Background: Conducting process evaluations of health promoting interventions, and measuring the effectiveness of specific intervention components, may help in the understanding of program failure or success. The purposes of the present study were to examine adults' exposure to and involvement in specific components of a three year long pseudo-experimental community-based physical activity intervention, and to examine the relationship between such exposure and participation and changes in stages of change in physical activity and psychosocial mediators.
Methods: 1497 persons in the intervention group attended the baseline survey in 2000 (50.6%) and 1204 (80.4 of baseline attendees) provided data on the outcome variables of the present study. In 2003, 1089 were still living in the area, and were re-invited to follow-up assessments. Current analyses are based on the 603 persons (mean age 49 +/- 10 years) who provided baseline and follow-up data for the current purposes (56.6% follow-up rate). Process data, stages of change in physical activity, and potential psychosocial mediators of change in physical activity were assessed by questionnaires. The theory-based intervention was composed of communication, physical activity, environmental and participatory components. Data were analysed using frequency and descriptive statistics, Chi-square and t-tests, and regression analyses.
Results: Exposure and participation rates in the various intervention components varied greatly (1.5-92.7%). Participation in walking groups and aerobic exercise groups, as well as having seen the "Walk the stairs"-poster were significantly and positively related to change in stages of change in physical activity (beta = .12, p = .011; beta = .211, p < .001; beta = .105, p = .014, respectively). Additionally, having used the walk path was significantly and positively related to change in stages in women (beta = .209, p = .001) but not in men (beta = -.011, p = .879), and in Western people (beta = .149, p = .003) but not in non-Westerners (beta = -.293, p = .092). Observed significant relations were partly mediated by positive changes in psychosocial factors as social support from friends, perceived control, and physical activity identity.
Conclusion: Findings revealed that particular intervention components, such as participation in physical activity groups, were more strongly related to forward transition in stages of change in physical activity than others. These findings together with results indicating that such transitions were mediated by specific psychosocial influences may improve theory and help to prioritize among specific intervention components in future programs.