Pierce, Sarah PT, DPT; Fergus, Andrea PT, PhD; Brady, Brooks PT, DPT, ATC; Wolff-Burke, Melissa PT, EdD
Children and adolescents with amputations face functional and psychosocial difficulties.1–3 Approximately two-thirds of pediatric amputations are congenital. The remaining one-third of pediatric amputations are acquired, with 75% a result of trauma and 25% secondary to disease including osteosarcoma.4 When treating a person with an amputation, valid tools must be used to determine an individual's function and quality of life.
Functional testing is commonly used in rehabilitation to obtain a baseline value for physical function and to determine an individual's future functional ability.5 Functional testing is clinically important because it provides health care professionals with information regarding a participant's level of fitness. This information can be used to assess future risk for disease or injury.6,7 Individuals with amputations are at a higher risk for developing comorbidities including cardiovascular disease, diabetes, depression, arthritis, osteoporosis, obesity, and cancer since the daily life activities of a person with an amputation can be restricted compared with the daily life activities of an individual without an amputation.3,6–8 Physical inactivity has been found to be a major contributing factor in the deteriorating physical health of people with disabilities.3,6–8 Being physically active is crucial to diminish the disadvantages of limb loss in the amputee population.6
A child's ability to participate in activities of daily living, hobbies, and sports, and his or her functional ability, have been identified as the outcomes of greatest importance to children and their families.1,9 However, there are few appropriate tools to adequately measure these constructs in children with lower limb amputations compared with adults with amputations.1,9 The majority of studies examining the functional ability of children and adolescents with amputations use subjective measures and parent report rather than objective, functional measurements.3,10,11 Although subjective information is important, it may not be sufficient to measure functional ability.12,13 It is conceivable that a combination of subjective and objective measures could give the most accurate evaluation of an individual's functional mobility. Marchese et al14 designed a tool, the Functional Mobility Assessment (FMA), to test the functional mobility of children older than 13 years following lower extremity sarcoma resection. The FMA incorporates valid and reliable objective measures of function, such as the Timed Up and Go (TUG)15 and Timed Up and Down Stairs (TUDS),16 in addition to subjective measures to determine functional mobility. The FMA's reliability and validity are supported for individuals with lower extremity amputations caused by osteosarcoma between the ages of 10.4 and 42.4 years. It would be useful to assess the validity of the FMA in subjects with varying causes of lower extremity limb loss, not just osteosarcoma. Specifically, it is essential to assess the FMA's ability to discriminate between individuals of varying functional levels. A study by Waters et al17 examined the role of amputation on energy expenditure and gait parameters. The authors examined subjects with typical development and subjects with amputation at various levels. Findings from their study revealed that subjects with typical development scored significantly better on all parameters than subjects with amputations. Therefore, the present study will examine the differences between subjects with amputations and subjects with typical development.
Being able to measure the functional mobility of a child with an amputation using subjective and objective measures, such as the FMA, allows a more accurate measure of the child's physical abilities and perspectives. Using a defined age group of children and adolescents will assist in examining the usefulness of this tool in a pediatric population. The information gained from the FMA can be used to set goals to promote physical activity and to assess and decrease future risk for disease or injury in children with amputations. The purpose of this study was to assess the validity of the FMA in children and adolescents between the ages of 8 and 19 years with lower extremity amputations who were high-functioning and independently ambulatory.
Amputation Group and Setting
An outdoor adventure camp for 36 children with amputations provided the sample group for this study. Children with amputations, ages 8 to 17 years, could attend this camp upon completion of an application, which included medical/health information, and indications of ability to navigate uneven terrain with limited additional assistance. All children with lower extremity amputations (n = 30) received information and consent forms explaining this research project. Approval by the Institutional Review Board of Shenandoah University was obtained before the beginning of camp and data collection. Signed informed consent forms were returned with the camper application and collected prior to data collection. Thirty campers with amputations were eligible for inclusion in the study and all agreed to participate in the study. The activities of the camp limited opportunities for data collection as subjects were removed from camp activities to participate in data collection. Data were also collected from 2 counselors with amputations, aged 18 and 19 years, before collecting data on the campers to ensure proper application of the FMA. The counselors had previously attended the camp and met the inclusion criteria established for the campers. The counselors signed a consent form at camp. Their data are included in the study to incorporate greater variability in the amputation group. Complete information was collected on 25 of the 30 participants. Inclusion criteria to participate in the study were (1) being a camper or counselor; (2) having a diagnosis of a lower extremity amputation that was congenital, or due to cancer, infection, trauma, or disease, or surgery of the lower limb; (3) being between the ages of 8 and 19 years; and (4) having no current injury to the lower extremity (including lesions of the residual limb tissue).
