Research Article
Assessing Older Adult’s Gait Speed with Wearable Accelerometers in Community Settings: Validity and Reliability Study
@INPROCEEDINGS{10.1007/978-3-031-34586-9_10, author={Antonio Cobo and Elena Villalba-Mora and Rodrigo P\^{e}rez-Rodr\^{\i}guez and Juan E. Medina and Paula Robles-Mateos and \^{A}ngel Rodr\^{\i}guez-Laso and Leocadio Rodr\^{\i}guez-Ma\`{o}as}, title={Assessing Older Adult’s Gait Speed with Wearable Accelerometers in Community Settings: Validity and Reliability Study}, proceedings={Pervasive Computing Technologies for Healthcare. 16th EAI International Conference, PervasiveHealth 2022, Thessaloniki, Greece, December 12-14, 2022, Proceedings}, proceedings_a={PERVASIVEHEALTH}, year={2023}, month={6}, keywords={gait speed wearabe accelerometer validity reliability}, doi={10.1007/978-3-031-34586-9_10} }- Antonio Cobo
Elena Villalba-Mora
Rodrigo Pérez-Rodríguez
Juan E. Medina
Paula Robles-Mateos
Ángel Rodríguez-Laso
Leocadio Rodríguez-Mañas
Year: 2023
Assessing Older Adult’s Gait Speed with Wearable Accelerometers in Community Settings: Validity and Reliability Study
PERVASIVEHEALTH
Springer
DOI: 10.1007/978-3-031-34586-9_10
Abstract
We present the preliminary results of a validity and reliability study of two different state-of-the-art algorithms to estimate the gait speed of older adults in free-living conditions from the data collected by the ActiveUP wearable device. We described the ActiveUP wearable sensor together with its integration in a smart environment for frail and pre-frail older adults via an edge-computing architecture. A cross-sectional observational study was conducted. A sample of 18 people, 77.89 (6.47) y.o. 11 women, was recruited and their movement signals were recorded during short (2.4 m) and long (6 m) walking bouts. Validity, agreement, and reliability were assessed with Pearson’s correlation coefficient, with SEM and Bland-Altman limits of agreement (LOA), and with an ICC(A, 1) model of the intra-class correlation coefficient, respectively. Validity and reliability seemed to be good. However, the small size of our sample results in broad confidence intervals for the estimators. The agreement seems not to be good enough to trigger therapeutic responses. More studies are necessary to test whether the threshold for clinical tests is applicable under free-living conditions.
