Exposure to ambient air pollution in Canada and the risk of adult leukemia.

There is a paucity of studies investigating adult leukemia and air pollution. To address this gap, we analyzed data from a Canadian population-based case-control study conducted in 1994-1997. Cases were 1064 adults with incident leukemia and controls were 5039 healthy adults. We used data from satellites and fixed-site monitoring stations to estimate residential concentrations of NO2 and fine particulate matter (PM2.5) for the period prior to diagnosis, starting in 1975 and ending in 1994. We modeled the average annual exposure of each subject. Odds ratios (OR) and their 95% confidence intervals (CI) were estimated using logistic regression, adjusted for age, gender, province, smoking, education, body mass index, income, and self-reported exposures to ionizing radiation and benzene. We found an 'n-shaped' response function between exposure to NO2 and all forms of leukemia: from the tenth percentile to the median (4.51 to 14.66 ppb), the OR was 1.20; 95% CI: 0.97-1.48 and from the 75th percentile to the 90th (22.75 to 29.7 ppb), the OR was 0.79; 95% CI 0.68-0.93. For PM2.5 we found a response function consistent with a linear model, with an OR per 10 μg/m(3) of 0.97 (95% CI 0.75-1.26). For chronic lymphocytic leukemia we found response functions that were consistent with a simple linear model, with an OR per 5 ppb of NO2 of 0.93 (95% CI 0.86-1.00) and an OR per 10 μg/m(3) of PM2.5 of 0.62 (95% CI 0.42-0.93). In summary, for chronic lymphocytic leukemia we found no evidence of an association with air pollution and with all forms of leukemia we found weak evidence of an association only at low concentrations of NO2. It is possible that these inconsistent results may have arisen because of unaccounted urban/rural differences or possibly from a selection effect, especially among controls.