All testing of the subjects took place during the time of the camp. The 5-day camp was held in an outdoor mountainous area with terrain that is uneven and rocky and includes hills, grassy areas, and walkways that are concrete, wooden, or gravel. All camp activities were accessible to all campers; however, all campers had to be independent in self-care, be able to navigate the terrain with minimal assistance, and be able to participate in all camp activities. Camp activities included a high ropes course and swing, whitewater rafting, horseback riding, tennis, waterskiing and boating, swimming, and activities that are physically less stressful such as gem mining, board games, and a talent show. All children participated in all activities to the best of their ability and assistance was provided as needed. Testing of subjects in the amputation group took place on outdoor wooden steps (TUDS) or on an indoor even surface (TUG, 9-minute walk/run).
Control Group and Setting
The control group included 12 subjects and was a sample of convenience comprised of children without limb deficiencies recruited from the local community. Informed consent forms were provided, signed, and collected prior to data collection. The same demographic information and objective portions of the test were used with the control group as were used with the subjects with amputations. Inclusion criteria for the control group were (1) having no amputation or diagnosis of lower limb amputation, (2) being between the ages of 8 and 19 years, and (3) having no current injury to the lower limbs. Testing of the control group subjects took place after the camp, indoors at the researchers' academic institution. The institutional setting, with its tiled floor and concrete staircase, was a more controlled environment than the camp. All subjects were able to be tested within a 3-hour time period.
The investigators, including 2 physical therapists and 2 physical therapist students, collected all data. The students and a faculty researcher performed data analysis. The investigators discussed the functional tests and subjective interviews before implementing data collection for consistency in testing and questioning. During data collection, the investigators assigned each subject an identification number. After data collection, this number was used to blind the investigators to the subject's identity during data analysis. Demographic information was collected from all subjects and included ethnicity, age, gender, and diagnosis. Both groups were evaluated using the FMA tool, which consists of 3 subjective questions and 6 objective measures including the TUG, TUDS, 9-minute walk/run, pain, heart rate (HR), and the Borg Rating of Perceived Exertion (RPE). The subjects with amputations were also asked for age at the time of amputation, age at revision, type of surgery, type of prosthesis, and 3 open-ended questions regarding activity participation and the effect of camp. The objective tests were performed in a random order to limit order effect. A pain rating and HR were measured immediately before and immediately after the TUG and TUDS. For the 9-minute walk/run, HR and pain were measured at 4 intervals including immediately prior, mid-task, immediately post-, and 3 minutes post-task. The RPE was taken immediately following each functional mobility test.
The FMA is reported to have good internal consistency (Cronbach alpha = 0.75) and excellent intra- and interrater reliability (ICC > 0.97) when used on individuals diagnosed with osteosarcoma who are 13 years of age or older.14
The TUG, which is one of the components of the FMA, is a measure of mobility in which the subject rises from a chair and walks 10 ft to a cone and around this cone to return to the chair and sit.15 The psychometrics of the TUG have been well supported in the literature.15,16 When used with adults with unilateral transtibial and transfemoral amputation, the TUG has strong intrarater and interrater reliability (r = 0.93 and 0.96, respectively).15 In children with disabilities, the TUG also demonstrates strong reliability.16 Specifically, in children aged 3 to 12 years with a diagnosis of spina bifida and cerebral palsy, the within-session reliability and test-retest reliability are good (ICC = 0.89 and 0.83, respectively).16 However, no psychometric evidence has been reported supporting the use of the TUG for children with amputations.
The TUDS is a measure of functional mobility and balance in which the subject ascends and descends 1 flight of stairs (12 steps).18 The TUDS has excellent intrarater, interrater, and test-retest reliability (ICC > 0.94) and moderate to high concurrent validity (Spearman r = 0.78) with the TUG in children from 8 to 14 years of age with and without cerebral palsy.18 The investigators of the present study were unable to locate literature assessing the use of TUDS in children with amputations.
