Change isn't easy: Identifying motivational profiles during physical activity behavior change in obese patients

To facilitate sustained physical activity (PA) in obese patients it is necessary to consider their motivational profiles, especially during the challenging initial phase of behavior change. Based on organismic integration theory (Deci & Ryan, 2000), this study intended to characterize PA motivational profiles of obese patients, who changed their PA behavior following a 3-week rehabilitation program. 237 obese patients (33.6% female; age M = 46.43, SD = 9.88; BMI M = 39.03, SD = 8.55) provided behavioral regulation data 6 months post-rehabilitation. From pre- to 6-months post-rehabilitation, 50.2 % did not change their insufficient PA behavior (i.e. Godin-Shepard-Score < 24; Godin, 2011; inactive-group); 18.6 % remained sufficiently active (active-group); and 31.2 % changed from inactive to active (PA-change-group). Motivational profiles were generated by usage K-means non-hierarchical cluster analyses with scores on behavioral regulations as criterion variables at 6 months. Differences in motivational profiles between inactive, active, and PA-change-group were analyzed by chi-square tests. Three motivational clusters were identified: high autonomous cluster (autonomous high, controlled regulation low), low-autonomous cluster (autonomous low, controlled regulation low) and high controlled cluster (autonomous low, controlled regulations high). Distributions of these clusters were significantly different across the PA groups (X2(4, 237) = 44.20, p<.001): 86% of patients in the active-group belonged to the high autonomous cluster, 58% in the change-group and 31% in the inactive-group. Patients in the PA-change-group endorsed relatively high levels of controlled regulations, were not as autonomously motivated as the active-group, but were still more favorably motivated than the inactive-group. Future research is warranted to analyze if controlled regulations in exercise beginners have a more functional, positive effect on behavior change or if they inhibit the change process. Additionally, the clusters identified provide crucial information regarding PA intervention planning, as different intervention approaches may complement different patient profiles and, therefore, impact on PA engagement.