S. Ji1, H.-W. Jung1, S. Yoon2, H. Roh2, M. Kim3, H. Jung4, R. Jang5, H. Ha6, J.Y. Baek1, I.-Y. Jang1, E. Lee1
1. Division of Geriatrics, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea; 2. Dyphi Research Institute, Dyphi Inc., Daejeon, Republic of Korea; 3. Department of Biomedical Science and Technology, College of Medicine, East-West Medical Research Institute, Kyung Hee University, Seoul, Republic of Korea; 4. Department of Biomedical Science and Technology, Graduate School, Kyung Hee University, Seoul, Republic of Korea; 5. Coreline Soft, Seoul, Republic of Korea; 6. University of Ulsan College of Medicine, Seoul, Republic of Korea.
Corresponding Author: Hee-Won Jung, MD, PhD, Division of Geriatrics, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, 88, Olympic-ro 43-gil, Songpa-gu, Seoul 05505, Republic of Korea, Tel: +82-2-3010-1852, Fax: +82-2-3010-1852, Email: dr.ecsta@gmail.com
J Frailty Aging 2024;in press
Published online May 28, 2024, http://dx.doi.org/10.14283/jfa.2024.48
Abstract
BACKGROUND: There is currently no standardized protocol for the measurement of gait speed in older adults, particularly regarding the choice between static versus dynamic start.
OBJECTIVES: This study aimed to assess the impact of removing the initial phase on gait speed measurement and compare the correlation of different measurement methods with other physical performance metrics.
DESIGN: A cross-sectional study.
SETTING: A geriatric clinic in a tertiary hospital in Seoul, Korea.
PARTICIPANTS: Adults aged 65 years or older who underwent physical performance examinations during comprehensive geriatric assessments (n = 511).
MEASUREMENTS: A one-dimensional light detection and ranging device was used to obtain real-time gait signal data and measure the participants’ gait speed. Descriptive statistics were obtained for the acceleration phase. Various lengths of the initial phase were removed to determine the point at which gait speed plateaued. Correlations between four-meter gait speeds, with different initial phase lengths, and chair stand and balance test results were examined.
RESULTS: The mean ± standard deviation of the acceleration phase (m) was 0.92 ± 0.51. The removal of various initial phase lengths showed that 2 m gait speed based on dynamic start (0.5 m) significantly differed from static start (0.7 m/s versus 1.05 m/s, p<0.001). Gait speed showed an increase with the removal of longer initial phases but plateaued after removing 1 meter. This change lacked clinical significance after removing 0.5 meters. Dynamic start gait speed exhibited superior discriminative ability in chair stand and balance tests compared to static start gait speed.
CONCLUSION: Static start underestimates gait speed, while dynamic start aligns better with other physical performance results. An acceleration phase of 0.5 to 1 meter appears sufficient, but further studies are needed to optimize gait measurement methods.
Key words: Gait speed, acceleration phase, light detection and ranging, physical performance.
Introduction
Mobility loss is a major component of human aging, which considerably impacts the quality of life of older adults (1). Measurement of an individual’s gait speed, a simple physical performance parameter, is considered an important aspect of evaluating the mobility and overall health of older adults (2, 3). Typically, gait speed is measured by an examiner with a stopwatch over a range of 4 m to 10 m (4-6). It has been associated with future survival and functional outcomes in older adults (7, 8), making it a functional vital sign and a reliable measure for assessing overall health (9). Furthermore, it is widely considered an essential examination in assessments of physical frailty, sarcopenia, and the risk of falls (10-12).
Despite its significant clinical implications, there are heterogeneities and controversies in defining optimal measurement protocols for capturing gait speed in older adults, particularly concerning the choice between static start versus dynamic start. Static start methods include the initial acceleration in the calculation of gait speed (13), while dynamic start methods commonly exclude the initial 1–3 m of the distance walked (14). A prior systematic review on gait speed in geriatric assessments showed that among studies specifying static or moving starts, 22 used a static start, while 23 studies used a dynamic start (2, 15, 16). Concerns have been raised about including the acceleration phase in gait speed measurements, with previous studies showing that static starts remarkably underestimate gait speed (9, 17-22). The effect of the initial acceleration profoundly affects the calculated gait speed of frail older adults (23).
