Вы находитесь на странице: 1из 255

January 16, 2001

Translation: June 2002

Mobile Telephones, Base Stations and Health

Current state-of-knowledge and recommendations

A report to the Director General of Health of France

Translated from the original French text by MCL

17B Woodford Road, London E18 2EL


MCL makes no warranty, expressed or implied, with respect to the use of any information or
procedure presented in this Report. MCL disclaims all liability for any loss, damage, injury or
other consequence whatsoever arising from any use howsoever caused, including any
resulting from error, omission or negligence in its application.




1.1 Requirements for assessment of the scientific literature 7

1.2 Biological effects and health effects 7

1.3 Should the Precautionary Principle be applied to RF and health? 8

1.4 Selection criteria and analysis methods in recent reports and documents
concerning mobile telephones and health 11



2.1 Introduction 14

2.2 Basic physics of electromagnetic fields 14

2.3 Biological effects of radiofrequency electromagnetic fields. 17

2.3.1 Thermal effects 17
2.3.2 Non-thermal effects 18

2.4 Mobile telephony: technical specifications 19

2.5 Deployment of base stations and exposure of people 23

2.6 Consideration of the question of reflection and amplification of waves. 32

2.7 The question of co-location of base stations 33

2.8 Results of field measurements with hand-free kits 33

2.9 Technical regulations and instrumentation 33





4.1 Fundamental reviews 37

4.1.1 Analysis of the ARCS Report (Austrian Research Centre Seibersdorf) 37
4.1.2 Analysis of the Stewart Report 42
4.1.3 Analysis of the "McKinlay" and COST 244 bis Reports 50
4.1.4 Analysis of the Report Of The Royal Society Of Canada 57
4.2 Additional reports 64
4.2.1 Analysis of the Report of the French Academy of Science 64
4.2.2 Analysis of the Report by Essor-Europe 75
4.2.3 Analysis of the COMAR Report 77
4.2.4 Analysis of the Report to The Swedish Government: “Review of
Electromagnetic Fields and Health” 79
4.2.5 Analysis of the Colloquium of the National Assembly “Mobile Phones: a
Danger To Health?” (19 June 2000) 80
4.2.6 Analysis of the Sage Report 81
4.2.7 Analysis of the paper by J M Elwood 88

4.3 Recent scientific reports 92

4.3.1 General articles and experimental work: 92
4.3.2 Epidemiological Studies 98

4.4 Children and Exposure to RF Associated with Mobile Telephones 103

4.5 Expert hearings 105

4.5.1 Joe WIART 105
4.5.2 Jean-Claude CARBALLES 108
4.5.3 Yvette SEGALA 114
4.5.4 Elisabeth CARDIS 117
4.5.5 Jean-Pierre CHEVILLOT 121
4.5.6 Madeleine BASTIDE 125
4.5.7 Jean-Marie ARAN 127
4.5.8 Pierre BUSER 130
4.5.9 Philippe HUBERT 133
4.5.10 J P VAUTRIN 151
4.5.11 Michèle RIVASI 154
4.5.12 Mrs. Gaëlle PATETTE and Mrs. Carole MATRICON 158
4.5.13 Professor Jacques FOURCADE 160
4.5.14 George CARLO 165
4.5.15 Marc SÉGUINOT 166
4.5.16 Laurent BONTOUX 167



5.1 Outline of recommendations from recent reports 169

5.2 Recommendations of the group of experts 170


6.1 Actual state of research at the international level 175

6.2 The expert group's recommendations for research 180

6.2.1 Studies of biophysical interactions 182
6.2.2 In vitro studies 182
6.2.3 Animal studies 183
6.2.4 Human laboratory studies 184
6.2.5 Epidemiological studies 185
6.3 Funding and organising the research 189














The expert group would like to thank Michel Le Diraison (PIOM, CNRS Bordeaux), Yvette
Lacoste (UJF Public Health Laboratory), and Isabelle La Groye (PIOM, CNRS Bordeaux)
who have contributed to the preparation of this report. Also acknowledged are the editors of
the scientific journals who agreed to communicate to the expert group articles in press.

MCL would like to thank Emilie Henein, who transcribed this translation.
Executive Summary
"Do the conclusions of recent reports summarising our current knowledge of the health
hazards related to the use of mobile telephones and their accessories justify the adaptation
of the risk management regulations recently adopted by the French and European
authorities?". This is, in effect, the crux of the question submitted to the expert group brought
together by the Direction Générale de la Santé (Health General Directorate).

The frequency range used for mobile telephony varies according to companies and their
technologies, ranging from 850 to 1900 MHz. The range will be extended to 2200 MHz with
the new UMTS technology, and to the 400 MHz waveband with the TETRA system, currently
under development. These are part of the much wider range of radio frequencies present
everywhere in our environment, at home (microwave cooking, etc.), at work (industrial
heating systems, etc.), or in public places (radio and television transmitters, burglar alarm
systems and remote-control devices, etc.), especially in urban areas.

The development of telecommunications has been followed by research into the effects of
radio-frequency electromagnetic fields (RF) on biological systems. Work first started in this
field after the Second World War. This research focused particularly on mechanisms that
could link exposure of human cells to the development of cancers. It is still too soon,
however, to assess any long-term effects. The highly-complex physical and biological
phenomena necessitated the development of new experimental, measurement and
observation procedures that were not always completely controlled in the early research
projects. This is why it is still difficult to draw clear conclusions, in spite of the enormous
volume of scientific work on the subject. Some research evidenced short-term modifications
in certain physiological or biochemical parameters, or even fine neuro-sensory functions,
while other work contradicted these results. The significance of these observations in
predicting the occurrence of long-term effects is debatable.

The public are naturally concerned by this difficulty in drawing conclusions. The issue of
potential health hazards resulting from exposure to RF takes on a very special importance
when it is considered that 30 million people are users of mobile telephones in France and
that the expected market is 44 million within 4 years. Even if the individual risk were very
small, the very number of people involved would produce a considerable impact in terms of
public health.

On the other hand, mobile telephones also contribute to health and safety. The speed with
which the alert can be given in case of accident, fire, or other dangerous situations, and the
effectiveness of emergency services have been considerably improved by the widespread
availability of this technology, which has already saved many lives worldwide. An evaluation
of the trade-off between risks, if there are any, and potential advantages, was not part of the
expert group's brief, which focused solely on risk assessment, based on the analysis of
scientific data alone.

The first chapter of this report describes, in detail, the critical synthesis approach used to
develop the expert group's opinion and recommendations. The second chapter presents the
sources and characteristics of the electromagnetic fields associated with mobile telephony,
and the known or investigated mechanisms by which they interact with living matter. The
third chapter gives the threshold limit values for public exposure to RF associated with
mobile telephony and explains the scientific basis for the figures.

The fourth chapter is the longest. It summarises the expert group's analysis of the current
state of scientific knowledge. Several scientific bodies have recently produced reports
containing comprehensive analyses of the biological and medical effects of RF. These
bodies, consisting of top-level experts in the various scientific fields involved, have analysed

all the scientific data available at the time. The expert group used five summary documents,
covering several hundreds of articles published in the scientific literature, to carry out their
mission. In addition to these ‘basic reports’, seven 'additional documents' were taken into
consideration: symposium proceedings and summary articles providing interesting
information. The expert group also made sure to take the most recent published work into
consideration, right up to the day this report was completed. Finally, the group interviewed
about twenty people from scientific organisations, administrations, industry, associations,
and politics, both to obtain further information and identify society's concerns on this issue
more fully.

The expert group had two main objectives in carrying out this assessment of scientific

To define the areas in which there were convincing scientific data proving the existence or, a
contrario, the absence of biological and medical consequences following exposure to RF
related to the use of mobile telephones and the operation of base stations (i.e. ‘what we

To highlight the areas in which currently-available scientific data does not exclude the
possibility of biological and medical effects, without necessarily confirming their existence
(i.e. ‘what is uncertain’).

The expert group's conclusions and recommendations are presented in the fifth chapter.
They are based on the following considerations:

There is considerably less personal exposure in the vicinity of base stations – with the
exception of exclusion areas – than there is when making a call with a mobile phone.

Scientific data indicate, with comparative certainty, that, during exposure to RF from a
mobile phone, a variety of biological effects occur (eg. electroencephalogram profile,
reaction time, etc.) at energy levels that do not cause any local increase in temperature.
However, in the current state of knowledge of these non-thermal effects, it is not yet possible
to determine whether they represent a health hazard.

Although this assertion is backed up by little scientific argument, the hypothesis that certain
medical effects are caused by the low-level RF fields associated with mobile telephones
cannot be completely excluded, in the current state of knowledge. Experimental and
epidemiological research into a range of health problems, including brain cancers and
headaches, is currently in progress; the role of exposure to RF in these symptoms or
diseases has not yet been clarified. However, in view of the exposure levels observed, the
expert group does not back the hypothesis that there is a health risk for populations living in
the vicinity of base stations.

If future research were to validate this hypothesis, i.e. demonstrate the existence of health
hazards, the risk, at an individual level, would probably be very low. Indeed, it is reassuring
to note that it has not yet been demonstrated, in spite of the considerable amount of work
done over the past several years. However, if mobile phone radiofrequency fields were
hazardous, the very high number of mobile telephone users could mean that, even if the
individual risk were very low, the impact on public health could be considerable.

The risk of accident and fatality associated with the use of mobile telephones when driving
has definitely been established. In the current state of knowledge, this is the only known
health risk, albeit a very serious one.

For all of these reasons, and in view of the brief they were given, the expert group
recommend a risk management approach based on the precautionary principle, aimed at
reducing public exposure to RF associated with mobile telephony to the lowest possible level
compatible with service quality and justified by current scientific data. The various measures
recommended are described in the full report. The objective is also to ensure that users and
the public have access to comprehensive information on their exposure. The expert group
consider that these recommendations would make it possible to apply the precautionary
principle in an enlightened way, i.e. on the basis of a rational approach.

The sixth and last chapter is devoted to recommendations for advanced research to
elucidate the remaining uncertainties in priority areas. Proposals are made concerning ways
of funding research that would guarantee the scientists' independence from the various
interests involved.

The expert group would like to emphasise that they have been able to work completely
independently, both from industry and public authorities. The Direction Générale de la Santé
(Health General Directorate) provided them with the effective, discreet support necessary to
complete their task.

“All scientific work is incomplete – whether it be observational or experimental. All scientific
work is liable to be upset or modified by advancing knowledge. That does not confer upon
us a freedom to ignore the knowledge we already have, or to postpone the action that it
appears to demand at a given time.”

Sir Austin Bradford Hill,

The Environment and Disease; Association or causation? 1965

"Do the conclusions of recent reports summarising our current knowledge of the health
hazards related to the use of mobile telephones and their accessories justify the adaptation
of the risk management regulations recently adopted by the French and European
authorities?". This is, in effect, the crux of the question submitted to the expert group
brought together by the Direction Générale de la Santé (French Health General Directorate).

The electromagnetic fields associated with mobile telephones occur in an environment which
already contains a wide range of electromagnetic frequencies (radiofrequencies, RF, from
30 kHz to 300 GHz), particularly in urban areas. This is true of the domestic environment
(microwave cookers, radio and television signals), the work environment (industrial heating
systems, medical diathermy equipment) as well as in public places (radio and television
transmitters, radars, emergency services and taxi communication systems, burglar alarms,
and remote controls). The frequencies used by mobile telephones, depending on the
operator and the technology used, range from 850 to 1900 MHz, and will reach 2200 MHz
with the development of the new UMTS technology, and 400 MHz with the projected TETRA
system. Mobile telephone systems have two specific features which have aroused
legitimate concern among the public: for handsets it is the immediate proximity of the
telephone antenna to the head during conversations, and for base stations it is the
proliferation of antennas in the immediate environment. The huge number of users requires
an increasing number of cells to ensure optimal coverage, especially in built-up urban areas
where there are many physical obstacles. These antennas are visible for all to see: 'macro
cells' on roofs or towers, and 'micro' or 'pico cells' on the façades of buildings or even inside
public premises. As of 28 December 2000, 29,416 base stations had been installed in
France (macro cell, micro or pico cell antennas). In the course of the last three months of the
year 2000, more than 1,664 new base stations were installed and 589 modified, while 403
were abandoned.

The development of telecommunications has been followed by research into the effects of
radiofrequency electromagnetic fields on biological systems. Work started in this field after
the Second World War and a considerable body of scientific literature has been published.
Research was given a new impetus in the United States and then all over the world following
legal proceedings taken in 1992 by a citizen who accused RF of being responsible for the
death of his wife from brain cancer. Work has focused on this type of pathology, exploring
the biological mechanisms which may link the RF exposure of human cells to the
development of carcinogenic processes. Results of experiments on animals or isolated cells
have been published in a variety of scientific journals. However, the timescale of these
experiments is, by necessity, to short to allow the identification of any possible long-term
effects. Some authors have reported a possible link between certain forms of brain cancer
and the use of mobile telephones, whereas other recent publications have demonstrated the
opposite. All experts recommend continuing this research before drawing any conclusions
as to the consequences, which, similar to those of chemical substances and ionising
radiation, may appear only after many years of exposure. The highly-complex physical and
biological phenomena necessitated the development of new experimental, measurement
and observation procedures that were not always completely defined in the early research
projects. The necessary replication of results, particularly replication under identical

experimental conditions, is then very difficult. This is why it is still not easy to draw clear
conclusions, in spite of the enormous volume of scientific work on the subject. Some
research has indicated short-term modifications in certain physiological or biochemical
parameters, or even fine neuro-sensory functions, while other work has contradicted these
results. The significance of these observations in predicting the occurrence of long-term
effects is debatable.

That there is scientific debate comes as no surprise to specialists familiar with studies of
environmental risks, who know just how difficult it is to demonstrate the harmful effects of a
chemical substance or a physical or microbiological agent under everyday exposure
conditions. The public are naturally concerned by this difficulty in drawing conclusions. Is it
not precisely in such situations where the risk of serious consequences are uncertain that
the ‘Precautionary Principle’ applies? The answer is certainly yes, if the scientific indications
of possible “serious and irreversible” effects are sufficient to establish “reasonable doubt”,
but if this is the case, how far should the precautions go? Which aspects of exposure to RF
should they cover? Public and media pressure is focused on base station antennas, but the
exposure level is much less than that during a conversation on a mobile telephone.

The issue of potential health hazards resulting from exposure to RF takes on a very special
importance when it is considered that 30 million people (29,681,300 subscribers on 31
December 2000) are mobile telephone users in France and that the expected market is 44
million within 4 years. This is a worldwide phenomenon with more than 27 million users in
Great Britain and around 80 million in the United States, for example. Even if the individual
risk were very small, the number of people involved could produce a considerable impact in
terms of public health. The search for an answer to this question is thus an urgent one.
However, the number of people concerned is not enough to establish the existence of a
hazard, if RF is not dangerous, under current exposure conditions. A rapid response may be
required, but it will necessarily be dependent on the current state of knowledge.
Furthermore, mobile telephones also contribute to health and safety. The speed with which
the alert can be given in case of accident, fire, or other dangerous situations, and the
effectiveness of emergency services have been considerably improved by the widespread
availability of this technology, which has already saved many lives world-wide.

The brief of the expert group convened by the Direction Générale de la Santé did not cover
other aspects of the development of this communication technology in business, or the fact
that it facilitates exchanges between people, although competent national and international
authorities will certainly consider these factors. The management of proven or strongly
suspected risks is thus a part of a balance of costs and benefits, as no technology likely to
induce a risk, however small, could be justified if it did not also provide substantial

Consideration of such a risk-benefit analysis was not part of the expert group's brief, which
focused solely on assessing the risks associated with using mobile telephones and their
accessories. The group did not take into account scientific work relating to low or extremely
low frequency fields; other technologies using RF – radio, television, and radar – will only be
mentioned in the context of health impact studies which provide indications for future

In recent years, several health bodies have produced reports summarising the state of our
current knowledge of the biological and health effects of RF. Several groups of top-level
experts from the various scientific disciplines involved have worked for months to collect and
summarise all the literature available at the time. Examples include the commission led by
Professor W. Stewart in Great Britain (May 2000) or the Royal Society of Canada report
(March 1999). In preparing their report, the French expert group examined these documents
and many others (see full list in Appendix). They also updated the corpus of knowledge,

taking into consideration several works published after the completion of the work of these
commissions. Finally, the group interviewed a number of well-known people from scientific
organisations, administrations, industry, associations, and politics, both to obtain further
information and identify society's concerns on this issue more fully.

The expert group had two main objectives in carrying out this assessment of scientific

To define the areas in which there were convincing scientific data proving the existence or, a
contrario, the absence of biological and medical consequences following exposure to RF
related to the use of mobile telephones and the operation of base stations (i.e. ‘what we

To highlight the areas in which currently-available scientific evidence does not exclude the
possibility of biological and medical effects, without necessarily confirming their existence
(i.e. ‘uncertainty factors’).

Recommendations were made for research into priority areas, with the objective of
eliminating the remaining uncertainties. Guidelines were also defined to ensure that users
and the public benefit from the highest level of health protection justified by current scientific
data, as well as to encourage manufacturers and operators, as well as users, to take every
possible measure to reduce exposure to a minimum.

The expert group would like to emphasise that they have been able to work completely
independently, both from industry and public authorities. The Direction Générale de la Santé
(Health General Directorate) provided them with the effective, discreet support necessary to
complete their task.

1 The Expert Group's Approach

1.1 Requirements for assessment of the scientific literature

Scientific knowledge is, by nature, imperfect and ever changing, as science and
technologies evolve. When it comes to levels of exposure to environmental factors – in this
case electromagnetic fields at radio and associated frequencies – which can be qualified as
“weak”, the biological and health effects produced are often of modest intensity and
expressed in a variety of ways. Effects observed depend on individuals and circumstances,
and some may only appear after several years. These aspects characterise many
environmentally-related health issues and can lead to the publication of apparently
discordant or inconclusive scientific works at any given moment.

In such a context, it is a delicate matter to summarise knowledge with the aim of bringing out
the general meaning and formulating recommendations for managing any risks that may
have been identified. This is generally referred to as “assessment”, i.e. the art of translating
current scientific data into guidelines for action or information. This work can only be
collective. It must satisfy quality and reliability criteria, as must the scientific work on which it
is based. Readers who are keen to understand the scientific approach will see the difficulty
of discerning a clear message from a limited number of inconsistent results by looking at two
examples presented in the appendices. Both are taken from epidemiological studies of the
risk of brain cancer after exposure to RF. Brain cancer is a morbidity which particularly
focuses research and public emotions. The appendices describe two large-scale studies that
have been interpreted or presented in recently published articles by individuals or groups of
experts in a highly contrasting or even contradictory manner, as illustrated by quotations and
comments relating to their work. Although it is now widely recognised in the scientific
community – not yet the case in every field – that an isolated study can never be held up as
a demonstration or definite confirmation of a complex physical or biological phenomenon,
these examples also show how strict an approach must be used to analyse and interpret the
results of scientific work, even in the framework of a joint assessment. Epidemiological
literature is particularly subject to this interpretation bias, due to the multi-factor nature of the
biological and health phenomena observed. We would be mistaken, though, in thinking that
the experimental approach does not face the same difficulties, particularly as results are
highly sensitive to measurement conditions. This state of affairs is certainly regrettable, but it
is also real and generally applicable. Everyone can thus understand the importance of the
collective and explicit character of this process of critical reading and reviewing of all the
scientific literature available at a given moment on a scientific subject which is not yet

1.2 Biological effects and health effects

The results of the scientific work on which our assessment is based must therefore be
analysed with a critical, panoramic view, covering all the data available at a given time. It is
also necessary to decide on the interpretation of various ‘biological effects’ observed after
exposure to radiofrequency electromagnetic fields in terms of their risk to health. Adopting
the point of view of various international bodies working on the subject1, the expert group
uses the term ‘biological effects’ to refer to changes of a physiological, biochemical or
behavioural nature which are induced in an organism, tissue, or cell in response to external
stimulation. A biological effect does not necessarily represent a threat to the health of an
individual; it may simply be the normal adaptation response of the cell, tissue, or organism to

See, for example, the report of the Royal Society of Canada of March 1999 (p15) or the
introductory document to the EMF program of the WHO (can be viewed on the internet site

this stimulation. A ‘health effect’ is a biological effect which may endanger the normal
working of an organism in that it goes beyond the framework of the ‘physiological’ responses
to the action of the external agent. This distinction is important and easy to understand.
Nobody could confuse the hearing of a sound (the enchanting song of a nightingale or the
disturbing noise of a dust-cart early in the morning), which is a biological effect (resulting
from a complex chain of elementary biological effects: mechanical, biochemical and
electrical), on the one hand, and the gradual (or sometimes sudden) loss of hearing after
prolonged exposure to noises of great intensity, such as that suffered by youngsters who go
to concerts with powerful amplifiers or in a metalworker after years of hammering without ear
protection: this loss is a serious adverse health effect.

Likewise, many people appreciate the biological effects induced by moderate exposure to
the sun, leading to the tanning of the skin via an increase in pigment production (melanin) by
the specialised cells in the skin, but they make a clear distinction between this effect and the
painful burns that come after excessively fast exposure to UV radiation without skin
protection, as well as the induction of a melanoma, both health effects which pose a serious
threat to the health of those who enjoy prolonged, intense sunbathing.

The duration or intensity of exposure to the agent causing the biological effect, and the
nature of the cell, tissue, or organism in which the effect is manifest, as well as other aspects
that have not all been elucidated, have a considerable impact on the link (or absence of one)
with a possible health effect. A wide range of biological or functional measurements are
undertaken when studying the effects of exposure to RF; some showing biological effects
according to the definition given above. Those which may be predictive of a health effect
remain to be determined. These 'hazardous biological effects' are causes for concern and, if
established, should be subject to preventive measures. Proving their existence is not an
easy task. First of all, they must regularly precede the occurrence of the suspected health
disorders or be associated with them. They may also constitute a stage in the chain of
biological effects leading to these disorders, in the human species in general or only in some
of its representatives (susceptible individuals) or, failing that, in several other species of
laboratory animals. An example which illustrates this, with regard to the non-thermal effects
of RF, is the increase which has frequently been described in the biochemical activity of
ornithine decarboxylase, an enzyme which may play a role in the development of cancerous
cells. This will be discussed later.

1.3 Should the Precautionary Principle be applied to RF and health?

World Health Organisation memorandum no. 193, published on 28 June 2000, states: “It is
clearly established that all the proven effects (our own italics) of exposure to
radiofrequencies are related to this warming [due to the thermal effects of RF, ed]”. This
principles behind this statement are followed also by the International Commission on Non-
Ionizing Radiation Protection (ICNIRP), an independent scientific body whose 1998
recommendations are founded only on the effects established by the available scientific data
(see Chapter III). Likewise, the Recommendation by the Council of the European Union of
12 July 1999 concerning the limitation of exposure by the public to electromagnetic fields
stipulates that “only proven effects have been used to establish the recommended exposure
limitation”. But what of those unproven effects which are currently the object of much
scientific research and could, according to some hypotheses, result from non-thermal
mechanisms? Must we wait for them to be either proven or formally refuted before decisions
are taken to manage these hypothetical risks? This is the question asked of industrial and
political protagonists concerned by the effects of RF.

The Precautionary Principle is a policy tool for the prudent management of unproven risks. It
can be applied whenever plausible mechanisms or experimental or epidemiological
observations provide a minimal scientific foundation. It is essentially a matter of risk

management and not of evaluation; evaluation must try to keep to the area of facts or
scientifically-founded hypotheses provable by experimentation, or strict epidemiological

This separation between 'objective science' and 'political management' can, however, turn
out to be rather theoretical in practice. In cases where facts have not been proven by
scientific means, the mere act of summarising knowledge leads the scientific community to
form judgement criteria on the existence or absence of links between exposure to the agent
studied and the biological or health effect under consideration. These criteria, however
objective they may be, are not always totally free of extra-scientific considerations. What
constitutes scientific ‘proof'? What is the degree of proof required to accept (or refute) the
hypothesis of causality? In their desire to protect health, scientists include elements of
caution in these criteria. But it is understandable that when they come to examine this
‘evidence’, they do their utmost to take into consideration the actual nature (serious and
irreversible, or benign and short-lived) of the health effect under consideration, otherwise
their judgement may be biased. The fact that the exercise is a collective one is invaluable,
but it does not provide an absolute guarantee.

In contrast, ‘decision-makers’ must take into account the nature of the potential health
effects, as well as the number of people exposed in the population (today and in the
foreseeable future). It is also their responsibility to consider the availability and cost of
solutions available to reduce exposure of the population. In a context where millions of
people in France already use mobile telephones and the number is constantly increasing,
and families see them as a means of staying in close contact with their children, the dilemma
of 'the lesser evil' is particularly acute. It is necessary to assess the balance between the
risks linked with developing a technology that may one day be found to have dangerous
health effects, on the one hand, and the damage (including health considerations, as we are
reminded by the European Union Council recommendation dated 12 July 1999) that could
result from unjustified restrictions obstructing its development, on the other hand. One of the
World Health Organisation's working documents2 thus warns the health authorities against
taking measures which, although inspired by a legitimate desire to reassure public opinion
and prevent hypothetical health hazards, would have no scientific foundation. Indeed, there
is a great risk that varying levels of pressure from public opinion could lead governments to
enact disparate ‘safety standards’ which could ruin all efforts to achieve international
harmonisation. The confusion and fears of the public would then be artificially exaggerated,
and there would certainly be conflicts about ‘arbitrary rulings and unfair competition or
constraints on commercial transactions'.

The Precautionary Principle cannot therefore justify measures without any rational
foundation. The demonstration of hazardous biological effects, if there are any, would
certainly require action to prevent the possible consequences, but that would not suffice to
provide a rational basis for an efficient approach if the physical parameters requiring action
were not yet fully understood. Exposure to RF, and the resulting energy absorbed by the
body, depends on many factors, such as intensity of the field (which depends on the position
of the device in relation to the base stations and on the position and type of antenna), field
modulation – which makes it possible to transmit information – exposure duration (long when
one is near a base station, but at very low intensity, intermittent yet more intense with the
telephone itself and variable over time in both cases). For example, a precaution which
appears to be ‘common sense’ (such as installing a physical protection around the antenna
of the mobile telephone to ‘protect’ the head) is in fact counter-productive, for it causes the
automatic power control of the telephone to increase field intensity to compensate for
weaker reception. There are, in fact, many ways of reducing exposure to RF, but if the

Draft Fact Sheet for Final Review. Electromagnetic fields and public health cautionary
policies. (6 July 2000); the document can be viewed on the site http://www.who.int/peh-emf/

health risks were clearly established or strongly suspected, we would need to act specifically
on the physical parameters responsible for the deterioration of the working of cells or
tissues, which are not necessarily the same as those which enable communication between

In its text on the Precautionary Principle3, the European Commission proposed certain
guidelines with the aim of “finding an adequate balance that makes it possible to make
proportionate, non-discriminatory, transparent and coherent decisions – (through) – a
decision-making process which is structured, based on detailed scientific data and other
objective information”. It reminds us that “the Precautionary Principle which decision-makers
use essentially in the context of risk management must not be confused with the element of
caution applied by scientists to the evaluation of scientific data”. One of these guidelines
which is particularly relevant in this case is the desire to ensure that the measures
implemented in accordance with the Precautionary Principle are:

· proportionate to the desired level of protection (and therefore to the level of risk to be

· consistent with similar measures already adopted in comparable fields;

· based on an examination of the potential advantages and drawbacks of acting or not


· re-examined periodically in the light of new scientific data.

The expert group was consulted about the existence of biological effects or health hazards
resulting from exposure to RF associated with mobile telephones. This scientific body has no
brief to decide on the size of the safety margin that would be required, if such effects were
proved to exist today or were only seriously suspected, in the light of the current data, to
achieve a level of risk which could be defined as ‘acceptable’. This is a matter for political
judgement legitimised by public debate on these issues. In contrast, the expert group will
determine whether the current state of our knowledge justifies a reduction in the exposure of
mobile telephone users or people who frequent places within the field of base stations. In
this case, they may make scientifically-founded recommendations concerning ways of
reducing exposure. Their report will indicate whether, despite any remaining uncertainties,
the facts appear to be sufficiently well-founded and serious, in health terms, to adopt a
precautionary approach. It will present the main lines of research required to eliminate these
uncertainties. It can be seen that ‘experts’ and ‘decision-makers’ have distinct, yet closely
related ambits.

Although the Precautionary Principle, used by public authorities for the prudent management
of potential hazards as part of a wide range of more-or-less restrictive measures – e.g.
regulatory, administrative, informative, etc. – is a public policy tool, various exposure-
reduction measures can be taken by industry or even individual users. The “prudent
avoidance” concept can be defined as the full set of voluntary measures which can be taken
by private individuals to minimise any unnecessary and/or easily avoidable exposure. For
example, telephoning under poor radiocommunication conditions (e.g. in certain closed
spaces) leads to a substantial increase in the radiation received. If we are aware of this and
have means available for acting on the information, it is then up to each individual to behave
appropriately, in an enlightened, responsible manner. The expert group thus considered a
set of measures – compulsory, recommended, or voluntary – which would contribute to
reducing personal exposure. Once they have explained their recommendations, they will

Communication from the Commission about recourse to the Precautionary Principle, 2
February 2000

indicate the measures they consider the most appropriate in the light of current knowledge of
the risks.

1.4 Selection criteria and analysis methods in recent reports and

documents concerning mobile telephones and health
There is a certain international consensus on the basic rules for ‘good assessment practice’
and this was adopted by the expert group4. Any group conducting this type of assessment
must represent a range of scientific specialities, given the complexity of the issue at stake,
and members should also have expressed a variety of opinions on the subject. All the
different points of view on the issue must be taken into account, including any divergences,
either within the expert group itself or through hearings or other forms of communication.
This requirement for plurality, in conjunction with transparency regarding any conflict of
interests that may exist in the group5, aims to ensure that the final opinion of the group is not

The criteria governing the selection of the scientific materials to be reviewed must be
explicit. The expert group thus selected 5 summary reports written by committees of experts
that met the following criteria:

· the expert committees included scientists from several disciplines connected with RF,
sometimes along with specialists from disciplines not directly involved with RF, and
produced their reports for national or international health authorities,

· the review of scientific evidence carried out by the committees of experts was based on
publications in scientific journals with a reading committee ('peer-reviewed' cf. below),
and was intended to be exhaustive at the date of review,

· the committees' criteria for assessing the literature were explicit,

· these reports have been published since 1996 (date of the McKinlay report for the
European Union), as it was considered that a sufficient amount of scientific data had
been published on the biological and health effects of RF by that date.

The expert group chose to accept only summary reports based on articles published or
accepted for publication in scientific reviews with a peer-review committee; this rule ensures
that the work in question has been scrutinised by specialists in the same subject, who were
not a part of the project itself. Although this does not provide an absolute guarantee of
quality, and even less so of truth, this rule is widely accepted in various international
assessment bodies and makes it possible to base the synopsis on information which meets
minimum quality conditions, thus avoiding fanciful or purely anecdotal documents and
limiting the amount of non-validated work. In our opinion, work refused for publication in the
(many) scientific journals available could have any claim to being superior, and this view
must also be accepted by those authors who consider it unnecessary to submit themselves
to external scrutiny. Despite the delays which these ‘peer review’ procedures cause
(sometimes as long as 1 year), the vast majority of topics considered in this report have
See for example “Evaluation and use of epidemiological evidence for environmental health
risk assessment”, WHO-Euro, Copenhagen, 2000
All the members of the group of experts filled out an information sheet on which, following
the example of the experts assigned by other health safety agencies such as the AFSSAPS,
they declared the scientific works carried out in conjunction with or financed by companies
involved in the development of mobile telephony, as well as any business interests they
might have in such companies.

been studied for long enough for the group of experts to obtain sufficient publications related
to each subject. Except for recently published works, the expert group considered that it was
neither useful nor feasible, in the time it had to give its opinion, to go back over each of the
hundreds of articles which had been analysed in detail in the summary reports it was
studying. Some particularly innovative recent works may, however, be exempted from this
rule. They were analysed on a case-by-case basis and their inclusion in this report was
clearly explained. Each of the ‘basic’ summary reports was subject to detailed critical
analysis by the expert group, who then gave an opinion on the scientific relevance of each
set of conclusions.

This report presents the critique of all the basic reports, following a single plan. In order to
make it easier to read this synopsis, this plan systematically follows the various morbidities
studied in the most recent summary report directed by William Stewart (May 2000).
Whenever possible, the reader will find, successively, studies concerning the nervous
system and behaviour; those concerning cancer, reproduction and development; the
cardiovascular system; the immune and blood systems; other miscellaneous disorders that
may have been studied; interference with biomedical implants and, finally, the risks linked
with driving a car when using a mobile telephone, with or without a hands-free system. For
each theme, the conclusions of the authors of the summary report are cited in the
introduction, in an identifiable manner, followed by a short presentation of the main work
justifying these conclusions. The judgement expressed by the expert group is mentioned at
the end of each theme. A summary of the opinion of the expert group is presented at the end
of each chapter devoted to a given report.

Besides the ‘basic reports’, a certain number of ‘additional documents’ were also taken into
consideration. They do not satisfy the criteria mentioned above but do provide other
interesting information (cf. list in the appendix). The points included in these documents
which were inadequately or differently covered in the ‘basic reports’ have been highlighted.

The expert group also collected all the scientific literature published after the most recent
summary report. Besides the usual exercise of bibliographical monitoring via computerised
databases, the group also contacted the main scientific publishers who were likely to receive
articles on the subject (about forty journals were identified on the basis of the articles
published in recent years), asking them to inform them of any article accepted for publication
(and therefore positively reviewed by peers) and intended for publication by the end of the
year 2000. Many editors responded (see list in the appendix of journals contacted and those
which responded). These recent original articles were added to the list of additional
documents studied with the ultimate aim of enabling the expert group to give a global
judgement of all the scientific material available.

Finally, the group interviewed a number of well-known people from scientific organisations,
public bodies, industry, associations, and the political sphere, both to obtain further
information on aspects not covered in scientific literature and identify society's concerns on
this issue more fully.

The principle of transparency also applies to the criteria used by the group in forming their
judgement on the state of knowledge, so that third parties would have an opportunity to
criticise their choices and tools. The criteria chosen by the expert group should be
understood in light of the main question they were asked: "Do the recently published
summary reports provide comprehensive, reliable information which the French health
authorities can use as a basis for updating their principles and rules for managing the risks
associated with the use of mobile telephones and base stations?” The expert group
considered that any response to this question would necessitate answering three logically
connected questions:

· What are the demonstrated biological effects of RF?
· Among these biological effects, which can be considered reasonably predictive of a
health effect, in the light of our current knowledge?
· Given what we know today, can we determine RF exposure levels and/or conditions
which would reduce or eliminate these possible hazardous biological effects?
The guidelines adopted for preparing this “expert judgement” were those stipulated more
than 30 years ago by the famous British epidemiologist Bradford Hill: “Is there any other way
of explaining the set of facts before us, is there any other answer equally likely, or more
likely, than cause and effect?”6.

The main criteria selected by the expert group for assessing the quality of the knowledge
summaries were as follows: the exhaustive nature of the literature review (at the date of
publication), the relevance of the critique of the articles upon which the opinion was based,
the degree of consistency of the different results observed in the literature – which implies
waiting for the replication of observations before pronouncing on their reality – and their
coherence (these results must be part of a logical sequence covering a chain of
mechanisms and/or be found in different animal species, including man). It should be noted
that criteria relating to the intrinsic quality of the studies published (experimental or
observation protocol, data analysis, and account taken of interfering factors, comparison
with pre-existing scientific data) are not on this list, as they have already been taken into
account, usually explicitly, in the summary reports on the literature that the expert group
analysed. These criteria were nonetheless applied to articles published too recently to be
included in the summary reports.

Hill, AB. The environment and disease: association or causation. Proceedings of the Royal
Society of Medicine, 1965, 58: 295-300

2 Radio Frequency Electromagnetic Fields and Health:
Elements of Physics and Biology

2.1 Introduction
Mobile telephony has developed into a major component of modern society in the last few
years. It uses electromagnetic fields in a well-defined frequency range, referred to as the
radiofrequency (RF) range. Radiofrequency electromagnetic fields are also used in other
applications, including :

· domestic: microwave ovens,

· occupational: high frequency presses, welding, radiocommunications
· general public: control or identity badges, transport passes
· medical diagnostic (MRI: magnetic resonance imaging) and therapeutic (physiotherapy,
hyperthermia, etc.).

Evaluation of possible health effects of radiofrequency electromagnetic fields requires

knowledge of the physics of electromagnetic fields and their interaction with the body, as
well as technical standards and principles of operation of mobile telephony.

2.2 Basic physics of electromagnetic fields

Definition of an electromagnetic field
An electromagnetic field is an association of an electric field and a magnetic field that vary in
time and propagate together in space. These fields are capable of displacing electrical
charges. Electromagnetic fields can be characterised their physical properties and
particularly by frequency or wavelength, strength (intensity) or power density.
Frequency. The frequency of an electromagnetic field is the number of variations of the
field per second. It is expressed in the unit hertz (Hz) or cycles per second, and extends
from zero to infinity. A simplified classification of frequencies is presented below, and some
examples of applications in each range are indicated.

X and gamma rays can break molecular bonds and cause ionisation, which can lead to
carcinogenisis. Ultraviolet, visible and infrared radiation can change the energy levels of
bonds within molecules. Radio frequencies do not have sufficient energy to disrupt
molecular bonds.

Frequency Range Application
0 Hz Static fields Static electricity
50 Hz Extremely low Electric power
frequencies (ELF) transmission and
domestic power supply

20 kHz Intermediate Video display units,

frequencies induction heaters

88-107 MHz Radiofrequencies FM Broadcasting

300 MHz-3 GHz Microwave Mobile telephony

400-800 MHz Analogue telephony
(Radiocom 2000),
900 MHz and 1800 GSM (European
MHz standard)
1900 MHz – 2.2 GHz UMTS (standard for
enhanced telephony
services including
mobile internet)
3-100 GHz Radars
Intruder detectors,
102 – 105 GHz Infra-red remote controls
105 – 106 GHz Visible Light, lasers
(0.8 - 0.4 mm)
0.4 – 10-1 mm Ultra violet Sun, phototherapy
10-1 – 10-2 mm X rays Radiology
10-2 mm and less Gamma rays Nuclear physics

Intensity and Power

Field strength (intensity) can be expressed in different units:

· the electric field is expressed in the unit volt per metre (V/m)

· the magnetic field is expressed in the unit ampere per metre (A/m) or the tesla (T), 1 A/m
= 1.27 µT

· the electromagnetic field strength can also be expressed as an equivalent power density
(PD, in W/m2 ). The PD is proportional to the product of the electric and magnetic field

PD = E x H = E 2 / 377 = 377 x H2, or: E = √377x PD

· The total power contained in an electromagnetic field can also be expressed in watts

Other properties:
· Polarisation: orientation of the electric field vector in the electromagnetic wave
· Modulation:
amplitude (AM),
frequency (FM),
pulsed wave (PW),
no-modulation = continuous wave (CW)
When the emission is modulated, one must differentiate between the peak power
(maximal power) and the average power. For example, in a pulsed radar emission
with a 1 ms pulse every second, the average power is one thousand time less than
the peak power.
· Field uniformity

Interaction mechanism of RF with biological systems

All living matter contains electrical charges (ions, molecules…) and insulating materials; it is
therefore a weakly conducting medium (called dielectric). When the tissue is subjected to a
RF field, part of the field is reflected, and part penetrates the tissue. The exposure of the
body produced by this interaction must be quantified, because it may lead to biological
effects. Many factors can influence the interaction:

1 Physical transmission parameters

· frequency
· incident power (peak or average)
· polarisation
· modulation
· uniformity of field
· proximity to the transmitter
· dimension and nature of the exposure chamber

2 Physical parameters of biological systems

· the dielectric properties of tissues
· size, shape, position and orientation of the exposed biological system, particularly
with respect to the wavelength, allowing differentiation between local and whole
body exposures.
· spatial distribution of exposed organisms

3 Environmental factors
· temperature
· humidity

Other factors that affect the resulting biological effects

1 Variables related to the biological investigation:

· measurement techniques
· response base line
· metabolic and functional disorders
· genetic predisposition

2 Experimental variables
· acclimatisation procedures
· animal – researcher relationship
· condition and confinement of animals (anaesthesia, stress,…)
· exposure interval during the day
· duration of exposure
· number of exposures
· time elapsed between exposure and measurement

In tissues, the electric field can displace free charges like ions, or orient polar molecules like
amino acids. This induces reaction forces in the exposed medium that are proportional to
the viscosity of the medium. Part of the electromagnetic energy is therefore transformed into
heat (thermal effect).

The field that penetrates the tissues can be calculated using electromagnetic models. The
validity of calculations can be confirmed by field measurements in “phantoms” containing a
tissue-equivalent material. The dose of energy absorbed by transformation into heat is
quantified by the power absorbed per unit mass of exposed biological material. It is termed
the specific absorption rate (SAR) and is expressed in W/Kg (cf. dosimetry annex). The field
level corresponding to a given SAR can be calculated when a well-characterised body is
exposed in the “far field”, a long distance from the source. This method is pertinent to the
characterisation of the exposure of the public to fields from base stations but not easy to
apply when the biological system is exposed close to the source (“near field” exposure),
which is the case when using a mobile telephone. The estimation of the absorbed power
necessitates the use of complex modelling methods.

2.3 Biological effects of radiofrequency electromagnetic fields.

2.3.1 Thermal effects

At radiofrequencies, high intensity fields, like those emitted by radar, induce well-defined
thermal effects.

Accidental overexposures leading to excessive heating. Accidental overexposure is

rare but accidents do happen instigated by, for example, the unexpected powering up of a
source or a deficiency in a safety system. The release of heat following such exposures
would often trigger a reflex defensive mechanism resulting in the withdrawal of the exposed
part of the body away from the harmful fields. Nevertheless there are cases where the
withdrawal was not possible and people suffered burns. In cases of exposure to the head,
headaches were reported.

Experiments in man: triggering of thermoregulation during exertion. The basal

metabolic rate of a man at rest is on the order of 1.5 W/kg. During moderate effort it
increases to 2.5 W/kg. A localised exposure to a SAR of 8 W/kg for 45 minutes for subject
at rest instigates physiological thermoregulation manifested by an increased bloodflow to the
skin without sweating, while maintaining a stable central temperature. When the exposure is

preceded by a moderate effort, the same exposure provokes increased bloodflow to the skin
with sweating [Adair 2000]. The effects of a given exposure thus depend on the metabolic
state of the individual.

Animal experiments: permeability of the blood-brain barrier

Experiments have been conducted at various levels of thermal exposure using several
different molecular radioisotope tracers. These experiments have shown, on the one hand,
an increase in the cerebral blood volume, and, on the other hand, a change in the uptake
depending on the nature of the molecule. These two phenomena are related to the increase
of body temperature. The authors have concluded that the increase of permeability of the
blood-brain barrier under the influence of microwaves is simply related to the temperature
variation in the brain [Barenski, 1973:Lin and Lin, 1980].

Animal experiments: cataract. Animal studies have shown that a cataract can be
provoked by an exposure of one hour at a surface power density of 100 to 150 mW/cm2
[Williams et al., 1955. Carpentier, 1960; Zaret et al., 1975]. Such high intensities are never
observed in the environment, of the general public and rarely in occupational exposure.
Even during an overexposure, not one case of cataract has been noted in humans.

Remark on thermostatation7 in vitro. In some studies, the heating of the exposed sample
is prevented by a cooling system. There is therefore no measurable increase in
temperature. Nevertheless, the energy absorbed in the exposed medium may be
considerable and it cannot be said that the exposure in these studies was “of weak intensity”
[Maes et al, 1993].

2.3.2 Non-thermal effects

Microwave auditory effect. A specific effect of radio frequencies is the auditory perception
of microwave emitted by radars or “microwave hearing”. When the microwaves arrive at the
skull, the energy absorbed is converted into heat and produces a very fast but weak (10-6 ºC
in 10 µs) increase in temperature. The thermal gradient generates a thermoelastic pressure
wave in the brain tissues, which propagate up to the cochlea where it is detected by the cells
of the inner ear [Rissman and Cain, 1975; Cain and Rissman, 1978; Chou and Guy 1979;
Lin 1981; Chou et al.,. 1985]. This is therefore a “micro-thermal effect” characterised by a
weak average power that does not lead to an increase in the overall temperature of a tissue.
With mobile telephones, the energy in the pulses is too weak to instigate a hearing effect.

Indirect effects: induced currents from touching a metallic structure exposed to an

electric field. When an large metallic structure is exposed to a radiofrequency field, it can
become charged and produce electric discharge currents when a person touches it; this
could result in painful, disagreeable sensations and accidents related to the inability to let-go
of objects. A person who touches a vehicle exposed to a field of 200 V/m in the frequency
range 10 kHz to 300 MHz (radar) would feel such discharges [Chatterjee, 1986].

Interactions with cardiac pacemakers. The interference of radiofrequencies with cardiac

pacemakers is theoretically possible. It is a matter of electromagnetic compatibility8,
complicated by a biological environment, the patient’s body [Gagny 1994]. Numerous
experiments have been carried out with mobile telephones. No effects have been observed
when the radiotelephones are held at more than 10 cm from the devices. At smaller
distances, minor perturbations of the ECG have been observed with some type of
telephones. New models of pacemakers are currently equipped with electronic filters
making them immune to fields from telephones.

Other non-thermal effects. Other non-thermal effects have been reported [Thuery, 1989;
de Seze & Veyret, 1996]. They are at the heart of the debate on health effects of RF. The
analysis of scientific reviews, reports and recent literature in this field, in Chapter IV of this

report, is mainly concerned with these biological effects, which are only mentioned here for
the record.

2.4 Mobile telephony: technical specifications

Principles of mobile telephony
The handset transforms voice into radiofrequency fields that propagate from the handset
antenna to a relay antenna (base station). The signal is then transmitted through the wired
telephone network to the destination. Each relay antenna or base station covers a limited
area or “cell”, which is where the name cellular telephony comes from.

GSM System (Global System for Mobile communications)

The characteristics of fields used in mobile telephony are specified in standards for different
operating systems used in different regions and in different countries. In France, the two
systems currently in operation are the GSM 900, implemented by the operators Itineris and
SFR, and the more recent GSM 1800 system implemented by the operator Bouygues

In the GSM 900 system, the carrier frequency is in the 900 MHz range. It extends from 872
to 960 MHz. In the GSM 1800 system, the carrier frequency is in the 1800 MHz range. It
extends from 1710 to 1875 MHz.
Within these ranges, the base station assigns to each user a narrow band of 0.2 MHz for
each conversation (frequency division). This allocation is not fixed and can change when
the user is moving as the call is relayed from one cell to another. Within each 0.2 MHz user
band, there is a time division whereby the information is transmitted in pulses, at the rate of
one 576 µs pulse every 4.6 ms (pulse repetition frequency of 217 Hz; duty cycle of 1/8).
This allows each narrow frequency band to be utilised by 8 different users simultaneously.
The conversation is then “reconstituted”, after decoding, all in sufficiently short time for it to
appear as continuous.

Commercial mobile telephone handsets operate as GSM 900, GSM 1800, or both (dual
band telephones). Some are even compatible with the North American system (tri-band).
The various models differ in performance, size and weight. The use of the weakest possible
signal for communication gives the best possible performance and makes possible a
smaller, lighter battery; this leads to minimum emissions from the telephone.
The highest possible power output from handsets is 2 W for GSM 900 and 1 W for the GSM
1800 system. The pulsed emission means that the average power is 8 times less, i.e. 0.25
W for the GSM 900 and 0.125 W for the GSM 1800 system. The power output is further
regulated in relation to the proximity to the base station; it is inversely proportional to the
quality of the link (250 mW several kilometres from the base station antenna, 10 mW in
close proximity).

The field emitted by a handset operating at maximum power is of the order of 400 V/m at a
distance of 2-3 cm. It decreases very rapidly with distance.
Interaction between handset and user.

The power absorbed in the head is about 40% of the total power output of the handset
(100mW maximum for the GSM 900). For 3 kg head, the corresponding average SAR is of
the order of 30 mW/Kg. However, since the power absorbed decreases exponentially as
function of depth, the local SAR is the more important factor as it is calculated in a small
volume. For a 10 g mass it ranges from 0.4 to 1 W/kg.

Two technical constraints tend to reduce the SAR in the head of users:

· good communication requires that most of the radiation emitted by the handset be in the
direction of the base station, which implies that the part absorbed by the head of the user
be minimal;

· handsets can use adaptive power control to optimise their output

The purpose of power control is to reduce interference between users in the same cell, and
to allow the reduction of the volume of the battery due to reduced power consumption.
When a mobile user is first connected to the network, the power emitted is set to a high level
for optimal immediate communication, and then the power control reduces the level in steps
of 2 dB in a few seconds, until it stabilises at the minimum level compatible with a good
quality of communication. Therefore the field to which the user is exposed varies at any
location, as a function of time (on a scale from 20 to 30 seconds). The displacement of the
user (while walking or in a moving vehicle for example) instigates the pick-up of the relay by
several successive base stations, each of them starting its communication at a higher level,
than lowering the power. Therefore, the exposure profile is a series of different exposure
levels varying from a few hundredths to the order of 0.5 to 1 W/kg. It is therefore during the
use of a mobile while moving that the exposure is the highest, or during a conversation in a
location with mediocre reception, which compels the base station and the handset to
maintain high power levels.

Base Stations
Different types of relay antennas or base stations are in use, depending on the area covered
and the number of calls made:

· macrocell installations: the most common of this type can emit a maximum power of 20
to 30 watts per frequency band. In rural settings the power will be elevated to cover
extended areas (10 - 30 km) with a limited number of user frequency bands, whereas in
urban areas the power will be distributed among several user bands within a smaller
area (500 m).

· microcells have less power and are used to cover limited areas with high user density
such as train stations or shopping centres, for example.

· picocells are installed in the interior of buildings such as offices.

The field in the vicinity of antennas is distributed as follows:

· The field strength is 50 V/m at a distance of 1 metre from a microcell installation,

immediately in front of the antenna. The reference levels set by the 1999/519/CE
European recommendation are 41 V/m at 900 MHz and 58 V/m at 1800 MHz. To ensure
that these levels are met a minimum access distance on the order of 1.5 m should be
maintained for microcell antennas; for macrocell antennas the appropriate distance is
2.5 m. The results of field measurement are presented elsewhere in this report (cf II-

· Behind the antenna, a metallic plate reflects completely the fields emitted in this
direction. A 50 cm distance is however recommended in order to ensure compliance
with the recommended levels.

· Away from the axis of the antenna, either above or below (antennas are usually mounted
at a height of about 20 m) the field is of the order of 1 to 2 V/m at most. The beam is
directional, slightly tilted downwards, with a large horizontal span of the order of 120°

and a narrow vertical range of a few degrees. It reaches ground level some 20 to 200
metres away from the base station depending on the height and tilt of the antenna (see
the figure below).



Figure 1 Main beam from a macrocell base station antenna

New frequency utilisation: emerging and future technologies

The GSM protocol was introduced in 1992; it is a TDMA signal (time division multiple
access). The repetition frequency is 217 Hz. The duty factor is 1/8 for the mobile phone
and varies from 1/8 to 8/8 for the base station (except for the control channel (BCCH) where
the duty factor is always 8/8).

HSCSD (High Speed Circuit Switched Data), allowing data rates of 38.4 kbps (email, fax,
etc.) and GPRS (General Packed Radio Service) allowing data rates of 115 kbps result
from improvements to the second-generation protocol.

UMTS (Universal Mobile Telecommunications System)

UMTS is the European version of the universal IMT-2000 (International mobile

Telecommunications-2000) protocol. This protocol allows the introduction of important new
services (multimedia, etc.). It is expected that three types of cells will be used (macro, micro
and pico).
Type of cell Macro Micro Pico
Beam <20 km < 1000m <100 m
Antenna Rooftops Sides of Buildings Ceilings, Walls
Applications Countryside High call density Buildings,
Urban areas City centres
Service Limited All services
Data rate <144 kb/s <384 kb/s <2 Mb/s

Two complementary protocols are used: W-CDMA and TD-CDMA (Wide-band or Time
Division Code Division Multiple Access), in the frequency bands 1900-1920 MHz and 2010-

2025 MHz respectively. The power used is up to 50 W for the base station and 2 W
maximum for the mobile phone (in practice much less due to power control).

DECT (Digital Enhanced Cordless Telecommunications)

DECT cordless phones have been used since 1988 in homes and offices. The frequency is
between 1.88 and 1.9 GHz. The protocol is TDMA, as for GSM, the average powers are
250 mW for the base station and 10 mW for the handset, the range is approximately
300 m.

TETRA (Terrestrial Trunked Radio)

TETRA is the new private communication system developed in 1995 for the police,
ambulances, etc. It is a TDMA protocol with 4 channels (compared with 8 for GSM).
Frequency bands used in Europe are 380-383 MHz and 390-393 MHz for the emergency
services and up to 921 MHz for other applications. The power of the base station is 15 W
and that for the terminal is 1 W.

TFTS (Terrestrial Flight Telecommunications System)

The TFTS system allows communication between aircraft in flight and the terrestrial
telephone system. The frequencies used are 1800-1805 MHz for the downlink and 1670-
1675 MHz for the uplink. The power is 10 W approximately. The cell height is very large
(350 km).

Other RF applications

Numerous short-range applications are in use or under development:

· Remote control (cars, engines, toys, etc.)

· Wireless connections for HiFis or videos,
· Telemetry and identification of people, vehicles, etc.
· Proximity radar

Several protocols for the transfer of numerical data at short distances are under

Hiperlan/2 (High Performance Radio Local Area Networks)

A high data rate RF link (< 54 Mbps), allowing great flexibility of use (voice, data, video).
The frequency is within the bands 5150-5250 MHz and 17.10-17.30 GHz. A CDMA protocol
is used and the average power is 100 mW.

Bluetooth (from the name of a 10th century Viking king) is a short-range communication
system. The idea came from within a group founded by the companies Ericsson, IBM, Intel,
Nokia and Toshiba. The frequency used is in the band 2400-2483.5 MHz, the power is
typically 1 mW and the range is 10 m only (communications between appliances in the
house or between mobile phone and ear piece). The data rate can reach 1 Mbps.

Progressively, communications based on analogue technologies will be replaced by digital

systems. This will enable faster communications as well as more efficient use of the
spectrum. The power emitted can be reduced because of the reduced interference
associated the digital transmission. Taken together, the development of the new
technologies described above will lead to the proliferation of sources in our environment.
However, it is possible that background levels will not increase significantly even if an
increase can be expected initially as a result of the introduction of new technologies in
addition to those already existing. The contributions of multiple weak sources at close
proximity should be similar to that of higher power sources at a distance (radio and television

broadcasting), whose power will not decrease significantly even for digital transmission.
(20 kW).

2.5 Deployment of base stations and exposure of people

a- Field measurements in the vicinity of base stations

The expert group asked the three mobile telephone operators in France to provide data on
call traffic or the results of site surveys, to be included in their report (by letter of 11
November, 2000). The same request was made on 5 October, to the National Agency of
Frequencies (Agence Nationale de Fréquences). At the time of writing the report, only
Bouygues Telecom had submitted some of the data as requested (see later) and, to a more
limited extent, France Telecom Mobiles. The group of experts was surprised at this
response, which does not allow it to present the expected information to the public. The
National Agency of Frequencies (Agence Nationale de Fréquences) has indicated to the
group of experts, by correspondence to its Director General on November 13, 2000, that, in
the absence of national measurement protocol to enable the undertaking of well-defined
surveys, the data available to the agency cannot be considered representative. The Agency
set for itself the objective of developing such a measurement protocol, covering the whole
radio frequency range, to provide a basis for a campaign of measurements at number of
typical sites that are representative of the situation in our country. The company Cégétel
provided some data obtained on 2 sites by a technical control body and by the ANFR. The
two sets of data were obtained using completely different protocols: in one case with a low
sensitivity broadband probe, and in the other case with a selected narrow band antenna.
These different methodologies do not allow a comparison to be made.

However, different organisations in other European countries have carried out surveys and
published data in scientific papers or reports. Data from these sources are summarised in
this section.

Cartography of typical GSM 900 sites, communicated by France Telecom Mobiles

France Telecom Mobiles delivered to the expert group a copy of a simulation study of typical
GSM-900 sites. The study took into consideration local cartography and included absorption
by walls, but not reflection. Irrespective of this, according to antenna manufacturers,
disparities associated with side and secondary lobes mean that such simulations can
provide order of magnitude data only. The reported simulations were obtained assuming full
traffic on 4 channels without taking into consideration variation during the day.

For a 900 MHz macrocell antenna (Kathrein K736863) producing a beam with a vertical
width of 8° and a horizontal width of 90°, the safety distance corresponding to an exposure
limit of 41 V/m is 2 m in front of the antenna, 20 cm behind the antenna as well as above
and underneath, and 1 m at the side of the antenna. The simulation of the antenna on a 23
m mast gives the following results: 15 V/m at 10 m in front of the antenna, 7 V/m at 20 m in
front of the antenna, 3 V/m at 50 m in front of the antenna, 2.25 V/m at 5 m under the
antenna at a horizontal distance of 20 m from the foot of the mast and 1 V/m at 10 m under
the antenna at a horizontal distance of 20 m from the mast.

For an antenna 1 m from the edge of a roof covered with cement, simulations show a
maximum value of 0.5 V/m at 2 m below the antenna (in the storey below). For a wall-
mounted antenna, the field strength is 15 V/m at 1 m from the sides of the antenna and
1.5 V/m at 1 m behind the antenna, taking into consideration a 10 dB absorption by the wall.

For wall mounted microcell antennas (Kathrein K736350), the safety distance corresponding
to the limit of 41 V/m is 10 cm around the antenna as well as above and below. The field
values of 3 V/m at a distance of 1 m behind the antenna, 10 V/m at 1m on the side of the

antenna and 1.5V/m at 15m in front of the antenna. For picocellular sites, the safety
distance defined by the same conditions is 10 cm around the antenna and 5 cm above and
below the antenna.

Measurements have been made on-site by France Telecom; the results are dependent on
the measurement instrumentation (sensitivity, isotropy, frequency range). Using an isotropic
frequency selective probe from Melop Thomson, measurements were made in 9 busy public
places in Paris. The maximum value of power density measured ranges from 0.72 to 0.0056
mW/m2 in the GSM 900 band and from 0.13 to 0.018 mW/m2 in the DCS 1800 band. The
sampling was repeated with the inclusion of power densities in the frequency range 85 to
1900 MHz covering, in addition to the base stations, FM radio and television broadcasting;
the power densities range from 19 to 1.2 mW/m2; the highest value, obtained close to the
Eiffel Tower, corresponded to a field strength of 2.7 V/m. On the basis of these data, mobile
base stations contribute only a very small part of the total power density from all radio
frequency sources. In fact the ratio between the maximum power density in the GSM 900
band and the power density in the band 85 to 1900 MHz ranges from a maximum 0.142 at
Notre Dame to a minimum of 0.001 at Montmartre (14% to 0.1%). In the DCS 1800 band
the ratio is less variable it extends from 0.032 at Place de la Concorde to 0.0086 at
Montmartre (from 3.2% to 0.86%).

France Telecom Mobiles provided 3 reports on the field measurements made by technical
control bodies on rooftops or in apartments. These measurements were made with isotropic
probe of low sensitivity and very wide frequency range (Wandel Golterman or Chauvin
Arnoux); these cannot be considered typical of base stations only but of the whole radio
frequency spectrum.

Field survey in Parisian schools that in the neighbourhood of a base station (data
submitted by Bouygues Telecom).
At the instigation of Bouygues Telecom, the ETDE (a technical control body) carried out a
survey of electric field strength in Parisian schools near Bouygues Telecom base stations.
The results of this study have not yet been published, but were presented at a congress in
December 2000. They are intended for the information of the Directorate of Scholastic
Affairs, Paris and the Directorate General of Health.

A list of 100 nurseries and elementary schools, near a Bouygues telecom base station was
established from the 338 nurseries and 335 elementary schools in inner Paris, of which 69
were visited in the month of August 2000. For each school, 3 measurement locations were
identified: the centre of the playground, a classroom in the middle of the building and the
entrance hall. In the absence of a harmonised measurement standard at the national or
European level, the operators made use of the latest drafts of standards under development.
The electric field was measured in narrow bands in the FM, TV, GSM-900 and GSM-1800
bands, using a spectrum analyser and antennas specific for each frequency band. The
results were given in V/m and as percentages of the reference level specified in the
European recommendation of July 12, 1999.

The distance between the nearest GSM 1800 base station and the school ranged from 30 to
372 m; the distance to the nearest GSM 900 base station and to other RF sources was not
reported in this study. The maximum call traffic in the GSM 900 and GSM 1800 bands was
obtained by extrapolation from data in the service channel, taking into consideration the
number of channels per antenna (usually 4 in Paris). Signals from FM sources and
television broadcast were combined to determine the total signal level in these two bands.

The results were expressed as a percentage of the reference field strength: in the FM band,
the values were between 1/100,000 and a few percent of the reference level; for the TV
band and the GSM 900 band the values were between 1/100,000 and 1/1000 of the
reference level; for the GSM 1800 band, the values were between 1/1,000,000 and 1/10,000
of the reference level (see the figure below). The reference level of 58 V/m for the GSM
1800 band is higher than the 41 V/m in the GSM 900 band: this could explain in part the
observed differences as the result is expressed in percentage of the reference level. The
scatter of points in each band makes it difficult to differentiate between the field strength in
the classrooms, the playgrounds or in the entrance halls, except in the FM band where the
values appear systematically higher in the playgrounds. This observation warrants a closer
statistical analysis. The relationship between the distance of the nearest base station and
the field strength was studied for each site in the playgrounds, the halls and the classrooms.
In the playgrounds, the results agrees with the theoretical model: between 30 and 100
metres, the field is weak and steady, on average between 0.005 V/m, with the exception of
one point situated 70 m from a base station, where the field strength was 0.003 V/m,
probably due to the presence of a side lobe. Then between 100 and 150 m, the average
field strength increased, reaching 0.002 V/m; also observed is a large scatter in the data
points probably due to the presence of obstacles and reflectors. Thereafter, the field
decreases and reaches a low value of 0.0005 V/m at around 200m. In the classrooms, the
distribution of field is a little different: irrespective of the distance to the nearest base station
the field strength is below 0.0005 V/m. As previously observed at 70m there were a few
high values of 0.004 V m-1, and between 70 and 175 metres there were a few points above
0.001 V/m. In the halls the measured values were low because of absorption by the walls
and more-or-less constant, on the order of 0.0002 V/m irrespective of the distance, with a
few points (6) between 0.0006 and 0.001V/m at distances between 70 and 120 metres.

Electric field strength (% of reference level)

Total FM field playground (%) Total TV field playground (%) Total 900 field playground (%) Total 1800 field playground (%)
Total FM field hall (%) Total TV field hall (%) Total 900 field hall (%) Total 1800 field hall (%)
Total FM field classroom (%) Total TV field classroom (%) Total 900 field classroom (%) Total 1800 field classroom (%)

Figure 2 Results of field measurements in Parisian schools; comparison of

different RF fields
Source: Bouygues Telecom, August 2000

3/ Conclusions
This study was carried out over a relatively large number of sites; at each site three data
points were taken in places were different types of activity take place and the measurement
procedure was correct. It can be concluded from the data available that the exposure level
in Parisian schools is higher in the FM band than in the GSM bands and that the exposure in
the TV band is of the same order of magnitude as in the GSM bands. The relationship
between the distance and the average field strength in the playground agrees with
established models, it is weak and fairly constant at short distance, it increases with distance
from about 100 metres and decreases again from about 150 metres. Measured field
strength from the nearest base stations appeared weak in the schools, the average value is
below 0.001 V/m at the classrooms as well as in the playgrounds; it is below 0.0005 V/m in
the halls.

Analysis of NRPB report: Exposure to radio waves near mobile phone base stations
NRPB-R 321.
In the UK, the NRPB, National Radiological Protection Board, is a public body in charge of
radiation protection, including non-ionising radiation. The Stewart report, published in May
2000, made recommendations with respect to base stations, in particular the need for
independent audit of sites in order to ensure that the exposure limits are not exceeded in
the proximity of base stations outside the designated safety (exclusion) zone. It is within this
framework that the NRPB undertook a programme of measurements at 118 base station
sites in neighbourhoods where the level of emission had been questioned. This study was
the subject of the report NRPB R321, published in June 2000.

1/ Summary of methodology

a/ Technical aspects
There is a description in the report of the basic concepts of cellular telephony in which base
stations, at the centre of cells, with uplinks from the mobile handset to the base station and
downlinks from the base station to the mobile. The cells have variable sizes depending on
the geographic characteristics of the terrain, ranging from tens of kilometres in rural areas to
one kilometre or less in towns. For macrocells, coverage is provided by antennas located
on rooftops or on a mast. Short distances of the order of 100 metres or less are covered by
means of microcells, or picocells for the very short range. The report specifies the technical
characteristics of antennas and in particular the directional characteristics of the beam. The
horizontal beamwidth is large, on the order of 120°, and the vertical beam width is just a
few degrees. The beam is slightly tilted with respect to the horizontal; it reaches ground
level at between 50 and 300 m depending on the height of the installation and its tilt with
respect to the horizontal. Each sector of the antenna emits a maximum power on the order
of 10 watts. The study reports only on macrocells were the maximum emitted power in the
United Kingdom ranges from 25 to 70 Watts depending on the number of sectors. According
to the authors of the report, microcell antennas emit maximum powers on the order of 1 W
while in the case of picocells the power level is on the order of 0.1 or 0.2 W. The gain of
microcell or picocell antennas is low which means that the beam is less directive and of
lower intensity. The corresponding safety zone does not exceed a few tens of centimetres.

b/National and international recommendations.

The report summarises the different exposure levels recommended for the general public
and for workers. In this frequency range, the UK maximum exposure levels (NRPB
guidelines) expressed as power density are 33.2 Watts/m2 at 900 MHz and 100 Watts/m2 at
1800 MHz. The exposure limits reported in the European recommendations of 12 July
1999, itself based on the recommendation of the International Commission on Non-Ionizing
Radiation Protection (ICNIRP) for the general public, are 4.5 Watts/m2 at 900 MHz and
9 W/m2 at 1800 MHz.

c/ Instrumentation
The authors report that hand held isotropic probes designed for measuring exposure level
comparable to the recommended limits are not appropriate for this type of measurement
because of their lack of sensitivity. The measurement were therefore made with a spectrum
analyser and antennas specific for each frequency range, allowing a detection threshold of
less than 1 µW/m2 and a good resolution. The measurements uncertainties were evaluated
and tend to be quite significant, on the order of 3 dB, even with this methodology.

2/ Modelling
In order to verify the validity of their measurements the authors compared their data with
predictions from well-established mathematical models. They used two models to calculate
the distance corresponding the limit of exposure in front of an 80 watt, 12 element antenna.
Using the inverse square law as the basis of calculation, the limit of exposure in the
European recommendation of July 12, 1999 is met at 8 metres in front of a 900 MHz
antenna or 6 metres in front of an 1800 MHz antenna. According to a more realistic model
for near field calculation, the exposure limit is met at 2.5 metres in front of a 900 MHz
antenna and at 5.5 metres at 1800 MHz. There were great discrepancies between actual
measurements and predicted values. The difference is more significant near the antenna as
the model can overestimate the actual value of power density by up to 4 orders of magnitude
where the measurement point is not directly within the main beam or when an obstacle, such
as a wall, act as a screen. Models for power density at ground level for a 15 metre high
antenna predict a very low power density under the mast and up to 10 m around. It
increases gradually between 10 and 100 m (the increase is not linear but shows peaks and
trough in view of the presence of side lobes). The maximum power density, taking into
consideration reflections, is reached around 180 m from the mast, beyond which power
decreases according to the inverse square law, and at 300 m it is slightly higher than the
value estimated at 70 m. The lower edge of the main beam reaches ground level in this
model configuration at a distance of about 100 m

3/ Measurements
Measurements were made at each site in the frequency range 30 MHz to 2.9 GHz in order
to cover the whole spectrum of emissions from FM radio and television broadcast and also
telecommunications. It was possible to identify, at each measurement point, the signal
emitted by the nearest base station, other base stations and radio signals in the frequency
range under consideration. Measurements were made inside and outside buildings.

The power density corresponding to the nearest base station varies considerably,
particularly at very short distances. The values obtained outside the buildings are on the
order of 10 µW/m2 to 1 mW/m2, which is about one thousandth or one millionth of the values
of the European recommendation. The power densities measured inside buildings are
generally lower and more variable than the values obtained outside: they range from 0.1
µW/m2 to 1 mW/m2. The average power density tends to increase between 0 and 100
metres from the antenna and then to remain relatively constant at about 0.1 mW/m2. The
reason is that, as one gets further from the antenna, one gets progressively nearer to the
main beam and then enters it completely. At short distances, the scatter in the measured
data is consistent with the presence of obstacles between the beam and the measurement
point and that the measurement points vary with respect to the position of the main beam.

The range of power densities corresponding to all radiofrequency sources is less than that
for the main base station alone. If the power density from the nearest base station alone is
higher than 0.1 mW/m2 the addition of all other sources has very little effect on the total
value, while if the power density from the base station is less than 0.01 mW/m2 the
contribution of the base station to the total power density is insignificant. On the basis of
geometric averages, the power density from nearest base stations is little different from the
power density from all other sources: 33 µW/m2 against 21 µW/m2.

The report shows that the total power density at a measurement point is between one
thousandth and one ten millionth of the allowed levels in the European recommendation.
As predicted, there is a large scatter in the data near the antenna and a more or less steady
value of the order of 1/10,000 from 100m onwards, which probably means that the power
density from the base station cannot be distinguished from the background power density at
this distance. The ratio of the signal from the nearest base station to the signal from all
other sources is quite variable at short distances; it ranges from 0.001 to 100. From 100m
onwards it steadies to a value on the order of 0.1.

The power density from the antenna may exceed the reference level at a short distance in
front of the antenna in the direction of the main beam. Because of the directionality of the
antennas, and as by predicted mathematical models, the exposure limits are not exceeded
beyond at distances greater than a few metres in the direction of the main beam. In general,
the measurements do not show a decrease in the power density with distance. In the sites
considered in the study, the power density from all sources (other than the nearest base
station) is of the same order of magnitude as that from the nearest base station. The
geometric mean of power density, from all sources, at distances of 5 to 230 m, are of the
order of 18 millionth of the values recommended by ICNIRP. Excluding the power density
from the nearest base station, the geometric mean of power density from all other sources is
less than 5.5 millionths of the values recommended by ICNIRP; in both cases there is
scatter in the measured data.

Summary of the article: Electromagnetic field pattern in the environment of GSM

base stations (published in the Occupational Medicine and Environmental Health, vol.
12 47-58 1999)
This article originates from the work carried out in the electromagnetic hazard Department of
the institute of occupational medicine in Lodz, Poland.

1/Polish regulations
The Polish regulations specify different levels of exposure for the general public and
workers. For the public, in the frequency range 300 MHz to 300 GHz the maximum
permissible power density is 0.1 W/m2 (6.14 V/m). For workers there are three classes of
risk, necessitating 3 different levels of protection: > 100 W/ m2, 20 to 100 W/m2 and 0.1 to
20 W/m2. The last two categories correspond to exposure levels observed in the work place.

2/ Mobile telephone networks in Poland

The situation in Poland is not significantly different from that in France. The power rating of
base station antennas is on the order of 10 to 20 watts; these antennas are mounted on the
roof tops of residential buildings, hospitals and, particularly in rural areas, on masts. Specific
to Poland is the frequent use of factory chimneys for the installation of these antennas. The
antennas are of the usual type, with about 35 different models; the tilt of the panel with
respect to the vertical plane is about 3 to 9°. The nominal power is about 20 watts and the
power loss is on the order of 3 dB.

3/ Summary of measurements and simulations

The measurement procedures are not described in the article which refers to a Polish
standard and to a certification procedure cited as a reference. For an antenna with a panel
tilt of around 15 degrees on a 4.8 m rooftop mast, the reference level of 0.1 W/m2 is met for
a 1.8 m tall person on the rooftop at a distance of approximately 8.8 m from the foot of the
mast in the direction of the main beam. For a smaller tilt angle, the reference level is not
exceeded any distance from the mast. If the height of the mast is 3 m only, the reference
level is exceeded at a distance of 3.8 m from the foot of the mast if the tilt is 15 degrees and
at a distance of 11.3 m for a tilt of 3 degrees. This is consistent with the very directional

characteristic of these antennas in the vertical plane; it also demonstrates the necessity to
place them near the edge of the rooftops, which is not common practice in Poland.
Measurements were carried out at 20 sites near base stations by public authorities. For
base stations on the roof of residential buildings, on masts with heights between 1.65 m and
8.5 m and panel tilts of 0° to 9°, the reference level is exceeded at any point on the rooftop
only in 20% of cases. The measured values range from 0.025 to 2 W/m2. The power
densities in the upper floors of the building were below the measurement threshold. In
neighbouring buildings, the measured power densities were of the order of a quarter of the
Polish reference level. For antennas on towers or factory chimneys, the power density in the
immediate surroundings was below the measurement threshold. On towers or chimneys in
regions accessible only to maintenance personnel, power densities on the order of 0.66 to
1.7 W/m2 were observed in the vicinity of antennas. These values are higher than the Polish
reference levels but significantly lower than the reference level of the European
recommendation of July 12, 1999.

4/ Conclusions
The power density reference levels in Poland are 45 to 95 times lower than those in the
European recommendation. The Polish reference levels are exceeded on rooftops with
base station antennas in only 20% of cases, and in 5% of the locations where
measurements were made on masts and very close to chimney-mounted antennas. The
Polish reference levels were not exceeded in any public place in neighbouring residential
buildings or in buildings with base stations. Maintenance personnel may be exposed to
levels above the reference level. Moreover, in order that the Polish reference levels are not
exceeded on rooftops, the microwave data links between base stations must be installed at
a minimum height of 2 metres for a 23 GHz link and at a minimum height of 3 metres for a
7 GHz link.

Summary of the presentation. "Exposure close to base stations in Austria"

By G. Neubauer, Austrian Research Centre Seibersdorf A 2444 Seibersdorf Austria

This is a study carried out by ARCS; the results have been presented at a conference but
have not yet been published in the scientific literature. The study was conducted as part of a
European project under the umbrella of COST 244 bis. The aim of this project is to assess
the extent of compliance with the reference levels at various sites near base stations in
Austria, Belgium, Hungary and Sweden. The results will be published in a COST 244 bis

1/ Methodology
In 1999, there were about 8500 base stations in Austria, operating at 900 and 1800 MHz.
Narrow-band field strength measurements were carried out by technical control bodies at
202 sites close to base stations, of which 100 were in urban areas and 102 in rural areas.
The measurements were made with a spectrum analyser and antennas specific to each
frequency band; broadband measurements were also made. Distance and area plans are
not reported in the presentation.

2/ Results
The results, expressed as power density, show a maximum value, measured in the narrow
frequency band of the nearest base station, of 13.4 mW/m2 (2.24 V/m) or 0.28% of the
power density reference level in the European recommendation. The maximum broadband
power density is 66.3 mW/m2 (5 V/m); in this case the measured power density is mostly due
to other sources than the nearest base station.

Overall, 8 out of 202 measured narrow-band power densities were equal to or greater than
1 mW/m2 (0.6 V/m), 40 were between 0.1 and 1 mW/m2, 43 were between 0.01 and

0.1 mW/m2, 61 were between 0.01 and 0.001 mW/m2 (0.02 V/m) and 50 were below
0.001 mW/m2.

3/ Conclusions
These results are in agreement with similar studies involving narrow-band measurements
centred on the frequency of the nearest base station. Expressed as percentage of the
reference levels for power density, the measured values are of the order of one thousandth
to less than one millionth. Generally, the power density from sources other than the nearest
base station is of its same order of magnitude or even larger than that from the base station

Opinion of the expert group on these measurement surveys and their results.

