Associations between quantitative mobility measures derived from components of conventional mobility testing and Parkinsonian gait in older adults

Abstract

Objective: To provide objective measures which characterize mobility in older adults assessed in the community setting and to examine the extent to which these measures are associated with parkinsonian gait.

Methods: During conventional mobility testing in the community-setting, 351 ambulatory non-demented Memory and Aging Project participants wore a belt with a whole body sensor that recorded both acceleration and angular velocity in 3 directions. We used measures derived from these recordings to quantify 5 subtasks including a) walking, b) transition from sit to stand, c) transition from stand to sit, d) turning and e) standing posture. Parkinsonian gait and other mild parkinsonian signs were assessed with a modified version of the original Unified Parkinson's Disease Rating Scale (mUPDRS).

Results: In a series of separate regression models which adjusted for age and sex, all 5 mobility subtask measures were associated with parkinsonian gait and accounted for 2% to 32% of its variance. When all 5 subtask measures were considered in a single model, backward elimination showed that measures of walking sit to stand and turning showed independent associations with parkinsonian gait and together accounted for more than 35% of its variance. Cross-validation using data from a 2(nd) group of 258 older adults showed similar results. In similar analyses, only walking was associated with bradykinesia and sway with tremor.

Interpretation: Quantitative mobility subtask measures vary in their associations with parkinsonian gait scores and other parkinsonian signs in older adults. Quantifying the different facets of mobility has the potential to facilitate the clinical characterization and understanding the biologic basis for impaired mobility in older adults.

Figure 1. Timed Get Up and Go (TUG) Subtasks are Best Identified from Different Channels.

This figure shows the acceleration and rotation signals recorded with a whole body sensor during conventional mobility testing of TUG in the community setting. The top channel shows the acceleration signal from the Anterior-Posterior (Blue) axis. The second channel shows the rotation signal of Yaw (Green, rotation around the vertical axis). Third, is the Pitch signal (Red, rotation around the mediolateral axis). The current study focused on several TUG subtasks including transition from sit to stand (S1), transition from stand to sit (S2), Turn 1 during the middle of the TUG and a second, Turn 2 which occurs immediately prior to sitting back down (S2). Walking measures can be extracted but in this study were derived from a 32 ft walk. To facilitate subsequent analyses, marks were inserted in the recorded data by the research assistant to identify the beginning and the end of each of the 3 performances analyzed in this study. The black star (M1) shows the first mark inserted when the research assistant pressed a button on the device immediately prior to instructing the participant to begin moving for the TUG. A second mark (M2) was inserted at the end when the task was completed. The M1 and M2 marks were used to extract the entire TUG trial from the continuous recording of the entire mobility testing session. After extraction of the entire TUG trial, an automatic algorithm was then applied for detecting the exact start and end times of the TUG based on the start time of the sit-to-stand (S1) and end time of the stand-to-sit (S2) AP signal (solid black line on AP channel). The Turn subtasks are visualized best from the Yaw (green) channel [black solid arrows Similarly, the Transition measures (S1 & S2) are best visualized on the AP (blue) and Pitch (red) channels [solid black arrows on the pitch]. Gait measures were derived from the onset and offset of the turns and transitions which are illustrated as described in the text (Appendix S1).