This is the second part of a two-part article. In Part 1 we introduced the prevalence of cancer, the survival rates and the common causes of this medical condition. We discussed the hallmarks (traits) of cancer cells and the mechanisms that have been proposed by scientists for carcinogenesis.
In Part 2 we discuss the findings of researchers who have investigated the occurrence of radiation-induced cancers in defined groups of people using a science called radiation epidemiology. We will present the evidence for an association between ionising radiation and specific types of cancer known as radiogenic cancers and whether it can be proven that individual cases of cancer were caused by radiation.
Epidemiology of the atomic bomb survivors
The health of the survivors of the atomic bombings of Japan over the course of their lifetimes is being investigated in the Life Span Study by a Japanese-American research organisation called the Radiation Effects Research Foundation (RERF). RERF have compared the survivors’ health with a control group of people who were living in Japan at the time of the atomic bombings but who were not exposed to radiation. In particular, RERF have investigated whether there are relationships between health outcomes and the doses of radiation which individuals had received.
The investigators estimated radiation doses for the participants based on their locations when the atomic bombs exploded (Figure 1); being closer to an explosion increased the dose and being indoors reduced the dose due to the buildings providing some shielding1.
The RERF scientists have calculated the excess number of leukaemia deaths produced by exposure for different doses of radiation (Table 1)2.
For example, if we consider the third row of data in Table 1:
• 6,304 survivors received a dose to the bone marrow in the range of 200 – 500 mSv.
• 27 people in the survivors’ group died of leukaemia.
• The epidemiologists calculated that 17 people in a control group of the same size also died of leukaemia.
• Thus, there are 10 (27-17) excess leukaemia deaths due to radiation.
RERF researchers also calculated the excess number of solid cancers (non-leukaemia cancers) produced by exposure for different doses of radiation (Table 2)3.
Overall, the data from the Japanese atomic bomb survivor studies indicate that the higher the radiation dose received the higher the increased risk of cancer.
The Linear No-Threshold Model
The data points from the Life Span Study have been used to contribute to the Linear No-Threshold (LNT) Model, which states that there is no dose that is too low to progress to cancer (no threshold) and that increasing the dose increases the risk in a linear manner, e.g. doubling the dose doubles the excess risk of getting cancer (Figure 2)4. There is no relationship between the size of the dose received and the severity of the cancer.
The American research body the National Academy of Sciences have used this model to estimate that if 100 people received a dose of 0.1 Sv, one person may get cancer. This compares with 42 out of 100 people who would get cancer during their lifetime for other reasons (Figure 3)4.
The American National Council on Radiation Protection and Measurements (NCRP) have assessed epidemiology studies for different groups of people including X-ray workers, nuclear power workers and patients who had received CT scans5. NCRP found some of these studies were more supportive of the LNT model than others, however, there was sufficient evidence for the NCRP to conclude that overall, these studies supported the use of the LNT model for radiological protection. This means that this model should be used, for example, for setting dose limits for radiation workers.
The accuracy of the LNT model has been disputed by some scientists for doses below 100 mSv, partly because there are not enough data points. Epidemiologists have calculated that a study group of 50,000 is sufficient to assess the effects of a 100 mSv dose (approximately 87,000 survivors were recruited for the Life Span Study), but 5 million participants would be required to assess the effects of 10 mSv6.
Alternatives to the linear no-threshold model
Researchers have proposed alternatives to the LNT model for doses below 100 mSv with the intention of reconciling epidemiology with current knowledge about cell biology (Figure 4).
For example, if we consider the third row of data in Table 1:
• The hypersensitivity model states that health risks may be greater than LNT predicts at low doses.
• The threshold model implies a dose below which there are no health risks.
• The hormesis model suggests that low doses of radiation may be beneficial to health.
Researchers who are researching the possibility of hypersensitivity are considering the non-targeted7 effects of ionising radiation especially the bystander effect and genomic instability, and we have already discussed the latter (in Part 1). The bystander effect is when unirradiated cells show effects as a consequence of receiving signals from cells which have been irradiated (Figure 5).
