Objective: To assess the efficacy of motor imagery training for arm function in chronic stroke patients. The relation between mental processes such as attentional and perceived personal control over recovery, and motor imagery was additionally investigated.
Design and subjects: Twenty patients with long-term motor impairments (mean two years post stroke), were assessed before and after four weeks of training. Ten patients mentally rehearsed movements with their affected arm. Their recovery was compared with patients who performed nonmotor imagery (n =5), or who were not engaged in mental rehearsal (n=5).
Setting: Patients were recruited from the stroke database of Ninewells Hospital, Dundee. Assessment and training were performed at the patients' home.
Interventions: The motor imagery group was asked to practise daily imagining moving tokens with their affected arm. The nonmotor imagery group rehearsed visual imagery of previously seen pictures. All patients practised physically moving the tokens.
Main measures: The following variables were assessed before and after training: motor function (training task, pegboard and dynamometer), perceived locus of control, attention control and ADL independence.
Results: All patient groups improved on all motor tasks except the dynamometer. Improvement was greater for the motor imagery group on the training task only (average of 14% versus 6%). No effect of motor imagery training was found on perceived or attentional control.
Conclusions: Motor imagery training without supervision at home may improve performance on the trained task only. The relation between movement imagery, attention and perceived personal control over recovery remained unclear.