Population Attributable Risk and Projected Population Impact Fractions for Lung Cancer and Outdoor Air Pollution in Canada

摘要

Introduction: Outdoor air pollution is classified as a carcinogen by IARC primarily on the basis of a relationship between PM2.5 and NO2 and lung cancer risk. This study evaluates the population attributable risk (PAR) and population impact fractions (PIFs) for counterfactual exposure distributions for PM2.5 and NO2 on lung cancer risk in Canada. In addition, census dissemination area-level PM2.5 and NO2 exposure measures from 1980-2014 were used to calculate PAR estimates, analysis of PM2.5 and NO2 trends overtime, and counterfactual PIFs. Methods: Using relative risk estimates calculated from a Canadian-specific meta-analysis, the relationship between PM2.5 and NO2 and lung cancer risk was modeled. This analysis assumed a log-linear dose response for a continuous exposure, representing levels during the relevant window of 20 years, with a 5-year lag. Exposure at the dissemination-area level, and a geographically weighted regression model was used to calculate PM2.5 and NO2 distributions from 1990-2009, allowing for PAR calculation for 2015, change in PM2.5 and NO2 levels overtime, and PIF counterfactuals. Results: Meta-analyses of four Canadian studies for PM2.5 and NO2 and lung cancer risk resulted in pooled relative risk estimates of 1.07 (95% CI: 1.03-1.10) per 10 μg/m3 for PM2.5, and 1.05 (95% CI: 1.02-1.09) per 10 ppb for NO2. Calculated PAFs for PM2.5 and NO2 were 5.4% and 7.9% respectively. Analysis of trends overtime suggested levels have decreased by 0.27 per 10 μg/m3 for PM2.5 and 0.56 per 10 ppb for N02 per year from 1980-2014. In light of observed trends, a 50% reduction in PM2.5 and NO2 is considered feasible. Counterfactuals reducing exposure by 50% are associated with a PIF of 5.6% for PM2.5, and 4.0% for NO2. Conclusion: These results quantify the proportion of incident cancer cases in Canada that could be prevented through changes in modifiable environmental exposures. Planned analysis will calculate PIFs based on projected exposure scenarios.