Radiation emissions from a mobile phone handset can be 10 times higher in rural areas than in urban districts to compensate for the fact that base stations are located further apart.

The findings are based on a sample of 1,400 patients with brain cancer who were compared against 1,400 healthy people who had also been interviewed about their use of mobile phones.

But the scientists who carried out the research admitted that the overall number of cases involved was still small and that the findings do not prove that using mobile phones can cause brain tumors.

Professor Lennart Hardell, a cancer specialist at the University Hospital of Orebro in Sweden, said the results nevertheless point to a link between the dose of microwave radiation from a mobile and the risk of developing brain tumors.

"It's another piece of evidence, but of course we have to wait for further studies. This is a further step indicating that there is probably a problem and people should use the precautionary principle to limit their use of mobile phones, especially for children," Professor Lennart said.

The study, published in the journal Occupational and Environmental Medicine, investigated more than 1,400 Swedes aged between 20 and 80 who had been diagnosed with a malignant or benign brain tumor between January 1997 and June 2000.

The scientists found no link between the probability of developing a tumor and the time spent on the phone, but they did find a link between the risk of brain cancer and place of residence -- rural or urban.

Residents of rural areas who had been using a mobile digital phone for more than three years were three times more likely to be diagnosed with a brain tumor than those living in urban areas.

For those rural residents who had used a mobile digital phone for five years or more, the risk quadrupled compared to city dwellers. Yet the scientists found no such increased risk when they looked at older, analog mobile phones.

Professor Hardell suggested the reason was that digital phones use a system called adaptive power control, which automatically boosts the power output of the handset signals when base stations are located further away.

Radiation emissions from a mobile phone handset can be 10 times higher in rural areas than in urban districts to compensate for the fact that base stations are located further apart, he said.

"With analog phones the emissions are constant and we did not see this difference between rural and urban areas," Professor Hardell said.

For malignant tumors the difference was even greater, with rural residents running an eightfold increase in risk compared to those living in urban areas.

But Professor Hardell said that the absolute numbers involved are small and said that the findings must be treated with caution until further, large-scale studies are completed.

"The message is that people should use hands-free sets and limit their phone calls if possible," he said.

A spokesman for Britain's Health Protection Agency said that Professor Hardell's study was not designed to test the hypothesis that rural phone use is more dangerous than in the city.

"We do need to be precautionary about the use of mobile phones, especially by young children. We also need to be precautionary about this study, because other research has not found a clear link between mobile phone handsets and brain cancers," he said.

'We should wait for the results of the Interphone study being carried out at the moment. It is a large study being carried out in 13 countries and should give a good indication of whether or not there is a real cancer risk from mobile phone use.'