The 9-minute walk is a measure of cardiovascular endurance in which the subject ambulates in a designated 65-ft-long area for 9 minutes. The 9-minute walk is a test taken from a physical fitness battery developed by the American Alliance for Health, Physical Education, Recreation and Dance, which is commonly used to assess physical fitness in school-aged children.14 No study could be found assessing the validity of the 9-minute walk test in children or adults with or without disabilities.
The Borg RPE was used as a subjective measure of the subject's level of exertion following each functional test. Rating of Perceived Exertion uses a numerical scale from 6 (no exertion at all) to 20 (maximal exertion). The subjects were instructed in the use of the RPE prior to the initiation of the objective tests. The Borg RPE has been validated for use with adults, but there is little literature supporting the use of this outcome measure in children.19 In adolescent girls aged 14 to 16 years, the BORG RPE demonstrated an ICC = 0.78.19 No evidence could be found assessing the validity of the BORG RPE in children with amputations.
Heart rate was monitored and recorded using a finger pulse oximeter on the index finger of the subject. The same 2-pulse oximeters were used on both the subjects with amputations and the control subjects. Physiological Cost Index (PCI) was calculated as an indicator of walking efficiency14 using the following equation:
PCI (beats/min) = (HR while walking − HR at rest)/walking speed (m/min)
The validity of the PCI has been established in the literature for use in children with cerebral palsy20 but has not been examined in children with amputations.
Pain was measured using the Wong & Baker FACES scale.21 The FACES scale uses 5 cartoon-like faces with a descriptor for each face ranging from “no hurt” to “hurts worst.” The scale is a reliable tool to use with healthy children between the ages of 3 and 18 years.21 No psychometrics are available for children with amputations.
Descriptive and inferential statistics were calculated for all of the subjects' demographic information as well as outcome measures. The raw data were converted into scaled data using the FMA developed by Marchese et al.14 The scaled scores ranged from 0 (worst) to 5 (best) with a minimum total score of 0 and a maximum total score of 70. All data were analyzed using SPSS version 16.0 (Chicago, IL). Nonparametric analyses (Mann-Whitney U test, chi-square test, and Spearman ρ) were used when the data were nominal or ordinal. Otherwise, parametric analyses were performed. The α level was set a priori at 0.05. A Bonferroni correction factor was used when multiple comparisons were performed to decrease the likelihood of committing a type I error.
Thirty-seven subjects, those with amputations (n = 25) and subjects who were developing typically (n = 12), were included in the data analysis (Table 1). The total FMA score for 3 individuals from the amputation group could not be calculated because of uncollected RPE scores on the TUDS (1 subject) and 9-minute walk (2 subjects). The mean age of the amputation group was 12.36 ± 3.07 years (range, 9–19 years). The mean age of the control group was 10.25 ± 1.42 years (range, 9–13 years). The control group was significantly younger than the amputation group (independent Student t tests, P = .007).
The discriminant validity of the FMA tool was assessed by comparing the total FMA score and the components of the FMA tool between the amputation group and the control group (Table 2).
The control group had significantly higher FMA scores than the amputation group (Mann-Whitney U test, P = .001), indicating that the control group had higher functional mobility across several domains than the amputation group. The control group had significantly higher scaled scores (independent Student t test, P < .01) than the amputation group in the following subscores: TUG time, TUDS time, and the 9-minute walk/run test (Table 2). Higher scores on the TUG and TUDS indicate that the control group is more efficient with dynamic balance tasks than the amputation group. Higher scores on the 9-minute run/walk indicate that the control group has better endurance in long distance activities and can ambulate faster than the amputation group. In addition, there were significantly more reports of “very satisfied with walking” (Mann-Whitney U test, P < .01) and more reports of “participate in work or school/sports” (Mann-Whitney U test, P = .001) between the control group and the amputation group. In contrast, there were no significant differences between the 2 groups on PCI, indicating that this parameter is not effective in discerning differences in endurance between the 2 groups. There were no significant differences between the 2 groups in pain, the TUDS RPE, TUG RPE, 9-minute walk/run RPE, or assistive device use.