In most previous studies using the classical method of measuring gait speed with a stopwatch, making a direct comparison of static start versus dynamic start is challenging. One exception was a study that utilized gyroscopes and a 20 m walking distance, but this is longer than the typical examination for physical performance assessment in geriatric clinics (4 to 10 m) (4-6, 23). Furthermore, previous studies demonstrated that gait speed assessed via static and dynamic methods might yield different results, but the clinical relevance of this finding was not thoroughly investigated. With advances in measuring gait speed, light detection and ranging (LiDAR), a remote sensing method that uses laser beams to detect and measure distances, has enabled the precise measurement of not only gait speed, but also measurements of gait parameters, including the acceleration phase length (24, 25).
In the present study, we aimed to retrospectively investigate the pattern of initial acceleration in older adults undergoing gait speed assessments with LiDAR technology in a real-world geriatric clinic setting. As depicted in Figure 1, we compared various methods of calculating gait speed by removing the initial 0 to 1.5 m. Our objectives were to compare the static start and dynamic start methods and to determine the optimal length of the acceleration phase.
Figure 1. Layout of the measurement and the location of the sensor for the study
LiDAR, light detection and ranging.
Methods
Study design and population
This was a retrospective cross-sectional study reviewing the real-world records of examinations at a geriatric clinic within Asan Medical Center, a tertiary medical center in Seoul, Korea, with approximately 2700 beds. A Comprehensive Geriatric Assessment (CGA) is routinely performed in this clinic, typically for pre-operation risk evaluations and other indications. As part of the CGA process, the Short Physical Performance Battery (SPPB) is administered using the eSPPB kit (Dyphi Inc., Daejeon, Korea) to evaluate physical performance. This study included older adults aged 65 years or older who underwent physical performance examinations as part of the CGA from January 2022 to August 2022. The records of 635 patients were collected for analysis. Individuals with a life expectancy of less than one year, advanced frailty preventing walking with or without walking aids, or advanced cognitive dysfunction impairing their ability to understand and follow instructions were excluded from the study, as SPPB is not typically conducted in such patients.
Ethical considerations
This study was approved by the Institutional Review Board of Asan Medical Center, with the requirement for informed consent waived (IRB No. 2021-0519). The decision to waive informed consent was made due to several factors, including the low anticipated risks associated with participation and the retrospective nature of the study. This study was conducted in accordance with the ethical principles of the Declaration of Helsinki. The patients’ personal and health information was kept confidential, and all data were collected and handled in a manner that ensured their privacy.
Measurement of gait speed
Gait speed was measured by a one-dimensional LiDAR device in the eSPPB kit (26). The LiDAR was placed at the end of the test range to measure the longitudinal distance between the sensor and the trunk of the patients with a sampling rate of 50 Hz. The software was designed to record the datapoints of distance from the beginning of the whole test protocol. The starting point of examination was 7 m from the LiDAR sensor, and the usual gait speed calculation was performed using data points placed 6 m to 2 m from the LiDAR sensor (Figure 1). During these measurements, patients were instructed to walk at their usual pace within the test area, allowing for real-world conditions such as the use of canes and shoes. The signal data were filtered using a Savitzky–Golay filter.
In the present study, we used the stored raw data to draw time–distance, time–velocity, and distance–velocity curves (Supplementary material Figure S1). Various methods were employed to remove the initial phase for calculating gait speed from each individual’s data. Initially, we determined when individuals reached their maximal gait speed while minimizing the effect of deceleration by calculating their two-meter gait speeds. Subsequently, we computed the four-meter gait speed to derive the SPPB Gait and compared this with other physical performance metrics.
Definition of the initiation and completion of the acceleration and the measured distance of acceleration
As mentioned above, we defined the usual gait speed as the gait speed from 6 m to 2 m from the LiDAR. Further, we defined the initiation of the gait as the time taken for the individuals to reach 5% of their usual gait speed. We excluded 124 records in which the exact point of initiation of walking could not be defined. The software was designed to capture raw signals from 7 m, but some patients may have already started moving beyond that point, causing the initial acceleration phase to be left out of the recorded data. Hence, a total of 511 records were included in our analysis. We defined the completion of acceleration as the time taken for individuals to reach 95% of their usual gait speed. Finally, we defined the acceleration phase as the distance between the initiation (5%) and completion (95%) of the acceleration.
