Mechanical Hand Synergies during Dynamic Hand Movements are Mostly Controlled in a Non-Synergistic Way by Spinal Motor Neurons

Precise control of spinal motor neurons is crucial for voluntary hand and digit movements. However, the specific mechanisms by which motor unit ensembles govern dynamic synergistic and individual digit tasks remain poorly understood. We recorded synchronized 3D hand kinematics and high-density surface EMG (HD-sEMG) data from extrinsic hand muscles of twelve participants during 13 dynamic hand and digit movement tasks, consisting of single-digit flexion/extension and mechanically synergistic grasping tasks. We extracted single motor unit (MU) activity and identified identical MUs across tasks. We extracted 7.8 ± 1.8 MUs per task and participant and found 182 out of 554 total MUs active during multiple movements. Analysis of the MU discharge patterns revealed two groups of motor units that were categorized into prime mover MUs , showing strong correlation between firing rate modulation and digit kinematics, and postural MUs with little modulated activity. We found these motor units could switch between the two modes, showing either postural or movement encoding activation depending on the task. However, MUs acted as prime mover only for one specific digit. We further observed highly task specific recruitment of prime mover MUs. Across participants, we found only 9 ± 8.2 % of prime mover MUs active during a grasp task and any single digit task involved in the grasp motion. We draw three conclusions: (1) Single digits are controlled by distinct groups of MUs. (2) Unexpectedly, mechanically synergistic grasp movements are mostly controlled in a non-synergistic way by distinct groups of MUs. 3) Multiple manifolds construct the movement of the human hand, and each motor unit can flexibly switch between postural and dynamic modes. Significance Statement We investigated the neural control of motor unit ensembles during single-digit and synergistic grasping tasks in dynamic conditions. We found that motor units exhibited strong movement-correlated activity only for one specific digit. We further observed highly task specific recruitment of motor units during mechanically synergistic grasp movements, showing that on a motor unit level, mechanically synergistic movements are controlled in a non-synergistic way. The findings extend the knowledge of motor unit recruitment strategies in natural movements and have strong implications in the field of neurorehabilitation and control of assistive devices.