Gait symmetry and walking speed analysis following lower-extremity trauma.

Lower-extremity salvage procedures have become widespread, and limbs once amputated are now routinely treated with complicated reconstruction. (1,2) One of the primary outcomes in limb-salvage rehabilitation is the recovery of functional gait. (3-6) Currently, the most widely used informal measures are walking speed, gait symmetry, and stride duration. (3-5,7-10) Although there is extensive literature on limb function and the reorganization of gait after amputation, relatively little is known about the limb-salvage recovery process.

Rehabilitation of people with lower-extremity reconstruction typically ranges from 12 to 72 months, with a median recovery time of 30 months. (11) Smaller studies (<40 subjects) that have examined the recovery of gait have demonstrated that a slower preferred walking speed, a lengthened stride time, a deterioration of balance control, and an involvement of the knee joint are associated with longer salvage recovery times. (4,11,12)

One study (6) showed that poor muscle strength (force-generating capacity) was correlated with abnormal gait, as well as the significant prognostic factors of increasing age and female sex. In salvages from bone tumors, decreased knee extensor strength was associated with a step-to-step (nonreciprocal) stair-climbing pattern, decreased locomotion, and compensatory hip and ankle movement strategies. (5) Research suggests that muscle power is a primary determinant of gait, (6,13,14) and manual muscle testing is a common clinical assessment tool for patients with lower-extremity trauma. The functional role, optimal strength, and optimal range of motion (ROM) of the hip, knee, and ankle have been well documented for normal gait, but the association of these measures with the recovery of gait among patients with limb salvage remains questionable.

The purpose of this study was to identify clinical measures associated with gait asymmetry and impaired walking speed in patients with lower-extremity reconstruction. The attainment of independent and unimpaired ambulation is an essential element of rehabilitation for lower-extremity trauma. Based on prior limb-salvage research, we hypothesized that increasing age, nonreciprocal stair-climbing ability, poor single-leg balance, and a manual muscle test (MMT) grade of less than 4 for hip extension, knee extension, ankle dorsiflexion (DF), and ankle plantar flexion (PF) would be statistically associated with gait deviation and walking speeds of less than 4 ft/s (1 ft/s=0.30 m/s). Understanding the association of strength, ROM, and functional measurements with impaired physical mobility in patients with lower-limb reconstruction will assist clinicians with establishing effective plans of care and help inform treatment decision making and prioritization of interventions.

Method

Design and Subjects

In the present study, we used 24-month follow-up data collected as part of a larger, multicenter study, the Lower Extremity Assessment Project (LEAP). The LEAP study was originally designed to determine functional outcomes of patients with severe lower-extremity trauma resulting in reconstruction or amputation. A total of 601 patients, 16 to 69 years of age, were enrolled between March 1994 and June 1997 from 8 level I trauma centers (Carolinas Medical Center, Cleveland MetroHealth Medical Center, Harborview Medical Center, North Carolina Baptist Hospital, R Adams Cowley Shock Trauma Center, Tampa General Hospital, University of Texas Southwestern Medical Center, and Vanderbilt University Medical Center). (15) Lower-extremity trauma was defined as complex fractures (Gustilo grade IIIB and IIIC fractures, selected grade IIIA fractures), (16) dysvascular limbs, major soft tissue injuries to the tibia (degloving or severe crush or avulsion injury), or severe foot and ankle injury. (17) Exclusion criteria included the following: a score of less than 15 on the Glasgow Coma Scale on admission, spinal cord deficits, prior amputation, third-degree burns, inability to speak English or Spanish, documented psychiatric disorder, and active military duty. (17)

The current analysis excluded 32 patients with bilateral injuries and 161 patients treated by amputation. An additional 27 patients were lost to follow-up at 24 months, resulting in 381 subjects with lower-limb reconstruction. Patients lost to follow-up were more likely to be of lower socioeconomic status than those with complete follow-up. (17) Lower-extremity reconstruction surgery was performed primarily as a result of complex tibia fractures (51%), with the remaining 30% performed following severe foot and ankle injuries (Gustilo grade IIIB ankle fractures, all grade III intra-articular fractures of the distal tibia) and 18% following dysvascular injuries (knee dislocations, closed fractures of the tibia, or penetrating wounds with vascular injury). (15,17) Injury characteristics included articular involvement (35%), nerve damage (6%), moderate-to-severe muscle damage (48%), and bone loss greater than 2 cm (21%). The mean age of the primarily male (75%) subjects was 35.6 years (SD=12.3). The majority of the participants were white (71%) and had less than a grade 12 education (69%).

Outcome

This study was primarily interested in the recovery of functional gait, as measured by gait speed and symmetry. Walking speed of [less than or equal to] 4 ft/s and gait deviation were combined into one primary outcome measure. This combined measure was used not only to provide an indication of impaired physical mobility, (18) but to reflect a subject's poor quality of movement. (19) Below we describe how walking speed and gait deviation were evaluated in the LEAP study and how these evaluations were combined into one primary outcome measure.

Walking speed. Subjects were asked to walk 150 ft (45.7 m) on a level surface "as fast as they can" without an assistive device. The time it took for subjects to complete the task was measured with a stopwatch and recorded as feet per second. The use of a stopwatch has been found to yield data with excellent concurrent validity with the gold standard of infrared timing gates and an intraclass correlation coefficient (ICC) of at least .998 for tests of walking speed. (20) No other reliability testing was conducted on this measure.

Gait deviation. Subjects who participated in the 150-ft independent walking test also were visually monitored for gait deviations. Physical therapists were provided with a detailed description of each deviation. Participants were found to have gait asymmetry if they had one or more of the following common deviations: Trendelenburg gait, trunk asymmetry, leg circumduction, hip hike (increased hip abduction on the unaffected stance limb, with simultaneous pelvic elevation on the affected side during swing), knee hyperextension, no heel-strike, toe drag, uneven step length, or a limp not accounted for by the other listed deviations. Observational techniques have been found to yield data with moderate reliability, (21-23) while instrumented gait analysis is the criterion standard. (24) However, a 3-dimensional technique was impractical and too costly for the present study.

