Project 459113

In 2015, exposure to ambient fine particles in the air was responsible for 4.2 million deaths globally. Emitted mainly from motor vehicles, ultrafine particles (UFPs), having a smaller diameter, may have a higher potential to induce deleterious health effects than larger particles as they penetrate deeper into the respiratory tract and can reach the bloodstream. However, knowledge about populations' exposure to UFPs, in Canada and worldwide, and their cardiorespiratory effects remains limited. In fact, estimating population exposure is challenging. UFPs are rarely measured at government monitoring stations. Moreover, their concentrations vary widely spatially, temporally and with climate, thus models to estimate UFP concentrations are scarce and their transferability between cities is lacking. Specific objectives of my thesis include, firstly, to develop and compare the performance and transferability of various models using statistical learning methods in estimating urban UFP concentrations, as opposed to traditional methods. Secondly, estimated UFP concentrations will be linked with health data from population cohorts created from medico-administrative databases to estimate associations with childhood asthma onset and cardiovascular mortality. Finally, I will use the previously estimated associations to compute marginal effects (i.e., number of cases that would be prevented by eliminating UFPs exposure). My research will provide essential information on the intra-urban variability of UFPs in Canada. In addition, it will provide more accurate tools for modelling UFPs exposure, thereby improving risk estimates from epidemiological studies. Results of my epidemiological analyses will also contribute to the evidence linking UFPs to cardiorespiratory effects. Ultimately, these results will be communicated to public health stakeholders in Quebec and Canada, providing input to update risk assessment documents and air pollution standards to better protect public health.