Learning a novel movement requires functional connectivity changes within primary motor cortex (M1)

Introduction: Recent studies provided strong evidence that training, differential use and alterations in behaviour caused plastic adaptational changes in the functional architecture of the human brain. Here we asked whether learning of a novel synergic movement requires changes in functional connectivity between the central representations of the participating muscle groups inside the primary motor cortex (M1).

Methods: To address this issue we employed a co-contraction paradigm in which 19 participants learned a novel synergic movement consisting of contraction of the abductor pollicis brevis (APB) and the deltoid muscles as simultaneous as possible. A 40 minute training period was monitored using EMG recordings from the two muscles providing instant bio-feedback to the participant with regard to their performance in obtaining synchronous co-contraction. The aim was to reduce the muscle onset asynchrony (MOA) as much as possible after 40 minutes of training. In order to investigate the learning-induced changes not only on a behavioral but also on a cortical level, we performed a fast event related fMRI paradigm before and after the individual training period.

Results: We found a mean drop in MOA after 40 minutes of co-contraction of 16.58±9.11 ms (s.e.m.). This kind of motor learning was accompanied by a shift in center of gravity (CoG) of the corresponding muscle representations (APB and deltoid) as well as changes in functional connectivity between supplementary motor area (SMA) and deltoid muscle representation.

In a further analysis we subdivided the participants into those that already showed a minimum MOA in the first 10 minutes of training („early efficient“, n=12) and into those participants that acquired efficiency only over the period of training („late efficient“, n=7). The early efficiency group showed a small but significant gain of 10.83±4.63 ms after 40 minutes of training whereas the drop in performance of the late efficient group was 26.3±5.53 ms. Furthermore, event related fMRI in late efficient participants revealed a highly significant increase in connectivity between the APB and deltoid representations after motor learning whereas no such changes could be observed for the late efficient group.

Conclusions: These results provided evidence that functional connectivity changes between M1 muscle representations appears to be a basic central mechanism for establishing movement synergies.