Accuracy and cut-off points of different models of knee extension strength analysis to identify walking performance in individuals with chronic stroke

BackgroundAdequate muscle strength is essential for walking performance in individuals with stroke.ObjectiveTo investigate the accuracy of different forms of muscle knee extension strength analysis to identify high or low walking performance in individuals with chronic stroke.MethodsTwenty-eight participants with a chronic stroke for more than six months participated. Independence for walking was judged by measurement of walking performance assessed for comfortable walking speed (CWS), maximum walking speed (MWS), and the Six Minute Walk Test (6MWT). Peak knee extension torque of the paretic side, non-paretic side, sum of the sides (SS), and difference in the sides (DS) was assessed during concentric movements using an isokinetic dynamometer.ResultsThe equation with greatest predictive capacity for CWS and MWS included the DS as the main predictor (R² of 0.65 and 0.71, respectively, p < 0.05). The variable with the greatest predictive capacity for 6MWT was time since injury (R² of 0.68, p < 0.05). The highest percentile for CWS in the receiver operating characteristic curve of DS was 25 Nm/kg (cut-off: -12.75 for CWS of 0.498 m/s). The 75th percentile of the 6MWT (324.3 m) was used as the cut-off for the SS (2.1 Nm/kg). The area under the curve for CWS was 0.76 (p < 0.05) on the DS and 0.75 (p < 0.05) for 6MWT on the SS.ConclusionThe models of muscle knee extension strength analysis using the SS and DS presented moderate accuracy to identify walking performance in individuals with chronic stroke.