Objective
Depressive symptoms in knee osteoarthritis (OA) are associated with increased pain severity and declines in physical performance. This study examined whether pain severity mediates the association between depressive symptoms and physical performance in persons with radiographic knee OA.
Method
Three years of annual data from participants (n = 1,463) with radiographic knee OA in the Osteoarthritis Initiative (OAI) were analyzed. Depressive symptoms were measured using the Center for Epidemiological Studies Depression (CES-D) scale. Pain severity was evaluated with the Western Ontario and McMaster Universities Arthritis Index. Physical performance was assessed via standardized gait speed. Marginal structural models were used to assess the direct (unmediated) effects of depressive symptoms on physical performance and indirect (mediated) effects through pain severity.
Results
Direct and indirect effects for a difference in CES-D score of 0-1 were -0.0051 (95% confidence intervals (CI): -0.0053, -0.0049) and -0.0016 (95% CI: -0.0024, -0.0007) standard deviations in gait speed, respectively. Higher depressive symptom severity exhibited diminishing, incremental, direct and indirect effects and for a difference in CES-D score of 15-16 were -0.0045 (95% CI: -0.0047, -0.0042) and -0.0009 (95% CI: -0.0014, -0.0004) standard deviations in gait speed, respectively. Therefore, the magnitude of the mediated, indirect effect, was never larger than 24%.
Conclusion
Pain severity mediated approximately one-fifth of the association between depressive symptoms and physical performance in persons with radiographic knee OA, and the diminishing incremental effects may explain why unimodal treatment strategies with a single disease target are often ineffective in depressed OA patients.
This publication is part of the RAND external publication series. Many RAND studies are published in peer-reviewed scholarly journals, as chapters in commercial books, or as documents published by other organizations.
RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.