Other physical performance assessments
Other physical performance metrics of the SPPB were also measured using the eSPPB kit. The eSPPB kit is a sensor-based automatic measurement tool with previously published cutoff points (27), generating scores for balance (ranging from 0 to 4), gait speed (ranging from 0 to 4), chair stand (ranging from 0 to 4), and a total SPPB score (ranging from 0 to 12) (26).
In this study, we calculated the gait speed score using various methods for removing the initial phase. Specifically, we implemented cutoff points based on the time required to walk four meters distance: assigning 4 points for times less than 4.82 seconds, 3 points for times ranging from 4.82 to 6.20 seconds, 2 points for times between 6.21 and 8.70 seconds, and 1 point for times exceeding 8.7 seconds. Specifically, the scores were categorized as follows: SPPB Gait Static (gait speed score without removing the initial phase), SPPB Gait 0.5 m (gait speed score with the initial 0.5 meters removed), SPPB Gait 1 m (gait speed score with the initial 1 meter removed), and SPPB Gait 1.5 m (gait speed score with the initial 1.5 meters removed). To assess the alignment of each SPPB Gait score with other SPPB metrics, we combined the balance score and chair stand score (SPPB Balance + Chair Stand).
Data interpretation and analysis
Descriptive statistics were used to summarize the demographic and physical performance characteristics of the study population. The distribution of the distance of the acceleration phase was described as a histogram. To determine the relationship of the acceleration phase with age, usual gait speed, and physical performance, we conducted Pearson’s correlation test. Furthermore, we calculated the 2 m gait speed by excluding the different lengths of the initial phase, and independent t-tests were performed to statistically compare the mean gait speed according to the various lengths of the acceleration phase. To ascertain the clinical significance of any observed differences, we defined 0.1 m/s as the Minimal Clinically Important Difference (MCID) of gait speed (28).
Furthermore, linear regression analysis was conducted to model a linear relationship between each SPPB Gait score (ranging from 0 m to 1.5 m) as an independent variable and SPPB Balance + Chair Stand as the dependent variable. The coefficient of determination (R squared) was calculated for each model to evaluate the fitness of the regression. The analysis was performed using Python 3 and R software. Two-sided p-values of <0.05 were considered statistically significant.
Results
Baseline demographics, physical performance, and gait pattern of the study participants
As described in Table 1, among the 511 patients included in the analysis, 285 (55.8%) were female and 226 (44.2%) were male. Mean ± standard deviation of the age of the participants was 75.3 ± 7.5, while that of the SPPB score was 10.1 ± 2.5. The mean ± standard deviation of the usual gait speed (m/s) was 1.01 ± 0.30, while that of the length of the acceleration phase (m) was 0.92 ± 0.51. The distribution of the length of the acceleration phase is shown in Supplementary material Figure S2. We defined the individuals with an acceleration phase of more than 1 m as the long acceleration group for the following analysis.
Table 1. Baseline characteristics
Data are mean ± standard deviation. SPPB, Short Physical Performance Battery
When comparing individuals included (n = 511) and excluded (n = 124) due to technical limitations in capturing the acceleration initiation and completion, we found that those excluded exhibited significantly greater frailty, as indicated by lower SPPB scores and slower gait speeds, without significant differences in terms of age or gender (Supplementary material Table S1).
The length of the acceleration phase was negatively associated with age (Supplementary material Figure S3(A), correlation coefficient, -0.15; p<0.001) but positively associated with usual gait speed (Supplementary material Figure S3(B), correlation coefficient, 0.33; p<0.001) and with the SPPB score (Supplementary material Figure S3(C), correlation coefficient, 0.34; p<0.001).