Walking speed and gait deviation. A walking speed of 4 ft/s or less was considered an appropriate cutoff for impaired speed because the literature recognizes acceptable walking speeds of 3.2 ft/s for patients with lower-limb salvage 15 months postinjury and speeds ranging from 4.07 to 4.2 ft/s for fully recovered patients with transtibial and below-knee amputations. (3,25,26) Mean walking speeds for individuals without gait impairments range from 4.4 to 4.9 ft/s, depending on age, sex, and location of ambulation. (27-29) A walking speed cutoff of 5 ft/s also was examined to confirm associations between selected clinical variables and the primary outcome measure.

Potential Clinical Factors Associated With Abnormal Gait

A number of common impairment and functional measures were used as assessment tools to examine the factors associated with abnormal gait. We will describe these measures along with their scales and scoring systems.

Self-report pain scale. Pain level was assessed with a visual analog scale (VAS), using a 0- to 100-mm horizontal line. Participants were asked to rate their average daily leg pain by placing a mark on the line between "no pain" on the left and "unbearable pain" on the right. The VAS score was recorded by measuring the distance from the "no pain" end of the line to the participant's mark of personal pain intensity. A distance of 5 to 44 mm was considered mild pain, 45 to 74 mm was considered moderate pain, and 75 to 100 mm was considered severe pain. (30)

Stair-climbing and balance performance. Gait capacity can be estimated by a person's ability to go up and down stairs with a step-by-step (reciprocal) maneuver. (5) Participants were asked to climb 12 steps and then descend the same 12 steps at their preferred speed. A physical therapist noted whether the subject displayed a step-by-step pattern or a nonreciprocal, step-to-step pattern for both ascending and descending tasks.

Two parameters were used for standing balance: unilateral (single-leg) and tandem stance. For unilateral stance, subjects were asked to stand on one leg with their eyes open and arms crossed across their chest for 30 seconds. The number of seconds that the subject was able to stand before dropping the other leg was recorded for both the involved and uninvolved legs. Subjects were scored from 0 to 30 seconds. They then were asked to stand unsupported for 10 seconds with their eyes closed, arms crossed across their chest, and their injured foot touching the heel of their other foot. A score of less than 30 seconds for the unilateral stance and less than 10 seconds for the tandem stance was defined as functionally poor. (31-33) A score of 0 seconds was attributed to subjects being unable to perform either of the 2 balance tasks.

Girth measurements. Thigh and calf girth measurements were used to assess muscle atrophy. (34) Circumferential measurements have been found to have high intrarater reliability and interrater reliability, with ICCs of .82 and .72, respectively. (35) Data were recorded with a tape measure using the following landmarks: 8 cm proximal to the superior patella with knee in extension and 8 cm distal to the tibial tuberosity. Muscle atrophy of 0 cm, 1 to 2 cm, or 3 cm or greater was determined by comparing each subject's data for the involved leg with data for the uninvolved leg.

Range of motion. Hip flexion and extension, knee flexion, and ankle DF and PF were measured using a standard goniometric technique. (36) Subjects were asked to remove their shoes and socks and actively move the joint through the desired range. Physical therapists recorded the active range of motion (AROM) with the subjects positioned supine and then prone using a universal goniometer; measurements were rounded up to 5 for 3's and 4's and rounded down to 0 for 1's and 2's. Studies of the universal goniometer have shown high intrarater reliability for knee and ankle ROM, with ICCs for knee flexion and extension ranging from .97 to .99 and ICCs for ankle PF and DF ranging from .82 to .86. (37,38) Starting and ending positions of each joint, as recommended by the American Academy of Orthopaedic Surgeons (AAOS), (39) were used to record measurements. Norms were determined based on the averages published by the AAOS; ROMs below 120 degrees of hip flexion, 30 degrees of hip extension, 135 degrees of knee flexion, 20 degrees of ankle DF, and 50 degrees of ankle PF were considered restricted.

Strength. Hip flexion, extension, and abduction; knee flexion and extension; and ankle DF and PF strength were measured by a patented device for exercising and measuring strength of a person's limb. (40) The strength apparatus includes a pair of pivot clamps that were used to connect it to a physical therapy table. The pivot clamps enabled rotational and translational movement of the frame, which allowed the flame to be positioned in a desired location and orientation relative to the limb being tested. Subjects were asked to apply maximal force against the force plates in the direction of the desired movement. A force transducer produced an output that represented each subject's force, and the output was displayed on a digital panel meter. A force gauge has been found to measure strength more reliably and accurately than manual muscle testing.

Three measurements of each motion were recorded for both the involved and uninvolved limbs. The maximum effort was selected, and the ratio of the injured limb to the uninjured limb for each motion was calculated. The ratio scores were separated into 2 categories: less than 50% and greater than 50%. Because manual muscle testing (41) is a more widely used measure and the MMT scale of 0 to 5 is clinically meaningful, (42) the ratio categories were translated into MMT grades with the use of the percentage scores of Kendall and McCreary. (43) Normal strength or an MMT grade of 5 corresponds to a joint motion ratio of greater than 81%, a ratio of 51% to 80% is considered grade 4, a ratio of 21% to 50% is considered grade 3, and a ratio of 20% or less is considered grade 2. (43,44) An MMT grade of 4 is considered an average or good score, and an MMT grade of less than 4 indicates fair or even poor strength. (41) Because a good score is considered a necessity for fast walking, (45) subjects in this study with an MMT grade of less than 4 were considered to have impaired strength.

An additional variable of toe raises was used to measure functional ankle PF strength of the involved and uninvolved legs. (41) The number of times a subject performed full-excursion toe raises within 15 seconds, while keeping the knee straight and using one hand for support, was recorded. (41,46) If a subject was unable to perform a toe raise, then a score of zero was recorded for that subject. A normal score was 10; a score of less than 10 was considered functionally impaired. (33)

Data Analysis

Data analysis and interpretation of results were performed using Stata statistical software, version 8.0. * Bivariate relationships between demographic and clinical variables and outcome measures were assessed using chi-square tests for binomial data and Student t tests for continuous data.

