Trajectory analysis combining pain and physical function in individuals with knee and hip osteoarthritis: results from the French KHOALA cohort

Abstract

Abstract Objectives The aims of this study were to identify homogeneous subgroups of knee and/or hip OA patients with distinct trajectories of the combination of pain and physical function (PF) over time and to determine the baseline factors associated with these trajectories. Methods We used data from the Knee and Hip Osteoarthritis Long-term Assessment (KHOALA) cohort, a French population-based cohort of 878 patients with symptomatic knee and/or hip OA. Pain and PF were measured annually over 5 years with the Medical Outcomes Study Short Form 36 questionnaire. First, trajectory models were estimated with varying numbers of groups for each of the outcomes separately then fitted into a multi-trajectory model. We used multinomial logistic regression to determine the baseline characteristics associated with each trajectory. Results Univariate four-class models were identified as most appropriate for pain and PF. Comparison of separate trajectories showed that 41% of patients included in the severe functional limitations trajectory did not belong to the more severe pain trajectory (Cramér’s V statistic = 0.45). Group-based multi-trajectory modelling revealed four distinct trajectories of pain and PF. On multivariate analyses, female sex, older age, high Kellgren grade, low physical activity intensity, low psychosocial distress score (high distress) and low vitality score were associated with the more severe symptoms trajectory. Conclusion Over 5 years, we identified four distinct trajectories combining pain and PF. Management of weight, fatigue and psychosocial distress and the practice of physical activity seem important to maintain function and limit pain in patients with lower-limb OA.