Prof Dudley Goodhead, Director, MRC Radiation & Genome Stability Unit, said:

"It is well known that exposure to ionizing radiation can cause cancer. In more recent years, continuing follow-up of the survivors of the atomic bombs in Japan, have shown that acute exposures to radiation are associated with an increased rate of death from heart disease also. The findings of the present study clearly suggest that even chronic exposure to radiation, spread over long periods of time such as received by some radiation workers in the past, may also be able to cause increased heart disease. On its own this study cannot prove such a relationship, especially since no firm mechanism has yet been identified, but it is a valuable addition to the further research that is needed."

Prof Richard Wakeford, Visiting Professor in Epidemiology, Dalton Nuclear Institute, University Of Manchester, said:

"That the death rate for blood circulatory diseases among BNFL radiation workers as a whole is significantly lower than the rate for the general population emphasises just how important are the background factors (smoking, diet, exercise, stress, etc., etc.) in determining the overall risk of these diseases. The pattern of results for external radiation exposure found in this study suggests that these background factors may have influenced the radiation findings, so that the statistical association with external radiation exposure must be treated with due caution - a cause-and-effect relationship cannot be reliably inferred, especially in the absence of a convincing biological explanation, and more research is required to properly understand the results."

Prof Bryn Bridges, former Chairman of the independent Advisory Group on Ionising Radiation of the Health Protection Agency, said:

"This is an important study on the British Nuclear Fuels workforce. While it shows that mortality in radiation workers is no worse than in non-radiation workers (and may even be better as a result of a smoking ban in radiation areas) it does provide evidence suggesting an increase in death from circulatory disease as a function of cumulative radiation dose. Taken at face value it would mean that an effect previously found in the A-bomb survivors is also present in those exposed to relatively low occupational doses, and to a greater extent than might be predicted. If this proves to be a real consequence of radiation exposure it would need to be taken account of in radiological protection. However, there is already a fair amount of evidence in the literature that is not consistent with the BNFL study and I suspect that epidemiology alone may not be able to give a clear-cut answer. Moreover, our understanding of what mechanisms might be involved in a possible effect is much poorer than our understanding of the mechanisms involved in cancer induction.

"Quite independently of this paper the Health Protection Agency has asked its independent Advisory Group on Ionising Radiation (AGIR) to look at circulatory disease in those exposed to radiation, particularly at low doses. This will be no easy task, but whatever the outcome of the AGIR review, the nuclear industry workforce can take some comfort in the fact that their mortality from circulatory disease is 16% lower, and from respiratory disease 36% lower than age matched men in the general population in NW England."

Dr Mark Little, Reader, Department of Epidemiology and Public Health, Imperial College Faculty of Medicine, said:

"This is a well-conducted analysis of this important worker cohort. The report documents statistically significant positive trends of mortality risk for circulatory disease, as well as various related endpoints (ischaemic heart disease, acute myocardial infarction, stroke, chronic ischaemic heart disease, diabetes) with radiation dose. However, there are a number of reasons for caution in interpreting the findings as representing causal associations. The excess risk per unit dose found is rather stronger, by about a factor of four, than that observed in the Japanese A-bomb survivors. Also, the fact that most specific mortality endpoints of non-cancer disease are elevated to a similar extent suggests that there may be bias. As with most other studies of radiation-exposed cohorts (apart from the A-bomb survivors), there is little adjustment for major cardiovascular risk factors, in particular cigarette smoking, diabetes, obesity, high blood pressure and high levels of blood cholesterol or low density lipoprotein – only socioeconomic status (a proxy for some of these variables) is adjusted for here, using a crude industrial vs. non-industrial classification. Specific occupational factors, in particular stress (e.g., related to shift work, which may well be associated with radiation dose) also have the potential to confound, and therefore seriously bias the results. It is of interest that there is significant heterogeneity for certain endpoints by employment type and radiation exposure, which may be reflect confounding by some of these factors.

"In summary this report suggests associations between mortality from non-cancer causes of death, particularly circulatory system disease, and exposure to ionizing radiation in the BNFL workers. As the authors state 'the tentative nature of biological mechanisms … [and] inhomogeneities in apparent dose–response, mean that the results are not consistent with any simple causal interpretation.' "

Professor Sarah Darby, Professor of Medical Statistics, Clinical Trial Service Unit & Epidemiological Studies Unit, University of Oxford, said:

"There is compelling evidence that ionizing radiation can increase the risk of heart disease. This evidence comes from cancer patients who agreed to take part in trials to see whether radiotherapy was helpful in the treatment of their cancer. In these trials, all the patients received the standard treatments, such as surgery and drugs, and some were selected at random also to receive radiotherapy while other similar patients were not. Radiotherapy was shown to be effective in the treatment of many types of cancer, but in some circumstances, the heart received some incidental irradiation leading to an increase in heart disease and reducing the overall beneficial effect of the radiotherapy.

"For the cancer patients who took part in the randomised trials, much of the radiation-induced heart disease was caused by relatively high doses (eg above 5 Gray or 5 Sievert) and the question of whether lower doses of radiation can cause heart disease is not yet resolved. At least a dozen observational studies have looked at the issue. Some of these studies, including the BNF study, have found that the risk of heart disease increased with radiation dose, while for others the findings were not readily compatible with an increase, and some even found a decreased risk of heart disease with increasing radiation dose. The likely explanation for these apparently inconsistent results is that, unlike the patients in the randomized trials, the people in the observational studies who received the bigger doses of radiation differed from the people who received the smaller doses by other factors as well, and that these other factors also affected their risk of heart disease. For example the people with the higher radiation doses might be more likely (or less likely) to be cigarette smokers than the people with lower radiation doses, or they might be more (or less) likely to be obese. The presence of these other factors, which have a substantial effect on the risk of heart disease, make the interpretation of the observational studies, including the BNF study, difficult.

"One possible way in which further information may be gained on whether or not there is a risk of radiation-induced heart disease at doses below 5 Gray (or 5 Sievert) is to examine the mechanisms of radiation-induced heart disease, and several studies focusing on this are in progress. Another option is to look at heart disease rates in women who have been treated with radiotherapy for breast cancer. The heart is positioned slightly to the left in the chest, so that for many radiotherapy techniques women with left-sided breast cancer receive a slightly higher radiation dose (typically about two Gray) than women with right-sided breast cancer (typically about one Gray). There is good evidence that, until recently, women with left-sided and right-sided breast cancer were equally likely to be irradiated. Also, it seems unlikely that women with left-sided and right-sided breast cancer differ appreciably according to other factors that affect heart disease risk. Therefore, comparison of heart disease rates in women irradiated for left-sided and for right-sided breast cancer may well be the best way to find out whether radiation doses of about a Gray cause heart disease."