Degree of Contribution of Motor and Sensory Scores to Predict Gait Ability in Patients With Incomplete Spinal Cord Injury

OBJECTIVE: To identify different contributions of motor and sensory variables for independent ambulation of patients with incomplete spinal cord injury (SCI), and reveal the most significant contributors among the variables. METHODS: The retrospective study included 30 patients with incomplete SCI and lesions were confirmed by magnetic resonance imaging. Motor and sensory scores were collected according to the International Standards for Neurological Classification of Spinal Cord Injury. The variables were analyzed by plotting ROC (receiver operating characteristic) curves to estimate their differential contributions for independent walking. The most significant functional determinant was identified through the subsequent logistic regression analysis. RESULTS: Motor and sensory scores were significantly different between the ambulators and non-ambulators. The majority was associated to the function of lower extremities. Calculation of area under ROC curves (AUC) revealed that strength of hip flexor (L2) (AUC=0.905, p<0.001) and knee extensor (L3) (AUC=0.820, p=0.006) contributed the greatest to independent walking. Also, hip flexor strength (L2) was the single most powerful predictor of ambulation by the logistic regression analysis (odds ratio=6.3, p=0.049), and the model fit well to the data. CONCLUSION: The most important potential contributor for independent walking in patients with incomplete SCI is the muscle strength of hip flexors, followed by knee extensors compared with other sensory and motor variables.