Walking speed is an important aspect of gait and is commonly used as an objective measure of functional mobility in both clinical and research settings. Its importance lies not only in its implications for community ambulation but also because of its relationship to various health outcomes. Walking speed has been shown to be a key factor in determining rehabilitation needs1,2 and discharge location3 and has the potential to predict future functional decline4,5 and fall risk.6,7 Furthermore, a decline in walking speed is associated with several health-related factors such as disability, hospitalization, loss of independence, and mortality.8–11 Improvement in walking speed has been linked to constructive changes in quality of life12 and walking behavior.13 This importance, combined with its ease of use and objectivity, substantiates the use of walking speed as a practical clinical measure that offers more insight into an individual's overall functional capacity.
Walking speed can be quickly and easily assessed in most clinical and research settings, and measurements of walking speed have demonstrated good reliability across multiple patient populations and in individuals with known gait impairments.14–16 Great variation exists, however, in measurement methods used to assess walking speed. There is little consensus concerning optimal testing parameters such as starting protocol, pace, and timed walking distance. Often tests are chosen based more on tester preference and convenience, especially in clinical settings where space is limited.
The 10-Meter Walk Test is a commonly used measure for assessing walking speed.17–22 It requires a 20-m path that includes 5 m for acceleration and deceleration. Practically, however, a full 20-m walkway is not always available, so there are several shorter distances commonly used to assess walking speed including 3-, 4-, and 6-m assessments.23 Two studies with neurologic populations found significant differences in fast walking speed between 5- and 10-m test distances,24 and in self-selected walking speed between 10 m and “real-life environments” (eg, parking lot of a shopping center).25 Overall, however, research examining how subtle differences in test walking distance affect gait speed reliability and validity is limited both within and across patient populations. It is unclear whether shorter test walking distances provide as accurate, representative assessments of walking speed as longer distances.
The primary purpose of this study was to examine the validity of a 4-Meter Walk Test compared with the commonly used 10-Meter Walk Test when used to assess walking speed in healthy, older adults. An additional objective was to examine the reliability and validity of walking speed measurements obtained through 2 different methods: the use of a stopwatch and automatic timers. If reliable, representative assessments of gait speed can be achieved using shorter distances with a handheld timer, this might increase the use of walking speed measurements in clinical settings, thereby offering more insight into an individual's functional health status and imparting clinically meaning information to help guide and monitor patient treatment. Although reliability assessments across longer time intervals is more applicable for clinical outcomes, intrasession reliability is an important first step to determine reliability of a measurement without external influences (eg, time) and is appropriate for 1-time assessments such as when using walking speed as a vital sign.26
Design and Participants
A cross-sectional study design was used with comparisons of gait speed between 2 different walking tests. Forty-three participants were recruited from a local retirement community. Inclusion criteria included age 65 years or older, the ability to reliably follow 2-step instructions, and the ability to walk 20 m with or without an assistive device. Exclusion criteria included severe weight-bearing pain (rated >5/10 on the visual analog pain scale), severe visual impairment, and/or severe arthritis or orthopedic problems that limited ambulation ability. All participants gave written informed consent. The study was approved by the institutional review board at Palmetto Health.
Walking Procedure and Measurement
Walking speed was assessed at participants' self-selected walking pace using a 4-Meter Walk Test (with 2 m provided for acceleration/deceleration) and the 10-Meter Walk Test (with 5 m provided for acceleration/deceleration) (Figure 1). Distances were provided at the beginning and end of the timed walkway to allow participants space to accelerate/decelerate outside the data collection area to help reduce gait variability introduced during these phases.23,27 Each participant completed 3 consecutive trials for each walking test, for a total of 6 walking trials. Order of administration of the 2 different walking conditions was randomly varied among participants so that not all participants were performing the longer walking test at the end of the testing session. Participants were instructed to “walk at your comfortable, usual pace” until they reached the end of the marked path. Wireless timers (Brower Timing Systems) that send radio transmissions were used to record walking time. These timers were placed at the beginning and end of the timed walkway area and automatically started/stopped as the participant walked past them. In addition, a member of the research team simultaneously measured walking time with a stopwatch, starting the stopwatch as soon as the participant's lead leg (or assistive device) crossed the first marker and stopping it when the participant's lead leg (or assistive device) crossed the second marker. The same person performed all stopwatch measurements to prevent introducing interrater variability. Participants were provided rest breaks as needed throughout the testing session.
