Understanding the effects of a self-directed exercise program on depressive symptoms among adults with arthritis through serial mediation analyses

Abstract The burden of comorbid depression among adults with arthritis is of concern. The purpose of this study is to understand how a self-directed exercise intervention among adults with arthritis exerted its effects on depressive symptoms by performing serial mediation analyses. Methods Participants were randomized to a self-directed exercise program or to self-directed nutrition program (control). Self-reported physical activity, arthritis symptoms (pain, stiffness, fatigue), disability, and depressive symptoms were measured at baseline and 12 weeks. Serial mediation models were conducted to examine the indirect effects of the exercise intervention on changes in depressive symptoms through changes in physical activity and changes in arthritis symptoms/disability (i.e., intervention group→increased physical activity → decreased arthritis symptoms/disability → decreased depressive symptoms). Results A total of 315 participants with complete data were included in these analyses. Results from the serial mediation models found significant indirect effects of the exercise intervention on decreases in depressive symptoms through (1) increases in physical activity and decreases in pain, and (2) increases in physical activity and decreases in disability. There were no significant indirect effects through (1) increases in physical activity and decreases in fatigue, or (2) increases in physical activity and decreases in stiffness. Conclusions This study highlights the importance of physical activity in reducing depressive symptoms, and the means by which it helps (i.e., reducing pain and disability). Understanding these pathways may help researchers design more efficacious interventions.