THINKING ABOUT WALKING: A NEW APPROACH TO QUANTIFYING GAIT INITIATION USING A WEARABLE SENSOR. (11th November 2018)

• Record Type: Journal Article
• Authors: Hausdorff, J; Geffen, N; Gazit, E; Dawe, R; Mirelman, A; Curran, T; Buchman, A
• Is Part Of: Innovation in aging. Volume 2(2018)Supplement 1
• Page Start: 521
• Page End: 521
• Publication Date: 2018-11-11
• Subjects: Aging -- Periodicals; Gerontology -- Periodicals; 612.67
• DOI: 10.1093/geroni/igy023.1928
• Languages: English
• ISSNs: 2399-5300

Abstract

Abstract: Gait initiation, a task that demands the orchestration of cognitive and motor processes in order to transition from a standing posture to the onset of walking, can be partitioned into three phases: 1) The duration of motor planning extends from the signal to move until the onset of anticipatory postural adjustments (APA), time to APA. 2) The APA ends when the center-of-pressure returns to baseline; 3) Step duration, the "motor" implementation time, extends from the end of APA to heel-strike. To measure gait initiation metrics in large-scale studies, we developed an automated approach based on changes in the waveforms obtained from 3D accelerometer recordings from the lower back. To validate these metrics, we compared them with metrics extracted from concurrent instrumented gait-mat recordings in 41 healthy young adults (31.2 ± 5.0 yrs; %51 women) and 14 older adult women (mean age: 76.1 ± 6.3 yrs). Time to APA, APA duration, and step duration estimates derived from the wearable sensor and gait-mat were highly correlated (p<0.001) in both young and older adults, with small differences between the two approaches (e.g., time to APA difference: 38.7 ± 31.944 msec). Using both methods, time to APA was significantly longer in the older women, compared to the young women, when the dominant leg lead; APA duration and step time were not different in the young and older women. These findings confirm that the planning phase of gait initiation is longer among older adults and demonstrate the feasibility of using an automated, accelerometer-based approach to quantify gait initiation in large-scale studies.