Greenway accessibility and physical activity behavior in two Michigan cities.

Since public health initiatives have made important but modest gains with non-ecological health promotion efforts aimed at increasing physical activity, the previously overlooked built environment is now being considered for its mediating effect in either facilitating or hindering one's choice to be active. The multi-use greenway is an example of an environmental feature which can support physical activity, but the number of persons performing activity on the greenway may be influenced by the characteristics of the areas surrounding it. This locational relationship between the greenway and the land uses surrounding it influences its proximity to residences and, once on the greenway, the opportunities created by a variety of potential destinations. A community-level, unobtrusive, Geographic Information Systems (GIS) and mixed model regression analysis of two urban greenways in Michigan was conducted to address the following question: How does the accessibility of the greenway, as measured by proximity to home and destination opportunities, affect the rate of people using it as a forum for physical activity? Mapping the uses per segment revealed a tendency of persons to use parts of the greenway that have natural qualities and parts that are located in or near the downtown area. The main effects regression models revealed that the land use mixture within a walking distance of the greenway was a significant predictor of uses. The models also revealed that population density within both walking and biking distances were significant predictors of uses, but this relationship was not in the expected positive direction. The explanation for this counterintuitive phenomenon becomes apparent when examining the interaction of land use mixture and population density on the use of greenway segments. The interaction results reveal that increasing population density is negatively correlated with uses in areas of low land use mixture, but positively correlated with uses in areas of high land use mixture. The results allow us to confirm that smaller scales of analysis are better at predicting greenway use, and increasing population density may have a negative effect on use for physical activity if a variety of land uses is not also present.Ph.D.Health and Environmental SciencesKinesiologyPublic healthSocial SciencesUrban planningUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/126128/2/3237936.pd