Summary of main findings Leukaemia and non-Hodgkin's lymphoma in children of male Sellafield radiation workers Heather O. Dickinson, Louise Parker International  Journal of  Cancer  2002;99:436-43 Children of radiation workers compared to other Cumbrian children
Between 1950 and 1991, children of radiation workers at the Sellafield nuclear reprocessing plant had twice the risk of leukaemia compared to other Cumbrian children.
• Within the village of Seascale (adjacent to the Sellafield site) children of radiation workers had nine times the risk of leukaemia compared to other Cumbrian children.
• Much of this excess risk can be explained by population mixing.  It is known that there is a higher risk of childhood leukaemia in areas where the population is more mobile and the population of Seascale was very mobile.
• After allowing for population mixing, children of Seascale radiation workers had only about four times the risk of leukaemia compared to other Cumbrian children.
 
Dose Response
• Among  children of radiation workers, the risk of leukaemia increased with the radiation dose which the father received while working at Sellafield before the child was conceived e.g. if a father received 100mSv of radiation before his child was conceived, the child's risk of leukaemia was on average about 60% higher.  However, the increase in risk with dose was higher inside Seascale and lower outside (relative risk at 100mSv= 2.0 and 1.5 respectively).
• This relationship between the child's risk of leukaemia and the radiation dose of the father before the child was conceived wasn't explained by population mixing.  It didn't change after allowing for population mixing, because among radiation workers, there was little correlation of the father's radiation dose and population mixing.
Comparison with Gardner's findings
Professor Gardner, of the University of Southampton, first reported this statistical association between the child's risk of leukaemia and the father's preconceptional radiation dose in 1990.  He used a case-control study which included only 1098 children born in West Cumbria, whereas our study included over a quarter of a million children born throughout  Cumbria.  The increase in risk of leukaemia with increasing radiation dose that we report is much lower than that reported by Gardner (only about a quarter).
• This is partly because the controls which Gardner selected were not representative of the distribution of radiation dose in the population.  This meant that Gardner over-estimated the effect of the father's radiation by a factor of about 2.
• It is also partly because Gardner thought all of the increased risk of leukaemia in children of radiation workers was due to the father's radiation dose, whereas most of the increased risk can be explained by the high level of population mixing within Seascale.  This led Gardner to over-estimate the effect of the father's radiation by another factor of 2.
Conclusion
• Most of the excess risk of leukaemia among the children of Sellafield radiation workers is due to population mixing.
• Some of the excess risk of leukaemia among these children may be due to the radiation dose which their fathers received while working at Sellafield before the child's conception, although Gardner over-estimated this effect, probably by a factor of about 4.