The use of wearable inertial measurement units to assess gait and balance outcomes related to fall risk among older adults

Abstract

Due to the prevalence and associated health, social and economic costs of falls among older adults, this thesis originally aimed to identify a more robust and objective way of assessing fall risk factors with the use of wearable inertial measurement units (IMU). However, due to unforeseen circumstances, the direction of the thesis had to be changed. Therefore, the thesis aimed to investigate whether gait and balance outcomes related to fall risk, when measured with wearable IMUs are sensitive to conditions which may replicate clinical and habitual environments.

In Study one, a systematic scoping review was conducted to identify characteristic differences between fallers and non-fallers with the use of IMUs. The lower trunk was the most common anatomical location, whilst walking a predetermined distance indoors was the most common test used with IMUs to distinguish between fallers and non-fallers.

In Study two, seventeen older and seventeen younger adults performed multiple walking and standing tasks in a laboratory. Older adults had a lower root mean square of the IMU acceleration signal, harmonic ratio and greater step time asymmetry compared to younger adults. The use of a cognitive dual task caused gait to be slower and less symmetrical among older and younger adults. Trunk displacement to quantify trunk sway during quiet standing was greater among older adults and increased as standing conditions became more difficult. Older adults exhibited distinct differences in gait when walking indoors and outdoors. The results of Study two suggested that IMUs may identify differences between older and younger adults regarding walking speed and time to completion of clinical tests, even when a stopwatch could not.

In Study three, twenty older and twenty younger adults had IMUs attached to different anatomical locations during waking hours. There were differences in all gait variables when walking supervised in the laboratory and unsupervised in habitual indoor environments for both older and younger adults. There were also large differences in gait variables when walking indoors and outdoors. These results suggest the need for future studies in continuous, outdoor and unsupervised free-living conditions, with regards to fall risk assessments.

This thesis demonstrates that gait and balance outcomes related to fall risk, when measured using wearable IMUs, are sensitive to conditions resembling habitual and clinical environments among both older and younger adults. This could prove valuable for the enhancement of future fall risk research.