The main goal was to develop a multivariate logistic regression model of clinical measures correlating with gait deviation and impaired walking speed 2 years following major lower-extremity trauma treated by reconstruction. Separate analyses were conducted for gait deviation and walking speed at [less than or equal to] 4 ft/s and for gait deviation and walking speed at [less than or equal to] 5 ft/s. Each analysis was conducted in 4 phases. First, a bivariate logistic regression analysis was performed to assess the relationship between potential variables and the combined outcome of gait deviation and decreased walking speed. Second, an initial multivariate logistic regression analysis was conducted with all baseline sociodemographic characteristics and clinical variables. The results of the multivariate and bivariate logistic regression analyses were compared to ensure that no covariates had been incorrectly dropped from the analysis, and the presence of interactions and correlations between impairment and functional measures was investigated. Third, clinical factors with P values less than or equal to .10 using the Wald test were selected for additional analysis. These variables were confirmed with stepwise regression and goodness-of-fit techniques. Fourth, a final multivariate logistic regression analysis was performed with variables that were statistically significant at the P<.10 level or if removal of the variable resulted in substantial changes in the magnitude of other variables in the model. Data for subjects who were missing specific data points were kept in the analysis by using missing data categories, but these categories are not presented.

Results

Information on gait deviation and walking speed was available on 277 (73%) of the eligible 381 patients with lower-limb reconstruction. Subjects with incomplete outcome data were significantly more likely to be 45 years of age or older (P<.05) and without a high school education (P<.01). Participants were found to have the following deviations: Trendelenburg gait (8%), trunk asymmetry (10%), leg circumduction (6%), hip hike (8%), knee hyperextension (3%), no heel-strike (7%), toe drag (13%), uneven step length (20%), and a noticeable limp not accounted for by one of the other deviations (12%). The subjects' mean walking speed for 150 ft was approximately 4.7 ft/s (SD=1.9).

The distributions of demographic and injury characteristics and clinical variables by outcome are presented in Tables 1 and 2. The bivariate associations between the primary outcome and these clinical measures 24 months after reconstruction are shown in Table 3. Nonreciprocal stair-climbing pattern, <30 seconds of unilateral stance, <10 seconds of tandem standing balance, <10 toe raises, [greater than or equal to] 3 cm of calf atrophy, <120 degrees of hip flexion ROM of the involved limb, <50 degrees of ankle PF ROM of the involved limb, <20 degrees of ankle DF ROM of the involved limb, and all strength measures with an MMT grade of <4, except hip abduction, were independently related to gait deviation and decreased walking speed at P<.05. After adjusting for age, insurance, and injury, only a nonreciprocal stair-climbing pattern, impaired unilateral stance, decreased hip extension and knee flexion strength, and limited ankle PF ROM of the involved limb remained significantly associated with impaired physical mobility (P<.10).

From the final multivariate logistic regression model, the most significant factors associated with gait deviation and walking speed of [less than or equal to] 4 ft/s for subjects with lower-limb reconstruction were <50 degrees of PF ROM of the involved ankle, nonreciprocal stair climbing, and knee flexion strength with an MMT grade of <4 at P<.05 (Tab. 4). In addition, these associations were further confirmed with a multivariate logistic regression of gait deviation and walking speed of [less than or equal to] 5 ft/s.

Discussion

As expected, we found a significant association between pathological gait and age, nonreciprocal stair-climbing ability, and poor single-leg balance. Age has previously been found to be a significant predictor of poor gait (P<.001) and overall limb function (P<.002) in patients with limb salvage. (6) The literature discusses stair-climbing ability as an estimate of gait capacity (5,8) and balance control as an area of clinical concern, (3,47) but our study is the first study to show a statistically significant correlation between abnormal gait and both nonreciprocal stair-climbing pattern and poor single-leg balance of the involved and uninvolved limbs. Balance has been examined more thoroughly in patients with lower-limb amputation, with one study finding single-leg balance of the uninvolved limb to be a significant predictor of functional outcome in elderly people with amputations. (31)

Studies of gait recovery after limb-saving surgery secondary to malignant bone tumors and trauma have consistently shown an association between decreased knee extension strength and increased asymmetry and decreased walking speed. (3,5,11) The association between knee extensor strength and impaired gait directly relates to the extent of quadriceps femoris muscle excision during the tissue and bone resection process of tumor surgery (3,48) and the resulting quadriceps femoris muscle atrophy in patients with traumatic reconstruction. (11) In this study of patients with traumatic lower-extremity reconstruction, decreased knee extension strength and thigh atrophy were not significantly associated with gait deviation and decreased walking speed. We found that only 24% of the subjects had a knee extension MMT grade of <4 and 18% had thigh atrophy of [greater than or equal to] 3 cm. The limited quadriceps femoris muscle involvement in this patient population can be attributed to their distribution of injury characteristics. The majority of the reconstruction surgeries were due to tibia fractures (51%) and severe foot and ankle injuries (40%), where excision of the quadriceps femoris muscle is not common.

Instead, impaired knee flexion strength was found to be significantly associated with pathological gait. Approximately, 72% of the subjects with impaired knee flexion strength had a gait deviation, and 50% had a gait deviation and a walking speed of [less than or equal to] 4 ft/s. This high percentage of gait deviation is consistent with the functional role of the knee during ambulation. The knee absorbs energy during the stance phase of gait and is mainly involved with smoothing the gait pattern. (49) A smaller contribution is made to walking speed by the biceps femoris muscle, which flexes the knee during the swing phase to assist with flexion velocity as well as smooth foot clearance. (4)

Hip extension strength also was found to correlate with gait pattern and speed. Although the knee flexors are involved mostly with energy absorption, the hip extensors are responsible for propulsion and forward acceleration of the trunk. In normal gait, propulsion is initiated after heel-strike and is maintained during mid-stance though eccentric activation of the hip extensor muscles. (13) The hip extensors also lengthen the supporting limb and reduce excessive drop of the body's center of mass. (14) Propulsion is an essential component of gait speed, while control of the stance limb allows for a smooth transition into the swing phase of gait. A study by Sadeghi et al (26) of patients with amputation showed that increased hip extensor strength and early activation during the stance phase is needed to control perturbations and normalize walking speed in order to compensate for the lack of ankle function on the amputated leg.

An unexpected finding was the strong contribution of decreased ankle PF ROM to gait deviation and walking speed of [less than or equal to] 4 ft/s. Other researchers (50,51) have found ankle stiffness among patients with lower-extremity reconstruction and proposed that diminished ankle ROM contributes to impaired gait and stair-climbing ability, but a statistical association has not been reported in the literature.

