Accuracy and concurrent validity of a sensor-based analysis of sit-to-stand movements in older adults

doi:10.1016/j.gaitpost.2016.02.004

Gait & Posture

Volume 45, March 2016, Pages 198-203

https://doi.org/10.1016/j.gaitpost.2016.02.004 Get rights and content

Highlights

•
We studied the accuracy and concurrent validity of sensor-based sit-to-stand measures.
•
In general the accuracy of sensor-based sit-to-stand measures was limited.
•
Concurrent validity ranged from moderate to very high.
•
The sensor method cannot replace laboratory methods for the analysis of sit-to-stand.
•
The sensor-based analysis of sit-to-stand may be relevant for clinical assessments.

Abstract

Body-fixed motion sensors have been applied for the assessment of sit-to-stand (STS) performance. However, the accuracy and concurrent validity of sensor-based estimations of the body's center of mass (CoM) motion during STS are unclear. Therefore, this study investigated the accuracy and concurrent validity of sensor-based measures of CoM motion during STS in older adults. Accuracy and concurrent validity were investigated by comparing the sensor-based method to a force plate method. Twenty-seven older adults (20 females, 7 males; age: 72–94 years) performed five STS movements while data were collected with force plates and motion sensors on the hip and chest. Hip maximal acceleration provided an accurate estimation of the center of mass (CoM) maximal acceleration (limits of agreement (LOA) smaller than 5% of the CoM maximal acceleration; estimated and real CoM maximal acceleration did not differ (p = 0.823)). Other hip STS measures and the chest STS measures did not provide accurate estimations of CoM motion (LOA ranged from −155.6% to 333.3% of the CoM value; sensor-based measures overestimated CoM motion (range p: <0.001 to 0.01)). However, the hip sensor did not overestimate maximal jerk of the CoM (p = 0.679). Moderate to very strong associations were observed between sensor-based estimations and actual CoM motion (range r = 0.64–0.94, p < 0.001). Hence, sensor-based estimations of CoM motion during STS are possible, but accuracy is limited. The sensor-based method cannot replace laboratory methods for a mechanical analysis of CoM motion during STS but it may be a practical alternative for the clinical assessment of STS performance in older persons.

Introduction

Leg muscle power is a determinant of movement execution and an important parameter for measuring intervention effects in older adults [1], [2], [3], [4]. However, available methods for the measurement of leg muscle power, such as force plates and isokinetic dynamometers, have practical disadvantages (e.g. costs, difficult to transport, uneasy to use) that limit the application of these methods in clinical settings. Therefore, it is important that practical methods are developed for the measurement of leg muscle power in older adults.

Zijlstra et al. developed an alternative method for the measurement of leg muscle power based on small body-fixed motion sensors [5]. Results indicated fair to excellent concurrent validity of sensor-based estimations of the body's center of mass (CoM) peak power during sit-to-stand (STS) based on a comparison with force plate measurements. In addition, results showed that a sensor on the hip provided more accurate estimations of vertical CoM acceleration during STS than a sensor on the chest, which overestimated CoM accelerations and peak powers [5].

However, Zijlstra et al. [5] had a limited number of older subjects and only evaluated the accuracy and concurrent validity of sensor-based estimation of CoM peak power during STS, not the accuracy and concurrent validity of other sensor-based measures of CoM motion during STS. Recent studies developed sensor-based measures of CoM motion during STS in addition to peak power [6], [7]. However, the accuracy and concurrent validity of these additional sensor-based measures of CoM motion during STS (e.g. maximal vertical velocity, maximal vertical acceleration) are unclear. Accuracy and concurrent validity indicate respectively the closeness of agreement and the association of a measured value with an accepted reference value. When accuracy and concurrent validity are adequate, the sensor-based method can be used for a mechanical analysis of CoM motion during STS instead of laboratory-based stationary methods (e.g. force plates). Therefore, the aim of this study was to investigate the accuracy and concurrent validity of sensor-based measures of CoM motion during STS in older adults. For this purpose, we compared the sensor-based method to a standard laboratory method consisting of force plates under the chair and feet of the participants. Based on the findings of Zijlstra et al. [5], we hypothesized that hip STS measures have adequate concurrent validity and accuracy, and that chest STS measures have adequate concurrent validity but overestimate CoM kinematics resulting in inadequate accuracy.