Measurement methodology
The results of the different studies are in general agreement: measurements were made
using similar but not identical protocols, which can explain some of the differences in the
results. It is clear that it is necessary to develop, as a matter of urgency, harmonised
European measurement standards based on a well-established methodology. All the base
station measurements (published in scientific articles) were narrow-band measurements
made with dedicated antennas. This methodology is quite appropriate in the absence of
isotropic probes that are also frequency selective and highly sensitive, because of the very
low field strengths measured in areas accessible to the public and in view of the fact that
other sources of radio frequencies can produce field strengths greater than those from the
nearest base station. This is understandable when we recall, for example, that a 1 megawatt
television transmitter is, in terms of radiated power, about twice as powerful as all the mobile
telephone base stations in France. Measurement using broadband isotropic probes is not
recommended, as such measurements are not representative of the base station alone but
of the totality of the radio frequency spectrum. Also, measurements of small field strengths
made with such probes have high uncertainties. These measurement methods could be
used in an anechoic chamber for the certification of base stations.

Evaluation of maximum power

Base stations have a variable emission level and can use than one communication channel.
(The power is constant only in the control channel). Assessment of the maximum power
from an antenna can be made in a rigorous manner either from the field due to the control
channel multiplied by the square root of the number of channels or by maintaining the base
station at maximum power during the measurement. Any other harmonised methodology
may also be used: measurement standards must define practical procedures to determine
maximum power so that a realistic exposure assessment can be undertaken

Compliance with the reference levels close to antennas

Very close to an antenna and in its main beam, measured field values can be in excess of
the reference level in the European recommendation of 12 July 1999. In view of the highly
directional nature of these antennas in the vertical plane, the distance at which the
reference levels are exceeded depends on the tilt angle of the installation. Practically, this is
only of consequence on accessible rooftops and only when access is possible to the main
beam in front of the antenna. It is therefore necessary to define on a case-by-case basis,
when the situation requires it, an exclusion zone where one should not stay and where the
wearer of an active medical device (a cardiac pacemaker for example) should not go. The
way in which such exclusion zones are defined should be consistent with internationally
defined procedures. In order to insure that these restrictions are in place there should be
well-defined written installation procedures for antennas; adherence to them must be
mandatory. Such a document is being prepared under the aegis of the Scientific and
Technical Centre for Building and its publication is imminent.

Field strengths in areas accessible to the public, and effect of distance to the antenna
In places where people spend time (residential buildings, offices or educational
establishments) measured field strengths are very much lower than - on the order of one
thousandth or one millionth of - the reference levels given in the European recommendation.
Because of the shape of the beam, the field strength is very small directly below the antenna
and increases gradually as a function of distance from the antenna, reaching a maximum
value at a distance that depends on height of the antenna with respect of the measurement
point, on a tilt angle and on the presence of any buildings that may act as screens. The
measurements carried out by the Scientific and Technical Centre for Building show that a
sheet of reinforced concrete attenuates the beam by a factor of approximately 30 and that a
plaster partition attenuates by a factor of 5. It can be seen that buildings constitute a good
protection against radiation from antennas, particularly when the antenna is on top of the
building, or very close by, such that the building is not in the main beam. Therefore the idea
of siting antennas away from certain buildings (particularly schools) cannot be supported by
rational argument; the best protection for a building is to be directly underneath the antenna.
By contrast, the expert group sees an advantage in ensuring that the main beam of the
antenna is not oriented directly towards places, within 100 m of the antenna, where people
spend a considerable amount of time even if the field strengths are very low and cannot be
distinguished from the RF background, particularly in the radio and FM band. This is for two
reasons: firstly, to avoid directing the beam unnecessarily toward people nearby, and to help
achieve the least possible exposure of the whole population; and secondly for a common
sense reason: directing the beam toward buildings reduces the range of the antenna by
absorption or partial reflection. This proposal is perfectly compatible with the possibility of
the beam passing above these buildings even if the antenna is nearby.

2.6 Consideration of the question of reflection and amplification of

Certain people or associations have been circulating over the past few years, particularly in
France and in Belgium, the idea of a possible amplification of the field from base station by
metallic structures nearby. This assertion has no real scientific foundation.

Like any wave, RF electromagnetic fields can be reflected from certain surfaces. The
phenomenon of reflection applies also to optical and acoustic waves. The extent of reflection
of an RF field depends on the material and is strongest with metals. This effect can be
advantageous, for example when used to increase the local field strength where it is needed
in certain applications. This is the principal of operation of a parabolic antenna where the
radiation received over a large surface is directed to its focal point. There is conservation of
energy and no amplification of the signal. All parabolic surfaces can focus, to a greater or
lesser extent, a radiofrequency beam but for an optimum result the beam must be directed
toward a large surface of uniform curvature. Convex surfaces lead, by contrast, to a
divergence of the beam, a plain surface to a direct reflection with no effect on the
beamwidth. In practice, reflection phenomena lead to spatial inhomogeneities of the power
density that can be modelled, and so taken into consideration in the calculation of antenna
patterns. This was considered in the NRPB report on base stations and the conclusion was
that reflections could increase or decrease the local field strength by a factor of two at most.
These models are only valid in the main bean and at a distance of more than 10 metres from
the antenna: under other conditions they could significantly overestimate the phenomenon.
An amplification mechanism is quite different. When amplification occurs, the energy
emitted is more than the energy received. A theory by which metallic structures are capable
of amplifying radio frequency fields is not compatible with the basic principles in physics.

2.7 The question of co-location of base stations7
The question is whether, with respect of the exposure of people to radiation from base
stations (BS), it is preferable to have all the antennas of different operators on the same
mast or if, on the other hand, there is an advantage in not doing so.

The answer depends on the criteria chosen to define the exposure. Assuming, for the sake
of argument, a population that is uniformly distributed over an area. The exposure is higher
near the base station and decreases gradually with distance according to the laws of
electromagnetic propagation, although the power density distribution can be quite complex
in an urban site.

Irrespective of the directionality of emission, if the population is uniformly distributed then the
average exposure is the average of all signals emitted by the different operators (when all
the signals are not correlated, which is the case) irrespective of the location of antennas.
Under these conditions, the co-location of antenna or a more random distribution makes no

If, by contrast, the maximum exposure is used as exposure metric, there is an advantage in
a more random distribution.

2.8 Results of field measurements with hand-free kits

The only data obtained in France and made available to the expert group was provided by
Bouygues Telecom. The data were obtained from measurement on a phantom, the hands-
free kit was mounted in the normal use position while the mobile phone was operational;
different configurations were tested in an attempt to induce the maximum SAR in the head
(orientation of the telephone with respect to the hands-free kit, winding the wire of the kit
around the antenna of the mobile…). About 40 measurements were made at 900 and
1800 MHz, with 5 different mobile phones and 5 hands-free kits.

There were no instances were the SAR in head was increased with the hands-free kit. The
SAR with the hands-free kit ranged from 0.39 to 0.007 W/Kg depending on the mobile phone
and hands-free kit and the configuration in which it was used. The average SAR reduction is
of the order of a factor of 50 under normal conditions, the lowest reduction, or worst case
scenario, was a factor of 2 only. The efficiency of 2 hands-free kits, used on the same
mobile and under normal use conditions varied by a factor of 1 to 10.

Whilst these results are limited and the measurements need to be extended, the outcome is
consistent with that from in international trials under standard conditions (unlike the results of
other studies which were made public).

2.9 Technical regulations and instrumentation

The development of technical regulations for the certification of different mobile
telecommunications equipment and base stations falls within the competence of the
Telecommunications Regulations Authority (ART). However, the R&TTE directive (Directive
No 99-5 of the European Commission of 3 March 1999 concerning radio emissions and
telecommunications equipment) has not yet been incorporated in French law. The CE
marking of terminals and base stations in France is made without legislative or regulatory
base and without using a harmonised European standard, relying only on the
recommendations of the Council of the European Union of 12 July 1999.

M. Bach-Andersen’s contribution at the request of the group of experts (University of Aalborg

The first draft of a harmonised basic technical standard for measurement of specific
absorption rate resulting from the exposure of people to fields emitted by mobile phones has
been issued for consultation by CENELEC TC 211 in the last few months. The preliminary
inquiry for this standard ended in 5 October 2000, and it will not be published before early
2001. Publication was expected earlier but it was delayed to accommodate dialogue with
the IEEE group developing a similar American standard. Harmonisation would result in a
truly international standard so that results can be obtained using the same measurement
protocols and direct comparison becomes possible. Another standard, a product standard
for the compliance of mobile phones, was placed in circulation by the UTE in France, on 17
March 2000. It was also submitted for preliminary inquiry until 5 Oct 2000. This product
standard is expected to appear in the first part of 2001 together with the basic technical
standard, and to be adopted as a French standard.

The standard for assessing compliance of base stations (in the laboratory) is circulating
within the technical committees of CENELEC and it is soon to be issued for consultation.
Publication is not expected before the end of 2001.

The standard for testing the compliance of base station in situ is under development; it is not
yet available because of the technical complexity of such measurements. At the same time,
the National Agency of Frequencies (Agence Nationale de Fréquences) is considering a
measurement protocol based on the requirement of the European recommendation of 12
July 1999, using a spectrum analyser and a narrowband antenna. Measurements made
with a broadband isotropic probe are not sufficiently accurate and are not representative of
the base station under consideration because they are sensitive to the totality of the radio
frequency spectrum, including sources producing fields that may be bigger than those from
of the base station under test. Prior to the publication a harmonised European standard, the
measurement procedure developed by the ANFR, which is much needed by those in the
field, could be used as a reference procedure in France. Such a reference procedure is
needed now. In the absence of a standard procedure, measurements have been made
using ill-defined protocols and, occasionally, inadequate equipment. Results, some of which
have been inaccurate, have been published in the press. Some private technical control
bodies have been active in this field because of the demand for measurements created by
public concern; the results obtained are uncertain or at least difficult to compare or verify.
For the future, it will be necessary to have reference procedures from a national control
organisation, which could be the ANFR, and which would be in a position to certify technical
control organisations carrying out measurement in accordance with standard procedures.

3 Limits on the Exposure of People to Radiofrequency Fields:
Basic Principles of Risk Management

The exposure limits adopted in the 1999 European Council Recommendation are based on
the outcome of a long process of deliberation by groups of experts. The recommendations
were established by committees of scientists and engineers with academic, industrial and
government backgrounds. The most widely quoted are the 1991 IEEE and the 1998 ICNIRP

The risks associated with exposure to electromagnetic fields in general and radiofrequencies
in particular have been evaluated after an exhaustive review of the scientific literature on
their biological effects. Following this scientific review, the establishment of limits of
exposure is based on consideration of the most sensitive biological effects reported to date
(that is those that appear at the lowest level of exposure tested and have undesirable health
consequences). Extra factors are then applied to these values to account for uncertainties
and to provide a certain "safety margin". The scientific state-of-knowledge is constantly
reviewed, as are the limit values which are derived from it; the scientific database held by
WHO comprises more than 40 to 100 articles describing recent or ongoing studies.

The consensus of opinion from these committees concerning undesirable effects observed
in exposed animals is that changes in behaviour, as observed in rhesus monkeys and
rodents, have the lowest exposure threshold. The change in behaviour most commonly
observed is an increased difficulty in, or a complete inhibition of, the performance a complex
learned task when exposed to RF energy above a certain threshold. The experimental
results show that this is a thermal effect; it occurs when the power absorbed by the body,
under whole body exposure conditions, reaches or exceeds a specific absorption rate of 4
W/kg averaged over the whole body. With the inclusion of a safety factor of 10, the value of
0.4 W/kg was proposed as exposure limit for workers under conditions of whole body
exposure. This parameter, derived from the study of a biological effect with an undesirable
health consequence, is termed a “basic restriction”. An additional safety factor of 5 was
introduced for the general population, in order to take into consideration the fact that people
have no control or knowledge of their exposure and their environment, and to account for the
variation in response due to the physiological or pathological state of individuals. The public
limit for whole body exposure is thus 0.08 W/kg.

Dosimetric analyses have indicated that, during whole-body exposure, certain parts of the
body could absorb up to 25 times the average. It was thus decided that the localised
specific absorption rate should not exceed 2 W/kg for the head and the trunk (since they
hold vital organs such as the heart, lungs, intestine) and 4 W/kg for the extremities. The
"local" exposure limits are not therefore directly derived from biological considerations, but
are indirectly obtained from dosimetric observations and extrapolated from whole body
exposure conditions.

SAR is not a quantity that is easily measured; it was therefore necessary to establish by
means of "transfer equations" electric and magnetic field strengths that are capable of
producing, under worst-case exposure conditions, SARs corresponding to the basic
restriction. These fields values can be measured; they are termed “reference levels”.
Because the absorption of RF energy in the body is frequency-dependent, the reference
levels are also frequency-dependent. The electric field strengths capable of producing a
SAR of 0.08 W/kg during whole-body exposure are 41 and 58 V/m respectively for the
mobile telephony frequencies of 900 and 1800 MHz10. These field strengths correspond to

power densities of 4.5 and 9 W/m2 . The derivation of the reference levels incorporates an
additional safety margin with respect to the basic restriction: they were derived under worst-
case exposure conditions so that the basic restrictions will always be met when the field
strength is less than or equal to the reference level.

WHO has based its recommendations on those of ICNIRP; similarly the Council of the
European Union incorporated them into its 1999 Recommendation8.

In summary, the current standard includes 3 "safety factors": (1) the fact that the basis of
the assessment is the lowest threshold for a biological effect; (2) the additional safety
margins applied to these thresholds; (3) the way in which reference levels are derived from
basic restrictions.

It should however be noted that the safety margins incorporated are more (sic) restrictive
than those usually used for chemical substances. In this case, starting from the lowest dose
establishing an undesirable effect in animals (LOAEL) the first reduction factor (2 to 10
depending on the type of effect) leads to an estimate of a dose that has no undesirable
effect (NOAEL). A factor of extrapolation from animal to man (often 10) is then applied
followed by another coefficient (often 10) to take into account the sensitivity and different
responses between people. In total, a reduction factor of up to 1000 is then applied from the
LOAEL in the animal to the limit value for the exposure of the population while the sum of all
reduction coefficients is only 50 in the case of electromagnetic fields. This is because the
scientific uncertainty is judged to be less than that associated with toxic chemicals because
of the procedures leading to the establishment of exposure recommendations for fields.

8 th
EU Official Journal L1999/59 of 30 July 1999

4 State of Scientific Knowledge: Analysis of Reports and
Recent Documents on Mobile Phones and Health

4.1 Fundamental reviews

4.1.1 Analysis of the ARCS Report (Austrian Research Centre Seibersdorf)12

In Austria, the Seibersdorf Research Centre is the most important research establishment
outside a university. It has already produced, for the government or by contract with
industry, several reports on the biological effects of electromagnetic fields (reports NIR I to
V, between 1985 and 2000). The report considered here is the executive summary of the
NIR V report dealing with RF EMF. Published in May 2000, it deals with the origin and
nature of EMF, results of experimental and epidemiological studies, with special emphasis
on those dealing with cancer, electromagnetic interference with medical implants and reports
on national and international regulations.

1. Studies dealing with central nervous systems and behaviour.

a-Animal studies
Summary of conclusions: Degenerative effects on the eye and changes in EEG after
exposure to modulated fields have been observed, but the results are inconsistent and the
studies must be replicated.

EEG: Three studies, on rats, cats and frogs, were considered. The last of these used very
high exposure levels, outside typical human exposure levels (peak SAR close to 100 - 3000
W/kg). EEG appears to be modified by a 950 MHz pulsed field with a power density of 15
W/m2 but not for 3 W/m2 .

The eye: Results of 4 studies describing physiological or histopathological effects of single

or repeated exposure of monkeys are reported. The power densities are very high, on the
order of 300 to 430 W/m2 for the frequencies 1250 and 2450 MHz, with SARs of 4 to 20
W/kg9. Three authors found corneal or retinal anomalies, or physiological changes, after
repeated exposure while the fourth reported results after a single exposure to a SAR of 8 or
20 W/kg, but not to a SAR of 4 W/kg.

b- Human volunteer studies

Summary of conclusions: some, but not all, authors reported changes in EEG. Reductions
in reaction time were reported. Shortening of sleep time and in the duration of the
paradoxical (REM) sleep were also described but not confirmed in the course of two
replication studies by the same authors. The results are inconsistent and should be
replicated. Overall, these modest effects do not seem to affect well-being.

A variety of effects were reported in eight studies, five researchers did not find any EEG
changes under a range of experimental conditions (analogue or digital mobile phone; rest or
waking EEG, with or without visual or acoustical stimulation, continuous or intermittent
exposure; power densities ranging from 0.2 at 50 W/m2).

In one study, for a local SAR in the head of 1 W/kg, effects on sleep were observed,
including changes in sleeping patterns, shortening of waking time whether conscious or
unconscious, together with changes in the intensity of the EEG in the dream phase.

The report does not specify if it is localised or whole body SAR

The performance of cognitive tasks and the shortening of reaction times have been reported
in two studies; the reduction in the execution times of tasks requiring vigilance were only
modest and these results could be explained by a mild thermal effect. In two other studies
there were no significant differences related to exposure.

Comment of the expert group on effect on the central nervous system and behavioural
studies: The first lesson to be learned is that the experimental protocols followed in the
studies were very different; it was therefore not possible to make comparisons, and this is
true of human volunteer studies as well as animal studies. The occurrence of physiological
and behavioural effects cannot be confirmed from these studies because of a lack of
replication under identical experimental conditions. The hazardous implication of these
effects has not in any case been established. Investigation of effects on the eye should be
extended to include exposure conditions that are commensurate with the use of mobile
phone. Despite the reported limitations, these conclusions are based on the findings of a
large number of studies.

2. Studies related to cancer

a. Experimental studies
Summary of conclusions: Taken as a whole, the studies do not show any convincing
evidence of a risk of cancer. The relevance of the studies to the risk of cancer in man has
not been established. RF EMF has no genotoxic effect in vivo or in vitro, as least under
conditions that do not lead to thermal effects. Some modest indirect effects on genetic
replication and/or transcription were observed under certain specific exposure conditions,
but the health consequences of such effects have not been established.

Genotoxicity has been studied in vivo and in vitro. Although the majority of studies were
negative, an increase in DNA breaks in rat brain cells, following exposure to RF EMF at
2450 MHz in the form of short intense pulses, has been described, but this result has not
been independently reproduced. Breaks in DNA have also been observed in radio-sensitive
cell lines exposed to modulated fields around 813 MHz, but not at 836 MHz, with a SAR
between 2.4 and 26 W/Kg, suggesting that a DNA repair mechanism, that has not been
sufficiently studied to date, could play a role. Chromosome aberrations have been reported
in human lymphocytes exposed to fields from a GSM base station, but the same author
could not replicate this result. By contrast, micronuclei analysis has shown RF-induced
changes in plant tissue (exposure to a 10 to 21 MHz field) and in human lymphocytes (9
GHz); these frequencies are quite different from those used in mobile telephony, so these
results must be reproduced under more realistic exposure conditions.

The activity of the intracellular enzyme ODC (ornithine decarboxylase) was increased in
several experiments with amplitude-modulated fields, but not in the case of frequency
modulation or no modulation at all. In view of the possible role of ODC in the promotion or
progression of cancer, this could point to the possibility of some epigenetic mechanism of
carcinogenesis but this work needs to be replicated.

Studies of the expression of proto-oncogenes (genes implicated in the process of regulation

of cellular proliferation and differentiation) have not produced unequivocal results, despite
reports of positive results (neuroblastoma cells, 915 MHz). Conflicting results on cellular
proliferation are reported: out of 7 publications reviewed, two reported an increase in
cellular proliferation for SAR in the range 5 to 81 W/Kg, three others reported inhibition of
cellular growth (7.700 MHz or GSM 960 MHz, or non-modulated 835 MHz). Previous
studies reporting evidence of transmembrane calcium efflux were not confirmed by recent

Studies on the initiation and the promotion of tumours have been described in nine studies.
These include the study of Repacholi et al (1997), which showed an increase in the number
of lymphoma in a genetically-sensitive strain of mice after exposure for 18 months to 900
MHz EMF of modulated at 217 Hz. The results of this study have been discussed, with the
aim of eliminating interpretation problems arising from the heterogeneous exposure
conditions of the mice; replication studies are in progress, with a better-controlled
experimental protocol. The other studies reported were negative, including those dealing
with the joint effect of fields and known cancer initiating agents (diethylnitrosamine) and
promoting agents (aromatic polycyclic benzo-a-pyrene).

b-Epidemiological studies
Summary of conclusions: The majority of studies, all with different protocols, did not
characterise properly the exposure to RF EMF; some studies were not sufficiently large or
lacked information on potential co-founders. The studies were not consistent in the types of
cancers that they considered might be related to EMF. In view of these limitations, it is not
possible at present to conclude on the existence of a cancer risk to general population
related to exposure to RF EMF.

The ARCS report first reviews, in this very important chapter, the well established quality
criteria for epidemiological studies. These are the criteria that are always considered by the
epidemiological community. Eight studies are presented, of which four relate to
occupational exposure to RF EMF. Of these, one relates to the exposure of personnel in the
Polish army exposed to high frequencies (150 to 3500 MHz)1 - a study whose results are not
in agreement with other work on military personnel, particularly that on American Navy
personnel; another discusses Canadian policemen exposed to fields from radar, while a third
reports on Norwegian Merchant Navy radio and telegraph operators (405 to 25 000 MHz
fields)10. The results of these three studies, for which the exposures were very varied,
mostly characterised by job description and only in a few rare occasions by some
measurements, cannot not be extrapolated to the mobile phone situation; it is important to
note that the cancers linked to these exposures are very diverse (leukaemia, lymphoma,
malignant melanoma, prostate and testicular cancer, breast and uterus cancer).

Three studies were initiated specifically to investigate clusters of cancer cases. It is well
known that a cluster of cancer cases does not necessarily mean that there is a common
cause; if so, epidemiological studies cannot lead to a definitive conclusion. One cluster
occurred in Great Britain in the neighbourhood of a radio and television broadcast station,
where there was an apparent increase in adult leukaemia. The study showed an increased
incidence of lymphoma, skin melanoma and bladder cancer in the population within 10 km of
the station, but not for all cancers or for childhood leukaemia. Another study of cancers
around 20 broadcast stations in Great Britain was made by the same team, following the first
study; the results do not show any increase in cancer of the bladder, adult leukaemia or
melanoma, and thus it does not confirm the findings of the initial study. A similar study was
carried out around three masts in New South Wales, in Australia. When six municipalities
within 12 km were considered, an excess of childhood leukaemia was observed in regions
within 4 km, but there was no increase in adult or child brain cancer. When other nearby
municipalities were included in the analysis, the results were not changed except that in one
of the extra municipalities there was an excess of childhood leukaemia.

A case-control study conducted in Sweden suggest a non-significant link between brain

tumours (malignant or benign) and the use of an analogue mobile phone on the same side
of the head; the data could not be used to investigate a link with digital GSM phones, this
being a more recent technology. However, no association was found with the use of a

Additional data taken from Elwood, Environ Health Perspect, 1999 (referred to in the ARCS report)

mobile phone if the analysis did not take into consideration the side of the head on which the
cancer occurred. These results are based on small numbers and need to be confirmed, as
stated by the authors of the study (see the annex for the different ways in which these data
can be interpreted [Hardell et al 1999], in different articles and reviews).

One study compared the causes of death between mobile phone users and users of hands-
free car phones, were the antenna is distant from the head No difference in mortality was
observed, after one year of mobile phone use - which is a very short time – while the
mortality among mobile phone users was less than the population as a whole, probably for
social and demographic reasons.

Comments of the expert group on cancer-related effects: The history of mobile phone use is
not long, and this limits considerably the scope of epidemiological studies, as the latency
period for cancers can exceed 10 years (it could be shorter for leukaemia or thyroid cancer,
although the latter is not under consideration here; the latency could also be shorter if an
already-initiated cancer is promoted or progressed, which is the hypothesis currently under
consideration by the international IARC project). When the risk is low (practicably, a relative
risk less than 2), characterisation of exposure, to allow discrimination between different
groups, assumes great importance . The studies presented were somewhat lacking in this
respect, as observed in the Austrian report. Although the review of epidemiological studies
is not exhaustive, the studies not considered do not convey a different message.

The review of experimental data from in vitro and in vivo studies and from epidemiological
studies does not add up to conclusive evidence with respect to cancer, as reported in the
Austrian document. The case for pursuing further research is however justified because
even though the risk of cancer is not supported by genetic mechanisms, as suggested by
current data on the genotoxicity of RF EMF, epigenetic mechanisms cannot be excluded at

3. Studies related to the cardiovascular system

Summary of conclusions: No clear interpretation can be made of the scarce data available.

One study found no changes in 12 measured parameters of cardiac rhythm and

electrocardiogram pattern after exposure to a GSM-900 field. For a similar exposure,
another study found an increase in systolic and diastolic blood pressure at rest, with a
slowing in the cardiac rhythm. However, the study protocol did not allow randomisation, and
for this reason it cannot be considered a blind study. Because the effects may therefore be
either real or placebo, interpretation of this study is limited.

Comments of the expert group relating to cardiovascular system: the studies are very
limited in number; the data do not allow for firm conclusions to be made.

4. Studies on reproduction and development
Only two studies on rodents are described, suggesting a reduction in the weight at birth, but
the relevance of the results is limited by the quality of the exposure protocol. The authors do
not report teratogenic effects.

Comments of the expert group on reproduction and development; the studies are very
limited in number; these data do not allow for any conclusions to be made. This is a
sensitive biological end point necessitating further experimental studies.

5. Studies on the immune system

There are very few studies relating to the immune system. Three studies are discussed in
the report, of which two are on rodents. The first study finds, inexplicably, modification in
immune factors in male mice but not in female after exposure to a modulated or
unmodulated 2450 MHz field; the second (GSM 900) finds no effects on different
lymphocytes of rats. The last study relates to occupational exposure (personnel using
diathermy equipment), and reports no observable effect.

Comments of the expert group concerning immune response: The studies are very limited
in number but do not support an effect of RF EMF on the immune system.

6. Studies of interference with medical implants

Summary of conclusions: More attention should be given to electromagnetic interference
between RF EMF and electronic biomedical implants. The continuing increase in the
number of RF sources makes the possibility of interference more likely. New, more
appropriate, electromagnetic compatibility tests must be developed.

Cardiac defibrillators and pacemakers. The latest mobile phone technologies are more
likely to cause interference because the low frequency components associated with them
are close to cardiac electrical frequencies. In vitro studies (active implants not worn by
people) showed that signals from digital phones were more likely than analogue signals to
influence the functions of some of the pacemakers tested, at short distance. Cardiac
defibrillators implanted in people suffering from tachycardia and fibrillation showed similar
tendencies. Beyond 20 cm, none of the devices were perturbed. Similar results were
observed in a study of a human volunteers wearing pacemakers, with a distance of less
than 15 cm between the antenna and the implant.

Consequently, it is recommended that people wearing a pacemaker carry their mobile phone
15 cm away and that while making a telephone call they place the device on the opposite
side to the pace-maker. With these measures, the use of mobile phones should be without
risk. Care must be taken while going through electronic security installations at the exit of
some stores. People with cardiac defibrillators should not dawdle within the bounds of the
equipment; instead they should move rapidly through the installation without coming into
close contact with the antennas.

Neurostimulators. These implants are used to alleviate tremors of people suffering from
Parkinson's disease. Tests using signals from different 900 MHz mobiles showed no effect
on their function, probably because (unlike pacemakers) these implants are not designed to
react to physiological signals.

Audio prostheses. Electromagnetic interference leads to undesirable noise

Comments of the expert group on electromagnetic interference with medical implants: The
Austrian report highlights the great need to develop research on electromagnetic
compatibility in view of the significant increase in the number of sources and the number of
implants in the population in general. The immunity of implants must be improved, and the

information given to the people concerned must ensure that they know what precautions to
take to limit their exposure.

7. Other effects
Summary of conclusions: these are studies on people regarded a priori as more sensitive
(older people or people who report being unwell). The significance to health of the reported
endocrine effects is not clear.

In experiments on human volunteers, weak and transient variations in the concentration of

certain hormones in the blood were reported in two studies, one with an exposure profile of
two hours per day over 5 days and the other for one night (GSM 900).

A variety of subjective symptoms were reported by some users of mobile phones (headache,
local sensation of heat, fatigue…). One study was conducted in Sweden and Norway
among users of analogue GSM devices. This epidemiological study showed a higher
incidence of problems with analogue devices, which could be correlated with higher fields
emitted and to a thermal effect due to the high battery current.

Another cohort study was undertaken in the vicinity of a short wave source in Switzerland.
Complaints were more frequent in the immediate vicinity of the source (restlessness,
insomnia, fatigue…); the authors of the study recognised that it is difficult to conclude that
these manifestations are due to electromagnetic fields rather than to public concern.

Comments of the expert group on other effects: RF EMF could have effects on certain
hormonal functions. At present it is not possible to say whether these are physiological
adjustments in response to the exposure or an indication of possible health risk. These
studies should be conducted primarily on people described a priori as highly sensitive, for
example, people taking medication for psychoses, or people on anti epileptic or and
migraines drugs

General conclusion of the group of expert on the ARCS report: Although recent, this report
is limited in content by comparison with the Canadian and British documents. The actual
report of which this is the summary has not been consulted. The literature cited does not
appear to be exhaustive, and the criteria used for selecting papers for review are not always
given. The conclusions of the authors are however similar to those of the more
comprehensive reports just mentioned. This report is however interesting in as much as it
reports in greater details on the risk related to interference of electromagnetic fields with
implanted electronic prosthesis and gives recommendations to their users.

4.1.2 Analysis of the Stewart Report11

The Stewart report was written at the request of the UK Government in order to evaluate the
potential health effect of mobile phones and base stations. The committee was made up of
12 experts in subjects and 3 external observers; the committee comprised specialists in
oncology, physics, statistics and neurophysiology, one member from WHO and two
members from the advisory group of the National Radiological Protection Board (NRPB), two
members of non-scientific background (politics and social communication). The observers
were from NRPB, the Department of Health and the Department of Trade and Industry. The
committee decided to organise a large programme of consultation in the UK and abroad;
among scientists, members of the public, groups such as “Friends of the Earth”,
Powerwatch, the Northern Ireland Families Against Telecommunications Transmitter
Towers, companies involved in mobile telephony, operators and the Federation of the

Mobile Phones and Health. Report from the Independent Expert Group on Mobile Phones, May

Electronics Industry. The committee also relied on information from the press and 5
regional public meetings in order to be to be fully aware of the reaction and sensitivity of a
large percentage of the population. Invitations to public meetings appeared in the national
press: individuals and organisations were encouraged to present their views. Written
evidence was also received from a large number of people. Certain number people and
organisations (28) were invited to present their evidence in closed meetings with the
committee. The establishment of a health risk from exposure to RF fields depends on the
outcome of well conducted reproducible scientific studies. This is particularly important
when studying the effects of mobile telecommunications equipment, as they are very weak.
Such studies are mostly published in peer-reviewed journals; in addition, the committee of
experts considered information from all other sources of available to it.

The report has five chapters and three annexes:

Chapter 1: Summary and recommendations
Chapter 2: Introduction to the technical aspects of mobile telecommunications
Chapter 3: Public perception and concerns
Chapter 4: The physics of mobile telephony
Chapter 5: Scientific “evidence”: Mechanisms of interaction with biological tissues, in vitro
and animal studies, human laboratory and epidemiological studies, mobile phones and
Annexes: Current research funding, Precautionary Approach, Procedures of the deployment
of the telecommunications network, Bibliography, Glossary, Constitution of the Expert
Group, Written evidence (174), Public (5) and closed meetings

In vitro and animal studies

Studies on the nervous system and behaviour
Behavioural change is the most sensitive indicator of a health effect from exposure to
electromagnetic fields. It occurs at exposure levels well below those that can lead to
damage. Behavioural tests can be used to assess cerebral functions such as memory and
other cognitive functions. Much of the research, especially in the early studies, was
conducted with a high level of RF exposure, or low average levels using high peak power
pulses that are characteristic of radar and quite unlike emissions from telecommunication

Comment of the expert group: emissions from mobile phones may or may not be in
pulses but the mean power is not very significantly lower than the peak: in the GSM system,
it is 8 times lower; in radars, this ratio is usually around 1000.

Changes in calcium efflux were observed together with other different phenomena:
a decrease in the rate of formation of ion channels specific to calcium,
a decrease in the frequency of channel opening,
an increase in the occurrence of spontaneous action potentials.

Calcium is associated with the activity of a certain neurotransmitter receptor: (the NMDA-
receptor). More specifically, in the cortex and hippocampus it is associated with long-term
potentiation, which is thought to be associated with mechanisms of memory. In an in vitro
study, it was shown that phase transition in lipid membranes was facilitated by low level RF
EMF. However, occurrence of this phenomenon, which was observed at temperatures
between 17.7 and 25°C, is not thought to be possible in man at physiological temperatures.
Many other studies did not show any modification in transmembrane action potential, or in
calcium and potassium currents (36 articles and 2 reviews).

Several studies show effects on neurotransmitters and their metabolism; the work of Lai is
cited in this context, and the Stewart report suggests that changes in acetylcholine
metabolism could be related to thermal sensitivity of the hypothalamus 12 (7 articles).

One study showed a modification in the enzymatic activity of acetylchlolinesterase,

suggesting a specific effect due to 16 Hz modulation, as well as a window effect with respect
to the amplitude of the applied signal, which at present has not been confirmed or
independently replicated (10 articles).

Several studies showed changes in EEG, but the conditions under which these changes
take place are ill defined (7 articles). Learned task performance and memory can be
affected in rodents at whole body SAR levels of 2.5 to 8 W/Kg, due to heating of
approximately 1°C; these effects are correlated with an increase in the expression of c-fos: a
proto-oncogene implicated in neuronal plasticity. The sensitivity to such effects depends on
the frequency and on environmental factors such as external temperature and humidity. The
threshold of exposure leading to an effect, under the most adverse conditions, can be as
low as 1 W/kg The study by Lai stands out from the others by showing effects at 0.6 W/Kg.
Wood has even shown, in an in vitro study on hippocampus slices, a change in the long term
potentiation at 0.001 W/Kg! The statistical approach of Lai has been contested ( 5.88) (9
articles, 1 review).

Effects on the permeability of the blood-brain barrier have not been independently
confirmed (7 articles, 1 review).

Thermoregulation comes into action when the SAR is comparable to or greater than the
rate of production of metabolic energy, that is of the order of 1 W/Kg. RF EMF at
frequencies below 10 GHz is less effective in inducing thermoregulation because of the
smaller percentage of energy reaching subcutaneous receptors involved in the
thermoregulatory response (2 articles, 1 review). The level of physical activity is reduced to
lower endogenous heat production, this happens above 1 W/Kg at 2.45 GHz, and 3 W/Kg at
915 MHz. One study reported that activity levels were reduced after 6 weeks continuous
exposure at 0.4 W/kg before stabilising (1 review: UNEP/WHO/IRPA, 1993).

The auditory perception of microwave pulses is now a well-characterised phenomenon

that is know to occur with pulses of duration less than 35 µs when the energy content in the
pulse is greater than 1 mJ/kg, and a peak SAR of the order of 30 W/Kg during the pulse.
This phenomenon can generate stress in animals who perceive it and it is important to take
this into consideration in the interpretation of observed effects. This effect cannot occur
under mobile telephony exposure conditions (6 articles, 2 reviews).

Effects on the eye, which could not be confirmed, could only be observed at very high peak
powers (9 articles, 1 review).

An inhibition in the secretion of melatonin is frequently reported at low frequencies, but has
not been demonstrated at radio frequencies and specifically not in the frequency range 3 to
30 MHz or at 900 MHz in the rat, or at 900 MHz or 1800 MHz in man (5 articles, 2 reviews).

Comments of the group of expert on neurological effects and behaviour: In its

conclusions, the Stewart report appears to have underestimated the work of Lai on
neurotransmitters, attributing this to a thermal effect; in our opinion this is not the case.

Comment of the group of experts: this remark is surprising since the SAR used is 0.6 W/kg usually
considered non thermal, including by Lai himself

While it is true that the reported effects on memory have not been confirmed, they constitute
nevertheless the coherent set of studies that have not been refuted. However these effects
cannot be extrapolated to mobile telephony exposures because of the very short pulses and
high peak power that were used, and the possibility that the auditory effect might be

Cancer and Genotoxicity

The different phases of the development of cancer are recounted: initiation, promotion, and
progression. The role of the enzyme ornithine decarboxylase (ODC) in carcinogenesis is
described: this enzyme plays an important role in the metabolism of polyamines, which can
stimulate the synthesis of DNA, as well as cell growth and differentiation. Its concentration
and its activities are increased in the course of carcinogenesis, after the activation of genes
that synthesise it. Some authors reported an increase in the activity of ODC during
exposure to electromagnetic fields: a doubling in the activity of low intensity RF fields
modulated at 60 Hz. This increase was not accompanied by an increase in the synthesis of
DNA. Moreover, the increase in ODC activity caused by known promoting agents is in the
order of 500. The reported effect can than be considered to be minor.

Comments of the expert group: The authors conclude that there is an increase in the
concentration of ODC, in our opinion this has yet to be confirmed. Moreover the authors
indicated that the observed changes are of the same order of magnitude as those required
to cause promotion, they then qualify the finding as a relatively weak effect on the activity of
ODC. The authors estimate that it is very unlikely that the increase in ODC has a promoting
effect. We agree with their appraisal. However it is difficult to reach this conclusion without
the support of more detailed arguments. They refute the idea of a synergy with other
environmental factors without presenting a rationale for their hypothesis (9 articles, 1

With respect to gene expression, table 5.8 qualifies two studies as being positive: Goswani
et al (1999) have shown a slight elevation in c-fos but not in c-jun or c-myc at 0.6 W/Kg, and
Ivaschuk et al (1997) have shown an increase in c-jun but not in c-fos “at higher intensities”
(5 mW/kg). These results are described as “contradictory” while, in fact, they relate to
different models and different intensities; if there were a specific mechanism at radio
frequencies, could it not have a thermal threshold at 0.6 W/Kg on the c-fos gene and a
specific effect at 5 mW/kg on the c-jun gene ? (10 articles).

Changes in cell growth of the order of 10 to 20% were reported for yeast cells by Grundler
(1992) with 41-42 GHz fields, but these effects were not replicated by Gos (1997). A low-
level effect was reported by Stagg who observed an increase in DNA synthesis in an
intensity window of 6 mW/Kg, while Kwee and Raskmark found a decrease in their
proliferation. The authors of the Stewart report recommend further studies to compliment
those of Stagg (8 articles).

No mutagenic effect and no damage to DNA were observed in vitro. Some fragmentation
of DNA was reported in vivo by Lai using the comet essay after exposure to a signal very
different from that used in mobile telephony, this result is one of many by the same author
and his group showing similar effects. Other research groups were not, however, able to
confirm these findings. Two in vitro studies and one in vivo study showed an increase in
chromosomal aberration. The result of studies of sister chromatid exchange were
inconsistent, increased micronucleus formation has been refuted in in vitro studies but has
sometimes been reported after exposure in vivo. The health consequences of these
observations are not clear.

Studies on the incidence of tumours after long-term exposure are not numerous and mostly
negative (Chou et al., 1992), even on animal models with a high incidence of spontaneous

tumour formation such as C3H/HeJ mice. A co-promoting (“epigenetic”) effect was reported
by some research groups but not by others: it should be studied further. Finally, studies on
the proliferation of grafted tumours were negative. Out of 38 articles reviewed, only 4
showed increased tumour incidence. Of these, 3 pertain to high SAR levels while the 4th,
that of Repacholi (1997) is currently being repeated in 2 different laboratories (4 reviews).

Haematology, immunology and longevity

Weak and transient effects on the haematopoietic and immune systems have been
described at low exposure levels. Many studies on longevity show no effect while one
showed a decrease in life span for lifetime exposure of one hour per day at 7 W/Kg (10
articles, 1 review).

Reproduction and development

No effect was observed on reproduction; a small increase in the physical activity of second
generation offspring (in rats) was observed at high SAR levels. A decrease in the fertility of
male rats was observed after a prolonged exposure to 2 W/Kg, one study indicated a
decrease in the fertility of female mice exposed near radio and television broadcast
antennas. However these studies had no matched controls. While admittedly not a well
designed study it suggests important health implications and this area should be pursued
further (17 articles, 3 reviews).

Cardiovascular system
There are no reports of low-level RF fields on the cardiovascular system (4 articles).

Comments of the expert group on animal and cellular studies: some reports of cancer
co-promotion need verification. The authors of the Stewart report reported damage to the
ocular tissues, however these effects were due to very high intensity pulses. Also reported
are effects on the nervous system of the rat. These effects were minor and difficult to
interpret in terms of biology or health; they do not appear to lead to altered cellular function
or to have an impact on health.

2. Laboratory studies on human volunteers

Brain function (3 articles)
Among the concerns expressed about the use of mobile phones is the possibility that mobile
phone signals have undesirable effects on cognitive functions such as memory, attention
and concentration. However, with the exception of 3 studies, the data available is mostly on
the physiological aspects of brain functions such EEG rather than on performance and
cognitive function as such.

Preece et al (1999) exposed 36 volunteers to a 915 MHz signal of 1 W continuous wave, or

0.125W as a train of 217 Hz pulses. The volunteers' performance of tasks in terms of choice
reaction time and memory was graded. The exposure had no effect on memory or simple
reaction time. A shortening in choice reaction time (374 ms instead of 388 ms for right/left
button choice, in response to on-screen information, for example yes or no) was
demonstrated when the power output of the exposure system was 1 W, continuous but not
when it was 0.125 W pulsed.

Koivisto studied 48 volunteers with tests exploring the same functions with a 902 MHz signal
modulated at 217 Hz, with a power output of 0.25 W. He did not confirm Preece's finding of
an effect on choice reaction time at 0.125 W, but he found a significant effect on what its
termed “vigilance” reaction time. In another study on working memory, currently in press, he
found a decrease in the reaction time of the visualisation of a target letter in a sequence of
letters. Paradoxically this visualisation time is increased for the reaction to a target letter
presented after just one or two previous trials.

These results imply that future experimental conditions should reflect a more elaborate
hypothesis, in order to narrow the range of circumstances under which an effect occurs to
include a more limited number of psychophysiological variables. Moreover these
investigations deal with short-term effects, and it is therefore not possible to make any
deductions, on the basis of these findings, about long-term consequences of exposure.

Electroencephalogram (EEG) (10 articles)

The functional significance of the different frequency components of diurnal EEGs is not very
clear. Night-time EEG is better understood: it has well-defined characteristics corresponding
to various stages of sleep in a healthy individual. It is also possible to measure the electrical
activity of the brain associated with events resulting in sensory, cognitive or motor
responses. These are known as "evoked potentials". Various effects have been reported on
awake and sleep EEG, and sometimes these have been difficult to replicate within one
research group. There is no coherent picture and at times the data appear contradictory.

Krause et al (2000), found a change in the frequency content of the EEG between 4 and
12 Hz when hearing a word matched to four target words given at the beginning of the test.
This difference was considered by the authors as possibly correlated to behavioural
modification as described by Koivisto et al (2000, in press). Unfortunately, Krause et al
(2000) did not report the performance scores of the subjects, and so there is no way of
knowing if the exposure improved performance of memory-related tasks or if modifications in
EEG are really correlated to this effect.

Three other studies on evoked potential gave the following results: one study on visual
potential by Urban et al (1998) was negative, one by Freude et al (1998) showed a decrease
in amplitude of the response, and finally that by Eulitz et al (1998) on cortical auditory
potential associated with a vigilance task showed a decrease in the spectral power at high
frequencies (18-30 Hz).

Conclusion of the expert group on studies concerning brain function: The results of
Preece and Koivisto suggest that acute exposure to fields from mobile phones at levels
below the recommended limits produce effects of sufficient magnitude to modify behaviour.
The causal mechanism remains unclear, but may well be due to a small localised thermal
effect. It appears clear that exposure to mobile phones modifies certain brain functions, it is
not however possible to establish any clinical significance for such effects or to demonstrate
a possible health effect. Long-term exposure studies among groups of new mobile phone
users could be designed to identify changes in brain function associated with cumulative
exposure. Such studies should be a priority for future research.

Effect on the heart and blood pressure (9 articles)

Braune et al (1998) showed small effects on cardiac rhythm, blood pressure and capillary
perfusion, but the order of the placebo and real-exposes was always the same, and because
of this it can be considered that there was no real control group.

The expert group concludes that there is no reason to suspect an effect on the heart and the
circulation but they advise that Braune’s study should be repeated with an appropriate

Mobile phones and driving (15 articles)

The use of a mobile phone can disturb driving directly if one hand holds the phone and
becomes unavailable for driving. It also has an indirect effect by redirecting attention from
driving to the telephone conversation. Several laboratory studies have shown, surprisingly,
that the effects are equally important, that is the distraction due to a telephone conversation
is just as significant whether a hands-free or a handheld telephone is used. The effect of a

telephone call is more important than simply listening to the radio or even performing an
automatic task such as repeating words; the effect on driving is evident during a simple
conversation, increasing with the cognitive load of the discussion, and is more significant in
older drivers.

3- Epidemiological studies
Studies pertaining to the general population (6 articles)
The cohort study carried out by Rothman et al (1996) was prematurely interrupted. The
absence of difference of mortality between users of hands-free and handheld phones after
three years is not sufficient to draw conclusions. In fact if there were a cancer risk it would
need a longer period to manifest itself.

The case control study of Hardell et al (1999) in Sweden did not show any increased risk of
tumour formation related to the use of a mobile phone, either analogue or digital. A fuller
critique of this article is presented in the annex.

Hocking (1998) contacted 40 persons who described themselves as having symptoms

attributed to the use of a mobile phone. The main symptoms were pain, an uncomfortable
heating sensation, blurred vision, effects on hearing or vertigo. None of the people who
responded in the study reported epileptic fits. Hansson-Mild's group (Oftedal et al, 2000),
reported the same incidence, in users of analogue and digital mobile phones, of a range of
symptoms including headaches, fatigue, and sensation of heat behind the ear; their
occurrence was related to the number and duration of daily calls. The study used a postal
questionnaire and may well suffer from selection bias.

Occupational studies (4 reviews)

Numerous other studies have looked into the incidence of disease and death in relation to
occupational or hobby-related (amateur radio) exposure to RF fields. The diseases most
commonly investigated are lymphatic and haematopoietic cancers, and brain cancer (18

Cancer: Szmigielski (1996) reported a 6-fold increase of the risk of lymphoma and
haematopoietic cancer among Polish military personnel exposed to RF while at work.
However, this study is deficient in many ways that limit or even nullify its conclusions; firstly
the exposure of cancer cases was not evaluated in the same way as for the general
population, and secondly the statistical analysis is not well described, with important data
missing from the report. With this exception of this study, out of 9 other publications only
that of Tynes et al (1992) shows a significantly increased risk - of leukaemia in Norwegian
electrical personnel. Similarly in the case of brain cancer, apart from Szmigielski's study, 2
out of 7 studies, both case-controlled studies, showed a significantly elevated risk; that of
Thomas et al (1987) who evaluated the exposure from job title, and that of Grayson (1996)
which indicated a modest increase in risk even for the highest exposure (RR = 1.39; 95%
confidence interval = 1.01 - 1.90).

In summary, taken as a whole these studies do not show an increased risk of cancer from
occupational exposure to RF. However, the exposure parameters vary between the studies
and are not comparable to mobile phone exposures. Many of the studies have low statistical
power and some have deficient methodologies. This is why the absence of consistently
positive findings does not necessarily mean that RF radiation from mobile phones does not
constitute a risk of cancer.

Health aspects other than cancer (5 articles, 1 review): Several cohort studies of
occupational groups exposed to RF radiation have examined non-cancer mortality, and in
some cases morbidity; while these studies did not provide any evidence of health risk, it is
nevertheless reasonable to suggest that workers exposed to high intensity RF fields should

be monitored in the long-term. The experts recommend the setting up of register of
exposure of workers and the assessment of cancer risk and mortality to determine whether
there are any harmful effects.

Residential exposure in the vicinity of radio and television broadcasting masts (8

Some non-significant increases in the incidence of childhood leukaemia were observed in
the study of Maskarinec et al (1994), Dolk et al (1997a), and Hocking (1996). A larger study
by Dolk et al (1997b) was negative while McKenzie et al (1998) showed that the increase
observed by Hocking pertained to one out of three villages and could be described as a
cluster. All the studies of this type suffer from the limitations of ecological studies, which are
based on patterns of cancer and exposure within the population and not on individual cases.
Overall, there is no evidence from these studies that there is a risk to health of people living
near base stations or in places were the level of exposure is only a very small fraction of the
recommended values.

Mobile phones and driving

Epidemiological studies by Violanti (1997, 1998) and Redelmeier and Tibshirani (1997) have
shown a clear association between accident risk and having and using a mobile phone in the
car, for both hand-held and hands-free phones. The experts concluded that there is no
justification for any differentiation in legislation between the use of hand-held and hands-free
phones even though the use of hand-held phones is easier to detect. There is in fact a risk
of appearing to favour, or at least to tolerate to a greater extent, the use of hands-free

The general conclusion of the Stewart Report is given in its prologue: “The balance of
evidence to date does not suggest that exposure to RF radiation from mobile
telecommunication technologies is hazardous to the general population in the UK. There is
some preliminary evidence that exposure to the fields from these technologies can produce
subtle biological effects, but it is important to note that these effects do not necessarily
constitute a health hazard”.

The general conclusions of the expert group on the Stewart report: This report
presents a clear and accurate description of mobile phone systems and of the fields they
emit, in air and in the tissues of the head. The physical characteristics of electromagnetic
fields and the mechanisms of interaction, well established or hypothetical, are described
well13. The reference section is quite extensive but not exhaustive.

The expert group notes that the analysis of biological and health effects in the Stewart report
takes into consideration high- and low-level effects, which could be misleading. Exposure
levels are often not sufficiently precise, and often qualified (“very weak, weak, average or
high intensity”). This can be misleading, and allow errors in interpretation of the actual
values corresponding to these terms. For example a SAR level of 55 W/m2 appears to be
considered as “weak” (§5.138), while a few pages previously, 100 W/m2 is identified as “very
high intensity” (§ 5.128): the interpretation of biological effects occurring under such
exposure conditions can vary, constituting a non-negligible source of errors of judgement.
Similarly, the magnitude of the effects observed is not always indicated, which makes it
difficult to interpret them in term of health consequences.

There were however a few errors that slipped into the report, for example, the mass density of
3 3
biological tissues is 1000 kg/m and not 0.001 kg/m (§ 4.37 p.38), an electron is attracted by a
positive not a negative charge (§ 4.39 p.39), and some arguments are not clear or not sufficiently
precise (the dividing line between ionising and non ionising radiation corresponds, in fact, to the
ionisation energy of the hydrogen atom in the water molecule: 12.4 eV), as this is the most abundant
element in the body.

Some results have been considered in the report as not relevant to mobile telephony: for
example, health effects of radiofrequency fields modulated at 16 Hz are not considered; the
reason given is that this modulation is not used in mobile telephony. The expert group has
reservations on this point because it is perfectly conceivable, in terms of mechanism, that
effects occurring at 16 Hz could also occur at 217 Hz (frequency actually used in mobile
telephony) it is also possible that future developments in technology may one day justify the
use of such a modulation frequency. Similar reservations are directed to their interpretation
of phase transitions in the lipid membrane, and on effects on EEG. If the mechanism
underpinning these effects, and the repercussions on cell or body function, are not known
they cannot be considered as without any health consequences just because the exposure
or biological conditions under which there were observed do not correspond to exposures
from mobile phones.

The expert group agrees with the statement in the Stewart report that there is clear evidence
that exposures to radiation from mobile phones, at intensities below the ICNIRP
recommendations, have short term direct effects on the electrical activity of the brain and on
cognitive functions. It is important to study the consequences of such effects on health,
because, even if only one health effect is present, the recommendations should be changed
accordingly. It is also agreed that it is important to determine if these effects are due to a
slight localised elevation in temperature or, as is also possible, a non-thermal effect. The
group of French experts agrees with the conclusion of the Stewart report that the
experimental studies carried out do not suggest harmful effects on the heart, the blood, or
the immune system but they considered that the data on reproduction and development are

Some people report symptoms such as headaches, fatigue and the perception of heat
behind the ear while using, or shortly after the use of, a mobile phone. It is not clear if these
symptoms are actually due to RF fields. They call for new studies. The epidemiological
evidence to date does not suggest that exposure to RF causes cancer, this is in line with the
absence of mutagenic effects, whether initiator or promoter, from exposure to RF below the
recommended levels. However, the use of mobile phones is too recent to allow for a
decisive epidemiological evaluation, consequently the possibility of an association between
the mobile telephony technology and cancer cannot not be excluded at this stage.

One important health risk of mobile telephony has been clearly established: an increased
risk of road accidents for drivers using a mobile phone. The risk is the same for hand-held
or hands-free phones as it is due to the distraction caused by the telephone conversation
rather than to an impediment to driving the car or to a direct field effect on the brain.

In general the group of (French) experts considers the information in the Stewart report to be
quite considerable and relevant, and agrees with its main conclusions.

4.1.3 Analysis of the "McKinlay" and COST 244 bis Reports

Report submitted to DG XIII in September 1996 and its update by COST 244 bis in
The "McKinlay" report, which carries the name of the leader of the expert group which
produced it, was written in nine months by ten experts. It was mandated by DG XIII14 of the
European Commission, its aim was to provide research recommendations for a European
research program on the possible health effects of mobile phones. This was the first
important report of its kind. The experts were selected by the Commission and charged with
providing recommendations on the state of knowledge and on the research needed to

Directorate General in charge of telecommunications

safeguard the general public and the management of this research (including financial
management and the setting up of a "fire wall").

In 1998, DG XIII requested the members of COST “244 bis” to update the review and
recommendations of the McKinlay report. This report, referred to hereafter as the COST
244 report was submitted to DG XIII in June 1999, following a workshop entitled "Future
European Research on Mobile Communications and Health", that was held in Bordeaux in
April 1999. Researchers from academia and from industry worked together towards the
realisation of this task. There follows a critique of the two reports.
1. Studies on the nervous system and behaviour (47 studies in 1996; 18 supplementary in

a- In vitro studies (11 studies reported in 1996)

Summary of conclusions: the few studies carried out on nervous tissue preparations were
not conclusive because a thermal effect could be excluded, except where thermoregulation
was effected, in which case the results were negative.

Previously reported studies on calcium efflux from brain tissue preparations exposed to
microwaves are also not conclusive as the exposure conditions were not well defined and
replication studies proved negative.

Calcium: Several studies reporting calcium ion efflux from cellular membranes were carried
out on chick brain. The effect was only observed at modulation frequencies between 6 and
20 Hz. Some studies reported that the effect could only be observed for exposure to signals
within a certain power range, this is referred to as “window” effect. Some replications
studies proved negative.

Long term potentiation: Scott and Tattersall (1999) studied the effect of exposure to 700,
900 and 1800 MHz (continuous and GSM) for 10 minutes on rat brain slices. They reported
a change in their long-term potentiation (a memory related mechanism) which occurs
irrespective of the nature of the signal.

Gene expression: Ivaschuk et al (1997) exposed PC12 nerve cells previously treated by
NGF15 to an 836.55 MHz (signal TDMA16, SAR of 0.41 - 41 mW/kg) for 20, 40 and 60
minutes. There were no modifications to the expression of the oncogene c-fos while that of
c-jun was increased at the highest SAR level, but only after 20 minutes of exposure.

b- Animal studies (34 studies reported in 1996, 6 supplementary in 1999, 12

Summary of conclusions:
Electrophysiology: The “microwave auditory effect” is well studied and by now well
understood. It is a direct stimulation of the cochlea by a thermoelastic pulse generated by
the absorption of microwaves in soft tissues of the head. The technology of mobile
telephony does not produce pulses of sufficiently high energy content to instigate this

Neurotransmitters and hormones: The available data appear to indicate that the effects
observed are thermal in nature. Recent studies on the level of neurotransmitters in the rat
(dopamine, noradrenaline) proved negatives.

Blood brain barrier: Numerous studies have been performed on the possible effects of
microwaves on the permeability of the blood brain barrier (BBB). Most of the older studies

nerve growth factor
pulsed signal: time division multiple access

were carried out at exposure levels that illicit thermal responses, but the more recent studies
with mobile telephony signals are more difficult to evaluate. The effects are inconsistent
except at high exposure levels. It is therefore conceivable that the effects are due to an
increase in the arterial pressure caused by the stress of confinement in exposure systems.

Fritze et al (1997) showed that microwaves from mobile phones do not cause protein
leakage at SAR levels below 7.5 W/kg in rats with their heads exposed inside a “carousel”.

The group of Salford and Persson carried out a long series of studies in rats with whole body
exposures to different signals. They reported a 50% increase of permeability at SAR levels
of less than 0.3 W/kg (GSM 900 and 1800; Persson et al., 1999). Beyond this threshold, the
permeability increased rapidly with power.

Recently, Tsurita et al (1999) exposed rats in a carousel at 1.439 GHz and did not detect
any increase in the permeability of the BBB for SAR values up to 10 W/kg

Memory: One study, from 1994, showed deleterious effects on the learning of rats exposed
to pulsed microwaves (work memory). Since then, only one replication study has been
attempted with a weak signal at 900 MHz and this proved negative.

In 1999, Sienkiewicz et al tried to repeat the work of Lai on mice exposed for 45 minutes
(GSM 900, SAR of 0.05 W/kg). The animals were then placed in an eight arm maze, no
effect of learning was observed at this low power level.

Gene expression and stress studies: Several studies have been carried out on heat shock
proteins (ARNm). Only exposure to high intensity pulsed or continuous radiation resulted in
an increase in gene expression. In the brain, the expression of the c-fos gene corresponds
to stress of thermal or other origin. The level of c-fos was used as a metric in two studies
that showed effects only at thermal levels.

Fritze et al (1997) studied stress, papooses and cell proliferation in rat brains exposed to
mobile telephony signals. Morrissey et al (1999) carried out similar studies on mice. In both
cases, increases in the expression of c-fos and other genes were only observed at the
highest SAR level (about 7 W/kg).

c. Human volunteer studies (4 studies reported in 1996, 8 supplementary studies in

Because of the proximity of the mobile telephone to the head of the user, a significant part of
the energy absorbed is in tissues associated with the brain. Several electrophysiological
and neurofunction studies have been made, mostly on human volunteers.

Summary of findings: Studies on EEG and sleep showed minor changes. Such studies
should be repeated under more rigorous conditions to confirm these results and enable a
clear interpretation.

Comments of the expert group on effects relating to nervous system and behaviour:
The two reports are very comprehensive and show the significance of this field of study in
the context of health effects of mobile phones. None of the studies carried out in vitro
showed evidence of any effect at non-thermal levels; nevertheless, some of these studies
should be duplicated. For animal studies, it is agreed that the effects on the permeability of
the blood brain barrier should be studied in a consistent manner and that stress markers are
a useful measure of effect on brain function. In human volunteer work, it is reasonable to
recommend sleep studies with diurnal exposure, this has not yet been carried out.

2. Studies on cancer
Experimental studies
Summary of conclusions:
Genotoxicity studies in vitro (11 studies reported in 1996, 9 others in 1999, 2 ongoing
or expected).
Most of the results obtained prior to 1996 were negative for non-thermal exposures. Recent
studies on DNA damage, sister chromatid exchanges, micronuclei and cellular cycle were
equally negative. Some results indicate a possible synergy between microwaves and other
mutagenic factors.

Low-level effects were reported by Maes et al (1996, 1997), showing a possible synergy
between RF fields and a mutagenic agent (mitomycine C) on human lymphocytes (935.2
MHz CW and GSM, 2 hours exposure at 0.3 - 0.4 W/kg.)

Phillips et al (1998) observed opposite effects on lymphocytes (American mobile phone

signals). The SAR levels were very low (2.4 - 26 mW/kg).

Studies of genotoxicity in vivo (11 studies in 1996, 7 supplementary studies in 1999, 3

ongoing or expected).

Most studies were carried out at high SAR levels corresponding to thermal effects. However
one series of studies on fragmentation of DNA in rat cells exposed to pulsed microwaves,
produced positive results that have not been replicated since.

In 1996, Lai and Singh published the results of DNA fragmentation in rat cells. The
exposure was whole body, for 2 hours, to pulsed or continuous microwaves (2450 MHz,
SAR of 1.2 W/kg). The test method known as “comet assay” was used and the number of
breaks of DNA was increased significantly. In 1997, the same group reported an inhibition
of the effect by antioxidants.

Malayapa et al (1998) could not replicate the effect using the same protocol or the same
exposure system (continuous microwaves). Since then, the same group failed to replicate
the experiments of Lai and Singh with pulsed microwaves (Lagroye et al 1999). Similarly,
the experiment by Hook (1999) on the same biological model but with signals similar to
those of mobile telephony was negative.

In vitro cancer studies (7 studies reported in 1996, 9 others in 1999, 15 ongoing or

Cellular proliferation
Several studies on cellular proliferation were carried out on cultures exposed to microwaves;
the results were not consistent and, if positive, not confirmed.

The group of Cleary described effects on the proliferation of CHO cells (Cao et al, 1995) and
of lymphocytes (Cleary et al, 1996), exposed to high intensity microwaves (2450 MHz, 25
W/kg) under isothermal conditions. Subsequently, Shi et al (1999) were unable to observe
the same effect using a similar exposure system with better temperature control.

In experiments on mastocytes exposed to microwaves at 835 MHz and 8.1mW/cm2,

Donnellan et al, in 1997 observed an increase in proliferation. The corresponding SAR was
not reported and therefore the possibility of heating cannot be excluded.

Recently, Stagg et al (1997), exposed C6 glioma cells to 836.55 MHz (pulsed mobile phone
signal) for two weeks at a low level (mW/kg). A slight increase in the incorporation of

tritiated thymidine was observed in some of the trials; however, the proliferation, measured
by counting cells, was not affected.

Cellular transformation
Following the 1991 study of the Balcer-Kubiczek group, showing a positive effect of
modulated microwaves together with other tumour promoters on the transformation of
C3H10T1/2 cells, two other cellular transformation studies carried out with mobile telephone
signals reported negative results (Cain et al, 1997; Malyapa et al 1997).

Enzymatic activity
A series of studies carried out on low-level microwaves modulated at low frequency showed
changes to the activity of intracellular enzymes implicated in cancer promotion, without any
significant effect on DNA synthesis. This applies particularly to the work of Litovitz's group
(Krause et al, 1997; Litovitz et al, 1993, 1997) on the activity of the enzyme ornithine
decarboxylase (ODC). When such effects are observed, their magnitudes are much lower
than the magnitudes of effects produced by other cancer promoters.

In vivo cancer studies (10 studies reported in 1996, 14 supplementary studies in 1999,
1 ongoing or expected).
Different animal cancer models were used to determine the possible effects of microwaves
on the three phases of the development of cancer (initiation, promotion and progression).
Other models were also used, such as tumours induced chemically or by means of ionising
radiation and transgenic animals that are more susceptible to the development of tumours.

Taken as a whole, the results of studies on different animal models do not show a significant
effect of microwaves on the formation of cancer and in particular on the promotional phase
that was most studied.

In 1997, the Australian group led by Repacholi published positive results on transgenic mice
carrying lymphoma exposed for a period of 18 months to GSM signals (whole-body SAR of
0.008 to 4.2 W/kg). A significant increase in the incidence of lymphomas was observed.

Toler et al, 1997 and Frei et al, 1998 exposed mice that are susceptible to develop brain
tumours to 2450 MHz (up to 1 W/kg) for the lifetime of the mice. No difference was
observed in the frequency or the stage at which the tumour appeared, and no change was
observed in the longevity of the animals.

The results by the group of Adey (1999) on brain tumours induced in rats were all negative,
irrespective of the mobile telephone signal used for long term exposure of the animals.

Wu et al (1994) did not observe any change in the development of colon tumour in mice.
Imaida et al (1998) did not report any effects of microwaves on liver cancer in the rat.

The group of Bartsch and Anane, studying rats with DMBA-induced tumours subjected to
low-level GSM signals, reported reductions in the incidence of tumours or a delay in their

c- Epidemiological studies (11 studies cited on brain tumours, 16 on leukaemia)

Summary of conclusions:
A summary is given of the principles of epidemiology and its application to microwave
exposures. Most of the studies were carried out under conditions not sufficiently well-
defined to enable an evaluation to be made of the exposure, and only one study related to
mobile phone exposure. It is therefore impossible to reach a clear conclusion, but precise
recommendations have been made with respect to the type of study that needs to be carried
out: case-control studies of cancer in the head but not breast cancer or leukaemia.

Epidemiological studies related to exposure from base stations are not recommended. Also
not recommended is the study of other non-cancerous diseases and subjective ailments,
given the current state of knowledge (pending the results of studies carried out in Sweden
and Norway).

In 1996, Rothman et al published preliminary results of mortality rates of subscribers to an

American mobile telephone network (255,868 people). A comparison was made between
users of cordless versus mobile phones. For regular users during the 3 years leading up to
the study, the odds ratio for mortality was 0.86 (0.47-1.53). The short follow-up period of
one year and the fact that the cause of death was not reported constitute major weaknesses
of the study.

Comments of the expert group on effects related to cancer: The data presented are
exhaustive and take into consideration the studies ongoing in 1999. In the light of
information acquired in 1999, it is possible to concur with the authors' conclusion that
exposure to microwaves does not, in general, induce lymphoma and other tumours.
However, the question of promotion remains open. It is therefore important to settle the
issue as the results to date are inconsistent.
The recommendations of the report concerning epidemiological studies were incorporated in
the protocol of the international study led by IARC. The conclusions of the (French) expert
group agree with those in the report under consideration with respect to studies related to
base stations.

3. Studies on the cardiovascular system

Summary of conclusions: Few studies have been carried out that are relevant to mobile

In 1999, Braune et al published results showing a 10% increase in arterial blood pressure of
volunteers exposed to GSM signals for 35 minutes. Unfortunately, the interpretation of these
results is made difficult by the lack of randomisation in the sequences of real and sham

Comments of the expert group concerning the cardiovascular system: the effects
observed after exposure to high-intensity microwaves are probably due to heating. Studies
related to exposure from mobile telephony are very limited in number; it is therefore not
possible to give firm conclusions in this respect.

4. Studies related to reproduction and development

Summary of conclusions: the subject was not discussed.

5. Studies relating to the immune system

Summary of conclusions: In vivo studies on the immune system (8 studies in 1996, 2
more in 1999).

Most of the recent results from in vitro studies of immune responses were classified in the
chapter on genotoxicity and cancer. In these experiments no effects were reported on the
immune system (for example, Antonopoulos, 1997; Eberle, 1997).

In vivo studies of the immune systems (11 acute studies reported in 1996, 5 chronic
and 2 more in 1999).
Most of the immune systems studies to date have been on animals. Most were coupled with
studies of cancer promotion. Few studies related specifically to exposures to mobile
telephony signals. The conclusion of the report was that only high exposure levels, causing

thermal effects, could have a permanent effect on immunity. The role of stress should be
studied in detail and so should immune response in the skin.

In 1999 Chagnaud et al published some negative results from the study of certain immune
system parameters of rats exposed to GSM signals (activation and under-population of

Comments of the expert group concerning immunity: Given that only very few results
were relevant to the issue of immunity, it is not possible to conclude that mobile telephone
signals have an effect on the immune system. Nevertheless, there is a need for a better
understanding of thermally-induced stress (even in the absence of an increase in body
temperature). There is also a need to study possible effects on skin in its capacity as an
organ of the immune system.

6- Studies on interference with medical implants

Summary of conclusions:
Prior to 1996, most of the work in this area concerned pacemakers. Some studies have
shown that certain types of pacemakers do not function properly very close to a mobile

Several studies have shown that that there is no interference with pacemakers when there is
sufficient distance between the device and the mobile phone (Barbaro et al, 1995; Carillo et
al, 1995; Hayes et al, 1995; Meckelburg et al, 1996).

In 1999, the subject was not discussed further, to some extent because the problem was
considered to be resolved.

Comments of the expert group on implants: It is agreed that the question of interference
with pacemakers does not arise with modern devices and the advice to keep mobile phones
at least 15 cm away. However, the issue of cochlea implants has not been resolved.

(there is no #7 in the French text)

8. Studies on other effects
Inner ear (2 ongoing animal studies in 1999 and several human studies proposed).
Summary of conclusion
There are ongoing studies. Others must be conducted to ensure the safety of the auditory

General conclusions of the expert group on the McKinlay/COST 244 bis reports: These two
reports constitute a coherent and comprehensive view of the state of knowledge. The
participation of a very large number of experts in the drafting of these reports meant that a
large spectrum of opinions was obtained. It should also be pointed out that these reports,
and in particular that of 1996, served as a basis for the call for proposals by the European
commission within framework V.
Tables summarising the results of studies are sufficiently comprehensive. The tabulation of
all ongoing studies and details of the research groups concerned is a useful and original
aspect of these reports. The recommendations for research are carefully drawn.
The weaknesses of these two reports relates to the chapter on cardiovascular system and
that on reproduction and development, which is non-existent.

The expert group concurred with the general conclusions of these reports with respect to the
absence of confirmed health effect from mobile phone technology and on the direction of the
research program required to fill the gaps in knowledge and to provide for a rigorous risk

4.1.4 Analysis of the Report Of The Royal Society Of Canada
This report, submitted to the Minister of Health of Canada in March 1999, is entitled: "A
Review of the Potential Health Risks of Radiofrequency Fields from Wireless
Telecommunication Devices"
A group of eight North American experts led by Professor D. Krewski of Ottawa, prepared a
report on the potential health risks of RF fields from mobile telecommunication systems. It
is intended to inform the government on the validity of the scientific basis of the Canadian
"Safety Code 6" health standard, developed by Health Canada, the organisation in charge of
setting limits on the exposure of workers to various environmental factors.
The criteria for selecting members of the group were set up by a Canadian committee on the
selection of experts; successive versions of the report were submitted to the committee who
arranged for another independent expert group to peer-review them. The final report was
then published without any modifications by the mandating organisation, the Royal Society
of Canada.

It is recognised, at the beginning of the report, that the field levels measured near base
stations are significantly lower than the limit in Safety Code 6 while those from mobile
phones could be of a similar order of magnitude.

The committee dealt with a number of important questions, summarised here:

Do current safety limits protect both the workers and the general population from the
thermal effects associated with exposure to RF fields?
No harmful effect is known to occur below the limits for whole body exposure. On the other
hand, the limits for partial body exposure are clearly higher and allow longer exposure times.
Consequently, the committee cannot confirm that these levels of exposure fully protect
workers from thermal effects on all parts of the body. Further research is required to
ascertain the maximum duration for safe exposure.

What are the non-thermal biological and health effects?

There is evidence that non-thermal biological effects exist, but the data available are not
sufficient to rule out the possibility that these non-thermal biological effects could lead to
undesirable health effects. In view of the lack of knowledge of non-thermal mechanisms it is
not possible to set exposure limits on the basis of these biological effects.
No non-thermal health effects are expected from exposure to base stations. In contrast
mobile phone users may experience biological effects, but there are no known associated
health effects.

There is at present no scientific basis for concerns over cancer. Clinical studies have
investigated cancer, problems with reproduction or development, epilepsy and headaches
Taken as a whole these studies were not conclusive but more research is needed.

Is there evidence that non-thermal effects, if they exist, are more important for
children or other sub-populations?
Certain sub-populations such as children, pregnant women and the elderly are generally
more sensitive to certain environmental health effects. Very few studies have been carried
out on these groups in relation to exposure to RF fields. The committee has not found
evidence in support of a microwave health effect syndrome. However, it seems that some
people are able to detect that they are exposed to RF fields.

What are the conclusions of the committee with respect to exposure levels, in
particular, should non-thermal effects be considered in Safety Code 6?
Safety Code 6 protects workers and the public from thermal effects; the committee has
however pointed out that, under the partial body exposure conditions, some parts of the
body of the workers such as the head, the neck, the trunk and the extremities can

experience some heating. The committee recommends the revision of partial body
exposure levels and exposure time for workers.

Current knowledge of non-thermal effects does not provide sufficient evidence of health
effects that can be incorporated in a revision of Safety Code 6. Partial body exposure limits
should be applied to the eye.

What research is needed to better understand non-thermal effects of RF?

Research is needed on the potential health effects of RF fields and in particular on the
sensitivity of certain groups. It is recommended that epidemiological studies should be
carried out, even if it is difficult to evaluate the exposure, because the timescale over which
mobile telephone have been used is too short to allow the identification of possible long-
term effects.

In the following critique of the Canadian document the deliberations and conclusions of the
Canadian group are reported in the order in which they appear in the document.

1. Studies on the nervous system and on behaviour

The possibility that microwaves could interact with brain tissues resulting in nonthermal
effects was suggested by results of studies from the former Soviet Union. The main issue is
whether brain tissue, being the centre of electrical activity, is more susceptible to
microwaves than other tissues. If this is the case, could the exposure trigger or enhance
nervous problems?

a-Studies on animals: summary of conclusions:

Calcium ions: a long series of experiments carried out in several laboratories has
demonstrated an increase in calcium efflux from cellular membranes under certain

Blood brain barrier (BBB): increases in the permeability of the blood brain barrier have
been observed under exposure to microwaves. These are effects that could have important
health consequences.

Brain metabolism: A series of studies, from the same research group, have shown
changes in neurotransmitter receptors and imply a role for endogenous opiates in
microwave effects.

Behaviour: Some studies have described a memory effect on rats exposed to pulsed

Most of the effects observed in animal studies occurred at high (thermal) exposure levels,
except for those on calcium ion efflux. Behavioural studies are generally positive but the
extrapolation of their results to man is difficult.

An increase in calcium ion efflux from cellular membranes of brain tissue was reported by
Blackman et al (1979, 1980). The exposure was to low-level RF (147 MHz, 0.5 mW/cm2)
modulated at 16 Hz.

D’Andrea's group exposed rats for 90 days (7 hours/day) to microwaves (2450 MHz,
0.5 mW/cm2) and observed changes in their performance of repetitive tasks (DeWitt,
D’Andrea et al, 1997)

Lai, Horita et al (1994) studied the role of endogenous opiates on memory deficit induced by
exposure to RF in rats using a radial arm maze. The retarded learning could be blocked by
pre-treatment with either a cholinergic agonist or an opiate antagonist.

The group of Salford and Persson (1992, 1994) demonstrated that pulsed or continuous
microwaves (915 MHz, SAR of 0.016-5 W/kg, modulated at 8; 16; 50 or 200 Hz) increased
significantly the permeability of the BBB to albumen in exposed rats. The increase was quite
evident at the highest SAR, but still perceptible below 0.1 W/kg (that is below the values
recommended in Safety Code 6)

b- Human studies: summary of conclusions:

Epilepsy: To date, there are no reproducible data to support the hypothesis that exposure
to microwaves triggers or aggravates epilepsy.

Neurodegenerative diseases: In view of the role of the metabolism of acetylcholine in the

central nervous system (CNS), it is important to investigate if there is an association
between exposure to microwaves and diseases such as Alzheimer's. To-date, there are no
data to support a causal link between exposures to microwaves and Alzheimer's disease.
This is also the case for ALS (amyotrophic lateral sclerosis). In the case of ALS,
epidemiological studies have investigated the role of ELF exposure. Studies at RF
frequencies should be carried out.

Sleep disturbance: Results of studies on human volunteers exposed during their sleep
showed alterations in sleep pattern with exposure to mobile phone signals or signals from
lower frequency therapeutic devices. However, these results do not appear compatible with
those from animal studies.

Depression, suicide and behaviour: A microwave syndrome has been described but
never quantified. The most common symptoms are: irritability, fatigue, loss of appetite,
torpor, loss of memory, loss of concentration, emotional instability, depression and
headache. Experimental data do not confirm the existence of a link between exposure to
microwave and headaches. This is, though, the subject of numerous complaints.
Epidemiological studies on suicide have been carried out only for exposure to ELF.

