BACKGROUND
The RESIDential Environment project (RESIDE) commenced in 2003, and is a prospective study designed to evaluate the impact of a new state government sub-division design code on the walking and cycling behavior or local residents. Trial of the Livable Neighbourhood (LN) Guidelines commenced in 1997, providing a unique opportunity to undertake a natural experiment. Based on new urbanism principles, the guidelines incorporate six design elements, four of which are relevant to walking and cycling (i.e., community design, movement network, lot layout, public parkland). To accommodate the longitudinal research design and the need to evaluate an urban design policy, the Neighbourhood Physical Activity Questionnaire (NPAQ) was developed to contextualise and differentiate between transport and recreational walking undertaken within and outside the neighbourhood.
OBJECTIVES
The objectives of this paper are: (1) to outline the RESIDE study design and provide a rationale for the development of the NPAQ; and (2) to compare multivariate models that use self-reported environmental correlates to predict step-counts, overall self-reported walking, and more specific models of recreational and transport-related walking undertaken within the neighbourhood.
METHODS
1813 people building homes in 74 new housing estates (n=538 in 18 LN Estates (LNE), n=358 in 11 Hybrid Estates (HE), n=917 in 45 Conventional Estates (CE)) agreed to participate (33% response rate) in three surveys over four years: (1) baseline - before study participants moved into their new home; (2) 1st follow-up - 12 months after moving; and (3) 2nd follow-up - two years later. At baseline the participants completed a 31 page self-completed questionnaire based on an ecological model that includes the NPAQ; five scales from the Neighbourhood Environment Walkability Scale (NEWS); measures of other interpersonal and intrapersonal correlates of physical activity, and demographic characteristics. In addition, study participants wore a pedometer for 7 days and completed an activity diary. These measures will be repeated in follow-up surveys, and although not reported in this paper, objective measures of the environment are being developed using Geographic Informational Systems. Four dependent variables are examined (any walking, achieving 10,000 steps/day and walking within the neighbourhood for recreation and for transport (No=0, Yes=1)) and five environmental correlate scales (the number of retail and recreational destinations within 15 minutes from home and perceived access to services, neighbourhood surroundings, presence of footpaths and separation from traffic; and safety) adjusted for age, gender, education and type of estate.
RESULTS
Overall, at baseline, 68.9% of RESIDE study participants reported doing some walking in a usual week and 27.2% reported achieving 10,000 or more steps/day. About 30% of study participants reported usually walking outside their neighbourhood, and within the neighbourhood, 52.6% reported usually walking for recreation and 36.1% walking for transport. On average, RESIDE study participants reported walking for 128 minutes/week overall (SD 159), however, within the neighbourhood they walked for recreation 67 minutes/week and for transport 26 minutes/week. The strength of the associations between environmental correlates and outcome variables varied. Stronger associations were evident when there was a greater correspondence between the outcome variable and the environmental correlate of interest. For example, the odds of usually doing any walking, achieving 10,000 steps and walking for recreation within the neighbourhood increased by 4-5% for every additional retail destination within 15 minutes of the respondent’s home. However, for walking for transport within the neighbourhood, the odds increased by 13% for each additional transport-related retail destination within 15 minutes of home (95% CI 1.09-1.17). Similarly, perceived access to services was significantly associated with walking for transport within the neighbourhood (OR 1.44; 95% CI 1.18-1.77), but not any other outcome measure. The odds of reporting any walking and both walking for recreation and walking for transport were higher for study participants who had more positive perceptions of their neighbourhood surroundings (OR ranging from 1.32-1.47), however perceived access to recreational destinations within 15 minutes of home was only associated with walking for recreation (OR 1.11; 95% CI 1.01-1.22). Similarly, perceptions of neighbourhood safety was only significantly associated with walking for recreation within the neighbourhood, and not walking for transport.
CONCLUSIONS
RESIDE is unique because it seeks to evaluate the impact of the implementation of a state-government sub-division design code on the walking behaviors of local residents. Measuring walking that takes place within the neighbourhood is critical. In this study, 30% of reported walking, took place outside the neighbourhood which would weaken the relationship between neighbourhood-level variables and measures of overall walking. Moreover, these preliminary findings confirm the need for future studies to use behaviour-specific rather than generic environmental correlates.
