In the year 2008, one in 190 Americans were living with a loss of a limb, and it is estimated that by the year 2050, this will increase to a total of 3.6 million Americans (Ziegler-Graham et al.1). From 1988 to 1996, there was a mean of 133,735 hospital discharges for amputations per year, and during this time, dysvascular amputations increased 27% and accounted for 82% of limb loss.2 Although it has been reported that nontraumatic lower-limb amputations among the US adult diabetic population declined from 83,153 in 1996 to 70,139 in 2008, these amputations were still well more than the 52,868 that were performed in 1996.3 In addition, from 2000 to 2011, there were 1,795 major (foot, ankle, unilateral and bilateral transtibial and transfemoral) amputations in soldiers of the American Armed Forces.4
Unfortunately, persons with amputations have a higher risk for mortality from cardiovascular disease than their nonamputee peers.5,6 Limb loss, especially in the lower limbs, can hamper the mobility of an individual and ultimately lead to a greater sedentary lifestyle. After an amputation occurs, it is still very important for the amputee to maintain and increase physical activity levels7 to reduce the risk for cardiovascular diseases.
Specific training programs have been used to increase the fitness in persons with lower-limb amputations,8 thus demonstrating that these individuals can physiologically respond to physical activity as those with intact limbs. An increasing number of persons with amputations participating in sports9 indicate that many desire to remain active, which can also provide a viable means for these individuals to increase and/or maintain activity levels. Specific training programs and sports are not the only means for increasing and/or maintaining activity levels for health- and fitness-related purposes. Just as those with intact limbs, those with lower-limb amputations engage in leisure activity, such as walking, cycling, and swimming, in addition to activities such as housework and gardening.
It is just as important that persons with lower-limb amputations be able to monitor their own physical activity levels as any individual without amputations to reduce the risks of cardiovascular disease. Those with lower-limb amputations need physical activity to maintain and/or improve health just as those with intact limbs. If persons with lower-limb amputations maintain or increase a sedentary lifestyle, health problems will most likely be exacerbated.8 It is also important for persons with amputations to be able to objectively measure their physical activity levels because other more subjective methods, such as recall questionnaires, seem to have little accuracy.10 One method by which those with lower-limb amputations can objectively track their physical activity is with a pedometer.
Pedometers are inexpensive small devices that, when fitted to the individual as specified by the manufacturer, can measure the number of steps a person accumulates. Pedometers are an economical tool used to measure physical activity and have been shown to be valid not only in healthy individuals11–13 but also in older individuals,14 children,15,16 children with disabilities,17 and obese individuals.18 To date, there are only a few studies exploring the accuracy of pedometers in lower-limb amputees.19
If pedometers accurately measure activity levels in persons with lower-limb amputations, then this instrument can be a means by which persons with amputations, their physicians, and their caregivers can objectively prescribe, assess, and monitor physical activity. The objective of this study was to evaluate the accuracy of commercially available pedometers in persons with lower-limb amputations.
This study was approved by the Texas A&M University–San Antonio institutional review board, and all subjects signed a written informed consent before participating. All data were collected during a 2-day span at the Amputee Coalition of America 2010 National Conference in Irvine, CA, USA. Twenty-four male and 15 female individuals using lower-limb prostheses who were able to walk without the use of a cane, walker, or other assistive devices participated in the study. Most of the participants had transtibial amputations, with the second most common being transfemoral. Participant characteristics can be seen in Table 1.
A Seca S-214 portable height rod (Hanover, MD, USA) and a DR400C digital body weight scale (Webb City, MO, USA) were used to collect height and weight with shoes and prostheses. The participants were fitted with Velcro Walk4Life™ pedometer belt (Walk4Life Inc, Plainfield, IL, USA) around the waistline at the hip, then fitted with the following three pedometer models: a SW-701 Digiwalker™ (NEW-LIFESTYLES, Inc, Lee’s Summit, MO, USA) spring-levered pedometer on the anterior midline of the right hip, a New-Lifestyles® NL-800 (NEW-LIFESTYLES, Inc, Lee’s Summit, MO, USA) piezoelectric pedometer just laterally to the SW-701, and an Omron HJ-112 piezoelectric pedometer just laterally to the NEW-LIFESTYLES NL-800.