The eye: The potential effects of microwaves on the eye, and in particular the production of
cataracts, have been studied for 40 years. The studies have numerous experimental
difficulties and it is not always possible to determine the mechanism by which microwaves
cause damage to eye tissues, except at high power where thermal effects are evident.

In conclusion, with the current state of knowledge it is not possible to identify a neurological
effect or an effect on brain function that can be directly attributed to microwaves at non-
thermal levels. Headaches are a common complaint in relation to exposure to RF, however
this symptom is ill-defined and difficult to relate to neurochemical changes. It is important to
carry out studies on this subject. Similarly, the unique properties of the eye call for special

In a study by Reiser in 1995, a commercial mobile phone operating at 900 MHz was used
40 cm from a person. An increase in the amplitude of the EEG spectrum was observed in
the alpha and beta bands after 15 minutes of exposure.

Sobel and Davanipour (1996) have suggested that electromagnetic fields contributed to the
process of neurodegeneration while Feychting et al (1998) have suggested that alteration in
brain tissue by electromagnetic fields predisposes to Alzheimer's disease.
In a study on sleep by Mann and Röschke in 1996, volunteers were exposed for 8 hours at
night-time, showed no change in total amount of sleep and no change in total amount of
slow wave sleep. However, there was a reduction in the period of REM sleep. In 1998, the
same group failed to reproduce theses results at lower exposure level.

A recent study by Hansson-Mild in Sweden and Norway in 1998, seemed to confirm that
headaches are a subjective symptom often reported by users of mobile phones.
In 1997, Kues et al studied vascular permeability in monkeys exposed to 2450 MHz
microwaves at for 3 days. A dose related effect was found between the microwave-induced
vascular permeability and corneal lesions that developed subsequently.

Comments of the expert group on nervous system and behaviour: The conclusions of
the Canadian report identify gaps in knowledge, particularly in relation to headaches and the
eye. They reviewed and took into consideration most of the published studies but their
analysis of animal and in vitro studies is not very rigorous, and there are some errors in
interpretation. There are data supporting an effect of RF on animal behaviour, but the
interpretation of these results in terms of relevance to human exposure cannot be done on
the basis of current knowledge.

When the Canadian report was being drafted, some recent results on sleep and cognitive
tasks were not yet known. Consequently, some of their conclusions are not up-to-date.

2-Cancer Studies

a-Experimental studies: summary of conclusions:

Genotoxicity: Taken as a whole, the published results do not provide proof of the
existence of a genotoxic effect from exposure to microwaves from mobile telephony. No
change was observed in mutation rates in vitro. Studies on chromosomal aberration were
inconsistent (14 studies reviewed of which 5 were positive). The proportion of micronuclei in
exposed cells is sometimes significantly increased (9 studies reviewed of which 7 were
positive). Cell cultures exposed in vitro did not show alterations in the DNA (10 studies
reported, all negatives). By contrast, animal studies gave different results (10 studies
reported of which 6 were positive). Studies on cellular transformation did not give consistent
results (4 studies of which 2 were positive).

Cellular proliferation: The effect of microwave exposure on cellular proliferation has been
studied under a number of different conditions. Where the exposure conditions, and in
particular temperature, are rigorously maintained, the results were almost always negative.
ODC: Ornithine decarboxylase, a catalytic enzyme involved in the formation of polyamines
in cells, plays an important role in the promotional phase of the process of carcinogenesis.
An increase in ODC activity has been observed with exposure to low frequency modulated
microwaves; this could be a type of “window effect”, as the increase in ODC activity is
reported by certain authors at very specific modulation frequencies (5 studies reported
positive results with exposure to modulated microwaves).

Melatonin: Some studies carried out at ELF reported changes in the night-time level of the
hormone melatonin in rodents. In view of these results, some studies were later carried out
with exposure to microwaves. No effects were observed on people exposed during the
night. In the case of exposed rats, there were also no effects on night-time melatonin level.

Initiation, promotion and progression of tumours: Some studies have reported an

increase in the incidence of tumours while others have reported a decrease in cancer risk.
There is no evidence for a microwave effect on cancer promotion or on acceleration of the
progression of cancer. However, because the picture is mixed, additional studies are

One positive result has been reported on the induction of sister chromatid exchange in
human lymphocytes exposed to 167 MHz (Khalil et al, 1993). It is possible that, in this
experiment, the effect was due to RF heating.

Lai and Singh (1995), used the test procedure known as comet assay to observe an
increase in the fragmentation of DNA in rat brain cells exposed at 2450 MHz. These results
have not been independently replicated despite several attempts by other laboratories.

The results of Balcer-Kubiczek and Harrison (1985; 1989; 1991) on the transformation of
C3H/10 T1/2 cells showed that the transformation did not occur with microwave exposure
alone. However, microwaves could act in synergy with other mutagens or promoting agents.
C6 cells and stem cells of rats were exposed by Stagg et al (1997) to mobile telephone
signals (TDMA) for 24 hours at low SAR levels. An augmentation in DNA synthesis was
observed at 5.9 mW/kg but not at 0.59 or 59 mW/kg; these are very low SAR levels.

Litovitz's group, among others, have reported an increase in ODC activity in cells exposed
for several hours to RF (835-915 MHz, 2.5 W/kg). The increases were observed only when
the signal was modulated at 16 and 60 Hz. The increase in ODC activity was only

In a study on the rate of production of melatonin, Vollrath et al, (1997) did not find an effect
in rats exposed for 6 hours to low level microwaves (900 MHz, 0.06 to 0.6 W/kg).

In 1997, Repacholi et al published results on transgenic mice exposed to GSM 900 signals.
The exposure increased the incidence of lymphoma (x 2.4) at the end of 19 months of
exposure. The average SAR ranged from 0.008 to 4.2 W/kg. It is expected that these
experiments will be repeated in other laboratories.

b-Epidemiological studies: summary of conclusions:

Epidemiological studies on exposure to RF are very scarce, very diverse and rarely relevant
to exposure from mobile telephony. There are studies on child and adult cancer, on
reproduction and on congenital defects. For most studies, the evaluation of exposure is not
an easy task. Occupational exposures are usually the highest (military and police
personnel, electrical, medical and other industrial workers were studied), the results are not
consistent and none of the studies relates to mobile phones. None of the epidemiological
studies allows for an estimate to be made of the risk to children.

In a cohort study carried out in Poland in 1996 by Szmigielski, the dosimetry was carried out
on volunteers under well-defined conditions. An increased incidence of cancers of the
haematopoietic system, lymphatic organs and nervous system was observed. The
methodology of this study has been the subject of severe criticism. In the UK, following the
detection of a cluster of adult leukaemias in the neighbourhood of a radio and television
mast, two studies were conducted, one in the vicinity of the original mast, and the other
covering the whole of the country (Dolk et al., 1997). A non-significant increase of 3% in the
incidence of cancer was observed in the vicinity of the mast irrespective of the distance.
The only significant increase was for leukaemia (OR: 1.83 [1.22-2.74]). By contrast, in the
study carried out all over the country no association was observed. The problems with
such studies are in the difficulty of assessing the exposure of people as a function of
distance from the mast and in determining the population movement during long observation

Rothman et al, (1996) measured the overall mortality of 255,868 users of mobile phones
(high power) and cordless phones (low power). No difference was observed between the
two populations. This study was criticised because of the short follow-up period and the lack
of exposure details provided by the operators.

Comments of the group of experts on the effects related to cancer: A very large
number of references are quoted but the information is not very well organised: different
aspects of the process of carcinogenesis and health consequences are in different non-

consecutive chapters. The authors of the report have paid particular attention to the risk of
cancer. They discuss at length, and may have over-interpreted, experimental studies on
ODC. This is presented as a possible “missing link” in a carcinogenic mechanism. The
available data are confused, and a hypothesis alone is not a proof. Given the current state
of knowledge more experimental data are needed. In fact, the inconsistency in the results is
pointed out as is the fact that most of the studies were negative. More specifically, on the
basis of current knowledge, it is possible to conclude that microwaves do not have a
genotoxic effect. A promoting effect cannot be excluded, in fact this is the hypothesis at the
basis of the international study by IARC. It is a matter for regret that the possible synergy
between RF and mutagens, or other agents, was not sufficiently considered.

The importance of exposure conditions in the assessment of the relevance of experimental

and epidemiological work, including how experimental conditions compare with mobile
telephony exposures, is not sufficiently underlined.

3. Studies on the cardiovascular system: summary of conclusions:

The subject is treated very briefly. Studies carried out nearly 60 years ago in the former
Soviet Union showed alterations in arterial blood pressure and some anomalies in rhythm.

In a recent study (Bortkiewicz et al, 1997) reported that some electrocardiogram anomalies
were more frequent in exposed volunteers than unexposed volunteers, (71 employees were
included in the study).

Comments of the group of experts on cardiovascular system: It is not possible to make

any conclusions on the basis of the limited data available.

4. Studies on reproduction and development: summary of conclusions.

The Canadian report draws no conclusion on the subject; it is not treated in depth.

Eight studies are presented, of which 5 related to occupational exposure among

physiotherapists exposed to microwaves. The studies reported on a variety of reproduction
effects, such as miscarriages, congenital malformations, prematurity, or birth weight. The
exposure conditions are not those of mobile telephony, the results are inconsistent, and the
numbers too small to lead to a conclusion. Some of these studies are suggestive (in a non-
significant manner) of an increased risk in cases of exposure to RF from diathermy

One study, carried out in Lithuania, was instigated by an observation of a decrease in male
births in the neighbourhood of a radar station emitting 154-162 MHz fields, pulsed at 24.4
Hz. A number of psychometric and neurological tests were carried out showing differences
as a function of proximity to the station, but the mobility of children aged 9 to 18 years was
not taken into consideration: this dilutes the relevance of any conclusions drawn from this

Comments of the group of experts on reproduction and development: The subject is

not discussed fully. Some occupational studies of physiotherapists exposed to microwaves
are discussed in some detail, but the data are inconclusive. This is a weakness of this

5. Immune system studies: summary of conclusions.

This subject is not treated separately; some data on the immune system are reported in the
chapter on cancer.

Comments of the group of experts on the immune system: It is not possible to reach a
conclusion on the basis of existing data.

6. Studies on interference with medical implants: summary of conclusions
The subject was not treated in this report.

7. Other effects: summary of conclusions

Possible effects on the eye:
The very few available results of studies on the effect of microwaves on the eye are reported
in a short chapter: these studies originate from the two groups of Kues and Kamimura.
Kues reported on the incidence of cataract in the eyes of monkeys exposed to pulsed or
continuous microwaves at high power, while Kamimura did not observe any deleterious
effects of continuous microwaves. The results of Kues have not been reproduced
elsewhere, however it should be noted that there were no attempts at replication under
rigorous condition. In conclusion, in view of the low vascularisation of the eye and of its
prominent position in the face, it is recommended that further studies should be carried out
to assess the risk to the eye and to establish more limiting exposure conditions compared to
partial body exposure for other organs. Consequently, a limit of 1.6 W/Kg for workers and of
0.2 for the public is recommended.

Mobile phones and driving:

This is not a direct risk of microwaves but the authors of the report point to the increased risk
of car accidents on the basis of a Canadian study of 699 subjects involved in a car accident.
The risk of accident without casualties was as high as that of driving under the influence of
alcohol (level of alcohol in the blood at the authorised limit), with a relative risk of 4.3 [3.0-
6.5], irrespective of whether the driver was using a hand-held or hands-free device. It is the
distraction that is responsible for accidents

Microwave syndrome:
This syndrome, previously described, has not been objectively confirmed. However, it
would be desirable to undertake a double-blind study of the possible sensitivity of some
people to RF fields.

Comments of the group of experts on effects on the eye: The importance given to the
eye in this report is in part because the standard is primarily developed for workers who
under certain circumstances can be exposed to relatively high-power microwaves.
However, it is not reasonable to give so much importance to this subject for the following

· The results of Kues are contradicted by those of Kamimura,

· They were obtained at high (in comparison to the power levels of mobile telephones and
walkie-talkies) power levels, and they are largely contested.

· The exposure levels proposed are lower than those for other, less vulnerable, organs or
the extremities but quite compatible with the use of mobile phones. In fact, because of
its position, the eye is not directly exposed to the fields from mobile phones.

General conclusions of the group of experts on the Canadian report:

This report is not very up-to-date and many recent experimental data are missing. On the
whole, the report is not well-balanced: some subjects are treated in great detail while others
are only skimmed over. The bibliography is comprehensive but not well organised by
subject. This is probably due to the small size of the group of experts and the limited
number of fields of study represented.

This report dwells at length on biological effects resulting from exposures to RF at levels
well-below those known to give rise to thermal effects. Nevertheless, the report is prudent in
its interpretation of these observations in terms of adverse effects and health risks. The
possibility of differentiating between biological effects on the basis of RF modulation
characteristics is reported, and deserves further exploration. Certain mechanisms are well
developed but remain hypothetical, indicating that further research is needed to reduce
uncertainties related to localised exposures and long-term effects.

4.2 Additional reports

4.2.1 Analysis of the Report of the French Academy of Science

Based on the international symposium “Mobile communications, biological effects” held at
the Collège de France on 19 and 20 April 2000 organised by the Academy of Sciences, the
Council for the Applications of the Academy of Science (CADAS), the National Academy of
Medicine and the Foundation for the Development of Science and its Applications (FDSA)

The aim of the symposium was to consider, as exhaustively as possible, all issues related to
biological effects from exposure to mobile communications signals. It was organised in six

Session 1, dealing with dosimetry, was the responsibility of Joe Wiart of France Telecom
Research and Development. It was chaired by Pierre Aigrain of the Academy of Science
and of the CADAS.

Session 2 was dedicated to human studies; Jean-Louis Coatrieux and Alain Bardou of
LTSI-INSERM at the University of Rennes were responsible for this session. It was chaired
by Claude-Henri Chouard of the National Academy of Medicine.

Session 3, dedicated to in vitro animal studies, was the responsibility of Bernard Veyret of
PIOM, Bordeaux and was chaired by Charles Pilet of the National Academy of Medicine, the
Academy of Sciences and the CADAS.

Session 4, dedicated to epidemiology, was the responsibility of Elizabeth Cardis of IARC in

Lyon and was chaired by Raymond Ardaillou of the National Academy of Medicine.

Session 5, dedicated to technical, economic and regulatory affairs, was the responsibility
of Luis Miro of the Faculty of Medicine at Nimes and was chaired by Jacques Joussot-
Dubien of the Academy of Science.

Session 6 dedicated to risk management of a potential hazard, was the responsibility of

Claude Gilbert of the CNRS and was chaired by Pierre Fillet of CADAS.

In line with the aims of the group of experts, the review of this document will be confined to
reported work on biological and health effects of RF. This does not, of course, mean that the
other contributions are less valuable. However, they are more concerned with risk
management, and an interested reader can refer to the full report published by the Academy
of Science. The references cited in this review are given in the annex.

The first chapter in this section related to the principles of electromagnetism as they apply to
dosimetry and the interaction of RF with people. Joe Wiart states in the introduction that
dosimetry, or the quantification of fields and power induced in biological tissue, is an
important aspect of the study of the interaction of electromagnetic fields with people. As
previously reported by the World Health Organisation EMF project [917], any analysis of
possible biological effects of electromagnetic waves must refer to the levels of electric and
magnetic field strength and power induced in tissues. Many of the earlier studies of
biological effects of electromagnetic fields did not incorporate rigorous dosimetric analyses,
which makes their interpretation difficult. The experts of DGXIII [576] recommended in their
report that dosimetric analysis should be a requirement for all in vivo and in vitro studies.

Dosimetric studies are also essential for the testing of compliance to international guidelines
and standards of exposures from mobile phones and associated systems. The European
Commission has issued a Recommendation in relation to the limitation of exposure of
people to electromagnetic fields (1999/519/CE)[224]. This document is based on the
scientific recommendations [386] of the International Commission on Non-Ionizing Radiation
Protection (ICNIRP). This document defines basic restrictions, limits on power absorbed by
tissues that guarantee the protection of people from known health effects. It also defines
electric and magnetic field strengths that ensure that the power absorbed is less than the
corresponding basic restriction. For the general public, according to these
recommendations, the power absorbed by the whole body should not exceed 0.08 W/Kg.
Locally, the average power absorbed in any 10 grams of tissue should not exceed 2 W/kg.
In the frequency range of radio telephony, the electric field strength reference level for the
general public is 41.1 V/m at 900 MHz and 58 V/m at 1800 MHz.

After presenting the requirements for electromagnetic dosimetry, the author discusses the
basic aspects of numerical and experimental dosimetry, recalling the characteristics of
electromagnetic fields and the coupling between electric and magnetic fields via Maxwell’s
equations. These are the principles for the numerical and experimental quantification of
electromagnetic fields and energy absorption in tissue and, in particular, the specific
absorption rate (SAR).

The second part of this document relates to the role of dosimetric analysis in the
assessment of possible biological effects from radiofrequency fields. The physical principles
of temporal multiplexing (TDMA), power control and discontinuous transmission are
described because they collectively determine the actual power emitted from mobile phones.

One chapter is dedicated to the description of a numerical model developed by CNET,

consisting of 10 tissues, (skin, bone, cerebrospinal fluid, grey matter, white matter….) and its
use in the study of the distribution of energy in different tissues. The dielectric properties
[permittivity and conductivity] of tissues used in the model are those published by Gabriel
[289]. On this basis, an emission at 900 MHz with a peak power level of 100 watts gives rise
to an average SAR over 10 grams of 1.1 W/kg [388]. Dosimetric analysis shows that the
head absorbs approximately 50% of the power emitted by the radiotelephone, the skin
absorb 15%, muscle 10%, cerebrospinal fluid 5% and the brain 13%. Most of the
absorption is localised in the zone near the ear, 30% of the energy being absorbed in a
cube, 5 cm on a side, centred on the inner ear.

The evaluation of induced fields from exposure to base stations is also considered. The
interaction of fields from base stations with people can be analysed either in terms of
incident field or in terms of fields induced in tissues. The latter approach, based on SAR, is
the only approach that is valid when the person is very close to the antenna (“near field”). In
this case, an experimental or a numerical model is developed to allow quantification of the

interaction. The former approach, of using incident fields, allows for a comparison to be
made with the reference levels in the exposure guidelines. This approach demonstrates that
the surface power density decreases rapidly as a function of distance from the source.
Beyond a few metres, the induced fields and the associated power densities are significantly
lower than international recommendations. Base station antennas, often designed as an
array of dipoles, focus the energy in one direction, this beam is emitted principally in a
direction normal to the antenna. The metrological problems of on-site measurement
originates from the multifrequency and multipolar nature of the summed emissions from all
the systems that may be present in one location. The use of probes with an integrating
diode detector is not appropriate in this situation. Methodologies based on the use of
selective isotropic antennas covering a wide frequency band, and associated signal
analysis, are being developed.

In of conclusion, the author indicates that the dosimetry of interactions of electromagnetic

waves with people and experimental animals is essential for the quantification of SAR levels
induced in tissues. Such analyses and evaluation are necessary if studies on human
volunteers or epidemiological studies are to lead to any viable conclusions; the dosimetry
must be as well defined as the study protocol so that comparisons and, if necessary
replications, will be possible.

The second contribution in this dosimetry chapter is written by Om P. Gandhi. It constitutes

a report on research carried out at the University of Utah on the dosimetry of the coupling of
electromagnetic fields and the human body, as related to both mobile phones and base
station antennas. After summarising the most recent international recommendations in
terms of SAR [388, 386, 873], the author described two human head and neck models for
use, as in the previous contribution, in numerical studies of distribution of SAR in the body
due to exposure from mobile phones and base stations [298, 668, 865]. The author
considers primarily the variation in SAR distribution with frequency (835 to 1900 MHz),
antenna length (including the most commonly-used short and helical antennas), the angle
between the antenna and the head, and variation of head dimensions. The comparison of
the results of this type of analysis with those from direct experimental measurement show an
agreement of better than 20% which, in the opinion of the author, is a remarkable validation
of the models used.

This article also describes a portable dosimetric system [506] that enables the evaluation of
the electromagnetic coupling between base stations and the human body in less than one
minute. The precision of this technique is 5 to 10% compared to a numerical evaluation
using a full FDTD (Finite Difference Time Domain) simulation with a mesh size of 3 cm at
835 MHz and 1.5 cm at 1900 MHz. The numerical evaluation would require a one hour
computational time. Such a portable dosimetric system can be used to identify rapidly the
spatial region close to base stations where the SAR is in excess of that given in international

The third and final contribution in this section on dosimetry is from Niels Kuster and Nicolas
Chavannes who report the latest progress in experimental dosimetry for the evaluation of
human exposure and for the characterisation and optimisation of exposure systems used in
biological experiments. According to the authors, the evidence for biological and health
effects from exposure to electromagnetic fields is often questionable because of poor or
incomplete dosimetry. This is the case for both in vitro and in vivo experiments. Another
important consideration is the validity of available tools for the analysis and optimisation of
the daily exposure of people [53]. The purpose of this article is to review the state-of-the-art
in experimental dosimetry, bringing out its strengths and weaknesses with respect to the
assessment of fields within the body and in exposure systems used in biological

In the last few years, numerous systems based on high performance “diode-loaded” field
probes have been described. These are highly specialised probes designed for the
measurement of fields in vivo and in vitro [705], and the determination of field polarisations
[706], etc. Major advances have been realised in terms of spherical isotropy, spatial
resolution, sensitivity, bandwidth, linearity, secondary mode rejection, and precision of
calibration, etc. [704]. The strength of experimental dosimetry resides in its ability to validate
numerical data as it can precisely determine the distribution of fields under realistic
conditions without simplifications or modifications. Weaknesses are, essentially, that it is
restricted to homogeneous or liquid media, its spatial resolution is limited (bigger than 1
mm3), its applicability is limited when dealing with small structures.

In contrast, the strengths of the most viable numerical methods (Finite Difference Time
Domain technique or FDTD) are principally its applicability to non-homogeneous structures,
its ability to produce high spatial resolutions in 3D, and its ability to evaluate the sensitivity of
the results to a number of different parameters (anatomy, posture, tissue parameters, etc.).
Under these conditions, FDTD is perfectly appropriate for the tasks of optimisation of
antenna design. Its primary weaknesses are that electromagnetic simulations require initial
simplifications compared to actual exposure conditions, and also that the FDTD method
does not allow for an assessment of uncertainty in the distribution of internal field [117].

2/ Human Studies
The introduction to this chapter was given by Alain Bardou and Jean-Louis Coatrieux
who provided a general summary of the state-of-knowledge on biological effects from
telecommunication equipment, as described in a report from the Commission of Consumer
Safety to the President of the Republic and the Parliament in 1997 [60]:

· the possibility of induction of brain tumours. Most of the studies gave negative results
(Persson and Salford [762, 764], Adey [11], Juutilainen [426]); the only study with a
different result was that of the group of Michael Repacholi (Royal Academic Hospital,
Adelaide, Centenary Institute of Cancer Medicine, Sydney, Australia), who is currently at
a WHO. The group carried out experiments on transgenic mice, genetically modified to
be predisposed to the development of lymphoma; they showed that “the risk of
developing lymphoma is significantly higher in irradiated mice in comparison with the
control group”, [734].

· modifications in the permeability of the blood-brain barrier, where the results remain
controversial (Prato [10], Salford and Presson [763, 695], Fritze et al[278]).

· modifications in the electroencephalogram, more specifically on the alpha waves and

the spectral distribution (Thuroczy [861], Von Klitzing [901] and on sleep pattern
particularly at the level of REM sleep (Mann and Roschke [554, 748]).

· changes to DNA, particularly damage to DNA in brain cells of rats exposed to

radiofrequencies. Lai [487, 488], observed an increase in single- and double-strand
breaks in DNA after exposure to pulsed or continuous microwaves at 2450 MHz, while
Fritze [279] and Juutilainen [426] both stress, on the basis of their results, that it is not
possible for emissions from mobile phones to have a genotoxic effect.

· effects on pituitary hormones. René de Sèze, having shown that there is a “slight
modification in the level of TSH during a period of a mobile telephone call”; another
possibility is the change in the level of melatonin [211, 212]

· the metabolism of acetylcholine. Lai [480] suggests a possible decrease in the sodium-
dependent uptake of choline and an augmentation in cholinergic receptors. Hossmann

[373] reports effects on the neurotransmitter system at the level of variations in the
activity of acetylcholine-esterase.

The contribution of René de Sèze on human studies is mostly concerned with the effects of
mobile phones on hearing, sleep and the nervous system. With respect to sleep, Röschke
and Mann in a study of 12 sleeping volunteers, have shown a slight decrease in sleep
latency (9.5 min. instead of 12.3 min.; p <0.005) and also a decrease in the total period of
REM sleep from 17.1 to 13.9% (p<0.05) [554]. Borbely has also shown a reduction in
waking time during sleep from 18 to 12 min (p<0.05), and an increase in the spectral content
(of ECGs) during REM sleep [40]. The exposure was alternating with stop-go intervals of 15
minutes, with a modulation at 217 Hz as well as 2, 8, 17 and 36 Hz; these frequencies are
found to a small extent in emissions from actual telecommunication systems. The calculated
peak SAR in tissue, averaged over 10 g, was 1 W/kg.

With respect to hearing, no modifications in the latency of auditory evoked potentials or in

the interval between their occurrence were reported by de Sèze et al half an hour before
and after exposure from a radiotelephone [214].

With respect to the nervous system in general, an unpublished epidemiological study [610]
has shown an association between the use of cellular radiotelephones and benign
symptoms that nevertheless affected quality of life (headaches, heating, skin irritation etc). If
such effects are caused by radiotelephones, they must be preceded by a primary
physiological interaction leading to the manifestation of the symptoms, or a pathology, which
could possibly be detected by quantitative electroencephalography (Q-EEG). The results
are variable, but Reiser in 1995 [729] have reported an increase in the spectral content in
several EEG bands. Thuroczy has found contrary results, first in the rat (1995), than in man
[862], while Hietanen found no effect after 30 minutes' exposure [30], as did Röschke's
group after 3.5 minutes [748].

With respect to cognitive tests (memory, attention, vigilance) Preece [715, 714] showed that
choice reaction time was significantly modified, while there was no change in simple reaction
time, vigilance reaction time and the immediate memory of words, numbers or images, or
spatial memory.

Koivisto has found no significant modification in any visual reaction test [446]. Krause has
shown [452] that an exposure to a GSM mobile phone signal during a task requiring the
recall of pronounced words, involving auditory working memory, increased the relative
contributions to ECG in the frequency range 6-12 Hz, with a predominance in the 10-12 Hz

Ulf Berqvist's contribution to this topic addressed mainly the possibility of hormonal and
cardiovascular effects of mobile communication signals, and the possibility of perception of
such signals by hypersensitive subjects.

Concerning hormonal effects, the author referred mostly to melatonin as it has different
modes of action and is implicated in the inhibition of tumour growth as well as in the
nycthemeral cycle. Some studies have looked at variations in the level of melatonin, using
different frequency sources [216, 555], without relating observed differences to exposure.
Other studies have investigated levels of cortisol and other hormones, reporting some weak
effects due to exposure [555, 213]. It should also be noted that there are some rather
contradictory results relating to cortisol which, Berqvist thinks, can be considered as being
without biological significance.

Braune [106] has investigated changes in circulatory parameters. Diastolic arterial pressure
in man increased by 5 mm Hg with GSM exposure and systolic pressure increased by

10 mm Hg. No change was observed in cardiac rhythm or in capillary perfusion. It should
however be noted that this study has been criticised for its lack of randomisation (the
exposure always taking place after the placebo) and also for the fact that it was conducted
with a simple, not double blind, design. However, these results have been discussed at
length in view of their possible relevance to the headaches reported by some people (see
the work of Hansson Mild in the epidemiology section of this report) and also
cerebrovascular events. Another study has not however found any changes in the
circulation in the brain after exposure to GSM signals [862].

The ability to detect radiofrequency signals emitted from mobile phones or their base
stations has been studied in certain subjects who presented themselves as suffering from
electromagnetic hypersensitivity. Definitive recent studies carried out in Finland have shown
that these individuals were not able to detect GSM signals from mobile phones.

The author concludes on this point, saying that these studies are still ongoing and that some
aspects must be followed. This would include study of subjects from different age groups,
and people suffering from variety of symptoms from using a mobile phone, as well as people
suffering from electromagnetic hypersensitivity.

3/In vitro animal studies

In his introduction to this topic, Bernard Veyret underlined the importance of animal and
cellular models: their use is justified by the extensive amount of information that could not
otherwise have been obtained. However, this approach has its limitations, the main one
being the difficulty of extrapolation to man of experimental results obtained from animals and
cellular models. All experimental conditions, including dosimetry and biological models,
must be included in very well defined experimental protocols to enable objective
comparisons of different results to be made. It is noted that none of the published results
point to deleterious health effects, however some positive results need to be confirmed and
new studies need to be designed to fill existing gaps in our knowledge. There are numerous
ongoing studies in national research programmes and in research programs funded by
industry (cf chapter VI, “state of research”).

The contribution of Zenon Sienkiewicz of the UK National Radiological Protection Board

(NRPB) consists of a comprehensive review of animal studies relating to the biological effect
of microwaves. Reviews are provided of the following fields of study: the nervous system
(including gene expression, blood brain barrier, electrical activity, neurotransmitters and
memory); carcinogenesis (several of its aspects are reported such as genotoxic effects,
incidence of spontaneous tumours, promotion, progression of implanted tumours, and
melatonin); reproduction and development.

Genetic expression
Several studies, for example those of de Pomerai et al [208], show a possible increase in the
heat shock protein c-fos suggesting the possibility of a very localised rise in temperature.
According to the author, this line of research deserves further attention and an independent

Effects on the nervous system

EEG patterns and their spectral content appear to be modified by exposure to
electromagnetic fields [151, 721, 903]. The variations in spectral content appear to be
observed at exposure levels resulting in a local increase in temperature in the brain [859].

An effect on neurotranmitters has been reported in several isolated studies from different
laboratories. Acetylcholine is an important neurotransmitter associated with learning and
memory in different parts of the brain. Exposure to electromagnetic fields appears to alter

cholinergic activity in the hippocampus and the frontal cortex [473, 447]. A prolonged
exposure appears to bring about an autoregulating increase in cholinergic receptors [480];
these effects are correlated with effects on behaviour. It has been remarked [476] that
similar modifications can be induced by stress factors such as noise or confinement.
Current thoughts lean towards the possible mediation of corticotrophine “releasing factor”
and the intervention of endogenous opiates [479, 482, 490]. Different modifications in
behaviour and learning function can be observed under different types of irradiation [485,
912, 810, 715, 445]. The author mentions, without specific reference, that recent studies
suggest that microwaves can have specific effects on cognitive performance in man; he also
thinks that complementary studies should be made on the subject, particularly on primates.

Effects on carcinogenesis
One important concern is the lack of replication [775] of the experiments of Lai [549] on
damage to DNA after exposure to microwaves. Supplementary studies carried out in vitro
[548, 547] confirmed that the effect could not be replicated. It is also important to note, that
other studies using different indicators of DNA damage [73, 778, 893, 895] tended to
absolve the mobile phone from any genotoxic effects.

The great majority of studies of the incidence of spontaneous tumours have provided
negative results [160, 870, 272, 271, 842]. According to the author, the results of the study
of Repacholi [734], carried out on transgenic mice cannot easily be extended to non-
transgenic animals, including man.

Studies from about twenty years ago have reported effects on promotion and progression of
tumours induced in the skin of mice using benzopyrene. The promotion effect has been
studied very specifically in some recent studies on chemically-induced tumours in the colon
of mice [485] and in the liver of the rat [391, 392]. There was no effect on the incidence,
number and size of tumours. The conclusion was the same for exposures of brain and
medulla tumours in the rat, whether these tumours are spontaneous or chemically-induced
using ethylnitrosourea (ENU) [15, 942].

Few studies have looked into the progression of tumours induced by injection of cancer cells
in healthy animals. Whilst most of the studies show that exposure has very little effect on
the growth of this type of tumour [765, 660], one study did describe an increase in growth
due to exposure to microwaves [900].

Possible effects on melatonin have also been studied. Because this hormone acts as
oncostatic agent, a reduction may lead to some increased risk of cancer. Two studies
reviewed by the author do not show any effect on the level of melatonin after exposure; two
studies even showed a slight increase in saliva melatonin [39, 392]. It should be noted that
in this same work are two references to papers by René de Sèze showing a possible
variation of melatonin level following use of a mobile phone [211, 212].

The author concludes that, taken as a whole, the data on carcinogenesis leave little room for
the hypothesis of an increased risk of cancer due to exposure to radiation emitted by mobile
phones. However, the few studies reporting a positive effect prevent us from indemnifying
completely mobile communications (even if it is possible to criticise the poor experimental
methodology or the difficulty of extrapolating to man); this residual scientific uncertainty
justifies the need for further, high quality, research.

Effects on reproduction
The teratogenic and testicular effects of hyperthermia are well known, as is the fact that heat
from exposure to microwaves can induce a higher level of intrauterine mortality and foetal
defects, as well as a more-or-less temporary sterility [415. 665]. Because the thermal effects
from mobile phones are extremely weak, one should expect an absence of effects at the

level to which the public are exposed. However, a study carried out on mice exposed in a
commercial antenna park showed a rapid decrease in fertility as well as decrease in the size
of the embryos [545]. The absence of a control group, and the fact that this type of result
can occur in a stressful situation, calls into question the validity of these results. This type of
study should be carried out in laboratory under well-defined experimental conditions.

The contribution of Jukka Juutilainen deals with in vitro studies on possible carcinogenic
effects of radiofrequencies. In its introduction, the author points to the importance of the
determination of the SAR in this type of study, in order to distinguish thermal from non-
thermal effects. Genotoxicity is related essentially to DNA damage. Physical agents that
can cause damage to DNA are often carcinogenic: in vitro genotoxicity is often used as a
pre-screening process for the identification of potentially cancer-initiating substances. Some
non-genotoxic substances can also have carcinogenic effects (promoters), either by
preventing the repair mechanism of DNA, or by suppressing the protective responses
towards oxidative stress, affecting apoptosis or by stimulating cellular proliferation.

Most of the studies in the literature suggest that radiofrequencies have no direct genotoxic
effect [113]. However, two studies provide positive results: that of Maes et al, at relatively
low levels of SAR [542] and that of Phillips et al that suggests a specific effect due to
modulation (TDMA and iDEN) at low exposure levels.[160]. Whilst the experiment of Maes
can be criticised for the lack of control over exposure conditions, the results obtained by
Phillips appear interesting. The author of this presentation thinks that effects at low
exposure levels should be confirmed by independent studies before they can be considered

Several studies have not shown any increase in DNA damage from a combination of a
known genotoxic agent and exposure to radiofrequencies; no effect has been observed on
DNA repair mechanisms also [701]. Some positive results have been obtained [544, 543,
781]. It is interesting to note that positive results have been obtained only if the exposure to
radiofrequencies precedes the action of the mutagen and not if it is simultaneous or
afterwards [544, 543]. Scarfi et al [781] have also shown effects from pre-exposure to
radiofrequencies, but at high SAR levels so that the possibility of thermal effects cannot be
excluded. Nevertheless, according to the author, the combined effect of radiofrequencies
and genotoxic substances must be the subject of further research.

A group of researchers has shown conflicting results of exposure to GSM-RF signals on

cellular proliferation; an increase in proliferation is described in one study, while in a second
a decrease is observed. The two studies were conducted on the same cell-types with
identical exposure parameters [470, 887]. Several studies carried out on proliferation
markers showed modifications, either of thymidine (increases the synthesis of DNA) [830], or
on ornithine decarboxylase (ODC) (key enzyme for the synthesis of polyamine) [630, 527,
689]. The work on ODC has not however shown any evidence of an effect due specifically
to the modulation of the amplitude of the signal [527, 689], or of the frequency of the signal
[120]. The author of this article considers that the works on ODC are potentially very
important and that this field of study should be actively pursued.

Two studies suggest that genetic expression could be modified by an exposure to low level
radiofrequencies [323, 402]. These two studies giving positive but different results, one
showing an effect on the fos proto-oncogene , the other on the jun proto-oncogene, and they
cannot be directly compared because of the different cell types used as well as the
differences in the level of exposure. Although these two studies show a possible positive
effect, the author considers that, in practice, their significance is not very clear.

Three models for in vitro transformation show a heightening of a chemical promoter (TPA)
[49, 50, 51], but these three studies, originating from the same research group, have not

been replicated independently. A study by Cain et al, shows no effect at a low SAR level
[125]. However, as previously noted, the differences in exposure parameters and
methodologies make direct comparison between the different results impossible.

The general conclusions of the author of this chapter are:

· Available data do not suggest direct genotoxic effects, but studies on indirect effects and
other effects must be pursued.

· It is difficult to reach a conclusion with respect to non-genotoxic carcinogenicity, as very

few studies have been carried out in this field.

· The reported studies used different biological models and exposure parameters, making
it difficult to carry out a comparison.

· There is no evidence in the studies carried out to date to prove an effect of amplitude
modulation, however, it is not possible to exclude this either.

Consequently, future studies should focus on non-genotoxic carcinogenesis and co-

carcinogenesis. They should also be designed to clarify the problems of the dependence of
amplitude modulation and in general, as a matter of priority, the independent replication of
the positive results reported.

In her introduction to the subject, Elisabeth Cardis recalled that, before mobile telephones,
the two main sources of exposure to radiofrequencies were occupational and domestic and
that the information on health hazards at that time were limited. Such information is
considered today to be totally inadequate to determine the presence or absence of an
association between exposure to radiofrequencies and effects on health. The author
concludes her introduction by citing two epidemiological studies dedicated to exposure of
mobile phones, one by Kjell Hansson Mild on symptoms associated with the use of a mobile
phone and the other by Joshua Muscat et al more specifically targeted to brain tumours.

The presentation of Kjell Hansson Mild described two epidemiological studies: one dealing
with subjective symptoms and brain tumours, the other addressing the possible induction of
brain tumours. The author underlines that since 1995 many people have reported suffering
from a variety of subjective symptoms such as migraine, discomfort, heating of the ear and
difficulties of concentration while using a mobile phone [667, 365, 611].

A large epidemiological study, including 12,000 users in Sweden and 5,000 in Norway, was
set up to evaluate the incidence of this type of symptom. A comparison between users of
GSM and NMT shows that the latter complain much more of a sensation of heat behind the
ear than do GSM users. The same conclusions apply to users in Sweden with respect to
headaches and fatigue [610].

The author described the work carried out by Hardell et al [346] on possible induction of
cerebral brain tumours, in which he himself participated. This was a case-control study of
233 people suffering from brain tumours and living in Uppsala and Stockholm. An increase
in the risk of brain tumour as a function of mobile phone use was observed between tumours
in the temporal and occipital lobes of the brain of subjects who use the phone on the same
side of the head. This increase in the risk was observed only for the NMT system, the
observation time being too short to include effects from GSM. These results, not statistically
significant, are based on a small number of observations (13 cases, 10 with malignant
tumours, 3 benign of which 9 were exposed to NMT, and 3 to both NMT and GSM and only

one to GSM) The conclusion of the author is that an increase in the risk of brain tumour in
the anatomical regions where the mobile phone is used must be examined in the future.

The contribution of Joshua Muscat et al, presented the results of a case-control

epidemiological study carried out between 1994 and 1998 in 5 medical establishments in the
New York and Massachusetts areas. 469 man and women who had recently being
diagnosed with brain tumour were studied. The results indicated that the risk of developing
brain cancer does not appear to be associated with the average monthly use of mobile
phones (evaluating from billing information) nor to the number of years of mobile phone use.
The risk was slightly elevated for brain tumours in the region close to the ear. In contrast,
specialised histological analysis revealed an increase in the risk of a rare form of
neuroepitheliomas, this risk being the same for high and low mobile phone users. In this
study the spatial contribution of brain cancer was not related to the use of the mobile phone.
The author thinks that some possible effect of the mobile phone can occur in the phases of
promotion or progression of the development of cancer and that such effect can be
manifested in the relatively short time. Future studies are absolutely necessary to evaluate
the effects over longer induction period.

5/ Technical, economic and regulatory affairs

In the first contribution to this subject, Luis Miro described the evolution of possible health
risks from the use of mobile phones. He reasoned that, in parallel with the technological
evolution of mobile phones and universal market penetration, it is wise to think of possible
health risks and how these will evolve.

Technological evolution implies that:

· The brain region will receive less exposure in coming years
· Mobile communication frequencies will be closer to the region of 2 GHz
· The number of users of mobile phones will increase significantly, including in particular
children, the elderly and the sick.

This is taken into consideration in the various exposure regulations by an additional safety
factor of 5 compared to the exposure limit recommended for workers and in controlled areas.
The risks to health are under control.

However, if we consider this situation, we can see that it refers only to mobile phones. Their
integration in the totality of telecommunication developments going on around is not
considered. In fact, in the medium term, the mobile phone will be integrated in a
constellation of networks of wireless devices including “Wireless Local Area Networks”
(WLANs) such as office networks, networks for the control of the domestic environment and
even body-mounted WLANs installed on the person itself. This development is likely to lead
to a significant increase in the number of emitting devices and in particular in the number of
low power devices, some of which may be placed in direct contact with the body. However,
the intercommunication between networks will probably take place at higher frequencies, of
the order of tens of GHz: that is in the millimetric wave region.

This new situation could lead to:

· an increase in the exposure of the whole body to a large number of electromagnetic field
emitting devices

· the placing on certain parts of the body of low power devices that may be capable of
direct coupling to reactive components in the skin.

Such a scenario is not unreasonable if one considers that:

· studies of field distribution inside the body show that penetration decreases with
increasing frequency, correspondingly increasing absorption in the superficial areas of
the skin
· several biochemical, cellular, physiological and therapeutic studies have shown an effect
from such waves. The conclusions from these studies are not self-consistent but,
collectively, these studies seems to imply that such effects exist even if for now we do
not know the relevant parameters.
· a certain number of studies have shown the possibility of resonance effects at certain
frequencies, although these results are considered controversial and are the subject of

In conclusion, the author considers that it is important to initiate biological and medical
studies to evaluate closely the real, and possibly physiopathological, effects on man from
millimetric waves and from the increase of the “electromagnetic environment” in which man
is destined to live. The author thinks that this should take place simultaneously with the
development of the technology.

In his contribution, Jörgen Bach Andersen treats the impact of technological evolution on
the biological effects of mobile phones. The author considers the relationship between the
continuous evolution of microelectronics, the popularity of mobile phones and the “global
internet” to mean that the growth in the number of mobile phones will continue. One of the
major changes to be expected is the increase in the amount of data transmission (images,
electronic commerce, news, commercials, publicity, etc) compared to the transmission of
speech. The main requirement for this change of utilisation is the ability to transmit data at
high rates: specifically the transmission rate should increase from 9.6 kb/s to several
hundreds of kb/s, reaching 2 Mb/s for UMTS or IMT-2000 terminals. This evolution has
already begun with GPRS or EDGE, an incremental modification of the GSM family. This
means that the TDMA (Time Division Multiple Access) system will persist. New spectrum
has just been allocated in the 2 GHz frequency band for UMTS, but this UMTS system will
be a new departure in that it will be a code division system (W-CDMA): the signal will be
transmitted at constant power and will not have the repetitive nature of the TDMA signal.

The mode of use of UMTS terminals will have implications for biology and health. The
terminals, which will be video-enabled, will be used at a reasonable distance from the head
thus reducing the direct coupling with the brain that occurs with current handsets. However,
the voice communication mode will remain an important component of the new UMTS

At high data rates, transmission will be almost permanently at the maximum available power
of 2 Watts, which is not the case for current mobile phones. The specification of maximum
transmission power will have to take into consideration possible biological interactions. An
interesting new area in wireless communication is the increasing use of small low power
transmission devices (of the order of milliwatts).

Many devices will be built into every-day objects to achieve connectivity without cables to,
for example, a printer or a digital camera. These will be low-power devices operating in the
ISM band at a frequency of the order of 2.45 GHz. The use of hand-held mobile phones
close to the ear, at low power, will be reduced. The main mode of use will be at high power
but at a reasonable distance from the body.

In conclusion, the author predicts an enormous increase in wireless communication.

However, because of the combination of relatively low power and large distance from
sensitive tissues, public health should not, in his opinion, suffer.

Opinion of the group of experts on the report to the Academy of Sciences: this
document provides a snapshot view of the state-of-scientific-knowledge available in April
200017. The different areas of research relating to possible relationships between mobile
phones and health are clearly presented and comprise in particular:

· the possibility of inducing brain tumours,

· modifications in the permeability of the blood-brain barrier,
· alteration in DNA.
· electroencephalogram modifications, in particular in the alpha wave and the spectral
content and, during sleep, changes in REM sleep,
· cardiovascular effects, mostly on arterial pressure,

In general these studies provide evidence for numerous biological modifications without
allowing any health consequences to be attributed to them. With respect to the hypothesis
of a carcinogenic risk, the review of all studies leas to the conclusion that it is very unlikely
that radio frequencies have a direct genotoxic effect. However, several points need to be
further clarified, in particular:

· effects of signal modulation need to be studied in greater detail (Phillips et al.)

· a co-carcinogenic effect suggested by the results of several studies (Maes et al., Scarfi
et al.)

· changes observed in the levels of oncostatic agents (where a deficit can lead to an
increased risk of cancer), such as melatonin (de Seze et al.)

· changes observed in the level of proliferation agents such as thymide (increase in the
synthesis of DNA (Stagg et al.)) or ornithine decarboxylase (ODC) (Stagg et al, Litovitz et
al., Penafiel et al.), which are key enzymes in the synthesis of polyamines. The group of
experts agrees with the conclusions of Jukka Juutilainen who, in reporting these studies,
qualifies them as potentially important, and consider that this field of study should be
actively pursued

The group of experts observed that the issue of base stations, despite it being the subject of
media attention, has been considered only superficially in this work.

The general conclusion is that, while the results of the studies presented do not provide
evidence of an actual health risk, it remains necessary to evaluate possible implications of
the biological effects observed. Several international research projects, cited in the report to
the academy, are actually pursuing this aim.

4.2.2 Analysis of the Report by Essor-Europe

Critique of the report entitled “Physiological and environmental effects of
electromagnetic fields” drafted by Essor Europe (Jean-Pierre Chevillot, Jean-Pierre
Husson, Philippe de Montgolfier) in May 2000.

This report was initially mandated by STOA18 of the European Parliament. It deals with the
whole of the non-ionising electromagnetic spectrum and in particular the low frequencies
(ELF) that are relevant to power distribution and the radiofrequencies that are relevant to
mobile telephony.
the document was published in January 2001 but it reports on work presented to the Academy on
th th
the 19 and 20 of April 2000
Scientific and Technical Options Assessment

The information for the report was obtained from different sources: analysis of information
available on the internet, sending a questionnaire to 45 experts, analysis of international
reports and interviews with 25 people; the final revision of the report was undertaken by 3
international experts (U. Bergqvist, M. Repacholi and B. Veyret).

The report deals equally with research and risk management. The following points are
considered specifically:

· Adequacy of regulatory matters,

· Particular consideration of susceptible sub-populations,
· Knowledge of mechanisms of interaction of fields with the human body,
· Current knowledge of fields in the environment and public information.

Summary of conclusions of the report:

1-Data obtained from research

The general conclusion of the report is that the research outcomes on which standards are
based confirm the thermal nature of known biological effects. They do not provide evidence
for the existence of deleterious effects associated with low-level exposure as encountered in
normal occupational and residential exposures. However, some biological effects have
been reported at low exposure levels: these must be confirmed in order that they can be
considered in the revision of standards.

The effects that must be clarified as a matter of priority are:

· promotion of tumours,
· permeability of the blood brain barrier, EEG
· changes in arterial blood pressure,
· endocrine and immune systems,
· hormone levels and in particular melatonin,
· activity of the enzyme ornithine decarboxylase (ODC),
· transport of calcium ions,
· alteration of DNA

Special attention must be given to “co-factors” that can increase the sensitivity of a biological
system to electromagnetic fields. Three categories of co-factors are considered:

· Aggressive factors present in the environment, such as chemical products associated

with atmospheric pollution
· Ionising radiation
· Psychosomatic or physiological susceptibility of certain individuals. The stress factor
must take into consideration the declared hypersensitivity of certain individuals to

The question of the possible vulnerability of children is also considered. If no specific effect
has been observed in relation to mobile phones and base stations it is, according to the
report, in part due to the very small number of specific studies that have been carried out.
The question is complex because of the lack of objective data. However, it appears that the
additional factor of five reduction in the limits compared to occupational exposure is
adequate for the protection of children. To tackle this problem more fully, more objective
data are needed.

2-Recommendations of the report:

· The report recommends a general attitude of “educated and responsible awareness”

· The scientific data available do not justify a revision of the recent European
Recommendation. It would even be desirable that this recommendation becomes a

· Several research areas must be given priority: possible effects of low-level, long-term
exposures; possible effects on populations that are considered a priori vulnerable
(children, the elderly, sick people); clinical studies of symptoms; studies on mechanisms
of interaction.

· Public education in an atmosphere of transparency,

· Signage of sources and mapping of the emissions around powerful emitters,

· Establishing specialised laboratories competent in the measurement of fields as needed.

· The establishment of channels of communication between the users and providers. A

Delphi-type study could contribute to this dialogue.

Conclusions of the group of experts on the Essor-Europe report: This report covers most of
the electromagnetic spectrum and provides, in a tabular form, a summary of the main
scientific data. It is not a comprehensive compilation, and its approach is different from that
of the other reports considered. The objective of this report was to extract the main thrust of
the interpretation of scientific data and to provide recommendations and options. Its main
feature of interest is that it was written after consultation with a very large number of players
in this field, from academia and from industry.

The conclusions relating to the need to pay further attention to the exposure of children are
reasonable, despite the lack of an explicit scientific basis.

This report has not been disseminated very widely because of strong political reaction to its
conclusions. This is regrettable, because it complements, in a useful manner, the other
recent international reports.

4.2.3 Analysis of the COMAR Report

(IEEE Committee on Man and Radiation)

The Committee on Man and Radiation of the Institute of Electrical and Electronic Engineers
has produced two reports on the impact on human health of mobile phones and their base

I Human exposure to radio frequencies and microwave radiations from mobile

phones and other telecommunication equipment.

Several national and international organisations have established recommendations on

human exposure to radiofrequencies. These recommendations are generally in agreement.

Field measurements have shown that the exposure of people from mobile phones and other
wireless systems is, in general, within the recommended limits. Some of these systems may
perturb the function of active medical implants, such as cardiac pacemakers, when they are
placed at very close contact. Wearers of such implants must seek advice from their doctor
with respect to the risks of using telecommunication devices.

The licensed frequencies used in the United States are described. The power emitted and
the distance at which the equipment is used should be defined: hand-held equipment is low-
power and used near the body, while other mobile equipment operating at high powers is
used at greater distances.

The structure of the recommendations on human exposure defines two regimes:

occupational or public exposure, or sometimes controlled or uncontrolled environments.
There is, overall, a factor of five between the two exposure limits. The factors taken into
consideration for different exposure conditions are whole-body vs local exposure, exposure
duration etc. The IEEE standard excludes the need to control exposure from low-power
devices; this covers most cellular radio communications equipment.

The scientific basis for the determination of limit values is presented in chapter III “exposure
limits”. A non-verified hypothesis is that the perturbation of learning in primates, occurring at
an SAR of 4 W/Kg, can be extrapolated to man. Despite numerous scientific speculations,
no mechanism has yet been established through which electromagnetic fields at levels
below the recommended values could cause biological insults with pathological

Very little information is available on long-term human exposure. Two epidemiological

studies producing negative results are cited: those of Rothman and Hardell, with the
acknowledgement that the follow-up period of these studies is insufficient.

There is no evidence that exposure to radiofrequencies at levels below the recommended

limits is harmful to health. Cellular mobile phones must conform to the recommendations.
Because of the large safety factor included in the standards, the recommended values are
well below the threshold for known risks to health. Most phones comply with the standards,
however a manufacturer has had to recall mobile phones that produced SAR levels slightly
above the recommended limits. The current digital mobile phones operate at lower powers
compared to the older analogue ones and are much more likely to comply with the standard.

Conclusions: The SARs produced by cellular mobile phones and other radio
communication equipment do not exceed the recommended limits set by standardisation
bodies. The scientific evidence to date does not show any risk to health from these devices.
There is a possibility of interference when they are used close to active implants.

Comments of the group of experts: the follow-up period is in fact too short in the
Rothman study as well as that of Hardell, it is expected that the latter will be extended to
1000 volunteers.

II Health effects associated with base stations used for radio communications

In most cases, public exposure to RF fields from base stations is clearly lower than the
recommended limits. In some cases, when antennas placed on rooftops are mounted such
that they are easily accessible, the exposure levels could be exceeded. Access to such
sites should be prevented or at least there should be a warning sign. Personnel who are
likely to be exposed in the close vicinity of the antenna must follow operating procedures,
including where appropriate the use of dosimeters or alarmed devices. Alternatively, if
maintenance personnel absolutely need to get close to the operating antenna, they should
wear protective clothing to reduce their exposure. Moreover, exposure to the main beam
may result in interference with active implants such as cardiac stimulators. People fitted with
such implants must seek the advice of their doctor prior to working close to antennas.

There are two criteria for the siting of base stations: providing coverage within a cell and
producing the necessary traffic capacity. When a system is being extended, base stations
are sited in smaller and smaller cells to increase the capacity of the network, however they
will operate at lower power in order to limit possible interference between base stations. The
height of the antenna is critical: it is typically between 10 and 100 m.

In the United States, the Federal Communications Commission (FCC) regulates base station
power levels. The effective radiated power is limited to 500 watts per channel; a base
station may use 21 channels or more per sector (sometimes 96). In towns, many base
stations would use only 16 watts per channel. Inside certain buildings, micro-cells operate at
less than one watt per channel.

The beam radiating from a base station antenna is narrow in the vertical plane and wide in
the horizontal (typically 6º in height and 120º horizontal aperture). Outside the main beam,
as in the case immediately beneath the antenna, the intensity of RF fields is weak. The
maximum field to which an individual would normally be exposed is at the point where the
main beam reaches ground level, usually between 50 and 200 m from the base of the tower.
This maximum level is already far below the recommended limit. Inside buildings, the field
intensity is less than outside, whether the building is directly below the antenna or in its

Conclusions: the levels to which the public are exposed in the neighbourhood of base
stations are significantly below the recommended values. Base stations do not constitute a
health risk for the population in general, including the elderly, pregnant women and children.

Where it is possible for people to be exposed to levels in excess of the recommended

values, access to the vicinity of the antenna should be controlled or there should be a
warning sign and operating procedures to prevent over-exposure, or protective measures
should be put into place. There is a specific risk of interference with active implants.

The group of experts is in agreement with the conclusions of COMAR on base stations

4.2.4 Analysis of the Report to The Swedish Government: “Review of

Electromagnetic Fields and Health”
In 1997, the Swedish government commissioned an evaluation of scientific work carried out
in the country and at an international level on health risks associated with electromagnetic
fields. The recent report from the Institute Of Occupational Environmental Health is the
result of this ininiative19. Made up of two parts and several annexes, it covers parts of the
non-ionising electromagnetic spectrum; a summary has been written in English (4 pages),
and that is the only part that can be analysed here. What is considered below are the
conclusions relating to RF.

The methodology is similar to that adopted by the group of French experts. The great
diversity of research protocols is pointed out, and it is recommended that in order to reach a
conclusion it is necessary to take into consideration all studies based on the same
hypothesis and not only the result of a single experiment.

Considering firstly occupational exposures to RF (other than mobile phones), the authors of
the report observed that many effects are associated with temperature rises of 1 to 2º C
(effects on behaviour, on the neuromuscular system, performance deficits, etc….).

” Eloverkanslighet och halsorisker av elektriskaoch magnetiska falt. Forskningsoversikt och
uivardering”. Slutrapport fran arbetsgruppen vid. Radet for arbetslivforskning.
Ulf Bergqvist, Lena Hillert and Elisabeth Birke, November 2000

By contrast, with respect to exposures below the reference levels, that is exposure without
thermal consequences, there is no evidence to date of an increased risk of cancer either
from experimental animal studies or from epidemiological studies. Even under worst case
communication conditions, the exposures of mobile phone users are always below the
reference levels. Some studies have reported biological effects, but the scientific evidence
does not support health consequences from the use of mobile phones.

Part of the review is dedicated to the problem of RF hypersensitivity. The report concludes
that it has not been possible to associate any particular RF parameter with the reported
problems. The idea is not dismissed that a combination of internal factors (related to the
autonomous nervous system) or external factors (physical, chemical and/or psychosocial)
can, for some people, produce these unexplained effects. The fact that this can lead to
public health problems, and deserves special attention, is underlined.

In conclusion, the report points to some studies whose aim is to verify that these
conclusions, based on work carried out at RF frequencies in current use, can be
extrapolated to other RF frequencies associated with new radio communication

4.2.5 Analysis of the Colloquium of the National Assembly “Mobile Phones: A

Danger To Health?” (19 June 2000)
This colloquium was organised by a group of National Assembly deputies (A. Aschiéri, J.-F.
Mattéi, J.-P. Brard, M. Rivasi, F. Loos, P. Lellouche); Messrs. Loos and Mattéi did not take
part in the meeting. About 150 people attended. An outline proposal for a law that the six
deputies intended to submit to the Assembly was developed by P. Lellouche. This
colloquium was divided into three main parts: legislation, effects on health and risk
management. Only the second part was recorded in the proceeding: this analysis covers
only technical or scientific matters relating to health, falling within the competence of the
group of experts.

Contribution of R. Santini
R Santini, a university lecturer from INSA in Lyon, holds the view that current standards are
inadequate and that long-term health effects should be considered. He has recently carried
out a survey of low statistical significance on students in his institute. The purpose was to
determine whether users of mobile phones were more likely than non-user to have
symptoms such as headaches, fatigue, etc. The results of the study, which are essentially
negative, have been submitted for publication.

Contribution of G. J. Hyland
The presentation of G J Hyland focused on a theoretical mechanism to explain possible
health effects of RF: biocompatibility by electromagnetic compensation. Since the
colloquium, G J Hyland has published this theoretical perspective in a prestigious
international medical journal (G J Hyland, Physics and biology of mobile telephony. The
Lancet 2000; 356:1833-35); this article is discussed elsewhere in this report (cf. reviews of
recent articles).

Contribution Of G. Carlo
G Carlo was the president of Wireless Technology Research (WTR) which was a private
organisation charged with developing for the US a large, industry-financed, programme of
research. In the colloquium, he summarised the main results obtained within the framework
of the WTR programme. These results, (some were still provisional at the time) extended
over several fields of studies including genotoxicity, particularly in vivo, and epidemiological
studies. These results will not be reviewed at this stage, as they are mostly published and

are discussed elsewhere in this report. However, it is important to point out that G Carlo's
main conclusion, taking into consideration the scientific results he reported, is that mobile
phones present an actual health hazard and that a public health approach to the issue must
be taken. This would require research and observations based on the systematic gathering
of data (such as complaints from users, cohort studies of users, studies on children and
pregnant women), studies on the adequacy of SAR, on effects on the brain, and on the
behaviour of mobile phone users. Finally, G Carlo proposes recommendations regarding
the use of mobile phones that fall within the concept of “prudent avoidance” even though this
term is not used explicitly.

It is interesting to describe the organisation of this vast research programme financed by

industry, and the processes implemented in order to guarantee quality and independence.
They include a systematic peer-review of research projects and their results, the process
being co-ordinated by a prestigious University Institution (Harvard School of Public Health)
and governed by strict rules of quality and transparency. Many public organisations took
part in various stages of the process of setting up and implementing research protocols
(FDA, FCC, EPA, NCI, NIOSH, etc.).

Contribution of B. Veyret
The main purpose of this contribution was to present the French "COMOBIO" research
programme; this programme is detailed elsewhere, see chapter V-1..

Contribution of M. Bastide
M. Bastide (Laboratory of Immunology and Parasitology of the Faculty of Pharmacy at
Montpellier) presented the results of research on effects on embryonic mortality and the
hormonal system in animals from exposure to electromagnetic fields from mobile phones.
These results show very significant effects on the embryonic mortality of chicks and on
stress hormones in young mice; these results are novel and without parallel in the
international scientific literature. For this reason, they arouse a great deal of interest, and
require independent replication before they can be taken into consideration, moreover the
experimental protocols used in the study are described very briefly and do not correspond to
realistic exposure conditions.

Other presentations were made by Messrs. J Wiart, of France Telecom, J C Bouillet, of

Bouygues Telecom, G Dixsaut, of the Directorate General of Health and R de Sèze
(member of the group of experts).

4.2.6 Analysis of the Sage Report

Report of 15 October 1999 presented to the Scottish Parliament

In the first part of the report, Cindy Sage, a private consultant, presents to the Transport and
Environment Committee of the Scottish Parliament her views on what information the
committee should consider in formulating advice “based on the present state of knowledge”.
In the second part of the report, she reviews the scientific articles that she considers
important and which should be taken into consideration.

The development of radiocommunications has resulted in a long-term cumulative exposure
of people without precedent in human history. There is evidence that biological effects occur
from exposure to radiofrequency radiation (RF). There is no conclusive scientific evidence of
the safety of such exposures, and some studies suggest that serious health effects can
occur due to cumulative or chronic exposure. The committee should advise “public health
precaution” and urge that the exposure of people all over the world be kept to a minimum
until further research can clarify risks. The committee should also advice that cumulative

exposure of the public to radiofrequency fields can eventually be dangerous on the basis of
existing scientific results.

The Precautionary Principle is frequently promoted by public health advocates given the
massive public health risk that is possible if exposure is carcinogenic or has other adverse
bioeffects. Even if the risk to individuals is slight, the sheer number of people around the
globe who may be at risk makes this choice of policy of utmost importance. The virtual
revolution in science that is taking place now is based on the growing recognition that non-
thermal and low intensity RF exposure can be detected in living tissues and results in well-
defined biological effects. The biological effects that are reported to result from RF exposure
include changes in cell membrane function, metabolism, cellular signal communication,
activation of proto-oncogenes, and cell death. Resulting effects reported in the scientific
literature include DNA breaks and chromosomal aberration, increased production of free
radical, cell stress and premature ageing, changes in brain function including memory loss,
learning impairment, headaches and fatigue, sleep disorders, neurodegenerative conditions,
reduction in melatonin secretion and cancer.

According to Cindy Sage, the committee for transport and environment should require the
wireless industry to provide complete, honest and factual information to consumers, to
independently monitor any health effects of mobile phone use, and to strongly urge public
participation in policy-making and regulatory processes for RF exposures and technology.
The United States has a policy of “post-sales surveillance” for cell phones: that means that
cell phones can be sold to the public, and only after years of use will there be studies to
characterise what health consequences have arisen as a result. In short, “we are the
experiment” for health effects. The committee should reject post-sales surveillance as
inadequate to protect existing users.

While the scientific community continues to study and understand the physical bases for
electromagnetic effects on living systems there is little to protect or inform the public about
consequences of blind belief in these new technologies. For all the potential good that such
technologies bring to us we must be vigilant about possible undesirable consequences.

Comments of the group of experts: numerous studies have failed to demonstrate any risk
to health. It appears that, for this author, whilst a serious effect has not been demonstrated,
the studies are not conclusive. It is already possible to confirm that if a risk exists, it must be
weak, because of the many studies demonstrating negative results. A non-thermal
exposure can be detected in living tissues, but it is rather presumptuous to say that the
biological effect that results from it is well defined. Effects on membrane function are not
explicit, which makes it difficult to guess what other results the author refers to. Cellular
death has not been reported except at very high thermal levels.

As will become clear from the comments on the studies cited in reference, the effects
referred to are either far from being confirmed, or due to an exposure with characteristics
very different from those of radio telephones: radar pulses of high peak power or emissions
of high average power.

Many studies have been carried out in the United States in the last few years; the results of
epidemiological studies published to date are rather reassuring.

Important scientific articles

The evidence for an association between RF and biological effects in living systems relates
to all levels of organisation: the atomic (calcium ions), molecular (DNA), in humans and
other mammalian species. In the past 50 years, experiments carried out at frequencies
spanning the whole electromagnetic spectrum have found repeatable biological effects on
many species, from mice to man. The cascade of biological, chemical and physical events

that occur in living systems in response to RF is increasingly understood by the scientific
community as a mature multidiscipline. Disease is not the only end point of this research,
there are potential medical implications that offer an unparalleled opportunity for healing and
well being.

Comments of the group of experts: effects can be observed in all biological systems,
some weak effects are reproducible but the majority are not, nor are strong effects in
particular. Effects in the extremely low frequency range (ELF) are more consistent than
those reported in the RF range.

Effects on DNA
The author has reviewed the work of Lai on DNA breaks evaluated by comet essay after
exposure to RF fields at 2450 MHz, and those of Jerry Phillips. Phillips suggests that the
rate of DNA repair could be altered by RF. He found an identical effect at extremely low
frequencies (ELF) of 100 mT at 60 Hz. He postulates that ELF magnetic field exposure can
affect both augment DNA damage and inhibit repair mechanisms, and lead to cell death

Conventional wisdom has traditionally held that microwaves are not genotoxic unless high
temperatures are produced.

Blank and Goodman (1997) postulate that the mechanism of EM signal transduction in the
cell membrane may be explained by direct interaction of electric and magnetic fields with
mobile charges within enzymes. Recent studies on DNA show that large electron flows are
possible within the stacked base pairs of the double helix of DNA molecules. Therefore
gene activation by magnetic fields could be due to a direct interaction with moving electrons
within DNA. Electric fields as well as magnetic fields stimulate gene transcription and both
fields could interact with DNA directly. Previous work on a heat shock protein by Goodman
and Blank, cited in the 1997 article, show that one cellular response to EM fields is activation
of the same stress response system seen in heating, but at very much lower energy than the
response to heat shock. (see § “gene transcription and induction”).

Comments of the group of experts: The hypotheses of Phillips on apoptosis have not
been confirmed experimentally. Moreover, they relate mostly to ELF as do the hypotheses
of Blank and Goodman on the interaction with free charges either in enzymes or with DNA

Chromosome aberration and micronuclei

Maes et al (1993) have reported an increased frequency of chromosomal aberrations and
micronuclei at non-thermal levels. One type of damage observed (dicentric chromosomes)
is considered to be an indicator of ionising radiation damage. These results are consistent
with those of microwave radiation damage at other frequencies and power densities reported
by other workers (Leonard et al, 1983; Garaj-Vrhovac et al, 1990, 1991; Ambrosio et al,

Maes et al (1995) reported that whole-blood exposed to the radiation from a GSM base
station showed increased chromosomal aberrations when placed within a distance of 5 cm
or less with 2 hours exposure. Combined effects of 954 MHz radiation and the chemical
mutagen mitomycin C (MMC) were studied by the same authors using human lymphocytes
(reference not cited: Maes et al, 1996). Samples exposed to an estimated SAR of 1.5 W/kg
and to MMC showed a significant increase of one form of chromosomal aberration;
translocation (sister chromatid exchange) by comparison to MMC only. Single strand DNA
breaks were also reported.

Comments of the group of experts: the study of Maes et al, 1993 was carried out at
75 W/kg, an SAR that cannot be qualified as non-thermal, even with a continuous thermal
control to maintain a temperature of exactly 36.1ºC. The agreement with other studies not
cited by Sage is not entirely evident in the summary of articles published by Information
Ventures in their EMF bibliographic database: Leonard indicates that most of the studies on
mutagenicity were negative except at thermal levels, but that exposure at non-thermal level
could activate other mutagenic agents such as UV or chemical substances. Similarly, the
study of Ambrioso showed a mutagenic effect, at a clearly thermal level of 100 W/kg with an
increase in temperature of 5º C.

In his study of 1995, Maes states clearly that the observed differences are not significant. In
the 1996 study that combines exposure to RF and that to a mutagen: mytomycine C, the
results indicate a significant effect of MMC compared to microwaves only, but it has not
been shown that it is significant the other way round as well, that is MMC plus microwaves
compared to MMC only.

Effects on ornithine decarboxylase (ODC)

Ornithine decarboxylase is an enzyme whose concentration and activity are significantly
increased in growing tissues, and in particular tumours. Litovitz's group have reported in
several articles that the activity of this enzyme was significantly changed by an
electromagnetic field at 835 MHz and a SAR level of 2.5 W/kg, amplitude modulated at very
low frequencies between 16 and 65 Hz.

Comments of the group of experts: the magnitude of the effect observed is a factor of 2;
the increase in ODC activity in tumour cells (considered a promoting effect) is of the order of
a factor of 400 to 500.

Gene transcription and induction

Goswami et al (1999) reported an increase by factor of 2 of the concentration of an RNA
messenger for the proto-oncogene fos in fibroblasts exposed to electromagnetic fields from
cellular radiotelephones. The emission was continuous at 835 MHz, modulated in
frequency. A rather smaller increase was observed at 847 MHz with CDMA modulation

Comments of the group of experts: the resulting SAR is not given in this report. One can
assume that Fos is used instead of c-fos.

Stress response
Daniells et al (1999) found that some worms responded to microwaves as they would
respond to heating or to the presence of a toxic chemical substance. The model used
shows that lower intensities induced the highest responses (the opposite of simple heating).
Microwave radiation caused measurable stress and protein damage within cells (induction of
heat shock protein) comparable to damage from metal ions, which are known to be toxic.

Comments of the group of experts: it is important to specify the SAR obtained and the
amplitude of the corresponding response. Similarly, in the comparison with toxic metallic
ions, it is important to specify the equivalent concentration of the magnitude of the toxic
effect produced.

Cellular effects of microwave radiation

Calcium ion balance in living tissue is very important for the proper function of cellular
communication, cell growth and other fundamental life processes. W. Ross Adey and his
team have described a succession of events leading to the alteration of cellular function by

Adey (1993) has given a summary of explicit effects of microwaves at the cellular level that
support the concept of an athermal effect. He has discussed the importance of free radicals
in a number of pathological phenomena such as neurological diseases, cardiovascular
diseases or cancer. Microwave exposure at athermal levels may act as a cancer promoter.
He recommends that the biological effects of low-level microwaves require further
investigation particularly in the light of non-linear non-equilibrium co-operative processes.

Dutta et al (1989) have published results showing modification in the concentration of

calcium ions in cells of different animal species, with amplitude-modulated radio frequencies
at low level SARs of 0.05 or 0.005 W/kg. According to the authors, these results confirmed
that modulated RF could induce cellular responses in different animal species.

Comments of the group of experts: calcium ion equilibrium is very important, but its exact
concentration is not very critical compared to the intra- and extra-cellular concentrations of
sodium and potassium, for example. Only very large changes in concentration, of the order
of 100 to 1000, are associated with cellular functions such as secretion, contraction, the
generation of action potentials, division or differentiation. When the variation produced by
microwaves are of the order of 50 to 100% (a factor of 1.5 to 2), it is not possible to speak of
“changes to cellular functions”, as a response induced by the exposure. There are only very
weak fluctuations in magnitude compared to spontaneous variations that can be observed in
these systems.

It is actually very doubtful that Adey would have written that “the exposure to microwaves at
athermal levels can have the same effect as cancer promoters”. Even if that was the case,
which is not been verified, this statement was not supported by any evidence.

Immune system cellular effects

Lyle et al (1983) reported that RF modulated at 60 Hz inhibits the activity of cytotoxic
lymphocytes (capable of killing cells that are attacking the body).

Veyret et al (1991) found a significant change to the immune system from exposure to
amplitude modulated microwaves with an SAR of 0.015 W/kg.

Elekes (1996) found a slight increase in the production of antibodies in male mice (but not in
female!) with exposure to amplitude-modulated RF. The authors postulate that the very
weak effects observed may be due to the very short exposure time.

Comments of the group of experts: the magnitude of the biological effects produced is
not given in the Sage report. In the study of Veyret et al (1991), the variations observed
were on the concentration of antibodies produced by immunisation and are at most 50%
either way. Compared to variations by a factor of 1000 to 10 000 that take place during
slight inflammation of the throat this effect is not biologically significant, and has no health
consequence even if it is statistically significant and the mechanisms are of fundamental
interest. The same argument might be applicable to the work of Lyle and Elekes, but this has
not been verified. The comment by Sage relating the magnitude of the biological effect and
the duration of exposure seems to suggest a cumulative effect that has never been
demonstrated. By contrast, in certain studies such as those by Lai for example, the effects
appear to be rather “cyclic”, positive at one time than negative a few minutes later, without a
continuous increase as a function of exposure time ever having been observed.

The blood-brain barrier

The blood-brain barrier protects the brain by preventing toxic substances that are present in
the blood from reaching the more sensitive brain tissues.

Salford has shown a leakage through the blood-brain barrier produced by RF at 915 MHz,
either pulsed or continuous. Further research is needed to demonstrate whether or not this
constitutes a risk to health. At least 10 references cited in his article report effects on the
blood-brain barrier by RF.

Comments of the group of experts: the work of Salford is contradicted by the results of a
large number of other studies (Hossmann, …), but also supported by some others (Albert,
1977), although that was at higher power. Most of the studies carried out at that time did not
include an evaluation of the SAR. Some of the references relates to high-power exposure.
The molecules identified in the leakage are molecules that are normally present in the brain
but usually at lower concentrations.

It has been demonstrated that the electromagnetic fields produce biological effects, at all
levels of the organism, that can have deleterious health consequences. The basic functions
that control proper cell growth, proliferation, immune surveillance and toxin protection are
shown to be adversely affected in many cases at environmental levels of exposure. In
human and animal studies it has been shown that there is a relationship between exposure
and certain cancer conditions. The major concern with mobile telephony is its rapid growth
around the world, and the increasing evidence of brain tumour.

Guy et al (1984) made a long term study of rats exposed throughout their lifetime to
450 MHz RF at a low power density (1 mW/cm2; SAR of 0.15 to 0.4 W/kg). Numerous
previous studies did not address the question of risk to human health from low-level
exposure. He reported that the number of primary malignant tumours was (significantly)
elevated: four times higher in a group of exposed animals compared to the control group.

Repacholi et al (1997) found 2.4 times more lymphomas in mice exposed for 8 months to
fields similar to mobile telephony compared to the control group.

Hardell (1999) reported an increased risk of malignant tumours of the brain in analogue
mobile phone users, on the side of the head where the telephone is usually used.

Adey (1996) found a protective effect on brain tumour in rats, however this result was not
statistically significant.

Comments of the group of experts: the mice in the Repacholi study were transgenic, that
is genetically modified to produce lymphoma. If the results of this study were confirmed, it
would be important to study more carefully the possibilities of extrapolating these results to
man: two replication studies are ongoing. In the studies by Hardell and Adey a non-
significant modification does not mean a risk, or a protective effect, but rather the practical
impossibility of reaching a conclusion. In contrast to the interpretation by Sage, Hardell
himself describes his study as not showing a risk. The work of Adey was published in 2000.

Cerebral symptoms attributed to mobile phones

Mild et al (1998) reported a significant association between using a mobile phone and a
sensation of heat behind the ear, headaches and fatigue. GSM phones were less likely to
cause these effects than analogue phones.

Hocking (1998) has also studied the occurrence of symptoms in man, reported to be more
common with GSM phones than analogue phones.

Comments of the group of experts: given that there is an appreciable heating from
telephones, probably more significant with analogue phones operating at 600 mW compared
to GSM phones operating at 125 or 250 mW, the sensation of heat or even headaches may

be expected. Fatigue should be studied further. The results of Hocking are rather
surprising, showing a predominance of symptoms with GSM phones.

Nervous system effects

Effects on the nervous systems have been studied at various levels. At the ionic and
molecular levels, numerous effects have been published and replicated at non-thermal
power levels. These effects include modifications in calcium, neurotransmitters, behaviour
and sleep.

Lai (1994) has written a review of the normal function of the nervous system and of the way
in which it could be influenced by RF. Changes in behaviour could be the most sensitive of
the effects of RF exposure.

The movement of calcium ions in brain tissue is altered by RF. Calcium ions control
numerous functions in the brain and the body, including the secretion of neurotransmitters
and their action at the level of receptors, and all such modifications could have implications
for health.

Comments of the group of experts: only changes above the “physiological noise level”,
that is the spontaneous variation of the parameter under consideration, are capable of
having any repercussion on health.

