COMARE 10th Report: The incidence of childhood cancer around nuclear installations in Great Britain

Using a unique database (consisting of over 32,000 cases of childhood cancer that occurred in Great Britain between 1969 and 1993) COMARE studied the incidence of childhood cancer in the vicinity of all the major licensed nuclear sites (power stations and other nuclear installations) in Great Britain. Using the most appropriate statistical test for each site, we found no evidence of excess numbers of cases in any local 25 km area around any of the nuclear power stations. This distance would include either primary exposure to radioactive discharges or secondary exposure from re-suspended material.

Around other nuclear installations our study demonstrated similar results to previously published studies that showed excesses of some types of childhood cancer, with only one site being different. We found an excess of leukaemia and non-Hodgkin lymphoma (NHL) near Burghfield, Dounreay and Sellafield. Aldermaston, Burghfield and Harwell showed a significantly raised incidence of solid tumours in their vicinity. These results have been extensively discussed in previous COMARE reports.

One site differs from previously published results. This was Rosyth where the overall incidence of leukaemia and NHL in children living within 25 km of the site was close to the expected value, but there was evidence of a trend in risk with distance from the plant. This finding differs from previously published work that used similar but not identical methods. It is, therefore, not possible to comment on whether living near the site at Rosyth genuinely confers a higher risk of leukaemia and NHL. Clearly it is important to establish the nature of the differences between the two studies and detailed comparisons of the two data sets and methodologies are already under way. In contrast, no significant trend for the incidence of solid tumours with distance from the site was found, although there was a significantly increased incidence in the area within 25 km. This may well be a reflection of the previously reported high incidence of central nervous system tumours in the surrounding Fife and Lothian region.

Public and media concern about childhood cancer around nuclear installations has usually focussed on reports of ‘clusters’ rather than on analyses reporting high incidence rates. Such reports are typically not based on systematic analyses but reflect a perception that an unusual pattern of occurrence has been identified. Such perceptions may well be correct but the way in which most clusters come to notice does not permit any formal analysis of their importance to be made.

To avoid these problems we have, therefore, in this report considered the results of a systematic series of updated analyses concerning the incidence of childhood leukaemia and other childhood cancers in the vicinity of nuclear sites in England, Wales and Scotland. These analyses use new data and, for each site, a series of computations was carried out to determine the most appropriate statistical test. Analyses were then carried out to distinguish between a general increase in incidence in the area and a trend in incidence in relation to proximity to the site. Some statistically significant results are found in these new analyses but these are largely in line with previous findings. COMARE is engaged in an analysis of childhood cancer in the whole of Great Britain and this will be published as a COMARE report in the near future.

Press enquiries to Professor Bryn Bridges, OBE, Chairman of COMARE during the preparation of this report, Tel no: 01273-813627.

COMARE 10th Report "The incidence of childhood cancer around nuclear installations in Great Britain" Chairman: Professor B A Bridges OBE. ISBN: 0-85951-561-3 (£13.50). Downloads available from www.comare.org.uk and hard copies are available from the Information Office at the Radiation Protection Division of the Health Protection Agency.

Notes for editors

In COMARE’s Third Report (1989), the Committee concluded that the distribution of cases of childhood leukaemia, or other childhood cancer, around individual nuclear installations cannot be seen in a proper context in the absence of comparable information about the general pattern throughout the rest of UK. Of the five recommendations made in the COMARE Third Report, two related to this conclusion. Recommendation 4 stated that “studies of the geographical distribution of childhood cancer incidence on a nation-wide basis be carried out…thus enabling the patterns found around nuclear sites to be seen in the context of patterns in the rest of the UK.” Recommendation 5 of the Third Report went on to say that “once the results of the studies outlined in Recommendation 4 are available, this Committee should be asked to participate in a review of the evidence relating to the incidence of childhood cancer around nuclear installations.”

The work initiated in response to these two recommendations has now been completed. Using a unique database of over 32,000 cases of childhood cancer, for the time period 1969 to 1993, analyses were carried out at county, county district and ward level throughout Great Britain. The data were also analysed in relation to population data and socio-demographic variables. We understand that this is the largest database on childhood cancer (in the sense that it contains details of the largest number of cases currently available for the types of analyses done here) anywhere in the world. This large database and the associated analyses are to be the topic of a separate report already in preparation by this committee. However, the information gathered so far does allow the review of the incidence of childhood cancer around nuclear installations in Great Britain and this topic is the subject of this our tenth report.

Download the 10th Report in Portable Document Format 

Top