Each pedometer was fitted on the right side of the participant. Before data collection, pedometer batteries were replaced with new ones and then fitted on the investigator as described above, who walked 100 steps while counting to check for accuracy. If a pedometer registered greater than 3% error, which is suggested accuracy for laboratory-controlled conditions,20 it was not used. In each case, no pedometer registered less than 97 counts, and the same pedometers were used for every subject.
A straight 100-m path was marked using a measuring wheel along a level carpeted surface along an indoor hallway. A cone was placed at the end of the path to mark the turnaround point. After the pedometers were fitted on them, the participants were positioned at the start line, and all pedometers were reset to zero. The participants were instructed to walk to the end of the path, around the cone, and then back to the start line at a pace at which they would normally walk. Each step was tallied using a hand counter (model no. 77270; Lab Safety Supply Inc, Janesville, WI), and time was kept using a stopwatch (HW20; Extech Instruments, Waltham, MA, USA). At the end of the walk, the participants were asked to remain still while the steps, rating of perceived exertion (RPE; Borg 6–20 scale), and time were recorded.
All statistical tests were performed using Predictive Analytics Software for Windows (Chicago, IL, USA). Repeated-measures (RM) analyses of variance (ANOVAs) were used to determine differences between pedometer and actual counts (ACs) obtained while walking, with α set at 0.05. The Mauchly test of sphericity was used to test for the equality of variances of the differences between AC, the SW-701, the NL-800, and the Omron while walking. If significant, the Greenhouse-Geisser correction was used as the correct test of within-subjects effects. If the omnibus effect was significant, Bonferroni-adjusted value from the least significant difference (LSD) was used for pairwise comparisons among the AC, the SW-701, the NL-800, and the Omron while walking. α Value for these comparisons was set at 0.008 (0.05/6). Cohen d with the pooled SD was used to compute the effect size between the AC and pedometer counts.
Single measure intraclass correlation coefficient (ICC) from a two-way random effects ANOVA was used to assess the agreement between AC and pedometer counts, with 0.90 or greater considered high agreement; 0.80 to 0.89, moderate agreement; and 0.79 or lower, low agreement.21 Bland-Altman plots of AC versus counts registered by the pedometers were used to provide an indication of overrepresentation/underrepresentation of steps and agreement between the measures.22 Scores less than 0 indicate an overestimation by the pedometers and scores greater than 0 indicate an underestimation by the pedometers. These plots show the variability in pedometer scores while allowing for the mean difference score and the 95% limits of agreement to be shown. Error scores of 0 indicate that there are no differences between the actual steps taken and those registered by the pedometer. Percentage error was calculated as ([steps detected by pedometer − AC]/AC) × 100.
The Greenhouse-Geisser correction was used to determine significance for walking (F1.09,41.5 = 18.8, p = .001) because the Mauchly test statistic was significant (p = 0.001). Pairwise comparisons indicated that the SW-701 registered counts significantly lower than AC (p = 0.001, d = 0.95); however, there were no significant differences between the NL-800 and AC (p = 0.167, d = 0.20) and between the Omron and AC (p = 0.644, d = 0.05). Table 2 provides the mean and SD values for the AC and pedometer counts.
Among the three pedometers, only the Omron was in high agreement with AC, whereas the NL-800 and the SW-701 indicated a low agreement (Table 3).
Percentage error was greatest in the SW-701 (18.8%), followed by the NL-800 and Omron (2.4% and 0.69%, respectively). Mean error was also greatest in the SW-701, followed by the NL-800, and least in the Omron (Table 4). The Bland-Altman plots illustrate the mean error among the pedometers, and the tighter limit of agreement is clearly visible with the Omron compared with the SW-700 and the NL-800 (Figure 1).
Time to complete the walk was 181.2 (31.7) seconds (66.2 [23.6] m per minute), with an RPE rating of 9.9 (2.4), which falls between the “very light” and “light” description on the Borg 6 to 20 scale.
The purpose of this study was to determine the accuracy of three commercially available pedometers for persons with lower-limb amputations. The results indicate that, of the three pedometers tested, the Omron HJ-112 is the best pedometer for a person with an amputation wearing a prosthesis, the NL-800 is slightly less accurate, and the SW-701 is the least accurate. It has been suggested that 3% error in pedometer readings is acceptable in controlled studies, and no more than 10% error is acceptable in free-living conditions.20 The Omron and the NL each indicated less than 3% error, whereas the SW was well more than 18.8% in the current study.
