About Us
Contacts
Company Tour
Products
Non thermal studies
Our Research
Magnetotherapy
Electro Pollution
Endogenous Fields
Media File
The Coghill Challenge
The Natural World
Origins of Electricity
EMF Video Archives
TETRA
Photon Platinum
Review of 2003 NRPB Consultation Document
BIMT Prospectus
How to find us
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| RF/MW RADIATION OF ORGANISMS: IS SPECIFIC ENERGY ABSORPTION RATE (SAR) THE CORRECT DOSIMETRIC APPROACH? | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Roger W Coghill and Gerard Hyland EXECUTIVE SUMMARY There will be 1.3 billion cellphone users by 2005, according to Deutsche Bank forecasts. Scientific concerns have increased sufficiently, beginning with studies some forty years previous, over the possible hazard to health from using cellphones in close proximity to the brain, to persuade regulatory authorities in an increasing number of countries to label the strength of cellphone radiations as a guide to safety. The metric chosen for cellphone and other instrument or device labelling is the specific energy absorption rate ("SAR") which aims to represent the amount of thermal energy deposited in an incremental set mass of organic tissue over a specified set period of time. The construction of this metric is fraught with difficulties, and at present no standard approach has been defined. At cellphone frequencies hot spots over fifty times the average level of thermal energy deposition can occur deep in the brain, whose prediction by means of SAR is highly uncertain. There is moreover good support for believing that energy deposition is not the only parameter by which adverse biological effects can be engendered from weak radiofrequency (RF) or microwave(MW) radiations well below levels needed for thermal effects. In consequence the populations of China and Russia (some 1.4 billion) are protected by legislation and regulatory limits around a thousand times stricter than values recommended as maxima in the West. Urgent research programmes are beginning in the US the UK, Australia and elsewhere in order to establish the extent of non thermal risk from cellphone use, but will not report conclusions for some years. This paper examines the basis for using SAR as a safety marker, and argues that it is the wrong metric, and inappropriate for measuring non-ionising radiation-related effects, for four main reasons. Firstly, there are now over a hundred studies showing biological effects al levels well below the SAR levels regarded as threshold to health risk. These fall into at least twelve categories of effect. Secondly, the values of tissue conductivity on which SAR levels are based differ considerably between equally competent laboratories, making the basic calculation of SAR suspect. The same inconsistency goes for the averaging tissue mass to be considered and the averaging time of exposure. Thirdly, human phantoms used to assess electric field strength within simulated tissue are too imprecise and too crude to warrant reliance on their results. With every increase in model sophistication has come a realisation that SAR levels are higher than previously indicated by cruder models, yet even today these models are poor substitutes for real tissue. This is partly because at RF/MW frequencies the user is in the radiating near field of the transmitter, when plane wave conditions are not established and effects difficult to predict. Moreover living tissue has higher conductivity than dead tissues, on which some calculated fields are based. Finally, though improvements in numerical dosimetry have occurred over the years, even in the view of those close to this science the present four major numerical methods use techniques still exhibiting serious imprecision. Against that background evidence is beginning to emerge that a number of biological mechanisms which have been developed through evolution in an electric-field-free environment are being perturbed when exposure arises to electric fields. These include cardiac rhythmicity, energy synthesis (ATP), signal transduction (calcium efflux), the permeability of tight junctions (blood brain barrier), sleep disturbance and alterations in EEG, immune system dysfunction (leading to solid tumours of the brain), effects on key DNA repair enzymes and others, and on DNA integrity itself. Many of these effects have been reported at levels well below those of thermal effect. Unless the inadequacy of SAR is quickly recognised, a burgeoning number of cellphone users may be susceptible to serious disorders from excessive use. Fortunately the very researches which have identified hazard have also led to suggestions how to mitigate the adverse effects of these weak radiations. 14 A review of the studies reporting no biological or clinical effects from weak cellphone frequency radiation exposures appear to be at frequencies other than those used by cellphones, or at power densities much lower than those needed for cellphone transmissions, or funded by parties interested for reasons other than science. Copyright ã 2001 Roger Coghill INTRODUCTION Continued rapid global uptake of mobile telephony persuades industry forecasters that the number of handset users will rise from some 490 million in 2001 to an estimated 1.35 billion in 2003+ (source Deutsche Bank). The cellphone is arguably one of the most radiative domestic appliances yet invented, and its mode of use normally requires that it is held close to or touching the head, which in turn is arguably the most sensitive organ of the body. Such considerations have led some scientists to question whether there may be adverse effects on user health, and to seek what mechanisms of interaction are responsible for reported biological sequelae on humans, animals and organic cells from acute and chronic cellphone exposure (e.g. Adey, 1997). In response to these concerns several large health effects research programmes have been initiated, (e.g. World Health Organisation -$3 million, the UK Govt’s Dept of Health - $10 million, and others in Australia -$4 million, etc.), though the outcome of these will not be available for some years. Meanwhile the generally accepted metric for assessing any possibility of health risk is the Specific Energy Absorption Rate ("SAR"), which reflects the amount of dose in terms of energy deposited into the body’s tissues. Basically this aims to prevent such energy from causing damage by heating the body more than the capability of the blood circulation to dissipate the heat. In the US and UK as well as other countries, regulatory authorities are preparing to require that all cellphone handsets carry written indications of SAR levels, and the public are likely to accept this as evidence that the appliance is in compliance with accepted safety standards. The means of ascertaining SAR levels with any exactitude is however fragile. Moreover, towards a 100 studies have now reported biological effects, some potentially or actually adverse, at levels of energy deposition too low to have any significant thermal impact. This has caused some researchers to ask whether some other metric might be more appropriate, based not on thermal considerations, but on other parameters or a combination of metrics, and recognising the disturbance of organic processes in ways not related to heat. In the absence of an agreed mechanism of interaction though it is not yet possible to set standards based on non-thermal effects, it is equally invalid to rely on SAR alone as the sole metric of risk assessment. This paper reviews the suitability of SAR for standard-setting, and evaluates the evidence that other biological mechanisms than simple thermal insult may be important in risk management and public health. The historical basis of Specific Absorption Rate (SAR) Over the last fifty years the only unequivocal mechanism of interaction of RF energy with body tissues is thermal. Accordingly when microwave exposure limits were first put in place they were derived from the concept of tissue heating. Early studies on radar operators exposed to microwaves before and after the end of WW2 reported adverse biological effects e.g. on the haemopoietic system (Daily, 1943; Lidman and Cohn 1945; Follis, 1946), but even then these were not considered to be solely due to heating effects (Worden, Herrick et al., 1948), and some very early studies recognised that heating may not be the only cause of convulsive response to acute exposures (Duggar, 1936). Studies in 1953 by Schwan, Michaelson and others aimed to ascertain how much heat was necessary to lift body temperature by several degrees (Shiefelbein, 1979). This turned to be around 100mW/cm2. A tenfold safety factor was then imposed and the result 10mW/cm2 was incorporated unofficially with little or no debate by 1957/8 into US military standards, formally adopted in 1965, and approved in 1974. Dr Schwan himself consistently maintained that his dosage limit was safe probably for no more than one hour, but the same standard was adopted for whole-day occupational exposure. The first ANSI standard (USAS C95.1 -1966) that allowed 10mW/cm2 as averaged over any 0.1 hour period across the 10MHz to 100GHz frequency range was formulated explicitly to protect "mankind". The welfare of animals or the global ecology was not considered, nor was the possibility of any non-thermal effects. During the 1950s it became apparent that the US Embassy in Moscow was been systematically irradiated with very weak (4-18mW/cm2) microwave radiation. Moreover the health pattern of the Embassy staff showed elevated incidence of compromised immune systems. Project Pandora aimed to investigate the possibility that weak radiations of this kind might impact adversely on health, and the resulting studies (by Dr Abraham Lilienfeld and colleagues at Johns Hopkins University) appeared to confirm the non-thermal effects. The lessons of the US Moscow Embassy were ignored, partly because any meaningful downward revision of the existing 10000 mW/cm2 standard might incur liability claims from service personnel, and would have seriously curtailed the expansion of military EMR use (see Smith and Best, 1989, pp 212-217 for brief review of the issue). Reviewing the early research between 1940 and 1960 Cook argued that some of the prior research did not proceed in a professional or scientific manner, and therefore no conclusions could be drawn from papers on dosimetry presented at the suspect Fourth Tri-services Conference (1960). In effect, states Kane in 2001, Cook was indicating that the dosimetry studies had provided artificially optimistic findings (Cook, 1980). Meanwhile in the eastern bloc there were distinct indications from ongoing research that microwave irradiation produced adverse histochemical effects on rabbits and guinea pigs (Baranski et al., 1972a). These studies were also published in Western literature (Baranski, Arber et al., 1972b) and cellphone industry researchers as well as standards regulators must have been aware of them. By 1980 it had became apparent that the relation between radiation and dose was not linear, and that a new approach to protection should be adopted. Researchers like Easterly were beginning to propose that weak magnetic fields could cause cancer by proliferation of latent tumour cells (Easterly, 1981). The existing regulations were based (and still are) on thermal effects only, but SAR is effectively also a thermal measure, and originated with the dosimetry concept first suggested by the American National Standards Institute (ANSI) in 1982 based on previous considerations concerning ionising radiation dosimetry, where biological effects could robustly be related to energetic dose. It adds to the simple level of radiation intensity by incorporating tissue conductivity and density in order to arrive at a dose-related metric. Thus the concept of SAR in non-ionising radiation was born. But even in 1984 it was admitted that much remained unknown about the fields close to radiating elements (Borup and Gandhi, 1984). SAR was subsequently adopted during the period 1984-1995 by other international regulatory bodies (DIN/VDE in 1984, NCRP in 1986, NRPB in 1986, IRPA in 1988, TTC/MPT in 1990, and CENELEC in 1995). This blanket acceptance makes the SAR concept somewhat difficult to challenge, being already embodied in so many regulatory limits, guidelines and standards, and used for reference in many research studies. One major attempt to challenge the existing regulatory position, which echoed similar concerns expressed in 1971 by the Electromagnetic Radiation Management and Control committee (ERMAC), was made by the Office of Health and Environmental Assessment of the US Environmental Protection Agency in 1990, who published to a large number of interested individuals for their evaluation a draft 420 pp document reviewing the biological effects of electromagnetic fields, but which printed on every page the words "Draft: do not quote or cite" (EPA, 1990). This document arguably became the most quoted and cited of all reviews in the literature of bioelectromagnetics, and led explicitly to a major review in the UK by an Advisory Group set up under the auspices of the NRPB and chaired by Sir Richard Doll (Doll et al., 1992). The purpose of the Doll report was clearly to downplay the worrisome conclusions of the EPA draft, widely published in the media, that ELF electromagneic fields are possibly carcinogenic, and that "radiofrequency fields modulated at the same ELF frequencies that cause some of the effects noted above also result in the same responses". It is instructive to realise that the original dose limit was arrived at via a few behavioural studies on simians (macaque monkeys, trained to an auditory observing response task for pellet reward). Original studies by De Lorge and Ezell (1980) on rats consisted of seeing when work stoppage occurred in their response to increasing microwave radiation, and not necessarily relating to thermal effects at all. The limit of 4W/kg SAR was a mean derived from several results (SAR of 3.75 for 1280MHz, and 4.9 SAR for 5620MHz) with disruption occurring within 30-60 minutes of exposure. Subsequently de Lorge tested monkeys on the auditory response task, exposing the simians to 225, 1300, and 5800 MHz. (de Lorge, 1984). Disruption of performance was observed after 30-60 minutes at 8.1mW/cm2 for 225MHz (equivalent to an SAR of 3.2W/kg), at which point the monkeys’ task response rate had dropped to only 33 percent and body temperature had increased by one degree C, and was therefore an extremely crude outcome. For 1300 MHz the disruption occurred at 57mW/cm2 (SAR 7.4 W/kg) and for 5800 MHz at 140mW/cm2 (SAR 4.3W/kg). Whether this was a solely thermal effect or not is debatable. From these figures an average of ~4.0 W/kg was derived, and this crudely-derived short term value has dominated regulation guidelines ever since, without consideration for any longterm effects until after the 1985 cut-off point for the IEEE ANSI C95.1 (1999) standard. The de Lorge Studies were replicated by D’Andrea but he was actually a colleague of de Lorge, so the independence of the replication could legitimately be challenged. Nevertheless in an early study (D’Andrea and Gandhi, 1977) the authors found that rats trained to press a lever for food when exposed to 600MHz radiation stopped their pressing-for-food activity at 7.5mW/cm2. This is perhaps one third of the power of a cellphone at user distance. It was also early recognised that the RF frequency range of 750-900MHz was much more efficient at depositing energy in tissue than at 2450MHz, the frequency of microwave ovens, used in many early studies (Freidenthal et al., 1984). Moreover these early studies said nothing about long-term effects. In a later effort to ascertain long-term effects D’Andrea et al., (1986a) exposed rats to 2450MHz 7hrs/day, 7 days/wk, 0.7 W/kg and found that behaviour was disrupted. In a separate study (D’Andrea 1986b) exposed rats to 2450MHz for 7hrs/day, 7 days/wk, 90 days (SAR 0.14 W/kg) and also found a small disruption in behaviour. He concluded from these studies that the threshold for behavioural and physiological effects of chronic RFR exposure in the rat occurs between 0.5mW/cm2 and 2.5 mW/cm2, that is between 0.14 W/kg and 0.7W/kg (D’Andrea et al., 1986a, 1986b). These later studies however were not applied to the revised ANSI standard in 1991, because the cut off period was 1985. In many instances since the 1970s effects on behaviour were being reported in the peer-reviewed literature at SAR values of less than 4.0 W/kg. For example:
These reports continue to appear and in the 2001 Bioelectromagnetics annual meeting, discussed later, no less than 14 studies reported effects from RF/MW radiation at non- thermal levels. By 2001 a large number of studies, probably towards 100, had reported biological effects at SAR values or power densities too low to induce thermal effects. At the same time negative studies were also being reported, particularly from US Air Force Base research groups such as Brooks Air Force Base, and from Battelle. A recent review of the literature suggested that there was a clear need for harmonisation (Erdreich and Klauenberg, 2001). Significantly, a trawl through the studies of both viewpoints - some 29 in total - reported at the June 2001 Bioelectromagnetics Society meeting in St Paul, Minnesota revealed that all of the US originating studies were negative in outcome, whereas nearly all of the positive studies originated outside the US, Japan, and Finland (major cellphone producing countries). Moreover, most of those studies reporting positive outcomes were from exposure at about 900 MHz, whereas many studies with negative outcomes had exposed the target at around 1800 MHz. The tables below summarise these presentations. Positive Effects reported
Negative outcomes
Note: The 15 negative outcome experiments originated from only ten laboratories, and the 15 positive studies from 14 laboratories. Most of the positive studies were in vitro with a broader range of targets, whereas the negative studies tended to aim at replication of previously reported non-thermal effects (BBB,ODC, MN). A recently published literature review from a historical viewpoint (Kane, 2001, self-published) reviewed in depth many of the studies from the 1950s through the mid 1990s. The author argued that a large body of published research pointed clearly to the existence of hot spots at cellphone frequencies and that these were ignored by the regulatory authorities. Absorption of energy is at its most efficient at the frequencies chosen for cellphone transmissions, making the brain most vulnerable to adverse effects including DNA damage, chromosome damage, tissue damage, cataract formation, tumour formation, memory loss, motor skills degradation and other consequences. He cites over 130 studies to support his case, and most of these are set out with brief descriptive text in a separate Bibliography. The scientists whose work is most frequently cited by Kane include Adey, Balzano, Chou, Cleary, Durney, Guy, Gandhi, Lai, Lehmann, Lin, Schwan, and Stuchly, all researchers who have spent their careers in research in this field at reputable establishments. Altogether there appears to be a formidable body of evidence supporting the conclusion that SAR has not proved an adequate metric for health effects research, and moreover that the guidelines previously and presently in place are untenable against an overwhelming body of evidence that non- thermal effects below these guidelines can have adverse consequences for health. Differing regulatory standards, guidelines, and limits Some confusion arises with the SAR criteria, since they can vary considerably with different authorities, a fact recognised in the Stewart Committee report published in May 2000, and elsewhere. The table below compares the ANSI (US), CENELEC (European), International Committee on Non-Ionising Radiation Protection (ICNIRP), and Japanese SAR limits: TABLE 2: Differing regulations related to SAR in different countries
