Gait & Posture 27 (2008) 518–529 www.elsevier.com/locate/gaitpost

Elderly unilateral transtibial amputee gait on an inclined walkway: A biomechanical analysis Deborah R. Vickers *, C. Palk, A.S. McIntosh, K.T. Beatty School of Safety Science, The University of New South Wales, Australia Received 20 October 2006; received in revised form 21 June 2007; accepted 28 June 2007

Abstract The greatest population of amputees in developed nations are elderly dysvascular transtibial amputees. Conventional prostheses, e.g. the solid ankle cushioned heel (SACH) foot, create difficulties in walking on inclines. The aim of this study was to analyse the gait characteristics of elderly amputees walking on an incline, through quantitative three-dimensional biomechanical analysis, by comparing them to agematched controls. Participants walked up and down an inclined (58) instrumented walkway at a self-selected pace. A ViconTM System 370 was used to acquire gait data, including temporo-spatial characteristics, ground reaction forces (GRF), electromyography (EMG), kinematics, and kinetics of the lower limb. Compared to the age-matched controls, the amputees demonstrated reduced speed, knee and hip range of motion, hip moments, vertical GRF, along with increased amplitude and periods of muscle activation. The residual limb also had shorter single support stance phase, small stance phase knee moments, and the smallest moments and powers. These differences demonstrate instability in stance of the residual limb. The sources of this instability include the prosthesis’ limited range of ankle motion and ankle power generation, coupled with the residual limb’s limited proprioception and tolerance of force. For these amputees to regain a gait pattern equivalent to their able-bodied counterparts on inclined walkways, they must be equipped with a prosthesis that has a full range of ankle motion and active power generation at the ankle. Prosthesis design and rehabilitation training should also improve the proprioception of their residual limb and increase their tolerance of force through the residual limb. Crown Copyright # 2007 Published by Elsevier B.V. All rights reserved. Keywords: Gait; Inclined walkway; Kinetics; SACH; Transtibial amputee

1. Introduction The largest amputee population, in developed nations, are the elderly with peripheral vascular disease resulting in Abbreviations: A, power absorption; AM, ankle moment; AP, ankle power; A/P, anterior/posterior GRF; ASIS, anterior superior iliac spine; CMC, coefficient of multiple correlation; DF, dorsiflexion; E, extension; EMG, electromyography; F, flexion; G, power generation; GRF, ground reaction force; HM, hip moment; HP, hip power; KM, knee moment; KP, knee power; MVC, maximum voluntary contraction; PF, plantar-flexion; SA, single axis; SACH, solid ankle cushioned heel; S.D., standard deviation; TTA, transtibial amputee; UNSW, University of New South Wales; V, vertical GRF * Corresponding author. Tel.: +61 2 9385 5413; fax: +61 2 9385 6190. E-mail address: [email protected] (Deborah R. Vickers).

transtibial amputation [1,2]. They can potentially ambulate independently, however, age and health problems may hinder their walking ability [1–3]. Although there are many advanced prostheses available, the elderly are often prescribed conventional prosthetic feet, due to cost and lifestyle limitations. The most commonly prescribed prosthetic foot in the USA is the solid ankle cushion heel (SACH) foot [4]. Neither the SACH foot, nor the single axis foot (another conventional prosthetic foot), match natural foot and ankle function [5]. The literature shows that ambulation of transtibial amputees on level surfaces differs from the normal population. With a prosthetic foot and ankle, the amputee lacks active dorsiflexion, plantar-flexion, strength, power, and proprioception resulting in gait abnormalities [6,7]. The

0966-6362/$ – see front matter. Crown Copyright # 2007 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2007.06.008

D.R. Vickers et al. / Gait & Posture 27 (2008) 518–529

limitations of prosthetic feet affect the residual limb knee and hip mechanics and require compensation by the sound limb [5]. The SACH foot yields to posterior loads to produce a form of plantar-flexion during loading response, but the rigid ankle prevents rapid plantar-flexion at toe-off [5]. It has inadequate heel strike restraint and push-off propulsion [5,7]. To compensate, the residual limb requires increased hip and knee flexion during stance, and the sound limb requires larger hip extension, knee flexion, and ankle dorsiflexion, than in able-bodied walking [8]. For the transtibial amputee, EMG activity of the quadriceps and hamstrings is more intense and prolonged throughout the gait cycle [9]. Vastus lateralis has increased activity to compensate for the foot instability [9]. These muscles stabilise the knee of the residual limb, and the prosthetic foot [9]. For stability, elderly amputees cocontract hamstrings and quadriceps, therefore, these muscles contribute little towards propulsion. In a study of different prosthetic feet, there were no significant differences in EMG activity for the sound limb, but there were differences for the residual limb according to the foot selected [10]. The residual limb has reduced muscle volume and strength, following amputation [11], therefore, the sound limb muscles dominate in maintaining stability [11]. Due to the longer heel-only contact and load-bearing phase, the residual limb has longer muscle activation periods [9]. These gait compensations require increased muscular effort and energy expenditure [9]. The study of gait in normal and disabled populations has focussed on horizontal walking [12], however, daily activities often require walking on inclined surfaces. In

