performed to detect limb by group differences. Significance level was set at p 0.05. Descriptive comparisons were made to historic pre-operative and healthy controls. Results: Ninety-three subjects were included in this study. Sixteen subjects were lost to follow-up, resulting in 77 subjects included in the final analysis (YES: N ¼ 18, 6M/12F, age ¼ 67.11; NO: N ¼ 59, 31M/28F, age ¼ 66.80). There were significant main effects of group and limb (Table 1), with the YES group having substantially weaker knee extensors on both limbs (Fig 1). There were no significant predictors of contralateral TKA for either limb. However, the addition of operated limb quadriceps strength to the model improved the significance of the model and was nearly a significant addition to the model (p ¼ 0.077). Conclusions: Given that the subjects who had contralateral TKA were substantially weaker than those who did not, and were weaker than historical group undergoing TKA, these results suggests that quadriceps weakness may play a role in discriminating those who do and do not demonstrate symptomatic progression on the contralateral limb. Future work should evaluate rehabilitation protocols that not only restore operated limb quadriceps strength to at least pre-operative levels. Although knee adduction moment is predictive of OA progression, we did not find that in this analysis. However, previous literature suggests that non-normalized PKAM may be a more clinically relevant measure when analyzing knee OA progression. Therefore, future analyses of incidence of contralateral TKA may include non-normalized PKAM. Purpose: Knee joint load is an important factor associated with progression of knee osteoarthritis. The knee adduction moment (KAM) is an indicator of medio-lateral knee load distribution. However, KAM only includes frontal plane moment and has recently been found insensitive in subjects with mild to moderate osteoarthritis. The Knee Index has been developed to include moments from all three planes (frontal, sagittal and transversal) and was able to distinguish between pain relief induced by pla-cebo, NSAID or opioids. However, to help interpret the underlying biomechanical characteristics of the Knee Index, the respective contributions of the knee moments derived from the three planes are important to determine. The purpose of this study was therefor to investigate how the frontal, sagittal and transversal moments contribute to the Knee Index, a novel biomechanical index of joint load for the knee, in patients with mild to moderate knee osteoarthritis. Methods: The contribution of frontal, sagittal and transversal plane knee moments to the Knee Index was investigated in 24 subjects (13 women, age: 58 AE 7.6 years, BMI: 27.1 AE 3.0) with clinically diag-nosed mild to moderate knee osteoarthritis according to the ACR criteria. Three dimensional gait analysis was performed using a 6-camera Vicon MX (Vicon, Oxford, UK) movement analysis system (100 Hz) with the Plug-in-Gait marker set. Ground reaction forces were recorded (1000 Hz) by two AMTI force-plates (AMTI, OR6-7, Watertown, MA, USA) embedded at floor level. Subjects walked barefoot at self-selected walking speed. The trial (out of 5 trials) representing the median velocity was selected for further analysis. The first peak (approximately 50 % of stance phase) magnitude Knee Index (calculated by the root mean square of frontal, sagittal and transversal knee moments (for equation, see figure 1A) and the corresponding knee moments (at the same time points) from all three planes were calculated for the knee diagnosed with OA using inverse dynamics. Percentage distribution of the contributors of the Knee Index (for equation, see figure 1B). Results: Frontal plane kinematics contributed with 60.0% (SD 25.6) of the Knee Index while sagittal plane kinematics contributed with 40.5% (SD 26.1) and transversal plane kinematics contributed with 0.2% (SD 0.3). A substantial inter-subject variation in the relative contribution of the flexion and extension moment components to the Knee Index was observed (see figure 2). Conclusions: Our findings in these subjects with mild to moderate knee OA support the notion that the primary contributor to the Knee Index is the frontal plane kinematics (i.e. the knee adduction moment), and secondarily the sagittal plane kinematics (i.e. the knee flexion moment). The transversal plane moment did not contribute to the Knee Index. It is hypothesized that the Knee Index's sensitivity to pain comes from the inclusion of the sagittal plane. The present substantial inter-subject variation gives interest to inves-tigate the relative contributions as predictive of future clinical changes. The present findings add to the knowledge of knee joint load dis-tribution and OA.
Clausen, B., Andriacchi, T. P., Nielsen, D. B., Roos, E. M., & Holsgaard-Larsen, A. (2014). Composition of the knee index, a three-dimensional biomechanical index for knee joint load, in subjects with mild to moderate knee osteoarthritis. Osteoarthritis and Cartilage, 22, S95–S96. https://doi.org/10.1016/j.joca.2014.02.183