Gait speed measurement according to the methods to remove the initial phase
We calculated the two-meter gait speed after excluding various lengths of the initial phase and compared the results. As depicted in Figure 2 and detailed in Table 2, we observed an increase in measured gait speed as the length of the excluded initial phase increased. Specifically, the measured gait speed for a two-meter length reached saturation once the excluded initial phase reached a length of 1 meter. Notably, the measured gait speed with an excluded length of 0.5 meters was statistically different from the measured gait speed using the static method (1.01 ± 0.34 vs. 0.71 ± 0.21, p < 0.001), with a difference of 0.3 m/s, exceeding the MCID of 0.1 m/s. However, after removing 0.5 meters, the difference in measured gait speed fell below the MCID. Additionally, in the long acceleration group, a longer excluded initial phase yield a maximal gait speed. After removing 1 meter in this subgroup, the difference in measured gait speed fell below the MCID (Supplementary material Figure S4, Table S2).
Table 2. Relationship between the removed initial phase and the measured gait speed for 2 m (m/s)
* Independent t-test p-values with 0.5 m as the reference; Ref, reference
Figure 2. Relationship between the removed initial phase and the measured gait speed for 2 m
Note: Mean and standard deviation of the measured gait speed.
Alignment with SPPB Balance + Chair Stand
We calculated SPPB Gait using each method of the four-meter gait speed with previously published cut-off values (the distribution of the four-meter gait speed is described in Supplementary material Figure S5) (27). As depicted in Figure 3, SPPB Gait 0.5 m, 1 m, and 1.5 m were more aligned with SPPB Balance + Chair Stand (R2 = 0.52, R2 = 0.53, and R2 = 0.53, respectively) than SPPB Gait Static (R2 = 0.49). The static start method tends to underestimate SPPB Gait. These results were consistent across the entire population and the long acceleration group (Supplementary Figure S6).
Figure 3. Association of SPPB Gait calculated based on each method with SPPB Balance + Chair Stand
Note 1: Mean and standard deviation of the SPPB Balance + Chair Stand with linear regression plots. Note 2: R-squared values for each SPPB Gait method; SPPB Gait Static, R2 = 0.49, SPPB Gait 0.5 m, R2 = 0.52, SPPB Gait 1 m, R2 = 0.53, SPPB Gait 1.5 m R2 = 0.53
Discussion
We calculated the length of the acceleration phase and found the distribution among the patients in a real-world geriatric clinic in a tertiary hospital. The mean ± standard deviation of the length of the acceleration phase (m) was 0.92 ± 0.51. Our observations revealed that as longer initial phases were removed, the two-meter gait speed increased, reaching a plateau after the removal of 1 meter. Notably, the difference between gait speeds measured by static methods (without removing the initial phase) and dynamic methods (removing the initial 0.5 meters) was 0.30 m/s, which exceeded the MCID. However, after removing 0.5 meters, the difference fell below the MCID. SPPB Gait scores derived from dynamic start methods are more aligned with other physical performance parameters than static start methods. These findings suggest that employing a dynamic start method may more accurately capture gait speed that is reflective of an individual’s physical performance. Furthermore, removing the initial distances of 0.5 to 1 meter emerged as reasonable alternatives.
We advocate for the use of dynamic methods over static methods in measuring gait speed for several reasons. First, our results, consistent with previous research, indicate static methods underestimate the gait speed (17-21). The difference between the static and dynamic methods of removing the initial 0.5 meters was 0.30 m/s (0.71 vs. 0.01, p<0.001), which is clinically significant. Second, dynamic methods, by removing the acceleration phase known for its greater variability, address a problem in clinical practice, as suggested by prior research.(23, 29) Third, as evidenced by our findings, dynamic methods display a stronger correlation with other physical performance parameters, such as chair stand and balance tests, implying that gait speed measured by dynamic methods better reflects an individual’s physical performance compared to static measurements.
