The BBC's Chris Jackson: "It seems fears about the effects of radiation were unfounded"

The study, to be published in the British Journal of Cancer on Monday, also found that clusters of cases around industrial sites are the result of "population mixing" - with newcomers' children more at risk than those born to locals.

Childhood leukaemia is extremely rare and scientists have been investigating the higher incidence of the disease near sites such as Sellafield nuclear reprocessing plant in Cumbria, northern England, for around 20 years.

The study found in Cumbria that around half of all cases of two childhood cancers - acute lymphoblastic leukaemia (ALL) and the related disease Non-Hodgkin's Lymphona (NHL) - could be linked to an infection caused by outsiders moving into the rural community.

Researchers Heather Dickinson and Louise Parker studied the records of 119,539 children born in Cumbria between 1969 and 1989.

They found that youngsters were much more likely to develop either ALL or NHL if both their parents were born outside Cumbria.

Higher risk

"All children, whether their parents are incomers or locals, are at a higher risk if they are born in an area of high population mixing," Dickinson said in a statement issued by the Cancer Research Campaign, which publishes the British Journal of Cancer.

"But children of incomers seem particularly at risk because they are potentially being exposed to more new infections," she added.

Sir Richard Doll, the renowned cancer expert who first linked tobacco with lung cancer, said childhood leukaemia appeared to be "an unusual result of a common infection".

"You would get an increased risk of it if you suddenly put a lot of people from large towns in a rural area, where you might have people who had not been exposed to the infection," he said.

He added that the study supported a theory developed by Professor Leo Kinlen of Oxford University that childhood leukaemia was caused by infections introduced into a community by migrant workers.

Responding to the findings, Prof Kinlen said the next challenge was "to identify the infective agent, and at the moment scientists have little idea what it could be".