Psychoactive drugs
Changes in neurotransmitter function may occur under the influence of psychoactive drugs.
Lai has given some examples and concludes that endogenous opiates are activated by RF
(Lai, 1992, 1994b).

Comments of the group of experts: The hypothesis of Lai is that the action of RF on
endogenous opiates is an initial step that triggers an avalanche of other biological reactions

Serotonin activity is reported to be affected by RF hyperthermia (Lai et al 1984 and
Panksepp, 1973 cited by Lai, 1984). Man can suffer from a number of symptoms, known
collectively as serotonin irritation syndrome, that includes anxiety, headaches, blushing and
hyperperistaltism. There have not been any other studies on RF effects on serotonin.

Comments of the group of experts: This work relates to high power exposures. There is
a strange resemblance between the symptoms described and those often reported to occur
at low power levels and collectively known as “microwave syndrome”.

Ocular damage
Drugs can enhance the adverse effect of RF on the eye (Kues et al, 1992)

Comments of the group of experts: the pulses used in this study are similar to those used
for radar, with a pulse of 10 µs width every 10 ms, which means that the peak power used in
the study is 1000 times higher than the average reported. Therefore, the threshold of 0.05
W/kg for the observation of an effect is in reality obtained with a peak SAR of 50 W/kg. It is
surprising to find this study reported in an identical manner in the Stewart report which leads
us to think that some contributions from outside the group were taken by the committee of
experts without verification of the information obtained.

Behavioural changes
The performance disruption paradigm which is the basis for the acceptable level of RF
(below the limit for thermal effects) does include the effect of microwaves on cognitive

processes. Such effects will not necessarily completely alter behaviour but they could
modify it. (d’Andrea, 1999; Silverman, 1973; Raslear et al, 1993).

Comments of the group of experts: as for cellular functions, changes in cognitive

processes would only affect health if they occurred as a matter of course during exposure or
if their effects persisted after exposure. In this case, it would be necessary to review the
current standard.

Learning and memory

Lai et al (1994) describe a decrease in learning in a star-shaped maze. This effect was
inhibited by a drug that stimulate cholinergic activity.

Comments of the group of experts: Again, the RF used by Lai was emitted in the form of
very short-duration pulses (2 µs every 2 ms), which means that the peak power is 1000
times higher than the average power of 0.6 W/kg: 600 W/kg peak. One could certainly
conceive of effects under such conditions.

Cognitive function
Preece (1999) showed an acceleration in human choice reaction time. This acceleration in
reaction was a function of the power level used.

Comments of the group of experts: The reference cited by Sage does not correspond to
the study described. In fact, there was a significant effect only with analogue telephones
that are effectively more powerful than the GSM type. This is more indicative of a threshold
than of a dose-effect relationship as suggested by Sage.

A decrease in the period of REM sleep (dream period) was reported by Mann et al (1996).

Comments of the group of experts: these effects were not confirmed by the same
research group in two further studies: Wagner et al (1998) and Röschke et al (2000), or
reproduced in an identical manner by other groups (Borbely, 2000).

4.2.7 Analysis of the paper by J M Elwood

A critical review of epidemiological studies on radiofrequency exposure.
J M Elwood, Environ. Health Perspect. 1999, 107(supp1), 155-168

This article was submitted to the journal Environmental Health Perspectives in 1998; it
reviews critically those epidemiological studies published between 1988 and 1998 on the
risk of cancer associated with exposure to RF.

The range of RF exposures considered covers radar, radio and television signals in areas
close to transmitters or occupational exposure in the electronics and telecommunications
sectors. The first publications related to mobile telephones are concomitant. What is of
interest about this article, therefore, is that it encompasses other occupational or
environmental exposure to RF than that related to mobile phones.

The author is from a university cancer research centre in New Zealand and carried out this
literature review at the request of Telecom New Zealand. The EHP journal is one of the
most prestigious in the field of environment and health sciences. It is a publication of the
National Institute of Environmental Health Sciences (NIEHS) and reports on experimental or
epidemiological scientific works.

The authors divided the work analysed into four categories: investigations into clusters - 4
original publications; studies of the general population exposed to radio and television
transmissions and other similar fields (5); cohort studies concerning occupational exposure
(5); and case-control studies (6). The articles considered were identified by researching the
Medline database and also from published reviews. After a description of each study,
including the way in which exposure had been assessed, the main results are tabulated by
category of work to facilitate comparison. Finally, Elwood takes all this information and
applies an interpretation grid inspired by B. Hill’s causality criteria.

1- Investigations into Clusters.

From time to time, cases of rare diseases appear to be grouped in time and/or in space.
These phenomena, often due to chance, must be explored carefully in order to identify any
commonalities. It is now accepted that these clusters can allow the development of
hypotheses about risks, although it is not possible to draw conclusions without implementing
specific explanatory studies. Three reports have been published on cancer clusters near RF

A case-control study was initiated when 12 children living near a radio transmitter in Hawaii
were found to be suffering from acute leukaemia. Among children living less than 4.2 km
away, a (non-significant) excess of cancers was suggested, but the small number of cases
led the authors to conclude that the cluster had doubtless arisen by chance. 340 American
policemen equipped with radar speed guns was investigated: the 6 cases of testicular
cancer that occurred could not be linked to exposure. Another cluster occurred around a
radio and television transmitter near Birmingham in Great Britain. The investigation covered
all cancer cases occurring over 12 years within a radius of 10 kilometres around the
suspected source. An excess of adult leukaemia was suggested in the inner circle, within a
radius of 2 kilometres. This apparent trend was due to the lower-than-expected number of
cases in the second circle, compared with the population as a whole. The authors concluded
that it was not possible to attribute the cancers to the transmitter, but they did undertake a
wider-ranging study covering 21 RF transmitters across the country.

2- Living close to radio-television transmitters

Five studies have been published on the incidence of cancer in populations living close to
radio or television transmitters. The Birmingham cluster led to a comparative study of the
incidence of cancers over 12 years, in a radius of 10 kilometres around 21 British
transmitters (in the 430-890 MHz bands), including the transmitter which triggered the work
in the first place. This study involved 3.39 million people. The initial observation of an excess
of adult leukaemia was not confirmed: the number of cases in the inner radius of 2 km
around the 20 sites other than Birmingham was, on average, lower than expected, whereas
the risk in the 10 kilometre-circle as a whole was slightly higher (+3%) than expected.
However, the results varied depending on the type of cancer and from one site to another,
and one large TV transmitter near London showed a trend in adult leukaemia which
decreased with distance. Overall, the authors considered that their results gave, at best, only
a very weak indication of support for the initial cluster.

Another incidence study was conducted near three television transmitters (60 to 500 MHz) to
the North of Sydney, Australia. The maximum power density estimated at a distance of 1 km
was 80 mW/m2, with 2 mW/m2 at 4 km. The comparison covered cases of child and adult
cancer in the period 1972-1990, in relation to the distance from the transmitters (less than 4
km and from 4 to 15 km). An excess risk of leukaemia, but not brain cancer, was observed in
adults (RR = 1.18 [0.98- 1.42]) and children (RR = 1.58 [1.1-2.3]). These results therefore
contrast with the British observations, despite the distinctly weaker exposure levels. They
show no gradient effect and the possibility of differences linked to socio-demographic
structures of the populations cannot be excluded. This work was continued by another
author who extended the area covered by the study to other nearby areas. Field

measurements were taken, showing power densities varying from less than 2.5 W/m2 to
1,000 W/m2 at the foot of the transmitter. Although one of the three most-exposed zones
showed, as in the initial study, an excess of child leukaemia compared with more distant
zones, two others did not, thus suggesting the possible role of factors other than EM fields.

The incidence of leukaemia and lymphatic and brain cancer among subjects below the age
of 21, between 1973 and 1988, in the San Francisco area was analysed in relation to
distance from a television transmitter tower: no excess risk was revealed within an inner
circle of 3.5 km (RR = 0.73). Another study drawn attention to by Elwood is reported in a
review of studies carried out on the subject, but was not the subject of a referenced

3- Retrospective occupational cohort studies

Polish military staff were monitored from 1971 to 1985, using metrics indicating possible RF
exposure (mainly to pulsed signals from 150 to 3,500 MHz, with a power density less than
20 W/m2). The excess risk of all types of cancer was calculated for exposed vs non-exposed
military personnel (RR = 2.1 [1.1- 3.6]); it was highest for leukaemia and lymphomas (RR =
6.3), but also for certain cancers of the digestive system (oesophagus, stomach, colon and
rectum), an observation which has never been reported elsewhere. No excess was shown,
however, for bronchial cancer. The exposure information may have been biased, according
to Elwood, when the careful exploration of patients' cancer risk factors was carried out in
military hospitals.

An earlier study compared 20,000 radar operators in the American Navy with 20,000 other
sailors (who had undergone lesser exposure) between 1950 and 1954; several exposure
classes were defined in accordance with job descriptions. Cancer mortality was evaluated in
1974. It did not vary between the groups, nor between the particular categories of cancers of
the digestive organs, leukaemia or lymphoma. However, lung cancer mortality was higher,
with a gradient that followed the intensity of exposure.

The incidence of various cancers was studied in a cohort of male radio amateurs from the
states of California and Washington 1979 and 1984. For all types of cancers, mortality was
lower than expected, but it was higher for one of the 9 forms of leukaemia studied: acute
myeloid leukaemia, as well as for the ‘other lymphatic cancers’ category. Unfortunately, too
little information was available about the exposure of the subjects who, in their activities or in
their jobs (often connected with electronics) could be subjected to other chemical or physical
exposure. A valid interpretation of this study is therefore impossible.

A cohort of 2,600 radio and telegraph operators in the Norwegian Merchant Navy, active
between 1920 and 1980, was studied for the incidence of cancers. A slight excess risk was
observed for all types of cancer (RR = 1.2 [1.0-1.4]), as well as for malignant tumours of the
breast (1.5 [1.1-2.0]) and uterus (1.9 [1.0-3.2]). Leukaemia, lymphoma and brain cancer
were no more common than in the reference population (not specified in the Elwood review).
A case-control study was nested within this cohort study. The similarity of excess of cancers
in the breast and uterus, in the absence of excess of other forms of cancer whose link with
RF has sometimes been shown, suggests the role of reproductive factors. This excess
remained after the age of the women when their first child was born had been taken into
account. Some EM field measurements carried out on vessels still equipped with old radio
equipment showed magnetic field levels (> 8 MHz) above occupational exposure limits.

The last cohort study, considered by the author to be the most methodologically valid, was
of Canadian and French electricity company employees. In total, 2,679 cases of cancer of all
types were recorded, thus making it possible to carry out a case-control analysis within the
cohort. Exposure was characterised using job-exposure matrices and by measuring the
electric field over a week at the work location of 1,300 workers in 1991 and 1992.

The high-exposure classes corresponded to electric field strengths of more than 200 V/m in
the 5-20 MHz band, but could also include fields of 150-300 MHz and RF (radio
transmissions). An ‘all cancers’ excess risk was observed (RR = 1.39 [1.05-1.85]), as well as
for bronchial cancer (after adjustment for tobacco use and a wide range of other risk
factors). However, no link was demonstrated with cancers reported elsewhere to be linked to
EM fields (leukaemia, lymphoma, brain cancer and melanoma). Elwood notes that the EM
fields considered in this cohort study are, for the most part, a long way from the RF band.

4- Case-control studies
Specific mention of RF was found in 6 case-control studies. US Airforce personnel employed
from 1970 to 1989 who developed brain cancer were each compared with 4 controls (same
force, age, and ethnic category). Occupational exposure was characterised using job-
exposure matrices defined by a group. As well as a strong, unexplained association with
rank, a modest relation to RF exposure was observed (RR = 1.39 [1.01-1.90]). One of the
weaknesses of this study is the fact that personnel who had left the army were not included,
making selection bias a possibility.

In three regions of the United States, fatal cases of brain cancer (white men over 30 years
old) were compared with control groups (same age and area of residence) who had died
from causes other than cancer, epilepsy, cerebrovascular accident, or violent death. A close
friend or relation was asked about occupational exposure (difference in response rate
between the cases and the controls: 74% vs 63%). The jobs were classified by potential
exposure to RF. The risk of cancer was linked with the ‘exposed’ jobs (RR=1.6) only among
the employees in the electrical and electronics industries, but not among those with other
occupations (RR=1.0). This suggests that the risk factor could be due more to other aspects
of the jobs than to RF (solvents, vapours etc.).

Testicular cancer risk was investigated in 271 cases aged from 18 to 42 and in 259 controls,
in three hospitals (including two military establishments). Job categories and subjects'
statements were used to classify exposure to microwaves and other radio waves. The
results were inconsistent, with excess risk when the subjects declared that they were subject
to exposure, but not in relation to the job title - the jobs considered as being the most highly
exposed to RF were associated with a low risk (RR = 0.8).

Risk factors in the incidence of breast cancer in men, a very rare disorder, were studied in
227 cases and 300 controls in 10 regions of the United States. The fact of having worked in
a job involving RF defined exposure. Although the a higher risk was observed among
electricians, telephone line installation, and power station staff, it was not significantly higher
among workers in the radio and telecommunications sectors (OR = 2.9 [0.8-10.0]; for 7
cases). The participation rates in the study were described as low by Elwood, who
considered this work a preliminary study.

Risk factors in female mortality from breast cancer were the subject of a case-control study
covering a total of 33,000 cases and 117,000 controls in 24 States of the United States
between 1984 and 1989. The only exposure data available were derived from death
certificates, which were used to classify the women by a job-exposure matrix. Alongside
various potential exposures to chemical substances (styrene, chlorine-containing solvents
etc.), the probability of exposure to RF was estimated and categorised in four groups
according to a gradient. Although exposure classes 1 and 3 suggested the existence of an
effect compared to the non-exposed group (average OR = 1.14 and 1.15 respectively), the
intermediate class showed no excess risk (OR = 0.95). The authors concluded that ‘the
investigations showed no association with ionising or non-ionising radiation'.

The effect of RF on the risk of intraocular melanoma was studied in 221 white men in a San
Francisco hospital and 447 controls from the same geographical area. Many occupational
exposure factors were explored by job category. Only those that turned out to be associated
with cancer were reported, thereby preventing any evaluation of the possibility of fortuitous
associations. An association was demonstrated with exposure to microwaves and radar
signals (OR = 2.1 [1.1-4.0] out of 21 exposed cases). This result is not featured in the
summary written by the authors, who also point out the possibility of recall bias in this type of

Elwood reports other studies which could be of interest. One is a comparison of

chromosome damage between 38 Telecom Australia employees - line technicians subject to
exposure close to or below occupational limit values at frequencies of 400 to 20,000 MHz -
and 38 office workers who were not exposed. The biological tests were carried out blind and
showed no malfunctions in cell division among 200 metaphases for each subject examined
(OR showing an aberration = 1.0 [0.8-1.3]). Two other datasets are mentioned, but in such
an unclear manner that it is impossible to interpret them.

At the end of his review, the author recaps the observations and submits them to Hill’s
causality criteria. This leads him to conclude that individual studies are weak in terms of
methodology, particularly when it come to the characterisation of exposure, thus making it
impossible to interpret them clearly in terms of a cause and effect relationship. “The major
impression is that these studies are inconsistent. No type of cancer is linked consistently
with exposure to RF”.

Opinion of the group of experts on this article: The studies considered in this review do
not directly concern exposure to RF from mobile phones and base stations. The frequency
ranges and exposure conditions are distinctly different. However, these studies are relevant
because they may highlight categories of cancers that deserve particular attention in specific
epidemiological studies on mobile telephones.

The information provided by Elwood on the quality of the original studies is very variable.
There is sometimes a certain confusion between presentation and criticism of the studies,
which does not facilitate ‘objective’ reading. Despite this, we can only agree that the
message from this series of studies is far from convincing.

4.3 Recent scientific reports

This section is a compilation of the results of work reported after the publication of the
various reviews discussed in the previous section. These reviews could not therefore have
considered them. The studies are grouped in two main parts: a- general articles and
experimental studies; b- epidemiological studies, many of which are recent. Each article is
critically reviewed. At the end of the analysis, the group of experts gives a general appraisal
of these recent articles, putting them into the context of the totality of knowledge garnered
from the reports discussed in the previous section.

4.3.1 General articles and experimental work:

Non-thermal heat-shock response to microwaves
David de Pomerai, Clare Daniells, Helen David, Joanna Allan, Ian Duce, Mohammed
Mutwakil, David Thomas, Phillip Sewell, John Tattersall, Don Jones, Peter Candido. Nature,
25 May 2000.

A short article has recently been published in the prestigious journal Nature, describing the
results obtained by British and Canadian teams on small earthworms called nematodes, of
the species Caenorhabditis elegans. These animals were exposed to low-level microwaves.
The aim of the study was to detect indirectly the expression of heat shock proteins (HSP)
caused by exposure. These proteins are produced when an organism is submitted to an

insult such as heat or a toxic substance. Cellular proteins can become damaged, and HSPs
act as guides for these proteins to maintain or re-establish their three-dimensional structure.

The authors prepared transgenic nematodes to monitor HSP production experimentally.

They prepared two strains, bearing “reporter genes” (which express easily-detectable
genes: those of the ß-galactosidase enzyme and of a green fluorescent protein). These
reporter genes are regulated (expressed following the expression of HSP promoter genes).

Overnight, the worms were exposed to CW microwaves at 750 MHz in a TEM20 cell, one of
the most commonly used exposure systems for cells in culture. During the course of several
experiments carried out at increasing temperatures in a standard incubator, it was observed
that the exposed worms reacted very differently than did the control worms: the activity of
the ß-galactosidase enzyme increased rapidly with the temperature of the incubator, as if
they were heated by the microwaves. There was a difference of 3°C between the two
batches (cf. appendix 3). The SAR was estimated at 1 mW/kg, which corresponds to a very
small rise in temperature.

To explain their observations, the authors put forward three hypotheses:

1) microwaves act on the bonds which maintain proteins in their folded structures
2) production of reactive oxygen species
3) the signal21 has an effect on transduction processes.

In fact, none of these “explanations” is experimentally (or even theoretically) founded at

present. Nevertheless, the authors are going to test them using this simple, quick model.
Although these results are interesting, they cannot at present be extrapolated in terms of
public health. However, the authors do not hesitate to suggest that exposure limits should be
revised if such non-thermal biological effects exist; a bold leap from earthworm to man.….

Mobile-phone type electromagnetic fields do not influence genetic stability in yeast.

Gos P., Heyer W.D., Kohli J., and Eicher B. (1999). In: Proceedings of the Second World
Congress for Electricity and Magnetism in Biology and Medicine, Bologna, Italy, June 1997,
F. Bersani, Ed.

The Grundler and Kielman group have published results of work on the effects of millimetric
waves on yeast22. The Gos group in Bern have unsuccessfully attempted to replicate these
experiments23. More recently, a study financed by the FGF and Swiscom was carried out by
the same group. The aim was to determine any possible effect on the same biological model
of GSM-900 type microwaves. Cultures of Saccharomyces cerevisiae were therefore
exposed at 900 MHz under far-field conditions.

Prenatal exposure to 900 MHz, cell-phone electromagnetic fields had no effect on

operant-behaviour performances of adult rats.
Bornhausen M., Sheingraber H. Bioelectromagnetics, 2000, 21, 1-9.

transverse electromagnetic
propagation of messages from outside the cell to the cytoplasm and the nucleus
(23°C, SAR of 0.13 and 13 W/kg). The tests carried out were those of reverse and direct mutation in
response to canavanine. No effect was detected
Grundler W, Keilman F, Putterlik V, Strube D (1982): Resonant-like dependence of yeast growth
rate on microwave frequencies. Br J Cancer 45:206-208.
Gos P, Eicher B, Kohli J, Heyer W-D (1997): Extremely high frequency electromagnetic fields at low
power density do not affect the division of exponential phase Saccharomyces cerevisiae cells.
Bioelectromagnetics 18:142-155

This article concerns the possible effects on brain development of radiofrequency
electromagnetic fields used in mobile telephone technology. Gravid female Wistar rats were
continually exposed to 900 MHz radiation, modulated at 217 Hz, during gestation. This
exposure is representative of typical public exposure to mobile telephony signals. The total
SAR (whole body) corresponding to this exposure was between 17.5 and 75 mW/Kg. The
animals which, as embryos, were subjected to these conditions were then tested to assess
their cognitive faculties in terms of learning abilities (obtaining food). These tests are
described as being particularly discriminatory for two parameters: the frequency of lever
activation and the inter-response interval. The results obtained with these different tests,
particularly the two parameters described, showed that in-utero exposure produced no
measurable effect on cognition in the exposed animals compared to controls.

Biological effects of electromagnetic fields - Mechanisms for the effects of pulsed

microwave radiation on protein conformation
Laurence JA, French PW, Lindner RA, and McKenzie DR Journal of Theoretical Biology,
206: 291-298 (2000)

Laurence et al, in Australia, studied the effects of pulsed microwaves on the induction of
heat shock proteins. Induction of HSP-70 was observed in mouse cells after exposure to
bursts of microwaves at 2450 GHz, each lasting 6 minutes.

The amplitude of the effect increased with the dose (SAR from 12 to 58 W/kg). The authors
claim to have shown that the 6 minute averaging time recommended by ICNIRP is
inadequate. They also carried out simulations of the underlying mechanism for this effect.
The main hypothesis was that the synthesis of heat shock proteins is triggered by the
transitory warming of proteins without a macroscopic temperature rise. They estimate that
the time required for a protein with a diameter of 10 nm to reach thermal equilibrium with its
environment after absorbing microwaves is 1 nanosecond, whereas the time required for the
unfolding of a protein is 50 nanoseconds.

Comments of the group of experts: several of the hypotheses put forward in this article
are somewhat unreasonable, notably that the temperature of the water bonded with the
proteins is high after absorption of the microwaves without there being a change in the
temperature of the bath. In fact, the equilibrium between the two forms of water (free and
bonded) is established almost instantaneously. Likewise, the explanation of the existence of
power “windows” is not supported by the experimental observations nor by theoretical
considerations (incomplete triggering of the response to the heat shock). It is probable that a
defective dosimeter is the source of the observations made in this article and that the effects
are in fact likely to be thermal in nature.

Exposure to pulsed high-frequency electromagnetic field during waking affects

human sleep EEG
R Huber, T Graft, KA Cote et al (Neuroreport, 2000, 11, 3321-3325)

Volunteers were exposed to a mobile telephone field emitted by a planar antenna on one
side of the head or the other. After a short night limited to just 4 hours of sleep, the
volunteers were exposed for half an hour early in the morning, in a seated position, before
going back to sleep. A GSM signal was emitted, with modulation at the various frequencies
encountered in actual telephones: 2, 8, 217 and 1736 Hz, with a pulse duty factor of 87.5%.
The maximum local SAR in 10 g was calculated as being 1 W/kg, corresponding to an
average SAR on the hemisphere of the exposed side of 0.28 W/kg. The sleep parameters
(latencies, duration, efficacy etc.) were not significantly different. A significant difference of
10 to 12% in EEG spectral power density in the frequency bands 9.75-11.25 Hz and 12.25-
13.25 Hz was observed in the first 30 minutes of non-REM sleep. This effect was not
predominantly on the exposed side.

Comments of the group of experts: the exposure conditions are not at all physiologically
realistic. An effect under these conditions has no significance that can be related to a real
situation. The authors justify their protocol among other things by arguing that daytime sleep
was favoured because the subjects had been previously deprived of night-time sleep.
Generally, it takes just a few nights for volunteers to get used to an unusual environment
and to sleep correctly. The main interest in this study is that it shows a delayed effect after
exposure, thus suggesting the possibility of a cumulative effect. This hypothesis remains to
be demonstrated in more physiologically-realistic conditions, e.g. exposure in the evening
before going to sleep.

Exposure to electromagnetic fields by using cellular telephones and its influence on

the brain.
M. Petrides, Neuroreport, 11 (15), F15, 20 October 2000

This editorial in the journal Neuroreport was triggered by the article by Huber et al,
published in the same issue (cf. the critique of this article presented above). It places this
study in perspective by referring to 6 other articles published since 1998 of brief exposure to
RF on the physiology of the brain and cognitive function.
The mechanisms for these effects remain uncertain, but it has been proposed that that a
microthermal effect on synaptic transmission could play a facilitating role in cognitive
function. The editorial underlines the fact that the effects shown by Huber et al are no longer
visible after a 3-hour period of sleep and thus draws the conclusion that, in the light of our
current knowledge, it is not possible for these short-term manifestations to be predictive of
long-term consequences after repeated exposure. New work is required to answer this

Radiofreqencies and genotoxicity (FDA programme)

A joint call for tender by the Food and Drug Administration and the American telephone
industry (CTIA) has just been issued for the replication and extension of two recent studies
(submitted for publication) showing the induction of micronuclei (genotoxicity test) after
exposure of mammalian cells to radiofrequency signals used in mobile telephony.

The conditions under which this effect was demonstrated were unusual and caution is called
for when it comes to interpreting these data.

Investigation of DNA damage and micronuclei induction in cultured human blood

cells Hook G J, Donner M, McRee D I, Guy A W, Tice R R (article accepted by
Bioelectromagnetics )

The mobile telephony signals studied have a carrier frequency of about 837 MHz (analogue,
CDMA and TDMA) or 1900 MHz (PCS) and are either voice-modulated or otherwise.
Cultures of circulating human lymphocytes (2 donors) were exposed to SARs of 1, 2.5, 5
and 10 W/kg for 3 and 24 hours.

The two parameters studied were

(i) induction of DNA damage (single-strand breaks, alkali-sensitive sites) assessed by the
comet test: this test identifies damage to individual cells after electrophoresis. An affected
cell takes the shape of a “comet” with the intact DNA forming the head, while the fragmented
DNA which has migrated makes up the tail.

(ii) micronuclei induction (MN, chromosome fragments or whole chromosomes which do not
migrate correctly on division of the cell) assessed by the cytokinesis-blocked micronucleus
assay. Only those cells which underwent cell division after treatment (binuclear cells) were

considered, which increased the sensitivity of the test (M. Fenech, The in vitro micronucleus
test, Mutation Research, 2000).

In this study, the comet test revealed no induction of DNA damage whatever the exposure
conditions. The MN test was negative for all the signals after 3 hours of exposure. It is worth
noting that only 2 experiments were carried out in most cases. The signals tested at 5 W/kg
and 24 hours (TDMA and analogue) inducted MN in the lymphocytes (p<0.001, only one
experiment). Finally, all the signals (modulated or not) showed an ability to induce MN after
24 hours' exposure (p<0.001, two experiments per condition).

According to the authors, this effect could be due to the cells heating, as the SAR range of
the samples was very wide (7 to 31 W/kg - Guy et al, 1999, Bioelectromagnetics, 20, 21-39).
However, no hypothesis was formulated by the authors to explain the fact that the comet test
was negative under certain conditions although a significant increase in MN had been

The effect of radiofrequency radiation with modulation relevant to cellular phone

communication (835.62 MHz FDMA and 847.74 MHz CDMA) on the induction of
micronuclei in C3H 10T1/2 cells
Bisht K., Moros E.G., Straube W.L., Roti-Roti J.L (results presented to the Annual BEMS
Meeting (Munich, June 2000) and submitted to Radiation Research)

This is an investigation of the induction of micronuclei (MN) in C3H 10T1/2 cells (mouse
fibroblast cell lines) exposed to radiofrequency signals used in mobile telephony in the USA.
Cells were exposed to the FDMA signal (carrier frequency: 835,62 MHz) at SARs of 3.2 and
5.1 W/kg or to the CDMA signal (carrier frequency: 847.74 MHz) at SARs of 3.2 and 4.8
W/kg. The exposure was isothermal (37 ± 0.3°C) and exposure times were 3, 8, 16 and 24
hours for the cells in the exponential growth phase or in the plateau phase (G0 phase of the
cell cycle). Ionising radiation (137Cs gamma rays) was used as a positive control. The test
used was the so-called “cytokinesis-blocked micronucleus test” in which only the binuclear
cells, whose division is chemically blocked, are considered.

Under the experimental conditions of this study, the test proved to be capable of detecting
the inducted MN from 0.6 Gy gamma photons.

The results showed that there was no increase in micronuclei when the cells were exposed
to the FDMA signal, whatever the exposure conditions. Exposure of less than 24 hours to
the CDMA signal was also without effect whatever the exposure level and the cell growth

In contrast, an increase was noted in the number of MN (binuclear cells with micronuclei and
number of MN for 100 binuclear cells) in the C3H cells in the plateau phase exposed at the
highest level (5.1 W/kg) for 24 hours. The increase was around 20% and was significant
(p<0.05, Student t with n=6). It should be noted that in cells in the plateau phase, the 50%
increase in MN observed after a dose of 0.3 Gy of 137Cs gamma rays was not found to be
significant (n=3).

Physics and biology of mobile telephony

G.J. Hyland , The Lancet , 2000, 356:1833-1836.

Professor Hyland of Warwick University puts forward, in this article, a theory concerning the
non-thermal effects of mobile telephones. It gives a brief reminder of basic information about
mobile telephony and its thermal effects. He then touches on the non-thermal effects,
presenting his basic hypothesis, which is that the organism reacts to modulated waves as
the latter can interfere with the oscillations of certain biological processes. A comparison is

made with the interference phenomena encountered in electromagnetic compatibility. There
then follows a list of examples of observed effects, such as:

· epileptic activity in rat brain sections, observed by Tattersall (however, the effect was
observed at 700 MHz GSM as well as with non-modulated RF….).

· The effects of millimetric waves on the growth of Saccharomyces cerevisiae were

reported by Grundler in 1992, but Gos’ team were unable to replicate these results

In all the results selected, except those of Repacholi, the amplitude of the biological effects
was low and did not correspond to foreseeable health effects.

Comments of the group of experts: The argument is undermined by the selective choice
of articles from the literature, since the negative experiments, in particular in replication, are
not quoted and the examples are taken from the ELF, RF and millimetric bands without this
being explicit. Furthermore, certain quoted references relate to unpublished works. Thus, in
the second table, the work of the group of M. Bastide is quoted although it has not been
published and concerns ELF.

Faced with the impossibility of reproducing certain results, the author invokes the “non-
linear” character of the phenomena (chaos, “butterfly effect”) to explain that the results
depend so strongly on the initial conditions that it is not possible to reproduce them! This is
a scientifically unacceptable argument. The citations of epidemiological data are also biased
and the conclusion on the famous episode of the irradiation of the American Embassy in
Moscow is quite irrelevant to RF exposures from mobile telephony. In conclusion, it is most
surprising that such a scientific journal as The Lancet should have published this article
which fails to comply with the elementary rules of scientific communication in terms of
content and form.

Criticism of the health assessment in the ICNIRP guidelines for radiofrequency and
microwave radiation (100 kHz–300 GHz)
Cherry N, 2000. http://www.emfguru.com/ CellPhone/cherry2/ICNIRP-2.htm

Neil Cherry is a professor of climatology in New Zealand. For several years, he has been
actively fighting for low RF exposure limits to be applied. N Cherry has just published a
severe critique of the ICNIRP recommendations24 on the internet. The author’s main
argument is that ICNIRP maintains, despite all the evidence, that the only established and
conceivable biological effects are thermal in nature, whereas non-thermal effects should also
be taken into account when assessing health risks.

Despite the impressive length of this contribution (143 pages), which claims to cover all
biological and health aspects of the non-ionising electromagnetic spectrum, the presentation
which follows is short, as it is for the other recent works reviewed by the group of experts. A
few examples drawn from the report by N Cherry highlight the methodological and
theoretical weaknesses of his arguments:

N Cherry explains that the amplitude of biological effects increases with the frequency over
the whole of the electromagnetic spectrum (‘EMR Spectrum Principle'). This hypothesis is
not backed up either by knowledge of the mechanisms, which depend on the frequency, or
by the biological results obtained in the various frequency bands.

International Commission on Non-Ionizing Radiation Protection

Disparate studies are bundled together by the author without discrimination (thus, ELF and
RF fields are all considered together)

The results of experiments are not always taken into account (negative and positive).

The report often features badly-summarised or over-interpreted results.

In epidemiology, there is a real difficulty in estimating subjects' exposure to the fields,

especially in retrospective studies. This is also true of case-control or ‘ecological’-type
studies on which the author makes abundant comment. This may lead to an under-
estimation of the risk. N Cherry concludes that any indication of an excess risk, even non-
significant, implies a causal relation. He thus makes a very unorthodox reading of the
causality criteria proposed by B Hill which are evoked at length in support of his thesis. Many
studies analysed by specialist groups as being ‘non-suggestive’ of an association or as
suggesting a relation whose causality is subject to caution, are presented by N Cherry as
being demonstrative25.

This thesis is a good illustration of the fragility of an isolated critical approach in a scientific
field characterised by its great complexity. In its intermediate report, the group of experts
insisted on the necessity, in such a context, of mobilising skills in different disciplines, as well
as different points of view on the subject. This need for contradictory scientific critique is
clearly highlighted here.

4.3.2 Epidemiological Studies

Epidemiological evidence on health risks of cellular telephones

Rothman KJ, Lancet, 2000, 356 : 1837-1840

Editor-in-Chief of the journal Epidemiology and himself a famous epidemiologist, K Rothman

reports, for “The Lancet”, on the state of epidemiological knowledge of the risks associated
with RF.

In a short introduction, he gives a brief reminder of certain distinctive features of exposure to

RF in relation to very low frequency electromagnetic fields (ELF EMF), a domain which has
been the object of a very large number of epidemiological studies. As far as mobile
telephone users are concerned, exposure is localised and concerns clearly-identified
tissues, it has considerably increased in recent times, occurs over short periods (phone
calls), and can be measured indirectly (via billing) or directly (SAR). All of these conditions
should make the work of epidemiologists much easier than has been the case for ELF EMF.
However, the time-scale is too short to be able to give definitive answers for certain effects
which may be delayed, especially given the fact that the technologies have evolved in recent

First of all, the article examines cancer. The studies relating to RF from radio and television
antennas, and also in certain occupational environments (radar, the electronics sector) – 11
articles discussed – are instructive but have little value as far as mobile telephony is
concerned. Exposure near base stations poses, for K Rothman, “formidable problems” of
method and he expects little in the way of conclusive results from such studies, given the

“The studies cited by ICNIRP contain sufficient evidence to conclude cause and effect between
RF/MW and cancer across many body organs, especially leukaemia and brain tumours, and at
chronic lifetime exposures showing dose-response relationships pointing to a Level of No Observed
Adverse Effects threshold of about 20 nW/cm .”

influence of so many possible factors. Three studies concerning mobiles are presented –
including one by the author of the article – and are considered inconclusive (doubts are
expressed about the interpretation of the results of the Hardell study). The results of 3 other
studies are expected, two in the near future (completed case-control studies) and the last
one in several years time (the IARC Interphone project).

The effect which is the most clearly established by epidemiology (3 articles presented) is the
risk of accident when driving a vehicle, with an excess risk of over 100 %.

The conclusion of the author is that it is too early to pronounce a verdict on the risks
associated with mobile telephones, notably with regard to cancer. But K Rothman, on the
basis of the risk factors envisaged (for brain cancer) or demonstrated (for accidents),
estimates that even if mobile telephones were proven to represent a health hazard, the
number of cases expected would be much lower than those due to accident.

Radiofrequency exposure and mortality from cancer of the brain and

lymphatic/haematopoietic systems.
Morgan RW, Kelsh MA, Zhao K, Exuzides KA, Heringer S, Negrete W. Epidemiology, 2000,
11 : 118-127

An occupational cohort in the Motorola company, where the probability of exposure to RF is

greater than in the population as a whole, was monitored from 1976 to 1996. The RF under
study was related to the production of telecommunications device rather than being specific
to the use of mobile telephones. With 195,775 workers and 2.7 million person-years of
observation, this cohort represents the largest available to date in the study of RF and
mortality. The health variable studied was mortality, by cause of death, paying particular
attention to brain cancer, lymphoma and leukaemia, among 14 causes of death due to

A detailed analysis of the work history of these employees, using company records, allowed
them to be categorised by level of exposure (nil – that is to say equal to the general
population – weak, moderate, and strong) and duration: a study was carried out to validate
the job-exposure matrix developed for the study, comparing the classification obtained in this
way with on-site measurements. Two mortality comparison systems were adopted: external,
with the population at large of the 4 American states in which most of the plants studied
were located (calculation of SMR), and internal, the most valid system, by comparing the
exposure categories within the cohort itself. The very elaborate statistical analysis made it
possible to take into account different factors of latency and length of service in the

Overall, neither the external comparisons (a strong ‘healthy worker effect’ was observed,
with an SMR ‘all deaths' of 0.66 [IC95%=0.64-0.67)], nor the internal comparisons (by level,
duration, mode – usual, maximum or accumulated value – seniority, or latency of exposure)
suggested that RF exposure played a role, notably in the three causes of death which
initiated the study.

Although this study provides no argument in favour of the existence of a risk linked with
occupational exposure to RF in this population, the authors underline the modest proportion
of people classed as being “moderately or strongly” exposed (about 9%), the small number
of subjects who died (3.2 %), and the relative youth of Motorola staff and point out that, in
their opinion, the possibility of long-term effects cannot be excluded.

In an Editorial in the same journal26, RD Owen, the head of the Radiation Branch of the
United States FDA, expressed his satisfaction with this study which he qualified as a
‘beginning’, pointing out that it is not possible, as things stand today, to extrapolate data from
one RF band to another or to predict the existence or absence of long-term effects. He
insisted on the need for further research, with focusing particularly on the conditions for
estimating exposure, both in experiments and epidemiological studies.

Case-control study on radiological work, medical X-ray investigations, and use of

cellular telephones as risk factors for brain tumours.
Hardell L. Med Gen Med, May 2000.

This was a population case-control study carried out in Sweden and concerning various risk
factors for brain tumours. 209 subjects (men and women) with brain tumours (malignant or
benign) and 425 controls were included in the analysis.

The cases, selected from the National Cancer Register, came from 2 different regions of
Sweden between 1994 and 1996 (benign tumours were only considered in 1996) and the
patients were alive at the time of their inclusion in the study. Anatomopathology reports were
available for 197 cases (136 malignant tumours and 62 benign tumours). The controls were
matched by sex, age and region and were drawn at random from the National Population

Exposure to risk was evaluated using a self-report questionnaire sent to the homes of the
subjects (completed by a telephone interview if necessary). The collection and coding of the
questionnaires and the additional telephone interviews were carried out blind with regard to
case or control status. The risks considered were: occupational factors (profession,
exposure to ionising radiation for health workers, chemical exposure), radiological
examinations, use of cellular phones.

The results show some significant associations with certain professions and with X-ray
examinations of the head and neck (we will not comment on the results concerning these
factors). As far as cellular-phone use was concerned, a significant association was observed
(OR = 2.62 ; CI : 1.02 - 671), and confirmed after adjustment for all the risk factors, with the
occurrence of tumours in the temporal or occipital zones and the temporoparietal lobe on the
side where the user habitually held the telephone (most exposed area of the brain). In
contrast, no association was observed for tumours on the opposite side to the ear usually
used or tumours in general, whatever their location. These results are based on 13 exposed
cases (10 malignant tumours and 3 benign); 9 cases were exposed only to analogue
telephones and 3 to analogue and GSM telephones.

Comments of the group of experts: This high-quality study is extremely sound in various
aspects: case and control recruitment in a population register, blind, standardised collection
of data, inclusion of the main known or suspected risk factors in the brain.

The main arguments in favour of causality are a clear association, the fact that the main
confounding factors were taken into account, the overall quality of the study and, above all,
the fact that the excess corresponds to the location at the greatest risk, in principle, and was
not detected at locations further from the exposure, thus making the observed positive result
biologically likely.

The principal limitations are the small number of exposed cases on which the results are
based (although the study as a whole is of a good size). For this reason, it was impossible to
Possible health risks of radiofrequency exposure from mobile telephones. Owen RD. Epidemiology,
2000, 11 : 99-100

take into account the analysis of exposure-effect relationships, nor the temporal aspects of
exposure, nor the histological form of the tumours. The homolateral location of the cancer in
relation to the declared use of the telephone is striking, but it is quite possible, given the
conditions of the study, that this could be a recall artefact.

In conclusion, this study provides arguments in favour of the causality hypothesis, but
remains too limited to go any further. In particular, it is not possible to exclude a potential
bias in the statement of the usual side on which the telephone was used. The study is
continuing and it will obviously be interesting to have more results and to compare them with
the other studies published on the subject.

Handheld cellular telephone use and risk of brain cancer

JE Muscat, MG Malkin, S Thompson, RE Shore, SD Stellman, D McRee, AI Neugut, EL
Wynder, JAMA, 284 (23), 3001-3007.

This keenly-awaited article (it was presented in July 2000, by G Carlo – who had contributed
to its financing by the WTR - as “suggesting that there is a risk of brain cancer”, in particular
on the side on which the mobile is used27, as the work of Hardell tended to show in 1999),
was published in mid-December. It presents the results of a case-control study conducted
between 1994 and 1998 in 5 university hospitals on the east coast of the United States,
covering 469 subjects (aged from 18 to 80) suffering from brain cancer and 422 matched
controls. Exposure to the fields associated with mobiles was characterised by a
questionnaire and measured by the number of hours of monthly use and the number of
years of use.

Compared to non-users, and after adjustment for confounding factors, the 'Odds Ratio' (OR:
standard measurement for "excess risk") for cellular telephone use was 0.85 (95% CI: 0.6 -
1.2). The average length of use was 2.8 years for cancer sufferers, as compared to 2.7
years for controls. The ipso- or contra-lateral position of the cancer depended on the area of
the brain that was affected. All histological types of cancer had an OR less than 1, except for
one rare form: neuroepithelioma (OR = 2.1 [0.9-4.7]).

The authors concluded from this work that there was no excess risk of brain cancer
associated with the use of mobile telephones, but considered that further research was
required, particularly to take longer periods of use into account.

Cellular telephone use and brain tumours.

PD Inskip, RE Tarone, EE Atch, TC Wilkosky, WR Shapiro, RG Selker, HA Fine, PM Black,
JS Loeffler, MS Linet. New England Journal of Medicine, 2001, 344: 79-86 (available on
internet on December 19th 2000).

This case-control study, carried out between 1994 and 1998, involved 782 patients with
intracranial tumours (cancers of the brain, meningioma, and neurinoma of the acoustic
nerve) and 799 matched controls (residence, age, and sex) with non-cancer diseases, in
three cities in the United States. This is the largest-scale study to date.

The cumulative use of a mobile telephone for at least 100 hours was not associated with the
presence of a tumour (OR = 1.0 [95% CI = 0.6-1.5] when all forms of cancer were
considered. This result varied depending on the type of cancer, but the relationship was
never statistically significant, once the various confounding factors had been taken into
account. The authors did not find any link between the presence of cancer and the intensity
Scientific Progress - Wireless Phones and Brain Cancer: Current State of the Science. George L.
Carlo, PhD, MS, JD, and Rebecca Steffens Jenrow, MPH, Wireless Technology Research, LLC
Washington, DC, MedGenMed, July 31, 2000, Medscape.

of use (over 60 minutes per day or over 5 years), or between mobile telephone use and the
side that the tumour was located.

As in the previous article, the authors concluded that, while their research did not show any
link between mobile telephone use and brain tumours, it was not possible to come to a final
conclusion concerning the consequences of long-term exposure (only 8 % of the patients
had started using a mobile telephone before 1993).

In an editorial in the same journal, two leading specialists in cancer epidemiology expressed
the opinion that this research should be considered reassuring as it confirmed other
publications on the same topic, and was consistent with the weakness of empirical
observations and the lack of a theoretical basis for carcinogenic effects unrelated to heat.

Prevalence of headache among handheld cellular telephone users in Singapore: a

community study. Chia, S-E, Chia H-P, Tan J-S. Environ. Health Perspective, 2000, 108:

This cross-sectional, epidemiological study was carried out on a random sample of

inhabitants of one part of Singapore, with the aim of comparing the prevalence of various
subjective symptoms (headaches, dizziness, fatigue, loss of memory, etc.) with the subjects'
use of mobile telephones (MT).

The study population, consisting of 808 men and women between 12 and 70 years old,
made very frequent use of mobile telephones (44.5 %). Special attention was paid to
controlling bias in the selection and declaration of symptoms. Headaches were the only
symptom significantly associated with the use of MT (OR = 1.31 [95% CI: 1.00-1.70]), with
increasing prevalence in relation to the declared length of use (up to 1 hour per day). It was
quite remarkable that MT users equipped with hands-free systems reported fewer
headaches (41.7% if they used them all the time, 54.4 % for irregular users, and 65.4 for
non-users). The authors envisage two etiological hypotheses: the effects of RF waves on the
blood-brain barrier and on the dopamine-opioid system. In spite of the limitations of cross-
sectional studies, particularly the difficulty of establishing the time sequence of the factors
studied, this research indicates that RF may play a role in headaches in the general
population. It remains to be determined whether, in view of the environment under study
(electromagnetic radiation density, noise, atmospheric pollution, etc.), these results can be
extrapolated to other situations. It should be noted that there is a high frequency of
headaches in this population. For example, in the French GAZEL cohort (40-60 years old),
the rate (prevalence over the previous 12 months) was on the order of 15-20 % in men and
33-38 % in women.

The Possible Role of Radiofrequency Radiation in the Development of Uveal

Andreas Stang, Gerasimos Anastassiou, Wolfgang Ahrens, Katja Bromen, Norbert Bornfeld,
and Karl-Heinz Jöckel. Epidemiology, Volume 12, Number 1, January 2001,

A very recent article published in the journal "Epidemiology", presenting the findings of a
case-control study carried out by a German team, examined the relationship between
professional exposure to various sources of EMC, including RF, and the occurrence of uveal
melanoma, which affects eye tissue between the cornea and the crystalline lens. In a series
of 118 cases and 475 controls, an excess risk associated with RF from mobile telephones
was identified (OR = 4.2, IC95% = 1.2-14.5).

Although this scientific journal has a good reputation, the group of experts could not assess
this work as they had access only to the abstract, and not the full text of the article.

General conclusions of the group of experts on recent research: Recent literature does
not make any clear departure from previously available information. Publications describing
experimental work give further details concerning the effect of exposure on certain cognitive
functions in both animals and man. Is this a ‘microthermal’ effect? Does it involve hormonal
phenomena? It is difficult to say in the current state of knowledge, just as it is impossible as
yet to conclude that prolonged and/or frequent exposure to these phenomena represents a
real risk to human health. However, these results will certainly be considered in light of the
epidemiological study in Singapore, which suggests quite convincingly that intensive use of
mobile telephones may lead to headaches. As the protocol of this study was relatively
unsophisticated, the findings should be confirmed and validated using other approaches,
under different circumstances, before they are considered as hard evidence.

Publications describing the appearance of micronuclei in cells exposed to RF should also be

replicated. They were not the first to study the effects of relatively long-term exposure (24
hours continuously, or even several days), but the other studies produced negative results28.
However, while such long-term uninterrupted exposure is not very realistic in comparison
with everyday use, it does offer ways of investigating the effect of repeated exposure, as the
cumulative effects are not very well known. Ongoing research is currently exploring this

The various epidemiological studies investigating the risk of brain tumours are in agreement
and do not conclude that RF plays any role in the development of these forms of cancer,
under present observation conditions, i.e. after relatively short induction periods (maximum:
5 to 6 years). While these results are reassuring, they do not exclude the possibility of long-
term effects. They do not, however, give any indication that such effects exist.

4.4 Children and Exposure to RF Associated with Mobile Telephones

The Stewart Report recommended that children (under 16 years old) should be discouraged
from using mobile telephones, unless it was absolutely necessary. This opinion was based
on the following main arguments:

· the exposure of a young child's brain is considered to be higher than that of an adult, if
the cellular telephone is emitting the same power

· children are considered to be more sensitive to external agents than adults

· children's cumulative exposure will eventually be greater than adults', as mobile

telephones have only recently come into use.

The group of experts also studied this important issue.

No facts have yet been clearly established concerning the greater dose absorbed by
children's heads due to their smaller size. Research by Gandhi's team (1996) at the
University of Utah concluded that the SAR received by the brain of a 5 year-old child at a
frequency of 835 MHz was 3.3 times higher than that received by an adult, while the ratio
was 2.2 for a 10 year-old child; there was no age-related difference at 1900 MHz. In 1998,

Particularly: Adey et al (1999): Spontaneous and nitrosourea-induced primary tumors of the central
nervous system in Fischer 344 rats chronically exposed to 836 MHz modulated microwaves. Radiat
Res.152(3):293-302; and: (2000) Spontaneous and nitrosourea-induced primary tumors of the central
nervous system in Fischer 344 rats exposed to frequency-modulated microwave fields. Cancer Res.

Schönborn et al carried out a new modelling and simulation study at the same frequencies,
but using head phantoms that were more representative of children (3 and 7 years old,
respectively). The findings contradicted those of Gandhi et al. Among their criticisms of the
earlier research, these authors indicated that the models used were merely proportional
reductions of adult heads, which do not really correspond to children's. Other authors
(Kuster and Balzano [1992], Hombach et al [1996], and Meier et al [1997]) agreed with
Schönborn et al. Another experiment that Stewart seems to have considered was carried
out on rats of different ages (Peyman et al 2000)29. It showed differences between the
dielectric properties of the brain, salivary glands, and muscle mass of rats 10 and 20 days
old, but there was no further reduction in conductivity in rats over 20 days old. It is very
difficult to extrapolate these (not yet published) findings in animals to age-related
differences in humans.

The other arguments presented by Stewart et al are discussed and explained below. The
age at the time of initial exposure can certainly affect the risk of developing a delayed
pathology in the long term. The major risk is cancer, but any delayed effect may present the
same characteristics. Cancer epidemiology offers arguments that the age at the time of
initial exposure may affect the risk level for various reasons.

Greater sensitivity: children may be more sensitive than adults to the carcinogenic effects of
exposure for reasons related to development (developing tissue, etc.) and physiology
(greater activity is accompanied by higher absorption, although the relevance of this
parameter for RF radiation is not obvious). Perhaps the best-established example is that of
tobacco: the younger a person starts smoking, the higher the risk of lung cancer, all other
exposure conditions being equal (average quantity, cumulative dose, etc.). There are,
however, also some counter-examples. In the case of asbestos, all the available findings
show that age at the time of exposure probably has no effect, so the risk is the same
whether the first exposure occurs during childhood or adulthood.

"Mechanical" effects of time: whether or not there is any greater sensitivity during childhood,
the "lifelong" risk of developing a cancer caused by exposure is increased if exposure takes
place early in life. This is due to a combination of two phenomena that have a cumulative
effect: (i) in general, the earlier the exposure, the higher the cumulative lifelong exposure will
be (if exposure is continuous: this is obviously not true if exposure stops or decreases over
time): the relevant parameter for quantifying cancer risk is generally cumulative exposure
levels; (ii) the earlier the exposure the more time is "available" to develop an effect related to
this exposure. This is all the truer if there is a long latency period between exposure and the
occurrence of the effect. For example, there is almost no risk of developing a mesothelioma
of the pleura if even very intense exposure to asbestos occurs after the age of 80, as the
average latency period is approximately 35 years and the person will probably have died
before a mesothelioma has had time to develop. Conversely, if exposure occurs very early,
the risk will be much higher, even though children do not seem to have any intrinsically
increased sensitivity, as mentioned above.

These arguments should be tempered by the fact that the hypothesis on which they are
based is that exposure will be continuous over time. It is, however, clear that, as
radiocommunication technologies develop, exposures from devices tend to decrease. In
particular mobile telephones will soon cease to be in close proximity to users' heads (see
paragraph on technological developments in chapter II), although the increasing number of
radiation sources in our everyday surroundings may offset this positive trend.

Peyman, Rezazadeh and Gabriel, poster presented at the annual meeting of the BEMS
(BioElectromagnetics Society), in Munich, 9-14 June 2000. "Changes in the dielectric properties of
aging rat tissues".

These data led the group of experts to recommend a "precautionary approach", although
they did not consider that currently available scientific data justified any restrictive legislation.

4.5 Expert hearings

Interviews were held with people from the scientific, administrative and industrial
communities, including trade associations. The meetings took place in two plenary sessions
of the group of experts on the 27th of October and the 23rd of November, at the Ministry of
Social Affairs. Each interview was in two parts. First, the interviewee was invited to present
answers to written questions from the group of experts (sent by mail in September). A
discussion session with the group of experts then took place. The interviews were recorded
and transcribed, and then revised and validated by the interviewee.

Three hearings took a different format: George Carlo, who was the director of the WTR in
the US, responded to the group of experts in a telephone conference on the 23rd of
November in the evening; Marc Séguinot, from DG SANCO of the EU met with the group of
experts on the 14th of December 2000. On these two occasions, the meetings could not be
retranscribed, so they have been reported more briefly. Philippe Quenel, in charge of the
Department of Health & Environment of the InVS responded by mail to the question of
monitoring possible health effects associated with mobile telephony. For consistency, this
contribution, with which the group of experts agrees, is reported in the chapter dealing with
recommendations for research (Chapter VI). Finally, Mr Roger Santini of INSA of Lyon was
asked for his opinion; he responded by mail on the 5th of September 2000: he did not think it
was necessary for him to meet face-to-face with the group of experts to express his opinion.
This section also contains the response given by Laurent Bantoux, of the Directorate
General of Research of the European Commission, to the questions addressed to him by
mail on the direction of research on RF and health.

First session, 27 October 2000

4.5.1 Joe WIART

representing the GSM Association30

1 Response to written questions

I How have the mobile operators in France organised the monitoring of scientific literature
on possible health effects?

a. Follow-up on the research

A working group on the environment as well as a full-time post of Director of Environmental
Affairs have been created within the GSM Association. Their activity consists of:

· following up the work of WHO and of ICNIRP, of scientific symposia and of all meetings
related to these issues;

· follow-up of the scientific literature;

· gathering of scientific opinions on the results of new research;

· informing the members of the GSM Association;

The GSM Association represents the interests of 490 members from European countries and
elsewhere in the world (third generation GSM operators, administrative establishments,

· ensuring a liaison between government, industry, and consumers.

b. Financial support of research on possible effects of electromagnetic waves

The public is more-and-more interested in these matters. The scientific community is of the
opinion that the exposure levels resulting from the use of mobile phones and base stations
do not present a risk to health. Nevertheless, according to the European Recommendation
of 12 July 1999, it is necessary to carry out additional, complimentary research. Moreover,
WHO will analyse all the available results in 2003-2004. A complete database is needed for
such an analysis; to help achieve this aim the GSM Association, in co-operation with the
Mobile Manufacturers Forum (MMF), decided to finance studies within the fifth Framework
Programme of the European Union. This research, co-funded by the GSM Association, is in
addition to the participation of members of the Association in research projects at the
national level.

c. Research projects co-funded by the GSM Association (current or future)

Projects Co-Funded with the European Union, National Authorities and The MMF:
· Perform A: analysis of carcinogenesis in animals
· Interphone: epidemiological studies organised by IARC in several countries

Prospective Projects Co-Funded by National Authorities and the MMF

· Perform B: replication studies on animal and cell cultures
· Studies on DMBA in China

IV-What are the new technologies available in France or at the international level that could
lead, in years to come, to a change in the level or in the conditions of exposure of the public
to RF EMF from mobile telephony?

d. Evolution in the number of mobile users

The number of users of mobile phones continues to increase. For instance, in France in
1992 there were 200 users, as of September 2000 there were 26.2 million. The level of
penetration reached 40.4% in June 2000 and is predicted to reach 50% soon. This evolution
requires an increase in the capacity of the network.

e. The impact of UMTS and Bluetooth: evolution in the use of mobiles and
technological evolution
Major evolution will be driven by UMTS and Bluetooth. New modes of use of radio
telephones will take place - reading from and watching the screen - new frequencies will be
used and new technical characteristics will be developed.

The increase in the number of users and the density of cells will lead to a decrease in the
level of emissions from handsets and base stations. In fact, power-control is an important
determinant of the emission from a radiotelephone; the nearer the user is to a base station,
the lower the signal level needed to make a call. One of the technical constraints on mobile
telephony in general, originating from the need to reuse frequencies in non-adjacent cells, is
to limit as much as possible the possibility of interference. Therefore, in order to reuse a
given frequency within a few kilometres, emissions from mobile should be minimised.

f. Principal characteristics of UMTS and Bluetooth

Only UMTS classes 3 and 4 were specified in 1999: class 4 (21 dBm, corresponding to 125
mW) is reserved for voice and data communication, while class 3 (24 dBm) is reserved for
data terminals.

Bluetooth used close to the ear will be limited to class 3 emitters (1 mW), in the frequency
range 2.4 GHz.

V- What are the measures taken by French operators to reduce the exposure of the
public to RF EMF from telephones and base stations to "as low as reasonably achievable"

g. Matching emissions to requirements, and adherence to legislation

Capacity is a very important element in mobile telephone networks. It is important to reduce
the interference between cells or different geographical regions in order to maximise
capacity, within the current GSM technology and also in the future for the UMTS. Moreover,
power control is essential to maintain quality. In fact, for the signal-to-noise ratio to be
acceptable, the noise and hence interference must be reduced to the least possible. Mobile
phones and base stations, under network control, operate at the least possible power that
produces an acceptable signal: operators transmit only what is absolutely necessary.
Moreover, they adhere to the regulations in force in each country where they are sited.

h. The ALARA principle

According to WHO, the scientific community considers that mobile phones and their base
stations do not present a risk to health as long as the exposure of people is within
recommendations. In the WHO publication "Electromagnetic fields and public health:
cautionary policies" it is stated that "ALARA is not an appropriate policy for EMF in the
absence of any expectation of risk at low exposure levels and given the ubiquity of
exposure". Consequently the ALARA principle would not appear to be an appropriate
means of regulating electromagnetic fields.

2 Discussions with the group of experts

Q: Base stations emit in a directional manner (approximately 120º horizontal width).
Is it possible to have the same arrangement for the antennas of mobile phones, so that the
head is less exposed?
A: First, with respect to the level of exposure to the tissue near the antenna or the radio
telephone, we strongly adhere to the regulations; the great majority or even the totality of
mobile telephones emit below the recommended levels. Under these conditions, on the
basis of current knowledge, there are no contraindications for the use of a mobile.
Secondly, if mobile manufacturers designed a directional antenna, this could affect the
quality of a call: the user would have difficulty receiving the signal emitted from the base
station. In fact, if we direct emission, we direct reception equally. It would then be necessary
for the user to be aligned with the antenna; therefore, in use, one would need to constantly
reorient the telephone in order to maintain optimal quality.

Q: Do base stations operate at or near maximum levels or do they often reduce the power
A: A GSM base station has several signal sources:

· a control channel (broadcast channel, BCCH) always emitting at maximum power;

· other channels whose power varies depending on the number of users at any given time
Power-control is used on these channels in order to limit interference. Emissions -
including maximal power - are therefore limited in order that frequency can be reused as
soon as possible in other cells.

Q: In an urban site (saturated or nearly), do all the emitters operate at full power?
A: No. The site density is necessary to deal with traffic peaks. A given station does not
always emit at its maximum. The intensity of emission varies as a function of time with
periods of saturation and of scarcity of calls. The power input to the antenna depends on
the configuration, whether the cell is urban or rural and on the cell size. The fields emitted
by the antenna depend on the antenna's characteristics and its gain (concentration of the

energy into a beam). The fields around a base station antenna decrease rapidly: for
antennas sited on rooftops, beyond a few meters the field strength is below the reference

Q: In fact, a saturated base station means that all the channels are utilised and not that they
emit at full power. On average, what are the emissions of a saturated base station with
respect to its maximal power (40%, 60%, ..)?
A: This depends on the power-control implemented by the operator. There are no studies
on this subject at the moment. I know the effect of power-control on exposures from
radiotelephones. An article, accepted for publication in IEEE Transactions on EMC, deals
with the subject. The fields emitted are, on average, less than the maximum field because
of power-control and that is before taking into consideration DTX (that minimises emissions
when no-one is speaking). These are statistical studies.

Q: What is the average power of a base station in an urban area?

A: It is difficult to give a number. For antennas sited on rooftops, I would say 10 to 20 W.
With respect to the exposure and the protection of people from radiation, one has to
consider the power at the base of the antenna, that is the power input to be radiated. In the
Paris region, an emitter has the power of 5 W at the base of the antenna for a TRX.

Q: What is the distance at which the SAR values correspond to the Recommendation ? Is it
a few metres?
A: The distance that I mentioned for base stations corresponds to the reference levels. The
basic restrictions given by ICNIRP and the European Recommendation are defined in W/kg
(in SAR). Instead of SAR, it is possible to use reference levels that ensure that the basic
restrictions are adhered to. While it is possible to use reference levels with respect to base
stations, is it not possible to use them for radiotelephones because the exposure is in the
near field.

Q: What are the possible consequences of operation at high frequencies with UMTS and
Bluetooth? Has there been any epidemiological or experimental work?
A: The answer to this question lies with biologists, not the GSM Association. Actually
international recommendations cover all frequencies. UMTS devices should also give rise
to fields and SAR levels that meet the recommendations. The power emitted by UMTS
devices is of the order of 125 mW. Power control is more effective with UMTS (every 0.66
milliseconds) compared to GSM (a minimum of 60 milliseconds, but often several seconds).

Q: Are GSM operators within the Association involved in new protocols other than UMTS:
TETRA, DECT, local wireless loops, and are they going to deal with the effects of these
A: At present, local wireless loops are not strictly speaking a part of mobile telephony. The
organisations dealing with this technology will address the conformity of these systems with
international standards. At present, when we site a base station antenna we take into
consideration all existing wireless systems.

4.5.2 Jean-Claude CARBALLES

Representing the Mobile Manufacturers Forum (MMF)

I Presentation by the MMF

The MMF is an international association of mobile radio communication equipment
manufacturers founded in 1998 and counting among its members Alcaltel, Ericsson,
Mitsubishi Electric, Motorola, Nokia, Panasonic, Philips, Siemens and Sony. Its aim is to
support research in the field of bioelectromagnetism and to participate in the areas of

standards, regulations and in the dissemination of information with respect to health in
relation to mobile telephony. More precisely, it sets itself the following objectives:

i. Research on the health and safety of mobile telecommunication equipment

to encourage and finance independent research that meets the WHO Electromagnetic Field
Project quality criteria.

j. Harmonisation of national and International standards

to adopt a coherent global approach aiming at developing harmonised standards for
conformity tests in order to ensure that the exposure standards are based on the best
scientific data available.

to commission quality research in support of standardisation in order to compare different

measurement techniques.

k. Regulatory policy
to co-ordinate at a global level the development and presentation of the views of the mobile
telephone industry.

to communicate these views to the authorities, and in particular regulatory authorities.

to respond to the request for information or help, from national or international organisations
on the question of exposure to electromagnetic fields from mobile communications

l. Communication with the public

to give to the public precise information on the subject of exposure to electromagnetic fields,
and a clear analysis of the safety of mobile communications equipment and the health of
their users.

to provide assistance to national employees' organisations (in France, the FIEEC and the
GITEP) by providing them with information.

Some employees' unions within the FIEEC decided to draft a leaflet entitled "Facts on
electromagnetic fields" - which covered radio telephony - for the benefit of its members and
interested official organisations. I submit this booklet to you along with that of the MMF. I
invite you to refer to the MMF booklet to get acquainted with the research financed by the
MMF, and its views on the health issues, aspects that there is no time to present to you now.

II Responses to the written questions of the group of experts

What are the technological innovations currently available in France or elsewhere, that can
lead, in the future, to a change in the level or the conditions of exposure of the public to
EMF-RF from mobile telephony?

m. Extension of the second-generation network

The first technological development concerns the extension of the current network to
accommodate an increase in the traditional mobile telephony service. By the end of
September 2000, 26 million mobile phones were used in France. It is expected that the
number of subscribers will continue to increase. By the end of the year 2000 the penetration
rate is expected to reach 50%.

The SAR level (specific absorption rate) of second-generation mobile phones should not
change significantly. Only a change in the network could affect appreciably this parameter.

The cell size (geographic zone covered by a base station) depends on three parameters:

· the topology of the terrain (hills, buildings, etc.);

· the frequency band used (the higher the frequency the smaller the cell);

· capacity (number of calls that can be accommodated per cell).

Hence, GSM base stations are 0.2 to 0.5 km apart in town and from 2 to 5 km in the
countryside. The capacity or limit in the number of calls of base stations is an important
element in changing the exposure conditions. In fact, the increase in the number of users
will require an increase in the number of cells and hence it will decrease their size. The
smaller the size of a cell, the lower the power level emitted by the base station in order to
avoid interference effects with neighbouring cells. Moreover, in order to reduce the
consumption of the battery, each mobile phone, during the conversation adjusts its power
level to the minimum necessary for the base station to receive a signal (Adaptive Power
Control, APACE). Similarly, the base station regulates itself continuously to emit at the
lowest acceptable level. Hence, the shorter the distance between the phone and base
station the lower the emission levels from the phone (and hence the actual value of SAR).

Consequently, the increase in the density of the network translates itself for the user of
mobile phone into an actual reduction in exposure and not, as one would have thought, an

n. Development of a new broadband system (third generation)

The second technological development concerns the deployment of a broadband system for
voice, data, images and internet connection. This type of service is likely to be in demand
as the number of users of the internet increases exponentially. This development will take
place in two steps:

1. utilisation of the current network (GSM) to operate internet-enabled mobile phones

(Wireless Access Protocol - WARP)

2. the deployment of a third generation network referred to as UMTS (Universal Mobile

Telecommunication System) specifically designed to offer broadband services

Exposure Conditions and Power Emitted

UMTS technology differs from GSM in the choice of modulation protocol of the radio signal:
CDMA (Code Division Multiple Access). The signal will initially be characterised by a quasi-
constant emitted power (in contrast to GSM); the average power will be similar to that of

UMTS technology operates in two modes:

1. UMTS FDD (Frequency Division Duplex) using broadband CDMA;

2. UMTS TDD (Time Division Duplex) which is a combination of a CDMA technology and a
time division multiple access (TDMA).

The UMTS specifications give no indication as yet on the maximum power from base
stations. There are different power classes – as there are for GSM – that will be defined in
March 2001. Base station manufacturers have already announced that the power input to
the antenna of their base stations would be of the order of 20 W. This value is the result of a
compromise between two elements:

1. the capabilities of current wide band power amplifier technology

2. the operators' desire to use existing GSM/DCS radio sites to locate UMTS base stations.

UMTS will initially operate in the frequency range 1920 to 2200 MHz, with a view to a new
allocation in the band 2.5 GHz in a few years time. The transition from the second to the
third generation will take place gradually.

Exposure Conditions and Changes in the Use of Mobile Phones: Contribution of

Bluetooth Technology
The change of usage of mobile phones is an important element likely to affect exposure
conditions. The development of new multimedia-type applications will affect the design of
terminals, leading in particular to an increase in screen size and the use of an ear piece to
enable simultaneous voice communication and visualisation of data and images on-screen.

Bluetooth technology will eventually replace wire and infrared connections between devices
such as between the mobile phone and the ear piece, and also between the phone and
computers, cameras etc. It will operate in the frequency range 2.4 GHz, at very low power
levels (probably less than 2.5 mW for mobile phones); its data output may reach 1 Bit/s
within a 10 metre range.

All these technologies will be deployed to meet existing standards based on the conclusions
of scientific studies. This will enable the development of applications within well-established
safety standards.

In conclusion, changes in exposure conditions may well take place in the future due to
technological change and/or changes in usage, and not for health reasons (which do not
justify any changes). In fact, after the detailed analysis of scientific studies carried out over
many years, international experts have concluded that, based on current knowledge, mobile
telecommunications has no negative effect on public health.

VII- What measures are in force, or planned for the future by French mobile phone
equipment manufacturers, that will lead to a reduction in the exposure of the public to RF
EMF to "as low as reasonably achievable" (ALARA) for mobile phones and base stations?

In line with the position taken by the World Health Organisation, the MMF considers that
neither the Precautionary Principle nor the ALARA principle apply to electromagnetic fields
(EMF). Neither should, therefore, become a basis for the formulation of public health policy
on exposure to electromagnetic fields. Standardisation bodies, government agencies and
health authorities periodically evaluate the data from scientific studies on health effects. The
conclusions of their most recent evaluation are summarised in the document submitted
during the interview31. In general, international experts conclude that on the basis of
established scientific fact, there is no evidence of negative health effects. In agreement with
the conclusions of the experts, WHO states that "ALARA does not apply to the formulation of
public health policy in relation to exposure to electromagnetic fields. In fact, the policy does
not apply to electromagnetic fields (whether generated by transmission lines or at

The document that Mr. Carballes presented during the interview, refers to the ICNIRP report, that
of the European Commission (European Commission Expert Group, Possible Health Effects Related
to the Use of Radiotelephones. Proposals for a Research Programme by a European Commission
Expert Group [1996]), the report of the Royal Society of Canada (A Review of the Potential Health
Risks of Radio Frequency Fields from Wireless Telecommunication Devices [March 1999]), the
Stewart report (Independent Expert Group on Mobile Phones and Health, Chairman Sir William
Stewart [May 2000]) and the report of the WHO (Repacholi MH. Low-level exposure to radio
frequency electromagnetic fields: Health effects and research needs. Bioelectromagnetics 19:1-19
[1998]). These documents are also presented in this report.

radiofrequencies) in view of the absence of risk at low exposure levels and in view of the
widespread nature of exposure"32.

III Discussion with the members of the group of experts

Q: What is the effect on output power of high data transmission rates ?
A: There is no relationship between output power and high data transmission rates. Power
emitted by GSM and UMTS systems will be of a similar order of magnitude. The channel
bandwidth will be larger in with UMTS than GSM/DCS, which mean that UMTS can transmit
higher rates of data and images, including animations. As a result, the use of phones will
change and the devices will be adapted to this new usage. However, even with an
increased screen size, it will not necessarily follow that the internet will be accessed on a
mobile phone in the same way as on a computer.

Q: Do you think that audio from the mobile will be delivered by Bluetooth rather than by
direct contact with the ear?
A: Yes, in Bluetooth-enabled devices, all information, including audio, will go via Bluetooth.
Bluetooth technology will enable the easy use of an earpiece, avoiding existing problems
with this type of accessory.

Q: Is the constant power transmission of UMTS a technical necessity?

A: The main problems are the link budget and traffic density. Actually, the higher the
frequency the narrower the area covered. For very close frequencies, there will be no
difference in the link budget. When traffic is not dense, there are two possible options:
setting up base stations further away from each other emitting higher power levels, or setting
up a closer network of base stations with lower power levels. We have here both economic
and technical considerations: the number of calls to a base station being finite, when there
is a need to increase the traffic density of the network, it will be necessary to create
additional microcells (as happens in towns). This means lower power levels per base
station, and per terminal too. Generally, when the base station is small (small cell radius),
the mobile phone user is close to the base station. The power of the base station is
therefore weak for technical reasons: to avoid interference with other base stations.

Q: The public is concerned by the multiplicity of sources: mobile telephony, TETRA, DECT,
BLR… Are the manufacturing members of MMF concerned by the biological effects of
mobile telephony (GSM then UMTS) of other sources?
A: The majority of other sources, in particular DECT, emit very low power. Licenses for the
installation of local wireless loops were given recently by ART to new operators in order to
facilitate direct access. The frequencies used are in the range 3 to 40 GHz. Two
frequencies bands are used in France: 3.5 GHz and 26 GHz. The levels emitted are lower
than from conventional base stations. However, I do not have any precise information about
biological effects at 26 GHz.

Q: Is it possible to indicate the SAR values on the display of the user's phone, and to
establish technical specifications such that SARs are verifiable and consistent for all
A: SAR is a very complex technical parameter; it is difficult albeit necessary to explain it
simply to users. To simply indicate a value on the screen is not enough. Moreover, SAR is
directly proportional to emitted power. As previously observed, power is not constant with
time: it varies when the user moves, and even if the user is not moving it changes with time
during a call. There is also a "hand over" protocol which ensures that the user is connected
to a good cell. This is characterised by an increase to maximum power and a decrease to
minimum power. The exposure level varies during this time.


Technical remark: The greatest technical difficulty is that the SAR depends on the position of the
telephone with respect to the ear, and the way in which it is held in the hand. Therefore the SAR is
not necessarily the same as the value indicated. It is therefore possible to indicate the SAR,
calculated from the power necessary to reach the base station, only if it is possible to standardise the
way in which the telephone is held during the call. The coupling between the hand and the telephone
modifies the power emitted from the phone by anything up to a factor 10.

A: The standard procedure for measuring SAR uses a phantom with a specified shape, a
test liquid whose properties depend on the operating frequency, and predefined positions of
the phone with respect to the head with the device operating maximum power.

Q: A possible solution would be to indicate on the screen the SAR corresponding to that
measured for example, three seconds previously, not the instantaneous SAR.
A: This sounds very complicated and the results will not be very meaningful. Actually, the
standards specify threshold values that take into consideration safety factors. As long as the
values of SAR are below the threshold value, there is no difference in terms of safety for
users exposed to SAR of, for example, 0.1 or 2 W/Kg. The industries that I represent are
not against the publication of SAR values on condition that they are indicated clearly and
understood by the public, and only after harmonisation between the American and European
standards which would enable the provision of viable and consistent data. As for indicating
instantaneous values of SAR permanently on the mobile, this sounds to me as very difficult.

Technical remark: Notwithstanding the above, on-screen indication of SAR is technically possible but
the values indicated will be no more precise than, for example, the number of bars indicating the level
of reception on the screen. People often hold their fingers close to the antenna whilst using the
phone, and this causes SAR variations. What really matters is the SAR measured in the laboratory
and the average power emitted by the mobile phone during a call.

Q: The indication of SAR on the screen of the mobile phone is a useful piece of information
for the users because it would enable them to change the way in which they make a call in
practice if they know that their level of exposure depends on the condition of use.
A: The user is already given different types of information and recommendations for
example not to hold the mobile phone by the antenna. To measure SAR under all possible
conditions sounds technically impossible to me. What may be possible is to give some form
of dosimetric information by indicating the average power emitted, in the knowledge that the
SAR is related to this value.

Q: For some telephones the maximum SAR is 1.5 W/Kg, for others it is 0.2 W/Kg. It
appears that the lowest SARs are obtained for telephones that have higher technical
specifications, price and performance. Is it possible to standardise all telephones to a
comparable emission level ?
A: From the technical perspective, it is possible to conceive a “quality factor” that takes into
consideration both the efficiency of emission from mobile phone and the maximum SAR. A
very low efficiency phone (which, technically, is not good quality) would probably produce
very low values of maximum SAR (the values specified in the standard). It is therefore not
quite right to say that very low maximum SAR values are systematically obtained from the
best performing telephones. It is right to say that an important parameter is the efficiency of
the phone and this is somewhat related to the maximum SAR. Personally I think there are
three important elements:

1. the technology must aim to minimise SAR (the design of the telephone is an important
parameter in this respect, as the antenna can be made to orient away from the head: a
favourable design element);

2. the power (efficiency of the phone). There is a compromise between the

efficiency/maximum SAR value that must be optimised;

3. the technical specifications of the network (currently the network operates on the basis of
33 dBm , if it where to function at 30 dBm, the SAR would be halved.)

4.5.3 Yvette SEGALA

representing the European association for defence against the effects of electromagnetic
fields (Association europeenne de defense contre les effets des champ electromagnetique

The group of experts had been informed that Mrs Segala will be accompanied by Mr Le Ruz,
Adviser to the ADECEM, who was not able to attend.

I. Response to written questions

1. What are your main sources of information on health risks associated with RF EMF?
Do you trust recent international reports? How do you think the public can be better

a. Main sources of information

The Association has access to several sources of information on possible risks to health
related to RF EMF:

· Learned societies (EBEA, BEMS, SFRP, OTA, INRS, ABPE, ANTIGAUSS,

· the commissions of the European Council and the European Parliament
· WHO and ILO (International Labour Organisation);
· the committee of independent experts of Strasbourg;
· scientific reviews such as "Science and the Future";
· colloquia held at the National Assembly and abroad;
· the parliamentary group for the Study of Health and the Environment;
· the medical vigilance sub-directorate (Sou-direction de veille sanitaire) of the Ministry of

b International reports and improvement in public information

The Association has little confidence in the authors of recent international reports because of
their close links with lobbies and the inconsistencies between their reports and information
from other sources. In order to improve public information, the Association favours the
participation of competent experts, independent of lobbies, in studies on the health risks
related to RF EMF.

