Wade Allison does not question the dangers of high levels of radiation but says that, contrary to scientific wisdom, low levels of radiation can be easily tolerated by the human body.

Most scientists who have responded disagreed with Allison's conclusions but his comments have highlighted the lack of understanding of how the body deals with low doses of radiation, a crucial issue given it is increasingly used in modern medical procedures such as scanning and cancer treatment.

Nuclear crises, from the bombing of Hiroshima and Nagasaki to the meltdown of a nuclear reactor at Chernobyl, have created widespread fear and distrust of nuclear power — and global pressure to keep radiation at the lowest possible level, according to Allison, a particle physicist who makes his arguments in a self-published book, Radiation and Reason.

He says long-term data on the health of survivors of the atomic bombs have demonstrated how good the human body is at protecting itself from radiological and chemical attack.

"The ability to repair damage and replace cells, we discovered in the last 50 years, shows how radiation doesn't cause damage except under extreme circumstances," he says. "The radiation that a patient gets in one day from a course of radiotherapy treatment, it would take a million hours of exposure for someone standing in the radioactive waste hall of Sellafield. And, if you have radiotherapy, it goes on for several weeks."

Ionising radiation, the type from nuclear reactions, can break strands of DNA in cells and these can make a cell cancerous unless the body's machinery can fix the damage. Scientists have used data from Hiroshima and Nagasaki, plus that from experiments on animals and cell cultures, to create a measure of how much damage is caused by high levels of radiation. This has then been extrapolated back, in a straight line, to estimate the potential damage from low levels of radiation to create what is called the linear non-threshold (LNT) model.

"The problem with a lot of these discussions is that you eventually get to the point where you don't have any more data," said Professor Gillies McKenna of Oxford University, Cancer Research UK's expert on radiation oncology. "Even the data from Hiroshima and Nagasaki — there weren't enormous numbers of cancers created in those cases, so we have to extrapolate what we think would happen at low dose."

Since the end of the Second World War, scientists have worked on the basis that there is no dose of radiation so low that it is not dangerous. Allison, however, believes there is a threshold below which any radiation exposure is fully repaired by the body but this is a view mainstream scientists disagree with.

"I wouldn't say Allison's ideas are fanciful but when you weigh up all the evidence, the scientific authorities come to the conclusion that the LNT dose-response relationship for low doses is the best we can do," says Richard Wakeford, an epidemiologist specialising in the health effects of radiation at the University of Manchester.

Allison's hypothesis assumes that all of the DNA damage caused below a threshold of radiation dose can be fixed by the cells' internal machinery. "I can't see and nor do the majority of experts in the field how these processes can be 100 per cent effective," said Wakeford.

"Radiation is particularly effective at causing double-strand DNA breaks, which make it difficult for the repair mechanisms in the cells to repair them properly."

Where McKenna and other scientists do agree with Allison is that fear of radiation is a problem. McKenna's expertise is in the use of radiation to kill cancer cells. "People become so fearful of radiation that they avoid diagnostic tests that may save their lives or avoid radiotherapy when they have cancer that is much more likely to kill them than exposure to radiation. He [Allison] is right that it has become a little bit hysterical. People are now avoiding CT scans or avoiding building nuclear power stations when in most aspects, radiation is a very useful thing."

Half of cancer patients will be given radiotherapy and more than half of those will be cured by it, McKenna said. "In most instances, where you use radiation — certainly in medicine and in most other forms of industry — the benefits greatly outweigh the risks."

Treatment involves a dose of radiation directed at the cancer cells which is 10 to 20 times the dose that would be fatal directed to the whole body.

Some areas, such as Devon and Cornwall, have naturally high levels of radiation in the rock and yet they do not have high incidence of cancer.

"It would suggest to me that we can tolerate relatively higher doses of radiation, unless you add things such as smoking," said McKenna, adding that there were good scientists on both sides of the debate, "but you reach a point where you can't generate the data you need and I do think we need to be careful not to exaggerate the risks and increase the fears".

Nothing has generated quite as much cancer concern in the United Kingdom as the Sellafield power station in Cumbria. Concern about radiation leaks at the plant, known as Windscale when it was commissioned in 1956, grew over the years until in 1983, Yorkshire Television produced a documentary called The Nuclear Laundry, suggesting low-level radiation emissions posed a risk. In the 1990s clusters of childhood leukaemia cases were identified near the site.

Investigating those concerns has been the preoccupation of Comare, the government's expert committee on the medical aspects of radiation, since it was set up in 1985.

After years of painstaking work and many reports, it has yet to establish a link between radiation and childhood leukaemia. The evidence for some sort of infection, possibly caused by the movement from one area to another of people working at the plant, is far stronger.

Comare's chairman, Alex Elliott, a professor of clinical physics at Glasgow University, says there is a wide spectrum of views on the dangers of low-level radiation. "There are those who believe people like me are part of an international conspiracy to hide the dangers of radiation from the public," he said.

At the other end are the believers in "radiation hormesis", who say we live in a beneficent soup of low-dose radiation, which is essential for life and may even prevent cancer deaths.

Elliott steers a middle path. "The Comare view, along with the consensus worldwide, is that the present risk estimates are broadly correct," he said. "They keep being revised but if they are wrong, it is by no more than a factor of two or three in each direction." And, he said, "We believe the linear hypothesis should continue to be used."

It is almost impossible, he said, to carry out experiments that would prove that low-level radiation is dangerous or is not, because the risks are so small.

But radiation generates fear, he said. "Because we can't see, hear, smell or touch it, we are much less tolerant of radiation than anything else. We are definitely hysterical about radiation. We go to enormous lengths on the precautionary principle. I don't know how many people are killed on the roads each year but we live with that. We're not thinking of banning trucks. We're incredibly bad at risk-benefit analysis."

But Wakeford said that calculating the risks of low-level radiation is becoming increasingly important. "One of the big issues today is just how you manage these new, relatively high-dose diagnostic procedures such as CT scans.

This is probably the big issue as far as low doses are concerned. In the US, remarkably, the average citizen receives more dose from medical diagnostic procedures than he receives from background radiation, which is a dramatic increase from the last time this was assessed about 20 or so years ago. When you come to make an assessment about balance of risk about whether to give a child a CT scan or not, these are real considerations, not hypothetical at all."

Comare, in a rare respite from studying leukaemia clusters at nuclear installations, recently produced a hard-hitting report on sunbeds, calling for a ban on their use by under-18s. "At the minute, it would appear that more people are damaged by sunbeds than by nuclear power in the UK," Elliott said.