The objective of our study was not only to identify factors significantly associated with gait deviation and decreased walking speed, but also to provide measurement guidelines that would assist clinicians with treatment planning and goal setting for patients with lower-limb reconstruction. The independent associations with gait deviation and impaired walking speed suggest that treatment plans of care should consider strategies to address the following impairments of the involved lower extremity: calf atrophy greater than 3 cm, hip flexion ROM below 120 degrees, ankle PF ROM below 50 degrees, ankle DF ROM below 20 degrees, and an MMT grade of less than 4 for hip and knee extension and flexion strength and ankle PF and DF strength. In addition, functional goal setting should consider: a reciprocal stair-climbing pattern, unilateral stance of the involved and uninvolved lower extremities for greater than 30 seconds, tandem standing balance for greater than 10 seconds, and the ability to perform more than 10 toe raises on the involved and uninvolved limbs.

Results from the multivariate analysis help inform treatment priorities, which appear to be improving ankle PF ROM to greater than 50 degrees, improving knee flexion strength to an MMT grade of greater than 4, and improving a patient's ability to ascend and descend stairs with a reciprocal stair-climbing pattern. Additionally, we believe that priority consideration should be given to improving hip extension strength and single-leg stance of the involved and uninvolved limbs. Recommendations for future research include examining the predictive ability of these impairment and functional measures on impaired mobility and physical disability. Further investigation is needed to determine the timing of treatment strategies and to support the inclusion of ankle PF ROM, knee flexion and hip extension strengthening, stair climbing, and single-leg-stance balance training in goal setting for patients with lower-extremity reconstruction.

A number of limitations of our study must be acknowledged. Information on the primary outcome and the functional variables was available on 73% to 75% of the eligible subjects. Subjects who were lost to follow-up at 24 months were more likely to be of low socioeconomic status than those who completed the study. Thus, the results may be an underestimation of the sample's level of abnormal gait and the associative ability of the significant clinical predictors. However, data on strength and ROM were available for only 31% to 40% of our sample. Even though the missing data were accounted for in the analysis, the small amount of impairment data may have contributed to the insignificant findings of knee extension strength and ankle involvement. In addition, the reliability of visual observations of gait deviation was not determined in this study and also should be considered in the interpretation of results.

The generalizability of our results is limited due to the focus of the study on patients from level I trauma centers. Limb-salvage outcomes may be influenced by the expertise of the staff from these level I sites. Moreover, the clinical variables were measured after 24 months, a period that may not be representative of full salvage recovery. Resolution of functional limitations and changes in gait may occur during the 24- to 36-month recovery period.

Conclusion

The most significant clinical factors associated with gait asymmetry and decreased walking speed among patients with lower-extremity reconstruction were ankle PF ROM of less than 50 degrees, an MMT grade of less than 4 for knee flexion and hip extension strength, a nonreciprocal stair-climbing pattern, and single-leg standing balance of the involved and uninvolved limbs of less than 30 seconds. These results can be used to inform treatment planning, to help prioritize clinical interventions, and for effective goal setting to attain unimpaired physical mobility among this patient population.

This article was received January 31, 2006, and was accepted July 25, 2006.

DOI: 10.2522/ptj.20060035

References

(1) Hansen ST Jr. The type-IIIC tibial fracture: salvage or amputation. J Bone Joint Surg Am. 1987;69:799-800.

(2) Lange RH. Limb reconstruction versus amputation decision making in massive lower extremity trauma. Clin Orthop Relat Res. 1989;243: 92-99.

(3) de Visser E, Veth RP, Schreuder HW, et al. Reorganization of gait after limb-saving surgery of the lower limb. Am J Phys Med Rehabil. 2003;82:825-831.

(4) de Visser E, Mulder T, Schreuder HW, et al. Gait and electromyographic analysis of patients recovering after limb-saving surgery. Clin Biomech. 2000;15:592-599.

(5) Kawamura H, Fuchioka S, Inoue S, et al. Restoring normal gait after limb salvage procedures in malignant bone tumors of the knee. Scand J Rehabil Med. 1999;31:77-81.

(6) Robinson MH, Spruce L, Eeles R, et al. Limb function following conservation treatment of adult soft tissue sarcoma. Eur J Cancer. 1991;27: 1567-1574.

(7) Schoppen T, Boonstra A, Groothoff JW, et al. The Timed "Up and Go" Test: reliability and validity in persons with unilateral lower limb amputation. Arch Phys Med Rehabil. 1999;80: 825-828.

(8) Deathe B, Miller WC, Speechley M. the status of outcome measurement in amputee rehabilitation in Canada. Arch Phys Med Rehabil. 2002;83:912-918.

(9) Deathe AB, Miller WC. The L Test of Functional Mobility: measurement properties of a modified version of the Timed "Up & Go" Test designed for people with lower-limb amputations. Phys Ther. 2005;85:626-635.

(10) Miller WC, Speechley M, Deathe AB. Balance confidence among people with lower-limb amputations. Phys Ther. 2002;82:856-865.

(11) Hertel R, Strebel N, Ganz R. Amputation versus reconstruction in traumatic defects of the leg: outcome and costs. J Orthop Trauma. 1996;10: 223-229.

(12) de Visser E, Pauwels J, Duysens JE, et al. Gait adaptations during walking under visual and cognitive constraints: a study of patients recovering from limb-saving surgery of the lower limb. Am J Phys Med Rehabil. 1998;77:503-509.

(13) Sadeghi H, Sadeghi S, Allard P, et al. Lower limb muscle power relationships in bilateral able-bodied gait. Am J Phys Med Rehabil. 2001;80: 821-830.

(14) Sadeghi H. Contributions of lower-limb muscle power in gait of people without impairments. Phys Ther. 2000;80:1188-1197.

(15) MacKenzie EJ, Bosse MJ, Kellam JF, et al. Characterization of patients with high-energy lower extremity trauma. J Orthop Trauma. 2000;14:455- 466.

(16) Gustilo RB, Mendoza RM, Williams DN. Problems in the management of type III (severe) open fractures: a new classification of type III open fractures. J Trauma. 1984;24:742-746.