The Shapiro-Wilks test was used to test for normality of data. Intraclass correlation coefficients (ICC3,1) were calculated to examine the reliability of gait speed measurements across walking trials 2 and 3 for the 4- and 10-Meter Walk Tests for both stopwatch and automatic timer assessments. The ICC values were interpreted using the benchmarks suggested by Menz et al28: more than 0.75 excellent reliability; 0.40 to 0.75 fair to good reliability; and less than 0.40 poor reliability. To quantify the amount of change in gait speed that must be observed to be considered to exceed measurement error and variability, the minimal detectable change (MDC) was calculated at both the 90% and 95% confidence levels using the following formulas: MDC90 = 1.64 × SEM × √2 and MDC95 = 1.96 × SEM × √2, where SEM is standard error of measurement. The SEM value was determined using the formula [SD × √(1 − r)], where r is the test-retest reliability coefficient (in this case, ICC3,1) and SD is the standard deviation of the trial difference scores.29,30 The SEM is the estimated standard deviation of measurement error, or the difference between the observed values and the true values. The SEM was multiplied by 1.64 or 1.96 to reflect the 90% or 95% confidence intervals (CIs), respectively. This value was multiplied by the square root of 2 to account for the error associated with repeat measurements.29,31
The agreement between stopwatch and automatic timer assessments (for both the 4- and 10-Meter Walk Tests) and between 4- and 10-m gait speed assessments was examined using ICC2,1 and the Bland-Altman method32 (95% limits of agreement), with similar interpretation of ICC point estimates as previously described. Validity was examined for both single trial (second walking trial) and average (across all 3 walking trials) gait speed assessments. The Bland-Altman technique allows one to visually assess the agreement between 4- and 10-m gait speed assessments (or between stopwatch and automatic timer assessments) by plotting the difference in the measurement methods against the mean of the 2 measurements.32,33 The resulting plot shows the size and range of the measurement differences and their distribution around the mean. The 95% limits of agreement (mean difference ±1.96 SD of the differences between measurement methods) provide an indication of how far apart measurements by the 2 walking tests (or 2 timing methods) are likely to be for most individuals.34 A smaller range between these 2 limits indicates a better level of agreement, and how close the measurements have to be is a clinical question/decision (is the discrepancy between methods large enough to meaningfully affect the interpretation of results?), not based on statistical testing. When comparing the 2 timing methods, single-trial assessments of walking speed were used in the Bland-Altman analyses. When comparing 4- and 10-m gait speed assessments, stopwatch assessments of walking speed were used for the Bland-Altman analyses, as this from of measurement has more clinical utility than automatic timers. In addition, paired t tests were performed to test for systematic differences in gait speed between the 2 walk tests, with α < .05. All statistical analyses were conducted using PASW version 18.0 (SPSS, Chicago, IL).
Forty-three community-dwelling older adults (32 women, 11 men) with a mean age 84.3 years (SD = 6.9) participated in the study. Average walking speed (across all 3 trials) on the 10-Meter Walk Test varied between 0.50 and 1.43 m/s, with a mean walking speed of 0.96 m/s (SD = 0.23) per stopwatch assessment. Seven participants used an assistive device for ambulation. One participant was unable to complete a third ambulation trial due to personal time constraints, so reliability analyses were performed and walking speed/time was averaged across the first 2 trials for this participant. The Shapiro-Wilks test showed that gait speed measurements were normally distributed.
Reliability Across Walking Trials
Both 4- and 10-m gait speed measurements were shown to have excellent test-retest reliability, with ICC values ranging from 0.96 to 0.98 (Table). Reliability was similar for both stopwatch and automatic timer assessments, with SEM values between 0.004 and 0.008 m/s and MDC90/MDC95 values between 0.01 and 0.02 m/s.
