Summary : Our attitude to ionising radiation is irrational, and easing safety limits would do far more good than harm, says Wade Allison

The word "radiation" frightens people, and little wonder. Ever since the cold war, the prevailing view has been that ionising radiation can do real harm to us without being seen or felt - and should be avoided at all costs. In fact radiation is much less harmful than we feared. Given the availability of carbon-free nuclear power, this makes a sea change in our view of radiation rather urgent.

Fear of radiation grew alongside descriptions of what might happen in the event of a nuclear war. In earlier decades there was genuine scientific uncertainty about radiation's long-term health effects, and scientists were unable to be reassuring. So, driven by universal popular concern, tight regulation was imposed to minimise public exposure.

Since 1950, public dose limits have been tightened by a factor of 150. Currently, the internationally recommended limit is 1 millisievert per year above the natural background level of about 2.5 millisieverts per year. For comparison, a typical CT scan might give you a dose of 5 millisieverts and a simple dental or limb-fracture X-ray 1/100th of that.

Much has been learned over the past half century from clinical medicine, radiobiology and accidents like Chernobyl. There is no doubt that a very high single dose is fatal, as the fate of the initial 237 firefighters at Chernobyl illustrates. Within a few weeks, 28 died, and 27 of those had received doses in excess of 4 sieverts.

However, many people receive much higher doses than this, albeit under very different circumstances. When a cancer patient is treated with radiation in a radiotherapy clinic, the tumour dies after absorbing a dose of more than 40 sieverts. During the treatment, healthy tissue and organs near the tumour get an incidental dose of some 20 sieverts, which is 20,000 times the recommended annual limit and at least five times the dose that proved fatal at Chernobyl.

How can tissue survive this friendly fire? A radiation dose is the same in principle, whether received in a hospital or elsewhere. But the critical point is that the therapeutic dose is spread over four to six weeks, giving cells time to repair the damage. Each day the healthy cells receive about 1 sievert, and just manage to repair themselves. The tumour cells receive a higher dose, and just fail to do so.

So much for acute effects, but what about longer-term ones? Very rarely, the damage is misrepaired, and the resulting error may eventually lead to cancer. To find out how often this happens, we need to compare the lifelong health data of a large number of people, some of whom have received a significant radiation dose and some who have not.

The nuclear bombs dropped on the Japanese cities of Hiroshima and Nagasaki in August 1945 provide us with the data we need. About 66 per cent of the original inhabitants of the two cities survived to 1950, since when their individual health records have been extensively studied.

By 2000, 7.9 per cent of them had died of cancer, compared with 7.5 per cent expected from rates found in similar Japanese cities over the same period (Radiation Research, vol 162, p 377). This shows that the extra risk caused by radiation is very small compared with the background cancer risk, and less than the 0.6 per cent chance of an American citizen dying in a road traffic accident in 50 years.

Not surprisingly, those who received higher doses developed more cancers. But those subjected to doses less than 0.1 sievert showed no significant increase in solid cancers or leukaemias. Nor did they suffer an increase in the incidence of deformities, heart disease or pregnancy abnormalities. So there is a practical threshold of 0.1 sievert for any measurable effect due to a single acute dose.

Given what we now know, from radiotherapy to the legacy of the attacks on Hiroshima and Nagasaki, it is clear that radiation safety limits are far too conservative. Evidently, our bodies have learned through evolution to repair or eliminate damaged cells, with a low failure rate. I suggest the upper limit might be reset at a lifetime total of 5 sieverts, at no more than 0.1 sievert per month. That would be a fraction of a radiotherapy dose, spread over a lifetime.

Such a revision would relax current regulations by a factor of 1000. This may seem excessively radical to some, especially those in the safety industry who have spent 60 years trying to reassure the public by regulating against all avoidable sources of radiation - which, after all, is what society asked them to do.

But common sense says that extra precautions are most needed when we know least, and in a reasoned approach to any new technology we should start with a cautious limit which may be relaxed later, as instrumentation improves and our appreciation of it grows. The regulation of ionising radiation has resolutely gone in the opposite direction, driven by fear.

Changing the limits would bring practical benefits. Radiation safety is a major contributor to the cost of nuclear power, so any relaxation should lead to big cost reductions. Given that we urgently need to develop carbon-free energy sources, that is hugely beneficial.

It should also lead to a more sensible attitude to nuclear waste. If treated properly, the quantities are small, it become harmless after a few centuries, and it may be buried at moderate cost. In any event, the effect of radioactive waste is a small matter compared with the global influence of carbon dioxide and leaked hydrocarbons. We should re-examine the environmental risks of radiation with the same radical attitude that is required for our own health.

Wade Allison is a nuclear and medical physicist at the University of Oxford and the author of Radiation and Reason (YPD Books). He has no ties to the nuclear industry.