Both genomic instability and the bystander effects could potentially result in a low dose of ionising radiation having a greater biological effect than predicted by the LNT model. However, it is not understood how non-targeted effects impact the health risks of radiation.
Advocates of the threshold model maintain that the LNT model over-estimates the health risks of radiation8,9 because this model does not take into account that living organisms have evolved in the presence of low dose radiation. This means that certain responses of the cell to DNA damage such as DNA repair, cell death and pausing the cell cycle should be considered to be evolved defences against cancer. They maintain that cancer can only occur if a dose is large enough (above a threshold) to overcome these defences.
Researchers who support hormesis hold that radiation can be beneficial in small amounts. They contend that organisms can adapt to benefit from receiving a low dose of radiation, e.g. there have been some studies which show that treating cells with small doses of radiation reduces the effects of subsequent larger doses10. This may be due to the initial small doses activating the DNA repair response in the exposed cells.
Currently none of these three alternative models has sufficient evidence to replace the LNT model, which continues to be used by most scientists, though with an awareness of its potential limitations.
The results of the Japanese Life Span Study and other epidemiological studies indicate that receiving a dose of radiation increases the risk of getting certain types of cancer (called radiogenic cancers) and not others (Table 3)11.
The type of cancer that an individual may get due to ionising radiation depends upon the type of radiation and the route of exposure (Figure 6).
The survivors of the atomic bombings of Japan were externally exposed to gamma and neutron radiation, which had the ability to reach all of the body’s organs and hence cause multiple types of cancer.
In contrast, some types of exposures are more localised and have more specific health consequences such as when radioactive iodine, an emitter of beta particles and gamma rays, was released into the environment by the 1986 Chernobyl nuclear power plant accident. This radioactive iodine entered the food chain via grass eaten by cows, which then produced contaminated milk that was drunk by people.
This issue was most serious in parts of Ukraine, Belarus and the Russian Federation where the highest level of radioactive substances were deposited by the accident (Figure 7).
The human body naturally concentrates iodine in the thyroid gland located in the throat and the concentration of radioactive iodine in the thyroids of the Chernobyl accident survivors has increased the occurrence of thyroid cancer in this group. This is a more common problem for those exposed as children, because they had smaller thyroids than adults and hence received higher doses from drinking the contaminated milk12, 13.
Historically, radioactive iodine was a short-term hazard, because it has a half-life of 8 days, meaning that its abundance in the environment decreased to negligible levels a few months after the Chernobyl accident14.
People can reduce this particular risk by taking potassium iodide (iodine pills), which prevent radioactive iodine from being absorbed by the thyroid. In addition, thyroid cancer can be successfully treated by killing the cancer cells with the same chemical agent which brought them into being, radioactive iodine14.
Scientific knowledge of radiogenic cancers comes primarily from epidemiological studies. A limitation of this is that epidemiology cannot say whether an individual case of cancer was due to radiation. A goal of ongoing research is to find biomarkers (specific DNA mutations or patterns of mutations) which could prove this, as currently none are known.
In this article we have discussed the epidemiological evidence for the linear no threshold model which states that the probability of having cancer but not the severity of the cancer increases with increasing dose. We have pointed out that the LNT model is disputed by some scientists for doses below 0.1 Sv and have described some of the alternative models whilst recognising that none of them has sufficient evidence to replace the LNT.
We have described radiogenic cancers with a particular emphasis on thyroid cancer. We have shared the scientific consensus that it is not currently possible to know if an individual case of cancer was caused by ionising radiation, though the search for appropriate DNA biomarkers is ongoing.
We at the CHRC do hope you have found this article informative and references are included for further reading. Please also refer to the Basic Information which can be found on the CHRC website: Basic Information – Centre for Health Effects of Radiological and Chemical Agents (chrc4veterans.uk)
1. Radiation Effects Research Foundation (RERF), Physical estimates of dose, RERF, viewed 16 March 2020,
<https://www.rerf.or.jp/en/programs/general_research_e/raditiondose_e/> Estimates of radiation dose caused by the atomic bombings of Japan.
