Changes in the hemiparetic limb with training. I. Torque output.

The study concerns eleven patients who had suffered from a cerebro-vascular accident (CVA) less than 6 months previously and who were hospitalized for treatment. In addition to their multidisciplinary rehabilitation treatments, the patients consented to participate in an electromyographic (EMG) biofeedback program designed to increase their shoulder activity. From blocks of five individual isometric contractions, curves of both the mean torque and of its standard deviation (SD) were generated and from each of these, a mean value was obtained. These two measures were used to analyze the effects on the biofeedback training had induced on shoulder activity. The training induced negligible changes on the contralateral (CL) side. The mean torque value for the group was 31.02 post vs 30.99 Nm pre-training. However, for the hemiplegic (HEMI) limbs, some patients showed improvement while others did not. The treatment was considered a success when the mean torque values obtained in the last 2/3 of the training or at the post-training session were > or = 1.5 the values obtained during the pre-training session. Based on this criteria, it was found that six out of eleven patients did improve their muscular activity through the training. Among these, the improvement of four was greater during the 2/3 session than that observed at the post-training session. The six successful patients were found to have a pre-training upper extremity Fugl-Meyer score < or = 13. Two patients had a similar pre-training score of 11, and met the criteria of success, but the younger patient performed better than the older one. Patients with scores of 20 and higher did not meet the 1.5 criteria. A correlation (r = 0.80) was found between post/pre ratios of Fugl-Meyer scores and those of the mean torque: the test thus seems to have a limited usefulness for assessing local changes such as the effects of the biofeedback treatment on the shoulder.