Health professionals and urban planners are increasingly calling for new approaches that involve changes in the built environment to address complex health and environmental problems. In particular, community designs to promote walking and cycling are seen as potential solutions to the obesity epidemic in the U.S. Yet, the net health effect that results from neighborhood transformations is not known today. Competing risks may be involved, particularly when considering the effects of encouraging people to be active in areas with significant air pollution and fraught with risks of traffic injuries.

This dissertation proposes a conceptual framework for assessing risks and benefits that ensue from the improvement of the pedestrian environment, and investigates some of these relationships in a quantitative application.

The probabilistic model developed for this work consists in simulating the movement of individuals in a case-study area that undergoes hypothetical changes in land use and street network. Resulting changes in energy expenditure due to active travel and in pollutant inhalation dose are estimated. The model uses an activity database, travel models from the transportation literature, and ozone and PM₁₀ fields developed for this work using the Bayesian Maximum Entropy framework and a combination of monitored and modeled data. Daily individual inhalation intake is thus calculated accounting for specific activities, locations, and times of day. Uncertainty and population variability is analyzed through Monte Carlo simulation.

Results show great uncertainty associated with estimating risks and benefits. For example, two travel models yield a four-fold difference in predicting the fraction of population with significant increases in PM₁₀ inhalation dose. Conservative estimates demonstrate a significant increase in the fraction of days above a PM₁₀ threshold across the population, and potential for some individuals to more than double their inhalation intake of both pollutants on certain days. Clear benefits in terms of physical activity, however, cannot be established by the conservative exposure model.

This work is an innovative risk assessment method for analyzing health impacts of built environment policies. The dissertation concludes with suggested policies to address increased risks, and a research agenda for future work in this area.