Wan Xianglin, Zhu Zihao, Xu Feng, Li Qiujie
School of Sport Science, Beijing Sport University, Beijing 100084, China; Key Laboratory for Performance Training & Recovery of General Administration of Sport of China, Beijing 100084, China.
School of Sport Science, Beijing Sport University, Beijing 100084, China.
Gait Posture. 2025 Jul;120:9-16. doi: 10.1016/j.gaitpost.2025.03.022. Epub 2025 Apr 2.
In daily life, stroke patients frequently experience falls during obstacle crossing. Analyzing the gait characteristics of patients in high-risk falling scenarios can help identify and predict fall risks.
Exploring the predictive power of gait characteristics during obstacle crossing for fall risk in stroke patients.
Recruitment of 38 stroke patients with unilateral hemiplegia discharged from rehabilitation. A Qualisys motion capture system and two Kistler force plates were used to record the marker positions and the ground reaction forces during crossing an obstacle 4 cm in height with the affected limb as the leading limb. Gait spatio-temporal parameters, joint angles, and joint moments were calculated. Following a 12-month follow-up survey to collect data on falls among participants, independent samples t-test and binary logistic regression models were employed to identify predictors associated with future fall risk.
During the follow-up period, 13 participants experienced at least one fall and were categorized into the fall group; 14 participants did not experience any falls and were categorized into the non-fall group. Binary logistic regression analysis revealed that the toe-clearance distance of the trailing limb, as well as the peak ankle plantarflexion moment of the trailing limb during double support phase, are effective predictors of fall risk in stroke patients (P < 0.05). The overall correct prediction rate of the regression model incorporating both factors was 85.2 %.
Gait analysis during obstacle crossing holds potential clinical value in identifying future fall risk in stroke patients.
在日常生活中，中风患者在跨越障碍物时经常跌倒。分析高危跌倒场景下患者的步态特征有助于识别和预测跌倒风险。
探索中风患者跨越障碍物时步态特征对跌倒风险的预测能力。
招募38名单侧偏瘫且已从康复机构出院的中风患者。使用Qualisys动作捕捉系统和两个奇石乐测力板，记录以患侧肢体为先导肢体跨越4厘米高障碍物时的标记点位置和地面反作用力。计算步态时空参数、关节角度和关节力矩。在进行为期12个月的随访调查以收集参与者跌倒数据后，采用独立样本t检验和二元逻辑回归模型来识别与未来跌倒风险相关的预测因素。
在随访期间，13名参与者经历了至少一次跌倒并被归入跌倒组；14名参与者未经历任何跌倒并被归入非跌倒组。二元逻辑回归分析显示，后肢的离地间隙距离以及双支撑期后肢的踝关节跖屈力矩峰值是中风患者跌倒风险的有效预测因素（P<0.05）。纳入这两个因素的回归模型的总体正确预测率为85.2%。
跨越障碍物时的步态分析在识别中风患者未来跌倒风险方面具有潜在的临床价值。
