How Well Do Commonly Used Co-Contraction Indices Approximate Lower Limb Joint Stiffness Trends during Gait?

Muscle co-contraction generates joint stiffness to improve stability and accuracy during limb movement but at the expense of higher energetic cost. The quantification of joint stiffness generated from muscle co-contraction is difficult through both experimental and computational means for its benefit and cost to be assessed. Quantification of muscle co-contraction may offer an alternative path for estimating joint stiffness. By choosing the commonly used Co-Contraction Indices (CCIs) to represent muscle co-contraction, this study investigated the feasibility of using CCI to approximate lower limb joint stiffness trends during gait. A calibrated EMG-driven musculoskeletal model of a hemiparetic individual post-stroke from a previous study was used to generate the quantities required for CCI calculation and model-based estimation of joint stiffness. A total of 14 classes of CCIs for various combinations of antagonistic muscle pairs were calculated based on two common CCI formulations, each with 7 types of quantities that included variations of electromyography (EMG) signals and joint moments from the muscles. Correlations between CCIs and model-based estimates of sagittal plane stiffness of the lower extremity joints (hip, knee, ankle) were computed. Although moderate to strong correlation was observed between some CCI formulations and the corresponding joint stiffness, these associations were highly dependent on the methodological choices made for CCI computation. The overall findings of this study were the following: (1) the formulation proposed by Rudolph et al. (2000), CCI1, was more correlated with joint stiffness than that of Falconer and Winter (1985); (2) Moment-based CCI1 from individual antagonistic muscle pairs was more correlated than EMG-based CCI1; (3) EMG signals with calibrated electromechanical delay and joint moment generated by individual muscle without normalization to a reference value were the most correlated for EMG-based CCI1 and moment-based CCI1, respectively. The combination of antagonistic muscle pairs for most correlated within each CCI class was also identified. By using CCI to approximate joint stiffness trends, this study may open an alternative path to studying joint stiffness.