Forschungsgemeinschaft Funk e.V.

Edition Wissenschaft

Back to overview


Edition No. 19/E Juli 2003
In vivo studies on the biological effects of high frequency electromagnetic fields: a survey and future prospects
Dr. Thomas Tillmann and Dr. Jochen Buschmann,
The Fraunhofer-Institute for Toxicology and Aerosol Research in Hannover

People consider electromagnetic fields to be dangerous because they are not perceptible through our senses, they are generated to a large extent from impressive technical equipment and they transport energy. In order to counter this emotionally charged situation, a great number of investigations have already been carried out on the subject. A critical review of these studies reveals that this goal, for various reasons, was not always reached. Therefore, the subject of this lecture will be firstly a brief survey of the important studies that have been done on long-term effects on animals. As an example for these types of studies, I will present my own investigations on how pulsating high frequency electromagnetic fields effect the prenatal development of rats.
On the basis of this study various procedures and experimental approaches will be discussed, and the prospects for the future of an experimental strategy will be presented which seems to show the most promise.

1. Review of the Literature
The research being discussed here deals exclusively with long-term effects, since these are especially relevant with regard to human health. Long-term effects were defined as those with repeated irradiation and/or an observation time of four weeks or longer. This selection is based on an approach which has been derived from toxicological testing, e.g. of chemicals.

A few studies will be briefly referred to as follows:
Adey et al. (1999) apparently discovered that NADC signals had a "tumour protective effect" on spontaneous and chemically induced tumours of the central nervous system (CNS) in rats. However, these effects in a followup study resulted in no effects being demonstrated with the given field irradiation, and the experiment could not be replicated. (Adey et al., 2000). Unfortunately, only tumours of the CNS were investigated.

The investigations of Chagnaud et al. (1999) resulted in no effects being demonstrated during a twoweek irradiation of rats with GSMsignals on chemically induced tumours. Nevertheless, the exposure time was too short and the cohort size too small for the investigation to result in conclusive findings.

In a methodically well performed study by Chou et al. (1992) it was demonstrated that there was a significant increase in the occurrence of malignant tumours in rats caused by irradiation at 2450 MHz. The authors qualified their results because the number of benign tumours remained the same and the increase in occurrence in the group that was exposed was well within the historical control data.

Frei et al. (1998) found no effects on the development of tumours in mice with an irradiation of 2450 MHz in two different field strengths. In a much quoted study done by Repacholi et al. (1997) the authors found a significant increase in a specific type of tumour viz., lymphoma in transgenic mice after being irradiated at 900MHz. The results still have to be replicated and the method used in the study can be questioned, therefore the study was done again in Australia. The findings of Utteridge et al. were published in August 2002 and they could not confirm Repacholi's results.

Salford et al. (1993) found no effects on the development of induced brain tumours in mice which were irradiated at 915 MHz. Unfortunately, the investigations were only done on one type of tumour. Toler et al. (1997) in a methodically well written study found no effect on the development of tumours in mice which were irradiated at 435 MHz.

From the literature published it is evident that the data produced in animal studies concerning long-term effects, especially teratogenicity (the induction of birth defects) and fertility (being able to reproduce) and carcinogenicity (the triggering of cancer) is often contradictory. Nevertheless, it appears that despite all of the contradictory data concerning the final end-points for teratogenicity and fertility the effects that do occur are obviously thermally induced and the occurrence of athermal effects are not as likely.

Relating to the endpoint for carcino genicity, the total picture shows an even greater inconsistency, which is partly the result of methodical problems. To be able to deduct scientifically sound conclusions on the possible adverse health effects on humans based on animal experiments does not seem possible at the moment, even when the majority of the studies do not come close to revealing any health risks.

2. My own investigations
In order to better understand studies done with animal experiments; two of my own studies on the influence of pulsed high frequency EMF will be described in detail.

In the first study it was investigated whether or not pregnant rats exposed to typical GSM fields in the range of the human exposure limits (5.5 W/m2 ) would affect the prenatal development of their offspring. None of the investigated parameters demonstrated any indication of such effects occurring.
In a follow up study, experiments were done under improved irradiation conditions using a wave guide of considerably higher field strengths; the results showed no thermal effects (60 W/m2). The results indicated no clear effects, however, especially the loss of embryos within the irradiated group was clearly higher (even if it is of no statistical significance). However, in this study many of the parameters investigated for the control group went beyond the historical control data which was determined by our lab and concerning the irradiated group the parameters were even stronger influenced.
This difference could have been caused by differences in how the rats were kept in the wave guide used. Therefore, the effects are most likely to be placed in a grey area.

When both studies are considered together the conclusion is that the human limit values had no effect on any of the analysed end points in the animal experiments. If one accepts the effects, produced under our conditions using the highest field strength, which did not give rise to any thermal effects, as borderline effects, then the "no observed effect level" is somewhere between the two tested values. This practical example illustrates the problem involved in interpreting studies but methodical conclusions for future studies can be drawn from these studies. This especially entails carrying out a cage or a shelf control as well as the simultaneous use of several irradiated groups in one study. Further to this, when planning and carrying out such studies the experimental advantages and disadvantages of using free moving or immobile animals during irradiation must be taken into account.

3. Future Prospects
When comparing the different methods of approach, they can be divided into two essential categories: those whose goal it is to present an analysis of a concrete hypothesis and those whose goal it is in the first place to serve as a global risk assessment.

The first type is strongly marked by pure academic research; often with high scientific values and carried out using very sensitive model systems. An extrapolation to the target organism viz. humans, is often only possible in a limited capacity and it is (especially concerning negative results when analysing individual hypotheses) very difficult to make statements concerning general adverse health risks for humans.

The second type of study uses a toxicologically marked approach in the sense of applied science. In various experimental approaches, groups of experimental animals are exposed to different field strengths and as many as possible potential health relevant end points are examined. This approach is best described as "apical"; it is well suited for an extrapolation of the results to humans owing to the numerous parameters that are investigated; this allows for a estimation to be made concerning health risks. However, these types of experiments more often fail than the first type due to their broadband character because they are not always sensitive enough for specific health risks.

Encouraging communication between the two types of research groups could result in a great degree of synergy. When on the one hand, pure research would allow that target results could be replicated in applied tests and on the other hand, the results from applied investigations could be confirmed with mechanistic approaches and this could lead to a considerably better starting point when assessing health risks.

With regard to the special field that has been discussed here, e.g. investigations on the possible EMF effects on health, the Forschungsgemeinschaft Funk is predestined in future to play the role of the mediator and moderator between both types of research groups. This could effectively help to close existing scientific gaps and to more quickly achieve the objective that the discussion concerning experimental data becomes less emotionally charged in future.

Back to overview