519

particular, inclined walkways are commonly used in aged care facilities for improved access. Of those studies that have considered walking on an incline [5,13], few have assessed amputees. Anecdotal evidence from amputees indicates that it is difficult to walk on ramps with conventional feet, and that descending is more demanding than ascending [14,15]. Pandian and Kowalske [16] claimed that transtibial amputees could walk on ramps without difficulty, but still recommended diagonal walking, or sidestepping on steep inclines. Buczek et al. [17] found that disabled populations (including amputees) required greater slip resistance than able-bodied walkers on an incline. In Macfarlane et al.’s [14] comparison of flex-feet and conventional feet, there was no significant interaction between foot type and grade, however, no comparisons were made to normal feet. Current research needs in amputee gait, which would assist in prosthetic design and rehabilitative gait training, are:  quantification of gait characteristics which inhibit proficient walking [5,8],  consideration of an active lifestyle and activities of daily living (ADL) [5,6],  kinematic and kinetic analysis of amputee gait at different speeds, and on inclines [8,17], and  identification of compensatory strategies [5]. The aim of this study was to identify the differences in elderly gait on an incline between unilateral transtibial amputees and able-bodied walkers. By quantifying the effects of a conventional prosthetic foot, on gait dynamics on

Table 1 General characteristics of all participants Participant no.

Age (years)

Amputee participants 1 68 2 77 3 68 4 77 5 75 6 59 7 68 8 76 Mean (S.D.)

Cause of amputation

Side

Years since amputation

Foot type

Height (cm)

Weight (kg)

F M M M M M F F

Vascular Trauma Vascular Vascular Vascular Vascular Vascular Cancer

Right Right Left Left Right Right Right Left

1 6 1 1 2.5 0.5 1 4

SA SACH SACH SA SACH SACH SACH SACH

151.0 173.3 172.0 177.0 162.8 165.8 164.3 147.7

61.2 63.8 102.3 84.6 76.3 69.1 81.4 71.3

71 (6.0)

Able-bodied participants 9 71 10 51 11 74 12 80 13 74 14 72 15 65 16 70 Mean (S.D.)

Gender

2.1 (1.8) F F M M M M F M

69.6 (8.6)

SA, single axis; SACH, solid ankle cushioned heel.

164.2 (9.7) 156.5 171.0 169.2 179.3 176.6 174.0 163.5 178.5 171.1 (7.9)

76.3 (12.4) 59.5 66.7 77.2 95.7 76.6 88.1 73.0 104.7 80.2 (15.1)

520

D.R. Vickers et al. / Gait & Posture 27 (2008) 518–529

Fig. 1. Diagram of the inclined walkway set-up in the UNSW School of Safety Science’s Biomechanics and Gait Laboratory.

an incline, the source and compensations of the limitations can be identified. These findings will hopefully lead to new insights into prosthetic design and amputee rehabilitation.

2. Methods 2.1. Subject sample Five male and 3 female elderly subjects with unilateral transtibial amputations were recruited from Prince Henry Hospital, a teaching hospital of the University of New South Wales (UNSW). The selection criteria were unilateral transtibial amputation, well adapted to prosthesis with no stump pain while walking, and able to ambulate 20 m independently. Six participants had SACH feet and two had single axis ankles. Eight able-bodied subjects were recruited from the Sydney Metropolitan Region to provide age and gender-matched controls. They had no history of lower limb arthroplasty. Participant details are presented in Table 1. Due to equipment failure, subject 10’s EMG data could not be analysed, but kinematic and kinetic data were included. The study was approved by UNSW human research ethics committee (CEPIHS no. 98196). All subjects provided written informed consent.