In 1999 the International Non Ionising Radiation Committee (ICNIRP) guidelines for the public were adopted in a European Council Recommendation, which has been agreed in principle by all countries in the European Union. However in the UK the NRPB guidelines are different again, and much more tolerant of exposure levels. Whilst Western guidelines and standards are at least in agreement within one order of magnitude, in some other parts of the world quite different standards have been adopted. In the former USSR permissible operational levels for an 8 hour workday are 2 V/m-1 for frequencies between 30-300MHz and 5 mW/cm2 for frequencies between 0.3 and 300GHz., the latter being relevant to cellphones, and compare with 1-5milliW/cm2 in the USA, i.e. the US standard is some 1000 times less strict. Russian researchers who had been investigating RF/MW exposures since the 1960s came to the conclusion that the human nervous system is responsive to EMF at levels lower than thermal, and that EMF possesses an informational action in its effect on humans (see Kositsky, Nizhelska et al., 2001 for a detailed review). The most frequent typical reaction in humans is observed during changes in the exposure frequency (Andreev, Beliy et al., 1985), suggesting that the informational content is frequency rather than amplitude dependent. This effectively means that there is a diametrically opposed and unresolved difference in viewpoint of how RF/MW interact with humans between Western and the Russian researchers. Commenting on the Western regulatory differences Stewart (2000) recommended that as a precautionary approach the (stricter) ICNIRP guidelines for exposure be adopted for use in the UK rather than the NRPB guidelines. Though Germany had incorporated these into statute Stewart did not see the need to do so in the UK, because "We believe they are liable to change as more scientific information on positive health effects becomes available". Almost none of the USSR research was referenced by Stewart, who only cited four studies from the Russian literature, and these were from several decades previous, with no mention made of the large discrepancy between the Western and USSR exposure regulations. In China the regulation of microwave exposure likewise has been based on experimental evidence such as medical examination, and epidemiological analyses of personnel exposed to EMFs (eg Chiang, Yao et al., 1989). (See also Chiang, abstract of SCC28 presentation, Fall 1999). It had been demonstrated there that chronic exposure to EMFs are associated with a variety of non-specific symptoms including increased frequency of neuroses, lability of the vegetative nervous system and slight changes in peripheral blood, lens and non specific immune competence. The limits for public exposure in China at 0.3-300GHz are 10mW/cm2, and for occupational exposure 40mW/cm2. In terms of SAR these equivalate to 0.02W/kg for the public and 0.1W/kg for occupational exposure. This means that towards a quarter of the world’s population is protected from MW radiation upto 1000 times more strictly than those living in the West. The above illustrates the fragile regulatory nature of SAR as a governing metric in different countries, and the need to standardise the appropriate metric, "once it has been demonstrated to be scientifically sound" (Stewart, 2000, para 1.51) . SAR Calculation difficulties In order to understand the various units referred to the reader is directed to Appendix Five where a Conversion Table assists in following the power densities being discussed in different submultiples of their basic W/m2 equivalent. The physical amount of energy radiated from a cellphone is relatively easy to characterise. Typical maximum electric fields at 2.2 cm distance from a 2 Watt power handset in use and transmitting at 900MHz (whose wave length is 33.3 cm) are 400 Volts per metre (V/m-1) and for a 1 Watt power handset transmitting at 1800MHz (whose wave length is 16.7 cm) it is 200 V/m-1. The maximum power density at the same distance is about 200W/m2 (source IEGMP, 2000, 4.27) i.e. well above the exposure limit in China. Calculation difficulties inside the body Internal to the body the same precision is not possible, yet this data is essential if dose is to be accurately calculated. The maximum fields created inside the human body are said to be about three times smaller than external fields, and the average values are all appreciably less again: "The largest values of electric field E inside a model of a head whose surface is 1.4cm from the antenna were…about 120V/m-1 for a 900MHz antenna radiating 2 Watts and 70V/m-1 for a 1800MHz antenna radiating 1 Watt" (source Dimbylow and Mann, 1994). Most users hold the handset with the antenna or body of the device actually touching the head, however. This means that the user is in the near field, not the far field, of the handset, which results in a number of imponderables. In an earlier study Dimbylow et al. had reported that at 800MHz non uniform absorption would occur, in other words there was the possibility of hot spots. Though showing that at 1mW/cm2 the maximum from frontal radiation of a child’s head is 1.23W/kg, the editor curiously omitted the graphical data for 800 MHz, leaving a blank space where it should have been in the journal page (Dimbylow and Gandhi, 1991). This maximum is actually an average in a mass of 1 gram, leaving great scope for hot spots of much higher intensity inside that 1cm cube volume, which comprises billions of biomolecules. Electric fields are one orthodox way of describing the RF/MW radiation emitted in the far field of a device (i.e. several wavelengths distant) but do not reliably characterise near field radiations, that is, where a plane wave has not yet been established, and the magnetic field is not perpendicular to the electric field. Since typical carrier wavelengths from cellphone handsets are 16.7 and 33 cm long it is clear that the normal user is exposed to their radiating near fields, though people nearby the user may also be exposed to far fields. External electric field strength measurements moreover do not by themselves answer the question of how much of this radiation is actually absorbed by the body. In order properly to assess possible bio-effects, appropriate dosimetric concepts have had to be developed, extending practices originating in ionising radiation dosimetry. Accordingly the received opinion has been that impact of cellphone radiations can be assessed in the same way as ionising radiation. This concept of dose, specific energy absorption rate or SAR, expressed in Watts per kilogram, is conventionally derived from the internal electric field strength (rms) via the following formula (in its simplest form): SAR = sE2/r (1) where s is the conductivity of the tissue in Siemens/metre (or mho/metre), and r is its density in kg/m3. The Siemens is the SI unit of electrical conductance, equal to the reciprocal of the ohm and replacing the equivalent MKS unit (mho). However some experts use the formula s|E|2/2r, which is confusing. The difference relates to the difference between average whole body SAR and peak spatial SAR (which involves the square root of 2). Moreover, the conductivity in organic tissue is argued to be a composite of several different tissue types with differing permittivity (e), and the density of tissue can vary substantially, so this measure not only needs careful computation, but is subject to a number of other complexities discussed in detail below. One immediate problem is that different laboratories have derived quite different conductivities for body tissues: for example those presented by Gabriel in 1994 differed substantially from those from the University of Maryland (Davis et al., 1996). Moreover there have been a variety of materials used as substitutes for human tissue in models. None of these are particularly accurate: a homogeneous substance developed by Guy and known as 2/3 muscle tissue for example as used in some studies was considered as a tissue cocktail to represent human tissue when in fact it is non-representative and ignores boundary effects. The difficulties of creating a suitable model to simulate the complexities of living human brain tissues were recognised and described well before the 1980s, one example being the way the body can control local blood flow to keep specific regions cool (Foster, Kritikos et al., 1978). A good deal of the incident radiation is deposited in the user’s brain and converted to heat, but it is almost impossible to predict exactly where or how much in any one site (Gandhi, 1994). Guy himself had earlier questioned whether any useful numerical results could be economically obtained (Guy et al., 1976), saying "It appears doubtful, however, that any useful theroretical or numerical solutions can be economically obtained for figure shapes as complex as man.." Bernardi (1998) compared the electrical parameters for some of the head tissues as derived from various laboratories, and found an alarming spread of SAR calculations based on widely differing conductivity and permittivity values: TABLE 1: Differing SAR values from various laboratories(e/s)
The table shows variations from 20-80 percent, both for the permittivity and conductivity values, and illustrates well the fragility of SAR based on experimentally obtained values. Notice also that primarily SAR is energy- rather than frequency-dependent, whereas there have been reported many biological effects related to frequency at non-thermal levels of field strength. SAR Measurement difficulties There are a number of important factors which immediately make SAR more difficult to apply to non-ionising radiation (NIR) than was the case with ionising radiation. If the appropriate health risk assessment criterion relates solely to the body’s ability to tolerate heat, a straightforward definition of a dose which correlates well with heating effects is not possible. This is because whole body or local temperature increases not only depend upon the amount of energy absorbed and the effects of passive heat dissipation but also to a large extent on complex thermoregulatory processes in the body (Kuster and Balzano, 1997). Such processes in turn depend on a number of parameters - specific organs, environment, health status etc - and it is not the initial temperature increase in the tissue used to define the dose but the power absorbed per unit mass, in other words the Specific Absorption Rate is a more usable metric where energy is regarded as the prime parameter. The issue of hot spots Hot spot absorption is a term used to describe excess radiofrequency energy being deposited in some small localised region, particularly of the brain, but also of any other tissue mass. The factors determining where a hot spot occurs are complex, and involve head size, shape, curvature subcutaneous fat layer, thickness, internal skull structures, and ventricles or voids within the skull. No dosimetric model has ever mimicked these accurately, and even had they, the variation between individuals, and between children and adults, is too wide to predict accurately the location of any hot spot. Moreover hot spots are also dependent on the type of antenna, the structure of the telephone, and how the phone is held during conversation, as well as the frequency. For example the type of antenna used - e.g. l/2 dipole, l/4 monopole, l/2 monopole - can lead to a large variation in SAR (Hombach and Thielen, 1994). Nikola Tesla first drew attention to the fact that RF radiation could heat human tissue in 1890, and proposed its use as a therapeutic device: "In fact to put it broadly, it is conceivable that a person entirely nude at the North Pole might keep himself comfortably warm in this manner. "Without vouching for all the results, which must of course be determined by experience and observation, I can at least warrant the fact that heating would occur by the use of this method of subjecting the human body to bombardment by alternating currents of high potential and frequency such as I have long worked with. It is only reasonable to expect that some of the novel effects will be wholly different from those obtainable with the old familiar therapeutic methods generally used. Whether they would all be beneficial or not remains to be proved". The concept of diathermy was developed by Arsene D’Arsonval and used extensively in medical practice, not without concerns at possible side effects. Indeed, researchers such as J.F. Lehmann and "Bill" Guy have pointed to the existence of hot spots for four decades. Balzano (1978b) recorded that "The temperature profiles generated by both antennas inside the head of the simulated operator indicated the presence of a hot spot about one inch below the surface of the temporal bone". This study was at around 840MHz, within the normal range of a cellphone transmission frequency. At around the same time Durney (1979) reported that children absorbed some 50% more radiation than adults, and that thin men absorb some 33 percent more radiation than an average 70 kg man. Even before then Lin (1976) had suggested that hot spots could be ten times higher at certain areas in the centre of the head compared with near- surface levels, and Schwan had initially characterised hot spots in 1972 as a result of non-uniform absorption characteristics. His laboratory also found the presence of hot spots across a wide range of frequencies (Kritikos and Schwan, 1976), but most notably between 800 and 1000 MHz, the frequency region used by many GSM cellphones. Lin (1976) found that hot spots became more pronounced as head diameter decreases, so that women and children are more likely to sustain hot spot effects than adults. Such findings were confirmed by a number of other researchers, e,g, Johnson and Guy (1972) had reported intense fields in the centre of the head at 918MHz, rather than attenuating with increasing distance from the surface: "…for human brain exposed to 918MHz power the absorption depth 2.3 times the depth of penetration (3.2cm) is twice the absorption at the surface. This corresponds to a factor greater than 200 times that expected…" They explained this by saying that "The regions of intense absorbed power density are due to a combination of high refractive index and the radius of curvature of the model which produces a strong focussing of power towards the interior of the sphere". Johnson and Guy pointed out that the frequencies of most utility for diathermy were also potentially the most hazardous from the point of view of hot spots. In 1974 Guy had reported how diathermy treatment conducted at 750 MHz and 915 MHz (two of the frequencies authorised for this therapy) was noticeably greater in effect than at the low Gigahertz frequencies, and without significant heating in the surface tissues (Guy et al., 1974). Similar findings had been reported by Lehmann et al., as early as 1962. In later work, while researching the phenomenon known as microwave hearing, Lin and colleagues found that power densities of as little as 0.1mW/cm2 could result in SARs of 140W/kg in laboratory rats in the near field of 918 MHz transmissions, where they had previously determined that this should result in a SAR of only 0.09W/kg, in other words a hot spot of 1500 times the expected value (Lin et al, 1977, Lin, 1977). This level of SAR would cause immediate tissue damage. Other early research by Joines and Speigel (1974) where the authors had created a six tissue layer model of the brain, found that enhanced energy absorption compared with earlier and simpler models, and reported that the hot spot frequency range extends to 2100MHz, thus embracing the other major cellphone frequencies in the 1800 MHz range. Some experiments showed dramatic temperature increases even from short duration exposures. A five minute exposure to a suface density of as high as 100mW/cm2 could yield an internal temperature rise of 12 degrees, almost certainly fatal. Scaling this down to 10-20mW/cm2, the level experienced by most cellphone users, the rise would still be about 1.2 to 1.4 degrees C., excluding any hot spot effects (Gandhi, 1975). Many cellphone calls well exceed a five minute duration however. The situation is further exacerbated when there are nearby reflecting materials. In 1977 Gandhi pointed out that for corner shaped reflectors the increase could be as high as 27 times, and nearly five times for reflections from a flat surface (Gandhi, 1977). Not all scientists even then were convinced that SAR was a competent metric. Guy in 1976 had expressed concerns that any useful numerical data could be easily determined, saying "It appears doubtful, however, that any useful thoeretical or numerical solutions can be economically obtained for figure shapes as complex as man". (Guy et al., 1976). In 1978 Durney et al., had compiled a Dosimetry Handbook for radiofrequency radiation (Durney et al., 1978). Even though using a crude model of a rhesus monkey with a mixture of salt water and plastic powders, Iskander and colleagues reported in 1981 that "it is surprising that the average SAR of the rhesus model … is nearly three times the expected value based on the empirical data found in the dosimetry handbook of Durney and co-workers (1978)". It was clear by 1980 from the relatively crude models then used that near-field RF radiations were a cause for concern (Chatterjee et al., 1980). By the mid 1980s several other laboratories had confirmed the extent of hot spots in skulls with a radius from 0.1 to 8 cm, which embraces nearly all humans and many animals (Kritikos and Schwan, 1975; Wong et al., 1984). Reviewing the position in 1980 Gandhi advised that several other countries like Sweden and Canada were abandoning the 10mW/cm2 standard (Gandhi, 1980). He later also pointed out that the early IEEE/ANSI standard was based fragilely on a few behavioral studies in laboratory rats, where the threshold of effect was sometimes taken to be when the animal almost completely ceased activity (Gandhi, 1987). One problem in considering hot spots is that several minutes are needed for measuring any build up in temperature, but heat loss during this period is negligible (Guy, 1984), so the use of thermographic methods by some early researchers is questionable. Several authors were later to point out that unlike the case with rat studies where the whole body was irradiated, over 50 percent of RF radiation from a cellphone handset would be absorbed by the user’s head (Kuster, 1993; Gandhi, 1994; Lovisolo, Raganella et al. 1994; Hombach and Thielen, 1994). Lovisolo calculated that at 0.6W and 900MHz, the power density would average 1.9W/kg, implying that in any 10 gram mass there would be regions above ANSI’s regulatory limit of 1.6W/kg. "It was estimated that 72 percent of this was burnt off in the brain", the authors claimed. This absorption would in part produce hot spots from near field radiation. These near field exposures are the most likely source of irregular heating in localised areas i.e. hot spots (Hines and Randall, 1952), and studies at 433,750, and 918 MHz confirm that energy is readily absorbed from the induction fields in the near zone (Guy 1971). This absorption efficiency at around 900MHz was confirmed in later rat studies (Hjeresen, 1983), where the brain temperatture could be elevated enough to cause convusions while the surface temperature remained relatively low. Frequencies below 900MHz were found to be more penetrating (Guy, 1974), while absorption within the brain was found to be 20 times greater than in the skull and subcutaneous fat. Part of the reason for this is the focusing effect of the skull (Guy et al., 1986). In 1985 the research group at Utah (Borup and Gandhi, 1985) reported that the hot spot effect was also seen in plane wave induced radiation: in that instance the observed hot spots amounted to 0.6 SAR at an incident power density of only 1mW/cm2. Since power densities from cellphones are around 10-20mW/cm2 this implies SARs of some 6-12 W/kg, sufficient to cause substantial temperature rises in the hot spot regions. Back in 1955 Schwan and Piersol had speculated that if the power level was sufficient to induce "a feeling of warmth on the skin, the deep temperatures which are higher than the superficial ones, will be elevated to a point that may bring about tissue destruction". In the near field the same level of incident power as that reported by Gandhi’s group in the far field would result in a much higher SAR due to enhancement factors including a phenomenon known as "matching". Matching is explained by Kane (2001) as follows: "The subcutaneous fat layer in humans lies beneath the skin, and varies in thickness from one person to the next. Certain thicknesses actually cause more of the radiation to be absorbed deep within the body. The thickness of these layers, together with certain antenna distances, can establish what is known in the scientific community as a "matching" effect. Fat layers and bone may serve as matching layers to help with this enhanced absorption of energy. If fat and bone layers are about 1 cm in thickness it is possible to maximise the absorption so that nearly all the radiation is absorbed into the brain or muscle". This suggests that the human head, where the brain is enclosed within bone and subcutaneous fat, is ideally suited to efficiently absorb radiofrequency energy. The general point to be made here is that these early "hot spot" studies, though using what at best may be described as simple inhomogeneous models or exposures of laboratory animals, had reported the existence of hot spots at radio frequencies used by cellphones long before they became popular. Yet no mention or warning is given by manufacturers of this potential hazard, and even today hot spots remain a controversial yet unpublicised issue among those concerned with standard-setting. Largely as a result of hot spots there continue to be numerous studies evidencing biological effects from exposure below the limits recommended by IEEE/ANSI C95.1 1992 (spatial peak SAR of 1.6W/kg). Some of these studies are listed in the table below, including those evidencing higher than limit SARs from cellphone power densities. Not all are peer-reviewed since some were reported at Bioelectromagnetics Society annual meetings, and though scrutineered by a Program Technical Committee, are not fully refereed:
