in vivo studies

It consists in exposing living animals (mice, rats, guinea-pigs, etc.) to electric and magnetic fields.

Exposure time is variable, and can last as long as the animal’s life.Tests make it possible to determine if the fields have affected the embryonic development, growth, fertility, behaviour or physiology of the animal.

Advantages usually the same as for human studies, but :

• A large number of animals can be studied (contrary to humans)
• Relatively cheap compared to clinical studies
• More invasive procedures can be used to investigate biological effects from exposure to electromagnetic fields
• Chronic exposure can be studied (e.g., life-time exposures of mice and rats)
• Specific animal models can be used (e.g., extra sensitive strains, genetically modified animals, …)
• ….

It is indeed difficult to extrapolate the results of animal studies to the case of actual human exposure. Animals are not humans and the biological effects observed in the animal are obtained under very specific experimental conditions.

Here are some limitations that prevent results in animals to be automatically transposed in humans:

• Applied doses are not always equivalent between species
• Differences in life expectancy
• Differences in size
• Differences in diet
• Differences in genetic variability
• Differences in amount of antioxidants
• Ethical problems
• …

A crucial point in EMF experiments is to make sure that fields are actually the variable responsible of whatever effects tested. It means that two groups are needed: the first one, the exposed group, will be placed under EMF exposure, while the other, the control group, will be sham exposed. The only difference between both groups is the EMF exposure. All other parameters need to be accurately controlled. Researchers cannot avoid working with two groups in their labs. Control and exposed groups must be identical in all relevant ways except for the EMF exposure of the experimental group.

Studies that compared the results of their exposed group to the results of a control group found in the literature are not valid.

To avoid observer bias/error, all experiment stages should be made by the same scientist and, in any cases in double blind.

Double blind in animal studies means that neither the person that breeds animals, nor the person that takes a blood sample, for example, have information on which group is really exposed or not. Each group needs to be treated in the same manner. The purpose of a double blind experiment in animals is to avoid observer bias and even some placebo effects in animals.

Here is a caricature of a study that not reaches the double blind objective.  It is not valid.
“Contrarily to the picture, each group needs to receive the same treatment.”

It is obvious that each group also needs to be physically in the same conditions, but it is not always the case: vibrations, noise, higher or lower temperature, light… can disturb the life of the exposed group. Let us think for example of a power generator that should be noisy or even slightly vibrating.

It is important that researchers have accurate information on animal exposure. Unfortunately, studies are too often deficient with respect to the exposure assessment.

“Rats sleeping just next to the coil are over exposed compared to other rats.”

As intensity quickly decreases with distance, an animal could spend too much time just above the coil while another could be less exposed because of its movement in the cage.

The best solution to avoid differences in intensity should be to place animals in closed tubes surroundings by coils.

Of course a huge cage seems stressless than this kind of confined tube. However, by experience, scientists know that it does not create more stress in animals.
Comparison with a control group in the same condition is compulsory.

Characteristics of exposure system are another flaw in lots of published scientific papers. Poor information is provided concerning the generator, the frequencies used and their harmonics, the signal (continued or pulsed) …

Mice, rats or guinea pigs are classical animals in labs. According to the aim of the studies, they will be normal or transgenic. Transgenic means that animals are genetically modified to become more susceptible to certain pathologies, skin cancer or leukaemia for example.

According to the difficulty of having perfect experimental conditions and the fact that all parameters cannot be controlled, results of one study mean nothing.

“Results of a single study are not sufficient to validate a theory and to break open a bottle of champagne.”

“It is compulsory to replicate a study and to compare with results of other laboratories. “

According to the World Health Organization, a single study is accurate to formulate a hypothesis, not to validate a theory.

Statistical analyses are an important part of a research. Results need to be analyzed with cautious. Each step requires to be carefully defined: number of animals in each group, parameters to be evaluated, statistical tests… According to the group tested (size and characteristics) and the parameters taken into account, the statistical analysis will be different.

In some published studies, statistical analyses are difficult to interpret as details are not often provided.

“Statistical analysis is a full-time job: only looking figures is not enough. Data need to be thoroughly analyzed. Better not to play at being God!”

Moreover, let us remind that we can make statistics lean towards what we want it to lean towards. Statistical analyses always provide values. Common sense is crucial to evaluate their accuracies, before going further.

The interpretation of these values is another potential flaw. Researchers need to keep in mind the results of previous studies, the characteristics of their methodology… in order to brightly discuss the results.

• Information sur les types de recherches en PDF (882 Ko)

• Une seule étude, ce n’est pas assez!