摘要: |
This study seeks to investigate the role that individual (e.g., age, income, race/ethnicity, smoking status, diet, physical activity, health status) factors may play in confounding or modifying the health effects of traffic-related air pollution. The study will also explore aggregating these individual level factors to create socio-economic profiles and indicators of health risk due to traffic-related air pollution along transit routes.
This study will utilize a panel of individuals that participated in the 2015 or 2016 (if available) National Health Interview Survey (NHIS). The research team will complete a proposal to the Dallas-Fort Worth Federal Statistical Research Data Center to gain access to the restricted variables, which include geocodes. The abundance of data within the NHIS will allow the research team to control for individual and household level factors that may also contribute to the incidence of morbidity and mortality. The health outcomes of interest relate to the rate of occurrence of the chronic impacts of traffic-related air pollution, which include the following: (1) Asthma; (2) Lung cancer; (3) Type II diabetes; and (4) Low birth-weight.
Since the study�s primary goal is isolating the impact of traffic-related air pollution, the team will investigate two approaches. Both modeling strategies use the occurrence of a health outcome (see those discussed above) as the dependent variable and the air quality measures and individual health-related variables (e.g. smoking status, diet, physical activity) as independent variables. The first approach estimates a disaggregate logistic regression model for the aforementioned variables. The second approach uses treed regression models, which are combinations of Classification and Regression Tree (CART) models and stepwise logistic regression models to assess adverse health effects of exposure. In both cases, the researchers will characterize the risks related to traffic-related air pollution for different socio-economic profiles along transit routes. |