Increasing muscle co-contraction speeds up internal model acquisition during dynamic motor learning

Sci Rep. 2018 Nov 5;8(1):16355. doi: 10.1038/s41598-018-34737-5.

Abstract

During reaching movements in the presence of novel dynamics, participants initially co-contract their muscles to reduce kinematic errors and improve task performance. As learning proceeds, muscle co-contraction decreases as an accurate internal model develops. The initial co-contraction could affect the learning of the internal model in several ways. By ensuring the limb remains close to the target state, co-contraction could speed up learning. Conversely, by reducing kinematic errors, a key training signal, it could slow down learning. Alternatively, given that the effects of muscle co-contraction on kinematic errors are predictable and could be discounted when assessing the internal model error, it could have no effect on learning. Using a sequence of force pulses, we pretrained two groups to either co-contract (stiff group) or relax (relaxed group) their arm muscles in the presence of dynamic perturbations. A third group (control group) was not pretrained. All groups performed reaching movements in a velocity-dependent curl field. We measured adaptation using channel trials and found greater adaptation in the stiff group during early learning. We also found a positive correlation between muscle co-contraction, as measured by surface electromyography, and adaptation. These results show that muscle co-contraction accelerates the rate of dynamic motor learning.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Electromyography
  • Female
  • Humans
  • Learning*
  • Male
  • Models, Neurological*
  • Motor Activity / physiology*
  • Muscle Contraction / physiology*
  • Psychomotor Performance