatively young healthy
males. It cannot be presumed that sensitivity, or lack thereof, to radio-
frequency radiation exposure would be the same in the general population as in
working groups. The general population is more diverse with the full range of
ages, sexes, races, and other factors that could influence health status or
the development of disease. To resolve this issue, specific exposed popula-
tions could be identified and evaluated if such research appears to be
feasible.
5.10.5.4. Summary—
There are some serious methodological problems in the human studies
literature that meke results equivocal. These problems are not necessarily
insurmountable. But, at present, the data on human beings exposed to RF
radiation is not yet sufficiently developed to be very useful in determining
exposure limits for the general population.
5-310

-------
TABLE 5-31. SUMMARY OF SELECTED HUMAN STUDIES CONCERNING EFFECTS OF RF-RADIATION EXPOSURE
Exposure Conditions
Effects
Species
Frequency
(MHz)
Intensity
(mW/cm2)
Duration
(Years)
SAR (V/kg)
(est)
Reference
No significant change in health status
of exposed personnel
No differences in three major
diagnostic categories between
the two groups of microwave workers
No differences observed in clinical
evaluations More subjective
complaints in exposed group
No effect on life span or cause
of death
No effect on mortality in a
military population followed for
more than 20 years
Decreased number of sperm/ml
of ejaculate
Reduced percentages of normal and
motile sperm in ejaculate
Human
adul t
male
Human
adult
male
Human
adul t
male
Human
adult male
and female
Human
adul t
male
Human
adul t
male
400 (PV)
2,880 (PW)
9,375 (PW)
Radar
Radar
2,560-4,100
600-9,500
-4
-4 (avg)
-4 (avg)
<0 2 (avg)
>0 2 (avg)
6 0 (max)
<5
3,600-10,000
Tens to
hundreds
jjW/cm2
0-13
1-10
5-10
0 005 (max) 22*
0 018 (max) 0 5
200-5000 (est, PW) ~1
(routine)
100
(occasional)
1-17
8 (avg)
~0 16
~0 12
-0 12
<8x10
-3
n"3
>8 x 10
24 (max)
<0 2
-4
2 x 10
(max) .
7 x 10
(max)
<0 05
<5
Barron and Baraff
1958
(See also Barron et M
1955)
Czerski et at 1974,
Siekierzynski et
1974a, 1974b
Djordjevic et aj^ 1979
Li 1lenfeld et al 1978
Robinette et al 1980
0 3-4 x 10 2 Lancranjan et a^ 1975
'Number of years of irradiation of the embassy and length of the study period, but the average exposure of individuals is estimated to be 2-4 years

-------