Spatial averaging difficulties The use of SAR as a metric requires a further parameter in order to maintain the relation between the absorbed power and the induced heat, namely the SAR must be averaged over a specified time period, called the averaging time. Near field exposures can however lead to considerable local heating long before the whole body average SAR limit has been exceeded. Next a minimum mass of tissue over which the SAR must be averaged also has to be defined, in order to account for strongly localised heating at the skin, e.g. caused by spark discharges. This is called the averaging mass. Finally the SAR of the whole body average must be adjusted to protect partial body exposures, where resonant effects are an additional factor. The issue of partial body averages has only been superficially addressed in regulation. Time averaging difficulties The biological bases for defining an appropriate time period and a suitable minimum tissue mass are extremely tenuous, hence the sometimes tenfold differences between regulatory authorities. The choice of six minutes by some regulators is not based on any biological parameters (e.g. irreversible cell changes) but simply happens to be 0.1 of one hour. As an example of the effect of time, a five minute exposure at 10-20mW/cm2 (the level of energy experienced by a cellphone user) the equivalent rise within tissue woukd by 1.2 - 2.4 degrees C (Gandhi, 1975), not counting any absorptions due to hot spots. This duration - or even six minutes for that matter - is less than many cellphone conversation lengths. Local hot spots are in any case more likely at cellphone frequencies. Mass shape difficulties It is equally tenuous to settle on a cube as the appropriate shape, since a wide variety of tissue shapes exist within the body generally and the head in particular, with few if any being even remotely cubic in shape, despite the perceptions of certain impressionist painters. Kuster and Balzano (1997) argue that if a hot spot distribution is assumed to be the most general occurrence, a spheroid with the local peak SAR at the centre would be much more appropriate. They also point out that the SAR distribution caused by a given exposure can only be assessed using human equivalent phantoms and by drawing on considerable technical resources: "In contrast to ionising radiation, the relationship between the exposure and the induced dose or SAR distribution is significantly dependent on various exposure parameters, such as the frequency and the field polarisation, as well as on the exposed biological bodies. The human body has complex surfaces and internal geometries and it is composed of tissues with spatially varying dielectric properties. Furthermore the mere presence of the body significantly alters the field distribution. In the case of near field exposures the coupling between the body and the electromagnetic source can even alter the performance of the source". A subsidiary issue is that no matter what shape is chosen, if the object is close to reflecting surfaces such as automobiles or metal structures, there is an enhancement of energy absorption, sometimes as high as 27 times the norm with a corner shaped reflector (Gandhi, 1977). Difficulties with dosimetric studies Despite the need for human phantoms to establish dose, their use by separate and equally distinguished research groups has shown that there are great difficulties in performing accurate and reliable dosimetric studies. A good example comes from the research carried out by a number of groups in the 1980s (Chatterjee et al., 1995, Guy and Chou, 1986, Stuchly et al., 1986a, 1986b, 1987, Kuster and Ballisti, 1989, Cleveland and Athey, 1989, Fujiwara et al., 1990) to establish whether a previous regulatory exclusion clause for devices below 7 Watts should be abolished. From these as an example Cleveland and Athey reported that testing performed at 900MHz and 0.6 Watts power output provided averaged SAR levels of as much as 1.9 W/kg, which indicates that in the 10 grams of tissue taken for that average there must be areas well above this level (Cleveland and Athey 1989). However the authors used synthetic materials with only simple layering to measure the radiation dose, producing what is hardly a realistic representation of human head tissues. Some authors readily conceded that homogeneous phantoms were of little use in determining localised exposure: "…though it has been demonstrated by other researchers that homogeneous whole body phantom models with an electrical conductivity 2/3 that of muscle will provide the most realistic condition for determining whole body average SAR, this is not valid for local partial body exposures nor is it valid for determining SAR distribution within the model". (Guy and Chou, 1986). Partly for this kind of reason the results of these early research groups proved contradictory and difficult to interpret with respect to what they variously concluded about this exclusion clause. Only in 1992 did a study by Kuster and Balzano clarify the situation finally making it clear that the exclusionary clause was in clear contradiction of the spatial peak SAR limits, after which it was abolished, but even then not universally. Sweden and Japan still accept the 7 Watt exclusion clause. Despite the immense increases in computer power and various numerical techniques developed in the last two or three decades, general electrodynamic problems are still difficult to resolve. From the engineering viewpoint it is useful to distinguish between capacitative coupling, (i.e. dominantly electric field induced currents) and inductive coupling (i.e. dominantly magnetic H-field induced currents). The tissue type also complicates the issue: the most relevant biological tissues have a high water content, and therefore a relatively high permittivity (er >30). The capacitative coupling is therefore poor in most cases, whereas the inductive coupling is often found to be a dominant factor, especially at higher frequencies and in low impedance fields. There is also the important but complicated issue of cumulative dose, particularly applicable to cellphone use. Very few studies have addressed the question of how SAR relates to duration and repeated application. Secondary to this, few studies have investigated the importance of ambient fields from emitters in the vicinity, which may be very different in separate laboratories or even in the same laboratories at different times. Non thermal effects Introduction: non-thermal effects and the kT problem. As indicated, towards a hundred studies have now reported biological effects from RF/MW exposure at field strengths too low to have any thermal impact on the results. This poses a problem for physicists in understanding how such effects can occur and by what mechanisms. At around the level of human temperature (310 degrees Kelvin) molecules are subject to random movement described as thermally generated electrical noise, known also as Brownian motion. This is defined thermodynamically to be proportional to the square root of kT Du, where k is Boltzmann’s constant (1.381 x 10-23 J K-1, equivalent to 86meV per degree), T is the absolute temperature of living tissue, (310K) and Du is the frequency bandwidth over which the noise is considered. Since at around 300K the value of kT is some 26meV (see below), if this is much larger than the energy of the motion produced by the electric field of interest, then any effect of this field is likely to be masked by thermal noise. The issue of thermal noise (kT) within organisms raises in turn the question of possible effects on existing ionic processes and its intensity in relation to the intensity of the insulting artificial field. The energy quanta of radiation at cellphone frequencies is less than 10microelectron volts (10meV) - actually 4 and 7 meV for 0.9 and 1.8 GHz respectively. By contrast, energy of 1eV approximately is needed to dissociate the weakest chemical bonds in DNA. It would therefore seem unreasonable to expect RF/MW radiation has a direct fracturing effect on DNA. The organism may however have its own mechanisms for processing DNA changes, and a weaker electrical stimulus might still be capable of triggering DNA fractures. If so the organism would need to distinguish between them and the electrical noise which is always present in materials, called thermal noise or kT, and caused by the random collision and movement of ions, atoms and molecules in all biological tissues. Thermal noise consists of k, which is Boltzman’s constant multiplied by K the temperature in kelvins (i.e. 273 degrees plus degrees centigrade, so 37 degrees C is the same as 310 degrees K.). The value of kT at around 300K is 26 meV, and much larger than the energy from cellphone radiation, so any effect of the field on ion movements could be masked by kT. Weaver and Astumian (1990) point out that one implication of this is that an interaction mechanisms restricted to a narrow frequency band (a resonant type response) is able to respond to a much lower level of background signal because of the improved signal to noise ratio. This in turn implies that there are structural components in the organism prepared by evolution to receive these specific frequencies, and similarly that there are organs capable of emitting the relevant frequencies for the purpose of regulatory growth and maintenance control. We shall later argue that such structures can indeed be identified. In the case of ions driven to and fro by an oscillating field the extent of motion is reduced by the viscosity of the surrounding liquid. For a field of 100V/m Stewart (5.14) calculates that the movement is less than 10-14 - the diameter of an atomic nucleus - and the energy associated with this motion is less than the thermal motion of the ion by a factor of 1015. For example, the velocity of the ion is mE = mEo sin(2pvt) where m the mobility is about 10-7 m2/(V/s) for chloride, the ion of highest mobility, and v , the frequency, is 0.9 or 1.8 GHz. This leads to a maximum displacement of mEo/2pv, which for an electric field Eo = 100V/m,, equals 2 x 10-15 and 10-15 at frequencies of 0.9 and 1.8 GHz respectively. So the average kinetic energy of an ion of mass m in this field is mm2Eo2/4 or mm2E2/2 where E is the rms value of the electric field. For a chloride ion this is equal to about 10-17 eV, or about 10-15 kT. This energy is so small it is considered not to be capable of producing any biological effects. Notwithstanding this argument, it is very evident that biological effects are produced by fields of this order of magnitude and well below them, since it has been known for decades that elasmobranch fish and other fish too respond to moving magnetic fields and oscillating electric fields of vanishingly small field strength (Lissmann and Machin, 1958; Kalmijn, 1966). In Lissmann’s famous study at Cambridge University of object location in the fish Gymnarchus Niloticus (a fish living in the black sedimentary waters of the river Nile) he tested its sensitivities to small direct currents and found that it could detect potential gradients of as little as 0.03mV/m in the surrounding water. Other studies found that catfish (Amiurus nebulosus) and goldfish were also remarkably sensitive to such currents (Regnart, 1931; Abe 1935). Lissmann also observed that a static magnet could be used to train a similar fish (Gymnarchus carapo) to feed, and noted that a fish swimming through the field would have small currents induced in it in the same way a moving magnet will affect a stationary fish. Lissmann was able to demonstrate that the fish’s sensitivity was dependent on the water conductivity and not by the presence of any particular ions in the water, by gradually adding KCl to a pure water environment until the conductivity of the aquarium water was reached. Kalmijn’s lifelong studies of electrosensitivity in sharks and rays reported the same exquisite sensitivity (Kalmijn, 1966, 1999). This problem of animal sensitivity to electric fields below thermal excitation levels has led researchers to propose a number of different mechanisms, not involving thermal effects. Before considering these (in a separate paper), however, it may be first useful to categorise the kinds of non thermal effects being reported in human subjects and in cellular preocesses. Non thermal effects by category These may be divided into at least twelve major categories of effect. Inevitably there is some overlap, and the categories of effect are not mutually exclusive: for example effects on the immune system will impact on cancer propensity. In view of their wide diversity it is unlikely that all non thermal effects can be explained by one single mechanism. However in a separate paper is presented a plausible biological system which embraces most categories of effect, underpinned by the explicit structural arrangement of all cerebrate organisms and by robust experimental evidence. 1. Ion movement effects 2. Calcium efflux 3. Ion Cyclotron Resonance, and similar mechanisms 4. Effects on Ion receptor proteins 5. DNA, sister chromatid exchange (SCE) and micronuclei (MN) related effects 6. Cell membrane related effects 7. Effects on enzymes 8. Immune system related effects 9. Effects on EEG and sleep; cognitive effects 10. Effects on the blood brain barrier 11. Ocular effects 12. Cancer related effects 13. Effects related to cancer subtypes and solid tumour formation To recapitulate, the normal formulation for constructing SAR values comprises three components, one being the rms electric field, and the other two (conductivity and density) being descriptive of the exposed tissues. However in the near field (i.e. within say 16 cm of the transmitter) it is the magnetic H field which dominates the energy, and there is no direct relation between the magnetic and the electric field at radiating near field distances. Hence measuring the electric field can say nothing about the strength of the magnetic field in a particular and usually complex exposure near field condition. A large number of studies have reported significant biological effects when the RF/MW carrier signal is modified at ELF frequencies, a condition having no important impact on the density of energy delivered (hence on SAR values as derived in the preceding paragraph), whereas without the modulation (continuous wave or CW) no effects were found. 1. Ion movement effects Stewart offers the argument that even resonant effects are unlikely in such circumstances. Ions (from the Greek word travel or move) are charged particles, and may be electrons (negatively charged) or the atoms and molecules from which they have been dissociated (usually positively charged). Molecules and atoms with a net positive charge are called cations (e.g. Calcium 2+) and those with a net negative charge are called anions. Since electrons seek stability by acting in pairs with contra-rotating spins , when they are freed from their atoms they seek other unattached or attached electrons as a means of stability, and may thereby damage nearby stable molecules (free radical action). This universal physical law that like charges repel and opposite charges attract is used in many ways in organic processes. Though ions are driven to and fro by the motion of the oscillating electric field, for a 100V/m field the distance moved is calculated at only 10-14 metre - the diameter of an atomic nucleus- and the energy associated with this motion is less than that of thermal motion of the ion by a factor of 1015. Stewart argues that this makes any non thermal biological effects from ionic movements unlikely. However, if the ion is thereby unable to attach to the receptor of interest as a result of the oscillation, then the biological effect can be large. Adair (1994) concedes that the energy increases with the mass of the object, and though cells of average size (10 microns) are unlikely to be affected, ionic movement might be recognisable with larger masses [such as whole organs]. This is unlikely to be the whole story however. In the cell plasma membrane (and in the mitchondrial inner membrane) are ion channels and a myriad configurations of sialic acid residues on glycoproteins embedded in the plasma membrane which are prepared by evolution to receive very specific electrochemical messengers, (of which calcium 2+ is one). These messengers bring regulatory instructions to the cell and instruct for important intracellular processes, carried within the cytoplasm by microtubules, again prepared by evolution to transport these messages throughout the cell or to the cell nucleus. Enzymes such as Na+, K+, APTase regulate these channels so that ions may selectively pass through in either direction. We shall discuss the effects of electric fields on microtubules and these regulatory enzymes later. Meanwhile it was reported by Tsong and Astumian (1988) that Na+ pumping can be induced by exposure to a 1MHz field of 2kV/m. The same authors produce evidence to suggest that cells are capable of using electric fields in ion pumping. Lednev of the Russian Institute of Biological Sciences, Puschino, proposed that ions bound to cell membrane proteins act as spatial oscillators with a series of vibrational frequencies that depend on the bond energy and the charge and mass of the ligand bound ion ("ion parametric resonance": Lednev, 1990). Liboff of Oakland University, Michigan, from an earlier but different theoretical approach also suggested an ion cyclotron resonance mechanism, and offered some experimental evidence, as did Lednev, but these have not proved robust. A third contributor to the idea of ion-based mechanisms were Blackman of EPA and co-worker Janie Blanchard, who also offered experimental support for their variant (Blackman and Blanchard, 1994). These studies were all at ELF frequencies, but have relevance when the issue of modulated RF fields is considered. 2. Calcium Efflux studies Among the earliest research into weak RF/MW bioeffects was the work evaluating calcium efflux in response to applied EM fields carried out at Loma Linda during the 1970s by Suzanne Bawin and colleagues under the general direction of W. Ross Adey, and also later the same decade at the Environmental Protection Agency (EPA) in Maryland by Carl Blackman. (Bawin et al., 1975; Blackman 1979; Adey 1980; Adey, 1981). Both these groups found the effect was maximal at 16 Hz. At least three other laboratories have confirmed the calcium efflux effect of modulated RF exposure (Sheppard et al., 1979; Dutta et al., 1984, 1989; Kittell et al., 1996), while three others have found no increased efflux (Shelton and Merritt, 1981; Merritt et al., 1982; Albert et al., 1987). The Adey, Dutta, and Blackman studies (on chick brain) were at 147 MHz, (though Dutta also used 915 MHz continuous wave) modulated between 0.5 Hz and 35 Hz. The Merritt studies (on rat brain) were at 1 GHz and 2.45 GHz frequencies, with pulses of 10-or 20 ms and 16-32 pps for 20 minutes. Moreover these negative studies were at higher flux densities namely upto 150 W/m2, compared with only 10-20 W/m2 in the early Bawin studies. Commenting on these differences Stewart argues that the finding is of no particular relevance to mobile phone use because the amplitude modulation within the critical band is very small. Adey (1989, 1993) has suggested that the observed calcium efflux may be due to an amplification process in which weak electric fields might be set up in the tissue at the ELF frequency of amplitude modulation and that these may act as a trigger for the initiation of long range co-operative events within the cell membrane.It is important to note that two of the calcium efflux studies were conducted on animals in vivo rather than solely on cells in vitro. One consequence of such efflux is to dampen neuronal activity, by stabilising the neurons electrically. In line with this expectation Arber and Lin (1984, 1985) found an increase in membrane conductance and a decrease in the spontaneous firing of impulses in a snail model (Helix aspera) when exposed both to CW and amplitude modulated 2.45 GHz radiation at relatively high intensity and clearly depended on a temperature rise. Research by Somosy and co-workers at the Hungarian Academy of Sciences included a study of the ultrastructural distribution of calcium after ELF modulated microwave and GSM modulated RF radiation in the temporal cortex of rat brain (Somosy, Kittel et al., 1999). At one hour radiation with intensity 1mW/cm2, frequency 2.45GHz, and an estimated SAR of 0.6mW/kg a significant increase of calcium level was seen in the synaptic cleft and also around cells, but when the irradiation was at 900MHz modulated at 271 (?217) Hz, this effect was not observed. Ionic calcium performs an important role in signal transduction in a number of different organic processes, including nervous conduction across synapses, second messenger roles in intra-cellular fluids, and in the regulation of heart beat rate. In the cytosol calcium is normally held in stores in the form of calmodulin, so that its presence unless carrying a signalling function, is minimal. Accordingly disturbance of this cation could interfere with several vital cell functions. 3. Ion Cyclotron and parametric Resonance Certain membrane proteins ("pumps") actively transport ions across the plasma membrane using energy derived from ATP. Other channels serve as conduits through which organic ions (e.g. sodium, potassium, chloride, calcium) can move across the membrane, and thereby perform vital cellular functions. RF fields even at non-thermal levels can affect these movements: Repacholi recently concluded from a World Health Organisation review of the literature that RF fields continuous or pulsed could affect membrane channels, based on reports of decreased rates of channel formation, decreased frequency of channel openings and increased rates of burst-like rapid firing (Repacholi, Basten, et al., 1998). RF exposure has also been reported to influence the ATP-dependent sodium -potassium pump in the membranes of red blood cells (Allis & Sinha-Robinson, 1987; Liu et al., 1990) and there has been considerable debate as to the mechanisms involved. Some scientists have proposed that the interaction of the earth’s magnetic field is involved in creating ion cyclotron resonance conditions whereby oscillating magnetic fields may thereby interfere with the normal passage of ionic calcium and other ions across the plasma membrane. A number of resonance related mechanism have been proposed, from the original formal presentation by Liboff (1985) [ICR = 1/p .q/m. Bosc : see Liboff , McLeod et al., 1990 for review], but there are some disadvantages with most these (e.g. the path of an ion in such conditions could be a metre across and during this orbit it is almost certain to encounter collision) and though debated in detail in the literature have not proved acceptable to a consensus. Though originally expressed as a function of artificial magnetic fields, the ICR hypothesis can also be applied to electric fields (Liboff, 1995). 4. RF/MW effects on ion receptor proteins The body of evidence that RF affects ionic processes is large. This supports the view that the concept of SAR is not correct, since the frequency of the insulting radiation rather than its energy becomes important, evidenced by studies showing such ionic effects at levels below those likely to produce any thermal changes. Philippova et al., 1994 found that 900MHz radiation at SARs of 1 and 100W/kg specifically affects the binding of odorant molecules to receptor proteins in the membranes of olfactory receptor neurons in the rat. They attributed this to shedding of this particular protein from the membrane, probably because of increased peroxidation of membrane lipids (see Phelan, Lange, et al., 1992). Phelan, Lange et al. (1992) exposed a B-16 melanoma cell line to pulsed 2.45GHz 100 pps for 1 hour (SAR = 0.2W/kg) and reported changes in membrane ordering, due in part at least to the generation of oxygen free radicals. Liburdy and Vanek (1987) had also reported protein shedding from membranes as a result of RF exposure. Neurotransmitter receptors are also affected by very low intensity RF exposure. D’Inzeo, Bernardi et al., (1988) reported a decrease in the frequency of opening of sodium channels associated with acetylcholine receptors in muscle membranes as a result of exposure to 9.75GHz radiation at only 10-20mW/m2, which might cause a decrease in the excitability of the muscle. Another set of studies during the 1990s reported that the flux of positively charged sodium and potassium ions across cell membranes can be affected by RF exposure over a wide frequency range from 27 MHz to 10 GHz (Cleary (1990a, 1990b, 1995). Though mostly at thermal intensities Cleary found a temperature window between 17.7 and 25 degrees C., leading others to imply facilitation of lipid phase transitions in the membrane near the phase transition temperature (Tenforde and Liburdy, 1988). Cleary persistently claimed that RF exposure causes increased proliferation in glial brain cells when exposed to 2.45 GHz at 25mW/g SAR, but the work has not been replicated by the cellphone industry or elsewhere except at his own laboratory (Cleary, 1990, 1994). In 1988 Cleary offered an overview of RF nonthermal bioeffects, concluding "cellular studies provide convincing evidence that RF radiation, and other types of electric or magnetic fielods, can alter living systems via direct nonthermal mechanisms as well as via heating" (Cleary, 1989). Writing in a 1990 review Cleary further considered that "Not surprisingly in vitro brain cell sensitivities to RF exposure are among the highest recorded" (Cleary 1990). Stewart comments that since the human body temperature is well above Cleary’s in vitro temperature windows (37 degrees C) then such effects are not likely to occur in vivo. In relation to cellphone penetrations of the head it is useful to remember, as Stewart (2000) points out, that in the hippocampal region of the brain and in the cortex, changes in the level of intracellular calcium resulting from incoming synaptic activity can lead to long term changes in the strength of synaptic imputs onto the neurons, and are thought to be involved in mechanisms of memory and learning (see Kandel et al., 2000). In their study on ion channels. D’Inzeo, Bernardi et al. (1988) also reported that exposure to CW 10.75 GHz microwave radiation at levels as low as 50mW/cm2 decreased the opening frequency of nicotinic acetylcholine (ACh) receptor ion channels by 20-40 percent within 60 seconds of the onset of exposure in chick myotubes. An attempted careful replication by Tattersall (1999) however failed to confirm the finding. Tattersall ruled out differences in bathing solution depth (2mm v. 5 mm), the possibility of a power density window, or differences in sensitivity between the samples. Having reviewed briefly this ion-related literature Stewart conceded that "There is evidence that RF fields can affect membrane proteins and can change the movement of ions across membranes… This might cause subtle changes in cell function but the significance of such effects for human health is uncertain. Moreover these effects have not been independently confirmed, which is important given the frequent lack of reproducibility of RF biological effects". 