An evaluation of the 30-s chair stand test in older adults: frailty detection based on kinematic parameters from a single inertial unit

Journal of NeuroEngineering and Rehabilitation An evaluation of the 30-s chair stand test in older adults: frailty detection based on kinematic parameters from a single inertial unit Nora Millor 1 2 Pablo Lecumberri 1 Marisol Gmez 1 Alicia Martnez-Ramrez 1 Mikel Izquierdo 0 0 Department of Health Sciences, Public University of Navarra , Pamplona , Spain 1 Department of Mathematics, Public University of Navarra , Pamplona , Spain 2 Research, Studies and Sport Medicine Centre, Government of Navarra , Pamplona , Spain Background: A growing interest in frailty syndrome exists because it is regarded as a major predictor of co-morbidities and mortality in older populations. Nevertheless, frailty assessment has been controversial, particularly when identifying this syndrome in a community setting. Performance tests such as the 30-second chair stand test (30-s CST) are a cornerstone for detecting early declines in functional independence. Additionally, recent advances in body-fixed sensors have enhanced the sensors' ability to automatically and accurately evaluate kinematic parameters related to a specific movement performance. The purpose of this study is to use this new technology to obtain kinematic parameters that can identify frailty in an aged population through the performance the 30-s CST. Methods: Eighteen adults with a mean age of 54 years, as well as sixteen pre-frail and thirteen frail patients with mean ages of 78 and 85 years, respectively, performed the 30-s CST while threir trunk movements were measured by a sensor-unit at vertebra L3. Sit-stand-sit cycles were determined using both acceleration and orientation information to detect failed attempts. Movement-related phases (i.e. impulse, stand-up, and sit-down) were differentiated based on seat off and seat on events. Finally, the kinematic parameters of the impulse, stand-up and sit-down phases were obtained to identify potential differences across the three frailty groups. Results: For the stand-up and sit-down phases, velocity peaks and modified impulse parameters clearly differentiated subjects with different frailty levels (p < 0.001). The trunk orientation range during the impulse phase was also able to classify a subject according to his frail syndrome (p < 0.001). Furthermore, these parameters derived from the inertial units (IUs) are sensitive enough to detect frailty differences not registered by the number of completed cycles which is the standard test outcome. Conclusions: This study shows that IUs can enhance the information gained from tests currently used in clinical practice, such as the 30-s CST. Parameters such as velocity peaks, impulse, and orientation range are able to differentiate between adults and older populations with different frailty levels. This study indicates that early frailty detection could be possible in clinical environments, and the subsequent interventions to correct these disabilities could be prescribed before further degradation occurs. Inertial units; Frailty syndrome; Kinematic parameters; 30-s chair stand test; Signal analysis - Background Frailty occurs often in people older than 65 years (ranging from 7 to 16.3%), and its prevalence increases with age [1-3]. Frail individuals are at particular risk for poor outcomes such as disability, fall, death and hospitalization from minor stressors [4-7]. The diagnosis of frailty is based on several health domains, including physical impairments (e.g., low gait velocity, fatigue and low grip strength), weight loss, and low physical activity [2]. Despite some vagueness in its definition, clinicians have indicated that early detection is one of the most effective methods for reducing the severity of physical frailty and for improving a patients well-being. Functional ability assessments aim to detect mobility impairments such as physical weakness so that early interventions are possible. The 30-s CST is one of the most important functional evaluation clinical tests because it measures lower body strength and relates it to the most demanding daily life activities (e.g., climbing stairs, getting out of a chair or bath tub or rising from a horizontal position) [8-10]. Low levels of body strength are the primary cause of both balance problems and falls in the elderly population [11,12]. The 30-s CST, similar to tests such as the 5-stands test and the timed up and go test (TUG), is able to differentiate between subjects with different functional levels. However, the 30-s CST is also able to assess the fatigue effect causeg by the number of sit-to-stand repetitions. Indeed, the 30-s CST has been widely used in many studies not only to evaluate functional fitness levels [12-14] but also to monitor training [15-18] and rehabilitation [19,20]. Classically, the 30-s CST consists of manually counting the number of sit-stand-sit cycles completed during the 30 seconds of the test. Since the early 1990s, IUs have been increasingly used to measure kinematic and kinetic parameters [21]. This technology is a non-invasive, portable and economical method to capture accelerations and angular velocities in three orthogonal planes [22]. However, signal analysis is needed to separate out the sit-to-stand (SitTS) and/or stand-to-sit (StandTS) transitions from the entire test duration. Recently, a wide range of studies have positively shown that IUs can furnish accurate kinematic transition-related measures, particularly when a test subject is standing up or sitting down, [21,23-26]. There is no gold standard yet, but this task has typically been achieved [24-26], through the use of thresholds on either the angular velocity [27,28] or the acceleration information [29,30]. However, threshold values are hard to generalize, as they are influenced by noise and by movement artifacts. Thus, peak detection techniques, such as those considered here, seem to perform better [31]. Other authors have preferred to obtain transition durations from the orientation signal of the trunk, which is the angle between the vertical axis and the anterior wall of the subjects thorax. In this paper, the sinus function is used to soften the signal and the time of postural transition are defined from the previous to the posterior maximum from a minimum point which is the transition indicator [32]. A major difficulty associated with transition detection is the fact that movement patterns depend on the subjects physical condition. Healthy subjects do not show the same transition indicator as frail subjects, and frail subjects may perform several attempts before completing a valid cycle [33]. Thus, this manuscript uses a novel technique to separate the sit-stand-sit cycles and their phases from the remainder of the signal. First, the vertical position signal is used to clearly differentiate the cycles, and then, transition events are detected using both acceleration and orientation signals to separate the phases, which include impulse, stand-up and sit-down. Vertical posit (...truncated)