The interaction between the muscle fascicle and tendon components of the human soleus (SO) muscle influences the capacity of the muscle to generate force and mechanical work during walking and running. In the present study, ultrasound-based measurements of in vivo SO muscle fascicle behavior were combined with an inverse dynamics analysis to investigate the interaction between the muscle fascicle and tendon components over a broad range of steady-state walking and running speeds: slow-paced walking (0.7 m/s) through to moderate-paced running (5.0 m/s). Irrespective of a change in locomotion mode (i.e., walking vs. running) or an increase in steady-state speed, SO muscle fascicles were found to exhibit minimal shortening compared with the muscle-tendon unit (MTU) throughout stance. During walking and running, the muscle fascicles contributed only 35 and 20% of the overall MTU length change and shortening velocity, respectively. Greater levels of muscle activity resulted in increasingly shorter SO muscle fascicles as locomotion speed increased, both of which facilitated greater tendon stretch and recoil. Thus the elastic tendon contributed the majority of the MTU length change during walking and running. When transitioning from walking to running near the preferred transition speed (2.0 m/s), greater, more economical ankle torque development is likely explained by the SO muscle fascicles shortening more slowly and operating on a more favorable portion (i.e., closer to the plateau) of the force-length curve.

The human ankle plantar-flexors, soleus (SOL) and gastrocnemius (GAS), utilize tendon elastic strain energy to reduce muscle fiber work and optimize contractile conditions during running. However, studies to date have considered only slow to moderate running speeds up to 5 m/s. Little is known about how the human ankle plantar-flexors utilize tendon elastic strain energy as running speed is advanced towards maximum sprinting. We used data obtained from gait experiments in conjunction with musculoskeletal modeling and optimization techniques to calculate muscle-tendon unit (MTU) work, tendon elastic strain energy and muscle fiber work for the ankle plantar-flexors as participants ran at five discrete steady-state speeds ranging from jogging (~2 m/s) to sprinting (≥8 m/s). As running speed progressed from jogging to sprinting, the contribution of tendon elastic strain energy to the positive work generated by the MTU increased from 53% to 74% for SOL and from 62% to 75% for GAS. This increase was facilitated by greater muscle activation and the relatively isometric behavior of the SOL and GAS muscle fibers. Both of these characteristics enhanced tendon stretch and recoil, which contributed to the bulk of the change in MTU length. Our results suggest that as steady-state running speed is advanced towards maximum sprinting, the human ankle plantar-flexors continue to prioritize the storage and recovery of tendon elastic strain energy over muscle fiber work.

The aim of this study was to compare muscle-force estimates derived for human locomotion using three different methods commonly reported in the literature: static optimisation (SO), computed muscle control (CMC) and neuromusculoskeletal tracking (NMT). In contrast with SO, CMC and NMT calculate muscle forces dynamically by including muscle activation dynamics. Furthermore, NMT utilises a time-dependent performance criterion, wherein a single optimisation problem is solved over the entire time interval of the task. Each of these methods was used in conjunction with musculoskeletal modelling and experimental gait data to determine lower-limb muscle forces for self-selected speeds of walking and running. Correlation analyses were performed for each muscle to quantify differences between the various muscle-force solutions. The patterns of muscle loading predicted by the three methods were similar for both walking and running. The correlation coefficient between any two sets of muscle-force solutions ranged from 0.46 to 0.99 (p < 0.001 for all muscles). These results suggest that the robustness and efficiency of static optimisation make it the most attractive method for estimating muscle forces in human locomotion.