Abstract
We evaluated the short- and longer-term effects of exercise therapy in hip osteoarthritis patients (OA) at baseline, three, six, and 12 months in a randomized setting, followed by a non-randomized setting. The primary randomized intervention (E = exercise, P = placebo-ultrasound, C = control) was followed by a voluntary three-month exercise therapy for P and C (renamed P-E, C-E). Participants randomized to E were not offered treatment again (E-C). Effect sizes (ES; 95% CI) were calculated for within-group effects across time for bodily pain (SF-36) and WOMAC pain, function, and stiffness. ANCOVAs of post-treatment scores were used for group comparison after the group-specific exercise intervention phase. Exercise adherence was assessed and related to post-treatment scores of clinical outcomes. Data of 115 participants of the RCT eligible for follow-up and completing exercise therapy were included into our analyses. Small to medium beneficial long-term effects of cumulative interventional effects, including exercise training, persisted in all groups. Group E-C (n = 49) showed significant 12 months vs. baseline within-group ES in all outcomes (ES 0.39-0.59) except stiffness. Findings were less prominent for exercise therapy in a non-randomized setting (C-E, P-E, both n = 33). Differences are partially explained by adherence rates, highlighting the relevance of therapy compliance strategies. Short-term between-group differences (ANCOVAs) only showed statistically significant differences for WOMAC function between P-E and E-C in favor of E-C (6.4 (95% CI 1.6-11.2; score range 0-100)).