Movement-specific effects of transcranial magnetic stimulation over premotor and parietal cortices after stroke

Background: Transcranial magnetic stimulation (TMS) has been proposed to support post-stroke motor recovery, but the response may depend on the task and motor-network status. We aimed to compare how premotor and parietal stimulation affects performance across reach directions with different movement demands, in preparation for therapeutic application of TMS. Methods: 16 participants with chronic subcortical stroke performed planar reaches with the KINARM exoskeleton toward distal and proximal targets. Before movement, a 10-Hz triple-pulse TMS train was delivered over bilateral dorsal or ventral premotor cortex (PMd, PMv) or dorsomedial posterior parietal cortex (PPC) at subthreshold and suprathreshold intensities. The parietal midline was used as the control target. Participant-level bootstrap confidence intervals assessed changes in trajectory straightness and smoothness relative to the control target and differences between reach directions. We compared motor-network activity and connectivity using task-based and resting-state functional magnetic resonance imaging (fMRI) data from participants with stroke and neurologically intact controls. To explore response variability, we tested associations between TMS response and motor impairment, resting-state connectivity, and the change in task-evoked activity at the stimulated region from planning to movement. Results: Suprathreshold PMd and PPC stimulation in both hemispheres improved trajectory straightness or smoothness during distal reaches relative to the control target. Suprathreshold contralesional PMd stimulation increased trajectory deviation and reduced smoothness during proximal reaches. Baseline performance during distal reaches was more closely related to motor impairment, and contralesional PMd effects were more favorable during distal than proximal reaches across all kinematic measures. Participants with stroke showed lower bilateral PMv activation during movement and stronger interhemispheric PPC connectivity. Only the change in task-evoked activity at the stimulated region from planning to movement was related to TMS response. Smaller changes relative to controls were associated with more favorable kinematic effects during distal reaches. Conclusions: Movements more affected after stroke were the most responsive to this type of stimulation. The response to TMS was also associated with task-evoked activity differences at the stimulated region between movement planning and execution. Neuromodulation may therefore be most effective when designed around both the pathophysiology of the cortical target and the type of movement.