Santini R, Santini P, le Ruz P, Danze JM, et al. (2003):

This was a survey study using a questionnaire on people living near cellular phone base stations. The participants were recruited through information given by press, radio and web sites. The questionnaire asked questions about "non specific health symptoms". Participants were asked about age, gender, symptoms, and estimated distances from base stations. Comparisons were made regarding prevalence of symptoms between people living at different distances from the base stations. Those living more than 300 metres away were the "reference group".

Some complaints (nausea, loss of appetite, visual disturbances) were seen more frequently in participants living less than 10 metres from a base station. Others (irritability, depressive tendencies, loss of libido) were seen more frequently in those up to 100 metres away, and other complaints (headaches, sleep disturbances, feeling of discomfort) in those up to 200 metres from the base station.

The authors advise that "base stations should not be sited closer than 300 metres to populations". This would seem an excessive use of the precautionary principle, given that the participants were self-selected and could include a large number biased in their perceptions of the dangers of base stations. It is reasonable to assume that those responding to the appeal to participate would be more liable to have concerns about base stations, and that those living closest to a base station would have the greatest concern. A randomly selected population would have been more reliable, and comparisons could have been made to a more representative sample of the general population