Viewpoint

Radiation Risks

The effects of radiation on health have been the subject of intense debate for nearly 50 years -- and probably will be for at least another 50. Unfortunately, this debate has become very polarised. Because of this, selective reporting of the scientific evidence is to some extent inevitable in any review such as that undertaken last year by the government's Special Committee on the safety of nuclear-powered vessels. However, there is a fine line between being selective, and being deliberately misleading.

Official limits on radiation exposures of nuclear workers and the general public are based on the recommendations of the International Commission on Radiological Protection (ICRP). The ICRP believes that even low-dose radiation is likely to be harmful. Setting dose limits involves a compromise between the needs of the nuclear industry, and the need to minimise adverse health effects.

The Special Committee used a recent ICRP publication, called ICRP 60, in their assessment of radiation risks. One of the principle conclusions they reached was that "the relationship between radiation dose and cancer induction, as given for instance in ICRP 60, is soundly based and that no recent studies give cause to question the findings of this document".

However, there undoubtedly are recent studies that conflict with these findings. The question is whether, taken together, such studies give sufficient grounds to increase the estimates of disease resulting from exposure to low-dose ionising radiation. The Committee state that there are no scientific studies linking radiation doses from "normal activities associated with the nuclear industry" and cancer. Later they mention a study which did not show any increase in cancers in people living near nuclear installations. However there are several well-conducted scientific studies which do show such an increase. These studies were not mentioned by the Committee.

It can be argued that increased cancer rates in the vicinity of nuclear installations cannot be caused by exposure to radiation, because the estimated doses involved are too low to account for them. However, faced with a discrepancy between theory and the epidemiological evidence, it is prudent to assume that the theory is inaccurate rather than to discount the evidence. The true doses to the public from the nuclear industry are unknown, and the population exposed is potentially very large.

Thus even though, as the Committee state, "at low doses, only a very few of the exposed individuals will develop cancer", the number of people affected may still be substantial. Subsequently, referring again to epidemiological studies suggesting a link between radiation and cancer, the Committee suggest that "none of the recent studies has been able to conclude categorically that an effect does or does not exist".

In fact, it is never possible to come to such a conclusion from the results of a single study. A study either shows an effect which is statistically significant, or it does not. Causation -- or lack of it -- is inferred from an evaluation of all the relevant scientific evidence. Several studies, which again the Committee do not mention, show statistically significant increases in cancers associated with low-dose exposures to radiation.

The health of workers in the nuclear industry has been the subject of several studies. In a recent UK study, one of the conclusions was that "there is evidence for an association between radiation exposure and mortality from cancer". The Committee did not mention this. Instead they stated that the authors "concluded that their study does not provide sufficient evidence to justify a revision in risk estimates for radiological purposes".

Risk estimates are estimates of the number of cancers that will result from a given dose of radiation, and are used in setting dose limits. The Committee compare current ICRP risk estimates with those derived from the UK study and from a US study. They do not mention the results of the most recent US study, which suggested a risk estimate ten times greater. The Committee apply confidence limits to the ICRP risk estimates, giving a false impression of the accuracy and reliability of these estimates.

The ICRP risk estimates are extrapolations from studies of populations exposed to high doses at high dose rates (high doses received quickly). One of the ICRP's assumptions is that low dose, low dose rate exposure will be relatively less harmful. However, this assumption is not supported by the evidence from populations that have had this type of exposure. It is just this type of exposure that occurs as a result of nuclear discharges or accidents such as Chernobyl.

In a section on Chernobyl, the Committee state that "except for the workers involved at the very early stages of the accident, and who suffered from acute radiation effects, essentially the only physical effects expected will be those due to late cancers".

This is incorrect, as diseases other than cancer are also expected to occur in the children conceived by exposed adults. There is good evidence that this has happened; this evidence was not mentioned. The Committee go on to estimate the number of cancers expected to occur in people directly involved in the accident, or living in the surrounding area. They give no estimate of health effects expected to occur in people in other countries, merely commenting that in these people "no physical effects due to the Chernobyl accident will ever be discernible..."

Good evidence of such effects has already been found. Whilst it is true that many of the health effects of the Chernobyl accident will never be identified, long-term effects such as cancer will be numerically greatest not in those most directly involved, but in other countries. One conservative estimate (from the pro-nuclear US Department of Energy) of the radiation dose to the world population as a result of Chernobyl is 1.2 million person-sievert. Multiplying this by the ICRP risk estimate of 5% per sievert gives an estimated 60,000 excess fatal cancers. It is quite possible that both these estimates are too low, and much higher estimates of total deaths are quite plausible.

Finally, the Committee discuss a recent article regarding thyroid cancer in children exposed to fallout from Chernobyl:

[The authors of the study] comment that the short time between the accident and the observed increase is surprising in comparison with other studies. They point out further that there is no evidence that the diagnostic or therapeutic use of radio-iodine in man carries a carcinogenic risk and that Iodine-131 has provided a safe and effective treatment of Graves' disease In adults.

From this, the impression is given that the Chernobyl accident is unlikely to have been the cause of the childhood thyroid cancers. The Committee do not mention the conclusions of the authors:

We believe that the experience in Belarus suggests that the consequences to the human thyroid, especially in fetuses and young children, of the carcinogenic effects of radioactive fallout is much greater than previously thought.

The Committee conclude that "people's fears are far greater than the scientific facts warrant". But historically, scientists have consistently underestimated radiation risks, and ICRP risk estimates have repeatedly been revised upwards over the past 50 years. It is up to each of us to decide if we wish to accept a compromise between the needs of the nuclear industry and the military, and risks to our health from radiation.

An extended fully referenced version of this is in the report of the Alternative Committee on Nuclear Ship Visits, from the Auckland Centre for Peace Studies.

Dr Simon Hales is a public health registrar.