Cell Phone Research

Short-term effects of exposure to mobile phones - Effect of a 915-MHz simulated mobile phone signal on cognitive function in man

A W Preece, G Iwi, A Davies Smith, K Wesnes, S Butler, E Lim and A Varey

The Research:

There are a relatively large number of reports of short-term memory loss being associated with mobile phone use. There are other reports of longer term effects from chronic use of phones. It is surprising that, although no scientific evidence of harmful effects have been produced, a recent court case at Abergavenny concerned the possible need to affix a hazard notice on each phone. Indeed a conventional cell-phone is a small and powerful transmitter which is commonly held a few centimetres from the head. The limits (originally set down in 1988 in NRPB GS11 and updated in 1993) are 10Wkg-1 for small volumes and most phones are within this guideline. Two of the common types of phone are the analogue, usually 0.6 Watts constant wave but possibly up to 1 Watt, and the GSM. The latter is 2 Watts peak but has a pulsed 217Hz modulation, giving up to 0.25 Watts average. In practice this may be reduced in strong signal areas.

Short-term memory and attention can readily be assessed by a proprietary computer package designed by Cognitive Drug Research. This has been used in the department here for a number of years and already shown to be sensitive enough to demonstrate minor memory effects of a very large 50Hz field applied to the head. Accordingly experiments on 18 volunteers, done as a randomised three-way crossover trial comparing both an analogue and a GSM simulation, with no signal, was carried out last summer. Probably as might be expected, we detected no change in memory, attention or simple reaction. The hippocampus, thought to be a site of memory, is deep in the brain and since the typical penetration at 900MHz is only about 3cm, this organ is fairly protected. However, one parameter was accelerated and that was choice reaction time. This could have been a statistical fluke so the whole series was repeated with a fresh set of 18 volunteers. Additionally the test room was changed, and the experiments supervised by two different people. This time the possible confounders, caffeine, sleep, alcohol and medicines, were accounted. The second set of data was again analysed blind by CDR, with an identical result. The two data-sets were then re-analysed locally, and eventually combined, with a resulting improvement in statistical surety.


It was Dr Stuart Butler of the Burden Neurological Institute who looked at the data and offered an explanation - an improvement to synaptic transmission at the angular gyrus. This organ is behind the left ear and responsible for the interface between vision and speech needed for the choice reaction test. The immediately probable explanation seems to be a tiny increase in local perfusion which could be a direct effect of a fractional temperature rise. An alternative explanation suggested by other research workers was a non-thermal effect on heat shock protein production, said to be produced non-thermally in some simple animal models. A vascular surgeon has pointed out to me a number of papers showing a relationship between HSP and vascular vasodilatation - the end effect is the same as heating.


The refereeing by IJRB appeared to be thorough, if not tough, and resulted in worthwhile improvements to the paper. Copies of the draft were shown to the DOH, representatives of the FEI and a number of interested colleagues. Criticisms from the FEI concerned the use of a simulation rather than real phones, the thickness of the antenna (!), and the smallness of the change, in spite of its statistical robustness.


My own view is that the simplest explanation is for now the better explanation. There are a number of models suggesting that a phone transmission can raise brain cortex temperature by 0.01 to 0.1 degree depending on blood flow (Foster, USA) but I know of other studies that include temperature measurement, suggesting higher shifts than this. In each case equilibrium occurs in 10-15 mins. which is well within the time of the tests. The direct result of this study is that it indicates a need for further work. The increased blood flow hypothesis is easily tested (PET scanning, MRI, ultrasound). The effect on the angular gyrus merely requires the experiment to be repeated on the other side of the head. The possibility of non-thermal HSP production in mammalian tissue should be examined. We have already started ultrasound blood flow studies and detailed SAR measurement in the phantom head developed for the CEPHOS project.

Current Work:


The interest in TETRA prompted us to study 36 volunteers. Unfortunately we could not get access to real hand sets, so this was simulated, using audio modulation at 17.6Hz, 25% duty cycle into a 400MHz single-sideband source. Nothing showed! however, the Home Office kindly loaned us a real set to compare the SAR (exposure) levels. We were well short of the real thing. To be repeated.


Our colleagues on the Isle of Man with a co-operative set of 10-11 year olds, allowed themselves to be trained to test children with phones and CDR sofware. Looks like no effect, but we are still writing up for publication.