(17) Bosse MJ, MacKenzie EJ, Kellam JF, et al. An analysis of outcomes of reconstruction or amputation after leg-threatening injuries. N Engl J Med. 2002;347:1924-1931.

(18) Archer KR, Castillo RC, MacKenzie EJ, et al. Physical disability after severe lower-extremity injury. Arch Phys Med Rehabil. 2006;87: 1153-1155.

(19) Kwakkel G, Wagenaar RC. Effect of duration of upper- and lower-extremity rehabilitation sessions and walking speed on recovery of interlimb coordination in hemiplegic gait. Phys Ther. 2002;82:432-448.

(20) van Loo MA, Moseley AM, Bosman JM, et al. Inter-rater reliability and concurrent validity of walking speed measurement after traumatic brain injury. Clin Rehabil. 2003;17:775-779.

(21) Eastlack ME, Arvidson J, Snyder-Mackler L, et al. Interrater reliability of videotaped observational gait-analysis assessments. Phys Ther. 1991;71:465-472.

(22) Brunnekreef JJ, van Uden C, van Moorsel S, Kooloos J. Reliability of videotaped observational gait analysis in patients with orthopedic impairments. BMC Musculoskelet Disord. 2005;6:17-25.

(23) Krebs DE, Edelstein JE, Fishman S. Reliability of observational kinematic gait analysis. Phys Ther. 1985;65:1027-1033.

(24) Toro B, Nester CJ, Farren PC. The status of gait assessment among physiotherapists in the United Kingdom. Arch Phys Med Rehabil. 2003; 84:1878-1884.

(25) Isakov E, Keren O, Benjuya N. Trans-tibial amputee gait: time-distance parameters and EMG activity. Prosthet Orthot Int. 2000;24: 216-220.

(26) Sadegbi H, Allard P, Duhaime M. Muscle power compensatory mechanisms in below-knee amputee gait. Am J Phys Med Rehabil. 2001;80:25-32.

(27) Neptune RR, Kautz SA, Zajac FE. Contributions of the individual ankle plantar flexors to support, forward progression and swing initiation during walking. J Biomech. 2001;34:1387-1398.

(28) Oberg T, Karsznia A, Oberg K. Basic gait parameters: reference data for normal subjects, 10-79 years of age. J Rehabil Res Dev. 1993;30: 210-223.

(29) Waters RL, Hislop HJ, Perry J, et al. Comparative cost of walking in young and old adults. J Orthop Res. 1983;1:73-76.

(30) Jensen MP, Chen C, Brugger AM. Interpretation of visual analog scale ratings and change scores: a reanalysis of two clinical trials of postoperative pain. J Pain. 2003;4:407-414.

(31) Schoppen T, Boonstra A, Groothoff JW, et al. Physical, mental, and social predictors of functional outcome in unilateral lower-limb amputees. Arch Phys Med Rehabil. 2003;84:803-811.

(32) Kegel B, Carpenter ML, Burgess EM. Functional capabilities of lower extremity amputees. Arch Phys Med Rehabil. 1978;59:109-120.

(33) Magee DJ. Orthopedic Physical Assessment. Philadelphia, Pa: WB Saunders Co; 1997.

(34) Ross M, Worrell TW. Thigh and calf girth following knee injury and surgery. J Orthop Sports Phys Ther. 1998;27:9-15.

(35) Soderberg GL, Ballantyne BT, Kestel LL. Reliability of lower extremity girth measurements after anterior cruciate ligament reconstruction. Physiother Res Int. 1996;1:7-16.

(36) Norkin CC, White DJ. Measurement of Joint Motion A Guide to Goniometry. 3rd ed. Philadelphia, Pa: FA Davis Co; 2003.

(37) Brosseau L, Balmer S, Tonsignant M, et al. Intra- and intertester reliability and criterion validity of the parallelogram and universal goniometers for measuring maximum active knee flexion and extension of patients with knee restrictions. Arch Phys Med Rehabil. 2001;82: 396-402.

(38) Youdas JW, Bogard CL, Suman VJ. Reliability of goniometric measurements and visual estimates of ankle joint active range of motion obtained in a clinical setting. Arch Phys Med Rehabil. 1993;74: 111-118.

(39) Joint Motion: Method of Measuring and Recording. Chicago, Ill: American Academy of Orthopaedic Surgeons; 1965.

(40) Peindl RD, McCarthy ML, MacKenzie EJ, inventors; The Johns Hopkins University, The Charlotte-Mecklenburg Hospital Authority, assignee. Apparatus for exercising and measuring strength of a patient's limb and an adjustable pivot clamp. US patent 5662591. September 2, 1997.

(41) Kendall FP, McCreary EK, Provance PG. Muscles: Testing and Function. 4th ed. Baltimore, Md: Williams & Wilkins; 1993.

(42) Haigh R, Tennant A, Biering-Sorensen F, et al. The use of outcome measures in physical medicine and rehabilitation within Europe. J Rehabil Med. 2001;33:273-278.

(43) Kendall FP, McCreary EK. Muscles: Testing and Function. Baltimore, Md: Williams & Wilkins; 1983.

(44) McCarthy ML, McAndrew MP, MacKenzie EJ, et al. Correlation between the measures of impairment, according to the modified system of the American Medical Association, and function. J Bone Joint Surg Am. 1998;80:1034-1042.

(45) Perry J, Ireland ML, Gronley J, Hoffer MM. Predictive value of manual muscle testing and gait analysis in normal ankles by dynamic electromyography. Foot Ankle. 1986;6:254-259.

(46) Silbernagel KG, Thomee R, Thomee P, Karlsson J. Eccentric overload training for patients with chronic Achilles tendon pain: a randomised controlled study with reliability testing of the evaluation methods. Scand J Med Sci Sports. 2001;11:197-206.

(47) de Visser E, Deckers JA, Veth R, et al. Deterioration of balance control after limb-saving surgery. Am J Phys Med Rehabil. 2001;80: 358-365.

(48) Bendetti MG, Catani F, Donati D, et al. Muscle performance about the knee joint in patients who had distal femoral replacement after resection of a bone tumor: an objective study with use of gait analysis. J Bone Joint Surg Am. 2000;82:1619-1625.

