Maternal exposure to ambient air pollution and risk of early childhood cancers: A population-based study in Ontario, Canada.

Background: There are increasing concerns regarding the role of exposure to ambient air pollution during pregnancy in the development of early childhood cancers.

Objective: This population based study examined whether prenatal and early life (<1year of age) exposures to ambient air pollutants, including nitrogen dioxide (NO) and particulate matter with aerodynamic diameters ≤2.5μm (PM), were associated with selected common early childhood cancers in Canada.

Methods: 2,350,898 singleton live births occurring between 1988 and 2012 were identified in the province of Ontario, Canada. We assigned temporally varying satellite-derived estimates of PM and land-use regression model estimates of NO to maternal residences during pregnancy. Incident cases of 13 subtypes of pediatric cancers among children up to age 6 until 2013 were ascertained through administrative health data linkages. Associations of trimester-specific, overall pregnancy and first year of life exposures were evaluated using Cox proportional hazards models, adjusting for potential confounders.

Results: A total of 2044 childhood cancers were identified. Exposure to PM, per interquartile range increase, over the entire pregnancy, and during the first trimester was associated with an increased risk of astrocytoma (hazard ratio (HR) per 3.9μg/m3=1.38 (95% CI: 1.01, 1.88) and, HR per 4.0μg/m=1.40 (95% CI: 1.05-1.86), respectively). We also found a positive association between first trimester NO and acute lymphoblastic leukemia (ALL) (HR=1.20 (95% CI: 1.02-1.41) per IQR (13.3ppb)).

Conclusions: In this population-based study in the largest province of Canada, results suggest an association between exposure to ambient air pollution during pregnancy, especially in the first trimester and an increased risk of astrocytoma and ALL. Further studies are required to replicate the findings of this study with adjustment for important individual-level confounders.