Transport-related physical activity, health outcomes, and urban design: descriptive evidence
Environments that support physical activity (PA) engagement are now seen as an important part of the solution for accumulating daily physical activity which confers health benefits, particularly transport-related PA (TPA). Despite this recognition, understanding the determinants of health-related PA has traditionally focused on identifying associations between the social environment and non-specific PA domains. As such, prior to commencing this thesis little was known about the relationships between the built environment and overall PA engagement within the New Zealand context. Little was also known about associations between TPA attitudes and behaviours, and urban design. This research sought to overcome this paucity of knowledge by investigating associations between overall PA and TPA engagement and perceptions, health outcomes, urban design, and socio-demographic variables from a public health perspective. The body of research was initiated with two literature reviews that formed the context for the following six empirical chapters. Data in the empirical chapters were drawn from three self-report surveys using adult samples. The research commenced by examining associations between urban design and overall PA engagement, and then narrowed towards TPA-specific relationships. TPA relationships were identified through the development and implementation of a reliable survey that captured associated behaviours and attitudes. Several new findings were drawn from this body of work. Overall, 21% of adults recognised they could replace automobile journeys on at least two days per week, and walking was deemed an acceptable travel alternative. Adult TPA engagement levels were low; 7% commuted to place of work/study and 32% travelled to the convenience shop by TPA modes. Furthermore, when compared with those who commuted to an occupation by TPA, respondents who used motorised travel were less likely to be classified as active (odds ratio (OR)=0.5) and be of normal body mass index (OR=0.5). No significant health relationships existed for convenience shop travel. As well, 4% of adults reported no automobile availability. When this group was compared with those with unrestricted automobile availability, they were less likely to be classified as active (OR=0.3), but were more likely to engage in TPA modes to access destinations (occupation, OR=6.3; convenience shop, OR=9.8). Occupation-related commute distances also revealed interesting findings. Overall, 50% of respondents perceived they could, and 10% of the sample actually did, commute by TPA modes for distances less than five kilometres, and relationships were strongly mediated by distance. Other urban design variables were objectively assessed with TPA engagement for occupation-related commute distances less than five kilometres. Those who travelled along the most connected street networks were more likely to engage in TPA modes (OR=6.9) when compared with respondents travelling along the least connected networks. No relationships were found with other urban design variables and TPA engagement.This research substantially contributes to this research area by identifying associations between overall PA and TPA engagement and perceptions with urban design. It is likely that shifting the perceptions of adults who recognise they can employ TPA modes, into actual TPA engagement will result in promising population health gains. Future PA initiatives with adults should consider promoting walking for transport as a sustainable solution. A lack of TPA interventions is evident; however, methodological issues need to be resolved before developing such initiatives. Although strategies aimed towards changing perceptions may be valuable for future TPA interventions, it is likely macro-scale urban design modifications (e.g., improved job-housing balance, highly-grained street networks) and legislation (e.g., automobile restrictions) will have the greatest success for increasing TPA engagement in the adult population.