Background: We examined associations of individual, psychosocial and environmental characteristics with 4-year changes in walking among middle-to-older aged adults; few such studies have employed prospective designs.
Methods: Walking for transport and walking for recreation were assessed during 2003-2004 (baseline) and 2007-2008 (follow-up) among 445 adults aged 50-65 years residing in Adelaide, Australia. Logistic regression analyses examined predictors of being in the highest quintile of decline in walking (21.4 minutes/day or more reduction in walking for transport; 18.6 minutes/day or more reduction in walking for recreation).
Results: Declines in walking for transport were related to higher level of walking at baseline, low perceived benefits of activity, low family social support, a medium level of social interaction, low sense of community, and higher neighborhood walkability. Declines in walking for recreation were related to higher level of walking at baseline, low self-efficacy for activity, low family social support, and a medium level of available walking facilities.
Conclusions: Declines in middle-to-older aged adults' walking for transport and walking for recreation have differing personal, psychosocial and built-environment correlates, for which particular preventive strategies may be developed. Targeted campaigns, community-based programs, and environmental and policy initiatives can be informed by these findings.