Omron readings in this study were consistent with findings in previous research with individuals without amputations. For example, the Omron HJ-720 pedometer indicated 2% error in slower-than-normal, normal, and faster-than-normal walking speeds in 102 individuals aged 20 to 80 years who walked overground for 100 m.23 The Omron HJ-113 was also found to have less than 3% error in 42 men and women walking between 54 and 107 m per minute on the treadmill.24 In addition, the HJ-112 was found to be valid in 92 men and women aged 18 to 20 years who walked on a treadmill at 67.2, 80.4, and 93.6 m per minute.25 All of the abovementioned HJ pedometers use the same piezoelectric counting mechanism as the HJ-112, and because the results of the current study are comparable with those mentioned above, it is plausible to say that this pedometer can be used with confidence in those with lower-limb amputations who use a prosthesis for ambulatory purposes.
The Omron HJ-112 is generally less expensive than the NL-800 pedometer (approximately $25.00 US dollars compared with $50.00 US dollars, respectively); however, the extra cost for the NL-800 may not be advantageous for those with lower-limb amputations. Although both pedometers are piezoelectric, the NL-800 incurred more error than the HJ-112, albeit still under the 3% condition.
The NL-series pedometer, however, has been shown to be reliable and valid. For example, 28 male and female university students wore the NL-1000 for 1 day during their waking hours, and it was concluded that this pedometer was a viable instrument able to accurately record step counts in free-living conditions.26 The NL-2000 was also accurate in controlled trials after 21 older adults aged 65 to 87 years walked at five speeds between 40.2 and 93.6 m per minute on a treadmill.27 Whereas percentage error on the treadmill was more than 20% at the slow speed, percentage error was well lower than 3% at all other speeds in this sample. In overground walking at self-selected slow, normal, and fast-paced speeds in these same individuals, percentage error was also lower than 3%. The NL pedometers mentioned above also use the same counting mechanism as the NL-800, which is considered a piezoelectric pedometer that has no moving parts. On the other hand, the counting mechanism of the SW-701 uses a spring lever that contains moving parts, and this might attribute to the greater percentage error in the current study.
Generally, piezoelectric pedometers are more accurate at slower walking speeds compared with spring-loaded pedometers, such as the SW-701. For example, the SW-701 elicited a mean 10% error while walking on the treadmill at three different speeds, whereas the Omron HJ-112 averaged 1.2% error.25 The NL-2000 was more accurate at various walking speeds on the treadmill and overground trials compared with the SW-701, which had percentage error ranging from 7% to 42%.27 Furthermore, after 20 men and women walked on a treadmill at five speeds ranging from 27 to 107 m per minute, the NL-2000 was within 3% of actual steps at 67 m per minute, but the SW-701 did not obtain this accuracy until 80 m per minute.12 Although the SW-701 is generally less accurate than the NL-800 or HJ-112 pedometers, this indicates that these can be accurate at faster walking speeds of 80 m per minute or greater. Although the current study did not control for walking speed, the slower walking speeds of the participants (66.2 m per minute) might have contributed to greater error in the SW-701 pedometer readings. Only five participants in the current study walked at this speed or faster, and any recommendation for the use of the SW-701 in persons with lower-limb amputations should include this criterion.
To the authors’ knowledge, there are no studies exploring the accuracy of the HJ-112 and NL-800 pedometers and only one study using the SW-70 on persons with lower-limb amputations.19 Error in the SW-701 in the current study was greater than that which has been reported in previous studies. In 20 participants with transtibial amputations who walked 160 m, error was 6% compared with almost 19% in the current study.19 The greater error in the current study may in part be the result of a greater diversity of various levels of amputations (Table 1), thereby eliciting a greater diversity of gait disorders. Research has shown that gait disorders can lower SW-701 accuracy, as was demonstrated in nursing home and community-dwelling older adults.28 The gait of persons using prostheses in the current study most likely deviates from normal gait, and this deviation presumably contributes to the underlying cause of greater error in the SW-701 in addition to the slower walking speeds.
Studies have been conducted on persons with lower-limb amputations using activity monitors, which are accurate in free-living29 and controlled conditions.30,31 Unfortunately many of these devices are very expensive, ranging from $1,000 to $3,000 US dollars. The current study demonstrates acceptable error in two of the three tested pedometers, which are substantially less expensive and thus could be used to monitor activity in this population, whether it is for research or consumer purposes.
