ABSTRACT
PURPOSE: To identify which paretic and non-paretic muscular groups of the lower limbs were the best predictors of gait speed in chronic hemiparetic subjects. METHOD: Twelve hemiparetic subjects with ages ranging from 65 to 75 years (70.67 +/- 3.31 years) were included in this cross-sectional study. All participants had time since onset of stroke of at least six months and the ability to walk independently. Main outcome measures included self-selected gait speed and maximum isometric torques of both the paretic and non-paretic lower limbs. Pearson's correlation coefficients and stepwise regression analysis were employed to investigate the level of associations between the outcome variables. RESULTS: Mean gait speed was 0.65 +/- 0.33 ms(-1). There were significant differences (p < 0.01) in torques values between paretic and non-paretic lower limbs for all assessed muscular groups. Stepwise regression analysis revealed that the most important independent variable for predicting gait speed was the isometric torque of paretic knee flexors ([image omitted] ). No significant associations were found between any torque values of the non-paretic side and gait speed. CONCLUSIONS: The paretic knee flexors were shown to play an important role in predicting gait speed after stroke, but no influences were found for the muscular groups of the non-paretic side. Therefore, findings of the present study suggest that paretic knee flexors torque should be addressed in rehabilitation programmes aiming to improve gait speed after stroke.