Validity of Stopwatch Measurements Compared With Automatic Timer
Agreement between the 2 timing methods was excellent for both walking tests, with ICC values ranging from 0.99 (95% CI: 0.988–0.996) to 1.00 (95% CI: 0.999–1.00). The values for ICC were similar for both single-trial and average gait speed assessments. Figure 2 shows a Bland-Altman plot for the differences in gait speed between the 2 timing methods. While no obvious relationship between the difference and mean was observed for stopwatch and automatic timer assessments, there was a slightly better level of agreement between the 2 timing methods on the 10-Meter Walk Test (95% limits of agreement ranged from −0.02 to 0.02 m/s) than on the 4-Meter Walk Test (95% limits of agreement ranged from −0.05 to 0.05 m/s).
Validity of 4- and 10-m Stopwatch Assessments of Gait Speed
Gait speed measurements were not significantly different between 4- and 10-m walk assessments for either single-trial (P = .957) or average (P = .349) gait speed comparisons. When comparing the second ambulation trial, both the 4- and 10-Meter Walk Test resulted in a mean gait speed value of 0.97 m/s (SD = 0.22). When examining average gait speed across the 2 tests, the 10-Meter Walk Test resulted in a mean gait speed value of 0.96 m/s (SD = 0.23) compared with 0.95 m/s (SD = 0.22) for the 4-Meter Walk Test.
The ICC value for single-trial gait speed measurements between the 4- and 10-Meter Walk Tests was 0.93 (95% CI: 0.87–0.96) and for average gait speed measurements was 0.93 (95% CI: 0.88–0.96). Figure 3 shows a Bland-Altman plot for the differences in gait speed between the 2 walking tests. No obvious relationship between the difference and the mean was observed for 4- and 10-m gait speed assessments, with similar mean differences and 95% limits of agreement noted for both single-trial (mean difference: −0.0007 m/s; 95% limits of agreement: −0.17 to 0.17 m/s) and average (mean difference: −0.0118 m/s; 95% limits of agreement: −0.17 to 0.15 m/s) gait speed comparisons.
The use of automatic timers to record walking time is simple, but few clinical settings have such devices. Stopwatches, on the contrary, are a more accessible instrument and often used in both clinical and research settings to record walking time for calculations of gait speed across various distances. Our results indicate excellent agreement both within and between stopwatch and automatic timer assessments across 2 different walking distances (4 and 10 m) in healthy, older adults, with little difference in SEM values between the 2 timing methods. Our ICC values were on the upper end of similar studies (ICC values from 0.88 to 0.97) that have examined the reliability of gait speed measurements in this patient population.15,35,36 Furthermore, the Bland-Altman analysis displayed a small range between the 95% limits of agreement (±0.05 m/s or less), indicating a clinically acceptable degree of agreement such that the use of one timing method over the other would not meaningfully affect interpretation of gait speed results.
If walking speed is to be used as a vital sign in health care assessments, measurement methods associated with a small SEM value are important to ensure a small degree of measurement error when assessing baseline values. Both 4- and 10-m gait speed assessments had excellent test-retest reliability, with similar SEM values when examined across consecutive walking trials. For example, if an individual exhibited a gait speed of 0.98 m/s on the 10-Meter Walk Test, our results indicate that we could be 95% confident that this individual's true gait speed is between 0.97 and 0.99 m/s, or 2 times the SEM (0.005 m/s for stopwatch assessments); similarly, if the 4-Meter Walk Test was used as a screening tool, our results indicate that we could be 95% confident that this individual's true gait speed is between 0.96 and 1.00 m/s.
Another extrapolation is to look at gait speed changes over time (eg, from pre- to posttreatment). Our results indicate that a change in gait speed of 0.01 m/s or more or 0.02 m/s or more is necessary for 10- and 4-m walk assessments, respectively, to be 95% confident that a true change has occurred beyond measurement error in healthy, older adults. As several participants required an assistive device to walk, future research work could examine measurement reliability and determine MDC values specific to this subpopulation of individuals. Furthermore, MDC values calculated from measurements taken across longer time intervals, such as days or weeks, could be in a manner different from those that are calculated from repeat measurements taken within the same testing session.
