Abstract IA10: Neighborhood matters: Residential levels of air toxics and increased risk of early childhood acute lymphocytic leukemia

Background: The evidence of benzene and 1,3-butadiene as leukemogens in adults comes primarily from occupational epidemiologic studies. The major environmental sources of benzene and 1,3-butadiene, which are classified as hazardous air pollutants by the U.S. Environmental Protection Agency (EPA), include vehicular exhaust and emissions from industrial facilities. Exposures to these pollutants during in utero and early life may play an etiologic role in childhood leukemia. Hence, we conducted a population-based case-control study (Symanski et al., Environmental Health 2016;15:70) to examine the role of residential levels of benzene and 1,3-butadiene in early childhood leukemia in a state that has an abundant number of petrochemical plants and extensive road traffic in and around urban areas. Methods: We identified cases of acute lymphocytic leukemia (ALL) diagnosed in children under age 5 years from the Texas Cancer Registry for the years 1995-2011 and linked cancer registry data, where possible, to birth records in the state. Using Texas birth certificates, we selected 10 controls per case, frequency matching by birth month and year to allow for the same potential for exposure. Geocoded maternal addresses at delivery were linked to the U.S. EPA’s National-Scale Air Toxics Assessment modeled census-tract level estimates of benzene and 1,3-butadiene in ambient air. We evaluated maternal and infant characteristics that were available from birth certificates, along with neighborhood-level metrics of disadvantage, as potential confounders. We applied mixed-effects logistic regression models to evaluate associations between air pollutants and childhood leukemia in single and copollutant models. Results: In single-pollutant models, adjusted odds ratios were elevated for infants whose mothers lived in census tracts with higher benzene and 1,3-butadiene levels. In adjusted models that included benzene and 1,3-butadiene together, odds of ALL remained elevated for 1,3-butadiene (medium-exposure group: OR=1.22, 95% CI: 1.00-1.50; medium-high group: OR=1.28, 95% CI: 1.01-1.63; and high-exposure group: OR=1.40, 95% CI: 1.06-1.86) relative to the low-exposure group. In contrast, results were close to the null value for benzene in all exposure groups. Conclusions: Studies are needed to confirm our finding that early life environmental exposure to 1,3-butadiene (and not benzene) increases risk of ALL in young children. Our results add to the growing literature that health