The findings of this study mimic those found in previous research with individuals without amputations. Those wearing prostheses may mimic the altered gait of the elderly, in which the piezoelectric pedometer has been shown to be valid for step counting.32 It can be seen in Figure 1 that only two subjects fell above and one subject fell below 2 SD, and had these three readings fallen within the 2 SD range, the NL-800 would have certainly rivaled the Omron. It can be speculated that these three readings may be due to placement on the hip, and if so, this pedometer may be more susceptible to placement location than the Omron.
The least accurate pedometer is the SW, which operates mechanically and is sensitive to tilt placement. Although this study did control for tilt by using the pedometer belt, error in readings is most likely caused by irregular gait. This pedometer is the least expensive of the three tested, but caution should be used for recommendations for its use because of the 18% error indicated.
Both the NL-800 and the Omron HJ-112, which are piezoelectric pedometers, seem to accurately record step counts in persons using lower-limb prostheses. These have been found to be more accurate than the less sensitive spring-loaded pedometers in other studies as well.33 Although the Omron may be considered the better pedometer according to the present investigation, it must be noted that the NL-800 can be considered accurate as well.
Fitness levels of persons using prostheses are much lower than those without amputations.34 The pedometer can be used as a motivational tool to increase physical activity of various age groups35,36 and thereby increase health in persons without amputations37 and should presumably do the same in persons using prostheses.
The authors thank the Amputee Coalition of America and the study participants for their support.
1. Ziegler-Graham K, MacKenzie EJ, Ephraim PL, et al. Estimating the prevalence of limb loss in the United States: 2005 to 2050. Arch Phys Med Rehabil 2008; 89: 422–429.
2. Dillingham TR, Pezzin LE, MacKenzie EJ. Limb amputation and limb deficiency: epidemiology and recent trends in the United States. South Med J 2002; 95: 875–883.
3. Li Y, Burrows NR, Gregg EW, et al. Declining rates of hospitalization for nontraumatic lower-extremity amputation in the diabetic population aged 40 years or older: U.S., 1988-2008. Diabetes Care 2012; 35: 273–277.
4. Armed Forces Health Surveillance Center. Amputations of upper and lower extremities, active and reserve components, U.S. Armed Forces 2000-2011. MSMR 2012; 19: 2–6.
5. Hrubec Z, Ryder R. Traumatic limb amputations and subsequent mortality from cardiovascular disease and other causes. J Chronic Dis 1979; 33: 239–250.
6. Modan M, Peles E, Halkan H, et al. Increased cardiovascular disease mortality rates in traumatic lower limb amputees. Am J Cardiol 1998; 82: 1242–1247.
7. Bussman J, Grootscholten E, Stam H. Daily physical activity and heart rate response in people with a unilateral transtibial amputation for vascular disease. Arch Phys Med Rehabil 2004; 85: 240–244.
8. Chin T, Sawamura S, Fujita H, et al. Effect of endurance training program based on anaerobic threshold (AT) for lower limb amputees. J Rehabil Res Dev 2001; 38: 7–11.
9. Bragaru M, Dekker R, Geertzen JH, Dijkstra PU. Amputees and sports: a systematic review. Sports Med 2011; 41: 721–740.
10. Stepien JM, Cavenett S, Taylor L, Crotty M. Activity levels among lower-limb amputees: self-report versus step activity monitor. Arch Phys Med Rehabil 2007; 88: 896–900.
11. Bassett DR Jr, Ainsworth BE, Swartz AM, et al. Validity of four motion sensors in measuring moderate intensity physical activity. Med Sci Sports Exerc 2000; 32: S471–S480.
12. Crouter SE, Schneider PL, Karabulut M, Bassett DR Jr. Validity of 10 electronic pedometers for measuring steps, distance, and energy cost. Med Sci Sports Exerc 2003; 35: 1455–1460.
13. Le Masurier GC, Tudor-Locke C. Comparison of pedometer and accelerometer accuracy under controlled conditions. Med Sci Sports Exerc 2003; 35: 867–871.
14. King WC, Belle SH, Brach JS, et al. Objective measures of neighborhood environment and physical activity in older women. Am J Prev Med 2005; 28: 461–469.
