Psychosocial and environmental factors associated with physical activity among city dwellers in regional Queensland

Background. Research has recently adopted the use of social-ecological models in the study of physical activity. Few studies, however, have addressed the influence of the environment on activity using Geographic Information System (GIS)-derived measures of environmental attributes and self-report ratings of other environmental attributes. Even fewer have examined walking behaviors. Methods. Self-report measures of physical activity, social support, self-efficacy, and perceived neighborhood environment were obtained by means of a Computer-Assisted-Telephone-Interview (CATI) survey of 1,281 residents of Rockhampton, Queensland. Over 94% (1,215) of respondents' residential locations were successfully geocoded into the existing city council GIS database. The self-report data, along with GIS-derived measures, were used to determine the relationships among selected variables of the neighborhood environment for each geocoded location. Results. GIS-derived measures of street connectivity and proximity to parkland, the number of active people in a 1-km radius, and self-reported perceptions of neighborhood cleanliness showed associations with the likelihood of achieving sufficient levels of physical activity when adjusting for selected psychosocial variables. GIS-derived Euclidian distance to footpath networks, number of dogs in 0.8-km radius, network distance to newsagents, and perceptions of footpath condition were significantly associated with the likelihood of participating in any recreational walking. Conclusion. Environmental characteristics were found to have differential influences on the two selected measures of physical activity. Aesthetics and safety appear to be important influences of physical activity, whereas proximal footpaths showed increased likelihood of participation in recreational walking. It is proposed that the strength of association between the environmental and physical activity may be improved if future research utilizes a Geographic Information System approach to the study of restricted geographical areas. (C) 2004 The Institute For Cancer Prevention and Elsevier Inc. All rights reserved.