Abstract
BACKGROUND: Retraining individuals with medial compartment knee osteoarthritis to walk with a patient-specific change in their foot angle (ie, toe-in or toe-out angle) can reduce excessive joint loading related to disease progression. This study investigated the clinical, biomechanical, and structural efficacy of personalised foot progression angle modifications compared with sham treatment in patients with mild-to-moderate medial compartment knee osteoarthritis. METHODS: In this single-center, parallel-group, randomised controlled trial, we recruited individuals with symptomatic medial compartment knee osteoarthritis at the Human Performance Laboratory and Lucas Center for Imaging at Stanford University, CA, USA, using online and print media. Eligible participants (aged ≥18 years) were randomly assigned (1:1) by a computer to an intervention or sham group. During six walking retraining visits to a university gait laboratory, all participants received real-time biofeedback instructing them to walk consistently with a personalised target foot progression angle. The intervention group's target was the 5° or 10° change in foot progression angle that maximally reduced their knee loading, and the sham group's target was their natural foot progression angle. Participants and staff involved in data analysis were masked to group allocation; staff performing the gait analysis visits were not. Primary outcomes were 1-year changes in medial knee pain (numeric rating scale) and medial knee loading (knee adduction moment peak). Secondary outcomes were 1-year changes in cartilage microstructure estimated from MRI (T(1ρ) and T(2) relaxation times). We evaluated safety by monitoring the number and type of adverse events. Intention-to-treat linear regression analyses, comprising all randomly assigned participants, were conducted. People with lived experience of knee osteoarthritis were involved in the design and conduct of this study. This study is registered with ClinicalTrials.gov, NCT02767570, and is closed to enrollment. FINDINGS: Between Aug 1, 2016, and June 25, 2019, 1582 individuals were screened for eligibility. 107 participants completed an initial gait analysis and 68 were randomly assigned to either the intervention (n=34) or the sham (n=34) group. 41 (60%) of 68 participants were female, 27 (40%) were male, and 54 (79%) were White; mean age was 64·4 years (SD 7·6). After 1 year, participants in the intervention group had greater reductions in medial knee pain (between-group difference -1·2, 95% CI -1·9 to -0·5; p=0·0013) and knee adduction moment peak (between-group difference -0·26 % bodyweight × height, 95% CI -0·39 to -0·13; p=0·0001) than participants in the sham group. The MRI-estimated change in cartilage microstructure (T(1ρ)) in the medial compartment was less in the intervention group than the sham group (between-group difference -3·74 ms, 95% CI -6·42 to -1·05). There were no significant between-group differences in T(2). There were no severe adverse events; however, two (6%) of 34 participants in the intervention group and one (3%) of 34 participants in the sham group dropped out of the study due to increased knee pain. INTERPRETATION: Personalised foot angle modifications improve pain, reduce knee loading, and might slow osteoarthritis progression, making them a promising non-surgical treatment option for some individuals with medial compartment knee osteoarthritis. FUNDING: US Department of Veterans Affairs.