with an MA-200 system. The global coordinate axis system was maintained irrespective of the incline angle. Anthropometric data were measured (Table 1) and optical markers placed bilaterally on each subject (Fig. 2). Surface electrodes were placed on gluteus maximus, vastus lateralis, and lateral hamstrings bilaterally. The skin was prepared by shaving and wiping with sterilising alcohol. Maximum voluntary contractions (MVCs) were measured for these muscles for approximately 6 s each. For gluteus maximus and lateral hamstrings MVCs, subjects lay prone with their knee flexed to 458, and resisted an upward force applied close to the knee and to the ankle, respectively. Vastus Lateralis MVC was measured with the patient supine, with their knee slightly flexed, resisting a downward force applied close to the ankle. The subjects first walked up and down the incline to identify a comfortable speed. Subjects walked in their own shoes, at a selfselected pace. They ascended and descended the incline until three

2.2. Description of the inclinable walkway The study was undertaken at UNSW’s Biomechanics and Gait Laboratory on an instrumented, inclinable walkway. The 1.2 m wide inclined walkway was 7 m long with 3 m horizontal areas at either end (Fig. 1). The walkway was inclined at 58. The steel framed walkway had a vinyl covered plywood floor. A Kistler force platform (model 9284) was secured to the steel frame in the middle of the walkway, at the same incline angle. Guardrails were around the walkway for safety. 2.3. Measurement and analysis of gait dynamics AViconTM System 370 (Oxford Metrics, Oxford, UK) was used to acquire marker coordinates and for data collection. Video data were recorded at 50 Hz and analog data at 1000 Hz. Muscle activation was measured using surface electromyography (EMG)

Fig. 2. Anatomical locations of optical markers. Markers were also placed on the sacrum and heels.

D.R. Vickers et al. / Gait & Posture 27 (2008) 518–529

521

Table 2 Temporo-spatial data Mean (S.D.)

Ascending

Speed (m/s) Cadence (strides/s) Stride length (m) Toe-off (% of cycle) Single support timea (% of cycle)

Descending

Controls

Amputees

Averaged L and R

Sound limb

1.40 (0.13) 0.97 (0.08) 1.45 (0.16) 65.8 (2.2) 35 a

Residual limb *

0.71 0.72* 0.98* 71.9* (2.8) 34a

(0.15) (0.06) (0.16) 68.1* (3.8) 29a

Controls

Amputees

Averaged L and R

Sound limb

1.42 (0.13) 1.01 (0.08) 1.41 (0.18) 64.1 (1.7) 36 a

Residual limb *

0.66 0.76* 0.88* 72.9* (3.9) 33 a

(0.16) (0.08) (0.21) 68.0* (3.9) 29 a

Speed, cadence and stride length are measured for each subject, not for individual limbs. a Averages calculated in VCM, standard deviations unavailable, therefore unable to test for statistical significance. * Significant difference to the control subjects.

force plate strikes were captured for each limb, in each direction. For data analysis, of each subject, kinematics and kinetics were averaged over two walks, and EMG was averaged over six walks. For the residual and sound limb groups, eight limbs were averaged. For the controls, the left and right limb data were included which resulted in 16 limbs being averaged. Marker coordinate and GRF data were processed in Vicon Clinical Manager (Oxford Metrics, version 1.34) to calculate the gait dynamics, time-normalised and averaged against body weight. The coordinates of optical markers defined the coordinate axis system, for each limb segment. The limb segments were foot, shank, thigh and pelvis. The segment and joint kinematics were calculated according to Davis et al. [18] and Kadaba et al. [19]. The internal net muscle moments and powers at the ankle, knee, and hip were derived using inverse dynamics [18]. The ankle moments and powers of the prosthetic limb are only an indication of the mechanical resistance in the prosthesis. The inverse dynamics process assumed that the prosthetic limb behaved in the same manner to the human ankle. This may have caused some error in calculation of the knee and hip, moments and powers for the residual limb. GRFs were measured by the inclined force platform, and converted into vertical and horizontal components in the global coordinate axis system.

2.4. Variables Data are presented for the controls, averaged across both legs, and for the amputees’ sound and residual limbs separately. Biomechanical variables (Tables 2, 3, 5–8) were speed, cadence, stride length, single support time, vertical and anterior/posterior GRF, angles, moments and powers at the ankle, knee, and hip, and muscle activation. Variables were compared between the two amputee limb groups and the control group, at selected events in the gait cycle, e.g. heel strike, 50% of the gait cycle. These events were chosen to coincide with turning points of the curves. Eighteen percent MVC was the muscle activity threshold [20]. Required coefficient of friction (RCOF) was calculated by dividing shear GRF by perpendicular GRF [21]. The vertical and horizontal forces of Fig. 4 were converted to perpendicular and shear forces before calculating the RCOFs. RCOF quantifies the direction of the GRF force applied by the limb, and is an indicator of the likelihood of a slip [21]. 2.5. Statistics Coefficients of multiple correlations (CMC) [22] were calculated for kinematics, kinetics and EMG, to assess data repeatability

Table 3 Summary of sagittal plane kinematics Mean (S.D.)

