PURPOSE:​ The purpose of this investigation was to compare the gait kinematics and kinetics of older adults with radiographic unilateral knee osteoarthritis and those with radiographic bilateral knee osteoarthritis with accompanying pain in one or both knees. We hypothesized that the gait of those with unilateral disease would be significantly different than those with bilateral disease, and therefore, should be analyzed independent of those with bilateral disease in future research studies.

METHODS:​ Baseline data from a subset of participants (136 out of 454) enrolled in the Intensive Diet and Exercise for Arthritis (IDEA) study were used for this study. 68 participants were identified as having unilateral radiographic OA and were matched to 68 participants with bilateral OA based on age, gender, BMI, race, walking speed, WOMAC pain score, and WOMAC function score. Internal joint moments and knee joint reaction forces were calculated using a 3-D gait analysis with a 37-reflective marker set arranged in a Cleveland Clinic full-body configuration, a 6-Camera Motion Analysis System set to sample data at 60 Hz, and a torque-driven muscluloskeletal model (DeVita and Hortobagyi). Means for peak values of percent of the gait cycle spent in the stance phase, joint ROM at the hip, knee, and ankle, frontal plane joint moments at the hip, knee, and ankle, sagittal plane moments at the knee, vertical GRF, bone-on-bone forces at the knee, and power generated at the hip, knee and ankle were calculated for each individual. Comparisons were made between the most affected leg in the bilateral cohort (self-selected by the participant) to the affected leg in the unilateral population, and between the least affected leg in the bilateral cohort to the healthy leg of the unilateral cohort. Peak mean values were then used to calculate symmetry between limbs in each individual using the symmetry index (SI), and mean values for each group were compared.

RESULTS:​ After controlling for age, gender, K/L grade, BMI, presence of hip OA, and WOMAC pain score, significant differences were found in between groups in hip flexion moment (p=0.01) and the ankle plantar flexion moment (p=0.04) in the comparison of the healthy limb in the unilateral cohort to the least affected limb of the bilateral cohort, and in the ankle plantar flexion moment (p=0.04) in the comparison of the most affected limb in both cohorts. Significant asymmetry (greater than 10% difference between limbs) was seen in knee power in the unilateral cohort (SI = 18%, p=0.03). Significant differences were seen in SI between unilateral and bilateral groups in hip flexion moment (p=0.03) and hip extension moment (p=0.01).

CONCLUSIONS:​ Distinctions are not always made in the classification of unilateral or bilateral disease in those with osteoarthritic knee joints in statistical analysis. While it is known that OA gait differs from the gait of healthy, asymptomatic controls, it cannot be assumed that the gait of painful unilateral knee OA and painful bilateral knee OA patients are similar. Our results indicate few differences when comparing the gait mechanics of those with unilateral OA to those with bilateral OA with accompanying mild to moderate pain. Overall, only 6 variables were found to be significantly different in 59 comparisons. This amount of variance may be expected due to chance alone. Due to this lack of statistical differences between groups in multiple comparisons, unilateral and bilateral OA can be analyzed collectively as one OA group, and that results discovered in one population may be generalized to both unilateral and bilateral OA.