Effects of radiation
Our knowledge of radiation effects derives primarily from groups of people who have received high doses. The risk associated with large radiation doses is relatively well established. However, the risks associated with doses under about 200 mSv are less obvious because of the large underlying incidence of cancer caused by other factors. Radiation protection standards assume that any dose of radiation, no matter how small, involves a possible risk to human health. However, available scientific evidence does not indicate any cancer risk or immediate effects at doses below 100 mSv a year. At low levels of exposure, the body's natural repair mechanisms seem to be adequate to repair radiation damage to cells soon after it occurs.
Some comparative radiation doses and their effects 2 mSv/yr Typical background radiation experienced by everyone (average 1.5 mSv in Australia, 3 mSv in North America). 1.5 to 2.0 mSv/yr Average dose to Australian uranium miners, above background and medical. 2.4 mSv/yr Average dose to US nuclear industry employees. Up to 5 mSv/yr Typical incremental dose for aircrew in middle latitudes. 9 mSv/yr Exposure by airline crew flying the New York – Tokyo polar route. 10 mSv/yr Maximum actual dose to Australian uranium miners. 20 mSv/yr Current limit (averaged) for nuclear industry employees and uranium miners. 50 mSv/yr Former routine limit for nuclear industry employees. It is also the dose rate which arises from natural background levels in several places in Iran, India and Europe. 100 mSv/yr Lowest level at which any increase in cancer is clearly evident. Above this, the probability of cancer occurrence (rather than the severity) increases with dose. 350 mSv/lifetime Criterion for relocating people after Chernobyl accident. 1,000 mSv cumulative Would probably cause a fatal cancer many years later in 5 of every 100 persons exposed to it (i.e. if the normal incidence of fatal cancer were 25%, this dose would increase it to 30%). 1,000 mSv single dose Causes (temporary) radiation sickness such as nausea and decreased white blood cell count, but not death. Above this, severity of illness increases with dose. 5,000 mSv single dose Would kill about half those receiving it within a month. 10,000 mSv single dose Fatal within a few weeks.

Epidemiological studies continue on the survivors of the atomic bombing of Hiroshima and Nagasaki, involving some 76,000 people exposed at levels ranging up to more than 5,000 mSv. These have shown that radiation is the likely cause of several hundred deaths from cancer, in addition to the normal incidence found in any populationg. From this data the International Commission on Radiological Protection (ICRP) and others estimate the fatal cancer risk as 5% per sievert exposure for a population of all ages – so one person in 20 exposed to 1,000 mSv could be expected to develop a fatal cancer some years later. In Western countries, about a quarter of people die from cancers, with smoking, dietary factors, genetic factors and strong sunlight being among the main causes. Radiation is a weak carcinogen, but undue exposure can certainly increase health risks.
In 1990, the US National Cancer Institute (NCI) found no evidence of any increase in cancer mortality among people living near to 62 major nuclear facilities. The NCI study was the broadest of its kind ever conducted and supported similar studies conducted elsewhere in the USA as well as in Canada and Europe.
In the UK there are significantly elevated childhood leukaemia levels near Sellafield as well as elsewhere in the country. The reasons for these increases, or clusters, are unclear, but a major study of those near Sellafield has ruled out any contribution from nuclear sources. Apart from anything else, the levels of radiation at these sites are orders of magnitude too low to account for the excess incidences reported. However, studies are continuing in order to provide more conclusive answers.
Low-level radiation risks
A lot of research has been undertaken on the effects of low-level radiation. Many of the findings have failed to support the so-called linear no-threshold hypothesis. This theory assumes that the demonstrated relationships between radiation dose and adverse effects at high levels of exposure also applies to low levels and provides the (deliberately conservative) basis of occupational health and other radiation protection standards.
Some evidence suggests that there may be a threshold below which no harmful effects of radiation occur. However, this is not yet accepted by national or international radiation protection bodies as sufficiently well-proven to be taken into official standards.
A November 2009 technical report from the Electric Power Research Institute in USA drew upon more than 200 peer-reviewed publications on effects of low-level radiation and concluded that the effects of low dose-rate radiation are different and that "the risks due to [those effects] may be over-estimated" by the linear hypothesis4. "From an epidemiological perspective, individual radiation doses of less than 100 mSv in a single exposure are too small to allow detection of any statistically significant excess cancers in the presence of naturally occurring cancers. The doses received by nuclear power plant workers fall into this category because exposure is accumulated over many years, with an average annual dose about 100 times less than 100 mSv". It quoted the US Nuclear Regulatory Commission that "since 1983, the US nuclear industry has monitored more than 100,000 radiation workers each year, and no workers have been exposed to more than 50 mSv in a year since 1989."
In addition, there is increasing evidence of beneficial effect from low-level radiation (up to about 10 mSv/yr). This 'radiation hormesis' may be due to an adaptive response by the body's cells, the same as that with other toxins at low doses. In the case of carcinogens such as ionizing radiation, the beneficial effect is seen both in lower incidence of cancer and in resistance to the effects of higher doses. However, until possible mechanisms are confirmed, uncertainty will remain. Further research is under way and the debate continues. Meanwhile standards for radiation exposure continue to be deliberately conservative.