Section snippets

Participants

Participants were recruited from a health care center, a residential care home and sheltered houses. Older adults could participate in this study when they were able to rise from a chair in one attempt without using the hands, walk at least 10 m (with or without a cane or wheeled walker), and when they were at least 70 years of age. Participants were excluded when they had any cognitive, neurological, cardiovascular or respiratory disorder, lower extremity orthopaedic surgery or a stroke within

Results

Missing samples were observed in the chest sensor data of three participants. Therefore these data were excluded from further analysis.

Discussion

This study investigated the accuracy and concurrent validity of sensor-based measures of CoM motion during STS in older adults by comparing the sensor-based method to a standard laboratory method consisting of force plates. Results showed that only hip maximal acceleration has adequate accuracy in older adults. The other hip STS measures and chest STS measures have inadequate accuracy and overestimated CoM motion. Hip maximal jerk showed inadequate accuracy, however, this measure did not show a

Conflict of interest statement

All authors made substantial contributions to the conception and design of the study, data acquisition, data analysis, data interpretation, drafting the article or revising it critically for important intellectual content. Each of the authors has read and concurs with the content in the final manuscript. The authors have no conflicts of interest.

Acknowledgments

This study was financially supported by a grant from The Netherlands Organisation for Health Research and Development (ZonMw; program ‘Diseasemanagement chronische ziekten’; project number 40-00812-98-09014). The sponsor was not involved in the study design, in the collection, analysis and interpretation of data, in the writing of the manuscript, and in the decision to submit the manuscript for publication.

References (16)

W. Zijlstra et al.
A body-fixed-sensor-based analysis of power during sit-to-stand movements
Gait Posture
(2010)
G.R.H. Regterschot et al.
Sensitivity of sensor-based sit-to-stand peak power to the effects of training leg strength, leg power and balance in older adults
Gait Posture
(2014)
G.R.H. Regterschot et al.
Test–retest reliability of sensor-based sit-to-stand measures in young and older adults
Gait Posture
(2014)
A.L. Hof
Scaling gait data to body size
Gait Posture
(1996)
G.R.H. Regterschot et al.
Sensor-based monitoring of sit-to-stand performance is indicative of objective and self-reported aspects of functional status in older adults
Gait Posture
(2015)
A. Zech et al.
Functional muscle power testing in young, middle-aged, and community-dwelling nonfrail and prefrail older adults
Arch Phys Med Rehabil
(2011)
J.F. Bean et al.
The relationship between leg power and physical performance in mobility-limited older people
J Am Geriatr Soc
(2002)
J.F. Bean et al.
A comparison of leg power and leg strength within the InCHIANTI study: which influences mobility more
J Gerontol A: Biol Sci Med Sci
(2003)

There are more references available in the full text version of this article.

Cited by (17)