(49) Sadeghi H, Sadeghi S, Prince F, et al. Functional roles of ankle and hip sagittal muscle moments in able-bodied gait. Clin Biomech. 2001; 16:688-695.

(50) Georgiadis GM, Behrens FF, Joyce MJ, et al. Open tibial fractures with severe soft-tissue loss: limb salvage compared with below-the-knee amputation. J Bone Joint Surg Am. 1993;75:1431-1441.

(51) Pelissier P, Boireau P, Martin D, Baudet J. Bone reconstruction of the lower extremity: complications and outcomes. Plast Reconstr Surg. 2003;111:2223-2229.

* Stata Corp, 4905 Lakeway Dr, College Station, TX 77845.

Kristin R Archer, Renan C Castillo, Ellen J MacKenzie, Michael J Bosse, and the Lower Extremity Assessment Project (LEAP) Study Group

KR Archer, PT, MS, DPT, is a PhD candidate, Center for Injury Research and Policy, Bloomberg School of Public Health, Johns Hopkins University, 624 North Broadway, Room 545, Baltimore, MD 21205 (USA). Address all correspondence to Dr Archer at: [e-mail].

RC Castillo, MS, is Assistant Scientist, Center for Injury Research and Policy, Department of Health Policy and Management, Bloomberg School of Public Health, Johns Hopkins University.

EJ MacKenzie, PhD, is Fred and Julie Soper Professor and Chair, Department of Health Policy and Management, Bloomberg School of Public Health, Johns Hopkins University.

MJ Bosse, MD, is Director and Clinical Research and Orthopaedic Traumatologist, Department of Orthopaedic Surgery, Carolinas Medical Center, Charlotte, NC.

The LEAP Study Group is: Ellen J MacKenzie, PhD; Michael J Bosse, MD;James F Kellam, MD; Andrew R Burgess, MD; Lawrence X Webb, MD; Marc F Swiontkowski, MD; Roy Sanders, MD; Alan L Jones, MD; Mark P McAndrew, MD; Brendan Patterson, MD; Melissa L McCarthy, ScD; Thomas G Travison, PhD; and Renan C Castillo, MS.

Dr Archer, Mr Castillo, and Dr MacKenzie provided concept/idea/research design and data analysis. Dr Archer and Mr Castillo provided writing. Dr MacKenzie and Dr Bosse provided data collection. Dr Bosse provided project management, fund procurement, and subjects. The authors acknowledge the tireless efforts of the co-investigators, study coordinators, and physical therapists at each of the 8 LEAP study sites. Their dedication to the study's objectives and their commitment to high-quality data collection were essential to the success of the study. They include: Julie Agel, ATC; Jennifer Avery, PT; Denise Bailey, PT; Wendall Bryan; Debbie Bullard; Carla Carpenter, PT; Elizabeth Chaparro, RN; Kate Corbin; Denise Darnell, RN, BSN; Stephanie Dickason, PT; Thomas DiPasquale, DO; Betty Harkin, PT; Michael Harrington, PT; Dolfi Herscovici, DO; Amy Holdren, RNC, ANP, MSN; Linda Howard, PT; Sarah Hutchings, BS; Marie Johnson, LPN; Melissa Jurewicz, PT; Donna Lampke, PT; Karen Lee, RN; Marianne Mars, PT; Maxine Mendoza-Welch, PA;J Wayne Meredith, MD; Nan Morris, PT; Karen Murdock, PT; Andrew Pollak, MD; Pat Radey, RN; Sandy Shelton, PT; Sherry Simpson, PT; Steven Sims, MD; Douglas Smith, MD; Adam Starr, MD; Celia Wiegman, RN; John Wilber, MD; Stephanie Williams, PA; Philip Wolinsky, MD; Mary Woodman, BA; and Michelle Zimmerman, RN. Dr Archer, Mr Castillo, Dr MacKenzie, Dr Bosse, Dr Webb, Dr McCarthy, Dr Swiontkowski, Dr Kellam, Dr Burgess, Dr Sanders, Dr Jones, Dr McAndrew, Dr Patterson, and Dr Travison provided consultation (including review of manuscript before submission).

This study was approved by the institutional review boards at the coordinating center (Johns Hopkins University, Bloomberg School of Public Health) and each LEAP study site.

This research was supported with funds from the Johns Hopkins Center for Injury Research and Policy and National Center for Injury Prevention and Control, Centers for Disease Control and Prevention (grant no. CE000198-03), and the National Institute of Arthritis, Musculoskeletal and Skin Diseases, National Institutes of Health (grant no. RO1-AR42659).

Table 1.

Baseline Demographic and Injury Characteristics of Study
Sample by Main Outcome at 24
Months (N=381)

                                                      Gait
Variable                           n                  Deviation

Age (y)
  0-24                              82                51%
  25-34                            106                48%
  35-44                            105                53%
  [greater than or equal to] 45     88                63%

Sex
  Male                             285                54%
  Female                            96                53%

Race/ethnicity
  White                            271                53%
  Black                             80                52%
  Hispanic                          19                58%
  Other                             11                67%

Education (y)
  <12                              262                55%
  [greater than or equal to] 12    119                51%

Insurance
  None                             146                48%
  Medicaid                          25                59%
  Medicare                           4                67%
  Health maintenance                46                55%
    organization
  Commercial                       148                59%
  Other                             12                33%

Injury
  Dysvascular                       70                52%
  Foot/ankle/tibia                 114                49%
  Tibial fractures                 197                57%

Articular involvement
  No                               246                38%
  Yes                              135                41%

Nerve damage
  No                               358                38%
  Yes                               23                52%

Muscle damage
  Mild                             198                37%
  Moderate-severe                  183                41%

Bone loss
  <2 cm                            301                38%
  [greater than or equal to] 2 cm   80                41%

                                   Walking            Gait Deviation
Variable                           Speed [less than   and Walking Speed
                                   or equal to] 4     [less than or
                                   ft/s               equal to]4 ft/s

Age (y)
  0-24                             19%                21%
  25-34                            20%                17%
  35-44                            44%                30%
  [greater than or equal to] 45    43%                38%

Sex
  Male                             30%                25%
  Female                           38%                32%

