Friday September 13 2013
The study found little leukaemia risk from nuclear power stations
“Nuclear plants 'do not raise child cancer risk',” says BBC News – a headline which addresses a controversy that has been around since the 1980s.
The headlines are based on a study looking at more than 20,000 children in Great Britain who developed childhood leukaemia or non-Hodgkin lymphoma before the age of 15, between 1962 and 2007. It compared how close they lived to nuclear power plants when they were born with the same information in similar children who were cancer-free.
The researchers found no link between how close the children lived to the plants at birth and their risk of childhood leukaemia or non-Hodgkin lymphoma
The study benefits from using data on a large number of cases from across Great Britain over a long period. This increases the chance that they would be able to identify a link if one existed.
However, the results may be influenced by factors the researchers could not measure or take into account in their analyses. And while the number of cases analysed was large, not many people live near to nuclear power plants in the UK, which would make it more difficult to detect an effect if there was one.
This new UK evidence is in agreement with the most recent report from the UK’s Committee on Medical Aspects of Radiation in the Environment (COMARE) – see box above right.
The study authors and other sources have reasonably concluded that although the new UK findings are reassuring, it makes sense to continue monitoring to ensure that, if there is any risk, it will be detected.
Where did the story come from?
The study was carried out by researchers from University of Oxford and the University of Manchester. The work of the Childhood Cancer Research Group (CCRG) has been supported by the charity CHILDREN with CANCER (UK), the Scottish Government, and the Department of Health for England. One of the study authors was supported on the project by a legacy left in a will. The study was published in the peer-reviewed British Journal of Cancer.
The BBC News site covers this story in a balanced way.
What kind of research was this?
This was a case-control study which looked at whether there is a link between living near a nuclear power plant and leukaemia in young children.
A possible link between childhood cancer and nuclear power first came to public attention following a television report in the 1980s. This report suggested that there were more cases of cancer than expected in young people in the vicinity of what is now known as the Sellafield nuclear power plant. This led to the setting up of the Committee on Medical Aspects of Radiation in the Environment (COMARE), which analyses data on rates of cancer in children in the areas around nuclear plants in Great Britain.
A report from COMARE in 2005 found that although a number of excess cases of some kinds of childhood cancer were found for nuclear sites whose main function was not the generation of electricity they “found no evidence of excess numbers of cases in any local 25 km area” for 13 nuclear power stations. A 2011 COMARE report assessing published research and analysing British data concluded that “in spite of its limitations, the geographical analysis of British data is suggestive of a risk estimate for childhood leukaemia associated with proximity to an nuclear power plant that is extremely small, if not actually zero.”
There has been criticism that the UK research has so far specifically looked at geographical areas and the incidence of cancer in these areas. Results of this type of study are limited by the fact that childhood cancers are so uncommon that each area is likely to only have a few cases, and differences between areas may be hard to detect. Results may also be influenced by the fact that people move in and out of the areas and may be lost to the study.
In the current study, researchers used a case-control approach to assess the possibility of a link. This type of study is good for assessing potential causes of rare diseases, such as the childhood cancers. This is because the case-control approach allows researchers to gather together a larger group of individuals with the disease (cases) than would be found in a single area, and compare their past exposures with those of individuals without the disease (controls).
A recent case-control study from Germany found that young children living within 5km of a nuclear power plant were at increased risk of developing leukaemia by age five when compared to other areas, but no difference was found for other cancers.
What did the research involve?
The researchers identified all children diagnosed with childhood leukaemia or non-Hodgkin lymphoma in Great Britain between 1962 and 2007 (cases) and matched them to children who did not have cancer (controls). They compared how close the cases and controls lived to nuclear power plants at birth. They also compared where children with childhood leukaemia or non-Hodgkin lymphoma lived at diagnosis with children who had other types of cancer.
To identify cases, researchers used data from the National Registry of Childhood Tumours. This registry records diagnoses of malignant disease and non-malignant tumours in the brain or spinal cord in children aged under 15 living in Great Britain. It is estimated to contain over 97% of all of these diagnoses in Great Britain since 1970, and to contain at least 99% of the leukaemia diagnoses for the period assessed by the study. This information was linked with birth data for children born in Great Britain.
The researchers selected a ‘control’ child of the same gender and approximate age (two weeks to six months’ difference) for each ‘case’ child from the same birth registry. These controls had to be cancer-free at the age at which their matched case had been diagnosed. The children’s home addresses at birth and diagnosis were obtained from the registries.
Thirteen nuclear power plants in mainland Britain were considered in the analysis. These were:
- Dungeness A and B
- Heysham 1 and 2
- Hinkley Point A and B
- Hunterston A and B
- Sizewell A and B
The distance between the children’s residence at birth or diagnosis and the nearest nuclear power station was calculated. Plants which had closed were also considered due to the possibility of lingering radioactivity. The main analyses excluded the Sellafield plant, as this was the site which initially generated this interest and researchers wanted to see if an analysis of all other sites would support or refute a link. They also re-did their analyses including the Sellafield plant to see if this affected their results.