Sit-to-stand muscle power test: Comparison between estimated and force plate-derived mechanical power and their association with physical function in older adults
2021, Experimental Gerontology
Citation Excerpt :
However, assessing mechanical power in older people is not always an easy task. For example, several investigations have evaluated mechanical power during the sit-to-stand (STS) task with using force platforms (Lindemann et al., 2003; Regterschot et al., 2016; Chorin et al., 2016; Lindemann et al., 2007; Alvarez Barbosa et al., 2016), linear position transducers (LPT) (Alvarez Barbosa et al., 2016; Glenn et al., 2015; Fleming et al., 1991; Glenn et al., 2017; Lindemann et al., 2015; Lindemann et al., 2016) or 3D accelerometers (Regterschot et al., 2016; Millor et al., 2020). Unfortunately, these instruments can be expensive and may need periodic calibrations, technical support or offline data analysis, which can prevent clinicians and other health professionals from assessing mechanical power (Alcazar et al., 2018a).
This study aimed i) to assess the assumptions made in the sit-to-stand (STS) muscle power test [body mass accelerated during the ascending phase (90% of total body mass), leg length (50% of total body height) and concentric phase (50% of total STS time)], ii) to compare force plate-derived (FPD) STS power values with those derived from the STS muscle power test; and iii) to analyze the relationships of both measurements with physical function.
Fifty community-dwelling older adults (71.3 ± 4.4 years) participated in the present investigation. FPD STS power was calculated as the product of measured force (force platform) and velocity [difference between leg length (DXA scan) and chair height, divided by time (obtained from FPD data and video analysis)], and compared to estimated STS power using the STS muscle power test. Physical function was assessed by the timed-up-and-go (TUG) velocity, habitual gait speed (HGS) and maximal gait speed (MGS). Paired t-tests, Bland-Altman plots and regressions analyses were conducted.
Body mass accelerated during the STS phase was 85.1 ± 3.8% (p < 0.05; compared to assumed 90%), leg length was 50.7 ± 1.3% of body height (p < 0.05; compared to 50%), and measured concentric time was 50.3 ± 4.6% of one STS repetition (p > 0.05; compared to assumed 50%). There were no significant differences between FPD and estimated STS power values (mean difference [95% CI] = 6.4 W [−68.5 to 81.6 W]; p = 0.251). Both FPD and estimated relative (i.e. normalized to body mass) STS power were significantly related to each other (r = 0.95 and ICC = 0.95; p < 0.05) and to MGS and TUG velocity after adjusting for age and sex (p < 0.05).
Estimated STS power was not different from FPD STS power and both measures were strongly related to each other and to maximal physical performance.
Validity of various portable devices to measure sit-to-stand velocity and power in older adults
2020, Gait and Posture
Citation Excerpt :
Furthermore, unlike the LPT and App, STS velocity obtained by the IMU was not highly related to 30s chair STS performance (r = 0.58). Several studies have assessed STS power in older adults using different techniques, including force platforms [18,24], LPTs [9,25,26] accelerometers [27], or a combination of mass, height, and manually-recorded time [28,29]. Studies have also employed contrasting STS protocols, such as by allowing the use of armrests [26] and estimating single STS power based on a 5-repetition STS test [28].
Movement velocity and power in a single STS are related to functional performance in older adults. Identifying accessible tools that provide valid measures of STS velocity/power would allow practitioners to evaluate physical function in clinical settings where time, space and finances are limited.
Does a linear position transducer (LPT), iPhone application (App), and inertial measurement unit (IMU) obtain valid measurements of velocity and power during a single STS compared with 3D motion capture?
Twenty-seven community-dwelling older adults aged ≥60 years completed a single STS test with mean velocity and power simultaneously measured with 3D motion capture, an LPT, IMU and App. Acceptable validity was established if the Pearson correlation coefficient (r) was very high (≥0.7) and bias as a standardised effect size (ES) was small (<0.6). The relationship between STS velocity/power and 30s chair STS performance was also evaluated.
Measures of STS velocity obtained by the LPT (r = 0.94, ES = −0.21) and App (r = 0.89, ES = −0.19) were very highly valid when compared to 3D motion capture, and were very strongly related to 30s STS performance (r ≥0.74). The LPT (r = 0.87, ES = 0.13) and App (r = 0.74, ES = −0.12) also showed very high correlations and negligible bias for measuring STS power. Data collected by the IMU failed to meet our pre-determined threshold of acceptable validity for STS velocity (r = 0.72, ES = 1.00) or power (r = 0.61, ES = 0.34).
