How daily activities and built environment affect health? A latent segmentation-based random parameter logit modeling approach

Abstract

Transport studies have recently started focusing on the link between activity participation and general health. This study explores the relationship between general health and travel activities, physical activity, and built environment attributes. Individuals’ self-rated health is considered a general health indicator. Using data from the Impacts of Bicycle Infrastructure in Mid-sized Cities (IBIMS) survey of Kelowna, British Columbia, Canada, this study develops a latent segmentation-based random parameter logit (LSRPL) model. The motivation for adopting the LSRPL model is to capture multi-dimensional heterogeneity by allocating individuals into discrete latent segments (inter-segment heterogeneity) and then allowing a continuous distribution of the parameters within the segments (intra-segment heterogeneity). The model is estimated for two segments where segment one includes younger suburban dwellers; segment two includes older urban dwellers. The model results confirm the effect of travel, physical activity, and built environment on general health revealing multi-dimensional heterogeneity. In the case of inter-segment heterogeneity, results confirm that a higher share of travel duration using a car reduces the likelihood of excellent health in segment one. In contrast, a positive relationship is found in segment two. Whereas, for intra-segment heterogeneity results reveal that, a higher commute duration is found to be negatively associated with good health in segment two. However, this variable shows higher heterogeneity in mean for land use mix index, which indicates that urban areas coupled with higher land use mix increase the propensity towards good health. These findings will assist in developing transportation policies and investment decisions to build healthier cities.