Race/ethnicity
  White                            34%                28%
  Black                            31%                24%
  Hispanic                          8%                 8%
  Other                            33%                33%

Education (y)
  <12                              32%                24%
  [greater than or equal to] 12    32%                31%

Insurance
  None                             32%                24%
  Medicaid                         35%                29%
  Medicare                          0%                 0%
  Health maintenance               25%                22%
    organization
  Commercial                       34%                32%
  Other                            33%                11

Injury
  Dysvascular                      30%                22%
  Foot/ankle/tibia                 26%                21%
  Tibial fractures                 37%                32%

Articular involvement
  No                               22%                18%
  Yes                              26%                21%

Nerve damage
  No                               23%                19%
  Yes                              30%                30%

Muscle damage
  Mild                             24%                19%
  Moderate-severe                  23%                20%

Bone loss
  <2 cm                            22%                19%
  [greater than or equal to] 2 cm  27%                22%

Table 2.
Clinical Variables of Study Sample at 24 Months by Main
Outcome (N=381) (a)

                                                          Gait
Clinical Variable                          n              Deviation

VAS for pain
  No pain                                   57            32%
  Mild pain                                158            55%
  Moderate-severe pain                      73            66%

Ascend stairs
  Reciprocal                               212            48%
  Nonreciprocal                             74            69%

Descend stairs
  Reciprocal                               186            46%
  Nonreciprocal                            100            68%

Unilateral stance: involved side
  [greater than or equal to] 30 s           83            34%
  <30s                                     204            62%

Unilateral stance: uninvolved side
  [greater than or equal to] 30s           183            50%
  <30s                                     104            61%

Tandem standing balance
  [greater than or equal to] 10s           108            41%
  <10 s                                    179            61%

Toe raises: involved side
  [greater than or equal to] 10             86            27%
  <10                                      199            65%

Toe raises: uninvolved side
  [greater than or equal to] 10            141            43%
  <10                                      144            65%

Thigh atrophy
  0 cm                                     118            48%
  1-2 cm                                   114            56%
  [greater than or equal to] 3 cm           51            62%

Calf atrophy
  0 cm                                     122            47%
  1-2 cm                                    90            49%
  [greater than or equal to] 3 cm           71            70%

Hip flexion ROM: involved side
  [greater than or equal to] 120[degrees]  266            51%
  <120[degrees]                             24            83%

Hip flexion ROM: uninvolved side
  [greater than or equal to] 120[degrees]  275            52%
  <120                                      15            80%

Hip extension ROM: involved side
  [greater than or equal to] 30[degrees]   262            52%
  <30[degrees]                              28            69%

Hip extension ROM: uninvolved side
  [greater than or equal to] 30[degrees]   271            52%
  <30[degrees]                              19            82%

Knee flexion ROM: involved side
  [greater than or equal to] 135[degrees]  224            52%
  <135[degrees]                             66            69%

Knee flexion ROM: uninvolved side
  [greater than or equal to] 135[degrees]  272            53%
  <135[degrees]                             18            65%

Ankle DF ROM: involved side
  [greater than or equal to] 20[degrees]    93            39%
  <20[degrees]                             197            60%

Ankle DF ROM: uninvolved side
  [greater than or equal to] 20[degrees]   222            52%
  <20[degrees]                              68            58%

Ankle PF ROM: involved side
  [greater than or equal to] 50[degrees]   135            39%
  <50[degrees]                             155            66%

Ankle PF ROM: uninvolved side
  [greater than or equal to] 50[degrees]   264            52%
  <50[degrees]                              26            65%

Hip flexion strength
  [greater than or equal to] 4             221            51%
  <4                                        12            82%

Hip extension strength
  [greater than or equal to] 4             197            49%
  >4                                        36            70%

Hip abduction strength
  [greater than or equal to] 4             219            50%
  <4                                        14            92%

Knee flexion strength
  [greater than or equal to] 4             191            48%
  <4                                        42            72%

Knee extension strength
  [greater than or equal to] 4             185            50%
  <4                                        48            62%

Ankle DF strength
  [greater than or equal to] 4             141            42%
  <4                                        92            68%

Ankle PF strength
  [greater than or equal to] 4             160            44%
  <4                                        73            69%

                                                          Gait
                                                          Deviation
                                           Walking Speed  and Walking
                                           [less than or  Speed [less
                                           equal to]      than or equal
Clinical Variable                          4 ft/s         to] 4 ft/s

VAS for pain
  No pain                                  23%            21%
  Mild pain                                29%            24%
  Moderate-severe pain                     44%            37%

Ascend stairs
  Reciprocal                               23%            20%
  Nonreciprocal                            57%            45%

Descend stairs
  Reciprocal                               21%            18%
  Nonreciprocal                            52%            42%

Unilateral stance: involved side
  [greater than or equal to] 30 s           7%             6%
  <30s                                     42%            35%

Unilateral stance: uninvolved side
  [greater than or equal to] 30s           20%            18%
  <30s                                     53%            42%

Tandem standing balance
  [greater than or equal to] 10s           13%            15%
  <10 s                                    44%            34%

Toe raises: involved side
  [greater than or equal to] 10             8%             7%
  <10                                      43%            35%

Toe raises: uninvolved side
  [greater than or equal to] 10            22%            19%
  <10                                      43%            34%

Thigh atrophy
  0 cm                                     31%            24%
  1-2 cm                                   30%            26%
  [greater than or equal to] 3 cm          39%            33%

Calf atrophy
  0 cm                                     28%            25%
  1-2 cm                                   20%            17%
  [greater than or equal to] 3 cm          54%            41%

Hip flexion ROM: involved side
  [greater than or equal to] 120[degrees]  30%            51%
  <120[degrees]                            56%            83%

Hip flexion ROM: uninvolved side
  [greater than or equal to] 120[degrees]  31%            26%
  <120                                     47%            40%

Hip extension ROM: involved side
  [greater than or equal to] 30[degrees]   30%            25%
  <30[degrees]                             54%            38%

Hip extension ROM: uninvolved side
  [greater than or equal to] 30[degrees]   31%            26%
  <30[degrees]                             47%            41%

Knee flexion ROM: involved side
  [greater than or equal to] 135[degrees]  30%            24%
  <135[degrees]                            54%            37%