Although the PCI, a measure of gait efficiency determined by HR per distance, was not significantly different between groups (independent Student t test, P = .976), HR measurements were significantly different between the control group and the amputation group, indicating more efficient energy expenditure in the control group than in the amputation group. These significant differences include a higher HR in the control group following the 9-minute walk/run (independent Student t test, P < .001) (Table 3), and a greater decrease in HR during the 3-minute recovery period in the control group (independent Student t test, P = .004) (Table 4).
This study achieved the purpose of assessing the validity of the FMA in a younger and more diverse population of children and adolescents with amputations. The results indicate that the total FMA score can discriminate between subjects with and without lower extremity amputation. Most of the individual outcome measures revealed differences in functional mobility between the 2 groups. For example, subjects in the control group walked further during the 9-minute walk. Also, times on the TUG were greater in the amputation group than in the control group. A possible explanation for the significant differences between groups, including increased times on mobility tests, increased HR and a longer recovery period, is the increased energy cost of ambulation in those using a prosthesis.22,23 Subjects with amputations have a higher energy cost for ambulation, which is largely due to additional energy associated with the walking movement itself, but also the effort of maintaining balance and posture.23 In a study by Hoffman et al,23 individuals with bilateral above-knee amputations demonstrated VO2 values that were 55% to 83% higher than aged-matched controls who were able-bodied and had HRs 30 to 50 beats per minute higher across 3 different walking speeds. This would account for the higher HRs and thus a slower HR recovery seen in the subjects with amputations in this study following the 9-minute walk test.
Also noteworthy are the significant differences in the satisfaction with walking and participation portions of the FMA. As expected, the subjects without amputation scored significantly higher on the mobility tests and also had higher reports of satisfaction with walking and participated in more sports and school/work activities than subjects with amputations, suggesting a potential correlation between objective and subjective measurements. Future studies should expand on the psychosocial factors influencing children and adolescents with amputations and the effect they have on objective measures.
The outcomes that did not significantly discriminate between the 2 groups were the PCI, pain, RPE, and the subjective measure of assistive device use. The authors believe that the lack of significant difference on the pain and assistive device use portions of the FMA was due to the subjects with amputations being high functioning, independently ambulatory, not using assistive devices, and demonstrating little to no pain.
The investigators noticed that during testing, the younger subjects had difficulty grasping the concept of the Borg RPE. There were several occasions during testing where a subject, in either group, would report an RPE of “6” (meaning no exertion at all) after the 9-minute walk/run test even though the subject was clearly showing signs of fatigue and exertion such as sweating or breathing heavily. It was assumed that the subjects did not understand the RPE scale. According to Groslambert and Mahon,24 the correlation of the Borg RPE scale with HR is low (r = 0.45–0.79) in children aged 9 to 11 years, indicating that the Borg RPE may not be appropriate for subjects in the present study. Groslambert and Mahon study24 also report that exercise intensity in children must be high for the RPE to be accurate, because at slow to moderate walking speeds, children have difficulty with whole-body perceived rates of exertion. If the activities, such as the TUDS or TUG, were not intense enough, the children may not have been able to perceive accurate rates of exertion resulting in an inaccurate RPE rating. Recent research indicates that the Children's Effort Rating Table (CERT) scale is an appropriate and reliable tool (r = 0.69–0.79).19,25 The CERT has shown significant correlation with HR and exercise intensity in children between the ages of 8 and 12 years. In subjects aged 13 to 18 years, the OMNI Perceived Exertion Scale for Resistance Exercise has been shown to be more valid and reliable than the Borg RPE.19,26 Development of a new tool to examine functional mobility in children with limb differences should consider the above findings in the literature and this is clearly an area for future research.
It should also be noted that whereas the 9-minute walk/run test is a valid tool14 and was shown to be valid in discriminating between the 2 groups in this study, as the subjects without amputations walked further distances than those with amputations, several children were easily distracted and unmotivated to complete a test of 9 minutes duration. Other walk tests should be considered for inclusion within the FMA in place of the 9-minute walk such as the 6-minute walk test, which has been shown in the literature as an effective tool for children and adults with disabilities. The 6-minute walk test is reproducible, takes less time to complete, and has higher levels of evidence than the 9-minute walk test.26 Future studies should examine various walk tests to determine the best alternative to make the FMA a more time-effective clinical assessment tool.
