In 2013, I discussed several epidemiological studies providing good evidence of radiogenic risks at very low exposure levels.
A powerful new study (1) has been published in Lancet Haematology which adds to this evidence. However the study’s findings are more important than the previous studies, for several reasons.
First, it provides “strong evidence”, as stated by the authors, of a “dose-response relationship between cumulative, external, chronic, low-dose, exposures to radiation and leukaemia”.
Second, it finds radiogenic risks of leukemia among nuclear workers to be more than double the risk found in a previous similar study in 2005. The excess relative risk of leukaemia mortality (excluding workers exposed to neutrons) was 4.19 per Gy. In 2005, a similar study (2) among nuclear workers (also excluding those exposed to neutrons) in 15 countries by several of the same authors found an ERR of 1·93 per Sv. In other words, the new study’s risk estimates are 117% higher than the older study. The clincher is that the new study’s estimated risks are much more precise than before.
Third, it confirms risks even at very low doses (mean = 1·1 mGy per year). Unlike the Japanese bomb survivors’ study, it observes risks at low dose rates rather than extrapolating them from high levels.
Fourth, it finds risks do not depend on dose rate thus contradicting the ICRP’s use of a Dose Rate Effectiveness Factor (DREF) which acts to reduce (by half) the ICRP’s published radiation risks.
Fifth, it finds radiogenic leukemia risks decline linearly with dose, contradicting earlier studies suggesting a lower, linear-quadratic relationship for leukemia. It strengthens the Linear No Threshold (LNT) model of radiogenic risks, as it now applies to leukemias as well as solid cancers.
Sixth, the study finds no evidence of a threshold below which no effects are seen (apart from zero dose).
Seventh, the study uses 90% confidence intervals and one -sided p-values. In the past, 95% intervals and two-sided p-values were often incorrectly used which had made it harder to establish statistical significance.
Explanation for change
In an earlier version of this blog posted on June 29 2015, I’d written that the increase between the 2005 study and the present study was 50%, ie up from 1.93 to 2.96 per Gy. This was because the study’s Discussion section specifically compared these two studies and their risks, stating the older study’s leukemia risk was smaller and less precise.
However a detailed examination of the report reveals the following sentence in the para immediately before the Discussion section. “We assessed the effect of excluding people who had recorded neutron exposures; we showed a positive association for leukaemia … (ERR per Gy 4·19, 90% CI 1·42–7·80, 453 deaths)…”. To make sure readers get the point, the risk is greater when neutron exposed workers are excluded.
This is important because the 2005 study excluded workers exposed to neutrons. Therefore the correct comparison is between the risks for non-neutron workers, that is between 4.19 and 1.93 per Gy – an increase of 117%, rather than 50%.
I’ve written to the report’s authors about this but have not received any replies yet, perhaps due to summer vacations. I shall keep readers up-to-date on any progress.
The study’s credentials are pretty impeccable. It’s a huge study of over 300,000 nuclear workers adding up to over 8 million person years, thus ensuring its findings are statistically significant, ie with very low probability of occurring by chance. Also, it’s an international study by 13 respected scientists from national health institutes in the US, UK, and France, as follows.
- Centers for Disease Control and Prevention, US
- National Institute for Occupational Safety and Health, US
- Department of Health and Human Services, US
- University of North Carolina, US
- Drexel University School of Public Health, US
- Public Health England, UK
- Institut de Radioprotection et de Sûreté Nucléaire, France
- Center for Research in Environmental Epidemiology, Spain
- UN International Agency for Research on Cancer, France
Funding was provided by many institutions, including US Centers for Disease Control and Prevention, US National Institute for Occupational Safety and Health, US Department of Energy, US Department of Health and Human Services, Japanese Ministry of Health Labour and Welfare, French Institut de Radioprotection et de Sûreté Nucléaire, and the UK’s Public Health England.
This study powerfully contradicts the views of ill-informed and inexperienced journalists (including the UK writer, George Monbiot) and self-styled scientists who argue that radiation risks are over-estimated and even that radiation is somehow good for you. Hormetic effects are neither found nor discussed in this study: such irrelevant effects are regarded by real scientists as beneath their consideration. The impressive list of contributing scientists and their national institutions here should serve to make radiation risk deniers reconsider their views. This is particularly the case for US risk deniers, in view of the many US agencies and US scientists backing the study.
The study pointedly comments that “At present, radiation protection systems are based on a model derived from acute exposures, and assumes that the risk of leukaemia per unit dose progressively diminishes at lower doses and dose rates.” The study shows this assumption is incorrect. The authors therefore join with WHO and UNSCEAR scientists in their views that DREFs should not be used. The question remains whether the ICRP will accept this powerful evidence and scrap their adherence to using DREFs. I advise readers not to hold their breaths.
As regards the implications of their study, the authors interestingly choose to comment – not on exposures from the nuclear industry – but from medical exposures. They state “Occupational and environmental sources of radiation exposure are important; however, the largest contributor to this trend is medical radiation exposure. In 1982, the average yearly dose of ionising radiation from medical exposures was about 0·5 mGy per person in the USA; by 2006, it had increased to 3·0 mGy.2 A similar pattern exists in other high-income countries: use of diagnostic procedures involving radiation in the UK more than doubled over that period3 and more than tripled in Australia.4 Because ionising radiation is a carcinogen,5 its use in medical practice must be balanced against the risks associated with patient exposure.6
This is all correct and worrying, especially the revelation that medical radiation doses increased 6-fold in the US and doubled in the UK between 1982 and 2006. The authors add “This finding shows the importance of adherence to the basic principles of radiation protection—to optimise protection to reduce exposures as much as reasonably achievable and—in the case of patient exposure—to justify that the exposure does more good than harm.” The same, of course, applies to exposures from the nuclear industry – the actual subject of their research.
(1) Leuraud, Klervi et al (2015) Ionising radiation and risk of death from leukaemia and lymphoma in radiation-monitored workers (INWORKS): an international cohort study. The Lancet Haematology Published Online: 21 June 2015.
(2) Cardis E et al (2005) Risk of cancer after low doses of ionising radiation: retrospective cohort study in 15 countries. BMJ 2005;331:77.
I thank Dr Alfred Körblein and Professor Keith Baverstock for their insights in preparing this article. Any errors are my responsibility.