Knee flexion ROM: uninvolved side
  [greater than or equal to] 135[degrees]  32%            26%
  <135[degrees]                            35%            35%

Ankle DF ROM: involved side
  [greater than or equal to] 20[degrees]   22%            18%
  <20[degrees]                             37%            30%

Ankle DF ROM: uninvolved side
  [greater than or equal to] 20[degrees]   30%            25%
  <20[degrees]                             37%            31%

Ankle PF ROM: involved side
  [greater than or equal to] 50[degrees]   18%            16%
  <50[degrees]                             44%            35%

Ankle PF ROM: uninvolved side
  [greater than or equal to] 50[degrees]   31%            26%
  <50[degrees]                             38%            31%

Hip flexion strength
  [greater than or equal to] 4             31%            24%
  <4                                       64%            54%

Hip extension strength
  [greater than or equal to] 4             28%            21%
  >4                                       56%            47%

Hip abduction strength
  [greater than or equal to] 4             31%            24%
  <4                                       54%            46%

Knee flexion strength
  [greater than or equal to] 4             28%            20%
  <4                                       55%            50%

Knee extension strength
  [greater than or equal to] 4             27%            21%
  <4                                       55%            42%

Ankle DF strength
  [greater than or equal to] 4             23%            18%
  <4                                       47%            36%

Ankle PF strength
  [greater than or equal to] 4             23%            18%
  <4                                       55%            41%

(a) VAS=visual analog scale, ROM-range of motion, strength
testing=0-5 scale, DF=dorsiflexion, PF=plantar flexion.

Table 3.

Bivariate Regressions of Clinical Variables by Gait Deviation
and Walking Speed of [less than or equal to] 4 ft/s at 24
Months (n = 277) (a)

Clinical Variable                                OR    95% CI     P

Mild pain                                        1.2   0.57-2.6    .61
Moderate-severe pain                             2.2   0.97-5.0    .06
Nonreciprocal stair-climbing: ascend             3.4    1.9-6.1   <.001
Nonreciprocal stair-climbing: descend            3.2    1.8-5.7   <.001
<30 s unilateral stance: involved side           8.2    3.2-21.2  <.001
<30 s unilateral stance: uninvolved side         3.2    1.9-5.7   <.001
<10 s tandem standing balance                    2.9    1.5-5.3    .001
<10 toe raises: involved side                    7.1    2.9-17.1  <.001
<10 toe raises: uninvolved side                  2.3    1.3-3.9    .004
1-2 cm thigh atrophy                             1.1   0.60-2.0    .76
>3 cm thigh atrophy                              1.5   0.74-3.2    .24
1-2 cm calf atrophy                              0.63  0.32-1.3    .20
>3 cm calf atrophy                               2.1    1.1-4.0    .02
<120[degrees] hip flexion ROM: involved side     2.8    1.2-6.6    .02
<120[degrees] hip flexion ROM: uninvolved side   1.9   0.65-5.5    .24
<30[degrees] hip extension ROM: involved side    1.8   0.79-4.2    .16
<30[degrees] hip extension ROM: uninvolved side  2.0   0.74-5.5    .17
<135[degrees] knee flexion ROM: involved side    1.8    1.0-3.4    .05
<135[degrees] knee flexion ROM: uninvolved side  1.5   0.55-4.3    .82
<20[degrees] ankle DF ROM: involved side         1.9    1.0-3.6    .04
<20[degrees] ankle DF ROM: uninvolved side       1.3   0.74-2.5    .33
<50[degrees] ankle PF ROM: involved side         2.7    1.5-4.9    .001
<50[degrees] ankle PF ROM: uninvolved side       1.2   0.52-3.0    .62
<4 hip flexion strength                          3.8    1.1-13.1   .03
<4 hip extension strength                        3.2    1.5-6.9    .002
<4 hip abduction strength                        2.7   0.08-8.4    .08
<4 knee flexion strength                         4.0    1.9-8.2   <.001
<4 knee extension strength                       2.8    1.4-5.6    .003
<4 ankle DF strength                             2.5    1.3-4.5    .003
<4 ankle PF strength                             3.1    1.7-5.8   <.001

(a) OR=odds ratio, CI=confidence interval, ROM=range of motion,
strength testing=0-5 scale, DF=dorsiflexion, PF=plantar flexion.

Table 4.
Multivariate Logistic Regression Model of Gait Deviation
and Walking Speed of [less than or equal to] 4 ft/s at
24 Months (n=277) (a)

Variable                                  OR    95% CI     P

Age (y)
  0-24 (ref)                              1.0
  25-34                                   0.88  0.31-2.5   .81
  35-44                                   1.3   0.47-3.7   .60
  [greater than or equal to] 45           2.2   0.77-6.3   .14

Insurance
  None (ref)                              1.0
  Commercial/Medicare                     1.4   0.64-3.1   .41
  Other                                   1.3   0.51-3.1   .62

Injury
  Dysvascular (ref)                       1.0
  Foot/ankle/tibia                        0.39  0.13-1.1   .08
  Tibial fractures                        0.89  0.36-2.2   .81

Ascend/descend stairs
  Reciprocal (ref)                        1.0
  Nonreciprocal                           2.7    1.4-5.3   .003

Unilateral stance: involved limb
  [greater than or equal to] 30 s (ref)   1.0
  <30 s                                   3.1    1.0-9.7   .05

Unilateral stance: uninvolved limb
  [greater than or equal to] 30 s (ref)   1.0
  <30 s                                   2.0   0.99-4.0   .05

Hip flexion strength
  [greater than or equal to] 4            1.0
  <4                                      3.0   0.61-14.7  .18

Hip extension strength
  [greater than or equal to] 4            1.0
  <4                                      2.3   0.89-6.0   .09

Knee flexion strength
  [greater than or equal to] 4            1.0
  <4                                      2.8    1.1-6.7   .03

Ankle PF ROM: involved side
  [greater than or equal to] 50[degrees]  1.0
  <50[degrees]                            3.2    1.5-6.8   .002

(a) Model includes all variables in table, as well as missing
categories for balance, strength, range of motion, and stairs.
CI=confidence interval, ref=reference, PF=plantar flexion,
ROM=range of motion.