5. Effects on DNA, sister chromatid exchange (SCE) and micronuclei formation In everyday life breaks in DNA macromolecules within the cell nucleus are frequent as a result of cosmic irradiation, chemical or other insult, and for a diversity of other reasons. Normally these breaks are repaired by enzymes before any mutation can get underway, the single strand breaks presenting an easier task than double strand breaks. Micronucleus formation is also thought to reflect DNA damage and is regarded as a sensitive assay sinew the aberrant cells tend to accumulate especially if the cells of interest are slow-dividing. Even in normal tissues the presence of micronuclei can be high and variable. The earliest reports of MN formation in cells exposed to RF radiation were from Russia in the 1970s, e.g. Zalubovskaya and Kiselev, (1978) cited in McRee, 1980, who reported that exposed human and pig kidney cells developed pyknotic or vacuolised nuclei and damaged membranes, with a 30-50 percent reduced survival rate. Garaj-Vrhovac et al., (1990b) reported a variable increase in micronuclei in lymphocytes from human subjects (radar workers) with rather ill defined occupational exposure at 30-300GHz and between 1000-5000 W/m2. Other non thermal studies reporting increased MN formation were either at 2.45GHz (Vijayalaxmi et al., (1997b) or other frequencies dissimilar to cellphone frequencies (Balode, 1996; Antipenko and Koveshinkova, 1987) Vijayalaxmi et al., 1999 found no MN formations but the exposure was ulktra wide band with an estimated whole body SAR of only 37mW/kg for 15 minutes. Research by a number of laboratories has for several decades reported DNA breaks and sister chromatid exchanges (SCEs), but largely at other than cellphone frequencies and often at thermal levels of intensity. Four independent laboratories have also published data on ELF induced DNA strand breaks, confirming that ELF EMR damages DNA strands: Lai and Singh 1997a; Svedenstal, Johansson et al., 1998; Phillips, Ivaschuk, et al., 1998a; and Ahuja et al., 1997. Adey while at Loma Linda had drawn attention to reported damage to DNA from RF exposure at the annual meeting of BEMS in Copenhagen 1994, pointing to Sarkar’s findings at New Delhi of band pattern changes in DNA fragments from brain and testis tissue in mice exposed to 2.45MHz radiation at 1.18 W/kg (1mW/cm2) for 120-200 days. At the same conference Verschaeve and his colleagues in Belgium had reported "a clear in vitro synergism between 954MHz waves and the mutagen mytomycin C", reaffirming their poster presented in 1993 at the EBEA Congress at Bled (HRE P23), which had reported that a 2hrs exposure at 2.45GHz resulted in a statistically significant increase in chromosomal aberrations and micronuclei but not SCEs. Chegrinetz and Gotz had also reported aberrrations at 150-300 MHz and 5 mW/cm2 at the same Congress. The former authors later considered that thermal effects might have been responsible for their results (Maes, Collier et al., 2001) since they were unable to replicate the effect, but conceded that the study was a short term one, and that it "may eventually show that 900MHz electromagnetic firelds may have some, though very subtle biological effects that are insufficient to be clearly demonstrated by the cytogenetic methods". Most attention has focused on the mid 1990s Lai and Singh work and the failure of a later independent group (Roti Roti and colleagues) to repeat the reported damage. Accordingly a detailed comparison of the two studies is given below. Earlier examples of such studies include that of Maes, Collier et al., (1995) who reported a synergistic effect between microwaves from cellphone frequencies (954MHz) and Mitomycin C (MMC, a potent radiomimetic) on human lymphocyte cultures as assessed by the sister chromatid exchange assay. Similar results had been reported by an Italian group from Naples University (Scarfi, Lioi, et al., 1996), but in a later study with human lymphocyte cultures from five subjects exposed for 15 minutes to 1.748 MHz, no effect on micronuclei formation, an indicator of cancer initiation and implying mutational sensitivity, was observed. Immune system impairment - decreased NK cells and a lower value of T-helper/T-suppressor ratio - at higher than cellphone frequencies was also reported in microwave exposed workers (Dmoch and Mosczcynski 1998). These examples show the complexity of the conflict in different in vivo and in vitro studies reporting DNA breaks, SCEs, and MN formation. Stewart (2000) briefly reviewed in vivo laboratory experiments related to cellphone communications and listed 24 DNA, SCE and MN related studies. Of these much interest has centred on the work of Lai and Singh at Washington University, Seattle, who have developed an assay based on the extent of debris from DNA damage (the Comet Assay: Lai and Singh, 1995, 1996). They reported both single and double DNA breaks following MW radiation at non thermal levels, but attempts to replicate the work have proved unsuccessful (Malpaya, Ahern et al., 1997). There were five distinct differences in the study designs of Lai and Singh at the University of Washington, Seattle compared with the Roti Roti group at Washington University, St Louis, Missouri funded by Motorola (Malyapa, Ahern et al., 1997) which explain why, though both groups used 2450 MHz , the latter found no effects. The tail DNA fragments extend out to 250 microns in the assay developed by Singh (1994) whereas in the Malyapa study, using a method developed by Olive (1992) the DNA fragments extend to less than 40 microns. The following factors make the Lai-Singh assay more sensitive: The Lai-Singh assay achieved complete lysis of the DNA using highly concentrated salt and two detergents. Lai-Singh used proteinase K to remove positively charged bound protein from the negatively charged DNA strands so that the electrophoresis field produced greater migration. DNA and protein have opposite charges and so for the electric field to cause migration the protein must first be removed. Lai-Singh used antioxidants during electrophoresis Lai-Singh carried out electrophoresis for a longer time to allow longer tails to form in the "Comet". Their electrophoresis field was 0.4V/cm for 60 minutes compared with 0.6V/m for 25 minutes by Malyapa et al. Lai-Singh used YOYO-1 dye. YOYO-1 is 100-fold more sensitive when bound to DNA than the propidium iodide used by Malpaya et al. Two other laboratories who used the Lai-Singh methodology also found that EMR, including cellphone radiation at extremely low intensities, causes DNA breaks (Verschaeve, Slaets et al., 1994; Phillips, Campbell-Baechler et al., 1998). Another important difference was that Lai and Singh used brain cells from live animals, whereas Malyapa et al. (1997a) used a mouse fibroblast cell line (C3H/10T1/2). In a second study Malyapa et al. used 836MHz frequency-modulated and 848MHz CDMA modulated radiation (simulating the frequencies used in the US) but still found no effect. Stewart argues from the increased senstitivy of the Lai-Singh assay that the controls seem unduly high when comapred with other stuyudies (e.g. McKelvey-Martin et al., 1993) and also calls the Malyapa studies "more rigorous" without explaining why. Against this more detailed examination Stewart’s conclusion about the DNA studies that "the evidence of Sarkar et al., (1994) and Lai and Singh (1995, 1996) for DNA damage in mice is contradicted by a number of other studies in vivo and is not supported by in vitro work" seems not well founded. Rotkovskaya et al., (1993) used 34GHz pulsed radiation at 200mW/m2 (Police radar) and found that DNA synthesis in cells from irradiated corneas was reduced by 25 percent, though not with statistical significance. 6. Cell membrane effects Electric fields polarise cells so that one side becomes positively charged relative to the other. This makes the cell a dipole, and it will attract and repel similarly polarised cells, which may interfere with signalling between them.. Though at frequencies below 100MHz the energies are likely to be comparable to thermal noise in electric fields of 300V/m (which in any case is far above what might be found inside the body from cellphone radiation), at RF frequencies they are calculated to be much less, depending on the detailed biological structure. This may still mean that polarisation has adverse effects, if for example, glycoprotein strands become adhered to each other through enforced adjacency, or signal transduction relying upon charges on the sialic residues of glycoproteins are disrupted. Moreover, the depolarisation of inner mitochondrial membranes would adversely affect ATP synthesis, so polarisation cannot be ruled out. Membranes are known to be strongly non-linear: in other words when a voltage is applied to one side of a cell membrane the current that flows is not always proportional to the voltage (Montaigne and Pickard, 1984). Stewart points out that the membrane acts as a rectifier, and "if a voltage is connected across the ends of a wire the size of the current that flows depends solely on the voltage. If the polarity of the voltage is reversed the current changes direction but its size is unchanged. However if the polarity of the voltage across a rectifier is changed the current changes direction, but its size also changes. This means that the AC oscillating fields might produce a net DC current, hence a net flow of products through the membrane. There are difficulties with this hypothesis, in that the response times of the ion gates are much slower than the period of microwave frequencies, and using data from measurements on membranes (Montaigne and Pickard, 1984), it was shown that for electric fields of 200V/m the relative change in membrane potential is very small.( Adair, 1994). 7. Effects on enzymes One possibility in considering mechanisms by which weak electric fields can impact on DNA is by effects on repair enzymes or on other enzymes concerned with growth regulation. A number of studies using tritiated thymidine as a marker have reported changes in the rate of DNA synthesis with exposure to electric fields (Rodan, Burrett et al., 1978; Korenstein, Somjen et al., 1984, etc). This marker is a reliable indicator since it enters the precursor pools for DNA only and unincorporated thymidine can be recovered from cell lysates by precipitation with cold trichloroacetic acid, so the incorporated label represents only the DNA which has been newly synthesised or newly repaired. Both the above studies were using strong electric fields (e.g. 116.6kV/m and 1.3-5.3kV/m respectively) but other studies at far lower strengths (10-5V/m at 10 and 16 Hz also showed elevated thymidine incorporation after a 30 mins exposure (FitzSimmons, Farley et al., 1986). Protein kinase studies As with their ODC studies Byus, Lundak et al., (1984) found similar effects on protein kinase activity in human tonsil lymphocytes, but only when the RF 450MHz field (10W/m2) was modulated with ELF and the maximum effect (a 55% reduction in activity) was seen at 16 Hz., but the effect returned to control levels after 45 minutes of radiation. These and other protein kinases are important in signal transduction pathways. Mitochondrial enzymes Melnick, Rubenstein et al., (1982) studied ATP synthesis in isolated mitochondria during irradiation for 2 minutes with 34.92 GHz waves at power densities upto 10kW/cm2. Irradiation with highpower densities resulted in an uncoupling of oxidative phosphorylation. (See also effects on ATP synthesis below). Lysosomal Enzymes Investigations at ELF frequencies (15Hz) reported a decrease in the media levels of N-acetylglucosaminidase and b-glucuronidase.(Murray, Lacy et al., 1988). Acetylcholinesterase (AChE) Several studies have reported effects while others found no response to RF irradiation. Galvin, Parks et al., (1981) studied this enzyme for electric eel cells, and found that 10 minutess irradiation with 2.45 GHz microwaves at SARs of 1-100W/kg had no effect on this enzyme’s activity, but Olcerst and Rabinowicz (1978) reported that irradiation at the same frequency at intensities upto 1.25kW/m2 for 30 minutes. By contrast, acetylcholinesterase isolated from the electroplex of a ray fish (Narcine brasiliensis) failed to produce an consistent response, though the SAR was upto 1.43 x 106 W/kg. Most significant of the studies on this enzyme was that of Dutta, Das et al., (1992) from Howard University, Washington DC, who reported that exposure of neuroblastoma cells to 147 MHz radiation modulated at 16 Hz demonstrated enhanced AChE activity. The increase was observed in a time window between 7 -7.5 hrs after the cells were plated. Alkaline phosphatase Only studies at ELF are reported. These showed effects in primary cultures of chick sternuym chondroblasts at 15 Hz, with temperature controlled to within 0.1 degree C. Exposure of sternum chondroblasts also decreased the protein content of the cultures. z) Hz) Na+, K+, ATPase Blank and Soo (1992) studied mainly ELF effects on this enzyme which is the transmembrane enzyme of the "ion pump" in virtually all cells, activated by binding Na+ and K+ on opposite faces of the enzyme The authors found that the threshold for response of the Na+, K+ ATPase to electric fields in conducting media was as low as 5mV/cm, and to the magnetic field 0.2-0.3mT. This led Blank and colleagues later to ask whether electromagnetic fields could interact directly with DNA (Blank and Goodman, 1997). Blank proposes that electromagnetic fields act as stressors inducing stress genes and stress response proteins (e.g. increased production of hps70, and that EM-induced cytoprotection is similar to thermotolerance (see Blank (1998) for review). Serpersu and Tsong (1984) had also studied this enzyme, and found electrogenic effects of exposure to an oscillating electric field. Ornithine decarboxylase (ODC) studies By far the greatest number of enzyme studies have been done on ODC (23 tissues, 7 laboratories) and have generally indicated a 2 to 4 fold increase in enzyme activity (e.g. Litovitz Krause et al., 1991). In studies by Byus and colleagues during the 1980s (Byus, Lundak et al., 1984; Byus, Peiper et al., 1987) reported non thermal effects on protein kinase and ornithine decarboxylase (ODC). This effect on ODC was important, because ODC is a key enzyme in the synthesis of polyamines, and is essential for cell growth. The polyamines putrescine, spermidine, and spermine (which contain respectively 2,3, and 4 positive charges at physiological pH, and bind tightly to negatively charged sialic acid residues on membranes, to DNA, RNA and proteins). In the Byus group studies Chinese hamster ovary melanoma or hepatoma cells were exposed to 450MHz radiation modulated at 16Hz. resulting in a 50% increase in ODC activity.. The associated aliphatic polyamines are believed to play an important role in the stabilisation of the three dimensional structure of macromolecules due to their tight ionic binding to negative charges, and because in some macromolecules they cannot be replaced by other naturally occurring cations such as magnesium. The polyamines are progressively synthesised by decarboxylation from ornithine by a rate limiting enzyme ornithine decarboxylase (ODC). Given the importance of ODC for potential growth and differentiation of cells and its sensitivity, together with the potential involvement of these polyamines in disease processes including cancer, ODC was chosen as an indicator of bio-effect during RF exposure. The 450MHz signal delivered a calculated SAR of 0.08W/kg (peak field intensity 1mW/cm2) to a culture medium containing Chinese Hamster ovary cells, sinusoidally amplitude-modulated at 16 Hz. Under these conditions the ODC activity was assayed by the amount of 14CO2 liberated from 14C L-ornithine during a 60 minutes exposure. The ODC activity was found to increase 50-80% within the first hour of exposure to the 16Hz amplitude-modulated RF field compared with sham-exposed cell cultures. The same effect was seen using 294T human melanoma cells, and remained elevated for at least one hour after removing the cells from the field. The unmodulated 450MHz carrier field produced no measurable changes in ODC activity, nor did higher modulations between 40-120 Hz, the elevations being confined to frequencies between 10 and 20 Hz. A number of other laboratories have reported similar alterations in ODC activity not found when the signal was unmodulated (Byus, Peiper et al., 1987, Byus, Kartun et al., 1988; Somjen, Yariv et al., 1983; Cain et al., 1985, Cain, Thomas et al., 1993; Litovitz, Krause et al., 1991, 1994; Mattsson, Mild et al., 1992; Valtersson, Mild et al., 1995). These laboratories used a variety of frequencies including RF modulated, 60 Hz electric, and 50-60 Hz magnetic fields. Further investigations indicated that the field need be on as little as 10 seconds in order to affect the alteration of the enzyme. An expert panel of the Royal Society of Canada reviewed the ODC literature in 1999 and reported that many studies have shown that conventional 50-60 Hz electromagnetic fields (without an RF carrier frequency) can produce a similar increase in ODC activity. The Stewart committee commented that the maximum increase in ODC activity produced by amplitude modulated RF radiation (approximately a doubling) is much less than that elicited by known tumour-promoting substances, which can cause upto 500-fold changes in ODC activity in relevant tissues. However, there is evidence that a trebling of ODC activity is sufficient to play a causative role in tumorigenesis (Moshier, Dosescu et al., 1983) It was also interesting that the Litovitz group, which used frequencies of 835 MHz and 915 MHz and exposed L929 mouse fibroblasts to SAR values of approximately 2.5W/kg in their model, found ODC activity responses at frequencies of 16-65 Hz., unlike the Byus group. The Stewart committee conceded that "pulsed modulated RF fields from mobile phones may cause a slight increase in ODC levels and activity at non thermal levels" (para 5.115). They concluded nevertheless that such changes on their own or synergistically could have a tumour promoting effect. 8. Immune system effects Non thermal effects on the cellular immune system have been investigated in vitro more thoroughly than most other effect categories in view of its importance to health and the prevention of aberrant cell proliferation. Moreover the human peripheral blood lymphocyte is a well characterised cell type. Epidemiological studies by contrast are fragmented and few in number. The prime question is whether radiations from cellphones or masts at their characteristic frequencies and power densities can adversely compromise immune competence in vivo. Early epidemiological studies (Sigler et al., 1965; Lilienfeld et al., 1978; Lester and Moore, 1982 etc.) suggest that weak chronic RF/MW exposures can have such effects, but the frequencies and power densities are not well reported. Cell studies of lymphocytes need careful management since thermal effects both above and below body temperature show effects (Roberts and Steigbigel (1977); Smith, Knowlton et al., (1978). There have been fewer studies of RF/MW effects on lymphocytes than ELF studies. Cleary, Liu et al., (1990) reported lymphocyte proliferation in isothermal conditions following exposure of whole blood to either 27MHz (SARs of 0-196 W/kg-1) or to 2.45GHz (SARs of 0-50 W/kg-1) for 2 hours. At both frequencies SARs of 50W/kg resulted in a significant depression of PHA-induced proliferation, whereas exposures at only 25W/kg had a significant stimulation effect. Neither of these were below existing guidelines. Another lack of effect following RF exposure is B-lymphocyte capping. Antibodies are expressed on cell surfaces and when bound to antigen these aggregate to form small patches. The patches then migrate to one pole of the cell, where they form a cap prior to internalisation of the antibody-antigen complexes. This in turn leads to proliferation and differentiation, resulting in an amplification of antibodies against the antigen in question. In studies of exposure at 2.45 GHz for 60 minutes, and SARs of upto 45 W/kg no difference was found between exposed and control cells (Sultan, Cain et al., 1983). A further study by the same authors (1984) found that no athermal effects followed exposure to amplitude modulated RF at SARs between 4.2mW/kg to 2 W/kg. Much more important were studies exposing murine cytotoxic lymphocytes to amplitude modulated electric fields at RF frequencies, where within 4 hours of a 48 hrs exposure period depressive effects on the cells’ ability to respond to mitogenic challenge were noted, and also for ELF electric fields (Lyle, Schecter et al., 1983, Lyle, Ayotte et al., 1988). Over a dozen studies have investigated effects of RF/MW exposure on sister chromatid exchange, mainly in lymphocytes. Only three of these were at cellphone frequencies (c 900MHz only) but four of the five micronuclei studies reported significant changes as a result of exposure, while only two of the seven SCE studies found any differences between exposed and controls. Lymphocytes from workers at radar stations were found to have increased micronuclei formations where occupational exposure was across a wide range from 30- 300 GHz, and with density ranging from 1000-5000W/m2 (Garaj-Vrhovac et al., 1990). A study of exposure at 900GHz found increased SCE frequency and enhanced mutagenic effects in human lymphocytes when the culture at 17 deg C. was also exposed to the chemical mitogen Mitomycin C (Maes, Collier et al., 1997), but not to RF alone. Micronuclei formation is believed to reflect DNA damage and is a sensitive assay (in the view of the Stewart Expert Group), since the aberrant cells tend to accumulate, especially among non-dividing and slow-dividing cells. 