Distinguishing between positive and negative affect: A new approach to understanding the antidepressant effect of physical activity

Objectives: physical activity has been found to alleviate depression, but little is known about the process or mechanisms of change. A small number of potential mediators are repeatedly postulated (e.g. increased self-efficacy), but few studies have examined explanations or attempted to identify other theoretical frameworks that may further understanding. Drawing on the tripartite model of anxiety and depression (Clark & Watson, 1991), we suggest a novel explanation and propose that it may be useful to distinguish between the independent mood dimensions of depression, positive affect (PA) and negative affect (NA), as potential mediators. Physical activity may ‘work’ by both decreasing NA and by increasing PA – however, an increase in PA may potentially be more important. Physical activity is more strongly associated with PA than NA, and longitudinal research we conducted found larger effects for improvement in PA than NA in the early stages of increased activity among depressed individuals. Physical activity may therefore particularly tackle the low PA, anhedonia-related symptom of depression. The aim of this study was to use statistical tests of mediation to assess whether PA, NA and previously suggested potential mediators – coping self-efficacy, physical activity self-efficacy and exercise-induced feelings – cross-sectionally mediated the physical activity-depression relationship and the relative importance of each potential mediator. Design: this was a cross-sectional, one-off postal questionnaire with individuals currently experiencing depression or low mood. A model in which all the potential mediators were specified to be direct mediators of the physical activity and depression relationship was assessed in a multiple mediation analysis, controlling for covariates. Methods: participants (N = 164) self-selected to take part based on their subjective self-evaluation of current negative mood. Participants completed the Beck Depression Inventory-II and measures of physical activity, the potential mediators and the covariates. Results: increased physical activity was associated with lower levels of depression. In a product of coefficients with bootstrapping multiple mediation analysis, increased PA and decreased NA were found to be the only significant direct individual mediators and statistical comparisons suggested an equal mediating role. However, post-hoc analyses suggested that increased physical activity self-efficacy may indirectly mediate through improvement in PA but not NA. Conclusions: physical activity self-efficacy may indirectly mediate specifically through improvement in PA but not NA. While PA was not found to be a more important mediator than NA, the findings suggest that it may still be theoretically beneficial to distinguish between these dimensions of depression in understanding the process of change. Implications for future research will be discussed. Although the findings require replication, interventions or physical activity counselling with depressed individuals that target self-efficacy for physical activity may increase the likelihood of PA responses, which in turn, may improve depression outcome