Gestational and Early Life Exposures as Risk Factors for Childhood Lymphoma, Leukemia, and Wilms' Tumors:
[Thesis]
Marcotte, Erin
an Exploration of Birth Characteristics, Influenza and Respiratory Syncytial Virus Infections, and Pesticide Exposure
Ritz, Beate
2013
Ritz, Beate
2013
Childhood cancer is the second leading cause of death among children age 0-14 in the United States, but there are very few established risk factors and the etiology of most childhood cancers is still unclear. This dissertation research investigated the associations between birth characteristics and childhood lymphoma, exposure to infections and childhood leukemia, and pesticide exposure and childhood Wilms' tumor among children born in California between 1983 and 2007. We identified all cancer cases among children age 0-5 from the California cancer registry and matched them to California birth certificates. We randomly selected controls from the same birth certificate files. Our analyses included 478 lymphoma cases, 3402 leukemia cases, 863 Wilms' tumor cases, and 215,841 controls. We observed associations between several birth certificate variables, including parental race and some complications of pregnancy and labor/delivery, and childhood lymphoma. We also observed that early life exposure to infections, as estimated by proxies such as timing of birth in relation to community infections, was protective for acute lymphoblastic leukemia, particularly among first born children. Finally, we found associations between gestational and early life exposure to pesticides and Wilms' tumor. Our findings confirm previous reports on the relationship between race and childhood lymphoma, and Epstein-Barr virus infection may help explain this association. We found evidence in support of the hypothesis that delayed exposure to infections in early life is a risk factor for childhood leukemia. We also report novel associations between several individual pesticides and Wilms' tumor in young children. Due to the rarity of childhood cancers, each of these analyses is limited in sample size and further investigations are necessary to confirm our results.