Background: Although particulate matter is a known carcinogen, its association with childhood cancer is inconclusive. The present study aimed to examine the association between long-term exposure to particulate matter and childhood cancer. Methods: A retrospective cohort was constructed from the claims database of the Korea National Health Insurance Service, including children born in seven metropolitan cities in Korea between 2002 and 2012. Monthly mean concentrations of particulate matter with aerodynamic diameter <10 μm (PM10) and other air pollutants (NO2, SO2, CO, and O3) were calculated using data from the AirKorea. Monthly mean concentrations of particulate matter with aerodynamic diameter <2.5 μm (PM2.5) were estimated based on a data fusion approach. Cumulative exposure was assessed by averaging the monthly concentrations accounting for the residential mobility of the children. The occurrence of cancer was identified by the appearance of diagnosis codes in the claims database. Hazard ratios (HR) and 95% confidence intervals (95% CI) were calculated using Cox proportional regression, adjusting for potential confounders and O3 concentrations. Results: During the study period, 1,725 patients were newly diagnosed with cancer among 1,261,855 children. HR of all cancers per 10 μg/m3 increment in annual mean concentrations of PM2.5 and PM10 were 3.02 (95% CI: 1.63, 5.59) and 1.04 (0.74, 1.45), respectively. Conclusion: PM2.5 exposure was positively associated with childhood cancer in a large retrospective cohort with exposure assessment accounting for residential mobility.