15. McNamara E, Hudson Z, Taylor S. Measuring activity levels of young people: the validity of pedometers. Br Med Bull 2010; 95: 121–137.
16. McMinn D, Rowe DA, Stark M, Nicol L. Validity of the New Lifestyles NL-1000 accelerometer for measuring time spent in moderate-to-vigorous physical activity in school settings. Meas Phys Educ Exerc Sci 2010; 14: 67–78.
17. Beets MW, Coombs C, Morgan M, et al. Accuracy of pedometer steps and time for youth with disabilities. Adapt Phys Activ Q 2007; 24: 228–244.
18. Swartz AM, Bassett DR Jr, Moore JB, et al. Effects of body mass index on the accuracy of an electronic pedometer. Int J Sports Med 2003; 24: 588–592.
19. Dudek K, Khan O, Lemaire E, et al. Ambulation monitoring of transtibial amputation subjects with patient activity monitor versus pedometer. J Rehabil Res Dev 2008; 45: 577–585.
20. Schnieder PL, Crouter SE, Basset DR Jr. Pedometer measures of free-living physical activity: comparison on 13 models. Med Sci Sports Exerc 2004; 36: 331–335.
21. Vincent WJ. Analysis of variance with repeated measures. In Statistics in Kinesiology. 3rd Ed. Champaign, IL: Human Kinetics; 2005: 183–200.
22. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 301–310.
23. Dondzila CJ, Swartz AM, Miller NE, et al. Accuracy of uploadable pedometers in laboratory, overground, and free-living conditions in young and older adults. Int J Behav Nutr Phys Act 2012; 9: 143. doi:10.1186/1479-5868-9-143.
24. Giannakidou KA, Ageloussis N, Fatouros I, et al. The validity of two Omron pedometers during treadmill walking is speed dependent. Eur J Appl Physiol 2012; 112: 49–57.
25. Hasson RE, Haller J, Pober DM, et al. Validity of the Omron HJ-112 pedometer during treadmill walking. Med Sci Sports Exerc 2009; 41: 805–809.
26. McClain JJ, Hart TL, Getz RS, Tudor-Locke C. Convergent validity of 3 low cost motion sensors with the ActiGraph accelerometer. J Phys Act Health 2010; 7: 662–670.
27. Grant PM, Dall PM, Mitchell SL, Granat MH. Activity-monitor accuracy in measuring step number and cadence in community-dwelling older adults. J Aging Phys Act 2008; 16: 201–214.
28. Cyarto EV, Myers A, Tudor-Locke C. Pedometer accuracy in nursing home and community-dwelling older adults. Med Sci Sports Exerc 2004; 36: 205–209.
29. Bussmann HB, Reuvekamp PJ, Veltink PH, et al. Validity and reliability of measurements obtained with an “activity monitor” in people with and without a transtibial amputation. Phys Ther 1998; 78: 989–998.
30. Coleman KL, Smith DG, Boone DA, et al. Step activity monitor: long-term, continuous recording of ambulatory function. J Rehabil Res Dev 1999; 36: 8–18.
31. Ramstrand N, Nilsson KA. Validation of a patient activity monitor to quantify ambulatory activity in an amputee population. Prosthet Orthot Int 2007; 31: 157–166.
32. Marsh A, Vance R, Frederick T, et al. Objective assessment of activity in older adults at risk for mobility disability. Med Sci Sports Exerc 2007; 39: 1020–1026.
33. Crouter SE, Schneider PL, Bassett DR Jr. Spring-levered versus piezo-electric pedometer accuracy in overweight and obese adults. Med Sci Sports Exerc 2005; 37: 1673–1679.
34. Chin T, Sawamura S, Fujita H, et al. Physical fitness of lower-limb amputees. Am J Phys Med Rehabil 2002; 81: 321–325.
35. Clemes SA, O’Connell S, Rogan LM, Griffiths PL. Evaluation of a commercially available pedometer used to promote physical activity as part of a national programme. Br J Sports Med 2010; 44: 1178–1183.
36. Snyder A, Colvin B, Gammack JK. Pedometer use increases daily steps and functional status in older adults. J Am Med Dir Assoc 2011; 12: 590–594.
37. Bravata DM, Smith-Spangler C, Sundaram V, et al. Using pedometers to increase physical activity and improve health: a systematic review. JAMA 2007; 298: 2296–2304.