9. Effects on EEG rhythms and REM sleep A number of studies have investigated the effects of cellphone and other RF/MW frequencies on electroencephalographic records (EEG) and on sleep patterns. The majority of these report positive changes, but the major problem Is to assess whether such changes are adverse, against a background where the function of brain rhythms is so poorly understood. Animal studies At least three studies have reported effects on brain rhythms in animals, with a further two finding no effect. Bawin, Gavalas-Medici et al., (1973) reported changes in EEG rhythms in the brains of live cats exposed to 147 MHz amplitude modulated fields, and suggested the effects were the result of changes in calcium release. Shandala, Dumanskii et al., (1979) chose rats and rabbits as their models and found EEG changes with exposure to 2.375GHz 7 hrs/day for 30 days at a power density of 0.1-5mW/m2. Thuroczy, Kubinyi et al., (1994) used a frequency of 2.45GHz and rats exposed to CW or AM at 16 Hz, and reported changes in spectral power with SARs in the brain of 8.4 W/kg and above. Takashima, Kubinyi et al., (1978) found that at much lower frequencies (1-10 MHz there were brain rhythm changes in rabbits following long-term exposure to modulated fields. Unanaesthetised rabbits were also used by Chizenkova and Safroshkina from the Russian Academy of Sciences at Pushchino. In a 1993 study they found that exposure to 40mW/cm2 at around 800 MHz for as little as 1 minute was able to alter their EEG recording, manifested as an increase in burst activity and a sharp increase in evoked responses to peripheral and central stimulation. Von Klitzing in 1994 also reported alterations to the EEG of man from pulsed RF radiation at less than 1mW/cm2 at the Copenhagen meeting of BEMS, and published the study in 1995. Among negative studies McRee, Elder et al., (1979) described experiments by Rosensteig of the US Environmental Protection Agency who exposed rats to RF from the late fetal stage until adult, but saw no changes in either the spontaneous EEG or visual evoked responses. A later joint US-USSR study (Mitchell et al., 1989) at 2.45 GHz CW fields for 7 hrs and 2.7W/kg density found small but significant reductions in power in the EEG, but in different parts of the frequency spectrum. Human studies on EEG The consensus of these is that the use of cellphones influences brain function. According the Stewart expert group (5.193) "The evidence is sufficiently substantial to warrant further investigation notably with respect to the influence of GSM-like signals on sleep and event related EEG changes during the performance of cognitive tasks" . Reiser, Dimpfel et al., (1995) reported that exposure to GSM signals was associated with increases some 215 minutes later in the power of EEG frequencies of about 10Hz and above. Roschke, Mann and colleagues at the University of Mainz, Germany carried out a series of studies in the 1990s on EEG changes and also on sleep disturbance effects from cellphone frequency exposures. They found that exposure while the subjects were awake had no effect (Roschke and Mann, 1997), but led to a pronounced sleep inducing effect with shortening of sleep onset latency as well as suppression of REM sleep (Mann and Roschke, 1996). A later study by the same group (Wagner, Roschke et al., 1998) investigated 24 healthy young male subjects during exposure to a 900MHz, 217 Hz pulsed fields with a power flux density of 0.2W/m2. They found no spectral differences between exposed and controls, but the power density of the earlier study was much higher at 0.5 W/m and the antenna used in the later study was of a different (i.e. not circularly polarised) design. They considered that the inhomogeneity of the previous study’s linear polarised fields and mainly uncontrolled external reflections could have accounted for the differing outcomes, and did not regard the results as inconsistent. Krause, Sillanmaki et al., (2000) also found that GSM signals impacted on event related changes in EEG recorded during the performance of a memory scanning task. The principal finding was that the pattern of EEG changes changed in power compared with non-exposed controls. In three studies of event-related response (ERP) one (Urban, Lukas et al., (1998) found no effects while two others (Eulitz, Ullsperger et al., 1998: Freude, Ullsperger et al., 1998) reported suppression of high frequency spectral power (18-30 Hz) by an infrequent auditory stimulus, and a small reduction in amplitude in a visual monitoring task respectively. Further EEG effects from MW exposure were reported in 2001 (Ivanova, Martynova et al., 2001) a St Petersburg group who exposed cats frontally to 980MHz at only 0.01-0.05 mW/cm2 over a 20 minute period. The spectral power of the cat brains increased significantly in the 12-16 Hz range after one minute of exposure. In summary there are clear effects on EEG rhythms in human and animal subjects from microwave exposure well below thermal intensities, with few negative findings (e.g. Hietanen, Kovala et al., 1999 reported no EEG differences in a study of 19 subjects exposed to a 20 minutes cellphone signal, but chose a P=0.01 cut off for significance instead of the usual 0.05). The question remains unanswered whether these EEG effects have long term adverse consequences, either from acute or chronic exposure. Cognitive effects Preece, Iwi et al., (1999) in a small pilot study reported several cognitive effects out of 15 parameters, and Koivisto, Revonsuo et al., (2000) in an improved cross-over design, also studied the effects of modulated 902MHz signals in a working memory task, and found significant results in four out of fourteen measures, though the statistical basis has been criticised (see Stewart, 2000, 5.181). In a second study of cognitive tests Koivisto found a reduction in reaction time as had Preece, but the significance of these findings is unclear. On balance it is evident that GSM signals a can alter EEG rhythms at power densities well below regulatory guidelines, though whether there are any adverse sequelae from such exposure is not clear. 10. Effects on the blood brain barrier (BBB) The blood brain barrier permits the selective passage of material from blood to brain, and maintains the physiochemical environment of the brain within certain narrow limits which are essential for life. Increased permeability of the blood brain barrier as a result of microwave exposure was first reported by Allan Frey in 1975, at frequencies (1.2 GHz) not dissimilar to those of cellphones and at densities unlikely to cause thermal effects (200mW/cm2). Frey’s basic technique was to expose live animals to the frequency of interest then inject sodium fluorescein and several minutes later exsanguinate, perfuse, section and measure the fluorescence of the brain section under UV light. Fluorescence was seen in the diencephalon and to some extent in the mesencephalon and metencephalon indicating permeability in the exposed (but not the control) animals. Oscar and Hawkins ((1977) using different markers found the greatest permeability in the medulla followed by the cerebellum and then the hypothalamus. The same endpoints were permeabilised by RF radiation in studies by Albert (1977a, 1977b, 1979) using Chinese hamsters, with penetration also seen in the cerebrum. These findings were disputed in a 1978 study by Merritt, Chamness et al., in whose replication "no transfer of parenterally administered fluorescein across the blood brain barrier of rats after 30 minutes of 1.2GHz radiation between 2-75mW/cm2 was noted". When these results were reanalysed by several other scientists however the data actually confirmed the adverse effects. Preston, Vavasour et al., (1979) also claimed to find no effect from RF exposure on the BBB, and failed to confirm Oscar and Hawkins findings. Frey re-analysed their data using binomial tests to correct for the Type two error he uncovered and found the Preston et al. data were in fact consistent with those of Oscar and Hawkins. A type two error is the acceptance of the nul hypothesis when it is actually false. Frey’s reanalysis was published in the 1980 BEMS Newsletter No. 18 (pp4-5). Little regulatory attention was paid to these early studies, but more recently the issue re-emerged with the findings by Salford, Brun et al., of Lund University, Sweden, of similar effects at 915MHz and a intensities as low as 0.016 W/kg (Salford, Brun et al., 1993, 1994). Lin (1999) in an overview of biological effects of microwave radiation cites nine studies reporting BBB effects, but also another five not finding any positive effects. The five negative studies were Gruenau, Oscar et al., 1982; Lin and Lin, 1980; Preston Vavasour et al., 1979; Ward and Ali, 1985; and Williams Hoss et al., 1984. The nine positive studies are Albert and Kerns, 1981: Goldman, Lin et al., 1984; Lin and Lin 1982; Neubauer, Phelan et al., 1990; Neilly and Lin, 1986; Oscar and Hawkins, 1977; Salford Brun et al., 1993; Salford Brun et al., 1994; and Sutton and Carroll, 1979. The Lin studies were conducted with exposures equivalent to SAR of 165W/kg, which raised the rat target temperature to 42 degrees C, and the curious problem with the Salford studies is that effects were noticed at even the very lowest exposures, equivalent to SAR of only 0.016W/kg. Most of the positive findings can be explained by possible thermal effects, alteration in cerebral blood flow, or anaesthetic effects, but the issue is still very much open, with Persson, Brun et al., in 2001 presenting new evidence of albumin leakage after exposure to GSM 900 MHz radiation for 6 hrs/day during 3 weeks. Other recent studies (Fritze, Sommer et al., 1997, Nagawa, Tsurita et al., 1999, and two studies from Brooks Air Force Base Texas: Mason, Miller et al., 2001; Miller, Merritt et al., 2001) have failed to confirm the earlier positive studies. Ocular effects The advent of occupational RF exposure, and more recently the proximity to the eye of the cellphone when in use, has prompted a number of studies since WW2 designed to examine if adverse ocular effects are found. Some of the earlier of these were conducted on primates and used radiation levels differing in frequency and intensity from those of cellphone telephony, but nevertheless relevant. Among the earliest ocular /RF studies was that of Richardson Duane et al., (1948). They reported that 2.45MHz is "highly productive in producing lenticular opacities" but were using densities sufficient to raise the rabbit eye temperature to 46 degrees C. The same year a different group found that exposure of dog eyes sufficient to increase temperature by only 1.9 to 3.2 degrees was also capable of causing cataracts (Daily et al., 1948). Tissue necrosis and disorganisation of the pigment layer were also reported. That acquired posterior subcapsular cataracts could serve as a signature of non-ionising radiofrequency injury of the lens, especially when all other differential diagnostic criteria were satisfied, was discovered and first reported in 1964 by Milton Zaret, described in detail by him in 1988 (Zaret, 1964, 1988) and independently verified (Bouchet and Marsol, 1967). When a cataract is produced by a specific spectral band its location can be ascribed: posterior polar cataracts are induced by ionising radiation, brunescent cataracts by UV, and capsulopathy, an early stage of capsular cataract, from non-ionising radiation. These distinctions are based on clinical findings in variously exposed human beings. One early study funded by the US Army Surgeon General’s office was an epidemiological investigation in response to that of Zaret (1963). The authors (Appleton and McCrossan, 1972) claimed that the microwave environments at Fort Monmouth, New Jersey were not cataractogenic However, there were major flaws in this study: The authors designated as controls workers at risk of eye injury from radiations other than microwaves. The control group was composed entirely of individuals who worked with "…laser, xenon arcs, untraviolet and welding equipment". Also they used an unorthodox examination technique that renders some lens opacities invisible, and included a score for irridescence of the lens capsule, a finding that has no diagnostic significance nor comparative value and only served to dilute the significance of positive findings, according to Zaret (1988). The publishing journal did not permit these criticisms to appear in subsequent correspondence, claims Zaret. Another early epidemiological study worthy of mention is that of Hollows and Douglas, 1984, which also found that posterior subcapsular cataract (PSC) was the earliest indicator of impending occupationally induced non-ionising radiation cataractogenesis. Their study examined 53 exposed radio linesmen compared with 39 controls finding a significant 21 percent prevalence of PSC in eyes of exposed compared with only 8 percent in controls. The exposure frequency range was 558kHz to 527Mhz. By the mid 1980s Allan Frey had also reanalysed the original Appleton and McCrossan data and found that it actually revealed significant microwave-induced eye damage in humans (Frey, 1985). Frey also extended his work on the blood brain barrier to other barriers including the blood vitreous barrier, which regulates the composition of the vitreous humour, and is involved in controlling the ionic and metabolic environment of the retina (Frey, 1984). In this study animals were exposed to 75microW/cm2 for 25 minutes and then injected with sodium fluorescein. A small but significant increase in vitreous fluorescence was observed in the exposed animals. Other studies by Neelakantaswamy and Ramakrishnan (1978, 1979) had reported that RF radiation could induce bending moments and stresses in the tissue capable of upsetting physiochemical processes in the eye. Kues and Monahan (1992) et al. found that exposures at 2.45GHz were capable of forming microscopic pits in the lens surface of anaesthetised monkeys. (in the corneal epithelium) at SAR of 2.6 W/kg but only after several hours exposure, with pulsed fields being more effective than continuous-wave radiation. In 1994 Kues, D’Anna et al., of Johns Hopkins University presented evidence of ocular effects on macaque monkeys (macaca mulatta) at well below thermal levels, concluding that "SAR may not be the most appropriate indicator of a biological effect in a pulsed exposure situation". Lu, Mathur et al. (2000) found no changes in retinal structure at 1.25GHz with SARs ranging from 4.3 to 20.2 W/kg and intense peaks. They did however, using a rhesus monkey model, find a slight enhancement in electrical responses. Two other monkey studies (Kamimura et al., 1994) using 2.45GHz continuous wave exposure and D’Andrea et al. (1992) using pulsed 5.6GHz radiation also reported largely negative effects. The use of timolol maleate, a medication used in glaucoma, appeared to reduce the threshold to 0.26 W/kg and increased vascular leakage in a study by Kues (1985). Stewart comments that the differences between the studies may be accounted for by the peak SARs per pulse used: the Kues, Monahan study involved peaks of 50,000-130,000 W/kg, whereas that of D’Andrea was only 4000W/kg. In 1999 a Royal Society of Canada review concluded that eye research was a priority in EMF research, particularly on walkie talkies, whose radiations can be higher than cellphones, especially since the antenna is held in front of the eyes. Subsequent ocular effects research has concentrated on user response in interviews of cancer patients. One of the commonest forms of eye cancer is uveal melanoma, and this subtype formed an interview study of 118 patients (475 controls) by a team from the University of Essen investigating exposures to all forms of RF radiation (Stang, Anastassiou et al., 2001). Those using cellphones or "walkie talkies" occupationally for more than several hours a day were found to have a three-fold incidence of uveal melanoma (based on 16 cases and 46 controls: OR 3.0, 95% CI 1.4-6.3) whilst those with probable/certain exposure to mobile phones has an OR of 4.2 (95% CI: 1.2-14.5). Other sources of EMF such as high voltage power lines and VDUs were not associated with uveal melanoma. UV exposure was not included in the study however. 12. Cancer related effects The possibility that chronic exposure to cellphone radiations may result in leukaemias, lymphomas, and solid tumours of the brain is an enduring concern both of the media and researchers in the field. These concerns reflect a long history of reported occupational effects on humans at RF/MW frequencies other than those emitted by cellphones, and a large list of controlled experiments on cells and laboratory animals. In 1994 Cain, Thomas et al., from Adey’s group at Loma Linda reported that 837Mhz radiation of C3H/10T1/2 fibroblasts and their mutant daughter cells enhanced TPA-induced focus formation by 40 percent, whereas exposure at lower levels had no effect or even decreased the co-promotion. TPA is a known carcinogen, so this suggested that cellphone frequency radiations at well below cellphone power densities might copromote existing carcinogens. An early but little publicised example of the animal studies is the work by Andriyenko and Serdyuk from Kiev at the 1993 EBEA Congress, in Bled, Slovenia, where they reported increased malignancies in rats after exposure of only 0.1-2.5 mW/cm2 at 300MHz, much lower than cellphone emission densities. At the same conference Kubinyi, Thuroczy et al reported adverse in utero effects on mice embryos exposed to 2.45GHz radiation at 3mW/cm2. A major attempt to evaluate the long term effect of weak microwave irradiation in rats was made in the mid 1980s but not published fully until 1992 (Chou, Guy et al., 1992). The rats were exposed at 2.45GHz and 0.5mW/cm2 for upto 25 months, with 155 measures of health and behaviour collected. This $5 million study at the University of Washington, funded by the US Air Force School of Aerospace Medicine, evoked much criticism with some scientists such as Robert Becker complaining (Cross Currents, pp194-196) that it was as if the study design was deliberately aiming not to find any effects. The use of gnotobiotic (a term meaning germ and virus free, where the animals are raised in sterile conditions) rats had the effect of diluting the results since these animals are less likely to get cancer. (20 percent of human cancers are caused by viral infection). Even so the data showed a near fourfold (18/5) increase in primary tumours in the exposed group. The malignanies moreover were not the usual types but of organs likely to reflect stress, i.e. pituitary, thryoid and adrenal gland cancers. The exposure was at twenty times lower than the thermal limit for exposure. The results were buried in a voluminous USAF report. Originally the authors had concluded that low levels of microwave radiation can cause cancer in mice (sic). However in a sanitized 1986 Scientific American article, having first suggested that regarding environmental microwaves "little of this invisible energy ever reaches the public" -no longer true with the advent of cellphones - they downplayed the results saying "For one thing the total number of malignant tumors in the control group was lower than the number expected for the particular strain of rat; the rate of malignancies in the exposed rats was about as expected… "The comparison of rates of malignancy was just one of 155 different comparisons made in the study. "although some hazard from weak microwave fields might be proved in the future, there is currently little evidence for the presence of such a hazard". The article In 1992 when the results were finally published in Bioelectromagnetics Journal, (not so widely read by the public as Scientific American) it was conceded that corticosterone levels and immunological parameters were affected at 13 months exposure, but not confirmed in a follow up study (Chou et al., 1986). The lymphoid system was also challenged by various mitogens (PHA, LPS, Con A) and abnormal responses were found: "These results indicate a selective effect of exposure on the lymphoreticular system’s response to mitogenic stimulation. Mitogen-response data were not available from the 25 month exposure studies because the lymphocyte cultures failed to grow. Since lymphocyte competence is a prerequisite for tumour immunity (Rosenberg and Terry, 1977) to report that there was a significant alteration in lymphocyte response in the exposed rats is worrisome. Regarding the excess tumors, the authors wrote "the finding of a near fourfold increase of primary malignancies in the exposed animals is provocative. These data cannot be considered an artifact because different statistical analyses led to similar results". Finally since the malignancies did not occur until the 18th month this suggested evidence of the long term effect primarily being investigated. A second major rat initiation-provocation study was reported in 1996 by Adey, Byus, and colleagues at UCLA Riverside and Davis, again with equivocal results. This time there appeared to be a lower number of brain malignancies (4/13) than expected, when the experiment was terminated at 24 months. Having been injected in utero with the carcinogen ENU, the 236 rats were exposed to TDMA 836.55MHz radiation at around 0.7W/kg, on and off for 7.5 minutes during a daily 2 hrs period. The small number of experimental numbers emphasise caution in interpreting these data, the authors concluded. More frequently cited is the study of transgenic mice by Repacholi, Basten et al. (1998) at Adelaide which reported a near twofold odds ratio in lymphoma following exposure to cellphone frequencies. Despite this laboratory evidence few clinical or epidemiological studies have been conducted at cellphone frequencies, and a proven mechanism of interaction is still lacking for effects at less than thermal intensities. It has already been indicated that cellphone handset radiations can exceed the limits regulated by reference to thermal insult, and many anecdotal reports of solid tumours on the same side as the phone