BACKGROUND: Age-related deterioration in the nervous system results in the decline of motor and cognitive abilities, which both have been identified as contributing to fall risk in older adults. Dual-task gait, which involves walking while performing a secondary cognitive task, is a common way to assess the interactions between cognitive and motor function. Previous work has established associations between the cost of the cognitive load on gait parameters (e.g., velocity) and fall risk in older adults. However, to date, no study has explored the potential value of combining a direct measure of performance on the cognitive component of the dual-task with the gait measures in fall risk prediction modeling.
RESEARCH QUESTIONS: Does including measures of performance on the cognitive task in dual-task walking with the gait velocity measures enhance the capacity to predict fall risk. Is this predictive capacity different in models employing dual-task gait velocity versus models including the cost of the cognitive load on gait velocity?
METHODS: Thirty-two community-dwelling older adults (76 years ± 3.44) were classified as fallers (n = 17) and non-fallers (n = 15) based on self-report of having at least one fall in the past 12 months. They completed single-task and dual-task walking on a pressure-sensing electronic walkway system. A progressively enhanced series of logistic regression models were performed commencing with gait velocity during the dual-task (Loaded Gait Velocity, LGV) as the covariate in predicting fall risk. This model was subsequently augmented by adding a measure of cognitive performance covariate and then further augmented with the addition of the interaction variable between the LGV and the cognitive performance variables. This stepped series of modelling was then repeated with the dual-task cost gait velocity (DTCGV, difference in gait velocity between single and dual-task).
RESULTS: With the addition of the cognitive measures (CM) and the interaction variables between the GV and CM variables, in both the LGV and DTC_GV models, the Nagelkerke’s R square increased as did the models’ respective sensitivity. Notably, the model including the LGV, CM and the interaction variables achieved 88.2% sensitivity, 80% specificity, with an overall classification accuracy of 84.4%.
DISCUSSION: This study is the first to show that the ability to identify fallers and non-fallers is enhanced by using both gait and cognition measures as well as interaction variables between gait and cognition measures. Further, our findings suggest that the added value of the cognitive measures is best realized with LGV rather than DTCGV. It reasons that because DTC already encompasses the cost of the cognitive load on the motor performance (gait velocity), combining it with cognitive metrics does not enhance its predictive capacity. This work suggests there is clinical utility of including cognitive performance measures in fall risk modeling as well as it provides further evidence of the interplay between cognitive and motor function in fall risk. / Graduate
| Identifer | oai:union.ndltd.org:uvic.ca/oai:dspace.library.uvic.ca:1828/13444 |
| Date | 05 October 2021 |
| Creators | Mohsenirad, Mahsa |
| Contributors | Hundza, Sandra R., Klimstra, Marc D. |
| Source Sets | University of Victoria |
| Language | English, English |
| Detected Language | English |
| Type | Thesis |
| Format | application/pdf |
| Rights | Available to the World Wide Web |
Page generated in 0.0023 seconds