Ascending

Descending

Controls

Amputees

Controls

Amputees

Averaged L and R

Sound limb

Residual limb

Averaged L and R

Sound limb

Residual limb

Ankle—at events in gait cycle Heel strike 10 DF (2) 50% 20 DF (2) Toe-off 5 PF (2)

10 DF (3) 22 DF (3) 5* DF (3)

7* DF (3) 14* DF (3) 7* DF (3)

10 DF (3) 20 DF (3) 5 DF (3)

5* DF (5) 18 DF (5) 12* DF (5)

7 DF (4) 12* DF (3) 7 DF (3)

Knee—at events in gait cycle Heel strike 25 F (5) 50% 5 F (5) Max swing 60 F (5)

25 F (7) 10 F (7) 55 F (7)

20 F (7) 10 F (7) 58 F (7)

10 F (5) 15 F (5) 70 F (5)

10 F (10) 18 F (10) 60* F (10)

15 F (10) 18 F (10) 63 F (10)

Hip—at events in gait cycle Heel strike 55 F (5) 50% 5 E (5)

50 F (10) 5* F (10)

50 F (8) 10* F (8)

30 F (5) 5 E (5)

30 F (10) 6* F (10)

32 F (10) 8* F (10)

Units are degrees. 50% = 50% of gait cycle; DF, dorsi-flexion; E, extension; F, flexion; PF, plantar-flexion. Ankle angles of the prosthetic limb are only an indicator of mechanical resistance in the prosthetic foot and ankle, and should be interpreted with care. * Significant difference to the control subjects.

522

D.R. Vickers et al. / Gait & Posture 27 (2008) 518–529

between subjects. Within-subject CMCs are presented for the EMG data, but were unavailable for kinematics and kinetics due to averaging in VCM. Descriptive statistics are reported for the amputee group’s sound and residual limbs and the control group. Statistical significance ( p < 0.05) was assessed using a two-sample Student’s t-test [23], comparing the sound and residual limbs to the controls. ANOVAs were performed on the EMG timing of activation to compare between all three groups using SPSS 14.0 for Windows (Table 8).

3. Results 3.1. Temporo-spatial data Temporo-spatial data (Table 2) were consistent with previous research on walking on inclines [13]. In both directions, the amputees had significantly reduced stride length and cadence, which resulted in walking half as fast as the controls. The sound limb had a similar single support time to the controls, however, the residual limb had a shortened single support time (Table 2). 3.2. Joint kinematics Joint kinematics of the ankle, knee, and hip in the sagittal plane in Fig. 3 are summarised for selected gait events in Table 3. Between-subject CMCs (Table 4) show better repeatability for the controls (mean CMC = 0.89), than for the sound (mean CMC = 0.77), and residual limbs (mean CMC = 0.66).

At 50% of the gait cycle, ascending the incline, the sound limb was in 58 hip flexion, compared to 58 hip extension for the control (Table 3). At toe-off, ascending the incline, the sound limb ankle was in 58 dorsiflexion, compared to 58 plantar-flexion for the control (Table 3). The prosthesis had a reduced ankle motion range, compared to the controls and the sound limb, and had no plantar-flexion (Fig. 3). Descending the incline, the sound limb’s ankle dorsiflexion was reduced at heel strike and increased at toe-off, compared to the controls. The sound limb had reduced knee flexion at maximum swing, compared to the controls (Fig. 3). The sound limb had 68 hip flexion at 50% of the gait cycle, compared to 58 extension for the control, whereas the residual limb had 88 hip flexion (Table 3). 3.3. Ground reaction forces Time-histories for vertical and anterior–posterior GRFs are presented in Fig. 4 and their peaks summarised in Table 5. Table 6 lists the peak RCOFs during loading response and terminal stance. The control’s vertical GRF had a two-peak pattern, whereas the amputee had a flattened pattern bilaterally [24]. Therefore, their values for V1, V2, and V3 were similar, instead of V2 being much lower than V1 and V3 (Fig. 4). The residual limb had a significantly decreased vertical GRF at V1, and a decreased posterior GRF at A/P1, during weight acceptance (Fig. 4).