use predicate further urgent research in this area. Such studies as have been conducted include those commissioned by Wireless Technology Research (WTR) whose CEO Dr George Carlo reported an unpublished 50 percent increase in auditory nerve cancers (acoustic neuroblastomas) in long time users compared with controls (Carlo, 2001). In 2001 two new epidemiological studies appeared to indicate the absence of evidence of a tumorifacient effect, but on closer examination the tumour types were not pertinent to cellphone use (Muscat, Malkin et al, 2000; Inskip, Tarone et al., 2001). The Muscat study of 469 persons (18-80yrs) with cancer at 5 US medical centres between 1994 and 1998 actually found an elevated 2.1 OR incidence of neuroepitheliomatous cancers (95%CI 0.6 - 4.7), and noted that cerebral tumours were more frequent on the side of the head used in cellphoning. But the overall negative conclusions may have been due to the similar level and duration of usage by exposed compared with control groups (2.5 hrs/month by exposed, 2.2 hrs.month by controls: 2.8 years duration of use by exposed, 2.7 years by controls). Some cancers are slow growing and the study was confounded by the wide age stratum, meaning that onset of many cases was before cellphone popularity. The Inskip study, investigating 792 patients in Phoenix, Boston, and Pittsburgh during the period between 1994 and 1998, also claimed a negative result but nevertheless reported an elevated incidence of acoustic neuromas (1.4 OR; 95%CI 0.6 - 3.5) though meningiomas and gliomas were not over-represented. The lack of RF impact on the meninges may be due to the absence of any focusing effect of the skull in that region. A further large Danish cohort study (Johansen, Boice et al, 2001) examined the history of 420,000 users between 1992 and 1995, and found no differences between "exposed" and control groups, but this may reflect uncertainties over exposure parameters. Most recently a Swedish questionnaire study of some 1600 cellphone users has been conducted by Hardell, Mild et al. reporting a 2.7 fold incidence of acoustic neurinomas on the ipsilateral side. Taken together the reasonable conclusion from these studies is that present regulatory limits of SAR levels, to which cellphone manufacturers say they conform, do not sufficiently protect users against tumours of the acoustic neuroma subtype. Another cancer hazard may be connected with uveal melanoma, discussed under ocular effects. Effects on adenosine triphosphate (ATP) synthesis The inner membrane of the mitochondria (organelles found ubiquitously in organic cells, especially in motile or energy dependent cells such as spermatozoa, glial cells of the brain, and in muscle cells) is the site where ATP is synthesised. This biomolecule provides about 90 percent of the enrgy used by organisms and its normal synthesis is by oxidative phosphorylation. In this process electrons are collected in the inner membrane and at the point where the potential difference across the membrane is sufficient (about 220mV) the electrons are released to permit the binding of a third and energy rich phosphate to ADP. Spare and unpaired electrons are collected by molecular oxygen to avoid free radical damage, and the entire process is carefully controlled by homeostatic processes. If however an electric field is applied it has the effect of depolarising the membrane and thereby preventing ATP synthesis. These processes are regulated by enzymes, including Na+ K + APTase, , and studies on the impact of microwave exposure on these are referred to above. As a result of effects on ATP synthesis many non-specific organic energy-reliant processes can be affected, including cardiac dysrhythmia, altered brain rhythms, suppression of immune competence, lowered libido and sexual potency, asthenias etc. , and in all these categories of life processes there have been reports of adverse effects from EMF exposure at a wide range of frequencies (e.g. Robinette, Silverman et al., 1980). CONCLUSIONS A relatively comprehensive review of the literature as sketched out above suggests that the present use of SAR as a guide to cellphone safety is misplaced. The metric is difficult to measure or calculate, and its finer details also reveal fundamental problems with time and spatial averaging. The use of SAR as an indicator of cellphone safety will deceive the public into a false sense of security. Even if an accurate value were incorporated in labelling, the differences simply in the way the cellphone is held relative to the head, or the proximity of conductive scatterers such as earrings or spectacles could double the level of exposure in energy terms. However there is increasingly robust evidence that the impact of RF/MW radiation, particularly if amplitude modulated, can have important biological effects at power densities too low to be simply thermal. There are so many categories of effect at non thermal levels that it suggests that not just one mechanism is being affected by the interactions. Organisms have their own scarcely recognised and little understood needs for signal transduction and intercellular information exchange, and until this century such processes were undisturbed by the presence of oscillating electric fields. It would be plausible for nature to choose electric fields to optimise the signal to noise ratio, since interference from static and quasistatic geomagnetic fields is prevalent on the earth’s surface. In summary the use of SAR as a health related metric is questionable, and it should be replaced with a metric which recognises fully the weak levels of radiation capable of causing serious damage to organic systems. In view of the present ubiquity and future continued growth of cellphone users, especially by young persons, research effort should be made to develop a metric recognising the potential risks from even a low chronic level of exposure. Appendix One: Conductivity considerations Conductivity is measured in mho per metre or Siemens per metre (s), the Siemens being the reciprocal of the ohm, the unit of resistance. Various models have used inorganic materials to simulate tissue conductivities, and phantom heads made with these have been measured with probes to detect field strengths as a result of exposure to cellphones placed in specified positions adjacent to the model head. None of these models take any account of the natural emissions and currents flowing within the brain in vivo, but physics tells us that any one electric field will perturb any other. This is dealt with in a separate section above. A more obvious limitation of such models is that the conductivities of the remainder of the human body will also affect the absorption rate in the head or other individual organs of the body. Moreover for cross-sections significantly smaller than the wavelength (e.g. small animals or individual organs of differing conductivity) as Kuster and Balzano point out, the SAR distribution is similar to induced eddy currents with high absorption at the peripheral regions and small values in the centre. Homogeneous models, that is models using one single material as a tissue substitute, must depend heavily on the particular conductivity of that material. Liquids can be made to mimic physiological saline of brain tissue, but the results are inevitably crude. Nonhomogeneous models attempt to overcome this disadvantage by incorporating a number of different materials with differing conductivities. There also are strong differences in convective capabilities between relevant tissue types: for example between cerebral grey matter and cerebral white matter, since there are substantially higher densities of blood vessel capillary beds in cortical grey matter (Ranck, 1964). Conductivity is also affected by the radiation frequency, and whether the subject is grounded. In the upright position the grounded adult body has a longitudinal resonance of around 35MHz, whereas in the transverse and anterioposterior axes maximum absorption occurs at frequencies from 135 to 165 MHz. Ungrounded the resonance is around 70MHz. Generally the higher the frequency the less able electromagnetic radiation is to penetrate materials. However even millimetre waves penetrate irradiated skin to a depth of 1mm, but the microcirculatory system of the skin begins to function at only 150 microns, i.e. is fully accessible to EHF exposure. Lower frequencies can penetrate further than this. The point is that conductivity issues must be addressed more carefully since circulatory pathways are available for incoming radiation at depths very close to the skin surface. In far field antenna models for the whole body there will be a pattern of secondary excitation for the remainder of the body, but no studies todate have addressed this issue. Conductivity (s) is the conductance of an electrode of unit dimensions, and reflects the ease with which delocalised electric charge can migrate through material under the influence of the field’s influence, whilst permittivity (e’) reflects the extent to which "localised" charge distributions can be distorted under the influence of the electric field. The permittivity of free space (eo) is 8.854 x 10-12 Farads/metre. The conductivity of biological membranes is of the order of 1 Siemens/metre. When one measures the permittivity and conductivity of a cell suspension as a function of increasing frequency one finds that permittivity falls and conductivity rises in a series of steps known as dispersions. With cellphone frequencies the gamma dispersions are of particular interest, and these arise through the motion of water dipoles. Indeed it can be shown that practically all the dielectric behaviour of biological material arises from the motion of water dipoles and dissolved ions. Davey and Kell (1991) point out that biological membranes may be regarded (with respect to the intra and extra cellular spaces) as essentially non-conductors. On each side of this "insulator" are conducting ionic solutions (cell cytoplasm and surrounding medium) and so a cell membrane is analogous to a classical electrical capacitor. This means that when an exciting voltage is applied across a cell suspension the membrane capacitance (cm) is charged up by ions moving through the conductivities of the cell medium (so) and the cytolasm (si). Thus the membrane charging is equivalent to a resistor (reflecting so and si) in series with a capacitor (reflecting cm and the volume fraction of cells) and like such networks it posses a time constant (relaxation time) t (equivalent to CR), which reflects the time taken to charge up the membrane. The physical interpretation of this is that as frequency rises fewer and fewer ions have time to charge up the membrane before the field changes direction. Thus the charge stored by the suspension for a given exciting voltage falls and the capacitance and permittivity of the suspension drops. At low frequencies the admittance (conductance to alternating current) of the cell membranes is very low and so the cells behave as non-conductors suspended in a conductive medium. This means that most of the current must flow round the cells. As the frequency increases the membrane’s admittance rises and an increasing amount of current can flow through the membrane and the highly conductive cytoplasm of the cells. Thus the conductivity increases with frequency. The is because the energy from the electric field must either be stored (as reflected by e’) or dissipated as reflected by s’) . Finally the negative surface charges on cell types differ and also those of aberrant cells, which have lost their surface glycoproteins, differ from normal so that one tissue type will be dissimilar in its conductivity and permittivity from another in health and disease. Typical conductivities for relevant tissues are brain tissue 0.65-0.88 S/m skin 0.6-0.82 bone 0.05-0.33 fat 0.06-0.17 physiological saline 0.001-0.01 Since the aqueous content of biological material is substantial it is apparent that the mode of interaction between non-ionising electromagnetic radiation and tissue is highly dependent on the dielectric behaviour of water and dissolved ions at RF and MW frequencies. Gabriel (Gabriel, Sheppard et al., 1983) and Grant (South and Grant, 1972) have long investigated the dielectric behaviour of biological material over the frequency range from DC to a few tens of GHz and how it can be explained in terms of various dispersion regions. The alpha, beta and mu dispersions are due to the electrical properties of cells and macromolecules, while the gamma dispersion may be attributed to the motion of water dipoles. Other than effects due to atomic and electronic polarisation, practically all the observed dielectric behaviour of biological material at frequencies in excess of 1GHz arises from the motion of water dipoles and dissolved ions. The contribution to the overall conductivity from the ions diminishes as the frequency increases with the consequence that these effects have diminished to negligible proportions by 10GHz. Thus the permittivity (e’) and conductivity of tissues varies considerably, not only with pathology but also with frequency. In the case of skin for example there can be a 10-point difference between normal and wounded skin, and its permittivity can vary from around 50 at 1 GHz to 110 at 10 MHz (Gabriel, Bentall et al., 1987). The attempts to mimic these complex characteristics of human tissue are far from perfect, and a usually cell free fluids. See Meier et al., (1996b) and Hartsgrove et al., (1987) for a summary of relevant tissue types for the human head, and their simulating liquids and their recipes. Appendix Two: Permittivity considerations Most biological materials are dielectric and are affected by polarisation (i.e. the induction of dipole moments) and orientation of permanent dipoles. If an ensemble of cells rather than single cells are considered , i.e. mesoscopic entities, it is useful to describe this mesoscopic response in terms of the dielectric properties of the material (see Matthes, 1996 for details). (Cell ensembles are also weak conductors, and this is treated separately). The permittivity of a dielectric material is usually written e and the permittivity of free space is written as eo (numerically 8.854. 10-12 F/m), where eoc2 = 107/4p, and Ñ · E = r/eo. The refractive index of a dielectric is normally given by Ö e/eo, and a simple model of a dielectric will illustrate the frequency dependence of the permittivity (see Vanderlinde, Classical Electromagnetic Theory, p224 for an example of such a model). This means that the conductivity of organic tissue is frequency dependent, but the permittivity is not directly included in the SAR formula. Bernhardt and Vogel, (1996) point out that the absorption of high frequency radiation and subsequent distribution of energy in the body are not only strongly dependent on the size and orientation of the body but also on the frequency and polarisation of the incident radiation. Both theory and experiment show that high frequency absorption in the body approaches a maximum value when the long axis of the body is both parallel to the electric field vector and approximately equal to four tenths of the wavelength of the incident field. Since at 900 MHz the wavelength is 33 cm, this suggests that the maximum is absorbed in a body of length 0.4 x 33 = 13.2 cm,. This is approximately the same as the diameter of a human infant skull. In the frequency range 200-3000MHz refraction may furthermore result in focusing effects producing spatially limited "hot spots" in the body. For example a considerable temperature rise may occur in certain areas of the head (WHO, 1993, p 116). The general form of SAR does not deal adequately with near field exposures and the hot spots they give rise to. As M. Taki of Tokyo Metropolitan University wrote, reviewing standards in the proceedings of an 1996 ICNIRP conference (3rd Intl NIR Workshop, Baden, Austria, 1996): "Plane wave exposures or far field exposures were intensively examined in the 1970s, and the whole body resonance, where the whole body averaged SAR approaches a maximal value around the frequency with a wavelength of four tenths of the body height was recognised to occur… The derived limits in terms of electric and magnetic field strength are thus obtained from the far field exposure conditions. On the other hand, risk assessment for near field exposures, especially for the partial body exposures by nearby radiation sources has not been well-established. However most of the actual situations where human bodies are strongly irradiated by RF fields are in the near field partial body exposures. Risk assessment of these exposures is a matter of great concern". Elsewhere in the same review Taki concedes that the present limits are sometimes closer than comfort to actual cellphone SAR values and "The peak SAR value during the ordinary use of the commercially available portable radios [i.e. cellphones] with a power output of 0.6 Watts does not exceed the basic restriction of local peak SAR, but the margin is not sufficiently large". Permittivity falls with increasing frequency. Even so, for muscle tissue, the penetration depth (distance over which about two thirds of the radiation is absorbed) at 30MHz is 10 cm, and at 1 GHz it is still as much as 3 cm (Schwan, 1988). At 3GHz it drops further to only 1cm, but this is still sufficiently penetrating to reach the cortex and glial cells of the brain and the retina of the eye, and the saline conducting extracellular fluids of the body.. For example, at 1 GHz the relative permittivity of retina falls from over 60 to below 10 during this transition (Gabriel, 1989). Experiments in our laboratory confirm that signals from a cellphone at 2.2 cm from a fluid surface leading to a filled conduit can be almost losslessly received via the conducting saline fluid at distances of 1.5 metres. This implies that radiations near the head will be easily detectable at the other exteemity of the body, and at almost the same power density or field strength. This finding make a nonsense of models whose measurements are confined only to the head, since not only does the vascularisation of skin begin at 150microns from the surface, but the vascularisation of the skull is also substantial. Appendix Three: : Tissue density considerations These pose less of a problem for dosimetry, since most human tissues have a high (~ 75%) water content, and hence a high permittivity (er >30) and altering its conductivity (e.g. by adding sucrose or salt) makes little difference to the density of the material. There is nevertheless a significant range of permittivity across different tissues, though this is not incorporated into the SAR formula directly. Moreover, organic tissues are largely transparent to magnetic fields. Because of the very dominant inductive coupling where sources are close to the body, a more refined formula for SAR at the surface of a flat phantom incorporates permittivity considerations: SARs = s/r mwÖ s2 + e2w2 (1+ccorrg pw)2H2tinc where Htinc is the incident magnetic field and the tissue material is represented only by s and e. Appendix Four: Temperature related measurement The calculation of SAR is simply described by formula (1) above. However the imprecision of this formula is exampled when different components of the formula are considered. There are at least four numerical ways of arriving at SAR: the finite difference time domain (FDTD)method the sine-base fast Fourier transform (FFT conjugate gradient (CG) method the impedance method (at lower frequencies) the method of moments (MOM) as well as methods based on temperature change (SAR = c.DT/Dt), which are not widely used for SAR measurement in human phantoms in view of their extra complexity, due to the homeostatic nature of organisms. The temperature change may be characterised by a whole body averaged measurement, a point measurement, or a thermographic camera analysis of bisected phantoms, but in the near field at cellphone frequencies the localised heating effects are dependent on so many different factors (e.g. refraction or reflection of the radiation in a variety of tissue types, and the proportion of the total energy which is electric rather than magnetic) that accuracy is suspect. Appendix Five: Conversion of power density units Most of the technical arguments about cellphones and masts are likely to discuss the power density of radiation emitted. From the power density is estimated the amount of dose received by anyone near it, and this latter statistic is called the specific energy absorption rate or SAR. Some scientists do not believe that SAR is the right metric to evaluate health risk, and that vital organic processes are disturbed by much lower power densities depending on their frequency and other factors, but it would be enough to show convincingly that the sort of power densities transmitted are of the same level as those measured in laboratory and epidemiological studies reporting associations with ill health. The way scientists describe power density is by the amount of power in Watts flowing through a given area of space. Because space can be described in different ways e.g. in square metres (large area) or square centimetres (small area) the resulting statistics are sometimes confusing, if for example one study is using milliWatts per square metre while another is using units of microWatts per square centimetre. So one needs to be able to convert these for comparison purposes. The conversion table below is adapted from the World Health Organisation’s EHC 137 (1993; p35) a useful handbook about electromagnetic fields:
