The relationship between non-occupational environmental exposures and the risk of acute lymphoblastic leukaemia in children

[Truncated abstract] Background: The early age at diagnosis of childhood acute lymphoblastic leukaemia (ALL) suggests that prenatal exposures and those in early childhood may play a role. Diagnostic irradiation of the mother during pregnancy is the only known cause of childhood ALL while there is long list of potential environmental risk factors for ALL. The aim of this analysis was to investigate whether exposure of the mother before birth, of the father before conception or of the child to a range of environmental factors increased the risk of childhood ALL. Methods: This thesis used data collected in the Australian Study of Causes of Acute Lymphoblastic Leukaemia, which was a national, population-based, multi-centre case-control study of children aged younger than 15 years conducted from mid 2003 to early 2007. Cases were recruited from all paediatric oncology centres in Australia and controls were recruited by national random digit dialling. Data from 389 cases and 876 frequency-matched controls were collected about a range of environmental exposures: pest control treatments, painting around the home and the use of wood burners in the time period from the year before the pregnancy until the time the child was diagnosed for cases and the questionnaires were returned by controls; refuelling of vehicles by the father in the year before conception or by the mother in the year before and during pregnancy; abdominal x-rays before the child’s birth conception (fathers) or birth (mother), and the child’s life-time history of diagnostic x-rays; and whether the child or parents had ever lived on a farm. Data were collected from both parents, where possible, in mailed self administered questionnaires. Further details were collected about exposures to house paint and pest control treatments in computer assisted telephone interviews. Data were analysed using unconditional logistic regression, adjusting for study matching factors and potential confounders.