Fig. 3. Hip flexion, knee flexion and ankle flexion: time-normalised and averaged for all participants (sagittal plane). (a) Hip flexion, ascending; (b) knee flexion, ascending; (c) ankle flexion, ascending; (d) hip flexion, descending; (e) knee flexion, descending; (f) ankle flexion, descending.

D.R. Vickers et al. / Gait & Posture 27 (2008) 518–529

523

Table 4 Between-subject coefficient of multiple correlation (CMC) for kinematics and kinetics Kinematic/Kinetic

Ascending

Descending

Controls

Amputees

Controls

Amputees

Averaged L and R

Sound limb

Residual limb

Averaged L and R

Sound limb

Residual limb

Ankle Angle Moments Powers

0.84 0.96 0.78

0.72 0.80 0.72

0.54 0.92 0.69

0.88 0.87 0.86

0.81 0.88 0.76

0.51 0.82 0.52

Knee Angle Moments Powers

0.96 0.87 0.96

0.87 0.58 0.87

0.86 0.22 0.85

0.97 0.84 0.97

0.89 0.72 0.89

0.87 0.57 0.86

Hip Angle Moments Powers

0.96 0.87 0.86

0.81 0.58 0.71

0.89 0.22 0.84

0.95 0.84 0.83

0.75 0.72 0.71

0.77 0.30 0.56

Descending the incline, the sound limb had reduced vertical GRF at V1, and increased vertical GRF at V2, compared to the controls. This was due to the sound limb’s constant vertical GRF, throughout stance phase. The sound limb had reduced anterior force, at A/P2, compared to the controls, at terminal stance. The residual limb had decreased vertical GRF at V1 and V3, compared to the controls, also due to the reduced variation through stance phase. The residual limb had decreased anterior force at A/P1, and decreased posterior force at A/P2, compared to the controls (Fig. 4).

3.4. Joint kinetics Joint kinetics are presented in Figs. 5 and 6, and summarised for selected gait events in Table 7. Compared to the controls, the sound limbs’ moments and powers were reduced and the residual limbs’ moments and powers further reduced (Figs. 5 and 6). These differences may be due to differences in walking speeds. As discussed in the methods, the moments and powers of the prosthetic/residual limb should be interpreted with caution.

Fig. 4. Vertical and anterior/posterior ground reaction forces: time-normalised and averaged for all participants. (a) GRF vertical, ascending; (b) GRF horizontal, ascending; (c) GRF vertical, descending; (d) GRF horizontal, descending.

524

D.R. Vickers et al. / Gait & Posture 27 (2008) 518–529

Table 5 Summary of ground reaction force data Mean (S.D.)

Vertical V1 V2 V3

Ascending

Descending

Controls

Amputees

Averaged L and R

Sound limb

10.9 (1.6) 7.3 (1.6) 11.2 (1.6)

Anterior (+)/posterior ( ) A/P1 1.8 (0.7) A/P2 + 1.7 (0.7)

Residual limb

9.6 (1.6) 9.1* (1.6) 9.8 (1.6)

8.2* (1.6) 8.8 (1.6) 9.3 (1.6)

1.1 (1.0) + 1.1 (1.0)

0.7* (0.7) + 0.5 (0.7)

Controls

Amputees

Averaged L and R

Sound limb

13.2 (1.6) 7.3 (1.6) 9.1 (1.6) 1.9 (0.7) + 1.9 (0.7)

Residual limb

9.3* (1.6) 9.2* (1.6) 9.0 (1.6)

9.6* (1.6) 9.3* (1.6) 9.0 (1.6)

1.1 (1.0) + 0.7* (1.0)

0.4 * + 0.7*

Units are N/kg. V, vertical GRF; V1, first stationary point (max); V2, second stationary point (min); V3, third stationary point (max); A/P, anterior/posterior GRF; A/P1, first stationary point; A/P2, second stationary point. * Significant difference to the control subjects.

Table 6 Peak required coefficient of friction, during loading response and terminal stance Peak RCOFs during

Loading response Terminal stance

Ascending

Descending

Controls

Amputees

Controls

Amputees

Averaged L and R

Sound limb

Residual limb

Averaged L and R

Sound limb

Residual limb

0.19 0.24

0.14 0.14

0.10 0.08

0.15 0.30

0.12 0.10

0.04 0.11

Fig. 5. Hip flexion moment, knee flexion moment and ankle dorsiflexion moment: time and mass-normalised and averaged for all participants (sagittal plane). (a) Hip flexion moment, ascending; (b) knee flexion moment, ascending; (c) ankle DF moment, ascending; (d) hip flexion moment, descending; (e) knee flexion moment, descending; (f) ankle DF moment, descending.