Determining the appropriate distance for removal may present practical challenges, but our results suggest that an initial distance of 0.5 to 1 meter (equivalent to 1 to 2 strides) might be reasonable. This conclusion is supported by several observations in our study. First, as demonstrated in Figure 2 and Table 2, gait speed reached a plateau after 1 meter of removal. Additionally, after removing 0.5 meters, the difference in gait speed fell below the minimum clinically important difference (MCID) threshold of 0.1 m/s. As demonstrated in Supplementary material Figure S4 and Table S2, even among the long acceleration group, removing 1 meter resulted in differences below the MCID threshold compared with the plateau. Additionally, Figure 3 and Supplementary material Figure S6 suggest that removing an initial phase of at least 0.5 meters may suffice in terms of alignment with the SPPB Balance + Chair Stand scores, further supporting this approach. However, we acknowledge the limitations of our study. Further research is needed to optimize and standardize the measurement protocol to ensure robust and reliable assessments of gait speed.
There was a previous study that calculated distance to achieve a steady state of gait speed in frail older adults and arrived at somewhat different conclusions (23). This previous research differs in methodology from our study in several aspects. First, their gait analysis involved a 20 meters walk, whereas ours covered approximately 7 meters. A 20 meters gait analysis and 7 meters gait analysis would likely yield different results for frail older adults, particularly in terms of gait speed and the acceleration phase. Second, their study utilized gyroscopes at each shank and thigh, analyzing data based on units of stride, whereas our study used LiDAR and the analysis was grounded on the distance from the sensor. Third, the participants in the previous study were frailer and older than those in our study; the mean age of their population was 83.1 years compared to 75.3 years in our study, and 49% of the population required a walking aid. The mean distance of the acceleration phase was 1.43 m and the mean gait speed was 0.66 m/s, and the researchers recommended excluding the initial 2.5 meters. Further studies involving different populations and different sensors are warranted.
To the best of our knowledge, in terms of the technical aspect, this is the first study to determine the minimum required acceleration phase using real-time gait signal data. We used data from the LiDAR sensor of the eSPPB kit, which has been previously validated to assess the physical performance of older adults (24, 26). As a LiDAR can measure distance data with high time resolution (50 Hz in our study), we were able to calculate the time–distance and time–velocity data (Supplementary material Figure S1). From this analysis, we were able to separate the acceleration phase from the steady state of the gait speed, as well as to manually examine the pattern of the acceleration and steady-state walking.
Nevertheless, our study has several limitations. First, our study has limited generalizability, as we only included a single tertiary geriatric clinic. As mentioned above, adaptation of this method to non-ambulatory frail older adults might not be appropriate. Second, 123 participants were excluded as the initiation of the movement was difficult to define using only the raw data, due to the measurement algorithm of the sensor. We defined the initiation of movement as the time at which individuals reached 5% of their usual gait speed, with the recording protocol omitting data points farther than 7 meters from the sensor. As detailed in Supplementary material Table S1, the excluded individuals were frailer, although there was no significant difference in terms of age or gender. As the single dimension LiDAR beam tends to miss objects at a greater distance, we believe that further research using a 2D LiDAR with appropriate spatiotemporal resolution for gait speed recording would alleviate this limitation. Third, data on underlying diseases, such as movement disorders, and other anthropometric measurements were unavailable, due to limited availability with waived informed consent. Consequently, we could not examine the effect of these variables on the acceleration phase. Fourth, to get steady-state gait speed (Results 3.2.), we used a four-meter distance instead of the standard four-meter distance to exclude deceleration, as our participants walked a total of 7 meters. However, our comparison with other physical performance measurements (Results 3.3.) utilized the more commonly used four-meter gait speed measurement method. Finally, this study is a retrospective cross-sectional analysis, which is limited in its ability to demonstrate the clinical utility, such as prognostic and discrimination power, of each method. Consequently, the findings have restricted applicability in clinical practice. To standardize the measurement protocol, a prospective study with a more rigorous design is recommended.
Conclusion
In summary, our study determined the mean length of the acceleration phase among patients in a real-world geriatric clinic at a tertiary medical center to be 0.92 meters. Static start methods were found to underestimate gait speed, while dynamic start methods exhibited better alignment with other physical performance measures, suggesting their superiority in assessing physical performance. An acceleration phase removal of 0.5 to 1 meter emerged as a reasonable approach, although further refinement of gait measurement methods is warranted.
