Pain and functional trajectories in symptomatic knee osteoarthritis over up to 12 weeks of exercise exposure

Summary Objective Exercise is the recommended treatment for knee osteoarthritis (OA). However, heterogeneous patterns in treatment response are poorly understood. Our purpose was to identify pain and functional trajectories from exercise interventions in knee OA, and to determine their association with baseline factors. Methods Prospective cohort of 171 participants (mean age 61 years; BMI 32 kg/m 2 , 71% female; 57% white) with symptomatic knee OA from a randomized trial comparing 12-week Tai Chi and Physical Therapy. We analyzed weekly Western Ontario and McMaster Osteoarthritis Index (WOMAC) pain (0–500) and function (0–1700) scores using group-based trajectory models. Associations between baseline factors and trajectories were examined using multinomial logistic regression. Results We identified four pain trajectories: Lower-Early Improvement (43%), Moderate-Early Improvement (32%), Higher-Delayed Improvement (15%), and Higher-No Improvement (10%). We found similar trajectories for function, except that the lower function trajectories diverged into gradual (12%) or delayed-improvement (15%). Compared with the Lower-Early Improvement pain trajectory, moderate and higher trajectories were associated with poorer physical and psychosocial health. A similar pattern of associations were found among the function trajectories. Conclusions We found four distinct trajectories for pain and function over up to 12-weeks of exercise interventions. While most participants experienced improvements over a short-term exposure, subgroups with greater baseline pain/physical disability had either gradual, delayed, or no improvements. These findings help disentangle the heterogeneity of treatment response and may advance patient-centered care in knee OA.