2. Radiation Effects Research Foundation (RERF), Leukaemia Risks among Atomic-bomb Survivors, RERF, viewed 22 January 2020, <https://www.rerf.or.jp/en/programs/roadmap_e/health_effects-en/late-en/leukemia/>. Radiation and leukaemia.
3. Radiation Effects Research Foundation (RERF), Solid Cancer Risks among Atomic-bomb Survivors, RERF, viewed 20 January 2020, <https://www.rerf.or.jp/en/programs/roadmap_e/health_effects-en/late-en/cancrisk/>. Radiation and solid (non-leukaemia) cancers.
4. National Research Council, 2006, Health Risks from Exposure to Low Levels of Ionising Radiation: BEIR VII Phase 2. The National Academies Press, viewed 4 February 2020,<https://www.nap.edu/resource/11340/beir_vii_final.pdf>.
Radiation dose and cancer estimates using the Linear No Threshold Model.
5. National Council on Radiation Protection and Measurements (NCRP), 2018, Commentary No. 27 – Implications of Recent Epidemiological Studies for the Linear-Nonthreshold Model and Radiation Protection, NCRP, viewed 5 February 2020,
<https://ncrponline.org/wp-content/themes/ncrp/PDFs/Product-attachments/commentry/27/overview.pdf>. Discusses how different epidemiology studies support the Linear No Threshold Model.
6. Brenner, D.J. et al (2003) Cancer risks attributable to low doses of ionising radiation: Assessing what we really know, Proceedings of the National Academy of Sciences, 100 (24), pp. 13761-13766, doi: https://www.pnas.org/content/100/24/13761. Mathematical aspects of cancer risks and radiation.
7. Kadhim, M. et al (2013) Non-targeted effects of ionizing radiation-implications for low dose risk, Mutation Research, 752 (1), pp. 84-98, doi: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4091999/. Discussion of the bystander effect and genomic instability and the potential consequences for the linear no threshold model.
8. Scott, B.R. and Tharmalingam, S. (2019) The LNT model for cancer induction is not supported by radiobiological data, Chemico-Biological Interactions, 301, pp. 34-53, doi: https://www.sciencedirect.com/science/article/pii/S0009279718311013.
Support for threshold and hermetic models for radiation-dose response.
9. Tharmalingam, S. et al (2019) Re-evaluation of the linear no-threshold (LNT) model using new paradigms and modern molecular studies, Chemico-Biological Interactions, 301, pp. 54-67, doi: https://www.sciencedirect.com/science/article/pii/S0009279718310858. Discussion of the case for a threshold dose when considering radiation-dose response.
10. Shibamoto, Y. and Nakamura, H. (2018) Overview of Biological, Epidemiological and Clinical Evidence of Radiation Hormesis, International Journal of Molecular Sciences, 19, 2387, doi: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6121451/. Radiation Hormesis.
11. Industrial Injuries Advisory Council, 2016, Cancers due to ionising radiation, UK Government, viewed 6 February 2020,
<https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/502166/cancers-due-to-ionising-radiation-IIAC-report.pdf >. Types of cancer associated with radiation.
12. Weiss, W. (2018) Chernobyl Thyroid Cancer: 30 Years of Follow-up Overview, Radiation Protection Dosimetry, 182 (1), pp. 58-61, doi: https://academic.oup.com/rpd/article/182/1/58/5076213. An overview of the thyroid cancers produced by the Chernobyl accident.
13. Zupunski, L et al (2019) Thyroid Cancer after Exposure to Radioiodine in Childhood and Adolescence: 131I-Related Risk and the Role of Selected Host and Environmental Factors, Cancers, 11 (10), 1481, doi: https://www.mdpi.com/2072-6694/11/10/1481/htm. The relationship between thyroid cancer and the dose of radioactive iodine received.
14. World Health Organisation (WHO), 2006, Health Effects of the Chernobyl Accident and Special Care Programmes, WHO, viewed 12 March, 2020, <https://apps.who.int/iris/bitstream/handle/10665/43447/9241594179_eng.pdf?sequence=1>.
The health effects of the Chernobyl accident.