Notice that the last two columns are different. These are the conversions from the power density into the electric (Volts per metre) and the magnetic (Amps per metre) field strengths. In practise when measuring the radiation, meters measure the electric or the magnetic field and then it is converted into the power density. There is obviously a fixed relation between the electric and the magnetic field at cellphone frequencies (unlike those of power lines), so if one knows the electric field then one can always calculate the equivalent power density. With the aid of this table one can follow the technical arguments about what is a safe exposure level. Suffice it to say for now the NRPB believes 194 Volts per metre is the guideline limit for 1800 MHz transmitters (they call it the investigation level) whereas the ICNIRP guideline is far lower than this at only 58 Volts per metre, so the NRPB are already out of line with the rest of the world! One can point to this disagreement to show how confused the regulatory authorities are. At around 1800 MHz, the frequencies of GSM transmitters, these official guidelines translate into power densities of 100W/m2 for the NRPB (which as one can now see is equivalent to 10,000 mW/cm2) and 9 W/m2 (~ 1000 mW/cm2) for ICNIRP, over ten times lower! When these bodies discuss cellphone or mast emissions they often describe them as a fraction of the guidelines. But there are many studies showing bio-effects well below such levels. Appendix Six: Recent Russian research on RF/MW bioeffects In general the eastern bloc countries have carried out much more high frequency bio-effects research than the West. In the 1950s USSR research attention focused on mastery of the millimetre (EHF) part of the electromagnetic spectrum (see O.V. Betskyi, 1997 for review), conducted in several institutions (the USSR Academy of Science, and the Ukrainian SSR) under the general direction of academician N.D. Devyatkov. Under Devyatkov during the 1960s all necessary instrumentation was constructed for mm and sub-mm waveband research, and by 1965 he was able to conclude that all living creatures are unadapted to these frequencies, because it is practically absent on the planet due to strong absorption by the earth’s atmosphere (mainly water vapour). Among the many discoveries of that time were that conduction of signals through the nervous system is accomplished in the myelin sheath of the axons, that conduction through the humoral system is connected with the movement of generating cells through the blood and lymphatic systems, so that the transmission of signals is accomplished not by conduction of radiation nor by movement of charges, but by the displacement of generator cells, the oscillations of which reflect the information being carried. Low intensity EHF radiation, it was found, could accelerate the transport of sodium ions (at a power flux density of 1mW/cm2), a change in the permeability of erythrocyte membranes to potassium ions, (power flux density of 1-5mW/cm2 ) and an increase in the ion conductivity of bilayered lipid membranes (power flux density of 1-10mW/cm2). These power densities are well below those emitted by cellphones (around 20mW/cm2). The dependence of biological effects on frequency was also uncovered at that time through research by Smolyanskaya, Sevastianova et al., during the period 1968-1971. This led, via research by Petrov and Betskyi that the change in potential difference of the plasma membrane was identical to that which arises during exposure to photosynthetic active radiation, to the conclusion that mm wavelengths have the ability to stimulate the synthesis of ATP in the cell. It thus came about that the Russian approach to irradiation by EMF early abandoned the concept that effects were solely thermal in nature and should be measured in terms of specific energy absorption rate. As their research proceeded they also became aware that cellular responses occurred for specific frequencies A further complication is the effect of relaxation time (t). For example the t value of a beta dispersion is proportional to the cell radius (r), whereas that of the alpha dispersion is proportional to r2. If a given dispersive mechanism exhibits such a distribution in t (and hence fc (where fc = 1/2pt), this being the characteristic halfway point in frequency terms between the beginning and the end of the dispersion) then the fall in permittivity and rise in conductivity will be less steep. Two recent studies evaluated the problem of differing cell types, one for 900MHz (Hombach et al., 1996) and one for 1.8 GHz (Meier et al., 1997). The authors reported that most of the radiated power was absorbed close to the source, indicating that there was little obvious difference between homogeneous and nonhomogeneous absorption rates. However, local SAR distribution depended significantly on the local distribution of tissue properties, and varied greatly with individual humans and could even change with time. The electrical parameters of a human being clearly vary with levels of physical and metabolic activity, health, and age, and this means that a particular SAR level may have a greater or lesser effect dependent on the individual. Spatial peak SAR values also depend significantly on the local anatomy: overestimations can occur with homogeneous models as much as 25 percent for the 1 gram averaged values (though <10 percent in the case of 10 gram values). Small shifts (<10mm) of the source parallel to the body surface can also result in important variations of the spatial peak SAR in nonhomogeneous models. Moreover the hand can also hold the cellphone in an infinite variety of positions, and enhancement effects from external metallic accessories such as ear-rings, spectacles or other local jewellery are likely with nonhomogeneous models. Taking all these conductivity-related issues together it is apparent that present models both theoretical and experimental are only able to provide a crude idea of actual specific absorption rates, and that the margin for error is substantial, with several important parameters ignored. Not surprisingly different researchers have reported widely differing results from the same exposure criteria: for example in near field situations, where the energy deposition is largely induced by the magnetic component, Gandhi et al., 1994 results differed from those of Kuster et al., 1992. There is however a general consensus that in such near field models having strong magnetic coupling to tissue, maximum field levels occur in skin, subcutaneous tissues, bone, and immediately subjacent brain tissue (see Bach Anderson et al. (1995) for review). Appendix Eight: Review of negative studies The case for questioning the validity of SAR partly depends on a large number of studies showing biological effects at levels of SAR well below the present guidelines. These investigations are not wholly unchallenged however, and it is instructive to examine in greater detail those studies reporting lack of effects at non-thermal levels, often as attempts to replicate work reporting positive effects. There are at least fifteen studies failing to find biological effects at non thermal levels. These are summarised below. Appleton & McCrossan (1982) Data analysis reveals significant microwave-induced eye damage in humans. Allan Frey (1985) re-analysed the study reported in 1982 by Appleton and McCrossan, who undertook a study for the U.S. Army at Ft. Monmouth to determine if microwave exposure would cause cataracts. They concluded: "The comparison showed the groups (microwave exposed vs. not exposed ) to be essentially the same and did not support the hypothesis that human cataracts are being caused by chronic exposure to microwaves in the military environment in this country." There are three major flaws in Appleton and McCrossan's work. First, the exposed group likely included people with little or no exposure. This would tend to minimize the possibility of finding microwave effects. Secondly, their control group consisted of people working with equipment known to cause eye damage. This also would tend to minimize the possibility of finding microwave effects. Thirdly, and most important, they did not do a statistical analysis on their data. When the writer did one, it was found that Appleton and McCrossan have a statistically significant difference between groups, with the microwave-exposed showing more lens opacities than would be expected by chance. Thus, their conclusion should have been the opposite of what they stated. It is the uncritical acceptance of negative biological studies of non-ionizing radiation, such as this, that has contributed to the distortion of science in this area of research and has stimulated public opposition to the installation of such energy sources. This summary was taken from PMID: 3847507 [PubMed - indexed for MEDLINE] Appleton, Hirsch et al., (1975) Microwave lens effects in humans. II. Results of five-year survey. Appleton B, Hirsch S, Kinion RO, Soles M, McCrossan GC, Neidlinger RM. Arch Ophthalmol 1975 Apr;93(4):257-8 Individuals selected on the basis of likelihood of occupational exposure to microwaves were subjected to a biomicroscopic examination of the lens. Control personnel were also examined along with them, the examiners having no knowledge of the exposure history of any examinee prior to or during the examination. Objective evidence of lens abnormality (opacities, vacuoles, or posterior subcapsular iridescence) was recorded and a comparison made between the two groups on the basis of that evidence. The comparison showed the two groups to be essentially the same and did not support the hypothesis that human lens damage is occurring in the military environment in this country. Instead, the authors (who were funded by the US military) argued it tended to support the assumption that the existing safety level of 10 mW/cm2 is adequate. This study is unlikely to yield usable results relevant to cellphones, however, since there was no adequate means of assessing the exposure differences of the two groups, or the extent of overlapping exposure or frequencies to which they were exposed. The abstract was taken largely from the PMID: 1119967 [PubMed - indexed for MEDLINE] Chagnaud, Moreau et al., (1999) No effect of short-term exposure to GSM-modulated low-power microwaves on benzo(a)pyrene-induced tumours in rat. Chagnaud JL, Moreau JM, Veyret B. Int J Radiat Biol 1999 Oct;75(10):1251-6 PIOM Laboratory, ENSCPB, University of Bordeaux I, Talence, France. PURPOSE: In view of current interest in the biological effects of amplitude-modulated microwaves arising from the rapid development of mobile communications, the effects of low-level microwaves on cancer development were investigated using a rat sarcoma model. MATERIALS AND METHODS: Two-month-old female Sprague-Dawley rats were treated by injection of benzo(a)pyrene and irradiated with GSM (Global System for Mobile)-modulated 900-MHz microwaves in an anechoic chamber at 55 or 200 mW/cm2 (75 and 270 mW/kg average whole-body SAR, 2h daily for 2 weeks). Rats were exposed from day 20, 40 or 75 after carcinogen injection. Additional groups of rats were sham-exposed in a second anechoic chamber. Anti-phosphatidylinositol autoantibody levels were evaluated in sera to monitor malignant transformation. RESULTS: Microwave exposure had no effect on the development of tumours. No acceleration or delays in tumour onset were observed. Animal survival was not modified and serum autoantibody levels were similar in exposed and sham-exposed groups. CONCLUSION: Low-level GSM microwave exposure of rat bearing benzo(a)pyrene-induced tumours had no effect on auto-antibody levels, tumour appearance and survival. The low exposure levels used here correspond to exposure limits for whole-body exposure of humans. This abstract was from PMID: 10549601 [PubMed - indexed for MEDLINE] It is always difficult to argue from animal studies to humans, or to assume that the dose for one will be the same in effect as for the other. One has to ask whether a 2hrs/2 wks exposure of an animal likely to live 2 yrs normally is an adequate exposure. Other lifetime studies of rats Adey , Byus et al.) and mice (Repacholi, Basten et al.) show elevated tumour incidence in the exposed groups. Higashikubo, Culbreth et al., (1999) Radiofrequency electromagnetic fields have no effect on the in vivo proliferation of the 9L brain tumor. Radiat Res 1999 Dec;152(6):665-71 Higashikubo R, Culbreth VO, Spitz DR, LaRegina MC, Pickard WF, Straube WL, Moros EG, Roti JL. Radiat Res 1999 Dec;152(6):665-71 Radiation Oncology Center, Mallinckrodt Institute of Radiology, Washington University, St. Louis, Missouri 63108, USA. The intracranial 9L tumor model was used to determine if exposure to a radiofrequency (RF) electromagnetic field similar to those used in cellular telephone has any effects on the growth of a central nervous system tumor. Fischer 344 rats implanted with different numbers of 9L gliosarcoma cells were exposed to 835.62 MHz frequency-modulated continuous wave (FMCW) or 847.74 MHz code division multiple access (CDMA) RF field with nominal slot-average specific absorption rates in the brain of 0.75 +/- 0.25 W/kg. The animals were exposed to the RF field for 4 h a day, 5 days a week starting 4 weeks prior to and up to 150 days after the implantation of tumor cells. Among sham-exposed animals injected with 2 to 10 viable cells (group 1), the median survival was 70 days, with 27% of the animals surviving at 150 days. The median survival length and final survival fraction for animals injected with 11 to 36 viable cells (group 2) were 52 days and 14%, respectively, while the values for those injected with 37 to 100 cells (group 3) were 45 days and 0%. The animals exposed to CDMA or FMCW had similar survival parameters, and the statistical comparison of the survival curves for each of the groups 1, 2 and 3 showed no significant differences compared to sham-exposed controls. This abstract was from PMID: 10581537 [PubMed - indexed for MEDLINE] The study, funded by Motorola, does not appear sensitive enough to see a difference in tumour growth rates: effects are more frequently observed with modulated, not FMCW exposures.. Inskip, Tarone et al., (2001) Cellular-telephone use and brain tumors. Inskip PD, Tarone RE, Hatch EE, Wilcosky TC, Shapiro WR, Selker RG, Fine HA, Black PM, Loeffler JS, Linet MS. N Engl J Med 2001 Jan 11;344(2):79-86 Epidemiology and Biostatistics Program, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, MD, USA. inskippe@mail.nih.gov BACKGROUND: Concern has arisen that the use of hand-held cellular telephones might cause brain tumors. If such a risk does exist, the matter would be of considerable public health importance, given the rapid increase worldwide in the use of these devices. METHODS: We examined the use of cellular telephones in a case-control study of intracranial tumors of the nervous system conducted between 1994 and 1998. We enrolled 782 patients through hospitals in Phoenix, Arizona; Boston; and Pittsburgh; 489 had histologically confirmed glioma, 197 had meningioma, and 96 had acoustic neuroma. The 799 controls were patients admitted to the same hospitals as the patients with brain tumors for a variety of nonmalignant conditions. RESULTS: As compared with never, or very rarely, having used a cellular telephone, the relative risks associated with a cumulative use of a cellular telephone for more than 100 hours were 0.9 for glioma (95 percent confidence interval, 0.5 to 1.6), 0.7 for meningioma (95 percent confidence interval, 0.3 to 1.7), 1.4 for acoustic neuroma (95 percent confidence interval, 0.6 to 3.5), and 1.0 for all types of tumors combined (95 percent confidence interval, 0.6 to 1.5). There was no evidence that the risks were higher among persons who used cellular telephones for 60 or more minutes per day or regularly for five or more years. Tumors did not occur disproportionately often on the side of head on which the telephone was typically used. CONCLUSIONS: These data do not support the hypothesis that the recent use of hand-held cellular telephones causes brain tumors, but they are not sufficient to evaluate the risks among long-term, heavy users and for potentially long induction periods. (Comment in N Engl J Med. 2001 Jan 11;344(2):133-4) The data presented by these authors do not support their own conclusions. There was an evidently elevated odds ratio (1.4) for acoustic neuroma, and this tumour subtype is also seen in other user studies (Carlo, Hardell). The focusing effect of the skull has been identified as likely to produce hot spots in central areas rather than on the peripheries, so meningiomas would not be expected to be elevated. The statistical strength of this study is weak, however, since all the Confidence intervals pass through 1.0 and may therefoire be due to chance. Thids abstract was largely from PMID: 11150357 [PubMed - indexed for MEDLINE] Johansen, Boice et al., (2001) Response: re: cellular telephones and cancer-a nationwide cohort study in Denmark. J Natl Cancer Inst 2001 Jun 20;93(12):952-3 Johansen C, Boice JD Jr, McLaughlin JK, Olsen JH. C. Johansen, J. H. Olsen, Institute of Cancer Epidemiology, Danish Cancer Society, Copenhagen, Denmark. PMID: 11416121 [PubMed - in process] Comment in J Natl Cancer Inst. 2001 Feb 7;93(3):166-7 Response: J Natl Cancer Inst 2001 Feb 7;93(3):203-7 Cellular telephones and cancer--a nationwide cohort study in Denmark. Johansen C, Boice J Jr, McLaughlin J, Olsen J. Institute of Cancer Epidemiology, Danish Cancer Society, Copenhagen, Denmark. christof@cancer.dk BACKGROUND: Use of cellular telephones is increasing exponentially and has become part of everyday life. Concerns about possible carcinogenic effects of radiofrequency signals have been raised, although they are based on limited scientific evidence. METHODS: A retrospective cohort study of cancer incidence was conducted in Denmark of all users of cellular telephones during the period from 1982 through 1995. Subscriber lists from the two Danish operating companies identified 420,095 cellular telephone users. Cancer incidence was determined by linkage with the Danish Cancer Registry. All statistical tests are two-sided. RESULTS: Overall, 3391 cancers were observed with 3825 expected, yielding a significantly decreased standardized incidence ratio (SIR) of 0.89 (95% confidence interval [CI] = 0.86 to 0.92). A substantial proportion of this decreased risk was attributed to deficits of lung cancer and other smoking-related cancers. No excesses were observed for cancers of the brain or nervous system (SIR = 0.95; 95% CI = 0.81 to 1.12) or of the salivary gland (SIR = 0.72; 95% CI = 0.29 to 1.49) or for leukemia (SIR = 0.97; 95% CI = 0.78-1.21), cancers of a priori interest. Risk for these cancers also did not vary by duration of cellular telephone use, time since first subscription, age at first subscription, or type of cellular telephone (analogue or digital). Analysis of brain and nervous system tumors showed no statistically significant SIRs for any subtype or anatomic location. CONCLUSIONS: The results of this investigation, the first nationwide cancer incidence study of cellular phone users, do not support the hypothesis of an association between use of these telephones and tumors of the brain or salivary gland, leukemia, or other cancers. The period of this study (1982 to 1995) means that it mainly covers a period when few cellphones were in use. The abstract was largely derived from PMID: 11158188 [PubMed - indexed for MEDLINE] Klug Hetscher et al., (1997) The lack of effects of nonthermal RF electromagnetic fields on the development of rat embryos grown in culture. Life Sci 1997;61(18):1789-802 Klug S, Hetscher M, Giles S, Kohlsmann S, Kramer K. Benjamin Franklin Medical Center, Free University Berlin, Institute of Clinical Pharmacology and Toxicology. Rat embryos (9.5 days old) were exposed for up to 36 h to various radio frequency (RF) electric and magnetic fields (modulation frequency: 16, 60, 120 Hz; electric field strength: 60, 600 V/m; magnetic induction: 0.2, 2.0 microT). A resonator technique was used to generate standing waves thus fulfilling three conditions: The site of maximum electric and magnetic oscillations could be separated, the field strengths were known exactly and a high homogeneity over the sample volume was achieved. In each frequency region the transmitter power levels were set to give specific absorption rate (SAR) values spreading from far below to far above the values met in the field of telecommunication (0.2, 1.0 and 5.0 W/kg). The criteria used to examine the embryos on day 11.5 for possible structural effects consisted of a scoring system, photographs, histology using both light and electron microscopy and determination of the protein content. All these data have been taken as sets of different intermediate frequency (IF) amplitude modulation of the RF carriers. Neither the electric nor the magnetic fields tested interfered significantly with the normal growth and differentiation of the embryos in vitro. This abstract was from PMID: 9365226 [PubMed - indexed for MEDLINE] Exposure for periods upto 36 hrs cannot give a picture of the effects of chronic exposure. Lary, Conover et al., (1983) Absence of embryotoxic effects from low-level (nonthermal) exposure of rats to 100 MHz radiofrequency radiation. Scand J Work Environ Health 1983 Apr;9(2 Spec No):120-7 Lary JM, Conover DL, Johnson PH. Pregnant Sprague-Dawley rats were exposed to radio-frequency radiation at a frequency of 100 MHz and a power density of 25 mW/cm2 for 6 h 40 min daily on gestation days 6--11. The total exposure time was 40 h. The exposure resulted in a specific absorption rate of 0.4 W/kg. This value corresponds to the maximum permissible level for specific absorption rate in the 1982 American National Standards Institute (ANSI) standard for radiofrequency/microwave exposure. The exposure produced no increase in maternal colonic temperature. Irradiated rats did not differ from sham-irradiated rats with respect to the number of implantations per litter, percentage of implantations dead or resorbed, percentage of fetuses malformed. fetal weight, fetal crown-rump length, or fetal sex ratio. The irradiated fetuses had fewer minor skeletal variations than the controls. These results suggest that radiofrequency/microwave radiation is not teratogenic or embryotoxic for rats at the maximum permissible exposure level of the 1982 ANSI standard. This abstract was from PMID: 6648409 [PubMed - indexed for MEDLINE] The exposure frequency is well b elow that used in mobile telephony. Malyapa, Ahern et al., (1997a,1997b,1998) Measurement of DNA damage after exposure to 2450 MHz electromagnetic radiation. Malyapa RS, Ahern EW, Straube WL, Moros EG, Pickard WF, Roti Roti JL. Radiation Oncology Center, Mallinckrodt Institute of Radiology, Washington University, St. Louis, Missouri 63108, USA . Recent reports suggest that exposure to 2450 MHz electromagnetic radiation causes DNA single-strand breaks (SSBs) and double-strand breaks (DSBs) in cells of rat brain irradiated in vivo (Lai and Singh, Bioelectromagnetics 16, 207-210, 1995; Int. J. Radiat. Biol. 69, 513-521, 1996). Therefore, we endeavored to determine if exposure of cultured mammalian cells in vitro to 2450 MHz radiation causes DNA damage. The alkaline comet assay (single-cell gel electrophoresis), which is reportedly the most sensitive method to assay DNA damage in individual cells, was used to measure DNA damage after in vitro 2450 MHz irradiation. Exponentially growing U87MG and C3H 10T1/2 cells were exposed to 2450 MHz continuous-wave (CW) radiation in specially designed radial transmission lines (RTLs) that provided relatively uniform microwave exposure. Specific absorption rates (SARs) were calculated to be 