As a part of a larger project (DEMO RCT, Krogh et al., 2007) which was carried out at the Copenhagen University Hospital in Denmark, The current study intended to establish longitudinally the extent to which an integrated model of hierarchal model (Fox and Corbin,1989) and Physical Self-Perception sub-scales to depression model (Van de Vliet et al., 2002b) was tenable. Thus, the objectives were, (1) to investigate the associations of the five sub-domains of PSP with global self-esteem, depression and anxiety longitudinally across the three waves of time, (2) to test the effects of exercise on the changes of physical self-perceptions (PSPs) among clinical depressed samples over time, (3) to examine the contribution of self-esteem and PSW as mediators between physical self-perceptions and negative affect relationships among depressed patients over a 12-month period, (4) to determine whether a causal relationship exists between exercise and changes in self- perceptions, self-esteem and depression. Results demonstrated that most of the scales represent acceptable to excellent values of internal consistency and the reliability of scales employed (Cronbach alpha = 0.92 to 0.78 at all three time points). We found that correlations between Fitness (VO2max) and PSPP subscales were significant at all three time periods. The same associations, but inversely, were found for fitness and depression. PSPs influence on BDI and HAMA was mediated longitudinally by RSES and PSW across the three waves of time. Also the results accentuated the strong association between the level of physical self-perceptions, self-esteem, and the degree of mood disorder symptoms. Applying partial correlations, the impact of PSW and RSES were attenuated. Only the relationship between body attractiveness and depression was not extinguished at baseline and even in follow-up assessment. Exercise showed to influence mood through self-esteem and its perceived physical competences in a vertical continuum of specificity- generality, as the paths hypothesized by the models 3-6 (time2 and 3)were significant. This influence was also evident through longitudinal trajectories at model 7 through which the skill-development hypothesis was corroborated. The self-enhancement hypothesis also was supported through the top-down model-8 on depressed patients. Moreover, most direct and indirect hypothesized associations between depression, self-esteem, PSPs and exercise behavior were found to be significant, especially the descending effects from time 2 (the post intervention) to time 3 were more vigorous. All of the hypothesized relationships in the longitudinal data were substantiated and the overall model fitting results of the path analysis provided strong empirical support for both proposed theoretical models (i.e. skill development and the self-enhancement hypotheses). By showing strong direct associations with a high magnitude of coefficients in PAs with a theoretical basis, this study indicates that there is causal and mediating relationship. Moreover, due to the longitudinal nature of this study and because this study examined mediators at multiple time points (i.e., short and long-term time points), thus causal inferences can be made. These findings extend the previously established theories and models related to global self-esteem to the clinical population.

Keywords