Maby E, Le Bouquin Jeannes R, Faucon G. (2006). Short-term effects of GSM mobile phones on spectral components of the human electroencephalogram. Conf Proc IEEE Eng Med Biol Soc 1:3751-4.
Several studies have suggested that pulsed high-frequency electromagnetic fields emitted by mobile phones may have some adverse effects on the brain.
As part of France’s COMOBIO project, the aim of this study was to investigate whether the GSM (Global System for Mobile) signals affect the brain’s electrical activity.
The study consisted of two groups: nine healthy subjects and six patients suffering from right temporal lobe epilepsy, which was resistant to anti-epileptic drugs. The experimental session consisted of two sequential recording phases: in the first phase, there was no radiofrequency (RF) emission (sham-expose); in the second phase, the emission power was at maximum. The control session was also composed of two phases which were the same length as in the experimental session, but here the mobile phone did not emit any radiofrequency. Electroencephalogram (EEG) activity was recorded on the surface of the scalp. Analytic techniques were used to reduce EEG artifacts during recordings, due to perturbations, such as eye blinking. In a second aspect of the study, to evaluate the influence of RF on the spectrum global shape of the EEG signals, the authors analyzed the spectral correlation computed between two spectra of signals collected on the same electrode in phase 1 and phase 2. The studied frequency band was 0 - 40 Hz. Also taken into account were conventionally defined EEG rhythms corresponding to emotional level and sleep/wake states.
In healthy subjects, the test showed that the spectral correlations were statistically lower in the experimental session (0.984) than in the control one (0.991). In the epileptic patients, the radiofrequencies also showed a significant decrease of correlations when passing from the control session (0.990) to the experimental one (0.981). The authors interpret this to suggest that the radiofrequencies induced some modification to spectral components.
In healthy subjects, EEG spectral energy parameters in controls were higher than in the experimental ones for the frequency band of 0-40Hz. This EEG spectral power decrease took place in parietal, centroparietal, and central areas. In particular, this change was localized to the alpha-band spectrum on occipital electrodes. In contrast, for the epileptic patients, these modifications were demonstrated by an increase of the power spectral density of the EEG signal, with changes taking place over the entire scalp.
Discussion and Interpretation
For the healthy and epileptic patients, the correlation coefficients in the frequency domain showed some modifications caused by mobile phones. For the healthy subjects, results indicate for all frequency bands a significant decrease of the EEG signal energy in the presence of radiofrequencies. The authors indicated that a meaningful result was the EEG signal energy decreased for the alpha-band in the occipital region.
The authors conclude that based on the current analysis, GSM mobile phones do have a biological influence on EEG signals, but that these particular results are not sufficient to formulate a meaningful hypothesis for the electrophysiological effect.