Author’s contributions: Sunghwan Ji: Software, Validation, Formal analysis, Writing – Original Draft, Visualization, Hee-Won Jung: Conceptualization, Methodology, Writing – Review & Editing, Supervision, Funding acquisition, Seoungjun Yoon: Validation, Investigation, Data Curation, Writing – Original Draft, Hyunchul Roh: Investigation, Data Curation, Writing – Original Draft, Miji Kim: Validation, Project administration, Writing – Review & Editing, Heeeun Jung: Validation, Writing – Original Draft, Ryoungwoo Jang: Software, Resources, Formal analysis, Hyeonkyu Ha: Validation, Writing – Review & Editing, Ji Yeon Baek: Validation, Project administration, Writing – Review & Editing, Il-Young Jang: Project administration, Supervision, Writing – Review & Editing, Eunju Lee: Project administration, Supervision, Writing – Review & Editing.
Acknowledgement/funding statement: This research was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number : HR20C0026), and the «Korea National Institute of Health»(KNIH) research project (project number: 2023ER080701)
Conflict of Interest disclosure: Hee-Won Jung, Hyunchul Roh, and Seongjun Yoon are co-founders of Dyphi Inc., a startup focused on sensor technologies for human movement and robotics. All other authors declare no potential conflicts of interest.
References
1. Shafrin J, Sullivan J, Goldman DP, Gill TM. The association between observed mobility and quality of life in the near elderly. PLoS One. 2017;12:e0182920. doi:10.1371/journal.pone.0182920.
2. Peel NM, Kuys SS, Klein K. Gait Speed as a Measure in Geriatric Assessment in Clinical Settings: A Systematic Review. The Journals of Gerontology: Series A. 2012;68:39-46. doi:10.1093/gerona/gls174.
3. Studenski S. Bradypedia: is gait speed ready for clinical use? J Nutr Health Aging. 2009;13:878-880. doi:10.1007/s12603-009-0245-0.
4. Yanishi M, Tsukaguchi H, Kimura Y, et al. Evaluation of physical activity in sarcopenic conditions of kidney transplantation recipients. Int Urol Nephrol. 2017;49:1779-1784. doi:10.1007/s11255-017-1661-4.
5. Dent E, Morley JE, Cruz-Jentoft AJ, et al. International Clinical Practice Guidelines for Sarcopenia (ICFSR): Screening, Diagnosis and Management. J Nutr Health Aging. 2018;22:1148-1161. doi:10.1007/s12603-018-1139-9.
6. Studenski SA, Peters KW, Alley DE, et al. The FNIH sarcopenia project: rationale, study description, conference recommendations, and final estimates. J Gerontol A Biol Sci Med Sci. 2014;69:547-558. doi:10.1093/gerona/glu010.
7. Studenski S, Perera S, Patel K, et al. Gait speed and survival in older adults. Jama. 2011;305:50-58. doi:10.1001/jama.2010.1923.
8. Dommershuijsen LJ, Isik BM, Darweesh SKL, et al. Unraveling the Association Between Gait and Mortality—One Step at a Time. The Journals of Gerontology: Series A. 2019;75:1184-1190. doi:10.1093/gerona/glz282.
9. Middleton A, Fritz SL, Lusardi M. Walking speed: the functional vital sign. J Aging Phys Act. 2015;23:314-322. doi:10.1123/japa.2013-0236.
10. Cruz-Jentoft AJ, Bahat G, Bauer J, et al. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing. 2019;48:16-31. doi:10.1093/ageing/afy169.
11. Chen LK, Woo J, Assantachai P, et al. Asian Working Group for Sarcopenia: 2019 Consensus Update on Sarcopenia Diagnosis and Treatment. J Am Med Dir Assoc. 2020;21:300-307.e302. doi:10.1016/j.jamda.2019.12.012.
12. Adam CE, Fitzpatrick AL, Leary CS, et al. The Association between Gait Speed and Falls in Community Dwelling Older Adults with and without Mild Cognitive Impairment. Int J Environ Res Public Health. 2021;18. doi:10.3390/ijerph18073712.
13. Lou N, Chi CH, Chen XD, et al. Sarcopenia in overweight and obese patients is a predictive factor for postoperative complication in gastric cancer: A prospective study. Eur J Surg Oncol. 2017;43:188-195. doi:10.1016/j.ejso.2016.09.006.
