Geographical variation in mortality from leukaemia and other cancers in England and Wales in relation to proximity to nuclear installations, 1969-78.

The distribution of mortality from 11 causes of death (lymphoid leukaemia, other leukaemia, leukaemia of all types, Hodgkin's disease, other lymphomas, all lymphomas, multiple myeloma, lung cancer, other malignancies, all malignancies and all other causes) has been examined in three age groups throughout England and Wales over the period 1969-78. The reorganisation of local authority administration in 1974 meant that the smallest areas that could be examined were 400 county districts or (in some cases) approximate county districts formed by aggregating pre-1974 local authority areas. The variation in the numbers of deaths observed about the numbers expected was assessed using log-linear models to estimate the effect on the relative risk in each district associated with social class, rural status, population size, health authority region and proximity to one of 15 nuclear installations. Trends in risk with increasing proximity to an installation (as judged by the proportion of the population resident within 10 miles) were examined after adjustment for the other four variables. The results showed that in districts near to an installation there were significant excess mortalities in persons under 25 years of age from leukaemia (RR = 1.15, P = 0.01) and especially from lymphoid leukaemia (RR 1.21, P = 0.01) and from Hodgkin's disease (RR 1.24, P = 0.05) and a significant deficiency of mortality from lymphoid leukaemia in persons aged 25-64 years. No significant trends were observed with an increasing proportion of the population near to the installations and the greatest excess mortality from lymphoid leukaemia in young persons was observed in the districts with the intermediate proportion of the population (10.0-65.9%) near an installation