Do Changes in Mental Energy and Fatigue Impact Functional Assessments Associated with Fall Risks? An Exploratory Study Using Machine Learning. (2nd July 2020)

• Record Type: Journal Article
• Authors: Boolani, Ali; Ryan, Jenna; Vo, Trang; Wong, Brandon; Banerjee, Natasha Kholgade; Banerjee, Sean; Fulk, George; Smith, Matthew Lee; Martin, Rebecca
• Is Part Of: Physical & occupational therapy in geriatrics. Volume 38:Number 3(2020)
• Page Start: 283
• Page End: 301
• Publication Date: 2020-07-02
• Subjects: Older adults -- falls -- energy -- fatigue -- Berg Balance Scale -- machine learning; Physical therapy for older people -- Periodicals; Occupational therapy for older people -- Periodicals; Older people -- Rehabilitation -- Periodicals; 615.820846
• DOI: 10.1080/02703181.2020.1748788
• Languages: English
• ISSNs: 0270-3181

Abstract

Abstract: Using a crossover-design, we assessed changes in 30-second chair stand test (30 s-CST), Timed Up-and-Go (TUG) and Berg Balance Scale (BBS) and energy and fatigue in older adults (N = 11) after performance of mental tasks. A Wilcoxon Sign Rank Test and a Friedman's rank test were used to assess changes in 30 s-CST, TUG, BBS and energy and fatigue respectively. A linear mixed model was used to assess joint variance and random forest classifier and support vector machine (SVM) algorithms were used to verify results. Statistically significant declines in feelings of energy (p=.003), specifically mental energy (p=.015), and BBS (p<.001), specifically during the "standing with eyes closed" (SEC), was noted for participants on days when they completed mental tasks compared to days they did not. The random-forest and SVM algorithms predicted with 79% and 80% accuracy respectively whether the SEC item of the BBS was performed after a decline a mental energy.