It is good to see that stroke, and in particular stroke rehabilitation, is losing its Cinderella status in the world of neurology. In this issue, Cauraugh and Kim (p 000) present the results of some recent work on the rehabilitation of arm function in chronic stroke patients.¹ This work touches on several areas of interest to both clinicians and basic scientists.

Their work was in part motivated by suggestions in the motor learning literature that the repetitive practice of the same action (“blocked” practice) may be inferior to practising different actions (“random” practice). If this were so, then it would have important implications for everyday clinical practice, in which patients are typically required to perform the same action time and again (the equivalent of blocked practice). With the particular training schemes used, however, no difference was found between the two groups.

A more positive result came from a comparison of the performance of patients allocated to receive or not neuromuscular stimulation assistance. In the assisted group, once the EMG level in the voluntarily activated muscle exceeded a threshold value, electrical stimulation was delivered to that muscle to assist the action. Patients who received such assistance fared better than those who received traditional physical therapy. Others have reported similar beneficial effects of such assistance in stroke and paraplegia. We have used an active computer joystick to guide the responses of chronic stroke patients in a novel approach to cognitive rehabilitation.² It is likely that the use of such technologies will play an increasingly important role in stroke rehabilitation.

The authors also showed that the impaired arm tended to perform better when tested during bimanual actions than when used alone. We know, of course, that many everyday actions are bimanual, and that there is extensive neural machinery involved in the coordination of bimanual action. Rather than focusing on the retraining of the impaired arm alone, perhaps we should be adopting a more even handed—and cheap—approach to the rehabilitation of hemiparesis.

Following stroke in man, there is evidence of plastic reorganisation in both hemispheres. A consensus view is emerging that plasticity in the hemisphere contralateral to the affected limbs is more important for recovery than reorganisation on the other side.³ Indeed, more complete recovery appears to be associated with a more normal—that is, relatively restricted and contralateral—pattern of brain activation.⁴ Nudo et al have shown in non-human primates that the primary motor cortex (M1) can reorganise after a focal vascular lesion if there is motor skill retraining. In animals not trained after a stroke, there is a further reduction in the size of the hand representation in M1.⁵ Cauraugh and Kim showed that neuromuscular stimulation assistance improved the ability of chronic stroke patients to use their paretic arms. One might speculate that this assistance increases the likelihood of appropriate plasticity in the motor areas, particularly on the contralateral side. This suggestion could be pursued with transcranial magnetic stimulation and functional imaging studies.

These are exciting developments. Findings from animal neurophysiology, human behavioural neuroscience and imaging studies, and the use of technological assistance hold out the prospect of more rational and effective rehabilitation strategies for the victims of stroke.