Saikat Chakraborty1, Anup Nandy2, Trisha M Kesar3. 1. Machine Intelligence and Bio-motion Research Lab., Department of Computer Science and Engineering, National Institute of Technology, Rourkela, India. Electronic address: saikat.scgmail.com. 2. Machine Intelligence and Bio-motion Research Lab., Department of Computer Science and Engineering, National Institute of Technology, Rourkela, India. 3. Division of Physical Therapy, Department of Rehabilitation Medicine, Emory University, Atlanta, GA, USA.
Abstract
BACKGROUND: Studies have demonstrated that ambulatory children and adolescents with cerebral palsy demonstrate atypical gait patterns. Out of numerous gait variables, identification of the most deteriorated gait parameters is important for targeted and effective gait rehabilitation. Therefore, this study aimed to identify the gait parameters with the most discriminating nature to distinguish cerebral palsy gait from normal gait. METHODS: Multiple databases were searched to include studies on ambulatory children and adolescents with cerebral palsy that included gait (spatio-temporal, kinematic, and kinetic) and dynamic stability variables. FINDINGS: Of 68 studies that met the inclusion criteria, 35 studies were included in the meta analysis. Effect size was used to assess the discriminative strength of each variable. A large effect (≥ 0.8) of cerebral palsy on double limb support time (Standardized Mean Difference = 0.98), step length (Standardized Mean Difference = 1.65), step width (Standardized Mean Difference = 1.21), stride length (Standardized Mean Difference = 1.75), and velocity (Standardized Mean Difference = 1.42) was observed at preferred-walking speed. At fast-walking speed, some gait variables (i.e. velocity and stride length) exhibited larger effect size compared to preferred-walking speed. For some kinematic variables (e.g. range of motion of pelvis), the effect size varied across the body planes. INTERPRETATION: Our systematic review detects the most discriminative features of cerebral palsy gait. Non-uniform effects on joint kinematics across the anatomical planes support the importance of 3D gait analysis. Differential effects at fast versus preferred speeds emphasize the importance of measuring gait at a range of speeds.
BACKGROUND: Studies have demonstrated that ambulatory children and adolescents with cerebral palsy demonstrate atypical gait patterns. Out of numerous gait variables, identification of the most deteriorated gait parameters is important for targeted and effective gait rehabilitation. Therefore, this study aimed to identify the gait parameters with the most discriminating nature to distinguish cerebral palsy gait from normal gait. METHODS: Multiple databases were searched to include studies on ambulatory children and adolescents with cerebral palsy that included gait (spatio-temporal, kinematic, and kinetic) and dynamic stability variables. FINDINGS: Of 68 studies that met the inclusion criteria, 35 studies were included in the meta analysis. Effect size was used to assess the discriminative strength of each variable. A large effect (≥ 0.8) of cerebral palsy on double limb support time (Standardized Mean Difference = 0.98), step length (Standardized Mean Difference = 1.65), step width (Standardized Mean Difference = 1.21), stride length (Standardized Mean Difference = 1.75), and velocity (Standardized Mean Difference = 1.42) was observed at preferred-walking speed. At fast-walking speed, some gait variables (i.e. velocity and stride length) exhibited larger effect size compared to preferred-walking speed. For some kinematic variables (e.g. range of motion of pelvis), the effect size varied across the body planes. INTERPRETATION: Our systematic review detects the most discriminative features of cerebral palsy gait. Non-uniform effects on joint kinematics across the anatomical planes support the importance of 3D gait analysis. Differential effects at fast versus preferred speeds emphasize the importance of measuring gait at a range of speeds.
Authors: Sophie Wist; Lena Carcreff; Sjoerd M Bruijn; Gilles Allali; Christopher J Newman; Joel Fluss; Stéphane Armand Journal: PLoS One Date: 2022-06-22 Impact factor: 3.752