D.R. Vickers et al. / Gait & Posture 27 (2008) 518–529

525

Fig. 6. Total hip power, total knee power and total ankle power: time and mass-normalised, and averaged for all participants. (a) Total hip power, ascending; (b) total knee power, ascending; (c) total ankle power, ascending; (d) total hip power, descending; (e) total knee power, descending; (f) total ankle power, descending.

The sound limb had a significantly reduced plantarflexion moment, at AM2, compared to the controls. The sound and residual limbs had reduced knee moment range, and reduced hip moments, at HM1 and HM2, compared to the controls, in both directions (Fig. 5). The sound limb’s ankle had less power generation, at AP2, than the controls, coinciding with toe-off. The sound and residual limbs had reduced knee power range compared to the controls. The sound and residual limbs had reduced hip power generation, at HP1 and HP3, and reduced hip power absorption at HP2 compared to the controls (Fig. 6). 3.5. Muscle activation Muscle activations for gluteus maximus, lateral hamstrings and vastus lateralis are shown in Fig. 7, and their periods of activation listed in Table 8. There was good repeatability within subjects (CMC = 0.73–0.9) (Table 9). Between subjects, there was better repeatability for gluteus maximus (mean CMC = 0.68), than for vastus lateralis (mean CMC = 0.53), and for lateral hamstrings (mean CMC = 0.48). There was no significant difference in the muscle activation periods between ascending and descending, except for the controls’ gluteus maximus. For all three muscles, the amputees had longer periods of activation than

the controls. For the lateral hamstrings; the activation period was longer for the sound than for the residual limb (Table 8).

4. Discussion The main difference for amputees walking on an incline, compared to controls, is instability in stance. This is demonstrated by reduced single support time, GRF and speed in the residual limb. The source of instability is the reduced ankle motion range of the prosthetic limb, and the lack of ankle power generation. The residual limb attempts to compensate through shortened step length and increased muscle activity, while the sound limb stabilises the body position. The loss of proprioception in the residual limb can be problematic. 4.1. Ankle range of motion and power generation Limitations of conventional prostheses include reduced range of motion and lack of power generation. For incline walking, the prosthetic foot/ankle has a limited dynamic range of 108, but maintains the same pattern of dorsi- and plantar-flexing as the controls (Fig. 3(c and f)). According to Lehmann’s [25,26] analysis of the SACH foot, 1500 N should be applied to the forefoot to achieve 2 cm deflection, and 800 N should be applied for 2 cm deformation at the

526

D.R. Vickers et al. / Gait & Posture 27 (2008) 518–529

Table 7 Summary of joint kinetic data Mean (S.D.)

Ascending

Descending

Controls

Amputees

Averaged L and R

Sound limb

Controls

Amputees

Residual limb

Averaged L and R

Sound limb

Residual limb

Ankle moments and powers—at events in gait cycle AM1 / / AM2 1.4 PF (0.15) 1.05* PF (0.15) AP1 0.53 A (0.4) 0.33* A (0.5) AP2 1.87 G (0.4) 1.0* G (0.5)

/ 0.89* PF (0.09) 0.13* A (0.1) 0.15* G (0.09)/

1.05 PF (0.20) 1.15 PF (0.20) 1.43 A (0.4) 1.2 G (0.4)

0.8* PF (0.20) 0.95* PF (0.20) 0.67* A (0.3) 0.33* G (0.3)

/ 0.889* PF (0.14) 0.19* A (0.29) 0.11* G (0.15)

Knee moments and powers—at events in gait cycle KM range 0.85 (0.25) 0.60* (0.25) KP range 2.15 (0.4) 0.75* (0.3)

0.20* (0.30) 0.20* (0.2)

1.1 (0.20) 3.0 (0.4)

0.60* (0.25) 0.55* (0.2)

0.35* (0.30) 0.9* (0.25)

Hip moments and powers—at events in gait cycle HM1 1.30 E (0.3) 0.60* E (0.3) HM2 1.10 F (0.3) 0.60* F (0.3) HP1 1.74 G (0.5) 0.67* G (0.5) HP2 0.90 A (0.5) 0.47 G (0.5) HP3 1.74 G (0.5) 0.60* G (0.5)

0.73* 0.27* 1.27* 0.20* 0.47*

1.20 1.00 0.80 0.90 1.54

0.45* 0.45* 0.07* 0.37* 0.54*

0.40* 0.40* 0.40* 0.34* 0.34*

E (0.3) F (0.3) G (0.4) A (0.4) G (0.4)

