An ecologic study of childhood leukemia and population mixing in Ontario, Canada

Objective: To examine the association between population mixing and the incidence of childhood leukemia, specifically the acute lymphocytic leukemia (ALL) subtype among young children. Methods: This ecologic study was based on incidence rates of leukemia in children aged 0–14 years. The Ontario Cancer Registry was used to identify the residence of 1394 leukemia cases between 1978 and 1992. Ecologic units were composed of census subdivisions in a 5-year period. Percent population change, determined from the Census of Canada, was employed as a measure of population mixing. The relationship between population mixing and childhood leukemia was examined separately after stratifying by the level of geographic isolation, defined according to urban–rural status. Analyses were also conducted separately in specific age groups and for the ALL subtype. Results: Population growth in rural areas was associated with an increased incidence of leukemia, particularly for the ALL subtype in children aged 0–4 years (rate ratio = 1.8, 95% confidence interval 1.1–2.8, for a greater than 20% population change relative to no increase in population). In contrast, an elevated risk due to population mixing was not observed in urban areas. Conclusions: Results from this study are consistent with results from similar studies conducted in the United Kingdom, which are suggestive of a role for an infectious agent in the etiology of childhood leukemia, as proposed in the Kinlen hypothesis.