Abstract

The increasing tendency for people to work outside their home community--one of the most striking of modern demographic changes--has relevance to a recent aetiological hypothesis about childhood leukaemia: that a community's immune response to an underlying infection can be disturbed by increases in new social contacts. This was tested in the only 28 former county boroughs in which accurate comparisons of workplace data from the 1971 and 1981 censuses are possible--because their boundaries were left unaltered by the major reorganisation in 1974. After ranking the districts according to extent of commuting increase, a significant trend in leukaemia incidence was found at ages 0-14 (P less than 0.05) and a suggestive one at ages 0-4 (P = 0.055). Among ten similar sized groups of county districts ranked by commuting increase, the only significant increases (P less than 0.001) of leukaemia in 1972-85 at ages 0-4 and 0-14 were in the highest tenth for commuting increase. These excesses persisted after excluding Reading, a major part of an area where an excess of leukaemia has been linked to the nearby nuclear establishments at Aldermaston and Burghfield. This whole area has experienced greater commuting increases than 90% of county districts in England and Wales. The findings are consistent with other evidence supporting the above hypothesis; they also suggest that contacts between adults may influence the incidence of leukaemia in children.