E (0.3) F (0.3) G (0.5) A (0.5) G (0.5)

E (0.3) F (0.3) G (0.2) A (0.2) G (0.2)

E (0.3) F (0.3) G (0.3) A (0.3) G (0.3)

Units are Nm/kg and Watts/kg. Ankle moments and powers on the prosthetic limb are only an indication of the mechanical resistance in the prosthetic foot and ankle, and should be interpreted with care. A, power absorption, AM, ankle moment; AM1, first stationary point (max); AM2, second stationary point (max); AP, ankle power; AP1 = min. stationary point; AP2, max. stationary point; DF, dorsiflexion; E, extension; F, flexion; G, power generation, HM, hip moment; HM1, max. stationary point; HM2, min. stationary point; HP, hip power; HP1, first stationary point (max.); HP2, second stationary point (min.); HP3, third stationary point (max.); KM, knee moment; KP, knee power; PF, plantar-flexion. * Significant difference to the control subjects.

Fig. 7. Amplitude and threshold of EMG of gluteus maximus, lateral hamstrings and vastus lateralis, for each group and condition. (a) Gluteus maximus, ascending; (b) lateral hamstrings, ascending; (c) vastus lateralis, ascending; (d) gluteus maximus, descending; (e) lateral hamstrings, descending; (f) vastus lateralis, descending.

D.R. Vickers et al. / Gait & Posture 27 (2008) 518–529

527

Table 8 Percentage of the gait cycle for which the muscle is activated, % gait cycle (S.D.)*,# Muscle

Gluteus maximus Lateral hamstrings Vastus lateralis

Ascending

Descending

Controls

Amputees

Averaged L and R

Sound limb

#

38 (13) 57 (21)# 43 (16)

#

62 (25) 90 (11)# 75 (29)*

Residual limb #

69 (19) 71 (21)# 74 (23)*

Controls

Amputees

Averaged L and R

Sound limb

Residual limb

21 (11) 48 (28)# 50 (24)#

49 (39) 85 (21)# 95 (10)#

50 (33) 70 (24)# 83 (16)#

Level of activation set at 18% MVC [20]. * Significant difference between this limb, and the control only, p < 0.05. # Significant difference between all three limb groups. Table 9 Within-subject and between-subject coefficient of multiple correlation (CMC) for EMG Muscle

Ascending

Descending

Controls

Amputees

Controls

Amputees

Averaged L and R

Sound limb

Residual limb

Averaged L and R

Sound limb

Residual limb

Within-subject CMCs mean (S.D.) Gluteus maximus 0.85 (0.07) Lateral hamstrings 0.87 (0.05) Vastus lateralis 0.82 (0.10)

0.87 (0.07) 0.88 (0.04) 0.79 (0.09)

0.89 (0.08) 0.88 (0.04) 0.90 (0.07)

0.81 (0.09) 0.86 (0.07) 0.73 (0.15)

0.83 (0.05) 0.86 (0.06) 0.74 (0.15)

0.86 (0.05) 0.88 (0.05) 0.80 (0.14)

Between-subject CMCs Gluteus maximus Lateral hamstrings Vastus lateralis

0.78 0.48 0.47

0.75 0.56 0.55

0.67 0.48 0.44

0.50 0.50 0.44

0.66 0.58 0.36

0.73 0.58 0.63

heel. These deflections are the SACH foot’s alternatives to dorsiflexion and plantar-flexion, respectively. The amputees do not achieve these forces (Fig. 4), as they would need stability on the prosthetic foot and their entire body weight loaded through the toe or heel. Klute et al. [26] agrees that the stiff heel region of the SACH foot may prolong the time between heel contact and foot flat, hence requiring additional stability of the residual limb. The amputee’s ankle moments are delayed by 10%, due to gradual application of GRF, and magnitude is slightly reduced (Fig. 5(c and f)). The prosthetic limb has almost no ankle power with only slight power generation at push off (Fig. 6(c and f)). Because of the limited ankle dorsiflexion, the knee on the residual limb side compensates with reduced mid-stance flexion, to maintain the body centre of mass position for forward momentum (Fig. 3(b and e)). This results in only a small knee extension moment and almost no knee power (Figs. 5(b and e) and 6(b and e)). The combination of reduced ankle range of motion and power, and its effects at the hip at terminal stance, contributes to inadequate step length and difficulty raising the centre of mass up the incline. 4.2. Stability The residual limb lacks stability in stance. This is demonstrated by its decreased single support time, speed, cadence (Table 2), and GRF compared to the controls (Fig. 4). This may be caused by the low force in the residual

