“Hundreds of nuclear workers have died of heart attacks and other circulatory diseases brought on by radiation,” The Times reported. Other newspapers including The Guardian also covered the story. They described a study of about 65,000 individuals employed between 1946 and 2002 at the four nuclear power stations, Sellafield, Springfields, Capenhurst and Chapelcross.
The study found that although the overall death rate among the workers was lower than in the general population, people receiving higher doses of radiation were more likely to die from circulatory system disease. The Guardian reported that the finding was “particularly surprising since there is no established biological mechanism that would explain how radiation exposure might cause heart disease”.
Many of the newspapers quoted the researchers as saying that the study didn’t take into account other factors, such as diet and exercise, which could have caused the result.
In their published report, the researchers call for “further work” and say that the results of their analysis are not consistent with a simple causal interpretation (that is, that radiation causes circulatory system disease). The limitations with this study mean that it is not possible to conclude that exposure to ionising radiation causes heart disease or is itself responsible for increases in cardiovascular deaths.
Some information was not available, so it was not possible to adjust for factors that have an established link with heart disease. Further studies which take these into account are needed before the strength of association, if any, between ionising radiation and cardiovascular deaths becomes clear.
Where did the story come from?
Dr Dave McGeoghegan and colleagues from Westlakes Scientific Consulting in Cumbria carried out the research. The study was initially funded (until April 1 2006) by British Nuclear Fuels plc (BNFL) and subsequently by the Nuclear Decommissioning Authority. No conflicts of interest were declared. The study was published in the peer-reviewed: International Journal of Epidemiology.
What kind of scientific study was this?
There is an established link between radiation exposure and risk of death from cancer, and a suggestion, from studies of survivors of the atomic bombs dropped on Hiroshima and Nagasaki, that there might be a link with non-cancer deaths. Here, the researchers used data on male employees at nuclear power stations in the UK to examine the link between radiation exposure at work and deaths from non-cancer causes.
Between 1946 and 2002, 64,937 people worked at Sellafield, Springfields, Capenhurst and Chapelcross. The researchers used this large group to carry out a retrospective cohort study. Deaths and the causes of death up until the end of 2005 were identified through searches of the records of the Office for National Statistics, the General Register Office, the National Health Services Central Register and the National Register (for deaths before 1952).
All workers had originally been classified either as “industrial” or “non-industrial” workers. “Industrial” workers generally had managerial and technical jobs, while “non-industrial” workers were generally skilled manual labourers. The workers’ employment records also categorised them according to whether they worked with radiation or not and there was detailed information available on the levels of exposure for most workers. The 42,426 workers for whom these details were available were included in the final analysis. As there were so few female employees at these sites during this time, the study only included men.
The researchers determined the overall death rate over time and cause of death of all the workers. They then compared these figures with the expected deaths for that time in the northwest region of England, taking into account the age and gender of the workers.
After these overall comparisons, they divided the workers into groups according to their levels of exposure to radiation. As different types of radiation have different effects, the most meaningful measure of exposure for humans is the sievert (Sv), which is a calculation of the absorbed dose of radiation multiplied by a measure of harm associated with that exposure (i.e. a weighting). Using sieverts allows exposure to different types of radiation to be meaningfully compared.
The researchers compared death rates according to the total number of sieverts that individuals had been exposed to until the time of their death.
What were the results of the study?
Overall, the researchers found that the workers had lower rates of mortality than the general population in northwest England during that time. Generally, this applied to all causes of non-cancer disease and was most pronounced for respiratory diseases where there were 36% fewer deaths due to respiratory disease in employees than the general population. For circulatory disease, there were 16% fewer deaths in the employees compared to the country as a whole. These results are not wholly unexpected as an employed population are likely to be healthier than a general population (which includes both sick and healthy people). This is called the “healthy worker effect”.
When assessing employees according to their “industrial” or “non-industrial” categories, the study found that “industrial” employees had much higher rates of death than non-industrial employees, including a 1.3 times increase in risk of death from circulatory disease (such as heart disease and heart attacks).
When analysing employees according to the dose of radiation received, the researchers found a “dose-response” indicating that people exposed to higher levels were at greater risk of heart disease than those exposed to lower doses. They calculated that there was a 0.65 times greater likelihood of dying from circulatory disease per sievert of exposure.
What interpretations did the researchers draw from these results?
The researchers say that their study shows evidence of an association between exposure to radiation and mortality from non-cancer causes of death with circulatory disease in particular.
They express caution in their conclusions however, saying that the inconsistencies in this pattern across different groups of workers (“industrial” and “non-industrial”) and the lack of a strong biological reason for this association mean their results “are not consistent with any simple causal interpretation”.
What does the NHS Knowledge Service make of this study?
This is a retrospective cohort study, drawing together data that had already been collected. There are limitations that should affect any interpretation of the results:
- Though the researchers found a link between high doses of radiation and deaths from circulatory causes, they found inconsistencies in this response across the different categories of workers. This may indicate that the differences between these groups had an influence on their risk of death. The researchers have not taken into account other “confounding” factors that can contribute to cause of death, such as diet or exercise, and say that these “may account for some or all of the apparent dose-response for ionising radiation”. The use of “industrial” and “non-industrial” as an indicator of socio-economic status is an attempt to control for this important confounder, but may not have been a sufficient adjustment. The researchers acknowledge this.
- Another important confounder is “shift work”. The researchers say that many studies have associated shift work with an increase in the risk of death from circulatory disease. Their study does not take into account the fact that people may be working shifts that increased their risk. This therefore remains an important potential confounder in this study.
- Even at the highest level of exposure, and ignoring the other possible reasons for the increase in risk, there were only three per cent more deaths in this group than expected. This is a small figure.
- Overall, the researchers found that risk of cardiovascular death was increased by 0.65 times per sievert. One sievert is a very high dose of radiation. For example, the UK government recommends maximum exposure of 20mSV (one fiftieth of an SV) per calendar year for employees aged 18 or over. Therefore, the applicability of this result to current practice in facilities where low levels of exposure occur is unclear. Since the 1950s, practice is likely to have led to reduced exposure, and people in the nuclear industry now probably receive lower doses.
The newspapers may have overstated the significance of these results, not mentioning the important failure of this study to take into account important confounding factors. The researchers say that the workers’ socio-economic status had a more significant influence on mortality than how much radiation they were exposed to. They also say that it’s possible that a combination of adverse lifestyle factors associated with socio-economic status, together with stress and other factors associated with shift work, “may at least contribute to the apparent dose-response for cumulative external radiation dose”.
Sir Muir Gray adds...
Even though this is a big study, it is quite small to answer this type of question. What we need is a systematic review of all the studies of people exposed to radiation.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
The Guardian, 5 March 2008
The Daily Telegraph, 5 March 2008
Daily Mirror, 5 March 2008
Channel 4, 5 March 2008
The Times, 5 March 2008
Financial Times, 5 March 2008
The Independent, 5 March 2008
Links to the science
Int J Epidemiol 2008 [Advance Access published online on March 4, 2008]