Upper limb (UL) paresis after stroke often leads to poor functional outcome. Application of transcranial direct current stimulation (tDCS) can modulate intra- and inter-hemispheric dynamics and promote treatment-related adaptive neural plasticity. However, the optimal stimulation mode for a patient with a given lesion pattern and a given level of motor impairment is unknown. Here we aimed to assess the likelihood of obtaining improvements in endpoint kinematics of the paretic UL, using different tDCS modes in stroke patients with variant residual motor functioning.

Twenty-eight subacute stroke patients performed planar point-to-point reaching movements with the paretic upper limb, before and after anodal-tDCS (excitatory), cathodal-tDCS (inhibitory) and sham-tDCS, in a randomized order. tDCS was applied over the left and right M1 hand area (C3, C4). Spatio-temporal kinematic variables derived from the minimum-jerk model were analyzed before and after training, comparing the 3 stimulation modes one with each other.

a-tDCS had a larger effect when applied on the lesioned hemisphere. Number of peaks, straight-line deviation and movement time improved (only the former two reached statistical significance). The Fugl Meyer (FM) score at baseline correlated negatively with the magnitude of improvement post-anodal stimulation of the contralesional hemisphere. This was reflected in number of peaks and in movement time.

a-tDCS applied on the lesioned hemisphere combined with motor training shows immediate positive effects on path smoothness (less fragmented paths) and path straightness (shorter paths). Higher FM scores at baseline likely imply smaller corticospinal damage in the affected hemisphere, and thus, greater likelihood that perilesional re-mapping will be the dominant mechanism underlying recovery. In the presence of higher corticospinal reserve, one would expect anodal (excitatory) stimulation of the contralesional hemisphere to exert a detrimental effect on function. It remains to be shown whether such stimulation is beneficial in the absence of corticospinal reserve.