14. Yuki A, Ando F, Otsuka R, Shimokata H. Sarcopenia based on the Asian Working Group for Sarcopenia criteria and all-cause mortality risk in older Japanese adults. Geriatr Gerontol Int. 2017;17:1642-1647. doi:10.1111/ggi.12946.
15. Kim S, Kim M, Won CW. Validation of the Korean Version of the SARC-F Questionnaire to Assess Sarcopenia: Korean Frailty and Aging Cohort Study. J Am Med Dir Assoc. 2018;19:40-45.e41. doi:10.1016/j.jamda.2017.07.006.
16. Fukuoka Y, Narita T, Fujita H, et al. Importance of physical evaluation using skeletal muscle mass index and body fat percentage to prevent sarcopenia in elderly Japanese diabetes patients. J Diabetes Investig. 2019;10:322-330. doi:10.1111/jdi.12908.
17. Macfarlane PA, Looney MA. Walkway length determination for steady state walking in young and older adults. Res Q Exerc Sport. 2008;79:261-267. doi:10.1080/02701367.2008.10599489.
18. Stuck AK, Bachmann M, Füllemann P, Josephson KR, Stuck AE. Effect of testing procedures on gait speed measurement: A systematic review. PLoS One. 2020;15:e0234200. doi:10.1371/journal.pone.0234200.
19. Phan-Ba R, Calay P, Grodent P, et al. A corrected version of the Timed-25 Foot Walk Test with a dynamic start to capture the maximum ambulation speed in multiple sclerosis patients. NeuroRehabilitation. 2012;30:261-266. doi:10.3233/nre-2012-0754.
20. Kim M, Won CW. Combinations of gait speed testing protocols (automatic vs manual timer, dynamic vs static start) can significantly influence the prevalence of slowness: Results from the Korean Frailty and Aging Cohort Study. Arch Gerontol Geriatr. 2019;81:215-221. doi:10.1016/j.archger.2018.12.009.
21. Krumpoch S, Lindemann U, Rappl A, et al. The effect of different test protocols and walking distances on gait speed in older persons. Aging Clinical and Experimental Research. 2021;33:141-146. doi:10.1007/s40520-020-01703-z.
22. Graham JE, Ostir GV, Kuo YF, Fisher SR, Ottenbacher KJ. Relationship between test methodology and mean velocity in timed walk tests: a review. Arch Phys Med Rehabil. 2008;89:865-872. doi:10.1016/j.apmr.2007.11.029.
23. Lindemann U, Najafi B, Zijlstra W, et al. Distance to achieve steady state walking speed in frail elderly persons. Gait Posture. 2008;27:91-96. doi:10.1016/j.gaitpost.2007.02.005.
24. Yoon S, Jung HW, Jung H, et al. Development and Validation of 2D-LiDAR-Based Gait Analysis Instrument and Algorithm. Sensors (Basel). 2021;21. doi:10.3390/s21020414.
25. Jung HW, Roh HC, Kim SW, et al. Cross-Comparisons of Gait Speeds by Automatic Sensors and a Stopwatch to Provide Converting Formula Between Measuring Modalities. Ann Geriatr Med Res. 2019;23:71-76. doi:10.4235/agmr.19.0016.
26. Jung HW, Roh H, Cho Y, et al. Validation of a Multi-Sensor-Based Kiosk for Short Physical Performance Battery. J Am Geriatr Soc. 2019;67:2605-2609. doi:10.1111/jgs.16135.
27. Guralnik JM, Simonsick EM, Ferrucci L, et al. A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol. 1994;49:M85-94. doi:10.1093/geronj/49.2.m85.
28. Bohannon RW, Glenney SS. Minimal clinically important difference for change in comfortable gait speed of adults with pathology: a systematic review. J Eval Clin Pract. 2014;20:295-300. doi:10.1111/jep.12158.
29. van Herk IE, Arendzen JH, Rispens P. Ten-metre walk, with or without a turn? Clin Rehabil. 1998;12:30-35. doi:10.1191/026921598667081596.
© Serdi 2024