Whereas the FMA discriminated between the groups, several of the items on the FMA were not discriminatory. Given the performance of the subjects in this study, the authors believe that these items and perhaps the entire tool may lack sensitivity to change in this population. Raw data were converted to scaled data ranging from 0 to 5. The scaling appeared to decrease the sensitivity of the data to change, which as a result did not appear to accurately reflect the range of scores in this population. Although the ranges provided may be appropriate for an adult population, they may be inaccurate for a younger population. For example, a HR greater than 167 is scaled as a “0” (worst) and a HR of 126 is scaled as a “5” (best). In a study by Marinov et al,27 the HR of children without amputations during a treadmill exercise test reached an average of 178.2 ± 7.5 beats per minute. For a child who is healthy and developing typically, this representative HR would score as a “0” according to the FMA data scale. The data scale ranges should be examined and adjusted to reflect the appropriate HR ranges for children with disabilities. Rescoring of the entire FMA with the incorporation of the continuous data obtained from the individual components may be an area for the future examination to improve the sensitivity and utility of this measure in the clinical setting.
There are several possible explanations for the differences in the findings of this study versus the original study performed by Marchese et al.14 They examined a homogenous group of subjects with lower extremity amputations resulting from osteosarcoma, while the present study examined a more heterogenous group of subjects with varying etiologies of lower extremity amputations. It should also be noted that whereas the present study had a small sample size, the FMA was able to significantly discriminate between the 2 groups arguing against a type II error. However, future studies should use larger heterogenous groups to determine the psychometrics of the FMA.
A limitation of the present study was the lack of variety in functional ability in the subjects with amputations. Even though a significant difference was found between the control group and the amputation group in total FMA score, the researchers believe that the subjects within the amputation group represent a highly active portion of the population with amputations thereby narrowing the range of FMA scores. All of the subjects with amputations in the study were able to ambulate independently with or without the use of an assistive device. Also, the subjects appeared motivated and were physically active enough to participate in the camp activities. A larger difference may be seen in total FMA scores if children from a broader functional spectrum were examined. The control group was a sample of convenience and had little variability within the group.
Interrater reliability statistics were not calculated for this study. Therefore, consistency among the 4 evaluators could not be determined and is another limitation of this study. The testing facility for the amputation group was a distraction for the subjects due to the intermittent presence of bystanders in the testing area. While the obstacles presented during the testing may be an accurate portrayal of ambulation within the community, these obstacles were not present during the control group testing, which may have increased the risk of error present within the study. In addition, it should be noted that the control group was significantly younger than the children with amputations. Logically, older children and adolescents would perform better than younger children. However, because the amputation group scored lower, the investigators do not believe that the age difference influences the conclusions that can be drawn from this study.
The present study highlights many areas for consideration and future research. Future researchers should consider tools that are valid, reliable, and appropriate for the age range being tested, such as the CERT or OMNI, or an alternate walk/run test, such as the 6-minute walk test. The data used in the scale should be modified to accurately reflect normative values for children. A larger population of children with amputations should be used with more variability in functional ability to ascertain the reliability and validity of a revised tool in a broader spectrum of children with amputations. The ability of the FMA to detect changes over time must also be examined. With the incorporation of our suggestions in future studies, another tool can be developed that accurately incorporates both subjective and objective measures to assess the functional mobility in children with lower extremity amputations. However, given the lack of discriminative ability of the individual components when not scaled, one must consider that the use of several focused separate measures may prove to be more valid and clinically meaningful to provide the current functional level of children with amputations.
The authors thank Colleen Coulter-O'Berry, PT, DPT, PhD, PCS, for her assistance with data collection.
1. Coster W, Khetani MA. Measuring participation of children with disabilities: issues and challenges. Disabil Rehabil. 2008;30:639–648.
2. Varni JW, Setoguchi Y, Rappaport LR, Talbot D. Effects of stress, social support, and self-esteem on depression in children with limb deficiencies. Arch Phys Med Rehabil. 1991;72:1053–1058.
3. Wetterhahn KA, Hanson C, Levy CE. Effect of participation in physical activity on body image of amputees. Am J Phys Med Rehabil. 2002;81:194–201.