0.7 and 1.9 W/kg. Temperatures in the RTLs were measured in real time and were maintained at 37 +/- 0.3 degrees C. Every experiment included sham exposure(s) in an RTL. Cells were irradiated for 2 h, 2 h followed by a 4-h incubation at 37 degrees C in an incubator, 4 h and 24 h. After these treatments samples were subjected to the alkaline comet assay as described by Olive et al. (Exp. Cell Res. 198, 259-267, 1992). Images of comets were digitized and analyzed using a PC-based image analysis system, and the "normalized comet moment" and "comet length" were determined. No significant differences were observed between the test group and the controls after exposure to 2450 MHz CW irradiation. Thus 2450 MHz irradiation does not appear to cause DNA damage in cultured mammalian cells under these exposure conditions as measured by this assay. This abstract came from PMID: 9399707 [PubMed - indexed for MEDLINE] The serious differences between this study and that of Lai and Singh are given in the main text. Measurement of DNA damage after exposure to electromagnetic radiation in the cellular phone communication frequency band (835.62 and 847.74 MHz). Radiat Res 1997 Dec;148(6):618-27 Malyapa RS, Ahern EW, Straube WL, Moros EG, Pickard WF, Roti Roti JL. Radiation Oncology Center, Mallinckrodt Institute of Radiology, Washington University, St. Louis, Missouri 63108, USA Mouse C3H 10T1/2 fibroblasts and human glioblastoma U87MG cells were exposed to cellular phone communication frequency radiations to investigate whether such exposure produces DNA damage in in vitro cultures. Two types of frequency modulations were studied: frequency-modulated continuous-wave (FMCW), with a carrier frequency of 835.62 MHz, and code-division multiple-access (CDMA) centered on 847.74 MHz. Exponentially growing (U87MG and C3H 10T1/2 cells) and plateau-phase (C3H 10T1/2 cells) cultures were exposed to either FMCW or CDMA radiation for varying periods up to 24 h in specially designed radial transmission lines (RTLs) that provided relatively uniform exposure with a specific absorption rate (SAR) of 0.6 W/kg. Temperatures in the RTLs were monitored continuously and maintained at 37 +/- 0.3 degrees C. Sham exposure of cultures in an RTL (negative control) and 137Cs gamma-irradiated samples (positive control) were included with every experiment. The alkaline comet assay as described by Olive et al. (Exp. Cell Res. 198, 259-269, 1992) was used to measure DNA damage. No significant differences were observed between the test group exposed to FMCW or CDMA radiation and the sham-treated negative controls. Our results indicate that exposure of cultured mammalian cells to cellular phone communication frequencies under these conditions at an SAR of 0.6 W/kg does not cause DNA damage as measured by the alkaline comet assay. Abstract from PMID: 9399708 [PubMed - indexed for MEDLINE] DNA damage in rat brain cells after in vivo exposure to 2450 MHz electromagnetic radiation and various methods of euthanasia. Radiat Res 1998 Jun;149(6):637-45 Malyapa RS, Ahern EW, Bi C, Straube WL, LaRegina M, Pickard WF, Roti Roti JL. Radiation Oncology Center, Mallinckrodt Institute of Radiology, Washington University, St. Louis, Missouri 63108, USA . This study was done to confirm the reported observation that low-intensity acute exposure to 2450 MHz radiation causes DNA single-strand breaks (Lai and Singh, Bioelectromagnetics 16, 207-210, 1995). Male Sprague-Dawley rats weighing approximately 250 g were irradiated with 2450 MHz continuous-wave (CW) microwaves for 2 h at a specific absorption rate of 1.2 W/kg in a cylindrical waveguide system (Guy et al., Radio Sci. 14, 63-74, 1979). There was no associated rise in the core body temperature of the rats. After the irradiation or sham treatments, rats were euthanized by either CO2 asphyxia or decapitation by guillotine (eight pairs of animals per euthanasia group). After euthanasia the brains were removed and immediately immersed in cold Ames medium and the cells of the cerebral cortex and the hippocampus were dissociated separately and subjected to the alkaline comet assay. Irrespective of whether the rats were euthanized by CO2 asphyxia or decapitated by guillotine, no significant differences were observed between either the comet length or the normalized comet moment of cells from either the cerebral cortex or the hippocampus of sham-treated rats and those from the irradiated rats. However, the data for the rats asphyxiated with CO2 showed more intrinsic DNA damage and more experiment-to-experiment variation than did the data for rats euthanized by guillotine. Therefore, the guillotine method of euthanasia is the most appropriate in studies relating to DNA damage. Furthermore, we did not confirm the observation that DNA damage is produced in cells of the rat cerebral cortex or the hippocampus after a 2-h exposure to 2450 MHz CW microwaves or at 4 h after the exposure. This abstract came from PMID: 9611103 [PubMed - indexed for MEDLINE] . The differences in study design and methods compared with Lai and Singh are discussed in the main text. Mittler (1977) Failure of chronic exposure to nonthermal FM radio waves to mutate Drosophila. J Hered 1977 Jul-Aug;68(4):257-8 Mittler S. A stock of Drosophila sc VI - YS/y ac oc ptg - YL/y ac oc ptg - YL/y sc S1 B In49 ct ns v sc8 that accumulated the recessive lethals on the X chromosome was exposed to a frequency of 98.5 MHz (wave length 3.35 m) and a field strength of 0.3 V/m. The flies were kept near the base of the 300-ft antenna of a 50,000 watt transmitter for 32 weeks. There was no significant difference in the percentage of lethals between the stock exposed to 4,020 hours of nonthermal FM radio waves and the controls. Abstracted from PMID: 925338 [PubMed - indexed for MEDLINE. The main beam of the transmitter was unlikely to provide much exposure so near the base of the mast. Moreover the frequency was very dissimilar from mobile telephony. Muscat, Malkin et al., (2000) Handheld cellular telephone use and risk of brain cancer. JAMA 2000 Dec 20;284(23):3001-7 Muscat JE, Malkin MG, Thompson S, Shore RE, Stellman SD, McRee D, Neugut AI, Wynder EL. American Health Foundation, 1 Dana Rd, Valhalla, NY 10595, USA. jmuscat2@earthlink.net CONTEXT: A relative paucity of data exist on the possible health effects of using cellular telephones. OBJECTIVE: To test the hypothesis that using handheld cellular telephones is related to the risk of primary brain cancer. DESIGN AND SETTING: Case-control study conducted in 5 US academic medical centers between 1994 and 1998 using a structured questionnaire. PATIENTS: A total of 469 men and women aged 18 to 80 years with primary brain cancer and 422 matched controls without brain cancer. MAIN OUTCOME MEASURE: Risk of brain cancer compared by use of handheld cellular telephones, in hours per month and years of use. RESULTS: The median monthly hours of use were 2.5 for cases and 2.2 for controls. Compared with patients who never used handheld cellular telephones, the multivariate odds ratio (OR) associated with regular past or current use was 0.85 (95% confidence interval [CI], 0.6-1.2). The OR for infrequent users (<0. 72 h/mo) was 1.0 (95% CI, 0.5-2.0) and for frequent users (>10.1 h/mo) was 0.7 (95% CI, 0.3-1.4). The mean duration of use was 2.8 years for cases and 2.7 years for controls; no association with brain cancer was observed according to duration of use (P =.54). In cases, cerebral tumors occurred more frequently on the same side of the head where cellular telephones had been used (26 vs 15 cases; P =.06), but in the cases with temporal lobe cancer a greater proportion of tumors occurred in the contralateral than ipsilateral side (9 vs 5 cases; P =.33). The OR was less than 1.0 for all histologic categories of brain cancer except for uncommon neuroepitheliomatous cancers (OR, 2.1; 95% CI, 0.9-4.7). CONCLUSIONS: Our data suggest that use of handheld cellular telephones is not associated with risk of brain cancer, but further studies are needed to account for longer induction periods, especially for slow-growing tumors with neuronal features. Abstract from PMID: 11122586 [PubMed - indexed for MEDLINE] It is noteworthy that despite relatively low use a nigh significant 2.1 OR was found for neuroepitheliomatous cancers, especially since there was scarcely any difference in exposure between cases and controls. Pakhomov, Dubrovick et al., (1991) Microwave influence on the isolated heart function: I. Effect of modulation. Bioelectromagnetics 1995;16(4):241-9 Pakhomov AG, Dubovick BV, Degtyariov IG, Pronkevich AN. Medical Radiology Research Center, Russian Academy of Medical Sciences, Obninsk, Kaluga Region, Russia. Dependence of the microwave effect on modulation parameters (pulse width, duty ratio, and peak intensity) was studied in an isolated frog auricle preparation. The rate and amplitude of spontaneous auricle twitches were measured during and after a 2 min exposure to 915 or 885 MHz microwaves and were compared to preexposure values. The studied ranges of modulation parameters were: pulse width, 10(-6)-10(-2) s; duty ratio, 7:100000, and peak specific absorption rate, 100-3000 W/kg. Combinations of the parameters were chosen by chance, and about 400 various exposure regimes were tested. The experiments established that no regime was effective unless the average microwave power was high enough to induce preparation heating (0.1-0.4 degree C). The twitch rate instantly increased, and the amplitude decreased, as the temperature rose; similar changes could be induced by equivalent conventional heating. The data provide evidence that the effect of short-term microwave exposure on the isolated heart pacemaker and contractile functions depends on pulse modulation just as much as modulation determines the average absorbed power. These functions demonstrated no specific dependence on exposure parameters such as frequency or power windows. Abstract from [PubMed - indexed for MEDLINE] Effects appear to be related to modulated signals. It is questionable whether such a short exposure (2 mins) would have any biological effect. The first author now works for the US Air Force who funds this research. Pakhomov, Prol et al., (1997) Search for frequency-specific effects of millimeter-wave radiation on isolated nerve function. Bioelectromagnetics 1997;18(4):324-34 Pakhomov AG, Prol HK, Mathur SP, Akyel Y, Campbell CB. Microwave Bioeffects Branch, U.S. Army Medical Research Detachment of the Walter Reed Army Institute of Research, Brooks Air Force Base, San Antonio, Texas 78235-5324, USA. Effects of a short-term exposure to millimeter waves (CW, 40-52 GHz, 0.24-3.0 mW/cm2) on the compound action potential (CAP) conduction were studied in an isolated frog sciatic nerve preparation. CAPs were evoked by either a low-rate or a high-rate electrical stimulation of the nerve (4 and 20 paired pulses/s, respectively). The low-rate stimulation did not alter the functional state of the nerve, and the amplitude, latency, and peak latency of CAPs could stay virtually stable for hours. Microwave irradiation for 10-60 min at 0.24-1.5 mW/cm2, either at various constant frequencies or with a stepwise frequency change (0.1 or 0.01 GHz/min), did not cause any detectable changes in CAP conduction or nerve refractoriness. The effect observed under irradiation at a higher field intensity of 2-3 mW/cm2 was a subtle and transient reduction of CAP latency and peak latency along with a rise of the test CAP amplitude. These changes could be evoked by any tested frequency of the radiation; they reversed shortly after cessation of exposure and were both qualitatively and quantitatively similar to the effect of conventional heating of 0.3-0.4 degree C. The high-rate electrical stimulation caused gradual and reversible decrease of the amplitude of conditioning and test CAPs and increased their latencies and peak latencies. These changes were essentially the same with and without irradiation (2.0-2.7 or 0.24-0.28 mW/cm2), except for attenuation of the decrease of the test CAP amplitude. This effect was observed at both field intensities, but was statistically significant only for certain frequencies of the radiation. Within the studied limits, this effect appeared to be dependent on the frequency rather than on the intensity of the radiation, but this observation requires additional experimental confirmation This abstract was from 9140663 [PubMed - indexed for MEDLINE. The efects reported were at fdensities one tenth of that of cellphones, and should really not be classified as a negative study. Roschke, Mann (1997) No short-term effects of digital mobile radio telephone on the awake human electroencephalogram. Bioelectromagnetics 1997;18(2):172-6 Roschke J, Mann K. Department of Psychiatry, University of Mainz, Germany. A recent study reported the results of an exploratory study of alterations of the quantitative sleep profile due to the effects of a digital mobile radio telephone. Rapid eye movement (REM) was suppressed, and the spectral power density in the 8-13 Hz frequency range during REM sleep was altered. The aim of the present study was to illuminate the influence of digital mobile radio telephone on the awake electroencephalogram (EEG) of healthy subjects. For this purpose, we investigated 34 male subjects in a single-blind cross-over design experiment by measuring spontaneous EEGs under closed-eyes condition from scalp positions C3 and C4 and comparing the effects of an active (0.05 mW/cm2) and an inactive digital mobile radio telephone (GSM) system. During exposure of nearly 3.5 min to the 900 MHz electromagnetic field pulsed at a frequency of 217 Hz and with a pulse width of 580 microseconds, we could not detect any difference in the awake EEGs in terms of spectral power density measures. The abstract was from PMID: 9084868 [PubMed - indexed for MEDLINE] The short exposure of awake subjects means that little biological effect could be expected. Longer exposure of sleeping subjects showed positive effects on EEG. Roti Roti et al., (2001) Neoplastic transformation in C3H 10T(1/2) cells after exposure to 835.62 MHz FDMA and 847.74 MHz CDMA radiations. Radiat Res 2001 Jan;155(1 Pt 2):239-247 Roti Roti JL, Malyapa RS, Bisht KS, Ahern EW, Moros EG, Pickard WF, Straube WL. Section of Cancer Biology, Radiation Oncology Center, Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, Missouri 63108, USA. The effect of radiofrequency (RF) radiation in the cellular phone communication range (835.62 MHz frequency division multiple access, FDMA; 847.74 MHz code division multiple access, CDMA) on neoplastic transformation frequency was measured using the in vitro C3H 10T(1/2) cell transformation assay system. To determine if 835.62 MHz FDMA or 847.74 MHz CDMA radiations have any genotoxic effects that induce neoplastic transformation, C3H 10T(1/2) cells were exposed at 37 degrees C to either of the above radiations [each at a specific absorption rate (SAR) of 0.6 W/kg] or sham-exposed at the same time for 7 days. After the culture medium was changed, the cultures were transferred to incubators and refed with fresh growth medium every 7 days. After 42 days, the cells were fixed and stained with Giemsa, and transformed foci were scored. To determine if exposure to 835.62 MHz FDMA or 847.74 MHz CDMA radiation has any epigenetic effects that can promote neoplastic transformation, cells were first exposed to 4.5 Gy of X rays to induce the transformation process and then exposed to the above radiations (SAR = 0.6 W/kg) in temperature-controlled irradiators with weekly refeeding for 42 days. After both the 7-day RF exposure and the 42-day RF exposure after X irradiation, no statistically significant differences in the transformation frequencies were observed between incubator controls, the sham-exposed (maintained in irradiators without power to the antenna), and the 835.62 MHz FDMA or 847.74 MHz CDMA-exposed groups. Abstract from PMID: 11121241 [PubMed - indexed for MEDLINE] It is possible that the heavy pre-irradiation with X rays masked any RF effects. Schmidt, Merritt et al., (1984) In utero exposure to low-level microwaves does not affect rat foetal development. Int J Radiat Biol Relat Stud Phys Chem Med 1984 Oct;46(4):383-6 Schmidt RE, Merritt JH, Hardy KH. Forty pregnant Sprague-Dawley rats were used in a study to determine low-level microwave effects on foetal development. Animals were continuously exposed to 2450 MHz radiation from day 2 to day 18 of gestation, at a specific absorption rate of 0.4 W/kg. Foetuses were removed, weighed and measured. They were then either fixed in Bouin's fluid, sectioned at 6 microns, stained with hematoxylin and eosin and examined by light microscopy, or cleared and stained with alizarin red. No significant difference was noted with respect to foetal weight, length, resorption rate or abnormality rate. Abstracte from PMID: 6334051 [PubMed - indexed for MEDLINE] Sienkiewicz, Blackwell et al., (2000) Low-level exposure to pulsed 900 MHz microwave radiation does not cause deficits in the performance of a spatial learning task in mice. Bioelectromagnetics 2000 Apr;21(3):151-8 Sienkiewicz ZJ, Blackwell RP, Haylock RG, Saunders RD, Cobb BL. National Radiological Protection Board, Oxfordshire, UK. zenon.sienkiewicz@nrpb.org.uk There is some concern that short-term memory loss or other cognitive effects may be associated with the use of mobile cellular telephones. In this experiment, the effect of repeated, acute exposure to a low intensity 900 MHz radiofrequency (RF) field pulsed at 217 Hz was explored using an appetitively-motivated spatial learning and working memory task. Adult male C57BL/6J mice were exposed under far field conditions in a GTEM cell for 45 min each day for 10 days at an average whole-body specific energy absorption rate (SAR) of 0.05 W/kg. Their performance in an 8-arm radial maze was compared to that of sham-exposed control animals. All behavioral assessments were performed without handlers having knowledge of the exposure status of the animals. Animals were tested in the maze immediately following exposure or after a delay of 15 or 30 min. No significant field-dependent effects on performance were observed in choice accuracy or in total times to complete the task across the experiment. These results suggest that exposure to RF radiation simulating a digital wireless telephone (GSM) signal under the conditions of this experiment does not affect the acquisition of the learned response. Further studies are planned to explore the effects of other SARs on learned behavior. Bioelectromagnetics 21:151-158, 2000. Published 2000 Wiley-Liss, Inc. Abstract from PMID: 10723014 [PubMed - indexed for MEDLINE] Veyret, Bouthet et al., (1999) Antibody responses of mice exposed to low-power microwaves under combined, pulse-and-amplitude modulation. Bioelectromagnetics 1991;12(1):47-56 Veyret B, Bouthet C, Deschaux P, de Seze R, Geffard M, Joussot-Dubien J, le Diraison M, Moreau JM, Caristan A. Laboratoire de Bioelectromagnetisme de l'Ecole Pratique des Hautes Etudes: ENSCPB. Irradiation by pulsed microwaves (9.4 GHz, 1 microsecond pulses at 1,000/s), both with and without concurrent amplitude modulation (AM) by a sinusoid at discrete frequencies between 14 and 41 MHz, was assessed for effects on the immune system of Balb/C mice. The mice were immunized either by sheep red blood cells (SRBC) or by glutaric-anhydride conjugated bovine serum albumin (GA-BSA), then exposed to the microwaves at a low rms power density (30 microW/cm2; whole-body-averaged SAR approximately 0.015 W/kg). Sham exposure or microwave irradiation took place during each of five contiguous days, 10 h/day. The antibody response was evaluated by the plaque-forming cell assay (SRBC experiment) or by the titration of IgM and IgG antibodies (GA-BSA experiment). In the absence of AM, the pulsed field did not greatly alter immune responsiveness. In contrast, exposure to the field under the combined-modulation condition resulted in significant, AM-frequency-dependent augmentation or weakening of immune responses. Abstract from PMID: 2012621 [PubMed - indexed for MEDLINE] It would not be true to say that this was a wholly negative study. Vijayalaxmi, Leal et al., (2001) Cytogenetic Studies in Human Blood Lymphocytes Exposed In Vitro to Radiofrequency Radiation at a Cellular Telephone Frequency (835.62 MHz, FDMA). Vijayalaxmi, Leal BZ, Meltz ML, Pickard WF, Bisht KS, Roti Roti JL, Straube WL, Moros EG. Department of Radiation Oncology, Center for Environmental Radiation Toxicology, The University of Texas Health Science Center, 7703 Floyd Curl Drive, San Antonio, Texas 78229. Vijayalaxmi, Pickard, W. F., Bisht, K. S., Leal, B. Z., Meltz, M. L., Roti Roti, J. L., Straube, W. L. and Moros, E. G. Freshly collected peripheral blood samples from four healthy human volunteers were diluted with RPMI 1640 tissue culture medium and exposed in sterile T-75 tissue culture flasks in vitro for 24 h to 835.62 MHz radiofrequency (RF) radiation, a frequency employed for customer-to-base station transmission of cellular telephone communications. An analog signal was used, and the access technology was frequency division multiple access (FDMA, continuous wave). A nominal net forward power of 68 W was used, and the nominal power density at the center of the exposure flask was 860 W/m(2). The mean specific absorption rate in the exposure flask was 4.4 or 5.0 W/kg. Aliquots of diluted blood that were sham-exposed or exposed in vitro to an acute dose of 1.50 Gy of gamma radiation were used as negative or positive controls. Immediately after the exposures, the lymphocytes were stimulated with a mitogen, phytohemagglutinin, and cultured for 48 or 72 h to determine the extent of genetic damage, as assessed from the frequencies of chromosomal aberrations and micronuclei. The extent of alteration in the kinetics of cell proliferation was determined from the mitotic indices in 48-h cultures and from the incidence of binucleate cells in 72-h cultures. The data indicated no significant differences between RF-radiation- and sham-exposed lymphocytes with respect to mitotic indices, incidence of exchange aberrations, excess fragments, binucleate cells, and micronuclei. In contrast, the response of the lymphocytes exposed to gamma radiation was significantly different from both RF-radiation- and sham-exposed cells for all of these indices. Thus, under the experimental conditions tested, there is no evidence for the induction of chromosomal aberrations and micronuclei in human blood lymphocytes exposed in vitro for 24 h to 835.62 MHz RF radiation at SARs of 4.4 or 5.0 W/kg. Abstract from PMID: 11228563 [PubMed - as supplied by publisher] Wang, Van Dop et al., (1991) References and Bibliography |