2. What effects have been brought to your attention by people reporting symptoms from
exposure to electromagnetic fields, and more specifically mobile phones?

We have distributed a petition on electromagnetic fields and can report the following:

· thermal effects;
· specific effects on the nervous and endocrine systems;
· an increase in the number of people hypersensitive to electromagnetic fields in Europe.

However, this petition was not specific to mobile phones and the Association has no answer
to this question yet despite the fact that it received many complaints from mobile phone

3 What action do you recommend on the use of mobile phones and the siting of their base
stations? What are the reasons behind these demands?

We recommend three types of measures:

· The implementation of the recommendations of the circular of the Directorate General of

Urban Housing and Construction (DGUHC/QC/9 N°99-31) and those in the
correspondence of the Directorate General for Health (DGS/VS3 N°187)33.

· The implementation of the limits of exposure to electromagnetic fields for the general
public proposed to the European commission and the Parliament, which is 1 V/m, at 400
kHz to 300 GHz;

· Taking into consideration the residents equipped with active and passive implants as
well as the residents who are hypersensitive to electricity.

These demands are made because of the very large number of court cases and by the
revelations in the media.

II Discussion with the members of the group of experts

Q: Have you had evidence from people suffering from hypersensitivity to electricity?
A: Absolutely. Some people are very sensitive to electricity manifested in relation to their
proximity to high voltage lines or work with a computer. This sensitivity leads to headaches,
problems with vision and hearing, dizziness. We have also started receiving some
complaints from people living close to base stations and from people suffering from
problems related to mobile phones and base stations.

Q: There are actually thousands of base stations in our environment, so that all the
population is exposed. Are you indicating to us that some people as more exposed than
others, or that some people can suffer more than others ?
A: This probably depends on the exposure, or the distance of the exposure, or the degree of
sensitivity of the persons.

Q: How often do you receive complaints and how serious are they? How many people have
complained about power lines and computers?
A: The number of complaints varies from week to week, I do not have the number of people
presenting common symptoms.

Q: How many members does your Association have?

A: In fact, many people contact the Association without asking to be members.

Q: On what basic principles do you base your replies and advise to the complainants (how
do you attribute a cause to a symptom)?

Technical note: The correspondence of the DGS (Directorate General of health) was brought to the
attention of DGUHC (Directorate General for Urban Housing and Construction) on the siting of base
station antennas on the balconies of some residential buildings in places that are directly accessible
to the public, in particular those fitted with pacemakers, this at a time when the European Commission
was preparing a recommendation limiting exposure of the general public. The DGUHC circular has
adopted the terms of this correspondence to communicate to HLM Organizations.

A: First we rely on the opinion of medical doctors and scientists. Secondly, the
complainants themselves try to understand the origin of their symptoms by reflecting on their
various activities (contact with electrical equipment, daily work on the computer etc.) or by
discovering if there is a base station antenna close to their homes.

Technical remark: R. de Seze has indicated that he receives similar queries through a professional
network of consultant pathologists at the INRS and that he redirects the complainants to the ADECEM
to determine a cause for their symptoms.

Q: A powerful 6 MW antenna is sited on top of the Eiffel Tower: this is the equivalent to 10
times all the base stations in France. Have you received complaints from residents in that
area that may be related to this antenna? Furthermore, there are about 40 FM radio
transmitters in Paris. The electromagnetic field strength due to the FM transmitters is
highest in Paris; it was there before television and GSM who contribute approximately
similar levels. Have you had more complaints in the vicinity of these FM stations?
A: We have some cases, but our Association is not the only one receiving complaints.

Q: In view of the lack of credible information, what type of organisation should, according to
you, provide trustworthy information (district, local council, DDASS)?
A: The District and Local Councils should ensure that base station antennas are sited in
locations that do not cause problems for residents.

Q: Base stations are, however, designed to be installed in the proximity of their users.
A: Of course, but they should not be sited near schools for example.

Q: However, according to the Stewart report published in the UK in May 2000, the siting of a
base station on the rooftop of a school reduces the level of exposure in the school.
A: I am not a scientist, I am not sufficiently qualified to answer this type of question. I am
sorry that Mr Le Ruz could not join me today, he would have given you a learned opinion on
this matter.

Q: How do you identify the independent expert that you need within the Association?
A: We contact the ABPE, for example. We also make inquiries at the European
Commission and we are in contact with foreign associations, in particular the Swedish. All
the associations have scientific councils.

Q: While not everyone has technical qualifications; it is necessary to provide the general
population with credible, well-founded information. This is why we would like to know what
type of organisation, or what course of action, you think would allow the provision of such
credible information?
A: There are laws and regulations on health matters.

Q: Do you therefore consider that the publication by the Ministry of Health of regulations to
form a solid and sufficient base ? According to you would the public trust such information?
A: I would like experts, independent of the Ministry, to provide an opinion in order to
exchange ideas.

Q: Results that are reported may be interpreted differently by different people. It is

therefore important that they are understood in the same manner by all people in order that
there is trust in measurements. What should be done to ensure that your members trust a
A: Our members do not ask all these questions. They trust the reported measurements.

4.5.4 Elisabeth CARDIS
International Agency for Research on Cancer (IARC) (World Health Organisation)

I Responses to written questions

IX- What do you think of the epidemiological studies that have reported so far?
The studies in the literature relate to a variety of radiofrequency sources: mobile phones,
occupational exposure, residential exposure. Their results are not consistent. Cohort
studies of occupational exposure present the widest scope. Their results are not however
consistent, for example some studies reveal an increase in the cases of leukaemia and brain
cancer, others do not.

o Study on occupational exposure and brain cancer

Three studies report an increase in the risk of brain cancer:

The study of Szmigielski (1995) in Poland shows a systematic increase in the risk of cancer
but it is very difficult to interpret because its methodology is not clear.

The Grayson study (1996) is a case-control study within a cohort in the United States; the
participants responded to a questionnaire. There were 230 cases and 920 controls from
within the US Airforce. The study shows a relative risk of brain cancer of 1.4 for those
exposed to radiofrequencies, but no dose-response relationship with respect to exposure.

The Thomas study (1987) is also a case-control study: it has 435 cases and 386 controls, it
used job titles to classify exposure. The relative risk of brain cancer was 1.6. A more
detailed analysis shows that this increased risk is only present in workers who, in addition to
exposure to microwaves, work in the electric and electronics industry and that the risk of
brain cancer is higher for workers in this industry even if they have not been exposed to

On the basis of case-control and cohort studies it is difficult to conclude that there is an
increased risk of brain tumour or leukaemia among workers. However, most of these
studies suffer from poor methodology and in particular none of the studies has really
characterised the individual level of exposure.

Studies on environmental exposure and risk of leukaemia and brain cancer

These studies deal with, for example, cancer risk in the neighbourhood of radio and
television transmitters etc. The majority of these studies do not show any increased risk,
and the levels of exposure were not very high.

Epidemiological studies on mobile phone users and risk of cancer

These studies are fairly recent, given that mobile phones themselves have not been widely
used for very long. It is however necessary to allow sufficient time for tumours induced by
radiofrequencies from mobile phones to developed and be diagnosed.

The study of Hardell et al. (1999) on brain cancer risk is based on 233 brain tumours. The
results show two elements:

· no risk of cancer in mobile phone users (relative risk of 0.98) compared to controls;
· a non significant increased risk of 2.5 on the basis of 13 tumour cases, three benign,
localised in the brain lobe near the antenna on the side of the head that the user places
the phone

However, it is difficult to draw firm conclusions on the basis of this study because of
methodological problems: it is based on a small number of cases and only on those cases

alive six month after that diagnoses, such that a very large number of eligible tumours were
not included in the study.

The study of Muscat et al (not yet published, but reported in the recent document from the
Academy of Science)34 on risk of brain cancer is based on 450 cases of brain tumours in the
United States. It reached two conclusions:

· no increase of risk in mobile phone users;

· a slight increase in one type of cancer.

In contrast to the conclusions of the Hardell study, this study does not show any increase in
the tumour on the side of the head near the antenna.

Nevertheless, this study has limitations of its own. It has too few cases, and relatively fewer
people use mobile phones in the United States than in Europe. Finally, the study was
carried out too early considering that the cases were recruited between 1994 and 1998, a
period were the number of mobile phone users all over the world was relatively small. The
fact that this study does not show an increase does not therefore mean that none exists.

The study of Dreyer et al (1999) on cancer risk had a cohort of 285,000 users in the United
States from among analogue phone users and car phone users. Two cases of death by
brain tumour and 4 cases of death from leukaemia were revealed, which corresponds to a
very small standardised mortality rate. It is difficult however to come to a conclusion
because the follow-up period is too short (only one year) and the number of cases too small.

The study of Morgan et al (2000) on Motorola workers deals with exposures from all
communication equipment, not only handheld devices. It has a cohort of 196,000 Motorola
employees followed up from 1976 to 1996. It defines three levels of exposure related to
different types of job description of the people involved, but it does not include any individual
assessment of exposure. Overall, it shows a decrease in the number of deaths compared to
the population as a whole. The same result is found for all cancers as well as for leukaemia
and brain tumours. No dose-response relationship was established.

A more detailed analysis showed that among those 196,000 subjects, few were exposed
(17,000 people exposed to high and moderate levels), with only 17 brain tumours (of which
three were among the highly exposed group) and 21 leukaemia cases (of which 5 were
among the highly exposed group).

The only conclusion is that it is too soon to reach a firm conclusion of any kind for RF
exposures with telecommunication equipment.

This study was since published in JAMA in December 2000 and it is reviewed among the recent
articles (see Chapter IV-E).

II. Discussion with the members of the group of experts:

Q: Is there a scientific basis for the reclassification of benign and malignant tumours in the
case control study of Hardell? Moreover, it is true that the follow-up period is too short to
show effects. However, according to Muscat, the hypothesis for carcinogenesis veers more
toward the role of promotion rather than initiation; under such conditions the time period
leading up to the appearance of cancer is shorter than if it was an initiating effect. Is this a
relevant parameter?
A: The two questions are related. It is thought that if RF has an effect, this effect will be
seen in the promotion or progression of the tumour, that is an increase in cell division etc.
which can occur with benign as well as malignant tumours. This is why it is important to
look into the two types of tumour. I think however that in the IARC study we will derive
conclusions for the two types of tumours separately. The study of Hardell has few cases, so
it is necessary to combine the two types of tumours.

Q: Reporting on the study of Hardell, Hansson-Mild points out that out of 13 cases of
tumour, 9 were exposed to NMT, 3 to NMT and GSM and 1 exposed to GSM only. Do you
see any grounds for discriminating between the effects of the different type of phones?
A: NMT phones are analogue telephones that monopolised the market up to 1992 when
digital telephony was developed. Consequently, the result reflects the follow-up period
between exposure and the appearance of tumours. In the United States, unlike in France,
the number of people in the whole population who have used NMT is higher than those who
have used GSM.

Q: What can we say about the hypersensitivity of certain populations?

A: According to Hansson-Mild some symptoms appeared to be related to the extent of use.
These are interesting results, but from only one study. I am often contacted by people who
complaint of headache, nausea, heating behind the ear, but this is not my field and it is
difficult to interpret these observations.

Q: Are the scientific data relating to the risk of cancer associated with RF EMF the
justification for the important IARC study? What is the time frame?
A: On the basis of the studies carried out on occupational exposure we cannot conclude
that there is no effect. However, if such effects exist, they would be very weak indeed at the
level of exposure of mobile phone users. Nevertheless, even a very small risk at the
individual level could mean a very large number of cancers all over the world in view of the
exponential growth in the use of mobile phones.

Q: In your criteria for the evaluation of the carcinogenic effects of RF, have you taken into
consideration experimental work that supports your hypothesis that effects could occur ?
What are the studies that support your hypothesis?
A: It is equally difficult to draw conclusions on the basis of experimental studies. The only
risk is probably the risk of promotion or progression. If this is the case, the latency period
could be short, of the order of five years, so we should have sufficient follow-up for the cases
to develop within the time frame of the epidemiological studies that are just starting.

Q: IARC has initiated an important international study; the results should be available in
three or four years.
A: We have not yet secured funding for the whole project. We have partial funding from the
European Commission and national bodies, and we are still looking for additional funds from
industry. The latter entered into discussion with the International Union Against Cancer
(Union Internationale Contre le Cancer (UICC)) about a year ago. The UICC could deal with
appropriate funds from industry but we had not yet received a contract that is acceptable to
us in as much as it guaranties the independence of our work. If we manage to secure the

funds required within the next few months, we should be able to produce the first results
from the national study in 2003 then the final international results in 2004.

Q: Our group of experts will be making recommendations on the means of providing funds
for research in electromagnetics; why have you met with difficulties in establishing clear
basis for securing funds from industry?
A: The contracts proposed by industry contain a number of conditions, such as the
possibility of stopping funding the study if the work is not carried out correctly, the
nomination of members of the scientific committee in charge of monitoring the progress, the
review of papers 60 days prior to their publication etc. However, according to our
international protocol the results must remain confidential until their publication. Only one
copy of the paper can be sent to the funding agency under confidential cover – with the
agreement of the journal –not more than one week before publication.

Q: The IARC evaluation of RF and cancer will be based on all the available data. How will
that fit with the schedule for the publication of results from this study?
A: The date of the evaluation of the individual studies is 2003. However, in one or two
years we will review the state of current knowledge and decide if it is too soon to carry out
such an evaluation. The Interphone study is independent of the individual studies. The
latter are the responsibility of another unit within IARC.

Q: Are there any other ongoing studies on the same subject?

A: There is a case-control study at the National Cancer Institute in the United States that
considers 800 cases and 800 controls diagnosed between 1994 and 1998, at a time when
the percentage of users of mobile phones was relatively low and only analogue phones were
used35. Two cohort studies of mobile phone users have also been conducted in Finland and
in Denmark.

We wanted to include leukaemia in the Interphone study but we did not have the funds to do
so. Independent studies on leukaemia may be therefore carried out elsewhere. Now that
ionising radiation and even extremely low frequencies are considered factors in the
development of leukaemia, some researchers thought that, by analogy that radiofrequencies
could also be responsible for leukaemia because the bone marrow of the skull is exposed to
RF while using mobile phones.

Q: How did you deal with combining data from questionnaires and the results of actual
A: In a feasibility study, we tried to identify factors that determine the level of exposure
(number of calls, duration of calls, use of power control etc). We have started a process of
quantification and we will be developing some sort of index of exposure.

Q: Are there any ongoing studies dealing with pathologies other than cancer in relation to
mobile phones (headaches for example)?
A: This is a field with which I am not familiar, I don’t know.

Q: Pending the results of this study, does the state of knowledge justify the adoption of
measures to reduce electromagnetic field exposures to below the levels recommended by
ICNIRP and the EU?

The exposure levels from mobile phones and base stations are actually well below the
recommendations in the standards. Most of the time, mobile phones emit less than 5% of
their maximum power: the emissions are therefore below the levels in the standard. While

This study by Inskip et al has also just been published in the New England Journal of Medicine and
is reported elsewhere (see recent articles)

there is some uncertainty relating to possible effects, one should adopt some form of
precautionary approach but it is difficult to know on what to base it. For example, it has
been recommended to limit the use of mobile phone by more vulnerable populations such as
children. This precautionary approach is based on analogy with the risk to children from
other agents, such as chemicals, ionising radiation etc.

Q: Industry and its representatives rely on the advice of WHO with respect to the
Precautionary Principle; they interpret it to mean: "provided that we respect
recommendations based on established science, it is not necessary for us to modify our
position". Has the WHO said that only well-established scientific data should form the basis
for regulations and that there is no reason to adopt a Precautionary Principle?

A: I would be surprised if WHO had made such a categorical statement.

4.5.5 Jean-Pierre CHEVILLOT

I Response to written questions

1 Do the current scientific data relating to the possible health effects of RF EMF justify
a revision of the European Recommendation of the limitation of EMF exposure of the public?
The response to this question is based on a report published in English in May 2000 entitled
"Physiological and environmental effects of electromagnetic radiation", by Messrs. Jean-
Pierre Chevillot, Jean-Pierre Husson and Philippe de Montgolfier from the Essor-Europe

p Methodology of the report

Initially written for politicians, this report has deliberately followed a didactic approach. In
order to acquaint non-specialists with basic principles of the subject matter and in order to
be able to formulate an objective opinion on the questions asked, it considered a very large
literature base encompassing diverse scientific and technical fields.

One part considers biomedical phenomena related to the interaction of people with the
electromagnetic environment. A comparative analysis and a summary of results as a
function of frequency are presented.

Another part of the report gives opinions that were gathered on the validity and scientific
soundness of the data that were considered. Particular attention was given to the sources of
information and to the criteria used in the presentation of the results. A distinction was made
between the results that were validated and those still under discussion or where
replications had not yet been completed. Results that had not been discussed by the
international scientific community were discarded. The report has two main parts: the first
gives the main subjects considered and the main conclusions and recommendations for the
benefit of Members of the European Parliament. The second part presents the arguments
in support of the first part. The last part of the report is a summary dealing with the basic
principles of electromagnetism and biology.

q Main options and recommendations

Main Scientific Options:

· complement the epidemiological studies – that do not show either the absence or the
presence of a risk – by clinical studies on people reporting symptoms in order to have a
better base for the interpretation of results;

· pay particular attention to vulnerable populations, particularly children, the elderly and
people undergoing medical treatment;

· look for possible physiological or psychological reasons for the hypersensitivity of certain
people to electromagnetic fields;

· address the lack of basic knowledge on interaction mechanisms for people in the
electromagnetic environment;

· continue to support pan-European interdisciplinary research and the co-ordination of the

research activities of different groups working in Europe on biomedical effects from
electromagnetic fields, an example of which is the COST 244 action that has carried out
such a task in a very satisfactory manner.

Main Political Options And Recommendations

· The results of research does not provide evidence of a probable risk that would justify a
change in the values of the ICNIRP recommendations which underlie the July 1999
European Union Recommendation and also the regulations in most European countries.
Indeed, it would be useful to change this Recommendation into a Directive in order to
remove any regulatory disparities between European countries. At the social level,
these disparities create concern and doubt within the population and at the political level,
they create difficulties for the manufacturers and operators of equipment.

· The Precautionary Principle as presented, notably by the European Commission, would

probably not provide any answers with respect to unproven risk. It would therefore be
useful to gain knowledge of the electromagnetic environment to be in a position to react
in an informed way: the exact term used in the report is "educated and responsible
awareness". This would enable the formulation of a policy of reasonable vigilance
against possible risks. In this context, it is important to increase the provision of public
information and education in order to allow the public to form objective opinions on the

· It should be possible to carry out a technical evaluation of the domestic or occupational

electromagnetic environment when people require it. Moreover, manufacturers should
provide some electromagnetic field exposure information for the use of domestic and
occupational devices. Finally, areas exposed to electromagnetic fields from large
installations must be surveyed and the data made available to the public.

· Considering the increased technical content of the questions related to electromagnetic

fields at the socio-economic level, it would be interesting to proceed with a Delphi-type
inquiry at the European level in order to create the right conditions for a constructive
dialogue between the general public, industry, scientists, related public authorities and
the media on the subject matter.

XII- What are the regulations adopted by the Member States of the European Union and
their neighbours with respect of the exposure of the public to RF EMF?

Notwithstanding the Recommendation adopted by the European Council in July 1999, the
exposure standards relating to electromagnetic field are not harmonised between different
European countries. The report details the special cases that apply in Italy and Switzerland.
The countries can be divided into six categories:

· Countries following the ICNIRP standard and the Council Recommendation. This
category includes Austria, Germany, France, Ireland, Norway, the Netherlands, Portugal
and Sweden.
· Countries following ICNIRP but which are in the process of developing more stringent
new regulations. The Walloon region in Belgium is preparing a standard on the basis of
3 V/m.

· Countries not following ICNIRP but having more stringent regulations. Italy with 6 V/m
(but 20 V/m in places where people do not stay longer than 4 hours per day). In
Switzerland, limits of 4 V/m for base stations at 900 MHz, 6 V/m at 1800 MHz and above;
5 V/m for base stations emitting in the two bands, are expected.

· Countries not following ICNIRP but having less stringent regulations. This is the case of
the United Kingdom with an SAR of 0.4 W/kg, which correspond to a power density of
26-33 W/m 2 for 800 – 1000 MHz and 100 W/m2 for 1800 – 1900 MHz.

· Countries which are not following ICNIRP and do not have regulations. This is the case
of Greece.

· Countries who have not yet taken any position. This category includes Denmark, Spain,
Finland and Luxembourg.

II Discussion with the members of the group of experts

Q: The European Union text on the Precautionary Principle, of February 2000, draws on
the necessity of having scientific data and on the principle of proportionality and consistency
(similar measures must be taken in comparable situations in other fields). It is noted that in
the case of mad cow disease, the Precautionary Principle was adopted in a very stringent
manner in 1996, with enormous consequences for farmers, and that was before the
availability of the recent scientific evidence. In the case of electromagnetic fields, we require
well defined scientific facts in order to change a regulation that is today based on thermal
effects only. Is there any consistency between the two approaches?
A: In the case of electromagnetic fields it does not seem appropriate to base standards on
hypotheses alone. There is no well-defined risk on which to base the approaches of the
Precautionary Principle. This is a basic difference between the case of BSE and that of
exposure to electromagnetic fields from mobile telephony. Invoking the Precautionary
Principle under these conditions appears to me to be inappropriate and the basis for it
insufficient. It is in my opinion more important to inform the public and to allow the public to
take conscious responsibility for its actions, this is significantly better than arousing
suspicion with false ideas on the subject of risk and creating a passive attitude of hiding
behind a Precautionary Principle with no defined objective. In the case of mobile telephony
for example, it would be useful if the general population had an attitude of "educated and
responsible awareness", such that they become aware of the way in which they use the
technology, particularly with respect to children who can use or misuse it.

Q: How can we better understand the advice of people in authority on the biological and
health effects of mobile phones on children?
A: To the question “do you think that the children are more vulnerable than adults to
possible risk?”, the responses of the experts consulted were as follows:

a) People within the industry do not consider themselves qualified to reply to such a
question and do not consider that the question should be asked while there is no
evidence of any effect. One however has said that it is not possible to say no.
Another gave an absolute positive response to the question. Another opinion was
that the response to the question depends on the hypothesis set out at the onset,
either we think that the effects are too weak to observe, or we think that there are

compensation mechanisms coming into action. In the latter case, the response is
certainly positive.

b) With two exceptions the scientists were in favour of the potential vulnerability of
children. One of the positive responses indicated that other people might also be
vulnerable if the effects reported on protein expression were true.

c) One scientist was of the opinion that current knowledge of the specific way in
which children interact with electromagnetic fields is largely inadequate

d) A representative of an international institution considers that the case of children

must be taken into consideration and points to specific facts such as the threshold of
their perception of contact current at 50 Hz, which is lower in children than in adults.

Taking into consideration the various arguments, often contradictory, reported in the
literature, the report concludes that it is necessary to deal very specifically with the case of
children. It was not possible to make very specific recommendations except to create the
conditions in which "educated and responsible awareness" becomes a means of active

Q: Does your report recommend studies of children to see if they are particularly sensitive
from a neurophysiological or other perspective? Why should the energy absorbed by the
skull of a child be weaker than that of an adult under similar exposure conditions?
Technical remark: This question relates to coupling: the head-handset impedance is lower, which
means that more radiation is reflected outward than toward the head given the morphology of the
head of a child.

A: Our report underlines the importance of supplementing the insufficient information

available on the interaction of people with EMF. This statement relates particularly to
specific conditions such as interaction with a child. It is difficult to conduct studies on
children for ethical reasons. However, special attention should be paid to epidemiological
studies, as is the case with studies dealing with childhood leukaemia. r Specific Biomedical
research should be undertaken in order to improve our knowledge of the possible
consequences of the exposure of children to EMF. The task in hand is made particularly
difficult by uncertainty about the long-term consequences of exposure to very weak fields.

Q: Have you debated the principle of optimisation, bearing in mind the possibility that the
European Recommendation of 1999 might become a Directive? Have you discarded the
ALARA principle? According to the people who gave evidence this morning, the
technological objectives lead to optimisation and thus to the reduction of doses: they
therefore adopted the ALARA principles without health justification.
A: Our report holds the principle of ERA ("educated and responsible awareness"). It implies
optimisation of the electromagnetic environment. We have gathered different opinions on
the ALARA principle: according to some it would not be applicable to the field of mobile

Q: In view of the lack of basic knowledge that you have referred to in your report, do you
recommend fundamental research on the growing animal?
A: Such studies would be useful. Those carried out to date have not shown a difference in
the results depending on the phase of development, but there have not been many studies
on the subject. In general, an improvement in our knowledge of mechanisms of interaction of
living organisms with electromagnetic fields would provide generally useful information.

Q: We always meet an ethical problem with research on children. Consequently how would
you direct the research (on a growing animal, on growing cells or on the gonads)?

A: The solution of this ethical question must come from the competency and responsibility
of the researchers themselves. One approach would be to start from basic knowledge of the
development of children and the interaction with electromagnetic fields, to develop a better
ability to analyse specific effects a priori, and determine whether such effects are possible or
not and at various stages of development.

Q: This means taking into consideration growth and risk cofactors simultaneously.
A: One issue needs to be clarified; the hypothesis referred to earlier of compensating
mechanisms that ensure a sort of homeostasis in the organism with respect to sensitivity to
the environmental electromagnetic field. This would explain some of the results obtained, in
particular from epidemiological studies, as due to a deficiency in such mechanisms in a
small percentage of the population. Moreover, in dealing with cofactors of risk, it can be
noted that the hypersensitivity of certain people to EMF is not related to EMF so much as to
these people's fear of EMF. It would be desirable to study people who are known to be
hypersensitive because one may find among them symptomatic cases that would make
possible the study of a number of parameters and would increase our knowledge of specific

Technical remark: It is important to distinguish the epidemiological and psychological aspects and
hence to differentiate between people who think they are hypersensitive to electricity but who
objectively are not and people who are sensitive to electricity because the field constitute, in their
case, a co-factor to their symptoms. Some people are also sensitive without knowing it.

4.5.6 Madeleine BASTIDE

Emeritus Professor, Faculty of Pharmacy, Montpellier

I Response to written questions

1. Among the recent experimental studies related to biological effect of RF EMF (heat shock
proteins, mortality of chick embryos, increase in the level of ornithine decarboxylase …),
which, in your opinion, are relevant to the possible risk to health of users of mobile phones?
Heat shock proteins are related to stress; they represent a defence mechanism against an
insult. These defensive molecules are simple. We have also carried out measurements of
stress hormones in mice.

We have followed very specific experimental conditions: for financial reasons we have
worked in call- rather than communication-mode with a telephone fitted with a mobicard, that
is, without subscription. The Faculty provided us with a silent line and a code number
allowing a repeat call every three minutes. We were therefore in an extreme exposure
situation system: our telephone will dial a number with a ringing phone for one minute then
rest for two minutes in total three minutes, then repeats the cycle over again etc.

Prior to the tests with the telephone, we carried out an experiment with computers. We
noticed a change in the ratio CD5/CD20 (immature B lymphocytes/mature B lymphocytes).
We used chick models that we had in the laboratory. Within three weeks of incubation, we
realised a 40% mortality rate in the eggs. We placed the survivors in a cage in front of the
screen of a computer, having covered the screen with a black cloth. We then proceeded to
measure antibodies, corticosterone and melatonin. We realised a very significant decrease
in melatonin and a decrease of corticosterone and specific antibodies.

Pleased with this experiment, we kept the same model for the experiments with mobile
phones and, as an end point the evaluation of the mortality of embryos.

2- How do you extrapolate from chick embryo to man? What risk to man would your results
predict? In your opinion, does your work relate to the problem of low frequency emissions
from mobile phones?
First of all, we have adopted a high intensity toxicological approach by using a call mode,
which is not the usual condition since the telephone is mostly used for communication.
However, before the establishment of communication, the users are also in call-mode albeit
for a shorter duration than we have used in our test.

We have carried out mutagenic test on bacteria (AMES’ test), which proved negative. A
positive mutagenic test on bacteria would suggest potential mutagenesis in man. The
mortality of embryos observed in our experiment originates from changes in the sequence of
development of the embryo. It is important to note that the chick embryo goes through the
same sequence of development as the human embryo, so extrapolation is more
straightforward than for bacteria. The group of Jocelyne Leal in Madrid have worked with
other electromagnetic fields (low frequency pulsed fields) applied over 48 hours. They
detected development anomalies and malformations in the first two days. In our work on
chick embryos, we did not analyse development anomalies for financial reasons, however
because we fixed and kept a series of embryos at the time of death in order to look for
anomalies, this work could still be carried out. We have therefore just carried out a
phenomenological study by counting the dead embryos. I think that their death is due to
developmental anomalies and structural modifications. The toxicological approach that we
have taken is similar to pharmacological studies carried out to predict possible toxicological
effects of drugs. Mobile phones could, in my opinion, be classified as teratogenic because
of the effect on the normal development of an embryo. If, on the basis of our results, the
mobile phone is classified as a teratogenic "drug" – a very special drug, used by a large
portion of the population - the advice should be not to give it to people at risk or to children.

II. Discussion with the group of experts

Q: What do you glean from these results about exposure levels?

A: It is difficult to reply to this question because our studies were carried out under very
specific conditions in call-mode to an external line without communication. It is necessary to
study what would happen in communication-mode for short and long duration calls.
Q: Can we for example say that a mobile phone carried at the belt by pregnant women puts
them at risk of embryonic mortality?
A: Our research did not cover this topic but on the basis of the experimental model it shows
the direction that new research should take. However, I have been in touch with nurses who
always carry their mobile phone at the belt position and who suffered spontaneous
abortions; embryonic brain damage could be seen but this may be a very special case.

Q: Your experimental model consists of a multiplicity of sequences of calls that are not
representative of the normal operations of a telephone. How is it possible to extrapolate
from this situation toward the general problem of the telephone?
A: It is necessary to carry out complementary studies. Our model can be used in a variety
of different configurations. With respect to our operational modalities, we have repeated our
experiment several times and always obtained the same results. We have always carried
out the experiment in parallel with an incubator containing the same eggs as those under
consideration and we have always obtained the same rate of mortality of exposed embryos
compared to controls.

The increase in temperature due to the operation of the telephone cannot constitute an
experimental bias because the experimental container used was thermally-regulated and the
temperature was monitored with 4 probes in the middle and around the tray of eggs and
there were practically no difference in temperature (38 ± 0.5°C).

Mr. Bouillet of Bouygues-Telecom has agreed to help us replicate this work with the
collaboration of Supélee and two different research groups, the Bioelectromagnetic
Laboratory of Jocelyne Leal in Madrid and that of INRA in Tours, both specializing in birds.

Q: You seem to be attributing the effects observed to the low frequency magnetic fields
from the telephones. Could you please explain that in more detail?
A: Because of our experiment in front of the screens of computers we knew that the effect
was not due to high frequencies (microwave effect). We have blocked the transmission of
high frequencies from the telephone using a 350 mm diameter copper mesh: we earthed the
mesh and placed it on top of the eggs without touching them before placing the telephone on
top of the grid. We observed the same mortality rate with and without the grid. We have
carried out similar experiments with mice and observed the same decrease in corticosterone
and ACTH as previously observed without the grid.

Q: Have you undertaken any studies with pregnant mice?

A: Not yet.

Q: Do you conclude from your experiments that there is a possible teratogenic risk?
A: I would conclude that our experimental results are repeatable and demonstrate an
important effect on the viability of chick embryos. We cannot say that the mobile phone is
safe. Computers appear to be more dangerous however, because their users spend several
hours in front of the screen and at proximity of between 50 and 80 cm as in the case of
exposed eggs. If the user spend several hours in front of the computer whilst using a mobile
phone, it synergic effects and enhancement might be possible.

4.5.7 Jean-Marie ARAN

Experimental and clinical audio biology laboratory, Pellegrin Hospital, Bordeaux

I. Response to written questions

1. Why ask questions about possible effects of GSM microwaves on hearing?
It was necessary to do so for several reasons, particularly the increasing number of users
(about 20 millions in France) and the increasing concern about possible undesirable
biological effects. Moreover, the ear is the organ most exposed to microwaves given that
the mobile phone in its current version is applied directly to the ear. In addition, if biological
effects in general are observed in tissue, they are just as likely to occur in the inner ear at
the same level of exposure (this remains to be demonstrated).

The inner ear is a highly differentiated structure, a specific micro-organ relatively isolated
from the rest of the body. It has a very high sensitivity to external acoustic stimulation, with
a threshold of detection of vibration of nanometre amplitudes. Reception of acoustic
vibration occurs through the intermediary of contractile proteins (prestine) that amplify the
vibration by 40 to 50 dB. The sensory epithelium is very heterogeneous and includes an
extracellular compartment for liquids of high potassium content (endolymph, in which the
filaments of the sensory cells bathe). There is a difference in potential of about 160 mV at
the interface between the filaments of the sensory cells and the liquid in which they bathe.
The electrical phenomena associated with mechanosensory transduction are of the order of
a few nanovolts. This is why, other than the biological effects common to all tissues, one
might expect possible perturbation of the electro-biochemical phenomena that are delicately
balanced and specific to the function of the ear

Finally, the ear is submitted simultaneously to microwaves from the mobile phone and to the
sound produced by it. It is possible that some synergy between microwaves and sound
could take place in sensory cells, and it is also possible that this synergy may prove

deleterious even if the two agents separately are harmless. The sensitivity of sensory cells
to microwaves may be different during their functional depolarisation.

2. What do we expect to see from the exposure of the ear to GSM microwaves?
If the energy absorbed at the level of the inner ear is sufficiently high, one would expect an
auditory sensation associated with electromagnetic stimulation, which is apparently not the
case with mobile phones. So, if the mobile phone induces an acoustic wave in the skull, this
would be below the intensity threshold for hearing. It would also be below ambient noise
and therefore without any risk to the ear. Electromagnetic waves could also stimulate
directly - in the same manner as an electrical stimulation - the sensory and nervous
structures of the inner ear that are highly sensitive to acoustic and electrical phenomena,
and result in an auditory sensation. This is evidently not the case.

Other possible effects are not necessarily specific to the ear but common to the effect of
GSM microwaves on biological tissues in general.

Q: Do you mean to say that in the absence of an auditory sensation there are no other
effects given that – according to you – the first manifestation is the development of an
auditory sensation?
A: The energy, not being sufficient to develop an auditory sensation, would also not be
sufficient to produce damage to the ear or to the retina for example. The same argument
can be made for equilibrium (dizziness).

Q: Some people complaint of dizziness after exposure to a mobile phone.

A: Vestibular function is more complex than the auditory function; it relates not just to the
ear but also to vision, to other somesthetic receptors and to central integration processes.
Dizziness is a central phenomenon. The peripheral origin (in the ear) of dizziness is difficult
to determine. There are no peripheral tests entirely specific to the vestibule in man.

Q: Some studies have shown that cortical auditory evoked potentials were modified only
when the people were carrying out an intellectual task. Can this have any consequences for
brain functions?
A: If sensory evoked potentials other than auditory (visual or somesthetic) are modified, we
can perhaps consider that we are dealing with central nervous system functional changes.

If only auditory potentials are modified, we should investigate further. In our experiments, we
do not exclude the possibility that a central effect occurs in addition to peripheral
phenomena. If a modification occurs in central responses via a chain of effects from the
periphery, we are in a position to carry out different types of tests (acoustical otoemission,
electrocochleography, cerebral evoked potential) that will enable us to localise the
disturbance and determine whether it is central or peripheral.

3. Studies on the specific effects of GSM microwaves on the ear

r Hungarian human volunteer study (1999)

This study showed an increase in the latency of the V peak of the evoked brain stem
potential during stimulation by a click in the exposed ear; no effect was seen in the non-
exposed ear. This increased latency corresponds to an increase in the threshold of hearing
of high frequencies of about 15 to 18 dB. This was confirmed by audiometric measurements
carried out immediately after the electrophysiological ones even though the subjects did not
notice any change in their hearing. This is an immediate effect; long-term persistence is not
reported. The authors think that sensory cells bathed in fluid are subjected to a local heating
or to transmembrane ionic movement induced by microwaves. This study is very limited and
deals with only 10 subjects without any statistical evaluation: in our opinion it should be

verified. It would also be interesting to measure the first peak and the interval between the
first and the fifth in order to determine the peripheral or central location affected.

s Carmela Marino's study using rats (2000)

The exposure system, dosimetry and physiological measurements in this study are well-
described. The study deals with otoacoustic emissions (distortions recorded in front of the
eardrum) for testing ciliary cells outside the cochlea. These cells are in fact the most
sensitive (they amplify vibrations inside the cochlea) and they are the most fragile (effect of
noise, ototoxicity). The whole of the rat head was exposed, with a SAR of 0.2 to 1 W/kg,
three hours per day for three days (0.2 W/kg) and five days (1 W/kg). No significant
difference was observed in the amplitude of the distortion products in the otoacoustic
emissions. If there are no changes in the otoacoustic emissions then there are no problems
at the level of external ciliary cells that are responsible for the high sensitivity and frequency
selectivity of the ear.

t Ongoing studies
Within the COMOBIO Project , we have studyed hearing at the peripheral level (ciliary cells
and the auditory nerve) and at the central level (brain stem) of hamsters exposed to different
SAR levels (1.2 and 4 W/kg) for one hour per day, five days a week for two months. We
used groups of eight animals. Permutations of the groups and of the animals within a group
were made in time and location in order to homogenise the exposure conditions. The
detection of otoacoustic emissions gives a measure of the function of ciliary cells outside the
cochlea, are involved in the first stage of the reception of sound. Detection of the action
potential of the auditory nerve gives a good measure of overall peripheral function (cochlea
and auditory nerve). The detection of evoked responses from the brain stem gives a central
measure of the auditory function. We are particularly interested in the threshold of
electrophysiological responses as a function of frequency and to the existence or absence of
acoustic distortion of otoacoustic emissions also as a function of frequency.

In the future, we intend to modify the experimental protocol because several animals lost the
cannula implanted in the brain to hold the bottom part of the detecting electrode. We will
implant permanent cannulas and, to detect the signal, we will insert needle electrodes under
the skin at the level of the vertex and the mastoids. With this arrangement we can detect the
evoked potential of the brain stem. By doing this, we will ensure that the detection of the
responses from the two ears are absolutely identical and that the animals that survive will be
available at all stages of the experiment. We have just validated this methodology in normal

4- Relevance of completed and ongoing research

There are three main criticisms of this type of study. Firstly, the total number of subjects in
any experiment is far lower than the number of actual and future users of mobile phones.
Secondly, the exposure durations (daily and in total) are much lower than the duration of
mobile phone use of many users. Thirdly, it is difficult to find hamsters with normal initial
hearing. Some subjects, experimental animals or human volunteers, have a genetic
predisposition, with physiologies that are more prone to the development of problems.

The following are responses to these three criticisms:

If small-scale experiments were sufficient to detect a risk, larger scale experiments would
not be needed. Instead, one would concentrate at looking into the underlying mechanisms.
Experiments with more realistic exposure conditions would demand tremendous resources,
far beyond those actually available. However, in vitro experiments are much easier to carry
out, even if they seem less than realistic, and they can be used to confirm risk. The main
problem is dosimetry.

Technical remark: One should add that in vitro experiments are not in line with the current tendency
in research, and risk the creation of new questions without necessarily addressing the main issue of
health effects.

Second session of interviews, 23rd November 2000

4.5.8 Pierre BUSER

Academy of Sciences

I Response to written questions.

XIII- Could you summarise the main conclusions of the Academy of Sciences colloquium on
It appears to me that the there are three different aspects to the mobile phone problem. The
first is the measurement of physical quantities, the electric field in V/m, the power density in
W/m2 , the magnetic field strength in A/m or the magnetic flux density in µT. These
measures are related to an important element which is the power absorbed in the tissues of
the body, evaluated via the Specific Absorption Rate (SAR). This quantity is by far the most
difficult to assess because it cannot be obtained directly and requires measurement inside
tissue, which therefore must be modelled; it is generally expressed in W/kg of exposed
tissue either in the whole body or in 10 g or even 1g of tissue. These dosimetric evaluations
are essential as they give us an objective means of measuring the fraction of RF energy
from an external source that interacts with the body.

The second aspect derives from public concern and problems reported by users. Radiation,
even non-ionising radiation, has long been the object of concern about possible health
effects; the mobile phone has reactivated this somewhat dormant fear. With the exception
of a few cases of tumours, reported clinical symptoms are subjective: headaches, dizziness,
changes in sleep pattern etc. These symptoms led a number of European countries to carry
out research on animals and human volunteers. In France, the research was carried out
under the umbrella of the COMOBIO Programme, instigated at the initiative of the RNRT
and financed from public and private funds.

The third, sociological, aspect concerns deep-rooted "panic and risk". There are, for the use
of mobile phones as for many other activities that carry risk, human and social elements that
should not be forgotten.

The colloquium organised by the Academy of Sciences had three main themes. The report
of the proceedings reflects these three aspects. The first chapter is dedicated to dosimetry.
A second chapter reports on human studies other than epidemiology, and a third chapter
deals with animal and in vitro studies. The following chapter specifically deals with
epidemiology and almost exclusively that of cancer. The fifth part is more speculative,
dealing with future technological development. Finally, the last chapter is dedicated to the
management of hazards where four sociologists describe their visions of risk and hazard
management. What, in my opinion, can we be summarised from this colloquium? I will
report mostly on dosimetry and then on human and animal studies.

In the field of dosimetry, we had three presentations reporting good data. They
demonstrated to us that several aspects of SAR measurement (not easy to do) were very
well advanced, but that there is also a great deal still to be accomplished, including further
development of models. Experimental phantoms exist, and very accurate probes can be
used. However, some doubt remains as to what constitutes an acceptable measure of
specific absorption rate. The scientists base their assessment on the fact that the telephone
does not always function at full power. GSM produces a peak power of 2W at 900 MHz and

1W at 1800 MHz. It is generally the case that that, particularly in urban areas, the power is
only about one tenth of these maximum values. These values are significantly below the
levels considered acceptable in recommendations from organisations such as ICNIRP
(international), CENELEC for the European Community, NRPB for the United Kingdom, and
IEEE-ANSI and the FCC for the United States. These studies are very important because
they point to the necessity of revising part of the early human and animal studies that often
suffered from poor dosimetry. It is important to add that these values remain below the
values eliciting thermal effects.

It should be noted that almost all of the report of the Academy is dedicated to the mobile
phone itself, that is to antennas placed next to the ear. The report does deal with base
stations, and these are considered low-risk. Having said that, it is right to recall public
concerns about the siting of these fixed antennas. We have here a complex fear syndrome
despite objective reassurances and discussions with significant contributions from
sociologists. The report also discusses what may constitute acceptable exposure levels; a
downward revision cannot be discounted, particularly in view of the initiative to lower the limit
values in Italy and Switzerland (to respectively 2 and 6 V/m as against 40 and 60 V/m that
prevail according to current European and international recommendations).

2-Which symptoms can be studied experimentally in vivo or in clinical human volunteer

studies, how can we study the triggering of headaches, excessive fatigue, insomnia … in
relation to exposure to RF EMF?
I would consider first of all human cancer studies. I would like to mention only the study of
the International Agency for Research on Cancer (IARC) under the direction of Mrs. Cardis.
This project considers brain tumours as well as acoustic neuromas, tumours of the parotid
gland and also leukaemia. This is a wider remit than most other cancer studies. There are
actually several studies that have already produced more or less negative results,
nevertheless these investigations must be continued.

Other studies have reported epidemiological analysis of less serious effects. For example, a
Scandinavian study used a rather rigorous protocol to uncover a high incidence of
headaches from prolonged mobile phone use. This type of study must also be continued.

Some human volunteer studies have investigated effects on sleep. They showed a
shortening of the latency time for sleep or a reduction in the duration of REM sleep. Sleep
studies require careful planning to avoid confounding effects, such as the effect of stress,
but the results are most interesting. I will not dwell too much on electroencephalographic
analysis. It is quite possible to observe changes in the spectral power using signal analysis
but this is phenomenology, in my opinion; it has little explanatory value. Studies in
experimental psychology have shown a reduction in choice reaction time (simple reaction
time is not modified) during exposure to GSM signals. I do not know, however, how to
evaluate these observations. It is also important to note that no changes were observed in
human auditory evoked potential; a detailed analysis of such effects would have allowed the
study of latency and thus change in auditory function. Other observed effects are somewhat
doubtful, for example cardiovascular effects such as modification in arterial pressure
observed by one researcher but criticised because of their lack of accurate dosimetry.
Finally it should be observed that reported human effects such as headaches are currently
the subject of experimental animal research under the frame of project COMOBIO.

II-Discussion with the group of experts

Q: The conclusion you reached on effects on man is that the work is not yet completed. Is
this opinion based on the legitimate and common professional concern of always pursuing
further research or does it signify fears that the researchers have?

A: I think that both interpretations are valid. It is easy to criticise, but the research on man
and on experimental animals is very difficult for different reasons. In man, there is always
fear, even if we are dealing with an effect quite different from the direct effect of mobile
phones. The media does not necessarily differentiate between such effects. However, it is
quite impossible to neglect sociological phenomena, fear of risk, and to eliminate the stress
that this fear can eventually create. Clinicians are quite familiar with psychosomatic effects
of different kind. These should not be forgotten. There is also an element of stress when
dealing with animals. Having said that, I am rather surprised that most experimental studies
are carried out mainly on rodents. This species is very different from man. It is regrettable
that there are absolutely no studies on monkeys.

Technical remark: The main reason for that is that it is extremely difficult in view of a different shape
of the skull to reproduce the exposure to man in the rhesus monkey. Years of research have been
dedicated to the design of special antennas for rodents; a system has yet to be developed for the
rhesus monkey. Moreover, we have a lot of experimental data on rodents.

A: I acknowledge these reasons are often used in other fields of studies, in addition to
financial reasons and the difficulty of manipulating primates. It is clear that the use of rhesus
monkey or even baboons requires completely different logistics, a great deal of money and a
lot of effort but I still think it will be worth doing.

Q: We have the impression that this field of study is rather immature, as was the extremely
low frequencies about 8 years ago. Today, research at low frequencies leads to consistent
results from one group of researchers to the others.
A: I agree. It is necessary to standardise protocols in order to stabilise results and facilitate
interpretation. It is important to note that several of the chapters of the report of the
Academy of Sciences conclude with a ritual sentence; "further studies should be carried out
because the results are not conclusive". In other words "it has not been demonstrated that
RF is a risk to health, but the absence of risk has not been demonstrated either".

Q: It could be interesting to work on human and animal models that are subject to certain
pathologies. For instance, to investigate whether electroencephalograms that are slightly
abnormal are more affected than normal ones.
A: Absolutely. Slightly abnormal EEGs may be quite a sensitive indicator. We often lack
such sensitive indicators. I think one of the COMOBIO projects considered the use of
epileptic subjects.

Q: In a review article published in the summer, K Foster and J Moulder stated that if non
thermal effects actually existed, SAR would not be the most appropriate measure of the
exposure. What do you think of this proposition?

Technical remark: There is a widely held idea that SAR is a measure of a thermal effect: this is not
the case. SAR is representative of the square of the electric field. To determine it, one should
measure the electric field at a given point. By taking the square of the field, we loose information.
The SAR takes into consideration absorption in different types of tissue because it includes the
conductivity of the material as a factor.

A: To respond to the technical remark, it is true that SAR is not a measure for thermal
effects only, it is more generally a measure of absorbed power. But a measurement of
electric field strength only is not sufficient to evaluate absorbed power because, as in the
technical remark, the conductivity of the irradiated tissue is a multiplying factor and its mass
density r is a quotient (SAR = sE2/2r). We have yet to determine if SAR is the best
measure of RF exposure given that we are dealing with situations that are below the thermal
threshold. I certainly repeat myself here, but I would like to stress the necessity of dealing
with not just the physical effects of the irradiation, but to place it into a wider context in which

psychological effects and the fear of risk are also included. Among all the messages
directed to the media this last one deserves special attention.

In conclusion, I would like to underline that the report of the Academy and the Colloquium on
which it is based represent a snapshot of the state of knowledge and that it has no
pretension of being anything more than a progress report of how things stood at the time.

I did not have time to report here the message of the sociologists with respect to risk
management, and what should and should not be done when facing a panic situation. I
have no doubt this would have been of interest.

4.5.9 Philippe HUBERT

Institute of Nuclear Protection and Safety (Institut de Protection et de Sûreté Nucléaire -

I Introduction

Technical Note: Philippe Hubert has thought deeply about the incorporation of scientific uncertainty in
decision-making. He is part of a group that considers risk evaluation within the national committee for
health and safety. From an epistemological perspective, the situation of mobile telephony today is
similar to that of ionising radiation about ten years ago, hence the interest in this interview.

I look after risk evaluation and the development of risk management at IPSN, mostly in
relation to ionising radiation. The department has a unit for the evaluation of impact, another
unit to deal with the epidemiology of radiation and finally a unit on risk management that
deals with the principles of management from a philosophical as well as from a risk
perception perspective.

Technical note: The following text was submitted by Philippe Hubert, it is followed by the transcription
of the discussions with him.

The management of risks from radiation is a very large field. It relates to the exposure of the
general public and to that of workers. It bears on activities using nuclear energy such as
nuclear reactors, nuclear-propelled ships, nuclear military installation. It relates to the
properties of ionising radiation as used in diagnostic radiography, radiotherapy, material
control, sterilisation, surface treatment, baggage and parcel inspection, fire detection. Also
included within the domain are activities leading to an increase in "natural" exposure such as
mines, air travel, certain rare earth and ceramic materials, geothermal spas and bottling of
mineral water. The field encompasses also the reduction of natural risks from, for example,
radon in buildings. Finally, “radiological protection”, the control of radiation doses to patients
in cases of radiological examination or treatment is also covered. In addition to a few
hundred thousand workers (in France about 250,000 people are monitored), the whole
population is affected to a greater or lesser extent (diagnostic radiology, radon).

The hazards that must be controlled fall in two groups. The first group contains the
threshold effects associated with high doses, giving rise to acute effects such as burns and
the destruction of tissue, problems with spermatogenesis, cataracts, teratogenesis; these
are referred to as “deterministic effects”. Remaining below the exposure limit values
eliminates this type of effect. The second group contains different effects that do not
necessarily manifest themselves but the probability of their occurrence increases with the
dose (cancers, congenital malformations). These are called “stochastic effects”. The
Exposure Limit Values do not protect against this type of risk, its management relies on the
concept of acceptability.

Since the identification of risks and the first warnings (1896 for burns) an elaborate system
of risk management has been developed and adopted at the international level. The
summary of this system and the way it functions are described below. The advantages and
disadvantages of this “model of risk management” will be discussed later and placed in the
perspective of the of management principle under discussion.

Development of a system of radiological protection

The evolution of the system has three main parts: development of industrial and institutional
practices, advances in the knowledge of the effects of radiation and finally the development
of risk management [CE 1996].

Development of institutions and practices

The first radiological practice was the application of “X-rays” to medical diagnostics:
discovery in 1896, commonly used two years later, incorporated in army equipment in 1914.
“Röntgen Societies” existed before the First World War and they provided recommendations.
The use of Radium (luminescent paint, the harmful effects of which were soon recognised)
was common in the twenties. The first Radiological Congress (London 1925) decided to
establish an International Commission for Radiation Protection (ICRP), previously known as
the International Committee for Protection Against X-rays and Radium. ICRP was officially
established in Stockholm in 1928. ICRP proposed its first standard in 1934 (Radiological
Congress in Zurich). The health (diagnostic and therapy) aspects were at that time the
primary concern.

The Second World War saw the birth of the military nuclear industry (mines, nuclear
reactors, ammunitions, arms, recycling). After the war, the civil nuclear industry developed
in turn, followed by other applications of radiation (sterilisation, smoke detectors, thickness
gauges). Little by little, the problems related to the nuclear industry lead to different
developments in radiological protection.

ICRP, reconstituted after the hostilities, produced formal publications from 1955 onward.
Since then, the changes have been documented, argued on scientific grounds and in terms
of management objectives. The US National Council on Radiation Protection (NCRP),
formed in 1946, played a similar role in the American continent. The proposals of the two
institutions were always quite similar. However, the first limits of public exposure were
proposed outside this frame (tripartite conference on nuclear industry EU, RU, Canada).

The fifties were characterised by nuclear weapons tests. The increase in fall-out and the
Bravo” test of 1st March 1954 (contamination of the Japanese fishing boat ‘Lucky Dragon”
and tests at Bikini Atoll) lead to the creation by the United Nations of the United Nations
Scientific Committee on the Effects of Atomic Radiations (UNSCEAR). The main activity of
the committee is the gathering of information, including data on effects and on sources of
exposure whether natural, industrial, military or medical The Committee then proceeds to
produce complete, critical reviews that constitute an important source of information without
equivalence for other physical agents.

The International Atomic Energy Agency is an intergovernmental organisation founded in

1957. The Agency publishes “Basic Safety Standards” jointly with many organisations
including the International Labour Office. In Europe, the Euratom treaty lead to the
establishment of basic standards (first Directive in 1959) of which the last edition appeared
in 1996 (Directive 96/29 Euratom Council, 13 May 1996).

Today we have scientific institutions (UNSCEAR) that support government organisations

(ICRP, NCRP) in the formulation of recommendations that are subsequently incorporated in
regulations, for example by the European Union (it is generally understood that basic

standards are based on the ICRP recommendations) or by intergovernmental agencies

From the identification to the quantification of health effects

Deterministic effects that appear almost immediately after exposure to radiation have long
been recognised. The first study reporting effects on the skin dates from 1896. During the
course of the 1914-1918 war there was very significant development in medical radiology
and the extent of radiation-induced damage was very high. The medical profession and the
“Röntgenologists” became aware of the danger. The appearance of skin erythema (first
degree burn) became a reference for the definition of limits for nearly 50 years (today we still
have a "skin" limit).

The other deterministic effects (cataract, decrease in fertility or sterility, foetal malformations
due to in utero irradiation) were identified early. There was a great deal of research on the
subject because all these effects and their mechanisms are not completely understood and
the stakes are high in situations of accidental or therapeutic exposure. However, the
management of these effects is not a controversial subject as none of the research studies
so far has put in question the pertinence of the limits in force.

Leukaemias and other blood diseases, and bone cancer, were identified in the twenties in
workers operating X-rays machine and radium sources. The Congress of Stockholm
concluded that “known effects that must be prevented are: (a) damage to superficial tissue,
(b) problems with the internal organs and changes in the composition of blood” [Norstedt
1929]. The mechanisms involved were not really considered to be significantly different from
acute effects and the knowledge available at the time did not lead to changes in risk

Before the Second World War, researchers identified hereditary effects (also known as
genetic effects). These are mutations that are capable of leading to malformation in the
offspring and descendants of exposed individuals. This had already been seen in animal
experiments (animal models: drosophila). This risk was known to researchers at the time of
the Manhattan Project . It was behind many of the research studies carried out in Hiroshima
and Nagasaki. Today, genetic effects are considered a lower risk than cancer (about 1/3 in
total in 1977 and 1/7 in 1990). There are still no epidemiological studies underpinning the
animal experiments. It is important to note that some genetic mutations can be recessive
and this has to be taken into consideration in the evaluation of risk at the societal level. The
unpredictable nature of genetic effects has introduced the notion of “stochastic” effects
(where the effect is not directly related to the irradiation in a deterministic manner);
stochastic effects are also characterised by the absence of threshold.
In the late fifties, ICRP stated that there were a number of identified effects that should be
avoided [ICRP 1958]:

· “Acute somatic” effects: burns

· “Deferred somatic” effects: leukaemia and other malignant diseases, cataracts, problems
with fertility, reduced lifespan (this latter effect was considered as one of the more
plausible effect)
· “Genetic damage”, that is congenital malformation in future generations.

It can be seen that all the effects had been identified even though their importance has
changed considerably since then, and so has the development and understanding of
mechanisms (damage to the genetic material inside the cells was not then considered as a
carcinogenic mechanism). Very few of these effects were supported by epidemiological
studies, a discipline that was not very informative and had little credibility among biologists.
For example, the epidemiological studies of A Stewart on the effects of in utero irradiation
were not well received. The effects that were most feared were those supported by

experimental studies [Muirhead 1989]. In man, leukaemia and bone cancer were identified
from clinical rather than epidemiological observations.

In the sixties, advances in epidemiology changed the picture. Several kind of cancers were
observed (skin, thyroid, breast, lung, liver, stomach etc) placing cancer at the top of the list
of hazards, well ahead of reduction in life span and, to a lesser extent, genetic effects that
have not been supported by the epidemiological evidence. Another important change is that
the discussion focuses on the quantification of risk. The linear dose-response relationship
without threshold value has been postulated as a tool for risk management. Questions now
relate to the estimation of risk parameters and to the principles involved in obtaining them,
for example, are the risk parameters the same for low and high dose rates? Can we
transport findings from one population to another? Scientists now make full use of
epidemiological studies, without neglecting animal and in vitro studies that are used in
discussions of dose level and equivalence between irradiation times. Studies on survivors of
Hiroshima and Nagasaki played an important role in the construction of a dose-response
relationship but they are not its only determinants. ICRP, for example, did not use this study
in its evaluation of cancer of the skin, thyroid, liver and bone (ICRP 1991). The BEIR
committee has used a large number of other sources in parallel with that of Hiroshima and
Nagasaki, in particular for breast cancer and cancer of the thyroid [BEIR 1990].

UNSCEAR has made a systematic comparison of the coefficients referred to in various

studies [UNSCEAR 1988], [UNSCEAR 1994]. The relationship between lung cancer and
radon is based on studies on uranium, iron and tin miners. The epidemiological studies can
be classified in three groups: studies that do not provide any significant results; studies that
provide a dose-response relationship, and studies that provide a positive result but do not
provide any information leading to the quantification of risk coefficients. In this last category,
the effect may be well demonstrated but the doses not well characterised (for example
doses to doctors exposed during the Second World War, cancer of the thyroid in children
around Chernobyl) or else it may provide results that have yet to be consolidated (for
example, studies on leukaemia in workers in the river Techa, and studies on workers in the
nuclear industry). In utero irradiation falls in between: the doses are not very well defined
and there are arguments about methodology [Muirhead 1989].

To appreciate the extent of the extrapolations needed, we can plot the actual exposure
conditions and the epidemiological study in a dose/duration plot (Figure 1).

Durée d'exposition (années)Radiographies in utero
Expositions usuelles
Etudes H&N
permettant Thymus
la quantification Teignes
plus faibles groupes significatifs Mastites
Etudes mettant en Minute
évidence un effet 0,00001 Spondylarthrites


Expositions professionnelles Thyroides
Iles Marshall
Populations Tcheliabinsk
Trav. Mayak
Travailleurs, CIRC

P.H. Novembre 196 100

0,01 0,1 1 10 100 1000 Dose (mGy)

Figure 3 Epidemiological studies and actual exposure situations; comparison of

doses and exposure durations

The public is exposed to doses on the order of 2.5 to 5 mSv per year, depending on location.
The dose accumulates throughout a lifetime and may exceed 200 mSv. For diagnostic
radiography, the average makes little sense because the exposure is of short duration and
targets specific organs. While generally low, the doses to specific organs may be high, for
example in angiography and body scanning. The exposure of workers is of the order of 2
mSv per year in the nuclear industry but the regulations allow up to 50 mSv per year, and
this is the case for some people.

The idea of “extrapolation to weak doses” may not be entirely correct. The dose levels in
epidemiological studies do not require significant extrapolation to relate to actual exposure.
For example, the average dose to a survivor from Hiroshima and Nagasaki is about 130
mGy (230 in the 35 000 people exposed), exposure in excess of 50 mSv are considered
significant [Pierce 1996], even though 100 mSv is considered a more plausible value. For
thyroid irradiation in children, the risk is significant at fairly low doses for example of the
order of 100 mGy for Israeli children treated for tinea or in the range of 0-300 for those
treated for thymus hypertrophy [Shore 1992], [Shore 1993]. These levels have reduced
significantly in the last two decades because of the statistical power accumulated throughout
an increased lifespan. In the seventies, no increased risk could be observed below 1 Gy. At
the time of the 1990 ICRP recommendation, that level was 0.2 Gy.

In contrast, dose rates are very different (except for diagnostic radiography); the extent of
extrapolation is no longer one order of magnitude, but rather 4 or 5. The relevance of the
first results relating to workers in the nuclear industry, residents and workers from the river
Techa, or even the cases of cancer of the thyroid around Chernobyl are no longer very clear.
Even if the results are not sufficient to estimate a risk coefficient, they show that the
induction of cancer occurs in these circumstances. The extrapolation is not more precise
but it does have a better foundation.

The same exercise can be carried out for exposures to radon; most studies on miners were
collectively reviewed [NCI 1994]. Smaller differences were observed for both rate and level
of exposure. For example, in his working life, a miner in the French study received the same
dose as a resident of the two or three most exposed regions would over his whole life. To
validate in the general population the estimation of risk deduced from the study of miners,
case-control studies were carried out in different countries. The first results did not lead to
the determination of a clear effect related to radon at the doses currently found in the
general population. Some studies were positive, others did not lead to any significant
results. A recent meta-analysis of the best eight studies in this field produced a risk factor
closed to that of miners [Lubin 1997].

The estimation of risk of death by radiation-induced cancer, from successive reviews, is

presented in Table 1. It is important to know the choice that we make when following an
epidemiological study in a living population. Should it be the absolute risk (excess cancer
per year per unit dose and per 100,000 persons-year) or the relative risk (proportion of
excess per unit dose with respect to the natural incidence) that will remain constant? By the
end of the eighties it was possible to close off the alternative by eliminating the “absolute
risk” model. The old values of the American Academy of Science (BEIR) show that if the
relative risk model were chosen at the time, the estimate would have remained fairly
constant. Unfortunately, this has not been the case. It is also important to note that the
estimates call for a "reduction factor" of 2 for low doses and low rates.

Table I: Evolution of risk coefficients

Probability of death in radio induced cancer: % per Sv

Acute irradiation, X and gamma rays Recommendation for
“current exposures"
“absolute risk “relative risk
constant” constant”
1972 : BEIR I 1.2 6.2
1977 : 2.5 non-estimated

1977 : ICRP 1.25

1980 : BEIR III 0.8/2.5 2.3/5

1988: 4/5 7/11

1990 : BEIR V No retention 8.85

1990 : ICRP 60 5

Discussions on management regulations

The change in the acceptable level of exposure should be discussed in the context of three
main questions: the status of the exposure limit, the concept of acceptability and the
principle of optimisation. At the same time, the concept of “dose” has changed significantly.

The concept of dose is not self-evident and its current status often leads to errors. Today,
“effective dose”, the quantity used to express the main exposure limit, is not a physical
quantity but rather an indicator of different exposures (to different organs, from different
types of radiation), as a function of their alleged effects and their weighting (one death by
radiation-induced cancer is equivalent, for example, to a serious anomaly in offspring). It is
more a management tool than a physical quantity.

The limits therefore apply to different doses and have had different objectives. The first
limits intended to eliminate risks. The first measure of “quantity of X-rays” was adopted in
1928 on the basis of ionisation in air, it was termed Röntgen (r then R). Other researchers
attempted to define a Skin Erythema Dose (SED) in the twenties. As can be seen, right from
the beginning, it was necessary to put together physical as well as biological concepts of
dose. It was possible to estimate that SED corresponds to 600 r (later this dose was
estimated to be 500 rad or 500 rem for X- and gamma rays and finally 5 Gy or 5 Sv). The
first value from ICRP (Zurich, 1934), was 0.2 Röntgen per day corresponding to about 500
mSv. That was considered “tolerable dose”.

In 1941, the concept of “maximum admissible dose” was proposed in the United States to
incorporate the concept that the limit is not an absolute guarantee. This value was reduced
to 0.05 r per day by the National Council on Radiation Protection and Measurement (NCRP)
around 1946, followed by ICRP in 1950), corresponding to around 125 mSv per year. These
values were established particularly to avoid hereditary effects. The concept of “limiting the
dose of the public” was proposed in 1949 (EU Conference, RU, Canada) for the first time
and was fixed at Harriman in 1953 (1.5 Röntgen per year: about 15 mSv, or about 1/10 of
the dose to the workers).

The first concepts of dose did not allow for the regulation of radium intake (a limit of 0.1 µg
was proposed in 1941). The concept of cumulative exposure was not explicit and the doses
were expressed on weekly or monthly bases.

In 1958, at the time of the first ICRP recommendations, the current concepts were
discussed. The “physical” dose was the rad, corresponding to the ionisation produced in
tissue rather than air (1 rad = 0.01 Gy = 0.01 J Kg-1). The primary biological equivalent was
the rem (it was noted that neutrons and alpha particles were more “biologically” active) and
the concept of cumulative exposure was introduced with limits over a lifetime. The concept
of collective dose was also developed on the basis of genetic effects. It became possible to
calculate dose from the concentration of radionuclei in the body and hence it became
possible to convert rems into admissible concentration. The concept of “whole-body-dose”
was outlined but was not valid except in cases where “the organs received a fairly similar

The concerns over “genetic effects” or “hereditary effects” led to two new concepts that were
formulated by ICRP in its 1958 recommendations. The limit value for reproductive organs is
a “ceiling” value deriving from a balance between cost and benefit. The concept of
“collective dose” was introduced indirectly. A limit of 5 rems (50 mSv to organs) applied to
the average dose to the population. The system is based on a linear relationship without
threshold. The concept of acceptability (health-socio-economic) was introduced and related
to the limit. ICRP was of the opinion that these values imposed a “heavy burden on society”
in view of hereditary damage but it was tolerable in view of the benefit to be expected from
nuclear energy. The need to keep the doses as low as possible was stressed.

For all other risks, the “permissible dose” had to restrict the risk to a “negligible” level that
could not be detected at the level of the population by “statistical methods applied to a large
group” and remained “probably hidden within normal biological variation”. We have here a
second concept of acceptability, different from the previous one related to “non detectability”
([ICRP 1958]§31). At the same time, the absence of threshold and the absence of “repair”
of organs tissue long after the exposure were postulated as prudent hypotheses. The limits
were more stringent for gonads, haematopoietic organs and the cornea (0.1 rem per week,
about 50 mSv per year). For the population (because of children) the limit for the organs
was fixed at 0.5 rem per year (5mSv).

Later, the linear relationship without threshold took on an increasing importance and “risk” of
radiation-induced cancer was judged as more important than hereditary effects. The
publication 26 of ICRP of 1977 [ICRP 1977] formalised the system with weighting
coefficients per organ and per type of radiation. The concept of central dose became “whole
body effective dose”. It did not apply for the estimation of acute effect for which the doses to
organ remained the only pertinent measure. It allowed the assessment in a uniform manner
of all exposure situations but requires some important assumptions (for example, that the
rate of exposure has no effect on carcinogenesis) that one tends to forget when using this

Recommendation 26 confirmed a limit of 50 mSv (on the idea of “whole body equivalent
dose”). For the first time it associated a risk to this limit (about 10-3 per year of exposure). It
was therefore a limit based on responsibility, defined by comparison to the risk of workers in
established industries (on the basis of the hypothesis that the average dose of the worker is
1/10 of the limit), and the risk was calculated using a dose effect relationship that is
considered as “reasonably prudent”. Except for acute effects, the protective character of the
limit was definitely rejected and the word "acceptable" was used instead.

The principle of optimisation –“all doses must be maintained as low as reasonably possible”-
was integrated in the recommendation and constitutes one of its main elements.

The new 1990 recommendation (publication 60 of ICRP [ICRP 1991]), in addition to lowering
the limits (20 and 1 mSv per year averaged over 5 years), did not qualify the limit values as
“acceptable” but as “the limits of what is unacceptable”. The exposures were now only
acceptable if they were below the limit and had been reduced to the minimum possible.
Acceptability and optimisation are therefore related and the acceptability no longer judged in
terms of risk level but on the quality of risk management. The text contains long
deliberations on how to judge optimisation and a framework of how to achieve it in practice.
The notion of genetically-significant dose had disappeared. It is only rarely that the
hypothesis is described as prudent. The need to re-evaluate the risk coefficient made ICRP
more prudent and less likely to make such an affirmation.

Limit values, concepts of dose and acceptability were modified with time and have resulted
in the current system of “risk management of acceptable risk” the outline of which is given

Recommendations in the process of being adopted.

The recommendations of publication 60 of ICRP [ICRP 1991] replaced that of ICRP
publication 26 [ICRP 1977]. These recommendations underlie the 1996 European directive
[EC 1996], which is in the process of being translated into national law in the countries of the
European Union. The scope of the directive is fairly similar to that of ICRP except for public
exposure to radon, which is the subject of a separate recommendation, and medical
exposure of patients which is also the subject of a separate document.

Basic principles: dangers and estimation of risk factors

The recommendation is based on data available in 1990 on “deterministic” effects
(previously called non-stochastic effects), on carcinogenesis (data collected by UNSCEAR
[UNSCEAR 1988]) and on hereditary effects. The hazards are those already described.
However, the risk coefficient for the probability of cancer per unit dose is significantly higher,
essentially because the hypothesis of absolute risk was discounted by analysis of data from
Hiroshima and Nagasaki (cf. Table I).

A dose-response relationship is produced by applying risk coefficients to the basic data;

these are then extrapolated and applied to other populations to calculate the “whole life” risk,
or to other exposure conditions. Certain hypotheses are related to demographics (the risk
increases with increasing life expectancy), others depend on the way in which the
extrapolation is made from one population to another, from young people to the rest of the
population or even from localised to whole-body irradiation etc. Data from animal and in
vitro studies were used, for example, to determine the equivalence of neutrons, X- and
gamma rays, or to establish the effect of dose rate. Value judgements are needed for the
choice of indicators (probability of death rather than loss of life expectancy, aggregation of
fatal and non-fatal effects)

The only explicit extrapolation rule (apart from that on the nature of the radiation) is that of
dose and dose rate effectiveness factor (DDREF). The epidemiological data are from
populations exposed to relatively high doses, delivered in a very short time (cf. Figure 1).
ICRP postulates that for X- and gamma rays the effect per unit dose is less important at low
doses and low dose rates. ICRP has used a DDREF of 2 since 1977. It should be noted that
higher values were suggested (3 to 10). The DDREF must be applied to doses less than 0.2
Gy or at dose rates less than 0.1 Gy h-1 .

The risk coefficients for of death from cancer after whole-body irradiation are evaluated in
terms of probability of death from radio induced cancers:

5% Sv-1 for a population of all ages at the time of exposure;

4% Sv-1 for a population 18 to 65 years old.

These coefficients equate to quite high levels of risk: 12,000 deaths by cancer can be
attributed to natural and medical irradiation in France. The coefficients of genetic risk for
future generations:

1% Sv-1 for a population of all ages

0.6% Sv-1 for a population 18 to 65 years old

A metric has been developed to include the risk of non-fatal cancer (essentially skin and
thyroid cancer) by weighting the probability of occurrence by an index of loss of quality of
life. This weighting factor is added to the risk factors for mortality and hereditary effects.

"Lifetime detriment" is calculated to be

7.2% Sv-1 for a population of all ages

5.5% Sv-1 for a population 18 to 65 years old

Hence, the measure of risk depends on management as well as on end-point. Loss of life
expectancy might perhaps have been chosen. Analysis of data from Hiroshima and
Nagasaki would have had very little influence on this metric, and the lowering in the
exposure limit values would have been minimal, which demonstrates the importance of “non-
scientific” choices.

Risk assessment; dose and dose equivalent

Certain coefficients have changed since 1977 but the general procedure remains the same.
It consist in deriving from a physical quantity (the absorbed dose), an “equivalent dose” so
that one can use the same quantity to evaluate exposure from different types of radiation in
different organs. These are indicators or end-points for risk management and are based on
equivalence of risk. They form the basis of practical regulations.

The absorbed dose is the basic dosimetric quantity. It is a physical quantity (DT for dose in
tissue), defined as the energy deposited per unit mass, expressed in the unit joule per
kilogramme ( or Gray, Gy).

Many biological effects depend on the nature of the radiation. ICRP defines weighting-
factors for the quality of the radiation, to calculate a weighted dose called organ dose
equivalent (HT = S RWTW RDT.R where DT.R is the dose absorbed in organ T from radiation
type R as described in Figure 2). The name given to this unit is the Sievert (Sv). The
weighting factors are derived from data obtained from animal and in vitro studies. As the
weighting factors are dimensionless, the unit is still equivalent to the joule per kilogramme,
but this is no longer an actual physical quantity (cf. Table II). Among the changes that could
have important practical implications, one should note the change in the weighting factors for
neutrons from 10 to 20 for a large part of the energy spectrum. Finally, the effective dose
allows for the summation of exposure to different organs. It is equal to the sum of the dose
equivalent to organs weighted by the weighting factor for organs:

E = å WT WR DT .R

On the basis of risk factors per organ, from epidemiological data, ICRP established relative
contributions of different organs (Table III).

Table II: Weighting factors WR

Nature – energy WR
Photons (all energies) 1
electrons, muons (all energies)
Neutrons < 10keV 5
10 keV – 100 keV 10
100 keV – 2 MeV 20
2 MeV – 20 MeV 10
> 20 MeV 5
Protons > 20 MeV 5
Particles (a, fission fragments, heavy nuclei) 20

Dose équivalente
à un organe =S Wr Dr
Dose efficace =S Wt Dt Dneutron
D thyroide Dalpha
D Poumon
D gamma
D estomac
D béta
D gonades
D peau

Table III: Weighting factors WT

Tissue or organ WT
Gonads 0.20
Bone marrow 0.12
Colon 0.12
Lung 0.12
Stomach 0.12
Bladder 0.05
Breasts 0.05
Liver 0.05
Oesophagus 0.05
Thyroid 0.05
Skin 0.01
Bony surfaces 0.01
Others 0.05

The collective effective dose is calculated by integrating the effective dose for all exposed
people over the whole exposure time. As with equivalent dose, it assumes linearity of the

dose-response relationship. It is an indicator of collective risk, measured in the unit Man-
Sievert (M.Sv).

An important characteristic of radiological protection becomes evident: the basis for practical
evaluation, the dose, has its origins in risk evaluation

Scope and principles

In defining scope, ICRP, and the European Union, differentiated clearly between two types
of human activities for which there are different rules:

- activities that lead to an increase in exposure from the use of existing sources, from new
sources (by definition controllable), from new exposure situations or exposure of previously
unexposed individuals. These activities are termed “applications”. The nuclear industry,
medical radiography, industrial gamma radiography are typical examples.

- activities that decrease the exposure of people by acting on sources, modifying the
exposures they produce or decreasing the number of people exposed. These activities are
termed “interventions”. The term applies to the management of accidents and also to the
reduction of exposure to natural radiation (for example, the case of radon in residential
areas) and also to the management of situations inherited from the past (for example sites
polluted by historical radium usage).

It is difficult to classify activities that do not themselves use any radiation, but lead to an
increase in natural irradiation: crewing of aircraft, mining, work in caves with radon are
examples of this. The European Directive requires that member states make note of the
situation but it allows them the freedom to apply part or all of the radiological protection

Control of acute effects is achieved by keeping doses below the relevant threshold. Control
of the risk of carcinogenisis is more complex. The protection system is based on three
principles first formulated in 1997:

· Justification: requires the exposure to result in a net benefit when balanced against the
radiological detriment.
· Optimisation: means that the doses form any source must be as low as reasonably
possible bearing in mind economic and social considerations.
· Limitation of doses (and risks): the exposure of people to sources that can be
controlled must be subjected to dose limits.

The text stresses that it is not good practice for the levels to be near the limits. The limits
are clearly defined as “bordering the unacceptable”. Exceeding the limits is not
acceptable. Below the limit, the level of exposure is only acceptable if all efforts have been
made to reduce the exposure and the principle of optimisation has been correctly applied.
ICRP suggests that very low doses may be ignored. It does not define what can be
neglected, but the European directive provides levels of exemption based on activity.

