| dc.description.abstract | Maintaining healthy gait into old age is key to preserving quality of life and reducing risk of falling. This thesis focuses on exploring the potential of nonlinear dynamic analysis (NLD) and standardized metrics for assessing gait characteristics associated with fall risk. The primary aim was to define the NLD metrics that best differentiate people with a recent history of falling from those who have not (fallers and non-fallers). By establishing these standardized metrics, the purpose is to enable their application in identifying individuals at risk of falls and in evaluating the effectiveness of interventions aimed at preventing falls. This research seeks to contribute to the testing of more targeted and evidence-based fall prevention strategies in the future.
The thesis begins with a systematic review of NLD measures that have been used to distinguish gait kinematics between healthy older adults and those prone to falling. This review highlights the need for standardized metrics to improve consistency and comparability across studies.
To address this research gap, three empirical studies were conducted. The first study aimed to define NLD metrics that effectively differentiate fallers from non-fallers. Thirty-four healthy female participants (17 with history of falling and 17 without history of falling in the last year), performed walking trials at their preferred walking speed (PW) and at ±20% of their PW on a treadmill (M-Gait, Motek Medical BV, Amsterdam, Netherlands). Kinematic data, obtained from motion capture (Miqus M3, Qualisys AB, Gothenburg, Sweden) and an inertial measurement unit (IMU, Blue Trident, Vicon Motion Systems Ltd, Oxford, UK) placed on the lower back, were collected during treadmill walking trials. A number of kinematic variables were analysed using the Short-term and Long-term Lyapunov Exponent (LyE) based on Rosenstein’s method, Lyapunov Exponent using Wolf’s methods and Multiscale Entropy (MSE). The findings demonstrated that short term LyE analysis using Rosenstein’s method applied to trunk acceleration in the anterior-posterior direction, collected via the IMU during preferred and slow walking speeds, better distinguished fallers from non-fallers.
The second study investigated the effects of different walking modalities (treadmill and overground walking) on metrics of gait stability. Similar to the first study, thirty-four female participants performed walking trials, and LyE analysis was applied to trunk acceleration data collected by the IMU. The results revealed that both overground and treadmill walking at preferred speed successfully differentiated fallers from non-fallers. Secondly, measuring preferred walking speed on the treadmill rather than using a fixed treadmill speed or relying on overground speed on the treadmill, is recommended for accurate assessments of gait dynamics.
The third study examined the effects of ageing on gait complexity using MSE analysis of trunk acceleration. A total of fifty-one female participants (17 young adults, 17 older adults with history of falling and 17 older adults without history of falling in the last year) with a mean age of 73 years (SD = 5.3) and 27 years (SD = 4.3) for older and young adults, respectively), completed overground walking trials. The findings indicated significant differences in MSE of trunk acceleration between fallers and non-fallers, supporting the use of MSE as a valuable biomarker for distinguishing healthy gait from unhealthy gait. However, the study did not observe significant changes in gait complexity related to age using MSE, suggesting the need for further exploration of alternative complexity metrics.
Overall, this thesis contributes to the field by proposing standardized metrics, particularly LyE of trunk acceleration, as a sensitive and valid tool for differentiating fallers from non-fallers based on gait kinematics. The findings highlight the importance of gait analysis in clinical settings and provide insights for developing targeted and evidence-based fall prevention strategies. By utilizing NLD measures and standardized metrics in the future, healthcare professionals can effectively evaluate gait dynamics in older adults, enabling early identification of fall risk and facilitating interventions to enhance their safety and well-being. | en_GB |