4. Dedmon BT, Davids JR. Acquired amputation of the lower extremity. J Bone Joint Surg. 2005;87:1054–1058.
5. Levin AZ. Functional outcome following amputation. Top Geriatr Rehabil. 2004;20:253–261.
6. Korkmaz A. The Physiological Effects of Sports in Amputees. Amputee Sports for Victims of Terrorism. Ankara, Turkey: Centre of Excellence Defense against Terrorism; 2007.
7. Legro MW, Reiber GE, Czerniecki JM, Sangeorzan BJ. Recreational activities of lower-limb amputees with prostheses. J Rehabil Res Dev. 2001;38:319–325.
8. Kelley R. Exercise a must for amputees. O&P Business News. June 15, 2006:17–24.
9. Vannah WM, Davids JR, Drvaric DM, Setoguchi Y, Oxley FJ. A survey of function in children with lower limb deficiencies. Prosthet Orthot Int. 1999;23:239–244.
10. Deans SA, McFadyen AK, Rowe PJ. Physical activity and quality of life: a study of a lower-limb amputee population. Prosthet Orthot Int. 2008;32:186–195.
11. Zahlten-Hinguranage A, Bernd L, Ewerbeck V, Sabo D. Equal quality of life after limb-sparing or ablative surgery for lower extremity sarcomas. Br J Cancer. 2004;91:1012–1014.
12. Pruitt SD, Varni JW, Setoguchi Y. Functional status in children with limb deficiency: development and initial validation of an outcome measure. Arch Phys Med Rehabil. 1996;77:1233–1238.
13. 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.
14. Marchese VG, Rai SN, Carlson CA, et al.. Assessing functional mobility in survivors of lower-extremity sarcoma: reliability and validity of a new assessment tool. Pediatr Blood Cancer. 2007;49:183–189.
15. Schoppen T, Boonstra A, Groothoff JW, de Vries J, Goeken LNH, Eisma WH. The Timed “Up and Go” Test: reliability and validity in persons with unilateral lower limb amputation. Arch Phys Med Rehabil. 1999;80:825–828.
16. Williams E, Carroll S, Reddihough D, Phillips B, Galea M. Investigation of the Timed “Up & Go” Test in children. Dev Med Child Neurol. 2005;47:518–524.
17. Waters RL, Perry J, Antonelli D, Hislop H. Energy cost of walking of amputees: the influence of level of amputation. J Bone Joint Surg Am. 1976;58:42–46.
18. Zaino CA, Marchese VG, Westcott SL. Timed Up and Down Stairs test: preliminary reliability and validity of a new measure of functional mobility. Pediatr Phys Ther. 2004;16:90–98.
19. Pfeiffer K, Pivarnik J, Womack C, Reeves M, Malina R. Reliability and validity of the Borg and OMNI Rating of Perceived Exertion scales in adolescent girls. Med Sci Sports Exerc. 2002;34:2057–2061.
20. IJzerman MJ, Nene AV. Feasibility of Physiological Cost Index as an outcome measure for the assessment of energy expenditure during walking. Arch Phys Med Rehabil. 2002;83:1777–1782.
21. Luffy R, Grove SK. Examining the validity, reliability, and preference of three pediatric pain measurement tools in African-American children. Pediatr Nurs. 2003;29:54–59.
22. Waters RJ. The energy expenditure of normal and pathologic gait. Gait Posture. 1999;9:2007–231.
23. Hoffman MD, Sheldahl LM, Buley KJ, Sanford PR. Physiological comparison of walking among bilateral above-knee amputee and able-bodied subjects, and a model to account for the differences in metabolic cost. Arch Phys Med Rehabil. 1997;78:385–392.
24. Groslambert A, Mahon AD. Perceived exertion influence of age and cognitive development. Sports Med. 2006; 36:911–928.
25. Ginsberg JP, Rai SN, Carlson CA, et al.. A comparative analysis of functional outcomes in adolescents and young adults with lower-extremity bone sarcoma. Pediatr Blood Cancer. 2007;49:964–969.
26. Geiger R, Strasak A, Treml B, et al.. Six-minute walk test in children and adolescents. J Pediatr. 2007;150:395–399.
27. Marinov B, Mandadjieva S, Kostianev S. Pictorial and verbal category-ratio scales for effort estimation in children. Child: Care, Health Devel. 2007;34:35–43.
adolescent; amputation/lower extremity; child; mobility limitation; physical exertion; test validity; walking
© 2011 Lippincott Williams & Wilkins, Inc.