Three population categories are considered: workers, the general public and medical
patients. For “applications” the three principles apply to all categories except medical
exposures. For interventions the limits are not applicable: an intervention must be justified
and optimised. Levels are given for guidance but do not have the status of limits. In the
case of radon, ICRP 65 [ICRP 1994] suggests that action may be appropriate above an
"action level" of between 3 and 10 mSv per year (200 to 600 Bq m-3). There is guidance for
accidental exposures (levels at which counter-measures are appropriate) in ICRP 63 [ICRP
1995]. Experience has shown that a value 1 mSv is very appropriate in post-accident
situations. In the case of “situations inherited from history”, the outcome is unclear.

This system makes it possible to treat very different problems within the same conceptual
framework. It was developed progressively to give coherence to many situations that would
otherwise have had to be treated on a case-by-case basis. The introduction of intervention
should enable the future development of a unified analysis of situations of increased natural
irradiation. The three principles have now been around for nearly 20 years. ICRP subjects
them to critical analysis periodically, and a new revision is under consideration.

Limit values
For workers, ICRP bases its limit on an evaluation of risk and proposes an average annual
dose of 20 mSv for a period of 5 years with, no more than 50 mSv in any one year. To
derive this value, ICRP calculated the consequences, in terms of several end points, of a
working-life exposure (18 to 65 years) to annual doses of 10, 20, 30 and 50 mSv. It should
be noted that the risk for one year of exposure to 20 mSv is about 10-3 (death by radiation-
induced cancer or equivalent detriment following non-lethal cancer or hereditary disease).
This value is quite high, effectively “bordering the unacceptable” which accounts for the
insistence of ICRP on increased optimisation. Notwithstanding subtle changes in the
decision-making process, the lowering of the values is essentially a repercussion of an
increase in risk coefficients. The Exposure Limit Value (ELV) (that is “effective dose”)
ensures against acute defects but an additional limit of 500 mSv (equivalent dose) is
necessary for the “skin surface”. The same approach was used to set limits for the general
public. This led to the proposal of a limit of 1 mSv per year (which in terms of risk to life from
one exposure is 7x10-5, which is quite high) and, in very exceptional circumstances, 5 mSv
in one year on condition that the average of 1 mSv/year is met over 5 years.

ICRP attaches great importance to the practical implementation of its recommendations.
For example, it identifies obligations on authorities and on operators, and defines
procedures for evaluating doses. It identifies the need for justification, and for the availability
of valid and efficient radiological protection protocols.

The final definition of "dose constraints" is left to the authorities. These are values, below
the limits, which authorities can use to ensure that optimisation will not lead to undesirable
individual doses. The constraint must ensure that optimisation will not lead to individual
doses that much higher than average (certain individuals will be exposed near the maximum
allowed, others not at all), must reflect the number of sources (people exposed
simultaneously to several sources could receive high doses), and should represent the
state-of the-art in radiological protection (the operators must be encouraged to implement
“good practice”). ICRP expects that, in the majority of cases, the real level of protection will
largely depend on the implementation of constraints. Operators have obligations with
respect to procedures. Some of these obligations are quite general, such as the
development of a safety culture and the creation of an organisation into which radiological
protection fits naturally. Other obligations are more specific. For example, there is are
obligation to appoint a trained "radiological protection advisor” and to classify occupational

Impact of the new European Directive

Notwithstanding all the changes described above, the lowering of the Exposure Limit Values
is the most apparent difference between the 1996 directive and the previous one. It is not
the most important change related to industrial and medical exposures. In France, the
number of workers exposed above 20 mSv was of the order of 1000 in 250,000 in 1996; it is
of the order of 100 today. Over the same time, uses of radiation have widened with a
number of occupational activities leading to an increase in natural exposure (for example
work in mines, thermal spa establishments, increased air travel, exposure to radon in
working environments etc).

One important change is the placing of a greater emphasis on the practical aspects of
radiological protection, with more rigour in the execution and evaluation of specific actions.
There are recommendations relating to the role and responsibilities of employers, users of
sources and Member States and their agencies. The text identifies a duty of care, which
makes effective optimisation imperative.

Main characteristics of the system of management

Evolution and efficacy of ground rules
The system described above is mostly based on the setting-up of ground rules for
optimisation, placing obligations on means rather than results, which makes it difficult to
manipulate. Doses close to the limit would mean that the system has failed; the records
show that it has actually operated well. Whether in the nuclear industry or for workers in
medical centres, the actual ELV is hardly ever reached except on rare occasions best
described as "incidents".

The exposure of workers can be measured by means of a personal “dosimeter”. We do not

yet have exact exposure statistics and one of the current challenges is to construct a
national database. A 1997 OPRI/IPSN report showed that of the 250,000 people being
monitored in France, most work in the medical sector (about 140,000) or in the nuclear
industry (65,000). The non-nuclear industrial sector (chemical industry where radioactive
materials are used, thorium welding, and most importantly non-destructive testing using
gamma rays) is not negligible, with about 22,000 people. The teaching and research sector
has about 15,000. In the near future, commercial aircraft personnel, who are exposed to
radiation at high altitude, will also be followed. The new European Directive on radiological
protection indicates that any annual doses of the order of 2 mSv/year must be considered
“attributable to human activity”. Even in the most exposed sector, veterans of the nuclear
industry, the average dose is of the order of 3.5 mSv, which is higher than the overall
average but clearly lower than the exposure limit value. In 1997 there were about 400
people who received a dose in excess of 20 mSv per year: that has been the maximum
permissible dose for a worker in the nuclear industry since the year 2000. Most of these
were workers in the nuclear industry, with about 100 from the medical sector. The number
of workers in this high exposure category has decreased in the last few years having been
about 1000 in 1995.

In spite of practical difficulties, it has been possible to manage simultaneously exposure of

the general public, workers and patients to natural radiation as well as human activity. For
example, almost all the European countries have a well-defined procedure for addressing
radon even if some national governments are more stringent than others [Massuelle 96].
The new European Directive enables all countries to apply the most advanced practices
concerning patient exposures from diagnostic radiography.

Although average doses have always been clearly lower than the ELV, the development of
optimisation has been slow. The formulation of the principle in 1976 was the result of
deliberation over a long period. It was not immediately applied practically. The early work in
this field considered “the financial value of the man-Sievert”: the question of the value of
human life. It was only after a number of case studies, retrospective rather than prospective,
that the approach was applied to exposures within Electricité de France (EdF) and other
companies in the electrical sector in the late 80s. Changing the way in which steam was
generated was considered a key step in the generalisation of the ALARA principle.

It should be noted that “good practices” were not widely used and, where used, they took
about a decade to be applied [Hubert 1994]. Management of the exposure of patients
during diagnostic radiography differs significantly between European countries, with doses
ranging from 1.65 mSv (France) to 0.3 mSv (Netherlands) [UNSCEAR 1994]. There is room

for a reduction of dose of about 99000 Man-Sv, and half of this would be possible at very
low cost. In medical centres, provisions are being made for the management of exposure
from very specific sources (surgical intervention under radiation for example). Outside the
nuclear industry very few steps have been taken towards optimisation. On the other hand,
the practical implications of the approach may not be well understood by all the interested
parties or by the authorities. The statutory basis of the principle of optimisation is not well-
established and there are few precedents and references to rely on [Hubert 1993].

Progress has still to be made on the improvement radiological protection via dosimetric
considerations. It is more important to reduce doses to exposed workers (100 Man-Sv) and
to the public around installations (the order of 60 Man-Sv) than exposures from radon (about
70000 Man-Sv) or diagnostic radiography (about 100000 Man-Sv). Operational radiological
protection has taught us that changes happen slowly. It has shown us that a management
regulation that is well accepted in principle cannot be placed in operation immediately. The
complete system of risk management, the exchange of information, the operational rules
and the procedures for implementation must be adopted and assimilated by all the players in
the field before it can be applied.

Novel approaches
Having realised that a no-risk situation was not possible (because of zero-threshold
carcinogenesis, and the presence of a certain level of natural exposure), the next step was
the abandonment of the idea that an acceptable or negligible dose level could be defined.
ICRP did not just define a level above which the risk would be “unacceptable” (about 1 in
1000 for one year for one worker, which corresponds to the ELV), it indicated that an
exposure above any level is unacceptable if the dose could be reduced easily. The
acceptable dose therefore is a dose that is below the limit and has been reduced to “as low
as reasonably possible”. Acceptability is therefore no longer judged by level of risk (which
may exclude but cannot prove acceptability), but on the implementation of management
practices. This approach, that can be described as “risk management of acceptable risk “
was developed very gradually and is now quite mature. It must be noted that it can be used
in other fields some of which may be far from the concept of zero-threshold carcinogenesis
(for example major industrial risk in the UK [HSE 1989] in the Netherlands [Kuijen 1988] or in
Switzerland [CFS 1991]).

These principles of management are not absolutely novel and they could have remained
principles of action without practical implications. In the case of radiological risk, the
principles could be implemented because the operators and authorities had at their disposal
conceptual and technical tools for their evaluation and management:

· the concept of effective dose, and more generally the concept of dose,
· the dose-response relationship,
· the definition of a standard man (weight of organs, rate of respiration etc),
· the publication of coefficients that allow the transformation of exposure to doses,
· standardisation of dosimetry,
· the means to carry out dosimetric investigations,
· the follow-up in real time of dosimetric evaluation,
· learning from experience,
· ALARA committees (As Low As Reasonably Achievable, an expression of the principle
of optimisation) within organisations.

Finally one should mention expert institutions that support public authorities, themselves
supported by the development of international expertise. In fact, a very well developed
national and international infrastructure has made available tools for the evaluation of
exposure as well as means for sharing information. The analysis of local practices was not

neglected in the development of recommendations. The most original aspect of radiological
protection is the balance between a principle of management, the management of an
acceptable risk and the practice of optimisation as well as the various tools made available
for its practical implementation [Hubert 1990]

Choice of subjects for discussion

It is difficult to look critically at the management of radiological risk without the subject being
clouded by discussions on the nuclear industry. The questions associated with radiation are
rarely regarded as such. The importance of the potential risk associated with the industry
explains this, but the situation does not reflect the dose levels of today. Radiological risk
management suffers from bias, stemming from public concern, government pressure and
even from research. For instance, epidemiological studies are planned around nuclear sites
but very few include people who have undergone diagnostic radiology (for example, the
detection of congenital dislocation of the hip in the new born), and very few studies include
therapeutic irradiation. The critical analysis that follows leaves aside the aspects related to
the nuclear industry and focuses on the radiological protection system. The system can be
criticised for both clumsiness and “an excess of subtlety” (cf. discussions in [CE 1996]). In
fact, an all-encompassing treatment of such diverse situations as incidents, work with
radioactive material or radon in local areas requires a great deal of intellectual agility. This
diversity cannot be reduced; the only option is to provide different types of management.

The principle of optimisation has an inherent inertia. For example, comparison with limits
can be made using brief but relevant calculations. Optimisation however requires
estimation of doses and risks with the highest possible precision: this requires great
expertise from operators and authorities. This investment in evaluation is the inevitable price
of greater efficacy, but the procedures should be simplified as soon as possible. In fact, the
theory of optimisation itself provides means for such regulation because the cost of obtaining
information must be taken into consideration.

There is a fine line between evaluation and management in the implementation of a

radiological protection system. The operation and the efficiency of the system rely, as
previously observed, on the close link between evaluation and management; the interested
parties should evaluate the situation and decide on the best option to take. A real-time
follow-up allows the strategy to be revised, if need be immediately. The conceptual tools are
also ambiguous. The dose-response relationship is described as a means of management
and not evaluation. The effective dose was also conceived as a means of management
even if it can be used to evaluate the efficacy of practices. It should be noted that the dose-
response relationship, with all its simplifications, may lead to action on the basis of criteria
which may themselves be less than obvious. For example, the use of a single risk
coefficient, irrespective of sex or age, ignores risk management policies that may distinguish
between them. Finally, while UNSCEAR is dedicated to evaluation, ICRP carries out both
roles. In fact, evaluation and management, in this context, are generally dealt with
separately unlike in the political arena. Evaluation carried out in the field is the
determination of dose; management is the application of pre-defined rules. The two
domains are quite separate even though ICRP deals with them both.

From a scientific perspective (for example assessment for an epidemiological study, analysis
of mammography) an evaluation necessitates relying on basic data and not on effective
dose or global risk coefficient. Similarly, decisions that do not rely on daily management
problems (distribution of iodine, criteria for the return to normal of polluted sites etc) there is
a separation between the analysis of risk and decision-making.

Finally, another originality of the system is the important role played by a non-governmental
organisation, ICRP, whose members are elected and which is not subject to any external
control. Some people consider this status to guarantee independence, others voice doubts

about the legitimacy of ICRP when it introduces public health regulation. However, there are
very few alternatives: intergovernmental organisations (European Commission), professional
associations, international bodies (UNSCEAR). It is not possible to say which is the ideal
situation but a certain equilibrium is achieved when different types of organisations actually
interact in the definition of modes of risk management.

The radiological protection system developed gradually over a period of more than a
century. Advancement in scientific knowledge went hand-in-hand with the development of
concepts and management regulations. Ionising radiation is treated as a ubiquitous zero-
threshold (or “inaccessible threshold”) carcinogenic agent. As usual in such cases,
Exposure Limit Values do not correspond to zero risk. In the case of ionising radiation, an
original idea was the definition of the ELV as a bordering the unacceptable, and the
insistence on the requirement that doses should be “as low as reasonably possible”. The
strength of the radiological protection system is that the means of carrying out such an
optimisation have been made available. This was achieved through co-operation, and by
the work of international organisations which became the focus of the main scientific

We can consider this system as a typical example of the “management of acceptable risk”,
however this is not a solution adaptable to all situations. The options selected may be
contested in themselves or because they cannot be adapted to other situations. In particular,
a full dosimetric survey is a very specific requirement and its equivalent is not always
available in all circumstances. Finally, the operation of this system has a very long history
from which we learned that a significant and sustained effort is needed to implement the
principles of risk management.

II. Discussion with the group of experts

Q: When was a consensus established on the shape of the dose-response function? Is
there a consensus today on all types of effect?
A: Some effects are no longer under debate. I refer, for example, to mental retardation of
children exposed in utero and to effects on growth. The key effects under discussion are
genetic effects and cancers.

The dose-response relationship between radiation and cancer was established on the basis
of human data obtained in the seventies. It allows for the quantification of risk. In terms of
management, the concepts of acceptability of risk and non-attainable limits were being
discussed in the seventies. Management started with deterministic limits: sufficiently low
doses to avoid deterministic effects. In the sixties, the exposure standard was still one tenth
of the dose at which skin erythema can be observed. In the fifties, other phenomena were
taken into consideration. It became evident that the standard did not ensure zero risk but
instead considered acceptability. ICRP developed all these concepts in 1958. Assessment
of population exposure, and the concept of collective dose was based on genetic effects,
that is effects on the descendants. It was considered immaterial whether there were few
people highly exposed (that is with very damaged genetic material) or more people less
exposed, hence the description of exposure in terms of collective dose. The limits were
determined such that the burden of exposure was acceptable to society. Later, the same
reasoning was applied to cancer. At the individual level, the probability of developing cancer
is never zero. An acceptable level had to be defined in the full knowledge that it does not
guarantee zero risk. The dose-response relationship was the basis of the determination of
acceptable risk. In 1972, the International Commission for Radiological Protection produced
the risk management system that we know today. This system is based on three principles.
Firstly, the limits are set in order to eliminate acute defects and maintain an acceptable level
of genetic effects and cancer. The second principle is that complying with the limit does not
guarantee acceptability in the case of zero-threshold effects. It is therefore necessary to

keep exposure as low as reasonably possible. The third principle is the principle of
justification, which states that one must balance the risks of exposure to the benefits deriving
from it. For example, it is not justified to dope jewels with neutrons so that they look more

Q: How does this last principle apply to people living in the neighbourhood of a nuclear
A: The question of justification applies to nuclear energy. Is it justified to have such a
practice, that is to have a human activity that creates exposure? The answer is eminently

Q: Can the statement that all levels of exposure carry a risk of genetic effect, and the
principle that the exposure must be as low as possible (ALARA) be dissociated? Is the
ALARA principle intrinsically linked to the consensus on the existence of a zero-threshold
effect? This point is important if one is to extrapolate this logic to the situation of mobile
A: I think that the definition of the word risk is ambiguous. In the case of ionising radiation,
researchers have absolutely no doubt that there is a risk. The concept of dose was invented
and introduced gradually in the case of ionising radiation. There is a physical quantity
expressed in J/kg corresponding to the energy deposited by ionising radiation. This quantity
is not used either in the evaluation or in the management. Another completely artificial
quantity, the effective dose, was invented. This takes into consideration the dose-response
relationship. It incorporates observations that neutrons are more efficient than gamma rays
which themselves are a little less efficient than X-rays. The system for evaluation of the
different types of radiation was developed from data from in vitro and animal studies. The
concept of dose equivalent for the whole body was introduced in order to be able to compare
exposure of the lungs, the thyroid or elsewhere in the body. The response of each of these
organs is different; the dose to each organ is weighted accordingly. The dose currently used
for ionising radiation is called the Sievert. It incorporates equivalence in risk of exposure to
different organs from different types of radiation.

The system has been stretched to the limit in the case of radon. In this situation, a
calculation of a dose is no longer the first step. Instead, scientists use a direct relationship
between the time-integrated concentration to which people are exposed and risk.
Consequently, the parameter "concentration-time" forms the basis of a dose that has
equivalence in risk. The unit of exposure to radon is the Bq/m3 which, in conjunction with
exposure time, gives the risk. The dose-response relationship from Hiroshima makes it
possible to express a risk in terms of milliSieverts equivalent. An actual physical dose does
not contribute to this process.

With respect to the dose-response relationship, I think we should distinguish between two
things. In fact, there are two different relationships. The first is obtained from modelling
epidemiological data and the second is used for prediction. These two relationships are
ultimately the same in terms of value but have very different status.

In the field of ionising radiation, there are three types of studies. Some of these studies
enable quantification when the doses and effects are well understood. The confidence
intervals on the risk coefficients are relatively small. It is therefore possible to quantify a
primary risk coefficient. Other studies say nothing at all. The third type of study is
somewhere in between. The best known studies in this category are those where it is not
possible to achieve quantification of the dose-response relationship because the doses are
not well-defined or because the confidence intervals are too high. The studies on thyroid
cancer after the Chernobyl accident fall within this category. Without any doubt, there has
been an increase in this type of cancer. However, scientists have not been able to quantify
the risk because the doses were not well defined. Some studies on workers in the nuclear

industry also fall within this category. The confidence intervals are so high that it is not
possible to progress the discussion on the dose-response relationship. There are therefore
a series of studies with an in-between status. They supply qualitative information but
nevertheless are very important.

The development of the dose-response relationship is based on three elements:

epidemiological studies, work by inference and extrapolation, and finally choice of and
decision on the measurement of the effect. Epidemiological work incorporates several
elements of interpretation and it is not always possible to compensate for parameters
incorporated into the data. For example, greater interest may be expressed in some
parameters such as age than others. Work by inference and extrapolation can be quite
sophisticated when dealing with human data. There is a body of work dedicated to the
transposition between different populations and between different types of radiation. The
dose-response relationship considered may be somewhat different from that fitting the data.
For ionising radiation, the factor for acute exposures is halved when applied to the
neighbourhood of nuclear sites or to workers. This is equivalent to a factor of two reduction
in dose rate. This effect is both endogenous and exogenous. Some relationships follows a
linear quadratic curve. In some animal experiments the response follows, more or less, the
dose rate. Finally the element of choice in determination of the effect or end point is very
important but is often neglected. For example, should we take as base-line a population of
men and women or a population of men from different parts of the world? Should the
variable be the loss of life expectancy or the probability of developing cancer in a lifetime ?
Depending on the choices made, the difference may be a factor of three or four.

Document 2 (p.7) [cf annexes] shows how the dose-response relationship has been
developed by different organisations. Institutions such as the Science Academy of the
United States, UNSCEAR, or ICRP developed a dose-response relationship on the basis of
epidemiological evidence, animal studies or theoretical predictions. The Academy of
Science of the United States indicates which hypotheses were discarded and points to the
mix of data from animal and human studies, and describes its decision-making process.
The strategies followed by the other institutions are not very different. ICRP and UNSCEAR
draw on data from Hiroshima and Nagasaki for the determination of coefficients. Other
studies are taken into consideration; the coefficients are modified on the basis of animal
studies and applied to a mixed population. The institutions could then conclude that there is
a 5% mortality associated with an exposure of 1 Sievert.

4.5.10 J P VAUTRIN
National Research and Safety Institute (Institut National de Recherche et de Securite

I Response to written questions

XIV- Are there any surveys carried out in professional circles in order to characterise the
exposure of workers from mobile phones and related equipment particularly base stations?

XV- Did the results of these surveys lead to the formulation of preventive regulations
intended to improve the safety of workers?

XVI- What are the technical and normative grounds on which your Institute bases its
preventive regulations, do you think these grounds are satisfactory?

INRS is managed by an egalitarian administrative council made up of 18 statutory members

representing employers (Medef) and employee syndicates (unions) (CFDT, CFTC, CGC,
CGT, CGT-FO). It is a non-profit association subject to state financial supervision, under the

auspices of the National Salaried Workers Health Insurance Fund (CNAMGS). It is an
essential link in the French system for the prevention of occupational risk. The headquarters
are in Paris but the Centre for Research is in Lorraine. The remit of this organisation
encompasses research, studies and information. It is also responsible for a general social
security regime run for the benefit of salaried workers and businesses. It is important to note
that the competence of INRS is unique.

The non-ionising radiation group of INRS is involved in diverse activities:

· For the past 20 years we have been carrying surveys on industrial sites in order to
characterise exposure.
· We carry out studies into remedial measures. At present, there are two ongoing
studies: the characterisation of fields emitted by resistance welders and the
improvement of shielding of high frequency presses.
· We participate in standardisation working groups.
· We respond to queries from industry and from health and safety advisors (health
and safety services within CRAM, occupational physicians, CHSCT).
· We publish articles and support the production of other information.
We do not carry out research ourselves. However, we do have the possibility of forming
partnerships with qualified research organisations. We have also established a network of
information and assistance for occupational physicians. Finally, and this is the most
interesting point as far as you are concerned, we are carrying out surveys of non-ionising
radiation in a network of sites across France, with the physical measurement centres of
CRAM (Regional Health Insurance Fund or Caisse Régionale d’Assurance Maladie). The
researchers in these centres are engineers and technicians in charge of the evaluation of
hazards from physical agents (acoustic, optical, thermal, vibration, ventilation etc). Some of
these centres have already acquired the necessary measurement equipment, others are in
the process of acquiring it. The whole operation has been planned. The participants from
the various centres have taken part in site surveys. As these measurements are very
specialised, we will initially take responsibility for the measurements. In fact, it is relatively
easy to make measurements close to microwave ovens or high frequency presses;
measurements in the vicinity of radars are however more demanding. In time, the
participating technicians will be able to carry out electric and magnetic field strength and
power density measurements in industrial sites in the proximity of most installations.
We have no intention of dealing with radiotelephones except for those intended for
occupational use. We will, though, be dealing with base stations because we have received
numerous queries about them. In my opinion base stations create two problems. The first
relates to the concerns of people who work in the neighbourhood of base stations. We were
able to reassure them that, in our experience, the field strength is on the order of 1 V/m at a
distance of about 10 – 15 metres. This is significantly below the limit. The second problem
relates to installation and maintenance personnel who need to work close to these base
stations and may be subjected to high fields. As far as we are concerned, this is not a very
high priority situation. There are potentially more hazardous uses of electromagnetic fields
in industry involving static as well as intermediate and high frequency fields. I do recall
measuring electric field strengths of 4000 V/m at 27 MHz around a high frequency press
used by a female operator! The situation now is much better and it is not often that we can
measure more than about 100 V/m which is somewhat higher than the standard (61.4 V/m).
The great majority of dielectric welding installations give rise to lower fields. There have
been problems with the electrolysis of aluminium; surveys were made and relatively high
fields were found, the health consequences resulting from this situation are not known.

The frequencies most often used in high frequency presses, for welding or PVC processing,
are mostly 27.12 MHz and 13.6 MHz. Five to six MHz is used for the polymerisation of
rubber and the drying of wood. In electrolysis, the electric field strength is low and the
magnetic field strength relatively high because high currents and low voltages (on the order
of 3 V) are used.
II Discussion with the Group of Experts
Q: Do you receive many queries from workers who use the mobile phones extensively?
A: We have received many queries about delivery personnel who use the telephone more
than average in their lorries or vans. There is perhaps a small problem to solve. The
solution is often to keep a distance.

Q: What makes you say that there is a problem?

A: "Problem" is not the right word. If the radiotelephone is very close to the skull it is
reasonable to ask questions about exposure time, but the people who contact us do not
complaint of thermal manifestations. They simply ask themselves questions. We often
receive queries about headaches in relation to radiofrequencies and electromagnetism in
general, however to my knowledge INRS has not received such queries about

Q: Has it been suggested to the delivery personnel to keep the antenna outside their
A: No, we are in the process of putting together typical responses to frequently asked
questions. This question will be included.

Q: Who has asked for measurement surveys around base stations?

A: The relevant sections of CRAM are not at present interested in this issue, except for
CRAM in Alsace who have personnel close to a base station. Most of the requests come
from occupational physicians, the CHSCT and the three operators (SFR, Bouygues and

Q: What measurement procedures do you use?

A: A procedure that we developed, although there is room for improvement. We should
establish more appropriate protocols as we have done in the case of high frequency presses
and induction heating. At present, we rely on our experience with mobile telephony. We
have to deal with very specific problems that must be tackled in the same manner by

Q: Are you aware of what the ANFR has done in this field?
A: Absolutely. The ANFR is perfectly aware of the possible effects of electromagnetic fields
on man. It shows its interest by participating in various meetings and having internal
discussions and deliberations on the subject.

There are other organisations working in this field. We cannot overlook the
Recommendation of 12 July 1999 of the Council of the European Union, this document is a
starting point even thought it was not designed for implementation in occupational situations
where the standards must be less stringent. For occupational exposure, one of the
references that we have used right from the very beginning is that of ICNIRP. At the
European level, the CENELEC standards 50166/1 and 2 were published and adopted in
France by the Standards committee C 18-600/C 18-610. We follow these two references.

The Council Recommendation stipulates 41.25 V/m for 900 MHz and 58 V/m for 1800 MHz.
For occupational exposure, the application of the limit values obtained from standard C 18-
610 gives, respectively, 92.1 and 130 V/m for the same frequencies.

Q: What are the rules of good practice aimed at limiting exposure during the installation and
maintenance of base stations?
A: We base our practices on ICNIRP. The first solution is distance. We do not allow people
to move freely on a site unless it is absolutely necessary. The second solution is signage of
two or three zones:

· a green zone in which one can stay permanently;

· an orange zone where care is needed;
· a red zone where it may be necessary to almost shut down the sources when
workers are there.

Q: Did you say "almost"?

A: In mobile telephony the fields are significantly lower than those observed, for example,
near the sources on the Eiffel Tower. However, I would tend to recommend the shutdown of
telephony systems when workers are in the red zone.

Q: Can you confirm that most maintenance personnel ask for shutdown of the antenna
during servicing?
A: I do not have sufficient information to respond.

Q: At present, there is no official regulation about this. Would you recommend the shutting-
down of the base station during maintenance depending on the power level?
A: It is difficult to respond to this type of question. A worker may be assessing the
radiation pattern of an antenna and hence be in the vicinity of the source, and the source
must be functional. Therefore one must consider the nature of the work. Together with
René de Sèze we will be thinking about this problem and analysing a range of likely
scenarios. I cannot elaborate further today.

Q: One should also take into consideration the fact that maintenance personnel may work
on a number of base stations in one day.
A: Of course. We would recommend to limit the exposure time, and to adhere to the six
minute average. I could also add that personal protection is a possible solution but I do not
think that it is appropriate for base stations. However, it is a solution I favour for high power
sources. It is not recommended for mobile telephony as it may end up generating
unnecessary concern.

Finally, I would like to draw attention to two points: firstly the problem of people with
implants. This problem is not so important today but one has to be careful about workers
with cardiac pacemakers. Secondly, we have realised that different operators do not follow
the same safety regulations. This creates a problem of unfair competition between
companies. The establishment of national or European regulation would resolve the
situation as well as ensure better safety for maintenance workers.

4.5.11 Michèle RIVASI

Member of Parliament (Deputy) for Drôme36
I Response to written questions
1. What do elected members expect from the scientific community in order that they can
manage risk related to mobile telephony in the context of current uncertainty on the reality
and magnitude of some biological and health consequences of exposure to RF EMF?

the comments of Michèle Rivasi on the re-transcription of the interview had not been received by
the group of experts at the date of the publication of this report.

2. Do you consider that the scientific data available today on the biological and health
effects of RF EMF justify the strategy for siting base stations and the use of mobile
We receive a lot of queries from citizens about mobile phones and their associated base
stations. I receive a considerable amount of mail from people who ponder on the dangers
and ask themselves if they should accept the siting of a base station antenna on their
rooftop. I have even received a delegation of co-proprietors [owner-occupiers of buildings in
multiple occupancy] and would like to refer to an article that appeared this week in the
Journal du Dimanche reporting that in Paris a rooftop can be rented out for between 20,000
and 110,000 francs. The operators use teams to identify available rooftops.

When we organised a colloquium at the National Assembly, I called the Directorate General
of Health and I was surprised to find that there were no regulations. I think that regulation
must cover a number of points.

There are many sources. I think that it is important to carry out measurements on these
sources. One must therefore decide on who will carry out the measurements and who will
monitor them. It could be government organisations if they are competent, identifiable and if
they publish their results. It is important to make a decision on the question of
measurements. Moreover, it would be desirable if the measurement procedures were the
same from one organisation to the other. In fact, there is actually a disparity in the units in
which they are expressed.

A study group has come up with a proposal for a law that covers several issues. I do not
want to enter into the specifics of it. My personal position is to say that fields have biological
effects and we do not have enough information to determine whether they also have an
effect on health. The biological research may well find genes that are radiosensitive, or
sensitive to magnetic fields. In my opinion concepts of threshold will develop in the next few
years. In the field of radioactivity a radiosensitive gene is practically identified. A threshold
of a few milliSieverts does not mean much to people who are radiosensitive. With respect to
antennas, in my opinion one should adopt an attitude of prudence as did the English, the
Austrians and others; the installation of sources in the proximity of public buildings should be
avoided. The proposed law will try to prevent the installation of base station antennas close
to public buildings and, a fortiori, schools.

There are two positions among the Deputies regarding doses. The first is to say that field
strengths must not exceed a certain limit value set by decree. The second position is to fix a
maximum field strength of 1 V/m for a base station antenna. One must also be aware of the
possibility of resonance in metallic infrastructures that may deflect beams.

A third problem is that of public information. When the sources are in large open spaces I
think that the minimum required is to inform the public that they are going through an
electromagnetic field.

The last issue relates to the protection of workers. When workers have to pass close to
antennas, the minimum requirement is that some means of assessment should be available.
It is possible to establish controlled areas defined in relation to a standard source. It is also
possible to equip workers with means of making measurements so that they themselves can
evaluate the danger in the place where they work. In case of danger, the worker can then
ask for the source to be shut down. It is also possible to combine the two solutions.

I would also like to inform the General Directorate of Health that there is an urgent need for
regulations. When there are no regulations, the law of money prevails.

II. Discussion with the group of experts
Q: You have concentrated in your discussion on base stations and their environment.
However, the field to which a person is exposed fairly close to a base station is lower than
that received by a user of a mobile phone. Have you extrapolated your argument to the
telephone itself?
A: With respect to mobile phones, I would wish to pass responsibility to parents, and to
provide information on the purchase of telephones for children. Advertising campaigns
urging parents to buy mobile phones for their children of 4 or 5 years old to keep them in
touch with their homes and their parents are quite absurd. I do not wish to forbid this type of
practice but it seems to me that we should say to parents that the purchase of mobile
phones for their children could carry risks. Ultimately the parent must have a choice. We
must at least provide information. Of course I am quite aware that the risk is not higher
close to sources (base stations). However there should be some controls on emissions
when a source is being installed.

Q: Effectively, we must avoid directing the beams towards public buildings. What do you
recommend with respect to radiofrequency sources other than mobile telephony? The Eiffel
tower has the equivalent of 30,0000 base stations. Here (ed: in the building of the State
Health Secretariat), we are exposed to fields of the order of 3 V/m from sources on the Eiffel
tower. You mentioned a value of 1 V/m for base stations. Could you explain the criteria that
determine this value? Does it depend on the frequency of the waves?
A: I have studied the existing foreign legislation, and proposals from scientific bodies. This
value emerged from that study. The Swiss government for example is close to 3 V/m (ed:
the Federal Ordinance has fixed a value of 4 V/m at 900 MHz and 6 V/m at 1800 MHz). I
made this proposal because I do not trust decrees. When you are elected, you make a law
on the basis of certain logic. If thereafter the values are fixed by order or decree it may fall
outside the logic of the elected members. We should avoid having an imposed decree or
order that allows high values when we wish to apply the Precautionary Principle.

Q: If, tomorrow, new scientific facts indicate that this value should be divided by two would
you ask the Parliament to reconsider?
A: It would be possible to have a new vote. There is a system that allows certain
modification of articles of law by vote.

Q: A standard is always related to the state of knowledge at the time it was developed. Do
you think that the Parliament can fix limit values for all the regulations that protect health and
can follow them up regularly?
A: I agree with you in principle. But even if certain levels are fixed by law, the levels can be
decreased by regulation. In radiological protection, there is a need to set a limit value of 1
mSv. Actually there is a discussion going on as to whether this value should be set in law
or in a decree. We would prefer this value to be set in the general law on radiological
protection. In fact decrees are not subject to discussion.

Q: You have invoked the Precautionary Principle with respect to base stations and the
definition of a limit value. Do you really think that the Precautionary Principle, in the strict
sense of the words, as defined in national and international interpretations, applies to the
situation where there is no knowledge of serious risk?
A: The level of 1 V/m came from different countries: there were also strong variations
between countries. However, studies show that some effects may occur in sensitive
populations. Adopting this value is therefore a compromise. If it is found later that effects
persist even at this level, we would be in a situation of acceptable risk.

Q: The idea of an acceptable risk requires that the risk be quantifiable. However at the
moment, in this field of interest, the risk is not quantified. We cannot state that we will save
lives by reducing the level from 40 V/m to 1 V/m as you suggest. We do not even know if it
will save any life at all.
A: No, but it can eliminate public concern. Studies on workers close to base station
antennas have taught us something. This is a precautionary approach.

Q: We are going to say to people that it is not good to be in a field in excess of 1 V/m.
However when we are passengers in a train or a bus and sitting right next to a person
constantly on the telephone, I could say that this person is exposing me to a field in excess
of 1 V/m during a call and ask the person to leave the vehicle to make the call. The law
must be consistent.
A: The argument would be the same for 5V or 10 V, etc.

Technical note: No, because when we are near a person on a mobile phone we can be subjected
easily to 1 V/m. However, under the same conditions we would probably not be submitted to fields of
5 or 10 V/m and certainly not to 40 V/m corresponding to the European recommendations.

A: What do you know at 40 V/m? You know very well that there have been many criticisms.
We found out that most of the studies were commissioned by mobile telephone operators.
You know that there are questions being asked nevertheless you adhere completely to this

Q: I would like to come back to the development of the standard of 41 V/m. There have
been numerous studies at radiofrequencies since the sixties and seventies. First, scientists
looked at high-intensity acute harmful effects. Then they looked at the possibility of long-
term effects on the development of cancer. No effect on cancer was found. The most
sensitive indicator is a change in behaviour. Researchers have also found some evidence
of effects on attention and learning in primates and baboons. These effects are induced at
levels ten times higher than 40 V/m. A safety factor has therefore been incorporated in
order to avoid the possibility of producing even changes in behaviour.

One has to be consistent in regulation. Look at the problem of glycol ethers today. The use
of certain ethers is forbidden in cosmetic and household products even though they were
only present in very small quantities in these products. However, we tolerate the exposure
of workers to higher levels. This is not consistent. Irrespective of the agreed value we must
take into consideration regulations for different exposure situations. A value of 1 V/m, given
the current state of affairs, would mean that mobile phones and other sources of radiation
may cause problems.

A: One can discuss levels. As you know a standard has scientific, economic and social
perspectives. There is now an urgency to act. I favour regulation of these antennas. We
must not let things happen anywhere anyhow. The British report is very interesting and
raises many questions. Some people are more sensitive than others. One must therefore
adopt an attitude of prudence. The British thought this way. We must inform the public and
identify areas where an electromagnetic field is present. This phenomenon must not be
trivialised. If we are not sufficiently rigorous at this stage, there is a risk that electromagnetic
pollution will get out of control.

Q: Have you any suggestions on how to conduct truly independent scientific research in
France? This is a general question. The National Assembly must make available means for
independent research in areas where the interest of the economy and the interest of health
may not coincide. One solution, recommended by the English, would be for industry not to
give funding directly to the laboratory but to an intermediate authority. The Parliament has a
very significant role to play in this matter. I think that Parliament should work on such
matters rather than on what limit value to adopt for electromagnetic fields.

A: This is a very difficult question. There is a need for financial independence. The
contracts given to university laboratories are often financed by the operators. I would favour
the solution you mentioned, as long as the money is subsequently given to several

4.5.12 Mrs. Gaëlle PATETTE and Mrs. Carole MATRICON

UFC - "Which to choose?" (UFC-Que choisir?)*
*UFC: Federal Consumers Association (Union Federale des Consommateurs)

I Response to written questions

XVII What are the main sources of information that you have at your disposal on the
possible risks of health effects in relation to RF EMF ? Do you trust the authors of recent
international reports? How do you think we can improve public information?

XVIII What are your recommendations about the use of mobile telephones and the siting of
base stations, and what is their basis?

The main sources of information at our disposal on the possible risks to health from mobile
telephony are the publications of WHO, peer-reviewed scientific papers, publications
intended for the general public and studies carried out by consumers associations in other
European countries. We are linked to these associations via the framework of ICRT, a
group of consumer associations that carry out tests in common. In this context, an English
association has carried out tests on hands-free kits. [Note: the French original does not
distinguish clearly between "hands-free" vehicle-mounted sets, and "pedestrian" hands-free kits in this
context.] The results of these tests were published on the Internet site of this association.
We were very interested in this study because it shows that there are problems with these

Technical note: The protocol used in this test is probably not good. The hands-free kit is 99%
effective in decreasing exposure. The authors of the English study should be challenged by
international experts in the measurement of fields from mobile phones. The scientific consensus
today is that these kits decrease significantly the level absorbed in the head, but there are sometime
significant differences in their efficacies: some are defective in their design.

UFC-Que choisir? will carry out tests on mobile phones in the near future. These tests will
include measurements of radiation. You asked us if we have confidence in the authors of
international reports. It appears to us that these reports are rather contradictory. It is
difficult to determine whether there is a hazard or not. An Australian study showing that
there was a hazard was reported in the press. We have also detected some conflict
between willingness to carry out serious studies and the fact that research groups were
financed by operators and manufacturers. This seems to us to be dangerous. Moreover,
the studies lack universality at present: there are sources other than radiation from mobile
phones. It would be interesting to carry out studies that take into consideration the overall
exposure of consumers. Finally we are eagerly waiting the results of the epidemiological
study just started by WHO.

II Discussion with the group of experts

Q: I am somewhat surprised by your arguments. I quite understand that you may be lost
reading the results of individual studies. We are lost ourselves. It appears to me however
that international reports are significantly more consistent. The international reports provide
a great deal of information after review and discussion, and also conclusions. They are not
at the same level as individual studies or articles in the press. The conclusion of all these
reports is that there is no evident risk to health.

A: Articles in the press place individual studies and international reports at the same level.
But the scientists in our organisation are perfectly aware that we cannot equate these two
sources of information. Nevertheless there remains the problem of consumer information.
This is a task for public authorities: information is received similarly by the consumer
whether the source is an international report or an individual study. People's perception of
hazard probably overestimates the actual hazard. If the consumer uses a telephone he will
accept a certain measure of risk. By contrast, he will accept any risk from a base station
antenna installed in his neighbourhood.
Q: You wish to see information on the state of scientific knowledge and on uncertainties.
What other forms of information should we place at the disposal of the public and
A: We stress the need for manufacturers and operators of mobile phones to inform
consumers about the level of radiation from the mobile phones by clearly marking them, as
well as providing the information within the documentation (user's manual, contract, etc).
More information would lead to a better understanding of the risks by the consumer. I think
that manufacturers have already taken an initiative in this respect37. I think we must also
consider preventive measures. In particular, we should avoid siting antennas close to
establishments for children.

Q: Should it not be the case that schools must not be in the main beam, because surveys
show that the greater the proximity the weaker the radiation at ground level and in the
building under the beam? Do you think that it is possible to explain that the exposure is
practically nil for people in the building on which the antenna is sited and that the exposure
is somewhat higher for people in the beam at 200 to 300 metres distant from the source?
For example, for schools in Paris the radiation from FM broadcasts is higher than that from
the mobile telephone base stations.
A: Absolutely. It is only a question of information. I think that the public should receive full
and clear information about electromagnetic fields so that they are in a position to evaluate
the risks. I think that a legal proposal to limit the use of mobile phones and the installation of
antennas must be drawn up.

I would also like to comment on the use of phones by children. It appears to me that nothing
has been done about this. The operators have phone contract targeted at children. A study
in Sofres, carried out in 1999, showed that 4% of children aged 8-10 and 14% of children
aged 11-13 possess a mobile phone. Since then, the numbers have probably increased.
Actually the operators are no longer actively recruiting more subscribers. If these phones
are potentially more dangerous for children, something has to be done to make the operator
refrain from these types of contracts.

Q: You mentioned in the document that you submitted to us the problem of using a mobile
phone while driving. Hands-free [vehicle-mounted] kits create a real problem in this respect.
In fact, currently, the Highway Code does not penalise the users of these devices. However,
scientific studies have shown that the risk of serious accident is 4 to 6 times higher when the
driver is using a hands-free kit. This risk is comparable to that of a driver with a 0.8g/l of
blood alcohol. Do you have any proposal about this?
A: Yes, we would like to see information for consumers about this issue, with leaflets
discouraging their use and also the putting in place of effective controls on this activity. I
would like to say that, so far, we have not received any complaints from consumers about
symptoms that could be linked to mobile phones. The last point that I would like to
emphasise is that the contract system encourages people to use the phone for the whole of
the prepaid period and therefore at times to use the telephone more than necessary.

Technical note: this action was instigated independently by industry in response to
recommendations by public bodies in different countries. A protocol is currently being finalised.

From a technical viewpoint, we would like to see a decrease in the level of radiation. This
level must be determined by an independent organisation, measured by a credible
procedure and indicated on the device. For base stations, France applies European
standards. Other countries are much more restrictive. This raises questions. Following the
Precautionary Principle, it seems consistent to adopt a standard as low as possible in order
to prevent possible risks.

Q: What do you think is the best way to inform the consumer about the device that he
purchases? Is it better to express the level of radiation as a value or as a percentage of the
recommended standard? In fact the actual value may be difficult to comprehend because
the unit of measure is totally incomprehensible to the lay person.
A: We have not yet taken a position on this subject.

Q: To my knowledge, you have not published in your magazine a long article on the
possible effects of mobile phones on health. Is there an article in preparation?
A: There will probably be an article dedicated to this subject when we report on the
forthcoming tests of mobile phones. A draft will probably be ready in the next few months. I
would like to state that the legal section of our association is concerned about publicity for
protective devices, but we are not thinking at this time of dedicating any space to these
Technical remark: These devices are either inefficient or counterproductive. When they actually
decrease the radiated power of the telephone, the telephone will automatically increase its power to
the maximum possible hence neutralising the benefit to the user. The net result of the development
of these devices will be the necessity of increasing the number of base stations in order to maintain
the function of networks.

4.5.13 Professor Jacques FOURCADE

President of the Consultative Committee for the Protection of People in Biomedical
Research in Nîmes. (Comite Consultatif de Protection des Personnes dans la Recherche
Biomedicale – CCPPRB)

Q: Does biomedical research about the risk to health from the use of mobile phone
generate specific constraints, and what are these constraints?
A: The government has organised biomedical research in France in accordance with the
Huriet-Sérusclat law of 20 December 1988.

The following text reports, in the light of these regulations, on the operation and decision-
taking criteria of one CCPPRB, reached through practice and experience.

I General conditions and research

Certain rules apply to all cases whether or not the research benefits directly the subject
taking part in it.

Necessity of developing a protocol in accordance with scientific rules

While it is not explicitly within the remit of a CCPPRB to verify the scientific aspects of a
project, it falls de facto within its role and even constitutes a prerequisite of its activity. In
fact, the methodology of a study is of great importance, in as much as a badly thought-out
protocol that may prove impossible to implement or give biased results, would have no
benefit at all. Even if the procedure carried no risk at all, a badly thought-out protocol would
produce an unfavourable risk/benefit ratio (because, as a minimum, there would be inherent
constraints on the subjects in any type of research) and would therefore be unjustifiable.

In this respect, there can be no prospective scientific research without a previously thought
out protocol. It should incorporate the following:

· a statement of the problem;
· the current state of knowledge;
· justification of the study in terms of scientific advances and/or returns in terms of health;
· clearly defined objectives and methodologies.

The following must be determined a priori:

· the duration of the testing period;

· the number of subjects to include;
· inclusion and exclusion criteria ;
· surveillance and exit criteria and a “rescue” procedure for excluded subjects;
· the duration of the study;
· analysis during the study to determine whether it is actually viable to follow the protocol;
· criteria for the analysis of results.

The practical feasibility of the protocol must also be demonstrated. It should include the
possibility of recruiting subjects for the study by the investigator. This should be realistic and
comparable to the number of subjects expected to be included within a certain period of

Necessity of ensuring the protection of subjects

This is the main mission of the CCPPRB. The following should be ensured:

· that the risk from the proposed research remains proportional to the predicted benefit
(the law defines more precisely “out of proportion”). These risks are not necessarily
zero, but they must not be higher than those normally taken in the same pathological
situation; the additional risk must be compensated by the potential benefit expected from
the agent being tested and by closer observation. Therefore when research has an a
priori potential benefit for the subject, a relatively high risk can be acceptable in the case
of a serious pathology that has not known cure. Similarly we can refuse an even modest
risk for a benign pathology that can be cured with existing means.

· the competence of investigators in the subject (hence the need for a Curriculum Vitae);

· the acceptable demonstration of a requirement for the research, on the basis of pre-
clinical trials (experimental animal, tissue tests, toxicological studies etc…) taking into
account the development of research capability.

The rules for testing a device are not fundamentally different from those applicable to a drug
whether the device is ultimately destined for biomedical research or for the general public. If
it is a device already in circulation, the standard to which it must ultimately adhere must be
well established (European standards for example). If the standard has not been developed,
the situation is that of a prototype, which is similar to that of a drug which has not been
licensed (AMM). In this case, guarantees must be given from the manufacturer of the device
or its various parts, for example, a certificate of conformity in relation to the protection from
electric power or from radiation. From their side the researchers must also endeavour to
use the device according to the recommendation of the manufacturer or to clearly indicate
that they intend to do otherwise (wavelength, power, duration of exposure etc) in the
framework of a non-conventional utilisation, which is a situation comparable to that of
research for a new usage licence for an established drug.

1 The need to guarantee the ethics of the project
The rules that apply to clinical research are not fundamentally different from “good clinical
practice” or the “quality care” that apply generally to all medical practice; research
constitutes a specific case but not a new medical practice.

i) Right of the patient for information and to give consent

The subjects of a study have a right to the most accurate and relevant information possible,
not necessarily complete but certainly honest, not defective by deception or omission with
respect to the purpose of the research, the conditions and the risk that may be expected to
be encountered. A recent change in jurisprudence, applicable to all actions that may
interfere with the integrity of a subject, requires doctors to divulge not only the most common
side effects, but also the rare and even the most exceptional side effects and in particular
the most serious as long as they had knowledge of them through the scientific literature.
This rule appears to me to be disproportionate and very constraining for research as well as
in normal circumstances; it may well prove to be very dissuasive towards the subjects of
research. This rule is based on prudence and whilst it can be accommodated in clinical and
diagnostic situations, it must be strictly adhered to in a research protocol particularly if the
research has no immediate therapeutic benefit, that is, where the ratio benefit/risk is not

The information must be part of a written text that is sufficiently explicit and does not exclude
the right of the subjects to solicit further questions in order that they receive appropriate
responses. Finally the subjects must be given sufficient time to reflect before they are asked
to give their consent in writing.

ii) The need to validate the aim of the research

One can imagine a research protocol that satisfies all the scientific criteria, ensures the
safety of the subjects and has their consent but whose aim does not result in the slightest
immediate benefit (studies with direct individual benefit) or long-term benefit (studies without
direct individual benefit) in the field of health. This is why article L.209.2 of the law stipulates
that biomedical research must aim at including not just biomedical knowledge but also the
condition of the human being. This is not a superfluous requirement; the risk of futile
research is quite real.

iii) The need to respect the dignity of the person

This condition follows from the rules previously specified, and goes further. Recollection of
Nazi experiments and other practices during the war show that scientific justification of an
experiment (such as researching the conditions aiming at improving the survival rate of
soldiers in the field) or the consent obtained from people deprived of their liberty, are not
sufficient to allow research to go ahead.

2 Procedure to follow when submitting a research project

Compliance with administrative rules
These derive from the law and its decrees:
· identification of the principal and secondary investigators;
· identification of the promoter of the study;
· knowledge of the law of appeal to the DRASS (Regional Director for Health and Social
· subscription to an insurance contract specific to the planned research;
· agreement of the head of department (in case of research within a hospital where the
principal investigator is not head of department), in order to ensure that the resources
needed for the research will be available;

· agreement of the Director of the health establishment who may also deal with the
allocation of resources. This agreement is a de facto requirement of the contract
between the promoter and the medical centre;
· submission by the promoter of the protocol to the Ministry of Health. The file must
include the advice of the CCPPRB.

Responsibility rests with the investigator, not the promoter, and can be questioned by the
CCPPRB. In fact, this should not impede good collaboration between the three parties.

In fact, monitoring of the execution of the protocol is not within the power of the CCPPRB
unless it had voiced reservations at the onset and reserved the right to follow up the
research. Of course, the investigator must communicate to the CCPPRB any undesirable
effect during the course of the research and to propose an amendment to the protocol. The
Ministry, who must be alerted by the promoter, may at any time exercise its power to stop
the research.

3 Special case of research without direct individual benefit

This is defined by the absence of immediate benefit to the subjects but, potentially, an
ultimate benefit to others from the information gained.

It is sometime difficult to classify the protocol as to whether or not there is an immediate

benefit. It is useful to ask the investigator to write in a short paragraph the criteria that made
him classify the study in one group or the other.

This type of research has, as we know, additional constraints;

· there is an obligation to carry out a study in a place agreed by the Ministry of Health; this
agreement is given to a certain person carrying out a certain type of research (physiological,
pharmacological, etc..) there may be some difficulty in realising “field” studies;

· submission of the volunteers to an initial medical examination; we are not dealing here
only with healthy volunteers, some studies without benefit may include sick volunteers who
may be happy to take part on condition that the study may potentially ultimately help other
similarly sick people;

· a special insurance on the basis of reversal of proof: the demonstration of the absence of
error is the responsibility of the investigator;

· registration of the volunteers in a file at the Ministry;

· participation of the volunteers in one study at a time, and indication of a period of

exclusion after one research project;

· agreeing on a reasonable indemnity to compensate for the inconvenience.

It is important to recall the special responsibility of the investigator given the absence of any
benefit to the subject, and therefore a necessarily disadvantageous benefit/risk ratio. This is
a situation that requires an absolute guarantee of safety (notwithstanding imponderables)
based on the hypothesis and the predictions of the protocol. An innocuous status requires
that the means of observation and any precautions implemented must be clearly indicated in
the protocol. It also requires that the investigator follow the protocol. The safety of the
subjects is also assessed from the pre-clinical file on the product to be tested. This file will
be examined exhaustively by the CCPPRB - hence the need for scientific expertise within
the committee.

One can therefore imagine a piece of research being carried out in a number of steps to
ensure prudence. For example, in order to conduct research on the possible effects of
mobile phones on the brain it is important to first carry out research on a population without
known brain pathology before testing subjects having a defined brain pathology such as
epilepsy or migraine conditions.

For certain category of subjects, the law has additional constraints for studies without direct
benefit. These categories include people at particular risk (eg pregnant women) and
situations where it is not practical to obtain consent (emergency cases, minors, wards of
court). These studies can only be carried out with these subjects if they will ultimately benefit
people in the same category, as well as being impossible to carry out within a normal

4 Research outside the Huriet law

A review of biomedical research shows that some studies do not submit their protocol to a
CCPPRB and therefore to the Ministry. These studies are not from the pharmaceutical
industry which in “difficult” cases probably prefers to carry out the so-called preliminary
protocols outside the country. Most appeared to come from “small” hospital investigations
who choose not to submit their protocols in order to be free of the constraints of the Huriet
law and in particular the most costly part, which is the insurance.

There are also misconceptions about the Huriet law even though it has now been in
existence for about ten years, such that some investigators may think, in good faith, that the
law does not apply to their research particularly when it deals with the evaluation of new
therapeutic responses of medicines that have previously been evaluated and licensed. The
absence of an industrial promoter aids such oversights.
There are however cases where the investigator has a good reason to hesitate before
submitting a research to the CCPPRB. Confusion is encouraged by the nature of the law
itself: it is not designed to encompass all research but rather to ensure protection of the
subjects of research. There is debate about an interpretation by the judiciary that research
that carries no risks (for example a non-invasive study) falls outside the Huriet law. The
Ministry itself leans toward this interpretation. It has a guide to investigators in which it gives
examples of biomedical research where the Huriet law does not apply. This indicates that in
ambiguous cases the choice (submission or not to the CCPPRB ) is definitely the
responsibility of the investigator, the courts being able to decide in a complex cases where
there is an appeal.
Borderline cases relating to devices include: tests relating to ECGs, ultrasound, MRI without
injection, hospital beds and wheelchairs. Certainly these devices are not invasive in the
strict sense of the term. One must not forget that hazards do not necessarily result from the
device being tested but also from the supplementary constraints required by the protocol
including collateral risks from, for example, the journey of the subject to be in a certain place
at a certain time. Not all biomedical research falls within the Huriet law but the distinction is
sometimes difficult and the problems of illegal research encourage researchers to be on the
safe side and not to operate outside the law.
We have encountered two phenomena. Sometimes investigators asked the CCPPRB to
determine whether their project falls within the Huriet law. Alternatively, some investigators
submit to the CCPPRB research projects that clearly do not fall within the Huriet law. While
the position of the different CCPPRBs are not unanimous on this point, our CCPPRB has
agreed that it is not within its remit to confirm or deny the requirement that a certain protocol
be submitted to it: this is the exclusive prerogative of the investigator who may be advised by
a member of the CCPPRB. Ultimately, the decision remains with the investigator and not the

However, we think that the CCPPRB is not qualified to refuse to examine a protocol
submitted to it as long as it falls within the area of biomedical research, even if it appears
that a submission to the CCPPRB should not have been made; to do otherwise would be to
give the CCPPRB a great responsibility that does not fall within its legal remit.

4.5.14 George CARLO

reported from a telephone conference with most of the group of experts on 23rd November
During the telephone conference, Dr Carlo proposed that the group of experts should use
the summary of his book to express his exact and current point of view on the subject matter
(with the kind permission of Philip Turner, of Caroll & Graf Publishers Inc., editor of the
book). This did not give the group of experts an opportunity to discuss, in his presence, the
author's hypothesis and statements. The rating and interpretation given here of scientific
work (also discussed elsewhere) are not shared by the group of experts.
The text that follows is an extract of the translation from pages 243 to 248 from the
book written by G. Carlo and M. Schram (Cell Phones. Invisible hazards in the
wireless age, 2001, Caroll & Graf Publishers Inc., New York)
Each of the alarming results of the studies on radiation from mobile phones are essential
pieces of information that fall into place in the puzzle relating to cancer. Some of the earlier
studies, that were in themselves non-conclusive or appeared difficult to interpret, now fit
equally well into the puzzle. These studies clarify the previously diffuse picture of cancer
and health risk. Here are the elements of scientific information that fit into the overall cancer
· Studies on human blood. These studies show a genetic effect in the form of
micronuclei in blood cells exposed to RF radiation. They demonstrate the biological
plausibility of the development of tumours following an exposure to RF radiation. These
studies consistently show chromosomic alterations in blood cells.
· Changes in the blood brain barrier. These studies integrate well in the whole picture
of cancer providing a two step explanation of the mechanism by which radiation from
mobile phones can cause cancer:
Step 1: A degradation of the blood brain barrier would allow chemical carcinogens
present in the blood (for example, by-products of tobacco, pesticide or atmospheric
pollution) to reach the brain and to affect sensitive tissues that were previously
Step 2: While different studies have shown that RF on its own could not affect
DNA, recent studies suggest that the repair mechanisms of brain cells could be
affected. Consequently, carcinogenesis induced by toxic chemicals could
propagate. This is an additional argument in favour of the biological plausibility of
the implication of RF in the development of brain cancer.
· Studies of cancer in users of mobile phones. Four studies were carried out among
users of mobile phones; the study of mortality by K Rothman and two studies by J
Muscat on brain cancer and acoustic neuroma, the study of Hardell on brain tumours.
These four studies, carried out by different investigators using different methodologies,
all give arguments in favour of an increased risk of tumours from the use of mobile
phones. This is the principle of consistency of results.
These four studies have shown an increase in the risk of developing a brain tumour. The
study of Muscat has shown a doubling of the risk of neuroepithelial tumours (and this
result was statistically significant). The study of Hardell has shown a doubling of the risk
of having cancer on the side of the head where the mobile phone is normally used (this
result was also statistically significant). The study of Rothman has shown that the users

of hand-held mobile phones have a risk that is twice as higher as that of users of car
phones (this result was not statistically significant). The study of Muscat on acoustic
neuroma indicates that the user of a mobile phone has a cumulative risk of 50% (this
result was significant only when the duration of use was taken into consideration). In
three epidemiological studies (those of Muscat and Hardell), the risk of tumour was
increased in the region of the brain close to where the mobile is held; this is an argument
in favour of a dose-response relationship.
Therefore the results of all of these studies are consistent.
The pieces of the puzzle fit well together and draw a picture that researchers, the
administrators and users can see for themselves. The picture is not complete and some
pieces of the puzzles are still missing. Enough is in place to realise that there is
sufficient reason to be concerned for the health of users of mobile phones. The radiation
emitted by the antenna can cause cancer and other health problems. This risk affects
hundreds of millions of people all over the world. One should use this information to
derive means of protection, and in particular the protection of small children whose skulls
are developing and who may be more vulnerable to the risk from radiation.

4.5.15 Marc SÉGUINOT

Report of the meeting between the group of experts and Dr Séguinot (EU-DG SANCO)
Paris, 14/12/2000
Note: Due to conflicting commitments, this meeting did not take place in accordance with
the procedures of other interviews with the group of experts. As it was not recorded, the
summary of the main points discussed is reported here.
The European Commission would like Member States to adopt as soon as possible the
framework of restrictions proposed by the Council Recommendation of 12 July 1999, by
translating them into national law, thus allaying public concern in relation to health on the
basis of internationally recognised scientific findings. For example, Germany and Sweden
have since adopted a regulation based on the exposure limits of the recommendation, the
Netherlands is soon to adopt the same scientific approach. Italy has chosen to set exposure
values that are significantly lower on the basis of political considerations described as
precautionary. Austria took a similar approach and adopted in certain of its provinces
ordinances imposing very strict levels of exposure of people to electromagnetic fields. The
Belgium government has just adopted intermediate values. The British government, in
response to the Stewart report, has accepted the values in the Recommendation without
however issuing regulatory procedures. However, the British government has actually
developed a programme of "in situ" assessment of the environmental exposure of the
population. It should also be noted that the mobile phone operators in Britain have decided
to comply voluntarily to the limits of exposure in the Recommendation. In early 2001, the
Commission will ask each of the member states to prepare a report on the implementation of
the Recommendation and will issue them with a detailed questionnaire; the responses will
be due in April 2001.
The Commission has requested its scientific steering committee to ascertain that the
scientific bases of the 1999 Recommendation are still valid. The committee will publish its
advice in July 2001.
The development of compliance standards for devices that fall under the umbrella of the Low
Voltage and Radio Telecommunication Directives are under development within CENELEC.
These standards will eventually replace national standards from March 2001. The
development of these standards is also an important element in implementing the
Recommendation without legislative regulation of the environmental exposure of people.
The Commission would like to organise, by the end of the first quarter of 2001, a forum of
discussion between governments of Member States, scientific organisations, the mobile
telephone industry and consumer associations in order to reach a consensus on the

management of risks associated with mobile telecommunications. Based on this, the
Commission would be in a position to submit, by the end of 2001, a new legislative proposal
to the Council and the European Parliament. The Recommendation of 1999 was under
discussion by the Parliament in 1994; the initial timetable predicted for the revision of this
Recommendation represents an acceleration of the rhythm of the procedure of gathering
information and reaching decisions which is quite appropriate in view of the rapid
development of new technologies that increase public exposure to RF. The Commission
favours the setting-up by mobile telephone operators of procedures for consultation with
residents prior to the installation of base stations. The situation in New Zealand is cited as
an example of this. The whole situation is framed in public involvement in decision-making:
it relies on the following argument that involves everyone:
"Would you like to be able to use mobile phones for communications? If yes, than there is a
need for base stations.
· This is how a base station works;
· Here are the standards of emission formulated to avoid any risk to the health of
residents, and here are the results of surveys carried out in the neighbourhood of
existing base stations.
· It is up to you (operators and residents) to come to an agreement on the siting of base
stations in your neighbourhood."
Such approaches incorporate rationalisation of the perception of risk related to the use of
these systems via a cost-benefit analysis, as required by the Council Recommendation.

4.5.16 Laurent BONTOUX

European Commission Director General of Research – Life Science Directorate (Health,
food and environment)
This is Mr Bonoux's written response (dated 12/12/2000) to the letter about the European
Union's research programme addressed to him by the group of experts .

Thank you for the information concerning the work of your group of experts. The points
made below do not represent the official position of the European Commission but give a
factual response to the precise questions that you asked.
During the preparatory work for Framework V, we became aware that the public has
legitimate concerns about the potential effects on health of mobile phones. The large
scientific uncertainty remaining in this field calls for new, more conclusive studies to be
done. This issue was therefore included in the work programme of the "Environment and
Health" Key Action under the umbrella of the "Quality of Life and Management of Human
Resources" Programme (1998-2002).
Initially, no priority was given to any specific area within the general theme of effects of
electromagnetic radiation on health. However, following the selection and funding of 5
projects after the 1999 call for proposals, this initial position was somewhat modified. In this
call, about 9M euros were allocated to projects aiming at elucidating potential carcinogenic
effects of RF. After consultation, in particular with the co-ordinators of our projects, we
decided not to include electromagnetic radiation in the research priorities of the 2000 call,
then to focus the 2001 call for proposals (currently open) on the potential for non-
carcinogenic effects of RF.
There is no a priori allocation of budget, in the 5th Framework, for the potential health risk of
RF specifically. The "Environment and Health" Key Action has a total budget of 160M
euros for the period 1998-2002. At the moment, following the calls of 1999 and 2000, 9M

euros have been allocated to research projects on RF and 1.4M euros to research projects
on radiation in the range of 1 Hz. We do not yet know how much will be allocated to studies
on non-carcinogenic effects because that depends on the number of proposals received by
March and on their success after evaluation.
For the period 2002-2004, nothing can be said because that depends on decisions related to
the 6th Framework.
The first projects financed within the 5th Framework started at the beginning of the year 2000
and all have durations of between 3 and 4 years. The most important results will be known
before 2004.
The results of these research projects will be used in the development of policies and
European recommendations and will be included in the WHO general evaluation expected in
2004-2005. Their relevance is ensured by two mechanisms. On the one hand, the work
programme of DG RDT and their annual reviews were developed in co-ordination with the all
the DGs concerned, in order to ensure that the themes covered correspond to the main
political concerns. On the other hand, the evaluation of the proposals includes an item on
"relevance" and on "added value" with the intention of maximising the usefulness of the
results for the Commission.
I do hope that you find some of this information useful. Please do not hesitate to contact me
if you need further details.

5 Conclusions of the Group Of Experts on Health Hazards and
Recommendations for Reducing Exposure of the Population to RF
Electromagnetic Fields
[Note: the translation of this section provided with the French text starts at Section 5.2. ]

5.1 Outline of recommendations from recent reports

This short section deals in a factual manner with the recommendations on matters related to
risk management given in the different reports considered by the group of experts. In the
following section the group of experts gives its own, complementary, recommendations.

The ARCS and COST reports do not give any recommendations on risk management. The
report of the Royal Society of Canada proposes that the exposure limit for workers be
reduced to 1.6 W/kg, the level used in Canada for public exposure38, because the limit of
8 W/kg for exposure of the head, neck and trunk is considered too great in view of the
possibility of risks to the eye. The 1998 WHO memorandum considered that, overall, any
health risk was unlikely. WHO does not make any recommendations for protection except
for common-sense measures (control of access to base station antennas, the avoidance of
EMC-related safety problems, the development of effective communication to "reduce
mistrust and anxiety"). The June 2000 WHO publication maintains a reassuring tone, but
does give some "active" recommendations: don't use a phone while driving; ask
manufacturers to reduce emission levels voluntarily; ask the public to limit the exposure of
themselves and their children; use a hands-free kit; consider visual impact and public
concern about the siting of base stations.

The Stewart Report is primarily concerned with giving recommendations. These are based
explicitly on a "precautionary approach" to the current state of scientific knowledge, except
for the risk of accident from using a phone whilst driving. Recommendations concerning
local authorities, industry and consumers are given.

These recommendations justify lower levels of exposure for the public than for workers for
various reasons: the existence of sensitive subgroups (old people, children, ill people);
differences in exposure duration; the possibility that children absorb more RF than do adults.
The Stewart Group also recommend the putting in place of a register of exposed workers
and a study of their mortality: this should be relatively easy in most circumstances (this is
done in France for all occupational risks).

The Stewart Group recommended the rapid establishment (within 12 months) of an

obligatory pre-installation regime for base stations, the setting-up of a national register of
base stations and their characteristics (with the additional advantage of facilitating future
epidemiological studies39), an independent measurement audit of compliance with exposure
standards, and that emission levels are the minimum compatible with good network
functionality. The Stewart Group recommended that beams from base stations should not
be permitted to fall on school grounds without the formal agreement of the school and the
parents, a measure difficult to put into practice. They also recommended the putting-into-
place of effective and identifiable exclusion zones, with specific signage, and
encouragement of operators towards site-sharing.

The recommendations addressed to industry concern the adoption of international standards

for the assessment of SAR, and the co-ordinated provision of SAR information for phone
Measured according to the ANSI standard
A similar French national database is managed by the National Frequency Agency

users. The Stewart Group particularly insisted that children's use of mobile phones should
be limited and that manufacturers and operators should stop all advertising aimed at

Specific recommendations for the public and mobile phone users: provision of information
on potential health effects, creation of the post of Ombudsman for base station installations.

5.2 Recommendations of the group of experts

The group of experts' brief was to express an opinion on the available scientific data and to
make recommendations in the field of public health. This put the group in a situation that
went beyond the usual role of scientific experts, whose expertise relates to risk assessment.
The group of experts accepted this mission. Thus, in this chapter, the group of experts
present their conclusions on health protection, based on all the information they obtained
and analysed. This included several summary reports, as well as the most recent scientific
research, and opinions expressed by the personalities they interviewed. The first paragraph
sets out the salient points on which they based their rationale, and the remainder of the
chapter consists of their recommendations.

Salient points:
1. International regulations, inspired by the work of the International Commission on Non-
Ionizing Radiation Protection (ICNIRP), are based solely on scientifically demonstrated
biological effects corresponding to health hazards. In the RF range, these consist of
certain effects due to heating generated by dielectric absorption. ‘Reduction factors’
have been applied to the lowest exposure levels that cause the most significant effect in
animals to transpose these to the human species; occupational as well as general public
exposures are considered. Exposure levels are expressed in terms of the quantity
Specific Absorption Rate (SAR), which was used to define the ‘basic restrictions’ for
exposure of the public in the European Union Recommendation of July 12th 1999.

2. Current scientific data do, though, indicate that a variety of biological effects occur at
energy levels that do not cause any localised temperature rise. From the current state of
knowledge on these non-thermal effects, it is not yet possible to state whether they
represent a health hazard.

3. Is it possible to state that there are no health risks ? No: although few scientific
arguments are available to back up this hypothesis, it is not possible to eliminate the
possibility of non-thermal health hazards associated with low-level RF fields on the
basis of our current state of knowledge. Furthermore, some potentially-serious effects
(e.g. promotion of brain cancer) are currently the subject of large-scale, international
epidemiological research which will not produce results for several years. Research is
also continuing into other potential effects (e.g. effects on hearing or the nervous system,
and headaches).

4. If future research were to validate this hypothesis, i.e. demonstrate the existence of
health hazards related to the usage of mobile phones, the probability, at an individual
level, would certainly be very low. Indeed, it is reassuring to note that no risk has yet
been demonstrated, in spite of the considerable amount of work undertaken over several
years. However, if there were a risk, the very high number of mobile telephone users
could mean that, even if the individual risk were very low, the impact on public health
could be substantial. In view of this, would it not be prudent to set new exposure
‘standards’ at lower values immediately, without waiting for the results of ongoing
research ? The group of experts consider that measures of this type would be justified if
they were really effective in reducing potential risks. This implies, firstly, that medical
effects resulting from exposure to RF fields could be identified and, secondly, that new
values could be defined that would guarantee reduction, or even elimination, of this risk.

This is not the case in the current state of knowledge. There is not, at present, any
reliable scientific information that could be used to optimise such measures. Therefore,
such exposure limits would be unscientific, arbitrary, misleading, and would probably
differ from one manufacturer or country to another, thus adding to public confusion and

5. It has been conclusively established that using a mobile telephone while driving, whether
with a hands-free kit or not, is a real accident risk factor. This risk is not associated
with electromagnetic fields, but is due to the loss of concentration resulting from the
telephone conversation. This is a major risk, in terms of both frequency of occurrence
and seriousness.

6. Mobile telephony has also been identified as a factor in security and medical safety (by
speeding up calls for help and, thus, the arrival of emergency services, etc), and there
are other advantages that were not covered by the group of experts' brief.

Consequently, the group of experts make the following recommendations:

1 They recommend adopting an approach based on the Precautionary Principle

(as the Principle is described in Chapter I of this report). The general overall objective
should be to reduce average exposure of the public to the lowest possible level
compatible with service quality. The following aspects should be taken into account
in implementing this principle:

a More intensive research into the biological and medical effects of exposure to RF is
required, to reduce the uncertainties and gaps in knowledge. Research priorities and
organisational recommendations are presented in the following chapter.

b Users should adopt prudent avoidance measures - simple steps aimed at reducing
superfluous exposure (e.g. minimise the use of mobile telephones when reception is
poor, use an earpiece, and avoid carrying mobile phones close to potentially sensitive
tissue – i.e. a pregnant women's abdomen or an adolescent's gonads).

c Manufacturers should continue their efforts to reduce mobile telephone emissions to

the lowest possible level compatible with service quality.

d The objective of reducing public exposure to a minimum is particularly important for

potentially sensitive populations, including children and sick people. For this reason, the
group of experts recommend that ‘sensitive’ buildings (hospitals, day care centres, and
schools), located less than 100 metres from a base station, should not be directly
in the path of the transmission beam40. This recommendation is not incompatible with
the installation of a base station antenna on the roofs of buildings in this category, as the
incident beam has little or no effect on the area immediately below it ("fountain" effect).
The group of experts feel that, if operators apply these measures, public fears, especially
those of parents concerned by their children's exposure in school, will be allayed,
especially keeping in mind that, in view of the exposure levels observed, the group of
experts does not back the hypothesis that there is a health risk for populations living in
the vicinity of base stations.

e Children are theoretically a sensitive population. From the current state of knowledge, it
does not appear that sensitive organs inside children's heads receive a higher dose of

The limit of the beam is reached when the field power has decreased by a factor of two. This beam
is defined in the horizontal axis and in the vertical axis.

microwaves during a phone call than those of adults. However, if mobile telephone use
starts at an early age, cumulative lifetime exposure will be higher than that of today's
adults (however, the constant, rapid changes in technology may lead us to reconsider
this). There are no scientific data establishing any risks from long-term exposure in
adults or children, but neither is it possible to eliminate that possibility based on the
current state of knowledge. For this reason, the group of experts suggest that parents
who feel it is necessary to equip their children with mobile phones should ensure that
they make reasonable use of this equipment. A recommendation to this effect should be
included in the instructions for use supplied with all mobile telephones.

f Exclusion zones in the immediate vicinity of base stations, where access is prohibited,
must be clearly marked, with a pan-European signage system.

g The group of experts do not support the proposal in the Stewart report concerning site
sharing. In keeping with the objective of reducing public exposure to the lowest possible
level, the group noted that calculations have shown that, while concentrating or
dispersing antennas does not affect average exposure, installing several in the same
place would tend to concentrate the electromagnetic fields in space, and, therefore,
lead to a more heterogeneous exposure for the population. The group is aware that this
point of view may be in contradiction with concerns for landscape preservation, but feel
that this problem can be solved by additional efforts to integrate (or hide) these antennas
in the landscape, at a relatively moderate cost to operators.

The group of experts felt that consideration of the proposal in the Stewart report
concerning mediation for the installation of base stations was not part of their brief.
They do not endorse the hypothesis that living close to base stations poses a health risk.
Furthermore, they felt that aesthetic or economic issues involving base stations were
outside their ambit.

The group of experts emphasises that the prudent attitude they recommend, given the
current state of knowledge and uncertainties, does not in any way constitute a validation
of the health hazard hypothesis. It is, rather, a matter of common sense advice justified
by the existence of a "reasonable doubt", pending further scientific investigation.

2 Public authorities should provide incentives for the implementation of these

principles by all the stakeholders:

a The European Commission's July 1999 recommendations should be applied in national

legislation as soon as possible, to clarify the situation for all those involved.

b Mobile telephone users should be able to find out the extent of their exposure41. This
requires two types of measures:

· Information on the power of mobile telephone emissions and on the local SAR in users'
heads, measured under standardised conditions, should be provided with every mobile
telephone purchased. This will facilitate a comparison between mobile telephones, taking
into account radiation efficiency, which affects local SAR.

· Telephone displays should inform users of the emission strength during conversations,
expressed in a simple way (e.g. % of maximum power, averaged over the duration of the
last telephone call). This would have an educational effect, showing users that making

It is recalled that exact exposure cannot be directly measured and must be estimated according to
some standardised parameters. True exposure varies considerably according to the environment and
the antenna position. Standardised assessment procedures are under development.

telephone calls under poor reception conditions increases the radiation they receive
quite significantly.

c The public should be able to find out the usual exposure in the vicinity of base stations.
Several measures would be required to implement this principle:

· The ‘Agence Nationale des Fréquences’ (ANFR – National Frequencies Agency) should
establish standardised rules for measuring electromagnetic fields in the vicinity of base
stations (prepared in the context of European protocols), as soon as possible. These
rules should be adopted by all technical monitoring organisations authorised to carry out
these measurements.

· An obligation to transmit the results of these measurements to the national monitoring

authority, presently the ANFR, should be included in the specifications of all
organisations authorised to carry out these measurements.

· Regular measurement campaigns carried out according to an annual, long-term

programme, on the initiative of the ANFR, using a sampling schedule taking population
density into account, in order to define maximum population exposure values (closest
buildings in base station emission beams)42.

· The ANFR should set up a register of results per site for all their own measurements and
those carried out by authorised private organisations, in the form of a data base
accessible to the public via the internet. The ANFR should publish an annual report
summarising the field levels measured throughout the entire country.

