Perinatal risk factors for childhood cancer

Previous research suggests that many of the carcinogenic events leading to childhood cancer occur prior to birth. The case-cohort study described in this dissertation was designed to examine pre- and perinatal risk factors for childhood cancer using information from public health surveillance systems. The New York State Cancer Registry was used to identify 6,332 cancer cases among children born in New York State excluding New York City. Electronic birth records were used to determine exposure status to a wide variety of perinatal factors. Controls were selected from the New York State excluding New York City birth cohorts, also using the electronic birth records. Analysis was stratified by type of cancer and by age at diagnosis where appropriate.; An advantage of the case-cohort study design is that many risk factor and disease relationships can be explored, creating an efficient means of screening hypotheses for further study. Other strengths of the study include the relatively large sample size compared to other studies of childhood cancer and the reduction of recall bias and selection bias by using existing records rather than contacting study subjects. Losses to follow-up of birth cohort members due to deaths from other causes and out-of-state migration as well as lack of data on postnatal confounders were potential weaknesses of the study design.; Several specific hypotheses were tested in this study. Of the nine hypotheses, six produced significant results, including associations between high birth weight and leukemia, neuroblastoma, and Wilms tumor, as well as a strong association between extremely low birth weight and hepatoblastoma. Increased risk of neuroblastoma among infants exposed to maternal hypertension and acute nonlymphocytic leukemia with older maternal age were also suggested. The hypotheses that were not confirmed were increased risk of leukemia with first-born status or history of maternal fetal loss and astrocytoma with high birth weight. New findings, which can potentially lead to additional hypotheses for future study, include increase risk of several types of childhood cancer among children those mothers were overweight or obese during pregnancy or whose mothers gained relatively more weight during pregnancy.