Abstract
OBJECTIVE: To explore whether pain beliefs and functional strength mediate the treatment effect of manual therapy (MT) and exercise therapy (ET) on the Western Ontario and McMaster Osteoarthritis Index (WOMAC) composite scores and its subscales in individuals with hip and/or knee osteoarthritis in the MOA trial. DESIGN: Secondary analysis of a randomised controlled trial that compared the incremental effects of supervised MT and ET in addition to usual care in patients with osteoarthritis of the hip or knee. 206 participants enrolled in the MOA trial were analysed. The primary outcome measure was the WOMAC composite score after 1 year. RESULTS: Pain belief mediated the effect of MT (b: -10.7, 95 % CI: -22.3, -0.9), ET (b: -14.5 95%CI: -26.0, -4.4). Functional strength did not mediate the effect of MT, ET, or MT + ET. Mediation sensitivity analyses suggest findings are likely to change if small confounding between those mediators and WOMAC composite score is present. CONCLUSIONS: We identified possible mediators of MT and ET. Future confirmatory studies could be designed to assess the mechanisms through which manual therapy and exercise cause improvements in pain and function scores in patients with hip or knee osteoarthritis.