Proximal and Distal Bimanual Motor Skill Learning After Stroke: Impairments, Trajectories, and Clinical Contributors

Background. Although coordination between the upper limbs (ULs) is essential for daily life activities and may benefit from early post-stroke plasticity, how patients early after stroke achieve bimanual motor skill learning (bim-MSkL) remains poorly understood. We investigated whether acute and subacute stroke patients retained bim-MSkL capacity, whether bimanual control policy and motor sequence were dissociable components, whether bim-MSkL was more impaired in the distal than proximal UL segment, and whether clinical factors predicted bim-MSkL. Methods. Ninety-seven (sub)acute stroke patients and 60 age-matched healthy individuals (HI) trained over three days on proximal and distal bimanual tasks in this randomized trial. Baseline motor and neuropsychological impairments were assessed. On day 3, participants were randomized into two subgroups on the proximal task to dissociate control policy from motor sequence. Learning trajectories, perturbation-related performance disruption, and predictors of improvement were analyzed. Results. Accuracy and coordination improved in patients on both the distal and proximal tasks over three days, although less than in HI, while proximal speed improvement remained limited. Control policy swap induced larger performance disruption than motor sequence change, whereas the relearning slopes did not differ in patients. Unlike HI, patients showed no proximal-to-distal gradient in bim-MSkL despite greater distal than proximal baseline impairment. Baseline tasks performances, subacute motor and attentional impairments were the strongest predictors of bim-MSkL impairment early after stroke. Conclusions. Most patients with mild-to-moderate impairment exhibited substantial residual proximal and distal bimanual motor learning capacity early after stroke. Impairment profiles support individualized bimanual rehabilitation strategies integrating both motor and cognitive demands.