Physical training is known to be an effective intervention to improve sensorimotor impairments after stroke. However, the link between brain plastic changes, assessed by transcranial magnetic stimulation (TMS), and sensorimotor recovery in response to physical training is still misunderstood. We systematically reviewed reports of randomized controlled trials (RCTs) involving the use of TMS over the primary motor cortex (M1) to probe brain plasticity after upper-limb physical training interventions in people with stroke.

We searched 5 databases for articles published up to October 2016, with additional studies identified by hand-searching. RCTs had to investigate pre/post-intervention changes in at least one TMS outcome measure. Two independent raters assessed the eligibility of potential studies and reviewed the selected articles’ quality by using 2 critical appraisal scales.

In total, 14 reports of RCTs (pooled participants=358; mean 26±12 per study) met the selection criteria. Overall, 11 studies detected plastic changes with TMS in the presence of clinical improvements after training, and these changes were more often detected in the affected hemisphere by using map area and motor evoked potential (MEP) latency outcome measures. Plastic changes mostly pointed to increased M1/corticospinal excitability and potential interhemispheric rebalancing of M1 excitability, despite sometimes controversial results among studies. Also, the strength of the review observations was affected by heterogeneous TMS methods and upper-limb interventions across studies as well as several sources of bias within the selected studies.

The current evidence encourages the use of TMS outcome measures, especially MEP latency and map area to investigate plastic changes in the brain after upper-limb physical training post-stroke. However, more studies involving rigorous and standardized TMS procedures are needed to validate these observations.