BACKGROUND: Wearable sensors (WS) can accurately measure body motion and provide interactive feedback for supporting motor learning. OBJECTIVE: This review aims to summarize current evidence for the effectiveness of WS training for improving balance, gait and functional performance. METHODS: A systematic literature search was performed in PubMed, Cochrane, Web of Science, and CINAHL. Randomized controlled trials (RCTs) using a WS exercise program were included. Study quality was examined by the PEDro scale. Meta-analyses were conducted to estimate the effects of WS balance training on the most frequently reported outcome parameters. RESULTS: Eight RCTs were included (Parkinson n = 2, stroke n = 1, Parkinson/stroke n = 1, peripheral neuropathy n = 2, frail older adults n = 1, healthy older adults n = 1). The sample size ranged from n = 20 to 40. Three types of training paradigms were used: (1) static steady-state balance training, (2) dynamic steady-state balance training, which includes gait training, and (3) proactive balance training. RCTs either used one type of training paradigm (type 2: n = 1, type 3: n = 3) or combined different types of training paradigms within their intervention (type 1 and 2: n = 2; all types: n = 2). The meta-analyses revealed significant overall effects of WS training on static steady-state balance outcomes including mediolateral (eyes open: Hedges' g = 0.82, CI: 0.43-1.21; eyes closed: g = 0.57, CI: 0.14-0.99) and anterior-posterior sway (eyes open: g = 0.55, CI: 0.01-1.10; eyes closed: g = 0.44, CI: 0.02-0.86). No effects on habitual gait speed were found in the meta-analysis (g = -0.19, CI: -0.68 to 0.29). Two RCTs reported significant improvements for selected gait variables including single support time, and fast gait speed. One study identified effects on proactive balance (Alternate Step Test), but no effects were found for the Timed Up and Go test and the Berg Balance Scale. Two studies reported positive results on feasibility and usability. Only one study was performed in an unsupervised setting. CONCLUSION: This review provides evidence for a positive effect of WS training on static steady-state balance in studies with usual care controls and studies with conventional balance training controls. Specific gait parameters and proactive balance measures may also be improved by WS training, yet limited evidence is available. Heterogeneous training paradigms, small sample sizes, and short intervention durations limit the validity of our findings. Larger studies are required for estimating the true potential of WS technology.
BACKGROUND: Wearable sensors (WS) can accurately measure body motion and provide interactive feedback for supporting motor learning. OBJECTIVE: This review aims to summarize current evidence for the effectiveness of WS training for improving balance, gait and functional performance. METHODS: A systematic literature search was performed in PubMed, Cochrane, Web of Science, and CINAHL. Randomized controlled trials (RCTs) using a WS exercise program were included. Study quality was examined by the PEDro scale. Meta-analyses were conducted to estimate the effects of WS balance training on the most frequently reported outcome parameters. RESULTS: Eight RCTs were included (Parkinson n = 2, stroke n = 1, Parkinson/stroke n = 1, peripheral neuropathy n = 2, frail older adults n = 1, healthy older adults n = 1). The sample size ranged from n = 20 to 40. Three types of training paradigms were used: (1) static steady-state balance training, (2) dynamic steady-state balance training, which includes gait training, and (3) proactive balance training. RCTs either used one type of training paradigm (type 2: n = 1, type 3: n = 3) or combined different types of training paradigms within their intervention (type 1 and 2: n = 2; all types: n = 2). The meta-analyses revealed significant overall effects of WS training on static steady-state balance outcomes including mediolateral (eyes open: Hedges' g = 0.82, CI: 0.43-1.21; eyes closed: g = 0.57, CI: 0.14-0.99) and anterior-posterior sway (eyes open: g = 0.55, CI: 0.01-1.10; eyes closed: g = 0.44, CI: 0.02-0.86). No effects on habitual gait speed were found in the meta-analysis (g = -0.19, CI: -0.68 to 0.29). Two RCTs reported significant improvements for selected gait variables including single support time, and fast gait speed. One study identified effects on proactive balance (Alternate Step Test), but no effects were found for the Timed Up and Go test and the Berg Balance Scale. Two studies reported positive results on feasibility and usability. Only one study was performed in an unsupervised setting. CONCLUSION: This review provides evidence for a positive effect of WS training on static steady-state balance in studies with usual care controls and studies with conventional balance training controls. Specific gait parameters and proactive balance measures may also be improved by WS training, yet limited evidence is available. Heterogeneous training paradigms, small sample sizes, and short intervention durations limit the validity of our findings. Larger studies are required for estimating the true potential of WS technology.
Authors: Kaleen M Lavin; Paul M Coen; Liliana C Baptista; Margaret B Bell; Devin Drummer; Sara A Harper; Manoel E Lixandrão; Jeremy S McAdam; Samia M O'Bryan; Sofhia Ramos; Lisa M Roberts; Rick B Vega; Bret H Goodpaster; Marcas M Bamman; Thomas W Buford Journal: Compr Physiol Date: 2022-03-09 Impact factor: 8.915
Authors: Karli Kondo; Katherine M Noonan; Michele Freeman; Chelsea Ayers; Benjamin J Morasco; Devan Kansagara Journal: J Gen Intern Med Date: 2019-08-14 Impact factor: 5.128
Authors: Daniel Kelly; Karla Muñoz Esquivel; James Gillespie; Joan Condell; Richard Davies; Shvan Karim; Elina Nevala; Antti Alamäki; Juha Jalovaara; John Barton; Salvatore Tedesco; Anna Nordström Journal: Sensors (Basel) Date: 2021-06-28 Impact factor: 3.576