The LPT and iPhone App, but not the IMU, are valid tools for measuring STS velocity and power in community-dwelling older adults. Clinicians can use STS velocity obtained by either the LPT or App as a simple and valid proxy for functional status, which could help identify patients at high-risk of incident disability.
The sit-to-stand muscle power test: An easy, inexpensive and portable procedure to assess muscle power in older people
2018, Experimental Gerontology
Citation Excerpt :
The sit-to-stand (STS) test (Csuka and McCarty, 1985) is an easy, rapid, and commonly used functional performance measure that involves measuring the time taken to stand from a seated position a certain number of times or recording the number of repetitions undertaken in a given period, with low space, material and time requirements. In addition, several studies have evaluated STS muscle power by the utilization of a force platform (Alvarez Barbosa et al., 2016; Chen et al., 2012; Cheng et al., 2014; Drey et al., 2012; Fleming et al., 1991; Lacroix et al., 2015; Lindemann et al., 2003; Lindemann et al., 2007; Regterschot et al., 2016; Zech et al., 2012; Zech et al., 2011), a linear position transducer (Alvarez Barbosa et al., 2016; Glenn et al., 2015; Glenn et al., 2017a; Glenn et al., 2017b; Glenn et al., 2016; Gray et al., 2016; Gray and Paulson, 2014; Kato et al., 2015) or a 3D accelerometer (Regterschot et al., 2016; Zijlstra et al., 2010). However, these procedures present the economic and technical limitations mentioned above for their applicability in large studies or in the clinical setting.
Skeletal muscle power has been demonstrated to be a stronger predictor of functional limitations than any other physical capability. However, no validated alternatives exist to the usually expensive instruments and/or time-consuming methods to evaluate muscle power in older populations. Our aim was to validate an easily applicable procedure to assess muscle power in large cohort studies and the clinical setting and to assess its association with other age-related outcomes.
Forty community dwelling older adults (70–87 years) and 1804 older subjects (67–101 years) participating in the Toledo Study for Healthy Aging were included in this investigation. Sit-to-stand (STS) velocity and muscle power were calculated using the subject's body mass and height, chair height and the time needed to complete five STS repetitions, and compared with those obtained in the leg press exercise using a linear position transducer. In addition, STS performance, physical (gait speed) and cognitive function, sarcopenia (skeletal muscle index (SMI)) and health-related quality of life (HRQoL) were recorded to assess the association with the STS muscle power values.
No significant differences were found between STS velocity and power values and those obtained from the leg press force-velocity measurements (mean difference ± 95% CI = 0.02 ± 0.05 m·s⁻¹ and 6.9 ± 29.8 W, respectively) (both p > 0.05). STS muscle power was strongly associated with maximal muscle power registered in the leg press exercise (r = 0.72; p < 0.001). In addition, cognitive function and SMI, and physical function, were better associated with absolute and relative STS muscle power, respectively, than STS time values after adjusting by different covariates. In contrast, STS time was slightly more associated with HRQoL than STS muscle power measures.
The STS muscle power test proved to be a valid, and in general, a more clinically relevant tool to assess functional trajectory in older people compared to traditional STS time values. The low time, space and material requirements of the STS muscle power test, make this test an excellent choice for its application in large cohort studies and the clinical setting.
Relationship between Acceleration in a Sit-To-Stand Movement and Physical Function in Older Adults
2023, Geriatrics (Switzerland)
Temporal, Kinematic and Kinetic Variables Derived from a Wearable 3D Inertial Sensor to Estimate Muscle Power during the 5 Sit to Stand Test in Older Individuals: A Validation Study
2023, Sensors
Analyzing Intra-Cycle Velocity Profile and Trunk Inclination during Wheelchair Racing Propulsion
2023, Sensors

View all citing articles on Scopus

View full text

Accuracy and concurrent validity of a sensor-based analysis of sit-to-stand movements in older adults

Highlights

Abstract

Introduction

Section snippets

Participants

Results

Discussion

Conflict of interest statement

Acknowledgments

Gait Posture

Gait Posture

Gait Posture

Gait Posture

Gait Posture

Arch Phys Med Rehabil

The relationship between leg power and physical performance in mobility-limited older people

J Am Geriatr Soc

A comparison of leg power and leg strength within the InCHIANTI study: which influences mobility more

J Gerontol A: Biol Sci Med Sci