Examining the validity of a shorter walk test compared with the commonly used 10-Meter Walk Test when determining gait speed is important, as space is often a limiting factor in clinical settings. While the ICC value quantifies the reliability of 2 methods, it alone is insufficient to evaluate patterns of discrepancy that may be present among differences in the data. The Bland-Altman method aids in the determination of whether 2 methods of clinical measurement agree sufficiently for them to be used interchangeably, or one in place of the other.32,33 The 2 methods may be used interchangeably if the calculated 95% limits of agreement are close enough, per clinical decision, such that a difference between measurement methods as extreme as described by the limits of agreement would not meaningfully affect interpretation of results.32,34 The ICC values for our data (0.93) indicated excellent agreement between 4- and 10-m walking speed assessments, as mean gait speed values differed by just −0.0007 m/s (SD = 0.09) and −0.0118 m/s (SD = 0.08) for single-trial and average gait speed assessments, respectively, with slightly higher gait speed values obtained with the 10-Meter Walk Test. While the mean difference in gait speed between the 2 measures was small, the range of the upper and lower 95% limits of agreement was ±0.15 to ±0.17 m/s. Several studies have demonstrated 0.08 to 0.14 m/s as the needed change for a meaningful improvement in walking speed in older adults.9,17,37 In addition, improvements in walking speed of 0.1 m/s or more have been shown to be a useful predictor for well-being whereas decreases in walking speed of the same amount have been linked with poorer health outcomes.26 Given these values of meaningful change in gait speed, the calculated limits of agreement in our study indicate that the degree of agreement between 4- and 10-m gait speed assessments is not sufficient to permit using the 2 walking tests interchangeably in assessments of gait speed in healthy, older adults. The discrepancy between measurement methods is large enough to potentially mask meaningful changes in gait speed over time if both methods are used. Therefore, although the reliability of both walking tests is excellent, the 4-Meter Walk Test does not exhibit a high enough degree of concurrent validity with the 10-Meter Walk Test to be used interchangeably for gait speed assessments in healthy, older adults.
This study is one of the few to investigate how subtle differences in test walking distance affect the validity of walking speed assessments. We did not attempt, however, to investigate the reliability and validity of different walking speed assessments in specific patient populations, which has been the focus of previous research.24,25 Our results indicate that although reliable assessments of walking speed in healthy, older adults can be obtained using a 4-Meter Walk Test, 4-m walking speed assessments cannot be used interchangeably with 10-Meter Walk Test assessments. Consideration must be taken when viewing our results, however, as they may have been influenced by a number of limitations present in this study. First, our study sample was relatively small, and it is possible that a lesser discrepancy between measurement methods might have been observed with a larger sample size. Furthermore, our sample included both individuals who ambulated with and without an assistive device, which improves external validity but increases variability in sample characteristics. Future studies with a larger number of individuals who use an assistive device could examine measurement reliability and validity in this subpopulation. In addition, the stability of gait speed measurements over time was not assessed in this study; reliability and validity analyses among gait speed measurements based on walking trials performed on separate days may result in different ICC and MDC values than the current study. Intrasession reliability needs to be assessed first, however, to decrease possible variations from external influences, followed by reliability evaluations across longer time intervals. Also, the participants in our study were all relatively healthy, older adults with a mean self-selected walking speed of 0.96 m/s (SD = 0.23), which is similar to 8,20 or somewhat lower15,36,38 than gait speed values obtained from other studies examining walking speed in older adults. These results might be different for older adults who ambulate at slower or faster walking speeds and/or who have specific pathologies (eg, stroke). Finally, although the differences in gait speed measurements between the 4- and 10-Meter Walk Tests greatly exceeded the MDC values of the current study (0.01–0.02 m/s), we considered MDC values from other studies to strengthen our conclusions; however, the amount of change that is considered meaningful by other studies may not really be meaningful if these values did not exceed measurement error and variability.