· The ‘Groupe interministériel RF’ (Interdepartmental RF Group) should issue the set of
technical specifications for the installation of base stations as soon as possible. This is
currently in preparation at the CSTB and its application should be made compulsory.
These national specifications should soon be replaced by a standardised European
reference manual.

3 In view of the frequency and seriousness of accident risks, there should be more
driver information on the danger of using mobile telephones while driving, with or
without a hands-free kit, and traffic laws on this subject should be made stricter. A
national information campaign on this theme should be launched in 2001.

4 The public should be given more extensive information on issues of legitimate


· The informative document currently being prepared by the ‘Groupe interministériel RF’
(Interdepartmental RF Group) to explain the physical and biological phenomena
associated with mobile telephony should be completed and circulated to the general
public as soon as possible. In particular, it is necessary to explain that exposure to the
electromagnetic field of base stations is insignificant compared to that associated with
mobile telephones themselves, even considering the exposure of the closest neighbours
of the base stations under the most unfavourable emission conditions.

· It is recommended that people who have electronic implants (pacemakers, insulin

pumps, neurostimulators, etc.) carry their telephone at least 15 cm away from their
implant and hold it to the ear on the opposite side when they call. If these measures are
applied, the use of mobile telephones should not present any risks. The group of experts

It follows from this sampling approach that average population exposure will be much lower than
these values.

noted that technical research into electromagnetic compatibility is continuing, particularly
to deal with new technological developments in RF, which will use different ranges of

· The group of experts do not recommend that users equip themselves with "anti-radiation
protection" systems, which have by no means proved their effectiveness.

5 The group of experts were not asked to examine the issue of occupational
exposure to RF. However, they recommend that appropriate steps should be taken
and their implementation monitored by the labour inspection department and
CRAM specialists, to ensure that maintenance and repair operations at base
stations are carried out only when the installation is shut down. This
recommendation is consistent with the concern to reduce exposure of the
population – including workers – to the lowest possible level compatible with
service quality.

The Royal Society of Canada report recommended the reduction of local exposure limit
thresholds for workers to the same values as those applicable for the public. The group
of experts consider that this issue should be settled by ad hoc occupational risk
management committees in France and the European Union. They are in favour of the
recommendation in the British report that a register should be put in place for certain
categories of highly-exposed workers, for epidemiological monitoring purposes.

6 In compliance with the rules recently laid down by the European Union governing
the application of the Precautionary Principle, a report on all available scientific
data should be prepared regularly by an ad hoc scientific body recognised by the
Union, in order to examine whether there are grounds to modify legislation on the
exposure of the general public and workers to electromagnetic fields associated
with the radiofrequencies used by mobile telephones and base stations, and make
appropriate recommendations to the political bodies responsible for these

6 Recommendations for Research

6.1 Actual state of research at the international level

Research activity on the question of effects of mobile phones is actually centred in Europe.
In fact, this type of research started in the United States in 1993, but the funding has dried
up in the last few years and European governments and organisations have taken the lead.

The table below shows the state of research, ongoing or recently published (published in
2000 or in press). We can see from this table that there is a balance between in vivo and in
vitro studies (in particular due to the contribution of the European Community), we can also
see that studies on cancer models prevail.

Type of study Completed Ongoing Total

Epidemiological studies 4 3 + 10 (CIRC) 7 + 10 (CIRC)
Long term genotoxic studies 4 7 11
Sensitivity models 8 8 16
Other in vivo studies 24 15 39
In vitro studies 28 22 50
Total cancer 68 55 123
In vitro studies 36 19 55
Acute in vivo studies 14 17 31
Human studies 26 23 49
Total non-cancer 76 59 74
Total 144 124 268

Main source: database of the WHO EMF project .
The usual source of information on results of research is the totality of the articles published
in the peer-reviewed literature. This is true for research in all scientific fields but in the case
of biological effects of electromagnetic fields, the number of specialised reviews and quality
papers is still small. In view of the diversity of the studies at the biological and physical
levels it is difficult to make a global interpretation and a review of all the knowledge acquired.
To facilitate the work, evaluations are made by different national and international
organisations. For example WHO publishes exhaustive databases on a regular basis.
Other informed databases are available44 as well as reports by experts.
The main in vivo and in vitro research programmes that are ongoing worldwide are briefly
described below.
Three types of research funding were identified: (i) programmes partially funded by a
government or the European Union, (ii) programmes carried out entirely by industry, and (iii)
programmes financed by industry through a foundation that shields the research from the
source of funds.


National programmes
There is a great deal of research in Germany, all the fields are covered by a number of
research groups that are funded by industry, the FGF foundation (see later) or the
government. Some examples are given below:

· Influence of exposure to RF fields (383 MHz, 900 MHz and 1.8 GHz) on the synthesis of
melatonin and reproductive functions in the hamster. (A. Lerchl, Institute of
Reproductive Medicine, University of Muenster).
· Effects of pulsed RF fields on the physiological parameters of the rat. Study of the
effect of an additional intense and acute exposure during gestation. (Dr. Buschmann,
Fraunhofer Institute for Toxicology and Aerosol research (ITA), Hanover, Dr. Hansen,
Chair of Theoretical Electrical Engineering, Bergische Universitaet-Gesamthochshule
Wuppertal; Dr. Chaloupka, Chair of High Frequency Engineering, Bergische
Universitaet- Gesamthochshule Wuppertal).
· Membrane receptors of electromagnetic fields (Prof. Boheim, Department of Cell
Biology Ruhr-University of Bochum; Prof. Hansen, Chair of Theoretical Electrical
Engineer, Bergische Universitaet-Gesamthochshule Wuppertal; Dr. Grosse,
Umweltagentur, Ruhr-University of Bochum
· Study of chemically induced tumours by DMBA in the rat subjected to low power GSM
signals (Bartsch, Tübingen, Deutsche Telekom).
· The study by Jensh on teratogenic effects of GSM signals on rats.
An informative database of recent results has been formulated (Dr. Silny, Institute for
biomedical Technology, Technical University (RWTH) of Aachen, http://www.femu.rwth-
Funding for research comes from a special tax on transmission licenses. The funds
collected are administered by an inter-ministerial committee, and the National Health and
Medical Research Council evaluate, recommend and supervise the research projects.
· Replication of the experiment of Repacholi on transgenic mice (Vernon-Roberts,
· L. Verschaeve is carrying out an in vitro study on the effect of GSM 900 signals on mice
with chemically induced tumours.
· A national cohort epidemiological study, carried out by Johanson, is in its final phase
(incidence and mortality of cancer).
· The university of Aalborg is very active in dosimetry (J. Bach Andersen).
· The Finnish research programme on possible health effects of mobile phones comprises
seven projects (including an animal study on skin cancer and two studies on cellular
culture). The research institutes taking part are: the University of Kuopio, the Finnish
Institute of Occupational Medicine, the Finnish Radioprotection Centre, the Finnish
Cancer Research Centre, and the Technical Research Centre of Finland. The
programme is funded to at a level of 1.1M euros per year, approximately.

· Juutilainen at Kuopio is studying the effect of exposure to GSM signals on transgenic
mice whose tumours were initiated by UV.
The COMOBIO Project (MObile COmmunications and BIOlogy) is currently underway within
the framework of the RNRT Programme (National Network of Research in
Telecommunications), financed by the Ministries of Research and Industry. It comprises
eight sub-projects of which two relate to the dosimetry of mobile phones and exposure
systems. The six sub-projects relating to biology are the following:
· Human studies on auditory evoked potentials, including studies on epileptic subjects
(J.L. Coatrieux, and G. Faucon, University of Rennes, University of Marseilles and of
Montpellier-Nîmes, ENST Paris)
· Hearing in the hamster (J.M. Aran, Inserm, University of Bordeaux)
· Brain metabolism (Bruno Bontempi, CNRS, University of Bordeaux)
· Memory and learning in the rat (J.M. Edeline, CNRS, University of Orsay)
· Neurotransmitters and receptors in the rat (R.de Sèze, University of Montpellier-Nîmes)
· Blood-brain barrier in the rat (P.Aubineau, CNRS, University of Bordeaux).
The total funding for the COMOBIO Project, involving about fifteen groups and the main
industrial players in mobile telephony in France, is about 12 MF over two years.
Other projects are ongoing, financed by the mobile telephone operators:
· Effects of GSM signals on the development of tumours chemically induced by DMBA (R.
Anane, CNRS, University of Bordeaux)
· Effects on the skin of GSM signals: study on cellular cultures, reconstituted skin and on
animals (B.Billaudel, CNRS, University of Bordeaux)
· Effects of GSM signals on cognitive tasks in man (R.de Sèze, University of Montpellier-
The majority of the projects ongoing in Greece relate to dosimetry (evaluation of fields in the
vicinity of base stations and terminals).
· Evaluation of health effects of mobile phones (K.S.Nikita, Biomedical Simulations and
Medical Imaging Unit, Department of Electrical and Computer Engineering, National
Technical University of Athens).
· PERFORM-A: studies of carcinogenicity in rodents (J.N. Sahalos, Radio
Communications Laboratory, Department of Physics, Aristotle University of
· Thuroczy's group is very active in different fields of study such as the effects of GSM
signals on the offspring of mice exposed to GSM signals.
There is a great deal of research activity in Italy. A multicentre research network was
created to co-ordinate the efforts of a number of universities (ICEmB). The Italian Ministry of
Research (MURST) has spearheaded a recent programme of research entitled: "Protection
of man and the environment from electromagnetic emissions". This vast programme
involves 59 institutes of research and is co-ordinated by the National Research Council
(CNR) and the Agency for New Technologies, Energy and the Environment (ENEA).

Several research projects are ongoing, in which animal and cellular models are used to
study the effects and their mechanisms.
Most of the studies were carried out under the umbrella of Framework V.
· Studies of the mechanisms of interaction of microwaves with tissues (group of d’Inzeo).
· Effects of the exposure on 3T3 and L929 cells (gene expression and oxidative stress).
Changes in the function of the internal ear of rats exposed to GSM signals (group of
· An epidemiological study on brain tumours in mobile phone users (PDC, 1.5 GHz) is
ongoing, led by Yamaguchi
· Sharai is pursuing a study on mammary tumours in mice (DNBA model, PDC at 1.5
The great majority of Polish research is carried out by the military health service:

· Effects of microwave fields on immunomodulation (E. Sobiczewska, Military Institute of

Hygiene and Epidemiology, Varsovie).
Other studies are mostly on dosimetry:
· evaluation of the exposure of workers and residents to systems of mobile telephony
(H.Aniolczyk, M.Zmyslony, Institute of occupational Medicine, Lodz)
A commission of experts is in the process of establishing a new national exposure standard
for the public and workers.
Czech Republic
Research activity in the Czech Republic deals mostly with field measurements and the
theory of interaction.
· Possible biological effects on mobile phones (J.Musil, National Institute of Public Health,
· Biophysical mechanisms of interaction of field with digital transduction signals
(J.Pokorny, Institute of Radio Engineering and Electronics, Academy of Sciences of the
Czech Republic, Prague).
· The main research activity in Slovenia is in the laboratory of D.Miklavcic at the Technical
University of Ljubljana. It relates mostly to dosimetry (for example, measurement of the
exposure of workers in the Mobitel company ).
· Study of SAR as a function of permitivitty (P. Gajsek, National Institute of Public Health
of Slovenia, Ljubljana and US Airforce research laboratory, Trinity University, Texas).
· The group of Hardell is carrying out an epidemiological study on brain tumours
· Persson and Salford, in Lund, are pursuing studies on the blood brain barrier in rats.

Research activity in the USA has significantly decreased in the past few years. There are
only a few studies funded by Motorola (groups of Roti Roti at St. Louis and of Anderson at
Battelle). Long-term studies on animals are envisaged by the NTP45

European Union and foundations

European Commission
Three multinational projects have just started within the 5th Framework Programme ("Quality
of Life" Action, "Environment and Health" Theme).

· REFLEX. This project is co-ordinated by the VerUm foundation in Munich, Germany. It

funds studies on the impact on the environment. The theme of REFLEX is the effect of
fields on cellular processes. The thrust of the work is to carry out in vitro studies on
models of carcinogenesis and neurodegenerative pathology. The different
complementary approaches include: direct and indirect genotoxic effects, effects on
cellular differentiation of stem cells, effects on gene expression, on the immune system,
on cellular transformation and apoptosis. Eleven European laboratories are partners
within this program; Benjamin Franklin clinic, Free University of Berlin,
Hindenburgdamm, Germany; University Clinic of Internal Medicine, Vienna, Austria;
Botanical Institute, Gatersleben, Germany; Ramon y Cajal Hospital, Madrid, Spain;
Radiobiology Laboratory, STUK, Helsinki, Finland; Institute of Biophysics, University of
Hanover, Hanover, Germany; University of Bologna, Bologna, Italy; Laboratory PIOM,
Talence, France; Department of Pharmacology, University of Milan, Milan, Italy; Institut
für Feltechnik und Höchstfrequenzietechnik, ETH, Zurich, Zurich, Switzerland. The total
funding for this project over three years is about 3M euros.

· CEMFEC. This project is coordinated by J. Juutilainen (Kuopio, Finland) its duration is

45 months. The objective is to determine whether or not electromagnetic fields can act
as co-carcinogens by enhancing the effect of carcinogens present in the environment.
In vivo and in vitro studies are incorporated in a complimentary manner. In the animal
study, the combined effect of exposure to fields and to a mutagen (MX) present in
drinking water will be studied. In vitro, two cellular lines will be exposed to field in the
presence or absence of MX or of the herbicide vinclozoline. Mechanisms of co-
carcinogenesis such as oxidative stress, proliferation and apoptosis will be studied.
Animal studies will be carried out at the University of Kuopio (Finland), in the Fraunhofer
Institute for Toxicology and Aerosol Research (Germany), in the Flemish Institute for
Technological Research (Mol, Belgium) and in the laboratories of the Radiation and
Nuclear Safety Authority (Finland). The cellular study will be carried out in Italy through
the interuniversity network based in Genoa “Interuniversity Centre for Interaction
Between Electromagnetic Fields and Biosystems”..

· PERFORM A. The most important of the three projects proposed by the industrial
consortium was funded; it tests the effects of RF on biological end-points related to
cancer in animals. The results of these studies will be used in the evaluation, in 2003,
of the outcome of studies by IARC into the carcinogenicity of RF. Rats and mice will be
exposed to GSM 900 and 1800 for 2 years (RCC in Switzerland and Fraunhofer ITA in
Germany). A complete histopathological study will be carried out on all the animals.
The ARCS laboratory in Vienna (Austria) will study chemically induced tumours by
DMBA in order to replicate the experiments carried out in Bordeaux and Tübingen. A
replication study of the Australian experiment of Repacholi on transgenic mice will be
carried out in the RBM laboratory in Italy. The group of N.Kuster in Zurich will be

National Toxicology Program

responsible for the dosimetry and the exposure systems including the challenge of
designing an 1800 MHz exposure system providing a well-defined SAR distribution in
the animals.

The FGF foundation was founded in 1992 to carry out research on the biological and health
effect of electromagnetic fields and to disseminate information on these matters, while
providing a buffer between funding agencies and researchers. The FGF has funded about
60 research projects over the past seven years at the cost of 50M F of which several were
animal (exposure system, melatonin, etc.) and cellular studies (calcium, etc).

The Electra 2000 foundation was set up to provide information on possible health effects of
electromagnetic fields and to co-fund research. It receives funds from the three Italian
operators Tim, Omnitel and Wind and funds Italian research groups at the recommendation
of an international scientific committee. Electra 2000 is currently evaluating several animal
and cellular studies.

The five-year WTR program (Wireless Technology Research) was funded by the CTIA
(Cellular Telecommunication Industry Association). Some resources were allocated to the
design and production of viable in vivo and in vitro exposure systems. Animal studies
investigated the fragmentation of the DNA in rat brain and on the formation of micronuclei in
lymphocytes exposed to high intensity fields. The program has now concluded and
publications are in press.

Further research is taking place within the framework of a co-operation between industry
and the FDA (Food and Drug Administration).

There is an important programme of scientific investigation already underway. In the next

section the group of experts will give recommendations on what it considers to be priority
areas for research. A large number of the recommendations are already incorporated in
previously described current research. In doing so, the group of experts is conscious that
some of its recommendations may be somewhat out of date and should be modified in the
light of the scientific evidence as it becomes available the near future.

6.2 The expert group's recommendations for research

The majority of the scientific work mentioned and analysed in this report is only indirectly
linked with the use of mobile telephones. The other data is often contradictory, making
comparative analysis of the results difficult, or even impossible, due to the diversity of
protocols and equipment used. These factors mean that assessments of the potential health
risks of mobile telephones can easily have a subjective bias.

It is also clear that some experts analyse the experimental findings through an implicit filter,
assuming that the biological effects of GSM microwaves must be due solely to an increase
in temperature in tissue, whereas other analyses envisage the possibility that microwaves
may have non-thermal effects, even if the precise mechanisms involved cannot be explained
scientifically at present. Depending on the importance given to experiments aimed at
demonstrating non-thermal effects, the main lines of research envisaged by various groups
of experts are quite different.

Following these prefatory remarks, the group of experts make three general
recommendations concerning research, to be started or completed, on the biological effects
of GSM microwaves:

· Experimental protocols and equipment should be standardised as soon as possible,

following discussion on a national and international level. This will facilitate comparison
of the findings of different studies (this is only partially being done in the context of the
WHO's EMF programme);

· in view of the very slight heating of tissues observed with mobile telephones in normal
use, special effort should be devoted to investigating effects that cannot be directly
explained by the thermal activity of microwaves (in vitro and in vivo);

· to reduce the risk of error or imprecision in interpreting results, a comprehensive

"monitoring" of the contingent physiological variables should be included in all new
protocols (stress assessment in animals, checking the vigilance of subjects during EEG
studies, etc.), particularly those for in vivo experiments on animals and humans. This is
especially true of experiments designed to demonstrate or confirm non-thermal effects of
microwaves (low- or very low-power exposure).

A review of all the available scientific literature shows that a number of research fields have
received little or no attention. This is particularly the case for:

· a possible synergy between the effects of microwaves and certain pre-existing or

concomitant chronic or acute pathologies (particularly skin and neurological conditions),

· the effect of microwaves on certain tissues that are directly exposed to varying degrees
(meninges, blood vessels, skin, etc.) under prevalent conditions of use (telephone held
to one ear), or probable future conditions related to the implementation of new
technologies like Bluetooth (telephone attached to a belt or carried in a pocket, which
leads to exposure of the skin, peritoneum, viscera, and sexual organs). Particular
attention should be given to the potential impact of RF on certain resident immune cells
active in a number of pathologies affecting the skin, aponeuroses, meninges, and viscera
and involving inflammation and/or pain (mast cells), both in vitro and in vivo.

· the particular effect of GSM waves on growing organisms and tissues (embryos,
foetuses, children, and adolescents);

· the possible effect of base stations during "whole body" exposure at average or relatively
strong intensities (installation and maintenance staff).

The group of experts felt that the major topics described in detail below should be given the
highest priority. Some of them require preliminary dosimetric studies. It is also important to
keep up with new developments in this technology and start research now into the potential
effects of the new frequency ranges that will be used in the near future (UMTS, Bluetooth,

Finally, the lack of epidemiological studies other than those targeting mobile phones as a
possible cause of brain cancer indicates a need for different types of research aimed at
identifying the possible effects of GSM microwaves on other pathologies, particularly
headaches, either in "ordinary" users or those with physiological conditions likely to make
them more sensitive to such effects.

The group of experts also made several specific research recommendations.

6.2.1 Studies of biophysical interactions
The COST 244b report recommended that large-scale work modelling these interactions
should be carried out prior to research in this field. It is obviously necessary to make a
precise determination of the type of field on a molecular level in order to predict its
macroscopic effects, which are based on the microscopic ion-ligand and membrane protein
interaction models.

This research may lead to an investigation into cellular mechanisms for detecting RF fields
(this work should initially concentrate on cells in the nervous systems of certain vertebrates
that are sensitive to magnetic fields).

6.2.2 In vitro studies

The biological effects of RF observed in vitro to date are of very small amplitude, which
perhaps explains the difficulty of reproducing them experimentally. Furthermore, if these
effects were confirmed, it would still be difficult to determine their possible health

By definition, in vitro studies observe isolated systems that do not take even the most
elementary interactions between the organic element studied and the rest of the system into
account. However, in vitro research is useful for studying the action of microwaves on
single-cell models (bacteria) and some cells isolated from animal and human organisms
(starting with immune and germ cells). These are justified in three specific cases:
1) replication of earlier positive experiments, 2) observing organisms that are difficult to
study in any other way, and 3) studying mechanisms that have not yet been investigated at

In these three categories, the following points should be highlighted:

· studies of the genotoxicity of microwaves (there are too few publications to form a
definite opinion). Among the experiments that should be replicated, for example, is the
"micro-nuclei" test which provides an evaluation of the extent chromosome of damage.
Another is the "comet" test, where fluorescence microscopy is used to identify
fragmented DNA (although a recent replication of this test in a French laboratory
produced negative results),

· the effect of microwaves on apoptosis or " programmed cell death" (no published work),

· gene expression (C-fos and C-jun) and nucleic acid synthesis. It is also important to
replicate recent experiments on worms showing that radiofrequencies modify the
expression of "heat shock" proteins without any increase in temperature using cultured
human cells. In fact, these proteins react to all types of cell "stress",

· changes in the synthesis and/or storage of neurotransmitters (brain slices);

· the effect of microwaves on intercellular transmission (brain slices);

· phenotype and functional modifications in immune cells (cultured cells).

All of the above experiments should be repeated in the presence of chemical and physical
(ionising radiation) mutagens to identify any interaction between these agents and RF.

In vitro studies of the effect of microwaves on the functional characteristics of inner ear hair
cells are justified due to the difficulty in carrying out in vivo research (relative appreciation of
the possible influence of microwaves on the various stages in the nervous system, from

these cells to the primary auditory cortex, and effects of prolonged exposure on their

6.2.3 Animal studies

The difficulties in comparing the results of different experiments mentioned in the
introduction to these recommendations are particularly acute in animal research. There may
be considerable differences in experimental conditions (exposure system, whether or not the
animal is anaesthetised, evaluation of SAR, etc. ) and some of them had crucial gaps, such
as not taking certain potentially important co-factors into account, including the stress of
tightly-restrained animals and the effect on their humoral, circulatory, or neurophysiological
condition. For this reason, a certain number of these investigations should be repeated using
stricter experimental protocols and these physiological variables, or at least a good index for
these variables, should be taken into account in analysing the results.

It is also true that several particularly important fields have received little or no attention,
although some are currently under investigation, e.g. in the French COMOBIO programme.

Among the data requiring confirmation, the group of experts give priority to the
following topics:
· effect on induced tumours (at SAR levels corresponding to GSM). Some controversial,
isolated experiments have shown that RF may reinforce the effects of certain
carcinogens or promote growth in transplanted tumours (cf. ICNIRP 1996, Repacholi
1998, Moulder et al 1999, Royal Society of Canada 1999); effect on DNA. Lai and
Singh's work (1995), showing DNA damage in animals exposed to radar signals,
requires validation;

· replication of experiments showing memory disorders in rodents, using behavioural tests

better targeted to assess different types of memory;

· effect on synthesis of neurotransmitters and their receptors in the brain;

· effect on neuron excitability (EEG, use of C-fos and C-jun markers);

· replication of studies investigating the permeability of the blood-brain barrier (relatively

large number of contradictory experiments at poorly-defined SAR levels, with no
monitoring of circulatory functions, using a variety of techniques to measure
extravascularisation, with differing levels of sensitivity that makes them difficult to

· effect on the inner ear. Intense radiofrequency fields produce an auditory perception
("click") that is interpreted as being due to a temporary increase in temperature that
produces a shock wave in the inner ear. No other effect has been shown to date and, to
our knowledge, none of the published work investigated the effects of transmissions at
power levels compatible with mobile telephones, although there is one ongoing study in
France. In particular, it would be advisable to study the potential effects of these
transmissions in conjunction with the "normal" acoustic stimulation involved in using a
telephone, taking into account the increase in temperature due to holding it against one
ear (independently of any thermal effect due to microwaves as such).

· long-term impact of repeated exposure on the appearance of cancerous tumours and

immune and endocrine system functions.

Fields that have received little or no attention to date:
· synergy with other types of radiation (UV, ionising radiations, etc) or chemicals with
known teratogenic, carcinogenic, or immunosupressor effects;

· synergy with chronic or acute pathologies (particularly certain inflammatory conditions46,

neuro-degenerative diseases, epilepsy, and the effect of alcohol or drugs);

· effect of RF on animals predisposed towards certain pathologies (cancer, high blood-

pressure, immune deficiency, etc.) or those that have been genetically modified (knock-
out for certain genes);

· effect on skin, paying special attention to melanocytes and resident immune cells, i.e.
mast cells;

· effect on the meninges (especially the dura mater, described as the site of migraine
diseases, also considered to be the lymphoid organ that protects the brain);

· long-term effect on the structure and functions of blood vessels in the skin, meninges,
and brain (paying special attention to the endothelium);

· effect on digestive tissues, gonads, and germ cells, in view of the likelihood that GSM's
will increasingly be worn on belts. For the same reason, research should be extended to
include the effect on embryos and foetuses (exposure of pregnant women). It is
particularly important to replicate the study by Magras and Xenos (1997) that
demonstrated a decrease in female fertility following exposure to low-intensity signals;

· general application of these research fields to immature animals (more suitable for
representing the susceptibility of children or adolescents).

6.2.4 Human laboratory studies

Any of the studies described below that are carried out in France will be required to comply
with the 1996 law on Bioethics and obtain approval from a CCPPRB.

The proposed research should be carried out using both healthy volunteers and patients
with a diagnosed pathology suspected of affecting the impact of GSM microwaves and other
frequency ranges under development on health. In most cases, these experiments will be
designed to answer questions raised by animal research using non-traumatic techniques
available in laboratories or hospitals.

Exposure of healthy volunteers:

· EEG (EEG and magnetoencephalography);

· neurotransmitters (positron-emission tomography);

On this subject, we note that there have been no complaints from users suffering from chronic skin
diseases, such as eczema or psoriasis, that affect the immune and nerve cells primarily involved in
other inflammatory pathologies, such as migraine, where the occurrence of attacks can be
exacerbated by GSM microwaves. If GSM microwaves are capable of acting on these deep cell
systems (meninges), they should a fortiori have an even greater impact on these same cells located
on the surface, as they are exposed to a much greater degree. In the same way, to our knowledge,
there have been no particular complaints from patients with vascular facial pain or characteristic
neuralgia such as that caused by damage to the sensory nerves linked to the trigeminal gland.

· immune and humoral systems (blood test);

· sleep;

· memory and cognitive tasks or tasks involving targeted associative brain regions;

· Immediate and delayed effects (repeated exposure) on sight and hearing;

· cardiovascular system. In particular, the experiment carried out by Braune et al (1998),

showing that a 35-minute exposure (with the telephone on the right side of the head)
caused a significant increase in arterial blood-pressure, accompanied by a decrease in
heart rate and capillary perfusion in the hand (signs of an increase in autonomous
sympathetic nerve activity), should be reproduced with a larger number of volunteers.

Part of this research should consist of non-invasive investigations, involving children and
adolescents, as well as patients with mild pathologies that may be aggravated (or attacks
may be caused) by the use of GSM, if this is ethically acceptable. These would focus
particularly on migraine, rheumatism of the joints, and inflammatory skin diseases, such as
eczema and psoriasis (which have not yet been studied, to the best of our knowledge). It is
also important to study the effects of mobile phone use on patients with more serious
pathologies: neurological syndromes (epilepsy, non-consolidated cerebral infarction, and
chronic or acute cerebral circulation disorders), cardiovascular problems (high blood
pressure), and ear or eye diseases (maculopathy, glaucoma, etc). In this field, we
emphasise that special attention should be given to the risk of epilepsy in children and
young adolescents. If the use of mobile telephones were shown to increase the risk of
epileptic seizure, the fact that the inhibiting mechanisms in young people's brains are not
completely developed should be taken into account, as it further aggravates this risk.

The same studies should also be carried out with a group of people who consider
themselves hypersensitive to RF, although no specific pathology has been identified, as well
as another group suffering from subjective symptoms (headaches, hot flushes, and attention
or memory disorders) associated with the use of mobile telephones (see study reports from
Scandinavia and Singapore). Double-blind studies could be carried out on these groups,
with or without exposure to GSM waves, under otherwise identical experimental conditions.

The group of experts also recommend that research should be carried out to identify a
possible nocebo effect, due to extensive media coverage of the potential dangers of mobile
telephones. The specific arrangements and protocol for such a study have not yet been
defined, but it could, for example, involve a group of healthy volunteers of the same sex and
age group, identified by a survey as being convinced that GSM phones are harmful or not.
These volunteers would be exposed to microwaves and subjected to the analyses described

6.2.5 Epidemiological studies

Apart from the risk of traffic accidents related to the use of mobile telephones while driving,
which is both clearly proven and serious, the other effects on human health are still only
hypotheses, backed to varying degrees by scientific arguments. It is therefore necessary, as
recommended by all the committees that have examined this issue, to develop further
epidemiological research, especially as some of the effects envisaged are serious, due to
their intrinsic severity and/or the high number of cases potentially attributable to the use of
mobile telephones.

Recommendations for epidemiological research should distinguish between the different

types of expected effects, particularly separating cancer from benign, short-term effects

(headaches, migraines, sleep disorders, "radiofrequency syndrome", etc). There are
considerable differences between protocols for epidemiological studies, including methods,
feasibility, and cost, depending on the effects to be studied.

Benign, short-term effects

The main priority should be given to studying exposure to mobile telephones, rather than the
proximity of base stations, as the latter are associated with very low exposure levels.
A variety of protocols may be used: cross-sectional, case-control, and prospective cohort

Cross-sectional studies are relatively easy to organise and may be completed in a short time
at low cost, but they have severe limitations in terms of causal interpretation. They may lead
to hypotheses, but never produce definite conclusions. Use data from operators for the
same subjects would strengthen the validity of these studies. A cross-sectional study based
on a sample where participants living close to base stations were clearly identified would be
useful for generating hypotheses, or disproving effects that are currently described without
any scientific basis.

Case-control studies are also difficult to interpret if they do not follow a particularly strict
protocol to avoid biased information, because the subjects under consideration are more
likely to attribute their disorders to the use of telephones at a time when the effects of
telephones are under close media scrutiny.

Prospective cohort studies (contemporary) are the best suited, as they study highly varied
effects, as well as the development over time of telephone technologies and methods of use
– if the monitoring period is long enough. The suspected benign effects are frequent and
short-term. Because of this, it is not necessary to set up very large cohort studies, and
reliable results can be obtained quite quickly, especially if data about the actual use of
mobile phones is available from operators. Existing prospective cohorts could be used, such
as SUVIMAX or GAZEL, with the addition on specific studies of the effects of mobile phones.
This would present several advantages (savings and speed), as they are already in place
and some of them have already had data collected on these effects for several years. This
methodological approach could easily and economically include studies on “well-being”, as
recommended by the Stewart report.

Susceptible or sensitive groups should be subject to specific studies, as should subjects who
are highly exposed in their working environment. Alongside general population studies, it
would therefore be wise to propose studies focusing on children, adolescents, and migraine
sufferers, as well as studies within suitably selected companies or professions.

Traffic accidents
Although the risk of accidents caused by the use of mobile phones while driving vehicles is
both clear and high, the group of experts recommends new epidemiological research in
France, for two reasons: (i) it would be useful to have epidemiological data comparing the
risks involved in the use of hands-free telephones with those of conversation with a
passenger, in order to confirm the results of experimental work on this point; and (ii) in terms
of prevention, the results obtained in the national context would obviously have a greater
impact on the public (and the public authorities), leading to a more effective implementation
of the necessary measures.

As far as base stations are concerned, the available data give no indication of a real risk.
Nonetheless, owing to demand, the group of experts would recommend verifying this point, if
possible. However, none of the epidemiological methods available (ecological, case-control,

or cohort) are capable of producing valid information due to the infinitesimal nature of the
risk, if it indeed exists, and the large number of potential confusion factors.

Several types of study can be carried out on mobile phones: ecological studies, population
case-control studies, cohort studies, and registers of exposed subjects. Ecological studies
do not seem to be appropriate in the current state of knowledge.

Population case-control studies are clearly the preferred protocol at present in attempting to
provide rapid answers to questions about the carcinogenic effects of using mobile phones.
As hands-free kits have only relatively come into use and we do not have sufficient
hindsight, this retrospective approach can only concern tumours of the brain, acoustic nerve,
and salivary glands. As the results of the huge, ongoing IARC study (‘Interphone' project, a
case-control study on tumours of the brain, acoustic nerve, and – although not in France –
parotid gland) in 13 countries, with a number of cases guaranteeing excellent power, will be
available in 3 or 4 years, it is unreasonable to propose developing new studies of this type in
France, especially as a French team is taking part in this international IARC study. The
funding of the French part of this study should be carefully examined.

On the other hand, it is important to stress the importance of the large-scale occupational
mortality cohort studies being carried out in various countries. This type of study is relatively
easy to set up in France, thanks to the many measures available for monitoring mortality.

The context is also theoretically favourable: many large companies have computerised
personnel files including full work histories, and a number of technical and research teams
have good knowledge of exposure to RF and other potential carcinogens. However, it will be
necessary to implement measures to guarantee the methodological quality (no in-house
epidemiological teams exist) and independence of the research. Certain categories of ‘highly
exposed’ workers, as recommended in the Stewart report, could be recorded in parallel with
the setting-up of occupational cohorts to form the base of a study, even though it would be
wise to supplement them with other types of users. These registers should obviously be
coupled with monitoring of mortality by cause.

The idea of a population cohort proposed by the Stewart report would seem difficult to
implement in terms of the risks of cancer, owing to the enormous number of people that
would have to be monitored over many years. In any case, this sort of effort is only
conceivable on an international scale (it should be remembered that IARC, quite rightly, did
not choose this protocol, preferring a case-control approach).

With a view to long-term monitoring, questions should also be asked about current and
forthcoming technological developments, as well as changes in the methods of use of
mobile phones, which are leading to exposure of other parts of the body. At the moment,
although it is clearly premature to envisage case-control studies of cancer in other parts of
the body, occupational environment prospective cohort studies could be considered the best
answer to this concern.

Other epidemiological research

Studies aiming to gain a better insight into exposure on a population level (including
“registers” of people who are ‘more exposed’, as indicated above) are necessary for various
reasons: (i) because of public concern about the possible effects of mobile phones, there is
good reason to give reliable, independent information about exposure within the population;
(ii) several of the epidemiological protocols envisaged here will be facilitated by the
availability of population exposure data.

This research could take several forms: individual dosimetry campaigns, and modelling
using data on base stations and the use of mobile phones.

Research in social sciences is necessary: quality studies of the psychological and
sociological aspects of mobile phone use would be of considerable importance if a “crisis
situation” were thought to be emerging.

Epidemiological monitoring
The group of experts' brief asked whether it would be advisable to set up a system for
monitoring the possible effects of exposure to RF. The group of experts consider that, like
future research in this area, priority should be given to the consequences of mobile
telephone use rather than the areas around base stations. The primary purpose of this
monitoring would be to produce information for decision-making purposes. Therefore, one of
the main criteria in considering the relevance of epidemiological monitoring programmes is
scientific evidence, which must be sufficiently sound to show that an increase in the
population's exposure to electric and magnetic fields resulting from the use of mobile
telephones is accompanied by an actual increase in associated health hazards.

However, epidemiological monitoring could sometimes be envisaged in the absence of

scientific evidence of such an association. Indeed, one of the other purposes of
epidemiological monitoring is to produce information used to generate hypotheses which
contribute to identifying disorder or pathology risk factors. In this context, epidemiological
monitoring would be one of the tools of descriptive epidemiology. In this case, its decisional
value is relatively low, but its relevance can be seen when the surveillance approach
efficiently collects the information required for the purpose of hypothesis generation.

Finally, when a potential hazard is an issue of widespread social concern and scientific
knowledge has failed to provide a satisfactory answer, epidemiological monitoring may also
be envisaged to collect further data on this social issue.

In the last two cases, the feasibility and cost-effectiveness of monitoring should be
compared with other approaches, including human or animal experimentation.

Does the issue of possible health effects associated with mobile phones fall into this

· As far as the risk of cancer is concerned, some départements already have registers
covering brain cancer. Cancer-related mortality is also covered by exhaustive records
throughout France. Scientific evidence on the role of exposure to RF associated with
mobile phones is very limited, as mentioned above. It would thus seem preferable to wait
for the results of the multi-centre case-control study co-ordinated by IARC before
deciding on any systematic monitoring of these pathologies using the national registry.
The specific association between exposure to electromagnetic fields and the occurrence
of cancers (or the percentage attributable to this exposure in the occurrence of these
pathologies) takes on particular importance here, in that one of the main objectives of
monitoring will be to assess spatial and temporal trends. Low specificity combined with
rare morbidity (with an annual risk of 10 -5 ), would make the interpretation of these
trends all the more difficult.

· On the other hand, with regard to self-declared subjective disorders, which at present
remain ill-defined (headaches, attention or memory disorders, heat sensations, etc.),
epidemiological monitoring may be envisaged in order to:

· better characterise this phenomenon;

· measure and monitor changes in the scale of this problem over time;
· generate etio-pathogenic hypotheses

In this context, the first stage could be to set up a descriptive survey based on the active
collection of self-declared events among mobile phone users via an active information
system developed in conjunction with the operators. The results of this survey would make it
possible to characterise the phenomenon more precisely and identify particular groups of
users to include in later analytical-type epidemiological studies (case-control), or groups for
targeted monitoring.

Should the phenomenon be confirmed, it would then be necessary to set up a cohort-type

survey in order to study its predictive value from the point of view of various health problems,
e.g. neurological disorders. At a later date, depending on the results obtained, this study
could either be repeated over time or a register could be made for declaring these
complaints. If scientific knowledge were to confirm the reality of this phenomenon, historical
data would thus be available to facilitating the monitoring of trends over time. This would
take into account future increases in prevalence and changes in the conditions of exposure
to electromagnetic fields attributable to the use of mobile phones.

6.3 Funding and organising the research

Funding for these studies should include a large contribution from companies operating in
the mobile telephony sector. Both equipment manufacturers and operators profit from this
industry, and it is therefore logical that they should contribute to research on mobile phones.
Possible measures could consist of a tax, which would be regularly revised according to the
number of subscribers and the financial needs of the research. This tax could either be
imposed by the public authorities or based on voluntary contributions from the
manufacturers and operators.
The group of experts recommends that a large part or even the majority of funding should be
provided by the public authorities, who would thus keep control over the research. Research
financed half by manufacturer/operators and half by public funds (major research bodies,
Public Health departments European Commission) could therefore be envisaged.

Whatever the measures chosen, they must always guarantee the absolute independence of
the programming and project-selection committees, as well as the research teams. For this
reason, it is crucial that contributions from the manufacturers and operators do not interfere
with the choice of research topics and follow-up. This implies that funding from
manufacturers and operators should either be channelled through the State or a structure or
“foundation” under state control.

For this reason, it seems important to form permanent “committees” of experts on a national
and European level. These experts should be chosen from various scientific disciplines and
governmental administrations concerned with the issue. In accordance with the rules already
applied in other research programs, the experts involved in projects submitted to these
authorities for funding should be excluded from discussions on those projects. On a national
level, this committee could be made up of members appointed by the major scientific bodies
(CNRS, INSERM) and the Public Health department. It would be responsible for several

· writing up a regular public report on current knowledge;

· determining priority research themes, depending on the conclusions of the report;
· publishing calls for tender corresponding to these themes;
· allocating private funding in response to applications from research laboratories.

This type of allocation should ensure complete transparency with regard to the
management, nature, operation, and progress of the research, as well as the content of the
resulting scientific publications.

On an international level, the national committee would also play a role in proposing and co-
ordinating research programs in liaison with any European committee(s) in this field.

The current level of funding for research in this field, which, including all public and private
contributions, comes to 7 million French Francs (excluding salaries), should be continued for
at least 5 years. Funding should be available not only for laboratory studies, but also for
epidemiological research in this field, which does not yet have any specific funding.


(A) CADAS Colloquium (2000)

(C) Royal Society of Canada (1999)
(E) COST 244 bis (1999)
(K) 'McKinlay' (1996)
(S) 'Stewart' (2000)
(R) Recent papers

1. Adair, E.R. and B.W. Adams, Microwaves Induced Peripherical Vasodilatation in Squirrel Monkey.
Science, 1980. 207(4437): p. 1381-1383 (C).
2. Adair, E.R., et al., Sensation, Subtleties, and Standards: Synopsis of a Panel Discussion. Microwaves
and Thermoregulation. E. R. Adair, ed. New York: Academic Press:231-238, 1983: p. (S).
3. Adair, E., Thermal physiology of radiofrequency radiation interactions in animals and humans. NATO
ASI Series A: Life Sciences. Klauenberg BJ, Grandolfo M, Erwin N, eds. Radiofrequency Radiation
Standards. Biological Effects, dosimetry, epidemiology, and public health policy, 1990. 274: p. 245-
269 (K).
4. Adair, R.K., Effects of weak high-frequency electromagnetic fields on biological systems.
Radiofrequency Radiation y Radiation Standards (B J Klauenberg, M Grandolfo and D N Erwin, Eds).
New York, Plenum Press, 1994: p. 207 (S).
5. Adair, E.R., et al., Human Exposure at Two Radio Frequencies (450 and 2450 MHz) : Similarities and
Differences in Physiological Response. Bioelectromagnetics, 1999. 20:(Suppl 4): p. (Suppl 4):12-20
6. Adey, W.R., Tissue Interactions With NonIonizing Electromagnetic Fields. Physiol Rev, 1981. 61(2):
p. 435-514 (S).
7. Adey, W.R., S.M. Bawin, and A.F. Lawrence, Effects of weak amplitude-modulated microwave fields
on calcium efflux from awake cat cerebral cortex. Bioelectromagnetics, 1982. 3(3): p. 295-307 (S) (C).
8. Adey, W.R., The extracellular space and energetic hierarchies in electrochemical signalling between
cells. Charge and Field Effects in Biosystems 2 (M J Allen, S F Cleary and F M Hawkridge, Eds). New
York, Plenum Press, 1989: p. 264 (S).
9. Adey, W.R., Biological effects of electromagnetic fields. J Cell Biochem, 1993. 51(4): p. 410-416 (S).
10. Adey, W., Electromagnetics in biology and medicine. Matsumoto H, ed. Modern science. Oxford;
Oxford University Press, 1993: p. 227-45 (K).
11. Adey, W.R., et al., Brain tumour incidence in rats chronically exposed to digital cellular fields in an
initiation-promotion model (Meeting Abstract). Bioelectromagnetics Society, 18th Annual Meeting, 9-
14 June, Victoria, British Columbia, Canada, 1996: p. 27 (A) (C).
12. Adey, W., Amplitude modulation dependent bioeffects in RF fields: mobile communications and
potential health effects. Proceedings of the General Assembly of URSI, September 1996. Lille,
France, 1996: p. K31 (K).
13. Adey, W., Bioeffects of mobile communications fields; Possible mechanisms for cumulative dose.
Kuster N, Balzano Q and Lin J, eds. Progress in assessment of mobile communications safety.
Chapman Hill, New York, in Press, 1996: p. (K).
14. Adey, W.R., et al., Brain tumour incidence in rats chronically exposed to frequency-modulated (FM)
cellular phone fields (Meeting abstract). Second World Congress for Electricity and Magnetism in
Biology and Medicine, 8-13 June, Bologna, Italy, 1997: p. 109 (C).
15. Adey, W.R., et al., Spontaneous and nitrosurea-induced primary tumors of the central nervous system
in Fischer 344 chronically exposed to 836 MHz modulated microwaves. Radiat Res, 1999. 152(3): p.
293-302 (A) (E) (S).
16. Ahlbom, A., Some fundamental aspects of epidemiology with reference to research on magnetic fields
and cancer. Non-Ionizing Radiation. Proceedings: Third International Non-ionizing Radiation
Workshop, Baden, Austria, April 22-26, 1996, R. Matthes, ed., 1996: p. 17-27 (E).
17. Ahlbom, A. and M. Feychting, Re: Use of cellular phones and the risk of brain tumours: a case-control
study [Letter]. Int J Oncol, 1999. 15(5): p. 1045-1047 (S).

18. Alam, M.T., et al., Cytological effects of microwave radiation in Chinese hamster cells in vitro. Can J
Genet Cytol, 1978. 20(1): p. 23-30 (S) (C).

19. Albert, E.N. and S.M. De, Do Microwaves Alter Nervous System Structure? Ann NY Acad Sci, 1975.
247: p. 87-108 (C).
20. Albert, E.N., Light and electron microscopic observations on the blood-brain barrier after microwave
irradiation. In Symposium on Biological Effects and Measurements of Radiofrequency Microwaves.
Washington DC: DHEW (HEW Publications), FDA77-8026, 1977: p. p294-304 (C).
21. Albert, E.N.a.D., M., Histological observations on central nervous system, in Biological Effects of
Electromagnetic Waves. Johnson C.C. and Shore, M.L. Eds. HEW Publ. (FDA) 77-8010. Department
of Health, Education and Welfare, Rockville MD, 1976, 1977. 1: p. 299 (C).
22. Albert, E.N., Light and electron microscopic observations on the blood-brain barrier after microwave
irradiation. Symposium on Biological Effects and Measurement of Radiofrequency/Microwaves, 16-18
February 1977, Rockville, MD, D. G. Hazzard, ed. Sponsored by BRH and WHO Collaborating
Center for Standardization of Protection Against Nonionizing Radiation. HEW Publication (FDA) 77-
8026:, 1977: p. 294-304 (S).
23. Albert, E.N., Reversibility of microwave-induced blood-brain barrier permeability. Radio Sci, 1979.
14(6S): p. 323-327 (S).
24. Albert, E.N. and J.M. Kerns, Reversible microwave effects on the blood-brain barrier. Brain Res,
1981. 230(1/2): p. 153-164 (K) (S) (C).
25. Albert, E.N., et al., Effect of amplitude modulated 147 MHz radiofrequency radiation on calcium ion
efflux from avian brain tissue. Radiat Res, 1987. 109(1): p. 19-27 (S).
26. Allis, J.W. and B.L. Sinha-Robinson, Temperature-specific inhibition of human cell Na+/K+ ATPase by
2450 MHz microwave radiation. Bioelectromagnetics, 1987. 8(2): p. 203-212 (E) (C).
27. Alm, H. and L. Nilsson, Changes in driver behaviour as a function of handsfree mobile phones - a
simulator study. Accid Anal Prev, 1994. 26: p. 441 (S).
28. Alm, H. and L. Nilsson, The effects of a mobile telephone task on driver behaviour in a car following
situation. Accid Anal Prev, 1995. 27: p. 707 (S).
29. Altpeter, E., et al., Study on health effects of shortwave transmitter station of Schwarzenburg, Bern,
Switzerland. ( Major Report). Bern: Univ Bern; BEW Publication Series, 1995. Study No 55: p. (K).
30. Anane, R., et al., Effects of whole-body exposure to GSM microwaves on rat bearing DMBA-induced
tumours (Meeting abstract). Bioelectromagnetics Society, 21st Annual Meeting, 20-24 June, Long
Beach, CA, Abstract No. 15-6, 1999: p. 90 (E).
31. Andersen, J., et al., On the possible health effects related to GSM and DECT transmissions. Aalborg;
Aalborg University: Report to the European Commission, 1995: p. (K).
32. Anderson, V. and K.H. Joyner, Specific absorption rate levels measured in a phantom head exposed
to radio frequency transmissions from analog hand-held mobile phones. Bioelectromagnetics, 1995.
16(1): p. 60-69 (C).
33. Anderstam, B., et al., Studies of possible genetic effects in bacteria of high frequency electromagnetic
fields. Hereditas, 1983. 98(1): p. 11-32 (S) (C).
34. Antipenko, E.N. and I.V. Koveshnikova, Cytogenetic effects of microwaves of non-thermal intensity in
mammals. Dokl Akad Nauk SSSR, 1987. 296: p. 724-726 (S) (C).
35. Antonopoulos, A., H. Eisenbrandt, and G. Obe, Effects of high-frequency electromagnetic fields on
human lymphocytes in vitro. Mutat Res, 1997. 395(2-3): p. 209-214 (E) (S) (C).
36. Apollonio, F., G. D'Inzeo, and L. Tarricone, Modelling of neuronal cells exposed to RF fields from
mobile telecommunication equipment. Bioelectrochem Bioenerg, 1998. 47(2): p. 199-205 (E).
37. Apollonio, F., G. D'Inzeo, and L. Tarricone, Neuronal cells under electromagnetic exposure: a link
between microscopic and macroscopic. Proceedings of the Second World Congress for Electricity
and Magnetism in Biology and Medicine, Bologna, Italy, June 1997, F. Bersani, Ed ; Klumer Academic
Publishers, 1999: p. (E).
38. Arber, S.L. and J.C. Lin, Microwave enhancement of membrane conductance: effects of EDTA,
caffeine and tetracaine. Physiol Chem Phys Med NMR, 1984. 16(6): p. 469-475 (K) (S) (C).
39. Arber, S.L. and J.C. Lin, Microwave-induced changes in nerve cells: effects of modulation and
temperature. Bioelectromagnetics, 1985. 6(3): p. 257-270 (K) (S) (C).
40. Arkin, H., L.X. Xu, and K.R. Holmes, Recent developments in modeling heat transfer in blood
perfused tissues. IEEE Trans Biomed Eng, 1994. 41: p. 97 (S).
41. Armstrong, B., et al., Association between exposure to pulsed electromagnetic fields and cancer in
electric utility workers in Quebec, Canada and France. Am J Epidemiol, 1994. 140(9): p. 805-820 (C).
42. Asanami, S. and K. Shimono, High body temperature induces micronuclei in mouse bone marrow.
Mutat Res, 1999. 390: p. 79-83 (S) (C).

43. Athey, T.W. and B.A. Krop, Millimetre wave induction of lamda prophage - dependent on growth
medium? (Meeting abstract). Bioelectromagnetics Society, 2nd Annual Meeting, 14-18 September,
1980, San Antonio, TX. Bioelectromagnetics, 1980. 1(2): p. 241-241 (S).

44. Auvinen, M., et al., Ornithine decarboxylase activity is critical for cell transformation. Nature, 1992.
360: p. 355-358 (C).
45. Auvinen, M., et al., Ornithine decarboxylase- and ras-induced cell transformations: protein tyrosine
kinase inhibitors and role of pp130CAS. Mol. Cell Biol, 1995. 15: p. 6513-6525 (C).
46. Auvinen, M., et al., Human ornithine decarboxylase-overproducing NIH3T3 cells induce growing,
highly vascularized tumours in nude mice. Cancer Res, 1997. 54: p. 3016-3025 (C).
47. Averbeck, D., M. Dardalhon, and A.-J. Berteaud, Microwaves action in procaryotic and eucaryotic
cells and a possible interaction with x-rays (Symposium Summary). J Microw Power, 1976. 11(2): p.
143-144 (S) (C).
48. Bach Andersen, J., et al., On the possible health effects related to GSM and DECT transmissions.
Report of a contract with the European Commission, 1995: p. (E).
49. Balcer-Kubiczek, E.K. and G.H. Harrison, Evidence for microwave carcinogenesis in vitro.
Carcinogenesis, 1985. 6(6): p. 859-864 (K) (A) (S) (C).
50. Balcer-Kubiczek, E.K. and G.H. Harrison, Induction of neoplastic transformation in C3H/10T1/2 cells
by 2.45-GHz microwaves and phorbol ester. Radiat Res, 1989. 117(3): p. 531-537 (K) (A) (S) (C).
51. Balcer-Kubiczek, E.K. and G.H. Harrison, Neoplastic transformation of C3H/10T1/2 cells following
exposure to 120-Hz modulated 2.45-GHz microwaves and phorbol ester tumor promoter. Radiat Res,
1991. 126(1): p. 65-72 (K) (A) (E) (S) (C).
52. Balode, Z., Assessment of radio-frequency electromagnetic radiation by the micronucleus test in
Bovine peripheral erythrocytes. Science Total Environment, 1996. 180: p. 81-85 (S) (C).
53. Balzano, Q., N. Kuster, and J.C. Lin, Mobile Communications Safety. Mobile Communications Safety.
London: Chapman and Hall. 296 pp., 1997. 4: p. (A) (E).
54. Balzano, Q., Exposure metrics for RF epidemiology: cellular phone handsets. Radiat Prot Dosim;
Proceedings of an International Workshop : Exposure Metrics and Dosimetry for EMF Epidemiology;
National Radiation Protection Board, Chilton, UK, 7-9 September, 1998, 1999. 83(1-2): p. 165-169
55. Banerjee, R., A. Goldfeder, and J. Mitra, Sister chromatid exchanges and chromosome aberrations
induced by radio sensitizing agents in bone marrow cells of treated tumor-bearing mice. JNCI, 1983.
70: p. 517-521 (C).
56. Baranski, S., S.L. Arber, and J.C. Lin, Histological and histochemical effects of microwave irradiation
on the central nervous system of rabbits and guinea pigs. Am J Physiol Med, 1972. 51: p. 182-190 (K)
(S) (C).
57. Baranski, S., et al., Influence of microwaves on genetical processes ofAspergillus nidulans. J. Microw.
Power, 1976. 11: p. 146-147 (C).
58. Baranski, S. and P. Czerski, Biological effects of microwaves. Stroudsberg, USA; Dowden,
Hutchinson, Ross, 1976: p. (K).
59. Barbaro, V., et al., Do European GSM mobile cellular telephones pose potential risk to pacemaker
patients? Pacing Clin Electrophysiol, 1995. 18: p. 1218-1224 (K).
60. Bardou, A., Avis relatif aux téléphones cellulaires. 13ème Rapport de la Commission de la Sécurité
des Consommateurs au Président de la République et au Parlement, Direction des Journaux officiels,
1997: p. 107-195 (A).
61. Barnett, S., CSIRO report on the status of research on the biological effects and safety of
electromagnetic radiation: telecommunications frequencies. Chatswood, Australia; CSIRO, 1994: p.
62. Bassen, H.I., et al., Evaluation of RF interference with medical devices from mobile communication
transceivers (Meeting abstract). Bioelectromagnetics Society, 16th Annual Meeting, 12-17 June,
Copenhagen, Denmark, 1994: p. 66-67 (K).
63. Bassett, A., Therapeutical uses of electric and magnetic fields in orthopedics. Biological Effects of
Electric and Magnetic Fields : Beneficial and Harmful Effects. D. O. Carpenter and S. Ayrapetyan,
eds., San Diego: Academic Press, 1994. II: p. 13-48 (K).
64. Baum, S.J., et al., Biological measurements in rodents exposed continuously throughout their adult
life to pulsed electromagnetic radiation. Health Phys, 1976. 30(2): p. 161-166 (S).
65. Bawin, S.M., R.J. Gavalas-Medici, and W.R. Adey, Effects of modulated very high frequency fields on
specific brain rhythms in cats. Brain Res, 1973. 58(2): p. 365-384 (S).
66. Bawin, S.M., R.J. Gavalas-Medici, and W.R. Adey, Reinforcement of transient brain rhythms by
amplitude-modulated VHF fields. Biological and Clinical Effects of Low Frequency Magnetic and

Electric Fields (J G Llaurado, A Sances and J H Battocletti, Eds). Springfield, Charles C Thomas,
1974: p. 172 (S).
67. Bawin, S.M., L.K. Kaczmarek, and W.R. Adey, Effects of modulated VHF fields on the central nervous
system. Ann N Y Acad Sci, 1975. 247: p. 74-81 (K) (S) (C).

68. Bawin, S.M. and W.R. Adey, Sensitivity of calcium binding in cerebral tissue to weak environmental
electric fields oscillating at low frequency. Proc Natl Acad Sci U S A, 1976. 73(6): p. 1999-2003 (C).
69. Bawin, S.M., Sheppard, A. and Adey, W.R., Possible mechanisms of weak electromagnetic field
coupling in brain tissue. Bioelectrochem Bioenerg, 1978. 5: p. 67-76 (C).
70. Bawin, S.M., W.R. Adey, and I.M. Sabbot, IIonic factors in release of 45Ca2+ from chicken cerebral
tissue by electromagnetic fields. Proc Natl Acad Sci U S A, 1978. 75(12): p. 6314-6318 (K).
71. Beaglehole, R., R. Bonita, and T. Kjellström, Basic epidemiology. Geneva: World Health Organization,
1993: p. 175 (E).
72. Beall, C., et al., Brain tumors among electronics industry workers. Epidemiology, 1996. 7: p. 125-130
73. Beechey, C.V., et al., Cytogenetic effects of microwave irradiation on male germ cells of the mouse.
Int J Radiat Biol, 1986. 50(5): p. 909-918 (K) (A) (S) (C).
74. Bell-Fernandez C, P.G., Cleveland JL., The ornithine decarboxylase gene is a transcriptional target of
c-myc. Proc. Natl. Acad. Sci. USA, 1993. 90: p. 7804-7808 (C).
75. Belyaev, I.Y., Some biophysical aspects of the genetic effect of low-intensity millimeter waves.
Bioelectrochem Bioenerg, 1992. 27(1): p. 11-18 (S).
76. Berman, E., J.B. Kinn, and H.B. Carter, Observations of mouse fetuses after irradiation with 2.45 GHz
microwaves. Health Phys, 1978. 35: p. 791 (S).
77. Berman, E., H.B. Carter , and D. House, Tests for mutagenesis and reproduction in male rats
exposed to 2450 MHz (CW) microwaves. Bioelectromagnetics, 1980. 1: p. 65-76 (S) (C).
78. Bernardi, P., M. Cavagnaro, and S. Pisa, Evaluation of the SAR distribution in the human head for
cellular phones used in a partially closed environment. IEEE Trans Electromagn Compat, 1996. 38(3):
p. 357-366 (E).
79. Bernardi, P., et al., Evaluation of SAR and temperature distribution in subjects exposed in the far-field
of radiating radio frequency sources. Proc. Int. Symp. on Electromagnetic Compatibility, Rome (Italy),
1998: p. 190-193 (E).
80. Bernardi, P., et al., A cell model to evaluate EM field absorption in biological samples: a sensitivity
and relevance analysis (Meeting abstract). European Bioelectromagnetics Assoc. (EBEA), 4th
International Congress, 19-21 November, Zagreb, Croatia, 1998: p. (E).
81. Bernardi, P., et al., Evaluation of the power absorbed in subjects exposed to EM fields in partially
closed environments by using a combined analytical-FDTD method. Proceedings IEEE MTT-S
International Microwave Symposium, Anaheim, California, 1999: p. 599-602 (E).
82. Bernardi, P., et al., SAR and temperature increases induced in the head of a cellular phone user
(Meeting abstract). Proceedings of the 26th General Assembly of URSI, Toronto, Canada, 1999: p.
83. Berteaud, A.J., et al., Action d'un rayonnement electromagnetique sur la croissance bacterienee
(Effect of an electromagnetic microwave radiation on the growth of bacteria). C R Acad Sci Paris,
1975. 281: p. 843-846 (C).
84. Betancur, C., G. Dell'Omo, and E. Alleva, Magnetic field effects on stress-induced Analgesia in mice:
Modulation of light. Neurosci Lett, 1994. 182(2): p. 147-150 (C).
85. Bianco, B., et al., Enhancement of the interaction between low-intensity R.F. e.m. fields and ligand
binding due to cell basal metabolism. Wireless Networks, 1997. 3: p. 477-487 (E).
86. Bisht, K., et al., The effect of radiofrequency radiation with modulation relevant to cellular phone
communication (835.62 MHz FDMA and 847.74 MHz CDMA) on the induction of micronuclei in C3H
10T1/2 cells. Bioelectromagnetics Society, 22th Annual Meeting, Munich, juin 2000, 2000: p. (R).
87. Blackman, C.F., M.C. Surles, and S.G. Benane, The effect of microwave exposure on bacteria:
mutation induction. Biological Effects of Electromagnetic Waves. Selected papers of the UNSC/URSI
Annual Meeting, Boulder, Colorado, October 1975 (C C Johnson and M L Shore, Eds). Rockville,
Maryland, US Department of Health, Education and Welfare, 1976. 1: p. 406-413 (S) (C).
88. Blackman, C.F., et al., Induction of calcium-ion efflux from brain tissue by radiofrequency radiation:
effects of modulation frequency and field strength. Radio Sci, 1979. 14 (S): p. 93-98 (K) (S) (C).
89. Blackman, C.F., et al., Induction of calcium-ion efflux from brain tissue by radiofrequency radiation:
effect of sample number and modulation frequency on the power-density window.
Bioelectromagnetics, 1980. 1(1): p. 35-43 (K) (S) (C).

90. Blackman, C.F., et al., Calcium-ion efflux from brain tissue: power-density versus internal field-
intensity dependencies at 50-MHz RF radiation. Bioelectromagnetics, 1980. 1(3): p. 277-283 (S) (C).

91. Blackman, C., et al., Induction of calcium ion efflux from brain tissue by radiofrequency radiation:
effect of sample number and modulation frequency on the power-density window.
Bioelectromagnetics, 1980. 12: p. 173-182 (C).

92. Blackman, C., Genetics and mutagenesis. Biological Effects of Radiofrequency Radiation, EPA-
600/8-83-026F; edited by Elder JA, Cahill DF. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, 1984: p. 5-94-5-105 (C).
93. Blackman, C.F., et al., Multiple power density windows and their possible origin. Bioelectromagnetics,
1989. 10(2): p. 115 (S).
94. Blackman, C.F., S.G. Benane, and D.E. House, The influence of temperature during electric and
magnetic field induced alteration of calcium-ion release from in vitro brain tissue. Bioelectromagnetics,
1991. 12(3): p. 173-182 (K) (C).
95. Blackshear, P., et al., High level, cell-specific expression of ornithine decarboxylase transcripts in rat
genitourinary tissues. Mol. Endocrinol, 1989. 3: p. 68-78 (C).
96. Blackwell, R.P. and R.D. Saunders, The effects of low-level radiofrequency and microwave radiation
on brain tissue and animal behaviour. Int J Radiat Biol, 1986. 50(5): p. 761-787 (S).
97. Blessing, M., et al., Transgenic mice as a model to study the role of TGF-b related molecules in hair
follices. Genes Deve, 1993. 7: p. 204-215 (C).
98. Blevins, R.D., et al., The effects of microwave radiation and heat on specific mutants of Salmonella
typhimurium LT2. Radiat Res, 1980. 82(3): p. 511-517 (S) (C).
99. Boheim, G.H., et al., Evidence for membrane bound water interaction with RF-fields (Meeting abstract
). Bioelectromagnetics Society, 21st Annual Meeting, 20-24 June, Long Beach, CA,, 1999: p. Abstract
No. 7-4, 46 (E).
100. Bohr, H. and J. Bohr, Microwave Enhanced Kinetics Observed in ORD Studies of a Protein.
Bioelectromagnetics, 2000. 21(1): p. 68-72 (S).
101. Borbely, A.A., et al., Pulsed high-frequency electromagnetic fields affect human sleep and sleep
electroencephalogram. Neurosci Lett, 1999. 275(3): p. 207-210 (A) (S).
102. Bornhausen, M. and H. Sheingraber, Prenatal exposure to 900 MHz, cell-phone electromagnetic
fields had no effect on operant-behavior performances of adult rats. Bioelectromagnetics, 2000. 21(1):
p. 1-9 (R).
103. Bortkiewicz, A., et al., Ambulatory ECG monitoring in workers exposed to electromagnetic fields. J
Med Eng Technol, 1997. 21(2): p. 41-46 (C).
104. Bosisio, A.V., J. Wiart, and A. Le Cornec, Preliminary analysis of the influence of system parameters
on RF exposure with GSM mobile phones (Meeting abstract). Bioelectromagnetics Society, 21st
Annual Meeting, 20-24 June, Long Beach, CA, Abstract No. P-137, 1999: p. 156-157 (E).
105. Bracken, T.D., L.I. Kheifets, and S.S. Sussman, Exposure assessment for power frequency electric
and magnetic fields (EMF) and its application to epidemiologic studies. J Expo Anal Environ
Epidemiol, 1993. 3(1): p. 1-22 (E).
106. Braune, S., et al., Resting blood pressure increase during exposure to a radio-frequency
electromagnetic field. Lancet, 1998. 351(9119): p. 1857-1858 (A) (E) (S).
107. Braune, S., et al., Radiofrequency electromagnetic field from mobile phones. lancet, 1998. 352: p. 576
108. Brookhuis, K.A., G. De Vries, and D. de Waard, The effects of mobile telephoning on driving
performance. Accid Anal Prev, 1991. 23: p. 309 (S).
109. Brown, D., A.H. Tickner, and D.C.V. Simmonds, Interference between concurrent tasks of driving and
telephoning. J Appl Psychol, 1969. 53: p. 419 (S).
110. Brown, R.F. and S.V. Marshall, Sister chromatid exchange in marrow cells of mice exposed to RFR
(400, 800, and 1200 MHzCW) (meeting abstract). Bioelectromagnetics Society (BEMS) "4 th
Bioelectromagnetics Society Annual Meeting - Abstracts." Los Angeles, June 28 - July 2, 1982, 1982.
Abstr. No. G31: p. (C).
111. Brown, R.F. and S.V. Marshall, Differentiation of murine erythroleukemic cells during exposure to
microwave radiation. Radiat Res, 1986. 108(1): p. 12-22 (K).
112. Brusick, D.J., An assessment of the genotoxic activity of radiofrequency radiation (Meeting abstract).
State of the Science Colloquium, sponsored by Wireless Technology Research, L.L.C. and The
International Committee for Wireless Communication Health Research, 13-15 November, Rome, Italy,
1995: p. (K).
113. Brusick, D., et al., Genotoxicity of radiofrequency radiations. Environ Mol Mutagen, 1998. 32: p. 1-16
(A) (E) (S) (C).

114. Brück, K. and P. Hinckel, Thermoafferent networks and their adaptative modifications. Encycl
Pharmacol Ther, 1990. 6: p. 129 (S).

115. Budd, R.A. and P. Czerski, Modulation of mammalian immunity by electromagnetic radiation. J
Microw Power, 1985. 20(4): p. 217-231 (K).
116. Burkhardt, M., et al., Numerical and experimental dosimetry of Petri dish exposure setups.
Bioelectromagnetics, 1996. 17(6): p. 483-493 (E).

117. Burkhardt, M. and N. Kuster, E-simulation of artifacts caused by FDTD-staircasing in RF simulation

invoiving dielectric objects. Electromagnetics and Guided Waves, 2000. (in press): p. (A).
118. Byus, C.V., et al., Alterations in protein kinase activity following exposure of cultured human
lymphocytes to modulated microwave fields. Bioelectromagnetics, 1984. 5(3): p. 341-351 (K).
119. Byus, C.V., S.E. Pieper, and W.R. Adey, The effects of low-energy 60-Hz environmental
electromagnetic fields upon the growth-related enzyme ornithine decarboxylase. Carcinogenesis,
1987. 8(10): p. 1385-1389 (C).
120. Byus, C.V., et al., Increased ornithine decarboxylase activity in cultured cells exposed to low energy
modulated microwave fields and phorbol ester tumor promoters. Cancer Res, 1988. 48(15): p. 4222-
4226 (K) (A) (S) (C).
121. Byus, C. and V. Wu, The level of substrate ornithine can alter polyamine-dependent DNA synthesis
following phorbolester stimulation of cultured hepatoma cells. J. Cell. Physiol, 1991. 149: p. 9-17 (C).
122. Byus, C.V. and L. Hawel, Additional considerations about bioeffects. Mobile Communications Safety
(Q Balzano and J C Lin, Eds). London, Chapman and Hall, 1997: p. 133-144 (S) (C).
123. Cain, C.D., et al., Pulsed electromagnetic field modifies cAMP metabolism and ornithine
decarboxylase activity in primary bone cells. International Conference on Electric and Magnetic Fields
in Medicine and Biology. Conference Publication number 257 (Institute of Electrical Engineers,
London and New York):, 1985: p. 9-134 (C).
124. Cain, C.D., D.L. Thomas, and W.R. Adey, 60 Hz magnetic field acts as co-promoter in focus formation
of C3H/lOTl/2 cells. Carcinogenesis, 1993. 14: p. 955-960 (C).
125. Cain, C.D., D.L. Thomas, and W.R. Adey, Focus formation of C3H/10T1/2 cells and exposure to a
836.55 MHz Modulated Radiofrequency field. Bioelectromagnetics, 1997. 18(3): p. 237-243 (A) (E)
(S) (C).
126. Cairnie, A.B. and R.K. Harding, Cytological studies in mouse testis irradiated with 2.45-GHz
continuous-wave microwaves. Radiat Res, 1981. 87(1): p. 100-108 (C).
127. Calle, E.E. and D.A. Savitz, Leukemia in occupational groups with presumed exposure to electrical
and magnetic fields (Letter). N Engl J Med, 1985. 313(23): p. 1476-1477 (K).
128. Canellakis, E., et al., Identification, cloning, and nucleotide sequencing of the ornithine decarboxylase
antizyme gene of Escherichia coli. Proc. Natl. Acad. Sci. USA, 1993. 90: p. 7129-7133 (C).
129. Cantor, K.P., et al., Re: Breast cancer mortality among female electrical workers in the United States
(Letter). J Natl Cancer Inst, 1995. 87(3): p. 227-228 (S).
130. Cantor, K.P., et al., Occupational exposures and female breast cancer mortality in the United States.
J Occup Environ Med, 1995. 37(3): p. 336-348 (C).
131. Cao, G., L.-M. Liu, and S.F. Cleary, Cell cycle alterations induced by isothermal 27 MHz radio-
frequency radiation exposure. Bioelectrochem Bioenerg, 1995. 37: p. 131-140 (E).
132. Cardis, E. and M. Kilkenny, International case-control study of adult brain, head and neck tumours:
results of the feasibility study. Radiat Prot Dosim; Proceedings of an International Workshop,
Exposure Metrics and Dosimetry for EMF Epidemiology, National Radiation Protection Board, Chilton,
UK, 7-9 September, 1998, McKinlay, M.H. Repacholi (Eds.), 1999. 83((1-2)): p. 179-183 (E).
133. Carillo, R., et al., Preliminary observations on cellular telephones and pacemakers. Pacing Clin
Electrophysiol, 1995. 18: p. 863 (K).
134. Carlo, G.L., I. Munro, and A.W. Guy, Potential public health risks from wireless technology. Research
agenda for the development of data for science-based decisionmaking. Scientific Advisory Group on
Cellular Telephone Res (504 pp.), 1994: p. 87-95 (K).
135. Carpenter, R.L., Ocular effects of microwave radiation. Bull NY Acad Sci, 1979. 55: p. 1048 (S).
136. Castillo, M. and R.M. Quencer, Sublethal exposure to microwave radar (Letter). JAMA, 1988. 259(3):
p. 355 (C).
137. CENELEC, Human exposure to electromagnetic fields: high frequency (10 kHz to 300 GHz).
European Committee for Electrotechnical Standardization; European Prestandard, 1995. ENV
500166-2: p. (K).
138. Chadwick, P.J., Occupational exposures to electromagnetic fields: the practical application of NRPB
guidelines. Chilton, NRPB-R301, 1998: p. (S).

139. Chagnaud, J.-L., B. Veyret, and B. Despres, Effects of pulsed microwaves on chemically induced
tumours in rats (Meeting abstract ). Bioelectromagnetics Society, 17th Annual Meeting, 18-22 June,
Boston, MA, 1995: p. 28 (C).

140. Chagnaud, J.-L., B. Despres , and B. Veyret , Effects of pulsed microwaves on the immune system
and on chemically-induced tumors in rats. Simunic D, ed. Proceedings of COST 244 workshop on
biological effects of amplitude modulated microwaves; Kuopio, Finland, 1995: p. (K).
141. Chagnaud, J.-L., J.-M. Moreau, and B. Veyret, No effect of short-term exposure to GSM-modulated
low-power microwaves on benzo(a )pyrene-induced tumours in rat. Int J Radiat Biol, 1999. 75(10): p.
1251-1256 (E) (S).

142. Chagnaud, J.L. and B. Veyret, No effects of GSM-modulated microwaves on immune parameters in
rats. Int. J. Radiat. Biol, 1999. 75: p. 111-113 (E).
143. Chapman, S. and W.N. Schofield, Lifesavers and Samaritans: emergency use of cellular (mobile)
phones in Australia. Accid Anal Prev, 1998. 30: p. 815 (S).
144. Chapman, S. and W.N. Schofield, Emergency use of cellular (mobile) telephones. Lancet, 1998. 351:
p. 650 (S).
145. Chazan, B., et al., Effects of microwave exposure in utero on embryonal, foetal and postnatal
development of mice. Biol. Neonate, 1983. 44: p. 339-347 (C).
146. Chen, K.M., A. Samuel, and R. Hoopingarner, Chromosomal aberration of living cells induced by
microwave radiation. Environ Lett, 1974. 6: p. 37-46 (S).
147. Chernovetz, M.E., D.R. Justesen, and A.F. Oke, A teratological study of the rat: microwave and
infrared radiations compared. Radio Sci, 1977. 12S: p. 191 (S).
148. Cherry, N., Criticism of the health assessment in the ICNIRP guidelines for radiofrequency and
microwave radiation (100 kHz-300 GHz). www.emfguru.com/CellPhone/cherry2/ICNIRP-2.htm, 2000:
p. (R).
149. Chia, S.-E., H.-P. Chia, and J.-S. Tan, Prevalence of headache among handheld cellular telephone
users in Singapour: a community study. Environ. Health Perspective, 2000. 108: p. 1059-1062 (R).
150. Chiabrera, A., et al., Quantum analysis of ion binding kinetics in electromagnetic bioeffects.
Electromagnetics in Medicine and Biology. C. T. Brighton, S. R. Pollack, eds. San Francisco, CA: San
Francisco Press, 1991: p. 27-31 (K).
151. Chizhenkova, R.A. and A.A. Safroshkina, Electrical reactions of brain to microwave irradiation. Electro
Magnetobiol, 1996. 15(3): p. 253-258 (A) (S).
152. Chou, C.-K., et al., Cochlear microphonics generated by microwave pulses. J Microw Power, 1975.
10(4): p. 361-367 (K).
153. Chou, C.-K. and A.W. Guy, Effects of electromagnetic fields on isolated nerve and muscle
preparation. IEEE Trans Microw Theory Tech, 1978. 26(3): p. 141-147 (K) (S).
154. Chou, C.-K. and A.W. Guy, Microwave-induced auditory responses in guinea pigs: relationship of
threshold and microwave-pulse duration. Radio Sci, 1979. 14(6S): p. 193-197 (K) (S) (C).
155. Chou, C.-K. and A.W. Guy, Carbon-loaded Teflon electrodes for chronic EEG recordings in
microwave research. J Microw Power, 1979. 14(4): p. 399-404 (K).
156. Chou, C.-K., A.W. Guy, and R. Galambos, Auditory perception of radiofrequency electromagnetic
fields. J Acoust Soc Am, 1982. 71(6): p. 1321-1334 (K) (S) (C).
157. Chou, C.-K., et al., Chronic exposure of rabbits to 0.5 and 5 mW/cm2 2450 MHz CW microwave
radiation. Bioelectromagnetics, 1983. 4(1): p. 63-77 (K) (S) (C).
158. Chou, C.-K., K.-C. Yee, and A.W. Guy, Auditory response in rats exposed to 2450 MHz
electromagnetic fields in a circularly polarized waveguide. Bioelectromagnetics, 1985. 6(3): p. 323-
326 (K) (S) (C).
159. Chou, C.-K., et al., Effects of long-term radiofrequency radiation on immune competence and
metabolism. USAF Sch. of Aerospace Medicine. Report USAFSAM-TR-85-105, 1986: p. 120 pp. (C).
160. Chou, C.-K., et al., Long-term, low-level microwave irradiation of rats. Bioelectromagnetics, 1992.
13(6): p. 469-496 (K) (A) (S) (C).
161. Chou, C.K., J.A. McDougall, and K.W. Chan, Absence of radiofrequency heating from auditory
implants during magnetic resonance imaging. Bioelectromagnetics, 1995. 16(5): p. 307-316 (K).
162. Chou, C.K., et al., Radio Frequency Electromagnetic Exposure: Tutorial Review on Experimental
Dosimetry. Bioelectromagnetics, 1996. 17(3): p. 195-208 (K) (E).
163. Chou, C.K., et al., Quantification of cellular telephone radio frequency dosimetry. Final Report
supported by Wireless Technology Research, USA, 1998: p. (E).
164. Chou, C.K., et al., Development of a rat head exposure system for simulating human exposure to RF
fields from handheld wireless telephones. Bioelectromagnetics, 1999. 20(Suppl 4): p. (Suppl 4):75-92

165. Ciaravino, V., M.L. Meltz, and D.N. Erwin, Effects of radiofrequency radiation and simultaneous
exposure with mitomycin C on the frequency of sister chromatid exchanges in Chinese hamster ovary
cells. Environ. Mutagen, 1987. 9: p. 393-399 (K) (S) (C).

