NEA spells out low dose impacts

31 March 1999

The Nuclear Energy Agency has issued a summary report on the state of scientific knowledge regarding the health impacts of low doses of radiation. More than a century has passed since radiation was first discovered, but there is still little understanding of its more subtle effects.

The basic scientific principle which underpins radiation protection limits for both workers within the nuclear industry and the general public, is the linear, no-threshold (LNT) dose-effect hypothesis. An application of the precautionary principle, it assumes that there is no radiation dose level which does not increase the risk of an individual developing cancer.

However at low radiation doses a direct link between radiation exposure and cancer is impossible to prove and the evidence of an impact is gleaned from animal experimentation and epidemiological studies of exposed groups such as the survivors of the atomic bomb blasts at Hiroshima and Nagasaki and populations around Chernobyl and Chelyabinsk. There remains considerable scientific uncertainty as to what health impacts, if any, low radiation doses actually have.


The NEA report summarises the current state of knowledge in this area. What is known for certain includes:

• The chief somatic effect of ionising radiation at low doses is the induction of cancer. At high doses, greater than 500 mGy, deterministic effects (such as erythema, cataracts, infertility) are known to occur.

• Ionising radiation at dose levels of interest for radiation protection is considered to be a weak carcinogen.

• There is firm evidence of radiation-induced cancer risk in humans at acute doses in excess of 200 mGy.

• No positive biological effects have been observed in humans exposed to acute doses of ionising radiation.

• Various tissues and organs exhibit a wide range of sensitivity to radiation-induced cancers.

• Radiation-induced, solid cancers have a long latency period, generally greater than ten years. Leukaemia and thyroid cancer in children can appear as soon as a few years after exposure.

• Various host factors (such as age at exposure, time after exposure, gender, genetic predisposition, etc) and environmental factors (such as cigarette smoking, infectious agents, etc) influence cancer risk at exposure levels where radiation effects have been observed.

• Cellular repair mechanisms are known to exist. However, misrepair and residual DNA damage occur.

• The yield of primary molecular and cellular events sometimes depends linearly on absorbed energy. However, many multi-step biological processes are known to be non-linear.

• Epidemiological studies alone will not provide definitive evidence of the existence or non-existence of carcinogenic effects due to low dose or low dose-rate radiation; the lack of epidemiological evidence for the existence of low dose and low dose-rate radiation-induced effects is not proof that such effects do not exist.

• The developing embryo/foetus is more sensitive to exposure to ionising radiation than are children and adults.

• Epidemiological studies have not detected hereditary effects of radiation in humans with a statistically significant degree of confidence.


The report also lists what remains unknown:

• The shape of the dose-effect relationship at low doses and dose rates for radiation carcinogenisis in humans, is in question.

• The roles of host factors (such as age at exposure, time after exposure, gender, genetic predisposition, etc) and environmental factors (such as cigarette smoking, infectious agents, etc) as determinants of radiation risk are uncertain.

• For the same absorbed dose, different types of radiation (alpha, beta, gamma, neutron) show different efficiencies at inducing biological effects; the basis of biological effectiveness of different radiations at inducing late effects in humans at low doses and low dose rates are not yet sufficiently understood.

• The mechanisms of carcinogenisis, whether induced by radiation or by other agents, is believed to be a multi-step process which is not fully understood. The origin of cancer is hypothesised to be the result of mutational events to critical genetic loci, and of other factors such as hormone status, age, immune function, etc. The effects of radiation on specific steps of carcinogenisis are not fully understood.

• Although damage to DNA is assumed to be a key step in radiation carcinogenisis, it is not known what critical lesions in DNA are responsible for gene or point mutations and chromosomal aberrations leading to cancer. The cause of an individual cancer cannot be specifically tied to a given insult, such as radiation exposure.

• It is not known how many tumourogenic cells are necessary to produce a cancer in vivo.

• It is unclear why organs and tissues vary in radiosensitivity. At present we do not know whether the sensitivity to radiation can be predicted form the spontaneous incidence of most cancers.

• We do not have methods to measure an individual’s radiation sensitivity.

