Correlation of knee extensor muscle torque and spasticity with gait speed in patients with stroke.

The purpose of this investigation was to determine in 17 stroke patients the correlation between two independent variables (knee extensor muscle torque [KET] and spasticity on the paretic side) and one dependent variable (gait speed). The patients had a mean age of 59 years; time since onset of first stroke was 51 days. Each variable was measured twice on each of two consecutive days. Both KET and spasticity were measured on a Cybex II dynamometer. Peak torque was measured during five-second maximal voluntary isometric knee extension efforts. Spasticity was defined by the relative angle of reversal (RAR) of the Cybex electrogoniometer curve obtained during pendulum tests. Gait speed was measured as patients walked 8.0 meters at their most comfortable speed while using their usual devices. The average of each day's KETs, RARs, and gait speeds was used in the calculation of Pearson product moment correlations. The correlations between KET and gait speed were significant (p less than .05) and were .574 (day 1) and .571 (day 2). The correlations between the RAR and gait speed were not significant (-.204 day 1 and -.262 day 2). All measurements had intraday and/or interday reliability coefficients (ICC 3,1) exceeding .930. The results of this study confirm that gait performance is correlated with measures of KET but not spasticity in stroke patients. Thus, at the knee, extensor muscle torque, unlike spasticity, appears to be justified as a practical and objective clinical measure.