RESUMEN

OBJECTIVE: To determine the main factor that predicts balance impairment in patients with chronic stroke. DESIGN: Cross-sectional study. SETTING: Inpatient rehabilitation hospital and research laboratory. PARTICIPANTS: A total of 57 patients (42 men, 15 women; mean age 55.7±12.2 years) with chronic symptoms after stroke. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Primary outcomes were ankle functions, including strength, range of motion, and proprioception, and balance, including Berg Balance Scale score and Timed Up and Go test values. Secondary outcomes included gait kinematics, Fugl-Meyer Scale score, and Fall Efficacy Scale score. RESULTS: According to the cutoff score <46 on the Berg Balance Scale and the Timed Up and Go test ≥13.5 seconds, 21 patients were classified as having a balance impairment (36.8%). Multivariable logistic regressions showed that ankle proprioception (odds ratio = 3.49; 95% confidence interval, 1.17-10.42) was a significant predictor when coupled with step length (odds ratio = 0.00; 95% confidence interval, 0.00-0.22). A cutoff score of 2.59 for the ankle proprioception value predicts balance impairment in patients with stroke (area under the curve 0.784). CONCLUSION: Ankle proprioception can be used to predict balance impairment in patients with stroke.