• The influence of repair processes on human radiogenic risk at low dose and low dose-rate is not fully understood; however, biological and chemical repair process of radiation damage are known to occur in cells. This contributes to the uncertainty in dose and dose-rate correction factors used to estimate radiogenic risk.

• It is unclear whether positive biological heath effects of low doses of radiation exists in humans.


These uncertainties have resulted in a debate that has continued for the last fifty years over the safe limits to radioactive exposure. The general pattern over time has been for the limits to be tightened, despite a number of radiobiologists opposing the LTN hypothesis. The gradual tightening of limits has been a result not only of scientific evidence; social, political and economic considerations have also been factors. A difficulty is that much of the scientific evidence is contradictory. The NEA summarises the difficulties:

“In spite of affirmations from different sides of the debate and in spite of the absence of a clear and definitive explanation of cancer induction mechanisms, we have to note that some data reinforce the use of a NTL model and some clearly demonstrate the existence of a threshold.... Moreover, it seems clear that the design of experiments could influence the shape of the relationship. The smaller the target the more the linear relationship is sometimes obvious, perhaps because primary molecular events are often linear (and may serve as a dosimeter), whilst on higher levels of biological organisation (tissues, organs, organisms) highly non-linear processes are likely and sometimes observed.”

Most studies have found difficulty in showing any statistical evidence of effects on cancer rates at exposure levels lower than 200 mSv, although some have claimed evidence as low as 50 mSv.

“In view of the current status of knowledge and of the ‘precautionary principle’, the use of the LNT assumption and of the current System of Protection is still justified where a unified approach must be applied to all sources and practices,” says the report.

However it does not dismiss the possibility that thresholds exist, but urges caution in their impact on policy.

“Information on the existence of thresholds or threshold-like effects in specific cases, essentially in the area of internal irradiation from radionuclides could be considered in the analysis and radiation protection management of those situations where this is relevant. The Committee on Radiation Protection and Public Health (CRPPH) should initiate a deeper reflection of the use of an ‘expertise approach’ in specific cases, noting that this is not inconsistent with the current System of Radiation Protection. This reflection should include the feasibility, and, in particular, the significance of the practical application and the potential misuse of such an expertise approach, perhaps illustrating this view with a series of case studies for which the expertise approach is appropriate.”

Another major issue is the difficulty in attributing a cancer in an individual to a specific event, such as radiation exposure. A number of other factors, such as genetic make-up and smoking form part of the equation and it is rarely possible to come to a definitive conclusion in any individual case. Work is being carried out into whether biomarkers may exist that indicate the early development of a tumour.

“Presently, it is not known whether tumour cells carry the signature of their causative agents. Biomarkers and biological dosimeters are not yet available at the dose levels relevant to radiation protection. However, developments in this area have a great potential to stimulate molecular epidemiological studies whose results, in turn, might significantly affect radiation protection.... Such developments will have implications in the area of the assessment of cancer causality in individuals. This will also have implications to national programmes in the areas of employment/employability, health insurance, and worker compensation.”

Genetic factors, and in particular, population sub-groups with increased susceptibility to radiation induced cancers is another issue which may, in the future, force exposure limits even lower. The NEA points out that the interaction of many environmental and individual factors, including genetics, smoking and the concentration of radon gas, combine to result in cancers. The report concludes with a summary of the difficulties of multi-causation:

“It can be said that interactions between radiation and other physical, chemical and biological agents are an important modifier in many biological processes and outcomes. Their implications for the limitation of individual and collective health risks in a unified concept including exposures from all important agents need careful consideration. To achieve this, fundamental research, as well as the development of conceptual models, are necessary. Based on these achievements, existing epidemiological data should be re-evaluated in view of these interactions and new studies should be developed to investigate specifically the effects of combined agents. However, in one particular case – radon exposure and cigarette smoking – new clear results could have public health policy, mitigation strategy and regulatory implications in the next few years. The CRPPH should continue to monitor progress in the area of combined effects research, and through this monitoring CRPPH members may attempt to foster international co-operation and collaboration among national research programmes.”

Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.