166. Ciaravino, V., M.L. Meltz, and D.N. Erwin, Absence of a synergistic effect between moderate-power
radiofrequency electromagnetic radiation and adriamycin on cell-cycle progression and sister-
chromatid exchange. Bioelectromagnetics, 1991. 12(5): p. 289-298 (K) (S) (C).
167. Cleary, S.F., L.M. Liu, and F. Garber, Viability and phagocytosis of neutrophils exposed in vitro to 100
MHz radiofrequency radiation. Bioelectromagnetics, 1985. 6(1): p. 53-60 (K).
168. Cleary, S.F., Cellular effects of radiofrequency electromagnetic fields. Biological Effects and Medical
Applications of Electromagnetic Energy (O P Gandhi, Ed). Englewood Cliffs, New Jersey, Prentice-
Hall, 1990: p. 339-356 (S) (C).

169. Cleary, S.F., Biological effects of radiofrequency electromagnetic fields. Biological Effects and
Medical Applications of Electromagnetic Energy. O. P. Gandhi, ed., Englewood Cliffs, NJ: Prentice
Hall, 1990: p. 236-255 (S) (C).
170. Cleary, S.F., L.-M. Liu, and R.E. Merchant, Glioma proliferation modulated in vitro by isothermal
radiofrequency radiation exposure. Radiat Res, 1990. 121(1): p. 38-45 (K) (S) (C).
171. Cleary, S.F., L.-M. Liu, and R.E. Merchant, In vitro lymphocyte proliferation induced by radiofrequency
electromagnetic radiation under isothermal conditions. Bioelectromagnetics, 1990. 11(1): p. 47-56 (K)
(S) (C).
172. Cleary, S.F., Effects of radiofrequency radiation on mammalian cells and biomolecules in vitro.
Advances in Chemistry Series 250: Electromagnetic Fields: Biological Interactions and Mechanisms.
M. Blank, ed., Washington, DC: American Chemical Society, 1995: p. 467-477 (S) (C).
173. Cleary, S.F., G. Cao, and L.-M. Liu, Effects of isothermal 2.45 GHz microwave radiation on the
mammalian cell cycle: comparison with effects of isothermal 27 MHz radiofrequency radiation
exposure. Bioelectrochem Bioenerg, 1996. 39: p. 167-173 (K) (E) (S) (C).
174. Cleary, S.F., et al., Effect of isothermal radiofrequency radiation on cytolytic T lymphocytes. FASEB J,
1996. 10(8): p. 913-919 (E).
175. Cleveland, R.F. and T.W. Athey, Specific Absorption Rate in models of the human head exposred to
hand-held UHF portable radios. Bioelectromagnetics, 1989. 10: p. 173-186 (C).
176. Clifford, A., et al., Role of ornithine decarboxylase in epidermal tumorigenesis. Cancer Res, 1995. 55:
p. 1680-1686 (C).
177. Cohen, S.M. and L.B. Ellwein, Genetic errors, cell proliferation and carcinogenesis. Cancer Res,
1991. 51: p. 6493 (S).
178. Cohen, B.H., et al., Parental factors in Down's syndrome-results of the second Baltimore case-control
study. In Population Cytogenetics. Ed. Hook, E.B., Porter, I.H. (New York: Academic Press), 1997: p.
301-352 (C).
179. Coleman, M., J. Bell, and R. Skeet, Leukaemia incidence in electrical workers (Letter). Lancet, 1983.
1(8331): p. 982-983 (K).
180. Conover, D., C. Moss, and C. Kardous, Health hazard evaluation for the Cardinal Pacelli School,
Cincinnati, Ohio. Letter report. National Institute for Occupational Safety and Health, Cincinnati, Ohio.,
1997: p. (C).
181. Corelli, J.C., et al., Effects of 2.6-4.0 Ghz microwave radiation on E-coli B. J Microw Power, 1977.
12(2): p. 141-144 (C).
182. COST 244bis, Proceedings of the Workshop on exposure set-ups and their dosimetry, Zürich. 1999:
p. (E).
183. Coughlin, C.T., et al., Interstitial hyperthermia in combination with brachytherapy. Radiology, 1983.
148: p. 285-288 (C).
184. Cui, W., et al., TGFa1 inhibits the formation of benign skin tumors, but enhances progression to
invasive spindle carcinomas in transgenic mice. Cell, 1996. 86: p. 531-542 (C).
185. Cullinan, W.E., et al., Pattern and time course of immediate early gene expression in rat brain
following acute stress. Neuroscience, 1995. 64: p. 477 (S).
186. Czerska, E.M., et al., Effects of continuous and pulsed 2450 MHz radiation exposure on spontaneous
transformation of human lymphocytes in vitro. Bioelectromagnetics, 1992. 13(4): p. 247-259 (K).
187. d'Ambrosio, G., et al., Genotoxic effects of amplitude-modulated microwaves on human lymphocytes
exposed in vitrounder controlled conditions. Electro Magnetobiol, 1995. 14(3): p. 157-164 (K) (E) (C).
188. D'Andrea, J.A., O.M. Gandhi, and R.P. Kesner, Behavioral effects of resonant electromagnetic power
absorption in rats. Biological Effects of Electromagnetic Waves. Selected papers of the UNSC/URSI
Annual Meeting, Boulder, Colorado, October 1975 (C C Johnson and M L Shore, Eds). Rockville,
Maryland, US Department of Health, Education and Welfare, 1976. 1: p. 257 (S).

189. D'Andrea, J.A., O.P. Gandhi, and J.L. Lords, Behavioral and thermal effects of microwave radiation at
resonant and non-resonant wavelengths. Radio Sci, 1977. 12(6S): p. 251-256 (S).
190. D'Andrea, J.A., et al., Physiological and behavioral effects of chronic exposure to 2450 MHz
microwaves. J Microw Power, 1979. 14(4): p. 351-362 (S).

191. D'Andrea, J.A., et al., Physiological and behavioral effects of prolonged exposure to 915 MHz
microwaves. J Microw Power, 1980. 15(2): p. 123-135 (S).
192. D'Andrea, J.A., et al., Behavioral and physiological effects of chronic 2450 MHz microwave irradiation
of the rat at 0.5 mW/cm 2. Bioelectromagnetics, 1986. 7(1): p. 45-56 (S).
193. D'Andrea, J.A., et al., Intermittent exposure of rats to 2450 MHz microwaves at 2.5 mW/cm 2 .
Behavioural and physiological effects. Bioelectromagnetics, 1986. 7(3): p. 315-328 (S).
194. D'Andrea, J.A., B.L. Cobb, and J.O. de Lorge, Lack of behavioural effects in the rhesus monkey: high
peak microwave pulses at 1.3 GHz. Bioelectromagnetics, 1989. 10(1): p. 65-76 (C).

195. D'Andrea, J.A., Microwave radiation absorption: behavioural effects. Health Phys, 1991. 61(1): p. 29-
40 (S) (C).
196. D'Andrea, J.A., et al., Rhesus monkey contrast sensitivity during exposure to high peak power 5.6-
GHz microwave pulses (Meeting abstract). First World Congress for Electricity and Magnetism in
Biology and Medicine, 14-19 June, Lake Buena Vista, FL, 1992: p. 62 (S).
197. D'Andrea, J.A., A. Thomas, and D.J. Hatcher, Rhesus monkey behaviour during exposure to high-
peak-power 5.62 GHz microwave pulses. Bioelectromagnetics, 1994. 15(2): p. 163-176 (C).
198. D'Andrea, J.A., Behavioral evaluation of microwave irradiation. Bioelectromagnetics, 1999. 20(Suppl
4): p. (Suppl 4), 64-74 (S).
199. D'Inzeo, G., et al., Microwave effects on acetylcholine-induced channels in cultured chick myotubes.
Bioelectromagnetics, 1988. 9(4): p. 363-372 (S).
200. Daniells, C., et al., Transgenic nematodes as biomonitors of microwave-induced stress. Mutat Res,
1998. 399(1): p. 55-64 (S).
201. Dardalhon, M., D. Averbeck, and A. Berteaud, Action des ondes centimetriques seules ou combinees
avec les rayons ultra-violets sur les cellules eucaryotiques. Ondes electromagnetiques et biologie,
Proceedings of International Symposium; 30 June B 4 July; Jouy en Josas, France, 1980: p. 17-24
202. Dardalhon, M., D. Averbeck, and A.J. Berteaud, Studies on possible genetic effects of microwaves in
prokaryotic and eucaryotic cells. Radiat Environ Biophys, 1981. 20(1): p. 37-51 (K) (S) (C).
203. Dardalhon, M., et al., Thermal aspects of biological effects of microwaves in Saccharomyces
cerevisiae. Int J Radiat Biol, 1985. 48(6): p. 987-996 (K).
204. Davanipour, Z., et al., Amyotrophic Lateral Sclerosis and occupational exposure to electromagnetic
fields. Bioelectromagnetics, 1997. 18(1): p. 28-35 (C).
205. Davis, R.L. and F.K. Mostofi, Cluster of testicular cancer in police officers exposed to hand-held radar.
Am J Ind Med, 1993. 24(2): p. 231-233 (S) (C).
206. de Lorge, J.O. and C.S. Ezell, Observing responses of rats exposed to 1.28- and 5.62-GHz
microwaves. Bioelectromagnetics, 1980. 1(2): p. 183-198 (S).
207. de Lorge, J.O., Operant behavior and colonic temperature of Macaca mulatta exposed to radio
frequency fields at and above resonant frequencies. Bioelectromagnetics, 1984. 5(2): p. 233-246 (S)
208. de Pomerai, D.I., et al., Mobile phone effects on nematode worms: animal in vivo studies using
Caenorhabditis elegens. Presented at Mobile Phones and Health Conference, September 1999,
Gothenburg, Sweden, 1999: p. 67-78 (A) (S).
209. de Pomerai, D., et al., Non-thermal heat shock response to microwaves. Nature, 2000. 25 may: p.
210. de Seze, R., et al., Effects of microwaves on rat neurotransmitters. Final results. Proceedings of the
Third EBEA Congress, Nancy France, 1996: p. (E).
211. de Seze, R., et al., Effects of microwaves emitted by mobile phones on hypophyseal hormones in
humans (Meeting Abstract). Bioelectromagnetics Society, 18th Annual Meeting, 9-14 June, Victoria,
British Columbia, Canada, 1996: p. 123 (K) (A).
212. de Seze, R., et al., Effects on human of microwaves emitted by mobile phones: Chronological Rythm
of ACTH (Meeting Abstract). Bioelectromagnetics Society, 18th Annual Meeting, 9-14 June, Victoria,
British Columbia, Canada, 1996: p. 56-57 (A).
213. de Seze, R., P. Fabbro-Peray, and L. Miro, GSM radiocellular telephones do not disturb the secretion
of antepitultary hormones in humans. Bioelectromagnetics, 1998. 19(5): p. 271-278 (A) (E).

214. de Seze, R., et al., Etude chez l'homme de l'exposition aux champs électromagnétiques émis par les
téléphones mobiles GSM sur les potentiels évoqués auditifs et les produits de distorsion acoustiques.
Actes du 1er Congrès National de Radioprotection, 1998: p. 47-51 (A).
215. de Seze, R. and L. Miro, Mobile phone exposure : studies on the human endocrine system.
Proceedings of the 26th General Assembly of URSI, Toronto, Canada, 1999: p. (E).

216. de Seze, R., et al., Evaluation in humans of the effects of radiocellular telephones on the circadian
patterns of melatonin secretion, a chrono-biological rhythm maker. J Pineal Res, 1999. 27(4): p. 237-
242 (A) (E).
217. Dec, S., E. Cieslak, and J.S. Miszczak, Electroencephalographic Frequency Mapping in Healthy
Subjects During Cellular Head Telephone Stimulation. Proceedings of the Second World Congress for
Electricity and Magnetism in Biology and Medicine, Bologna, Italy, June 1997, F. Bersani, Ed ; Klumer
Academic Publishers, 1999: p. (E).
218. Del Seppia, C., et al., Exposure to oscillating magnetic fields influences sensitivity to electrical stimuli.
I: Experiments on pigeons. Bioelectromagnetics, 1995. 16(5): p. 290-294 (C).
219. Delannoy, J., et al., Hyperthermia system combined with a magnetic resonance imaging unit. Med
Phys, 1990. 17: p. 855-860 (C).

220. Demers, P.A., et al., Occupational exposure to electromagnetic fields and breast cancer in men. Am J
Epidemiol, 1991. 134(4): p. 340-347 (C).
221. DETR, Department of the Environment, Transport and the Regions and the National Assembly for
Wales Code of Best Practice.Telecommunications prior approval procedures as applied to mast/tower
development. 1998: p. (S).
222. DETR, Government targets mobile phone drivers in new campaign. Press notice 0041, 21 January
2000, 2000: p. (S).
223. DeWitt, J.R., et al., Behavioural effects of chronic exposure to 0.5 mW/cm 2 of 2,450-MHz
microwaves. Bioelectromagnetics, 1987. 8(2): p. 149-157 (C).
224. DG XIII, Recommandation du conseil du 12 juillet 1999 relative a la limitation de l'exposition du public
aux champs électromagnétiques. Proposals for a research Programme by a European Commission
Expert Group, 1999. L199, 59 (1999/519/EC).: p. 27 (A) (S).
225. Dhabhar, F. and B. McEwen, Enhancement of immune function by acute stress. Proceedings of
Society of Neuroscience Annual Meeting; San Diego, 1995: p. abstract 548.6. (K).
226. Diederich, C.J., et al., Evaluation of a multi-element microwave applicator for hyperthermia. Proc 9th
Int'l Conf Radiation Research, Toronto Canada, 1991. (unpublished): p. (C).
227. DiGiovanni, J., Multistage carcinogenesis in mouse skin. Pharmacol. Ther, 1992. 54: p. 63-128 (C).
228. DiGiovanni, J., et al., Evaluation of the possible copromotion effect of a 60 Hz magnetic field during
chemically induced carcinogenesis in skin of Sencar mice (Meeting abstract). Annual Review of
Research on Biological Effects of Electric and Magnetic Fields from the Generation, Delivery and Use
of Electricity, 17-21 November, San Antonio, TX, 1996: p. 6-7 (C).
229. Dimbylow, P.J. and S.M. Mann, SAR calculations in an anatomically realistic model of the head for
mobile communication transceivers at 900 MHz and 1.8 GHz. Phys Med Biol, 1994. 39(10): p. 1537-
1553 (K) (S).
230. Dolk, H., et al., Cancer incidence near radio and television transmitters in Great Britain. 1: Sutton
Coldfield transmitter. Am J Epidemiol, 1997. 145(1): p. 1-9 (S) (C).
231. Dolk, H., et al., Cancer incidence near radio and television transmitters in Great Britain. 2: All high
power transmitters. Am J Epidemiol, 1997. 145(1): p. 10-17 (S) (C).
232. Donnellan, M., D.R. McKenzie, and P.W. French, Effects of exposure to electromagnetic radiation at
835 MHz on growth, morphology and secretory characteristics of a mast cell analogue, RBL-2H3. Cell
Biol Int, 1997. 21(7): p. 427-439 (E) (C).
233. Dreyer, N.A., J.E. Loughlin, and K.J. Rothman, Cause-specific mortality in cellular telephone users. J
Am Med Assoc, 1999. 282(19): p. 1814-1816 (S).
234. Drogichina, E.A., et al., On the problems of autonomic and cardiovascular disturbances under the
effects of ultrahigh frequency electromagnetic fields. ATD Report 66-124, Library of Congress,
Aerospace Technology Division. Gig Tr Prof Zabol, 1966. 10: p. 13 (S).
235. DTI, Electromagnetic compatibility aspects of radio-based mobile telecommunications. LINK
Research Communications Programme. London, Department of Trade and Industry, 1999: p. (S).
236. Durney, C.H., H. Massoudi, and M.F. Iskander, Radiofrequency Radiation Dosimetry Handbook.
USAFSAM-TR-85-73, Fourth Edition. Brooks Air Force Base, Texas: USAF School of Aerospace
Medicine, 1986: p. (C).

237. Dutta, S.K. and W.H. Nelson, Lack of genetic effects of short duration low power density 2450 MHz
CW and 8.5-9.6 GHz pulsed electromagnetic radiation on selected microbial tester strains. Environ.
Mutagen. Soc - Ann. Meet. Program Abstr. 9 th, 1978: p. 69-70 (C).
238. Dutta, S.K., et al., Effects of 8.6-GHz pulsed electromagnetic radiation on an Escherichia coli repair-
deficient mutant. Symposium on Electromagnetic Fields in Biological Systems, 27-30 June, 1978,
Ottawa, Canada. IMPI and IEEE-Microwave Theory and Techniques Society. (Edmonton, Alberta,
Canada), 1978: p. 76-95 (C).

239. Dutta, S.K., et al., Effects of chronic non-thermal exposures of pulsed microwaves on a repair-
deficient mutant of Escherichia coli(Meeting abstract). Mutat Res, 1978. 53(1): p. 91-92 (C).
240. Dutta, S.K., et al., Lack of microbial genetic response to 2.45-GHz CW and 8.5- to 9.6-GHz pulsed
microwaves. J Microw Power, 1979. 14(3): p. 275-280 (S) (C).
241. Dutta, S.K., et al., Microwave radiation induced calcium ion efflux from human neuroblastoma cells in
culture. Bioelectromagnetics, 1984. 5(1): p. 71-78 (S) (C).
242. Dutta, S.K., B. Gosh, and C.F. Blackman, Radiofrequency radiation-induced calcium ion efflux
enhancement from human and other neuroblastoma cells in culture. Bioelectromagnetics, 1989.
10(2): p. 197-202 (S) (C).
243. Dutta, S.K., et al., Dose dependence of acetylcholinesterase activity in neuroblastoma cells expopsed
to modulated radio-frequency electromagnetic radiation. Bioelectromagnetics, 1992. 13(4): p. 317-322
(S) (C).

244. Eberle, P., et al., Zellproliferation, Schwesterchromatidaustausche, Chromosomenaberrationen,

Mikrokerne und Mutationsrate. .FGF, Newsletter Edition Wissenschaft, 1996. 4: p. 5-15 (K).
245. EC and European Commission, Possible Health Effects Related to the Use of Radiotelephones.
Proposals for a Research Programme by a European Commission Expert Group. Brussels, EC.,
1996: p. (S).
246. EC and European Commission, Council Recommendation of 12 July 1999 on the Limitation of
Exposure of the General Public to Electromagnetic Fields (0 Hz to 300 GHz. Off J Eur Commun,
1999. L199, 59 (1999/519/EC).: p. (S).
247. EC and European Commission, Communication from the Commission on the Precautionary Principle.
Brussels, EC, COM(2000)1, February 2000., 2000: p. (S).
248. Edwards, G.S., et al., Resonant microwave absorption of selected DNA molecules. Phys Rev Lett,
1984. 53(13): p. 1284-1287 (S).
249. Eicher, B., et al., Effects of TDMA modulated hand-held telephones on pacemakers (Meeting
abstract). Bioelectromagnetics Society, 16th Annual Meeting, 12-17 June, Copenhagen, Denmark,
1994: p. 67 (K).
250. El Kanza, K., M. Lalancette, and R. Mandeville, Ornithine decarboxylase activity in animals exposed
to different doses of ENU and EMF (Meeting abstract). Annual Review of Research on Biological
Effects of Electric and Magnetic Fields from the Generation, Delivery and Use of Electricity, 17-21
November, San Antonio, TX, 1996: p. 72 (C).
251. Elder, J., D. Cahill, and eds, Biological effects of radiofrequency radiation. US Environmental
Protection Agency, 1984. EPA-600/8-83-026: p. (K).
252. Elder, J.A., Radiofrequency radiation activities and issues: a 1986 perspective. Health Phys, 1987.
53(6): p. 607-611 (C).
253. Elekes, E., G. Thuroczy, and L.D. Szabo, Effect on the immune system of mice exposed chronically to
50 Hz amplitude modulated 2.45 GHz microwaves. Bioelectromagnetics, 1996. 17(3): p. 246-248 (K).
254. Elwood, J.M., A critical review of epidemiologic studies of radiofrequency exposure and human
cancers. Environ Health Perspect, 1999. 107(Suppl 1): p. 155-168 (S) (C).
255. Engstrom, S., Dynamic properties of Lednev's parametric resonance mechanism.
Bioelectromagnetics, 1996. 17(1): p. 58-70 (C).
256. EPA, Biological Effects of Radiofrequency Radiation. EPA-600/8-83-026F. U.S. Environmental
Protection Agency,Research Triangle Park, North Carolina, 1984: p. (C).
257. EPA, Summary and Results of the April 26-27, 1993, Radiofrequency Radiation Conference. 1.
Analysis of Panel Discussions. EPA 402R-95- 009. Research Triangle Park, NO U.S. Environmental
Protection Agency, March 1995, 1995: p. 40 (C).
258. Eriksson, N., The psychosocial work environment and illness among office workers. Sweden; UmeD
University, Department of Sociology: Doctoral thesis, 1996. No 2: p. (K).
259. Erwin, D.N. and W. Hurt, Biological effects of ultrawideband emission. EMP Human Health Effects
Science Review Panel Proceedings, 16-18 March, Theater Nuclear Warfare Program, Crystal Plaza
5, Rm. 866, Washington, DC 20362-5101, 1993: p. 91-95 (S).

260. ETSI, Radio transmission and reception. European Telecommunications Standards Institute; ETSI-
GSM technical specifications, 1992. PHASE 1, Documents 0.5.05: p. (K).
261. ETSI, Radio transmission and reception. European Telecommunications Standards Institut; ETSI-
GSM technical specifications, 1993. PHASE 1, Documents 0.5.05-DCS: p. (K).
262. ETSI, European digital cellular telecommunications system (Phase 2); GSM electromagnetic
compatibility (EMC) considerations. European Telecommunications Standards Institutes, TC-SMG,
1994. Technical Report 108: p. (K).
263. Eulitz, C., et al., Mobile phones modulate response patterns of human brain activity. NeuroReport,
1998. 9(14): p. 3229-3232 (E) (S).

264. Farrell, J.M., et al., Effects of low frequency electromagnetic fields on the activyt of ornithine
decarboxylase in developing chicken embryos. Bioelectrochem Bioenerg, 1997. 43(1): p. 91-96 (C).
265. Fesenko, E.E., et al., Microwaves and cellular immunity. I. Effect of whole body microwave irradiation
on tumour necrosis factor production in mouse cells. Bioelectrochem Bioenerg, 1999. 49(1): p. 29 (S).
266. Feychting, M., et al., Dementia and occupational exposure to magnetic fields. Scand J Work Environ
Health, 1998. 24(1): p. 46-53 (C).
267. Finkelstein, M., Cancer incidence among Ontario police officers. Am J Ind Med, 1998. 34(2): p. 157-
162 (C).
268. Fitzner, R., et al., Growth behaviour of HL-60 cells under influence of high-frequency electromagnetic
fields: Investigation of potential cancer-promoting effects. Electromagnetic Compatibility of Biological
Systems. Biological effects of High-Frequency Electromagnetic Fields of Mobile Radiotelephone
Systems and Police Radio. K. Brinkmann, G. Friedrich, eds. Berlin: VDE-VERLAG GMBH, 1997. 5:
p. 178-203 (E).

269. Forum, U., UMTS/IMT-2000 Spectrum. 1998(December).

270. Foster, K.R., B.R. Epstein, and M.A. Gealt, "Resonances" in the dielectric absorption of DNA?
Biophys J, 1987. 52(3): p. 421-425 (S).
271. Frei, M.R., et al., Chronic exposure of cancer-prone mice to low-level 2450 MHz radiofrequency
radiation. Bioelectromagnetics, 1998. 19(1): p. 20-31 (A) (E) (S) (C).
272. Frei, M.R., et al., Chronic, low-level (1.0 W/kg) exposure of mice prone to mammary cancer to 2450
MHz microwaves. Radiat Res, 1998. 150(5): p. 568-576 (A) (S).
273. French, P.W., M. Donnellan, and D.R. McKenzie, Electromagnetic radiation at 835 MHz changes the
morphology and inhibits proliferation of a human astrocytoma cell line. Bioelectrochem Bioenerg,
1997. 43(1): p. 13-18 (E) (C).
274. Freude, G., et al., Effects of microwaves emitted by cellular phones on human slow brain potentials.
Bioelectromagnetics, 1998. 19(6): p. 384-387 (E) (S).
275. Frey, A., Auditory system responses to radiofrequency energy. Aerospace Med, 1961. 32: p. 1140-
1142 (K).
276. Frey, A.H., S.R. Feld, and B. Frey, Neural function and behaviour: defining the relationship. Ann N Y
Acad Sci, 1975. 247: p. 433-439 (K) (S) (C).
277. Frey, A.H., Headaches from cellular phones: are they real and what are their implications?
(Commentary). Environ Health Perspect, 1998. 106(3): p. 101-103 (C).
278. Fritze, K., et al., Effect of global system for mobile communication (GSM) microwave exposure in
Blood Brain Barrier permeability in rat. Acta Neuropathol (Berl), 1997. 94(5): p. 465-470 (A) (E) (S)
279. Fritze, K., et al., Effect of global system for mobile communication microwave exposure on the
genomic response of the rat brain. Neuroscience, 1997. 81(3): p. 627-639 (A) (E) (S).
280. Frohlich, H., The biological effects of microwaves and related questions. Adv Electron Electron Phys,
1980. 53: p. 85-152 (S).
281. Frohlich, H., Coherent excitation in active biological systems. Modern Bioelectrochemistry. F.
Gutmann, H. Keyzer, eds., New York: Plenum Press, 1986: p. 241-261 (S).
282. Fröhlich, H., Long-range coherence and energy storage in biological systems. Int J Quantum Chem,
1968. II: p. 641 (S).
283. Fucic, A., et al., X-rays, microwaves and vinyl chloride monomer: their clastogenic and aneugenic
activity, using the micronucleus assay on human lymphocytes. Mutat Res, 1992. 282(4): p. 265-271
(K) (C).
284. Fukumoto, G. and C.V. Byus, A kinetic characterization of putrescine and spermidine uptake and
export in human erythrocytes. Biochim. Biophys Acta, 1996. 1282: p. 48-56 (C).
285. Funch, D.P., et al., Utility of telephone company records for epidemiologic studies of cellular
telephones. Epidemiology, 1996. 7(3): p. 299-302 (K).

286. Furia, L., D.W. Hill, and O.P. Gandhi, Effect of millimeter-wave irradiation on growth of
Saccharomyces cerevisiae. IEEE Trans Biomed Eng, 1986. 33(11): p. 993-999 (S).
287. Furstenberger, G., et al., Stimulatory role of transforming growth factors in multistage skin
carcinogenesis: possible explanation for the tumor-inducing effect of wounding in initiated NMRI
mouse skin. Int. J. Cancer, 1989. 43: p. 915-921 (C).
288. Gabriel, C., et al., Microwave absorption in aqueous solutions of DNA. Nature, 1987. 328(6126): p.
145-146 (S).
289. Gabriel, C., Compilation of the dielectric proprieties of body tisues at RF and microwaves frequency.
Brooks Airforce Technical report, 1996. AL/OE-TR-1996-0037: p. (A).
290. Gadja, G., A. Thansandote, and D. Lecuyer, Report on cellular tower surveys. (Ottawa: Health
Protection Branch of Health Canada). 1997: p. (C).

291. Gadja, G., D. Lecuyer, and D. Wilkinson, Report on an electromagnetic field survey at three
residences on Blessington Road, RR1, Corbyville, Ontario. (Ottawa: Health Protection Branch of
Health Canada). 1998: p. (C).
292. Gadja, G., A. Thansandote, and D. Lecuyer, Report on a survey of radiofrequency emissions at five
Vancouver area schools. (Ottawa: Health Protection Branch of Health Canada). 1998: p. (C).
293. Gallagher, R.P., et al., Brain cancer and exposure to electromagnetic fields. J Occup Med, 1991.
33(9): p. 944-945 (K).
294. Galvin, M.J., D.L. Parks, and D.I. McRee, Influence of 2.45 GHz microwave radiation on enzyme
activity. Radiat Environ Biophys, 1981. 19(2): p. 149-156 (K) (S).
295. Gandhi, C.R. and D.H. Ross, Microwave induced stimulation of 32P incorporation into
phosphoinositides of rat brain synaptosomes. Radiat Environ Biophys, 1989. 28(3): p. 223-234 (K) (S)

296. Gandhi, O., ANSI radiofrequency safety guide: its rationale, some problems, and suggested
improvements. Biological Effects and Medical Applications of Electromagnetic Energy. edited Gandhi
OP. Prentice-Hall, Englewood Cliffs, New Jersey, 1990: p. 28-47 (C).
297. Gandhi, O.P., et al., Specific absorption rates and induced current distributions in an anatomically
based human model for plane-wave exposure. Health Phys, 1992. 63(3): p. 281-290 (E).
298. Gandhi, O.P., G. Lazzi, and C.M. Furse, Electromagnetic absorption in the human head and neck for
mobile telephones at 835 and 1900 MHz. IEEE Trans Microw Theory Tech, 1996. 44(10 pt. 2): p.
1884-1897 (A).
299. Garaj-Vrhovac, V., D. Horvat, and Z. Koren, The effect of microwave radiation on the cell genome.
Mutat Res, 1990. 243(2): p. 87-93 (K) (S) (C).
300. Garaj-Vrhovac, V., A. Fucic, and D. Horvat, Comparison of chromosome aberration and micronucleus
induction in human lymphocytes after occupational exposure to vinyl chloride monomer and
microwave radiation. Period Biol, 1990. 92(4): p. 411-416 (S) (C).
301. Garaj-Vrhovac, V., D. Horvat, and Z. Koren, The relationship between colony-forming ability,
chromosome aberrations and incidence of micronuclei in V79 chinese hamster cells exposed to
microwave radiation. Mutat Res, 1991. 263(3): p. 143-149 (K) (S) (C).
302. Garaj-Vrhovac, V., A. Fucic, and D. Horvat, The correlation between the frequency of micronuclei and
specific chromosome aberrations in human lymphocytes exposed to microwave radiation in vitro.
Mutat Res, 1992. 281(3): p. 181-186 (K) (S) (C).
303. Garaj-Vrhovac, V., et al., Effects of 415 MHz frequency on human lymphocyte genome. Proceedings
IRPA9 Congress. Vienna, Austra, 1996. 3: p. 604-606 (C).
304. Garaj-Vrhovac, V., Micronucleus assay and lymphocyte mititic activity in risk assessment of
occupational exposure to microwave radiation. Chemosphere, 1999. 39(13): p. 2301-2312 (S).
305. Garland, F.C., et al., Non-Hodgkin's lymphomas in US Navy personnel. Arch Environ Health, 1988.
43(6): p. 425-429 (S).
306. Garland, F.C., et al., Incidence of leukemia in occupations with potential electromagnetic field
exposure in United States Navy personnel. Am J Epidemiol, 1990. 132(2): p. 293-303 (K) (S) (C).
307. Garson, O.M., et al., A chromosomal study of workers with long-term exposure to radio-frequency
radiation. Med J Aust, 1991. 155(5): p. 289-292 (K) (S) (C).
308. Ghoda, L., et al., Prevention of rapid intracellular degradation of ODC by a carboxyl trauncation.
Science, 1989. 243: p. 1493-1495 (C).
309. Gillois, M., C. Auge, and C. Chevalet, Effet des Ondes Electromagnetiques Non Ionisantes sur la
Viabilite et l'Heredite des Cellules de Mammiferes en Culture Etablie. Int Symp (URSI, CNFRS) on
Electromagn Waves & Biol (Ondes Electromagn & Biol ), Jouy en Josas, France, 1980: p. 9-15 (K).
310. Gilmour, S., et al., Constituitively elevated levels of ornithine and polyamines in mouse epidermal
papillomas. Carcinogenesis, 1991. 12: p. 1619-1625 (C).

311. Gilmour, S., et al., Induction of ornithine decarboxylase in specific subpopulations of murine epidermal
cells following multiple exposures to 12-O-tetradecanoylphorbol-13-acetate, merzerein and ethyl
phenylpropriolate. Carcinogenesis, 1992. 13: p. 51-56 (C).
312. Glick, A., et al., Loss of expression of TGF-a in skin and skin tumours is associated with
hyperproliferation and a high risk for malignant conversion. Proc. Natl. Acad. Sci. USA, 1993. 90: p.
6076-6080 (C).
313. Goblirsch, M. and E. Vogel, St`rbeeinflussungen von Herzschrittmachern durch
Diebstahlsicherungsanlagen. StrahlenschutzPraxis, 1996. 2(1): p. S46-50 (K).
314. Goldman, H., et al., Cerebrovascular permeability to Rb-86 in the rat after exposure to pulsed
microwaves. Bioelectromagnetics, 1984. 5(3): p. 323-330 (K) (C).
315. Goldsmith, J.R., Incorporation of epidemiological findings into radiation protection standards. Public
Health Rev, 1991. 19(1-4): p. 19-34 (C).

316. Goldsmith, J.R., Epidemiological evidence of radiofrequency radiation (microwave) effects on health
in military, broadcasting and occupational studies. Int J Occup Environ Health, 1995. 1(1): p. 47-57
317. Goli, V., et al., Transesophageal echocardiographic evaluation for mural thrombus following
radiofrequency catheter ablation of accessory pathways. Pacing and Clinical Electrophysics 14
(November, Part II), 1991: p. 1992-7 (C).
318. Gonzalez, G. and C. Byus, Effect of dietary arginine restriction upon ornithine and polyamine
metabolism during two-stage epidermal carcinogenesis in the mouse. Cancer Res, 1991. 51: p. 2932-
2939 (C).
319. Gordon, Z.V., Problems of industrial hygiene and the biological effects of electromagnetic super-high
frequency fields.(in Russian). Medicina, 1966: p. (S).

320. Gordon, Z.V., Occupational health aspects of radiofrequency electromagnetic radiation. Ergonomics
and physical environmental factors. Occupational Safety and Health Series No. 21, International
Labour Office, Geneva, 1970: p. 159-172 (C).
321. Gos, P., et al., Extremely high frequency electromagnetic fields at low power density do not affect the
division of exponential phase Saccharomyces cerevisiae cells. Bioelectromagnetics, 1997. 18(2): p.
142-155 (S).
322. Gos, P., et al., Mobile-phone type electromagnetic fields do not influence genetic stability in yeast.
Proceedings of the Second World Congress for Electricity and Magnetism in Biology and Medicine,
Bologna, Italy, June 1997, F. Bersani, Ed ; Klumer Academic Publishers, 1999: p. (R).
323. Goswami, P.C., et al., Proto-oncogene mRNA levels and activities of multiple transcription factors in
C3H 10T1/2 murine embryonic nfirbroblasts exposed to 835.62 and 847.74 cellular phone
communication frequency radiation. Radiat Res, 1999. 151(3): p. 300-309 (A) (E) (S).
324. Gottlob, L.R. and D.J. Madden, Age differences in the strategic allocation of visual attention. J
Gerontol B Psychol Sci Soc Sci, 1999. 54: p. 165 (S).
325. Goud, S.N., et al., Genetic effects of microwave radiation in mice. Mutat Res, 1982. 103(1): p. 39-42
(S) (C).
326. Grangeat, C., Radio frequency radiation from mobile phones. Alcatel Telecommunications Review,
4th Quarter 1998, 1998: p. 298-304 (E).
327. Grayson, J.K., Radiation exposure, socioeconomic status, and brain tumor risk in the US Air Force: a
nested case-control study. Am J Epidemiol, 1996. 143(5): p. 480-486 (K) (S) (C).
328. Grin, A.N., Effects of microwave on catecholamine metabolis in brain. US Joint Pub Research Device
Rep, 1974. JPRS 72066.: p. (S) (C).
329. Grin, A., Effect of microwave radiation on the metabolism of catecholamine in the brain. (Rus). Vrach
Delo, 1984. 10: p. 129-130 (K).
330. Gruenau, S., et al., Absence of microwave effect on blood-brain barrier permeability to 14C-sucrose in
the conscious rat. Exp Neurol, 1982. 75: p. 299-307 (K).
331. Grundler, W., et al., Mechanisms of electromagnetic interaction with cellular systems.
Naturwissenschaften, 1992. 79(12): p. 551-559 (S).
332. Guberan, E., et al., Gender ratio of offspring and exposure to shortwave radiation among female
physiotherapists. Scand J Work Environ Health, 1994. 20(5): p. 345-348 (C).
333. Gunn, S.A., T.C. Gould, and W.A.D. Anderson, The effect of microwave radiation on morphology and
function of rat testis. Lab. Invest, 1961. 10: p. 301-314 (C).
334. Guy, A.W., et al., Microwave Induced Acoustic Effects in Mammalian Auditory Systems and Physical
Materials. Ann NY Acad Sci, 1975. 247: p. 194-218 (K).
335. Guy, A.W., et al., Long-term 2450-MHz CW microwave irradiation of rabbits: methodology and
evaluation of ocular and physiological effects. J Microw Power, 1980. 15(1): p. 37-44 (K) (S).

336. Guy, A.W., et al., Effects of long-term low-level radiofrequency radiation exposure on rats. Vol 9:
Summary (Final rept. Jun 80-Jul 84). NTIS Document No. AD-A159 512/3/XAB, 24 pp. Report No.
SR-29; USAFSAM-TR-85-64. Contract No. F33615-80-C-0612; 7757; 01., 1985: p. (K) (S) (C).
337. Guy, A.W., Bioeffects of long-term exposures of animals. In : Radiofrequency radiation standards -
Biological effects, dosimetry, epidemiology and public health policy (NATO ASI Series, series A : Life
Sci, 1994. 274: p. 311-326 (K).
338. Haidler, T., et al., Clastogenic effects of radiofrequency radiations on chromosomes of Tradescantia.
Mutat Res, 1994. 324: p. 65-68 (K) (S) (C).
339. Hall, A.S., et al., Observation by MR imaging of in vivo temperature changes induced by
radiofrequency hyperthermia. J Comput Assis Tomogr, 1990. 3: p. 430-436 (C).
340. Hamnerius, Y., et al., A negative test for mutagenic action of microwave radiation in Drosophila
melanogaster. Mutat Res, 1979. 68(3): p. 217-223 (K) (C).

341. Hamnerius, Y., A. Rasmuson, and B. Rasmuson, Biological effects of high-frequency electromagnetic
fields on Salmonella typhimurium and Drosophila melanogaster. Bioelectromagnetics, 1985. 6(4): p.
405-414 (S) (C).
342. Hamrick, P.E., Thermal denaturation of DNA exposed to 2450 MHz CW microwave radiation. Radiat.
Res, 1973. 56: p. 400-404 (K) (C).
343. Hamrick, P.E. and S.S. Fox, Rat lymphocytes in cell culture exposed to 2450 MHz (CW) microwave
radiation. J Microw Power, 1977. 12(2): p. 125-132 (K).
344. Hansen, V.W., Guidelines for experiments to investigate the effects of RF electromag,netic fields on
biological systems. RF aspects. Edition Wissenschaft 11/E, FGF. Nov 1997, 1997: p. (E).
345. Hansen, V.W., et al., A radial waveguide exposure system for the uniform exposure of a large number
of small animals with a 383 MHz Tetra25 standard signal (Meeting abstract). Bioelectromagnetics
Society, 21st Annual Meeting, 20-24 June, Long Beach, CA, Abstract No. P-163, 1999: p. 170-171

346. Hardell, L., et al., Use of cellular telephones and the risk for brain tumours: A case-control study. Int J
Oncol, 1999. 15(1): p. 113-116 (A) (E) (S).
347. Hardell, L., Case control study on radiological work, medical X-ray investigations, and use of cellular
telephones as risk factors for brain tumors.
04.hard.html, 2000: p. (R).
348. Harland, J.D. and R.P. Liburdy, Environmental magnetic fields inhibit the antiproliferative action of
tamoxifen and melatonin in a human breast cancer cell line. Bioelectromagnetics, 1997. 18(8): p. 555-
562 (C).
349. Hawel, L.r., R. Tjandrawinata, and C. Byus, Selective putrescine export is regulated by insulin and
ornithine in Reuber H35 hepatoma cells. Biochim. Biophys. Acta, 1994. 1222: p. 15-26 (C).
350. Hawel, L.r., et al., Biosynthesis and selective export of 1,5-diaminopentane (cadaverine) in
mycoplasma-free cultured mammalian cells. J. Biol. Chem, 1994. 269: p. 7412-7418 (C).
351. Hayes, R.B., et al., Occupation and risk for testicular cancer: a case-control study. Int J Epidemiol,
1990. 19(4): p. 825-831 (S) (C).
352. Hayes, D., et al., Effect of digital cellular phones on permanent pacemakers. Pacing Clin
Electrophysiol, 1995. 18: p. 863 (K).
353. Heikkinen, P., et al., Effects of radiofrequency radiation on the development of cancer in mice
(Meeting abstract). Bioelectromagnetics Society, 20th Annual Meeting, 7-11 June, St. Pete Beach,
FL, Abstract No. P-58B, 1998: p. 139-140 (S) (C).
354. Hermann, D.M. and K.A. Hossmann, Neurologic effects of microwave exposure related to mobile
communication. J Neurol Sci, 1997. 152: p. 1-14 (E) (S) (C).
355. Hertz, L. and A. Schousbee, Ion and energy metabolism of the brain at the cellular level. Int. Rev.
Neurobiol, 1975. 18: p. 141 (C).
356. Hibshoosh, H., M. Johnson, and I.B. Weinstein, Effects of overexpression of ornithine decarboxylase
(ODC) on control of oncogene-induced cell transformation. Oncogene, 1991. 6: p. 739-743 (S) (C).
357. Hietanen, M., et al., Influence of RF Fields Emitted by Cellular Phones on the Human EEG. XXV th Gl
Ass URSI, 28 Aug - 5 Sep, Lille, France, 1996: p. 567K3-6 (K).
358. Hietanen, M. and A.-M. Hämäläinen, Exposure to cellular phones and EEG activity. EMC 98 Roma,
International Symposium on Electromagnetic Compatibility, 1998. 1: p. 181-183 (E).
359. Hietanen, M., T. Kovala, and A.-M. Hämäläinen, Human brain activity during exposure to
radiofrequency fields emitted by cellular phones. Scand. J. Work. Environ. Health, 1999. (Accepted):
p. (E).

360. Higashikubo, R., et al., Radiofrequency electromagnetic fields have no effect on the In vivo
proliferation of the 9L brain tumor. Radiat Res, 1999. 152(6): p. 665-671 (A) (S).
361. Hill, A.B., The environment and disease: association or causation? Proc. R. Soc. Med, 1965. 58: p.
295-300 (C).
362. Hill, D., Human studies. Biological Effects of Radiofrequency Radiation., 1984. U.S. EPA-600/8-83-
026F: p. 112-21 (C).
363. Hill, D., A Longitudinal Study of a Cohort with Past Exposure to Radar : the MIT Radiation Laboratory
Follow up Study (Dissertation). Univ Michigan Dissert Serv (Ann Arbor), 1988: p. (K).
364. Hocking, B., et al., Cancer incidence and mortality and proximity to TV towers. Med J Aust, 1996.
165(11-12): p. 601-605 (S) (C).
365. Hocking, B., Preliminary report: symptoms associated with mobile phone use. Occup Med (Oxf),
1998. 48(6): p. 357-360 (A) (E) (S).
366. Hocking, B., Symptoms associated with mobile phone use: addendum [letter]. Occup Med (Oxf),
1998. 48: p. 472 (E).

367. Holly, E.A., et al., Intraocular melanoma linked to occupations and chemical exposures.
Epidemiology, 1996. 7(1): p. 55-61 (S) (C).
368. Holm, I., et al., Feedback control of ornithine decarboxylase expression by polyamines. Analysis of
ornithine decarboxylase mRNA distribution in polysome profiles and of translation of this mRNA in
vitro. Biochem. J, 1989. 258: p. 343-350 (C).
369. Holt, J. and A. Nelson, Four years of microwaves in cancer therapy. J Belge Radiol, 1979. 62(5): p.
467-476 (K).
370. Hook, G.J., Evaluation of potential genotoxicity of RFR in vivo using the single cell gel electrophoresis
assay. University of Washington. 2nd WTR State of the Science Colloquium. Long Beach, USA.,
1999: p. (E).
371. Hook, G.J., et al., Genotoxicity of radio frequency fields generated by analog, TDMA, CDMA, and
PCS cellular technologies evaluated using the single cell gel electrophoresis (scge) and the
cytochalasin B micronucleus (CB-MN) assay. Proceedings 21th Annual Meeting of BEMS, Long
Beach, USA, 1999: p. (E).

372. Hook, G.J., et al., Investigation of DNA damage and micronuclei induction in cultured human blood
cells. Bioelectromagnetics, 2001: p. sous presse (R).
373. Hossmann, K.A., Health aspects of mobile communication (Meeting Abstract). Second World
Congress for Electricity and Magnetism in Biology and Medicine, 8-13 June, Bologna, Italy, 1997: p.
5-6 (A).
374. Hoth, M. and R. Penner, Depletion of intracellular calcium stores activates a calcium current in mast
cells. Nature, 1992. 355: p. 353-355 (K).
375. Huang, A.T., et al., The effect of microwave radiation (2450 MHz) on the morphology and
chromosomes of lymphocytes. Radio Sci, 1977. 12(6S): p. 173-177 (K) (S) (C).
376. Huang, A.T.-F. and N.G. Mold, Immunologic and hemopoietic alterations by 2450 MHz
electromagnetic radiation. Bioelectromagnetics, 1980. 1(1): p. 77-87 (K).
377. Huber, R., et al., Exposure to pulsed high-frequency electromagnetic field during waking affects
human sleep EEG. NeuroReport, 2000. 11: p. 3321-3325 (R).
378. Hunt, E.L., N.W. King, and R.D. Phillips, Behavioral effects of pulsed microwave radiation. Ann NY
Acad Sci, 1975. 247: p. 440 (S).
379. Hurta, R., A. Greenberg, and J. Wright , Transforming growth factor b1 selectively regulates ornithine
decarboxylase gene expression in malignang H-ras transformed fibrosarcoma cell lines. J. Cell
Physiol, 1993. 156: p. 272-279 (C).
380. HWC and (Health and Welfare Canada), Guidelines on exposure to electromagnetic fields from
magnetic resonance clinical systems. Safety Code 26, 1987: p. (C).
381. Hyland, G.J., Non-thermal bioeffects induced by low-intensity microwave irradiation of living systems.
Eng Sci Educ J, 1998. 7(6): p. 261-269 (S).
382. Hyland, G.J., Physics and biology of mobile telephony. The Lancet, 2000. 356: p. 1833-1836 (R).
383. IARC and International Agency for Research on Cancer, Cancer Incidence in Five Continents. IARC
Scientific Publications No. 143, Lyon
Parkin DM, Whelan SL, Ferlay J, Raymond L, Young J, eds, 1997. VII: p. (C).
384. Ichiba, T., et al., Functional regions of ornithine decarboxylase antizyme. Biochem. Biophys. Res.
Comm, 1994. 200: p. 1721-1727 (C).
385. ICNIRP, Health issues related to the use of hand-held radiotelephones and base transmitters. Health
Phys, 1996. 70: p. 587-593 (K) (E) (S).

386. ICNIRP, Guidelines for limiting exposure to time-varying electric, magnetic and electromagnetic fields
(up to 300 GHz). Health Phys, 1998. 74(4): p. 494-522 (A) (S).
387. ICNIRP, Response to questions and comments on ICNIRP guidelines. Health Phys, 1998. 75(4): p.
438 (S).
388. IEEE, IEEE standards for safety levels with respect to human exposure to radiofrequency
electromagnetic fields 3kHz-300GHz. C95.1-1991, 1991: p. (K) (A) (C).
389. IEEE-USA and E.P. Statement, Human exposure to RF emissions from cellular radio base station
antennas. IEEE United States Activities. (Washington, DC: COMAR)., 1992: p. (C).
390. IEEE/ANSI, Standard for safety levels with respect to human exposure to radiofrequency
electromagnetic fields, 3 kHz to 300 GHz. New York, IEEE, 1991. C95.1-1991: p. (S).
391. Imaida, K., et al., Lack of promoting effects of the electromagnetic near-field used for cellular phones
(929.2 MHz) on rat liver carcinogenesis in a medium-term liver bioassay. Carcinogenesis, 1998.
19(2): p. 311-314 (A) (E) (S) (C).
392. Imaida, K., et al., The 1.5 GHz electromagnetic near-field used for cellular phones does not promote
rat liver carcinogenesis in a medium-term liver bioassay. Jpn J Cancer Res, 1998. 89(10): p. 995-
1002 (A) (E) (S) (C).

393. Inaba, R., et al., Effects of whole body microwave exposure on the rat brain contents of biogenic
amines. Eur J Appl Physiol, 1992. 65(2): p. 124-128 (K) (S) (C).
394. INIRC, Guidelines on limits of exposure to radiofrequency electromagnetic fields in the frequency
range from 100 kHz to 300 GHz. Health Phys, 1988. 54(1): p. 115 (S).
395. INIRC, Interim guidelines on limits of exposure to 50/60 Hz electric and magnetic fields. Health Phys,
1990. 58(10): p. 113 (S).
396. Inskip, P.D., et al., Study design for a case-control investigation of cellular telephones and other risk
factors for brain tumours in adults. Radiat Prot Dosim, 1999. 86: p. 45 (S).
397. Inskip, P., et al., Cellular telephone use and brain tumors. New England J Med, 2001. 344: p. 79-86
398. Irnich, W., et al., Ein Beitrag zur Sicherheit von Implantation Schriftenreihe der Bundesanstalt fhr
Arbeitsschutz. Fb 569, Dortmund, 1989: p. (K).
399. IRPA-INIRC, 1988 Guidelines on limits of exposure to radiofrequency fields in the frequency range
from 100kHz to 300GHz. Health Phys, 1988. 54: p. 115-123 (K) (C).

400. Isa, A. and M. Noor, Nonionizing radiation exposure causing ill health and alopecia areata. Med J
Malaysia, 1991. 46(3): p. 235-238 (C).
401. ITU, Final Acts of the World Administrative Radio Conference (WARC-95). Geneva, 1995: p. (E).
402. Ivaschuk, O.I., et al., Exposure of nerve growtn factor-treated PC12 rat pheochromocytoma cells to a
modulated radiofrequency field at 836.55 MHz. Effects on c-jun and c-fos expression.
Bioelectromagnetics, 1997. 18(3): p. 223-229 (A) (E) (S) (C).
403. Janes, E.D., et al., Effects of 2450 MHz microwaves on protein synthesis and on chromosomes in
Chinese hamsters. Non-Ionizing Radiat, 1969. 1: p. 125-130 (C).
404. Janne, J., L. Alhonen, and P. Leinonen, Polyamines: from molecular biology to clinical applications.
Annals of Medicine, 1991. 23(3): p. 241-259 (C).
405. Jauchem, J.R. and M.R. Frei, Heart rate and blood pressure changes during radiofrequency
irradiation and environmental heating. Comp Biochem Physiol, 1992. 101A(1): p. 1-9 (S).
406. Jauchem, J.R. and M.R. Frei, High-peak-power microwave pulses: effects on heart rate and blood
pressure in unanaesthetised rats. Aviat Space Environ Med, 1995. 66(10): p. 992-997 (S).
407. Jauchem, J.R., Exposure to extremely-low-frequency electromagnetic fields and radiofrequency
radiation: cardiovascular effects on humans. Int Arch Occup Environ Health, 1997. 70(1): p. 9-21 (S).
408. Jauchem, J.R., Health effects of microwave exposures: a review of the recent (1995-1998) literature.
J Microw Power Electromagn Energy, 1998. 33(4): p. 263-274 (S).
409. Jensh, R.P., W.H. Vogel, and R.L. Brent, Postnatal functional analysis of prenatal exposure of rats to
915 MHz microwave radiation. J Am Coll Toxicol, 1982. 1(3): p. 73-90 (C).
410. Jensh, R.P., I. Weinberg, and R.L. Brent, Teratologic studies of prenatal exposure of rats to 915-MHz
microwave radiation. Radiat Res, 1982. 92: p. 160-171 (C).
411. Jensh, R.P., I. Weinberg, and R.L. Brent, An evaluation of the teratogenic potential of protracted
exposure of pregnant rats to 2450-MHz microwave radiation. I. Morphologic analysis at term. J
Toxicol Environ Health, 1983. 11(1): p. 23-35 (S) (C).
412. Jensh, R.P., W.H. Vogel, and R.L. Brent, An evaluation of the teratogenic potential of protracted
exposure of pregnant rats to 2450-MHz microwave radiation. II. Postnatal physiologic analysis. J
Toxicol Environ Health, 1983. 11(1): p. 37-59 (S) (C).

413. Jensh, R.P., Studies of the teratogenic potential of exposure of rats to 6000-MHz microwave
radiation. I. Morphologic analysis at term. Radiat Res, 1984. 97(2): p. 272-281 (S) (C).
414. Jensh, R.P., Studies of the teratogenic potential of exposure of rats to 6000-MHz microwave
radiation. II. Postnatal psychophysiologic evaluations. Radiat Res, 1984. 97(2): p. 282-301 (S) (C).
415. Jensh, R.P., Behavioral teratologic studies using microwave radiation: is there an increased risk from
exposure to cellular phones and microwave ovens? Reprod Toxicol, 1997. 11(4): p. 601-611 (A) (E)
416. Johansen, C. and J.H. Olsen, Cellular telephones, magnetic field exposure, risk of brain tumours and
cancer at other sites: a cohort study. Radiat Prot Dosim; Proceedings of an International Workshop,
Exposure Metrics and Dosimetry for EMF Epidemiology, National Radiation Protection Board, Chilton,
UK, 7-9 September, 1998, 1999. 83(1-2): p. 155-157 (S).
417. Johansson, O., Electrical hypersensitivity and hypersensitivity to mobile phones. Results from a
double blind provokation study of the character of a method study. (Swed). Stockholm; Karolinska
Institute, Unit for Experimental Dermatology, 1995. Report no 2: p. (K).
418. Johnson, R.B., et al., Open field behavior and corticosterone. Effects of Long-term Low-level
Radiofrequency Radiation Exposure on Rats. Brooks Air Force Base, Texas, USAF School of
Aerospace Medicine, 1983. Volume 4: p. USAFSAM TR 83-42 (S).

419. Jokela, K., et al., Radiation safety of handheld mobile phones and base stations. Stockholm,, 1999.
STUK - A161.: p. (E) (S).
420. Jorritsma, J.B. and A.W. Konings, The occurrence of DNA strand breaks after hyperthermic
treatments of mammalian cells with and without radiation. Radiat Res, 1984. 98: p. 198 (S).
421. Joyner, K.H., V. Anderson, and M.P. Wood, Interference and energy deposition rates from digital
mobile phones (Meeting abstract). Bioelectromagnetics Society, 16th Annual Meeting, 12-17 June,
Copenhagen, Denmark, 1994: p. 67-68 (K).
422. Justesen, D., et al., Research on health effects of nonionizing radiation. United States House of
Representatives: Hearing Committee on Science and Technology, July 12, 1978, 1979. No. 52- 3620:
p. 356-366 (C).
423. Juutilainen, J., et al., Do pulse-modulated or continuous 900 MHz RF fields enhance the carcinogenic
effect of ionising radiation in mice? Proceedings 1 st International Medical Scientific Congress "Non-
Ionizing High-frequency EM Radiations: Researching the Epidemiology and Clinical Evidence", Rome,
November 1999, (in press).: p. (S).

424. Juutilainen, J., et al., Epidermal ODC and polyamine levels after UV and magnetic field exposures.
Proceedings of the 17th Annual Meeting of the Bioelectromagnetics Society, 1995: p. Abstract A-7-1
425. Juutilainen, J., et al., A study on the effects of pulsed or continuous 900 MHz radiation on the
development of cancer in mice (Meeting abstract). European Bioelectromagnetics Association
(EBEA), 3rd International Congress, 29 February-3 March, Nancy, France, 1996: p. (C).
426. Juutilainen, J. and S. Lang, Genotoxic, Carcinogenic and Teratogenic effects of electromagnetic fields
: Introduction and overview. Mutat Res, 1997. 387(3): p. 165-171 (A).
427. Juutilainen, J. and R. de Seze, Biological effects of amplitude-modulated radiofrequency radiation.
Scand J Work Environ Health, 1998. 24(4): p. 245-254 (E).
428. Kaiser, F., Theory of resonant effect of RFand MW energy. Biological Effects and Dosimetry of
Nonionizing Radiation. Radiofrequency and Microwave Energies. M. Grandolfo, S. M. Michaelson,
A. Rindi, eds., New York: Plenum Press, 1983: p. 251-282 (S).
429. Kallen, B., G. Malmquist, and U. Moritz, Delivery outcome among physiotherapists in Sweden: is non-
ionizing radiation a fetal hazard? Arch Environ Health, 1982. 37(2): p. 81-85 (S) (C).
430. Kamimura, Y., et al., Effect of 2.45 GHz microwave irradiation on monkey eyes (Letter). IEICE Trans
Commun, 1994. 77B(6): p. 762-765 (S) (C).
431. Kandel, E.R., J.H. Schwartz, and T.M. Jessel, Principles of Neural Science. New York, McGraw-Hill.
Professional Publishing, 2000. (4 th edition): p. (S).
432. Kaplan, I.T., et al., Absence of heart rate effects in rabbits during low-level microwave irradiation.
IEEE Trans Microwave Theory Tech, 1971. 19: p. 168 (S).
433. Kavaliers, M., et al., Opioid systems and the biological effects of magnetic fields. On the Nature of
Electromagnetic Field Interactions with Biological Systems. A. H. Frey, ed., Austin, TX: R. G. Landes
Co., 1994: p. 181-194 (C).
434. Kerbacher, J.J., M.L. Meltz, and D.N. Erwin, Influence of radiofrequency radiation on chromosome
aberrations in CHO cells and its interaction with DNA damaging agents. Radiat Res, 1990. 123(3): p.
311-319 (K) (S) (C).

435. Khalil, A.M., W.F. Qassem, and M.M. Suleiman, A preliminary study on the radiofrequency field-
induced cytogenetic effects in cultured human lymphocytes. Dirasat, 1993. 20: p. 121-130 (S) (C).
436. Khillare, B. and J. Behari, Effect of amplitude-modulated radiofrequency radiation on reproduction
pattern in rats. Electro Magnetobiol, 1998. 17(1): p. 43-55 (S).
437. Kim, Y., H. Pan, and V. AK., Non-AP-1 tumour promoter 12-O-tetradecanoylphorbol-13-acetate-
responsive sequences in the human ornithine decarboxylase gene. Mol. Carcinogen, 1994. 19: p.
169-179 (C).
438. Kim, Y.S., et al., Characteristics of EEG and AEP in human volunteers exposed to RF (Meeting
abstract). Bioelectromagnetics Society, 20th Annual Meeting, 7-11 June, St. Pete Beach, FL, Abstract
No. P-19A, 1998: p. 105-106 (E).
439. Kiselev, R.I. and N.P. Zalyubovskaya, Study of inhibiting effect of superhigh frequency millimetre
wave on adenoviruses. US Joint Pub Research Service Rep, 1976. JPRS L/5615: p. 71 (S).
440. Kitchin, K., J. Brown, and A. Kulkarni, Complementarity of genotoxic and nongenotoxic predictors of
rodent carcinogenicity. Teratogenesis Carcinog. Mutagen, 1994. 13: p. 83-100 (C).
441. Kittel, A., et al., Qualitative enzyme historchemistry and microanalysis revela changes in ultra-
structural distribution of calcium and calcium activated ATPases after microwave irradiation of the
medial habenula. Acta Neuropathol (Berl), 1996. 92(4): p. 362-368 (S) (C).

442. Klauenberg, B.J., et al., Radiofrequency radiation standards. Biological effects, dosimetry,
epidemiology, and public health policy. Radiofrequency Radiation Standards. Biological Effects,
Dosimetry, Epidemiology, and Public Health Policy. Nato ASI Series A: vol. 274, 1995: p. (K).
443. Koenig, H., et al., Polyamines and Ca 2+ meidate hyperosmolal opening of the blood-brain barrier: In
vitro studies in ioslated rat cerebral capillaries. J Neurochem, 1989. 13: p. 1135-1142 (C).
444. Kohli, M., et al., Calculated microwave absorption of double-helical B-conformation poly(dG).poly(dC).
Biopolymers, 1981. 20(4): p. 853-864 (S).
445. Koivisto, M., et al., The effects of electromagnetic field emitted by GSM phones on working memory.
NeuroReport, (in press).: p. (A) (S).
446. Koivisto, M., et al., Effects of 902 MHz electromagnetic field emitted by cellular phones on response
times in humans. NeuroReport,, 2000. 11: p. 413-415 (A) (S).
447. Kolodynski, A.A. and V.V. Kolodynska, Motor and psychological functions af school children living in
the area of the Skrunda Radiation Location Station in Latvia. Sci Total Environ, 1996. 180(1): p. 87-93
448. Kowalczuk, C.I., R.D. Saunders, and H.R. Stapleton, Sperm count and sperm abnormality in mice
after exposure to 2450 MHz microwave radiation. Mutation Res, 1983. 122: p. 155-161 (C).

449. Koza, R., et al., Constituitively elevated levels of ornithine and polyamines in mouse epidermal
papillomas. Carcinogenesis, 1991. 12: p. 1619-1625 (C).
450. Krause, D., et al., Enhancement of ornithine decarboxylase activity in L929 cells by amplitude
modulated microwaves (Meeting abstract). Bioelectromagnetics Society, 12th Annual Meeting, 10-14
June, San Antonio, TX, 1990: p. 94 (K) (C).
451. Krause, D., et al., Role of modulation on the effect of microwaves on ornithine decarboxylase activity
in L929 cells. Bioelectromagnetics, 1997. 18: p. 132-141 (K) (E).
452. Krause, C.M., et al., Effects of electromagnetic field emitted by cellular phones on the EEG during a
memory task. NeuroReport, 2000. 11: p. 761-764 (A) (S).
453. Krewski, D., et al., A review of the potential health risks of radiofrequency fields from wireless
telecommunications devices. The Royal society of Canada. Ottawa, Canada., 1999. RSC.RPR 99-1:
p. (E).
454. Kristensen, T.S., Cardiovascular disease and the work environment. A critical review of the
epidemiologic literature on nonchemical factors. Scand J Work Environ Health, 1989. 15: p. 165-179
455. Kubota, S., et al., Ornithine decarboxylase overexpression in mouse 10T1/2 fibroblasts. Cellular
transformation and invasion. J Natl Cancer Inst, 1997. 89: p. 567 (S) (C).
456. Kues, H.A., et al., Effects of 2.45-GHz microwaves on primate corneal endothelium.
Bioelectromagnetics, 1985. 6(2): p. 177-188 (S) (C).
457. Kues, H.A. and S.A. D'Anna, Changes in the monkey eye following pulsed 2.45 GHz microwave
exposure. IEEE Engineering in Medicine and Biology Society, Ninth Annual Conference, 13-16
November, 1987, Boston, MA,, 1987. Vol. 2: p. 698-700 (S) (C).
458. Kues, H.A., et al., Histological evaluation of microwave-induced vascular leakage in the iris (Meeting
abstract). Bioelectromagnetics Society, 10th Annual Meeting, 19-24 June, Stamford, CT, 1988: p. 49

459. Kues, H.A., et al., Microwave-induced electroretinographic changes in the primate (Meeting abstract).
Bioelectromagnetics Society, 13th Annual Meeting, 23-27 June, Salt Lake City, UT, 1991: p. 53 (S).
460. Kues, H.A. and J.C. Monahan, Microwave-induced changes in the primate eye. Johns Hopkins APL
Tech Dig, 1992. I3: p. 244 (S).
461. Kues, H.A. and J.C. Monahan, Pulsed microwave-induced ocular changes in the restrained non-
human primate (Meeting abstract ). First World Congress for Electricity and Magnetism in Biology and
Medicine, 14-19 June, Lake Buena Vista, FL, 1992: p. 62 (S).
462. Kues, H.A., et al., Increased sensitivity of the non-human primate eye to microwave radiation
following opthalmic drug pretreatment. Bioelectromagnetics, 1992. 13(5): p. 379-393 (K) (S) (C).
463. Kues, H.A., et al., Absence of ocular effects after either single or repeated exposure to 10 mW/cm 2
from a 60 GHz CW source. Bioelectromagnetics, 1999. 20(8): p. 463-473 (S).
464. Kumar, A. and A. Butler, Transcription factor Sp3 antagonizes activation of the ornithine
decarboxylase promoter by Sp1. Nucleic Acids Res, 1997. 25: p. 2012-2019 (C).
465. Kumlin, T., et al., Nocturnal Secretion of 6-Hydroxy-Melatonin Sulphate in Mice Exposed to 900 MHz
Radiofrequency Radiation or 50 Hz Magnetic Fields. Proceedings of the Second World Congress for
Electricity and Magnetism in Biology and Medicine, Bologna, Italy, June 1997, F. Bersani, Ed ; Klumer
Academic Publishers, 1999: p. (E).
466. Kunz, L.L., et al., Effects of long-term low-level radiofrequency radiation exposure on rats. Vol 8 -
Evaluation of longevity, cause of death, and histopathological findings. NTIS Document No. AD-A154

283/6/XAB, 69 pp. (Final rept. 1 Jun 80-31 Jul 84). Report No. SR-28; USAFSAM-TR-85-11. Contract
No. 15-80-C-0612; 7757; 01. F336., 1985. Vol 8: p. (K) (C).
467. Kuster, N. and Q. Balzano, Energy absorption mechanism by biological bodies in the near field of
dipole antennas above 300 MHz. IEEE Trans.Veh.Technol, 1992. 41: p. 17-23 (C).
468. Kuster, N., R. Kastle, and T. Schmid, Dosimetric evaluation of handheld mobile communications
equipment with known precision. IEICE Trans, 1997. E80-A(5): p. 1-7 (C).
469. Kuster, N., Dosimetric needs for laboratory and epidemiological research. Non-Thermal Effects of RF
Electromagnetic Fields. Proceedings: International Seminar on Biological Effects of Non-Thermal
Pulsed and Amplitude Modulated RF Electromagnetic Fields and Related Health Risks, Munich,
Germany, 20-21 November 1996, J.H. Bernhardt, R. Matthes, M.H. Repacholi, eds., 1997. ICNIRP
3/97: p. 27-34 (E).
470. Kwee, S. and P. Raskmark, Changes in cell proliferation due to environmental non-ionizing radiation
2. Microwave radiation. 2. Microwave Radiation. Bioelectrochem Bioenerg, 1998. 44(2): p. 251-255
(A) (E) (S) (C).
471. Lagorio, S., et al., Mortality of plastic-ware workers exposed to radiofrequencies. Bioelectromagnetics,
1997. 18(6): p. 418-421 (S) (C).

472. Lagroye, I., et al., Measurement of DNA damage after acute exposure to 2450 MHz microwaves in rat
brain cells by two alkaline comet assay methods (Meeting abstract ). Bioelectromagnetics Society,
21st Annual Meeting, 20-24 June, Long Beach, CA, Abstract No. 1-1, 1999: p. 9-10 (E).
473. Lai, H., et al., Low-level microwave irradiation affects central cholinergic activity in the rat. J
Neurochem, 1987. 48: p. 40-45 (K) (A) (S) (C).
474. Lai, H., et al., Effects of low-level microwave irradiation on hippocampal and frontal cortical choline
uptake are classically conditionable. Pharmacol Biochem Behav, 1987. 27(4): p. 635-639 (K) (C).
475. Lai, H., et al., A review of microwave irradiation and actions of psychoactive drugs. IEEE Eng Med
Biol, 1987. 6(1): p. 31-36 (C).
476. Lai, H., A. Horita, and A.W. Guy, Acute low-level microwave exposure and central cholinergic activity:
studies on irradiation parameters. Bioelectromagnetics, 1988. 9(4): p. 355-362 (A) (C).
477. Lai, H., et al., Low-level microwave irradiation and central nervous cholinergic systems. Pharmacol
Biochem Behav, 1989. 33: p. 131-138 (K) (A) (S) (C).
478. Lai, H., et al., Low-level microwave irradiation and central cholinergic activity: a dose response study.
Bioelectromagnetics, 1989. 10: p. 203-208 (S) (C).
479. Lai, H., et al., Corticotropin-releasing factor antagonist blocks microwave induced decreases in high-
affinity choline uptake in the rat brain. Brain Res Bull, 1990. 25: p. 609-612 (K) (A) (S) (C).
480. Lai, H., et al., Naltrexonew pretreatment blocks microwave-induced changes in central cholinergic
receptors. Bioelectromagnetics, 1991. 12: p. 27-33 (K) (A) (S) (C).
481. Lai, H., Research on the neurological effects of non-ionizing radiation at the University of Washington.
Bioelectromagnetics, 1992. 13(6): p. 513-526 (S) (C).
482. Lai, H., et al., Opioid receptor subtypes that mediate a microwave-induced decrease in central
cholinergic activity in the rat. Bioelectromagnetics, 1992. 13(3): p. 237-246 (A) (C).

483. Lai, H., et al., Single vs. repeated microwave exposure: effects on benzodiazepine receptors in the
brain of a rat. Bioelectromagnetics, 1992. 13(1): p. 57-66 (C).
484. Lai, H., et al., Effects of a 60 Hz magnetic field on central cholinergic systems of the rat.
Bioelectromagnetics, 1993. 14(1): p. 5-15 (K) (C).
485. Lai, H., A. Horita, and A.W. Guy, Microwave irradiation affects radial-arm maze performance in the
rat. Bioelectromagnetics, 1994. 15: p. 95-104 (A) (E) (S) (C).
486. Lai, H., Neurological effects of microwave radiation. Advances in electromagnetic fields in living
systems. New York: Plenum Press, Lin JC, ed, 1994. 1: p. 27-80 (K).
487. Lai, H. and N.P. Singh, Acute low-intensity microwave exposure increases DNA single-strand breaks
in rat brain cells. Bioelectromagnetics, 1995. 16(3): p. 207-210 (K) (A) (E) (S) (C).
488. Lai, H. and N.P. Singh, Single and double-strand DNA breaks in rat brain cells after acute exposure to
radiofrequency electromagnetic radiation. Int J Radiat Biol, 1996. 69(4): p. 513-521 (K) (A) (E) (S) (C).
489. Lai, H., Spatial learning deficit in the rat after exposure to a 60Hz magnetic field. Bioelectromagnetics,
1996. 17(6): p. 494-496 (C).
490. Lai, H., et al., Intraseptal microinjection of -funaltrexamine blocked a microwave-induced decrease of
hippocampal cholinergic activity in the rat. Pharmacol. Biochem. Behav, 1996. 53: p. 613-616 (A).
491. Lai, H. and N.P. Singh, Melatonin and a spin-trap compound block radiofrequency electromagnetic
radiation-induced DNA strand breaks in rat brain cells. Bioelectromagnetics, 1997. 18: p. 446-454 (E)
492. Lai, H., M. Carino, and N. Singh, Naltrexone blocks RFR-induced DNA double strand breaks in rat
brain cells. in press, 1997: p. (C).

493. Lai, H. and M. Carino, Intracerebroventricular injection of mu- and delta-opioid receptor antagnoists
block 60Hz magnetic field-induced decreases in cholinergic activity in the frontal cortex and
hippocampus of the rat. Bioelectromagnetics, 1998. 19(7): p. 432-437 (C).
494. Lamble, D., et al., Cognitive load and detection in thresholds in car following situations: safety
implications for using mobile (cellular) telephones while driving. Accid Anal Prev, 1999. 31: p. 617 (S).
495. Lancranjan, I., et al., Gonadic function in workmen with long-term exposure to microwaves. Health
Phys, 1975. 29: p. 381-383 (C).
496. Larsen, A.I., J. Olsen, and O. Svane, Gender-specific reproductive outcome and exposure to high-
frequency electromagnetic radiation among physiotherapists. Scand J Work Environ Health, 1991.
17(5): p. 324-329 (C).
497. Lary, J.M., et al., Teratogenic effects of 27.12 MHz radiofrequency radiation in rats. Teratology, 1982.
26: p. 299-309 (S) (C).
498. Lary, J.M., et al., Teratogenicity of 27.12 MHz radiation in rats is related to duration of hyperthermia
exposure. Bioelectromagnetics, 1983. 4: p. 249-253 (S) (C).
499. Lary, J.M., D.L. Conover, and P.H. Johnson, Absence of embryotoxic effects from low-level (non-
thermal) exposure of rats to 100 MHz radiofrequency radiation. Scand J Work Environ Health, 1983.
9(2): p. 120-127 (C).

500. Lary, J.M., et al., Dose-response relationship between body temperature and birth defects in
radiofrequency-irradiated rats. Bioelectromagnetics, 1986. 7(2): p. 141-149 (C).
501. Lary, J.M. and D.L. Conover, Teratogenic effects of radiofrequency radiation. IEEE Eng Med Biol
Mag, 1987. 42: p. (S).
502. Last, J.M.e., A dictionary of epidemiology. Edited for the International Epidemiological Association.
(Oxford University Press). 1983: p. (C).
503. Laurence, J.A., et al., Biological Effects of Electromagnetic Fields-Mechanisms for the Effects of
Pulsed Microwave Radiation on Protein Conformation. J Theor Biol, 2000. 206(2): p. 291-298 (R).
504. Laval, L., P. Leveque, and B. Jecko, A new system for 900-MHz microwave irradiation during in vitro
studies (Meeting abstract). Bioelectromagnetics Society, 21st Annual Meeting, 20-24 June, Long
Beach, CA, Abstract No. P-48, 1999: p. 197 (E).
505. Lazzi, G. and O.P. Gandhi, Realistically tilted and truncated anatomically based models of the human
head for dosimetry of mobile telephones. IEEE Trans Electromagn Compat, 1997. 39(1): p. 55-61 (E).
506. Lazzi, G. and O.P. Gandhi, A mixed FDTD-integral equation approach for on-site safety assessment
in complex electromagnetic environments. IEEE Trans on Antennas and Propagation, To appear: p.
507. Lebovitz, R.M. and L. Johnson, Testicular function of rats following exposure to microwave radiation.
Bioelectromagnetics, 1983. 4(2): p. 107-114 (C).
508. Lebovitz, R.M. and L. Johnson, Acute, whole body microwave exposure and testicular function of rats.
Bioelectromagnetics, 1987. 8(1): p. 37-43 (C).

509. Leonard, A., A.J. Berteaud, and A. Bruyere, An evaluation of the mutagenic, carcinogenic and
teratogenic potential of microwaves. Mutat Res, 1983. 123(1): p. 31-46 (K) (C).
510. Lerchl, A., et al., Investigations on the effects of 900 MHz electromagnetic fields on growth, melatonin,
and testicular cell composition in Djungarian hamsters(Meeting abstract). Bioelectromagnetics
Society, 20th Annual Meeting, 7-11 June, St. Pete Beach, FL, Abstract No. A-2-2, 1998: p. 11-12 (E).
511. Li, J.R., et al., TP53 Tumor Suppressor Protein in Normal Human Fibroblasts does not Resp