The analyses of leukaemia and non-Hodgkin's lymphoma focused on children under five years of age, as this was the group found to have an increased risk in the German study. Of the children in this age group with known birth locations, 10,071 were diagnosed with leukaemia or non-Hodgkin lymphoma. Of these, 9,821 were pair-matched with the controls with both having residences at birth determined with acceptable accuracy.
The researchers took into account other factors which could influence their results, such as social class (based on father’s occupation on birth certificate) and characteristics of the region where they lived (based on mother’s residence at the child’s birth), such as the level of deprivation, population density, and whether it was rural or urban.
The addresses of the child controls was only known at their time of birth. Therefore, the difference in their residential proximity to nuclear power plants at the time of the cases’ diagnosis could not be assessed. Because the previous German study suggested that only leukaemia risk was associated with proximity to nuclear power plants, the researchers decided to compare proximity to nuclear power plants at diagnosis between 10,618 children with childhood leukaemia or non-Hodgkin lymphoma with 16,760 children diagnosed with other forms of cancer. As the children with different cancer types were not matched, the analyses took into account differences between the groups in age at diagnosis, gender, region of residence (England, Wales, or Scotland), as well as social class.
What were the basic results?
There was no significant association between proximity of a child’s home to a nuclear power plant at birth and their risk of being diagnosed with leukaemia or non-Hodgkin lymphoma under the age of five years (odds ratio (OR) for risk associated with living within 5km of an nuclear power plant: 0.86, 95% confidence interval (CI) 0.49 to 1.52).
There was also no significant difference in how close under-fives with leukaemia or non-Hodgkin lymphoma were living to a nuclear power plant and how close under-fives diagnosed with other types of cancer were living to a nuclear power plant (OR for risk associated with living within 5km 0.86, 95% CI 0.62 to 1.18).
There was also no significant association for other child age groups. Inclusion of Sellafield in the analysis did not affect the results.
How did the researchers interpret the results?
The researchers concluded that their “results show little evidence of an increase in risk of [leukaemia or non-Hodgkin lymphoma] to children aged under five years from living in the vicinity of a [nuclear power plant]”. They say that their results do not confirm the findings of the recent German case–control study that found a link.
The current study is the first to use a case-control design to assess the possibility of a link between living close to a nuclear power plant and childhood leukaemia or non-Hodgkin lymphoma in Great Britain. In contrast to a recent German case-control study, it found no significant link.
The advantages of the study design are that it could collect a larger number of cases than would be possible in a study of a small area. The use of registry data allowed the study to include most childhood cancer cases in the whole of Great Britain over a long period, which gives it a better chance of detecting differences if there are any.
The study also has a number of potential limitations, several of which the researchers themselves discuss:
- They note that the absence of a statistically significant result represents only weak evidence that there is no effect, as it may be that a study is not large enough to detect a big difference. However, they point out that the fact that their results suggested a non-significantly lower risk of leukaemia or non-Hodgkin's lymphoma nearer to the plants in the under fives suggests that the study size is not hiding an increase in risk from living near to the plants.
- The control children were selected from the same birth registers as the cases, which may make the cases and controls more alike in terms of distance from a nuclear power plant. This would reduce the ability of the study to detect a difference, but the authors note that their study did cover the whole of Great Britain over a long period, which should counteract this.
- The fact that nuclear power plants in Great Britain tend to be located on the coast, away from heavily populated areas, means that fewer cases live near to the nuclear power plants than in Germany. This could also reduce the ability of the study to detect differences.
- Measuring distance of home address from the nuclear power plants is not a direct measure of radioactivity. However, as proximity to nuclear power plants is a source of concern for people, the authors say that it is reasonable to study this. They also say that monitoring near to nuclear power plants has suggested that radiation levels are not high enough to suggest that they would increase risk.
- The children’s address at birth and diagnosis may not reflect where they actually lived through their lives.
- Because the data was obtained from registries, there are a lot of factors that could be potential confounders that they could not measure or take into account in their analyses.
- Ideally the researchers would have been able to identify the addresses for the control group of matched children who did not have cancer, at the time the case children were diagnosed. As they were not able to do this, they had to use children with other types of cancer as a control group, which was not ideal.
- The study did not look at sites which dealt with nuclear material for purposes other than generating electricity.
Overall, the current study does not support the findings of the recent German study of a link between living near to a nuclear power plant and risk of leukaemia or non-Hodgkin's lymphoma in the under-fives. However, the authors note that the findings of the German study cannot be easily dismissed, and that it makes sense to continue to monitor populations that may be at risk.
Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.