Although 4- and 10-m walking speed assessments in healthy, older adults demonstrated excellent test-retest reliability and were highly correlated, our results indicate that there is insufficient agreement between the 2 walking tests to permit them to be used interchangeably in this patient population. We therefore recommend the use of the 10-Meter Walk Test to obtain the most valid clinical assessment of walking speed in healthy, older adults; however, a 4-Meter Walk Test can be used if space is a limiting factor, but the same walking test needs to be used for all subsequent measurements of gait speed for evaluations of meaningful change in gait speed over time. Our results also demonstrated that handheld stopwatches were as reliable as automatic timers in measurements of gait speed. Further research should continue to examine how subtle differences in walking test parameters affect walking speed assessments across different patient populations. If a shorter walking test can provide a high, clinically acceptable degree of agreement of gait speed measures compared with the 10-Meter Walk Test, this might increase the use of walking speed measurements in clinical settings to offer more insight into a patient's functional mobility and health status.
1. Purser JL, Weinberger M, Cohen HJ, et al. Walking speed predicts health status and hospital costs for frail elderly male veterans. J Rehabil Res Dev. 2005;42(4):535–546.
2. Goldie PA, Matyas TA, Evans OM. Deficit and change in gait velocity during rehabilitation after stroke. Arch Phys Med Rehabil. 1996;77(10):1074–1082.
3. Rabadi MH, Blau A. Admission ambulation velocity predicts length of stay and discharge disposition following stroke in an acute rehabilitation hospital. Neurorehabil Neural Repair. 2005;19(1):20–26.
4. Montero-Odasso M, Schapira M, Soriano ER, et al. Gait velocity as a single predictor of adverse events in healthy seniors aged 75 years and older. J Gerontol A Biol Sci Med Sci. 2005;60(10):1304–1309.
5. Brach JS, VanSwearingen JM, Newman AB, Kriska AM. Identifying early decline of physical function in community-dwelling older women: performance-based and self-report measures. Phys Ther. 2002;82(4):320–328.
6. Maki BE. Gait changes in older adults: predictors of falls or indicators of fear. J Am Geriatr Soc. 1997;45(3):313–320.
7. de Rekeneire N, Visser M, Peila R, et al. Is a fall just a fall: correlates of falling in healthy older persons. The Health, Aging and Body Composition Study. J Am Geriatr Soc. 2003;51(6):841–846.
8. Studenski S, Perera S, Wallace D, et al. Physical performance measures in the clinical setting. J Am Geriatr Soc. 2003;51(3):314–322.
9. Hardy SE, Perera S, Roumani YF, Chandler JM, Studenski SA. Improvement in usual gait speed predicts better survival in older adults. J Am Geriatr Soc. 2007;55(11):1727–1734.
10. Kuo HK, Leveille SG, Yen CJ, et al. Exploring how peak leg power and usual gait speed are linked to late-life disability: data from the National Health and Nutrition Examination Survey (NHANES), 1999–2002. Am J Phys Med Rehabil. 2006;85(8):650–658.
11. Penninx BW, Ferrucci L, Leveille SG, Rantanen T, Pahor M, Guralnik JM. Lower extremity performance in nondisabled older persons as a predictor of subsequent hospitalization. J Gerontol A Biol Sci Med Sci. 2000;55(11):M691–M697.
12. Schmid A, Duncan PW, Studenski S, et al. Improvements in speed-based gait classifications are meaningful. Stroke. 2007;38(7):2096–2100.
13. Bowden MG, Balasubramanian CK, Behrman AL, Kautz SA. Validation of a speed-based classification system using quantitative measures of walking performance poststroke. Neurorehabil Neural Repair. 2008;22(6):672–675.
14. Green J, Forster A, Young J. Reliability of gait speed measured by a timed walking test in patients one year after stroke. Clin Rehabil. 2002;16(3):306–314.
15. Steffen TM, Hacker TA, Mollinger L. Age- and gender-related test performance in community-dwelling elderly people: Six-Minute Walk Test, Berg Balance Scale, Timed Up & Go Test, and gait speeds. Phys Ther. 2002;82(2):128–137.
16. van Loo MA, Moseley AM, Bosman JM, de Bie RA, Hassett L. Test-re-test reliability of walking speed, step length and step width measurement after traumatic brain injury: a pilot study. Brain Inj. 2004;18(10):1041–1048.
17. Perera S, Mody SH, Woodman RC, Studenski SA. Meaningful change and responsiveness in common physical performance measures in older adults. J Am Geriatr Soc. 2006;54(5):743–749.
18. Ota A, Yasuda N, Horikawa S, Fujimura T, Ohara H. Differential effects of power rehabilitation on physical performance and higher-level functional capacity among community-dwelling older adults with a slight degree of frailty. J Epidemiol. 2007;17(2):61–67.
19. Avlund K, Rantanen T, Schroll M. Tiredness and subsequent disability in older adults: the role of walking limitations. J Gerontol A Biol Sci Med Sci. 2006;61(11):1201–1205.
20. Kressig RW, Wolf SL, Sattin RW, et al. Associations of demographic, functional, and behavioral characteristics with activity-related fear of falling among older adults transitioning to frailty. J Am Geriatr Soc. 2001;49(11):1456–1462.
21. van Hedel HJ, Dietz V, Curt A. Assessment of walking speed and distance in subjects with an incomplete spinal cord injury. Neurorehabil Neural Repair. 2007;21(4):295–301.
22. Jorgensen JR, Bech-Pedersen DT, Zeeman P, Sorensen J, Andersen LL, Schonberger M. Effect of intensive outpatient physical training on gait performance and cardiovascular health in people with hemiparesis after stroke. Phys Ther. 2010;90(4):527–537.
23. Graham JE, Ostir GV, Fisher SR, Ottenbacher KJ. Assessing walking speed in clinical research: a systematic review. J Eval Clin Pract. 2008;14(4):552–562.
24. Salbach NM, Mayo NE, Higgins J, Ahmed S, Finch LE, Richards CL. Responsiveness and predictability of gait speed and other disability measures in acute stroke. Arch Phys Med Rehabil. 2001;82(9):1204–1212.
25. Moseley AM, Lanzarone S, Bosman JM, et al. Ecological validity of walking speed assessment after traumatic brain injury: a pilot study. J Head Trauma Rehabil. 2004;19(4):341–348.
26. Fritz S, Lusardi M. White paper: “walking speed: the sixth vital sign.” J Geriatr Phys Ther. 2009;32(2):2–5.
27. Lindemann U, Najafi B, Zijlstra W, et al. Distance to achieve steady state walking speed in frail elderly persons. Gait Posture. 2008;27(1):91–96.
28. Menz HB, Latt MD, Tiedemann A, Mun San Kwan M, Lord SR. Reliability of the GAITRite walkway system for the quantification of temporo-spatial parameters of gait in young and older people. Gait Posture. 2004;20(1):20–25.
29. Haley SM, Fragala-Pinkham MA. Interpreting change scores of tests and measures used in physical therapy. Phys Ther. 2006;86(5):735–743.
30. Fritz SL, Blanton S, Uswatte G, Taub E, Wolf SL. Minimal detectable change scores for the Wolf Motor Function Test. Neurorehabil Neural Repair. 2009;237:662–667.
31. Wyrwich KW, Tierney WM, Wolinsky FD. Further evidence supporting an SEM-based criterion for identifying meaningful intra-individual changes in health-related quality of life. J Clin Epidemiol. 1999;52(9):861–873.
32. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1(8476):307–310.
33. Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res. 1999;8(2):135–160.
34. Bland JM, Altman DG. Comparing methods of measurement: why plotting difference against standard method is misleading. Lancet. 1995;346(8982):1085–1087.
35. Bohannon RW, Andrews AW, Thomas MW. Walking speed: reference values and correlates for older adults. J Orthop Sports Phys Ther. 1996;24(2):86–90.
36. Hollman JH, Childs KB, McNeil ML, Mueller AC, Quilter CM, Youdas JW. Number of strides required for reliable measurements of pace, rhythm and variability parameters of gait during normal and dual task walking in older individuals. Gait Posture. 2010;32(1):23–28.
37. Kwon S, Perera S, Pahor M, et al. What is a meaningful change in physical performance? Findings from a clinical trial in older adults (the LIFE-P study). J Nutr Health Aging. 2009;13(6):538–544.
38. Cesari M, Kritchevsky SB, Penninx BW, et al. Prognostic value of usual gait speed in well-functioning older people—results from the Health, Aging and Body Composition Study. J Am Geriatr Soc. 2005;53(10):1675–1680.