limb and/or the moments tolerance or limited proprioception. The knee of the residual limb has become the controlling mechanism of the calf, and the prosthetic foot. From studies of able-bodied inclined walking, we know the hamstrings decelerate the leg prior to heel contact, then work with quadriceps to stabilise the knee [27]. These muscles also elevate or lower the body weight. On an incline, the muscles work harder to stabilise the knee and ankle joint, and to increase speed, the amplitude of EMG increases (Fig. 7). The compensating muscle activity in the amputee shows the increased effort to stabilise the body on the incline. Particularly on the downhill, all three muscles measured in the residual limb, had increased amplitude compared to the controls (Fig. 7(d–f)). The amputees’ slower speed, may predict decreased EMG amplitude, but the opposite is true. For the amputees to achieve the controls’ walking speed, a further increase in EMG amplitude would be required. With peak amplitudes of 65–138% MVC, for the residual limb, it is difficult to comprehend how this reduced muscle could work any harder (Fig. 7). For the sound limb, the lateral hamstrings are activated throughout the gait cycle, and vastus lateralis only rests briefly (12–14% of gait cycle) during swing phase (Fig. 7(b, c, e, and f)). They also have increased amplitude compared to the controls. This compensation requires increased muscular energy, to support a stable body position, through the knee of the sound limb. Gluteus maximus of the sound limb has reduced amplitude compared to the controls, over an increased activation period (Fig. 7(a and d)). Therefore, it may be affected by the reduced speed and increased stance

528

D.R. Vickers et al. / Gait & Posture 27 (2008) 518–529

phase, rather than changes in other gait parameters (Table 2). The sound limb supports the body while the residual limb undergoes gradual weight acceptance. The prosthetic limb has heel-to-toe gait, for both uphill and downhill walking. At the end of stance phase, the ankle of the sound limb does not plantar-flex (Fig. 3(c and f)), as this would mean balancing the entire body weight through the toes of the sound limb, while transferring weight onto the prosthetic limb. This is difficult due to the lack of proprioception of the prosthetic foot and would require sufficient hip abduction and extension strength. 4.3. Forces and friction Hubbard and McElroy [24] found a significant relationship between vertical GRF pattern and speed. Walking slowly was associated with the flattened vertical GRF pattern seen with the amputees (Fig. 4). Jones et al. [28,29] showed a relationship between static weight bearing (SWB) and speed in elderly unilateral transtibial amputees. A higher tolerance of forces in SWB allowed greater speed during gait. The reverse is also true: by reducing the speed (which happens in the elderly) the vertical GRF at weight acceptance is reduced. If load intolerance at the stump/ socket interface is problematic, then it must be improved before speed can be increased. Studies [13,30] on healthy populations walking and running down an incline observed an increased braking force through an increase in both the first perpendicular and first shear GRF peaks. On the uphill the V1 GRF peak decreased, while the AP2 GRF peak decreased, with an overall increase in the push-off force (Fig. 4). Our controls demonstrated this pattern, however, the amputees did not. The sound limb AP forces were consistent with an increase in braking force on the downhill and increase in propulsive force on the uphill (Fig. 4(b and d)), but the vertical GRF did not change for uphill or downhill nor did it have distinct regions of braking and propulsion (Fig. 4(a and c)). This may be explained by the reduced speed [24] but may also relate to their tolerance of vertical GRF. The sound limb had a smaller AP force in both directions than the controls, and the prosthetic limb values were smaller still (Fig. 4(b and d)). On a 48 incline, McVay and Redfern [21] calculated RCOF to be 0.3 and 0.4 for downhill and uphill walking, respectively. He also showed that RCOF increased as angle of incline increased. Peak RCOFs for our controls were 0.30 and 0.24 for downhill and uphill, respectively (Table 6). Peak values for the prosthetic limb were calculated at 0.11 and 0.10 for downhill and uphill, respectively (Table 6). The prosthetic limb’s lower RCOF is due to a lower shear force. The RCOF increases during weight acceptance, to its maximum. This measure of RCOF, indicates the direction of forces applied by the prosthetic limb. The lack of shear force is part of what hinders the forward propulsion of the prosthetic limb. By reducing the RCOF, amputees have decreased the chances of a slip occurring. However, the RCOFs of the

controls, show that the floor surface and incline angle are not limiting factors for safe walking. Rather the amputees are lacking confidence, proprioception and/or ability to apply the forces in a helpful direction. Slips occur if shear forces overcome frictional forces at the shoe–floor interface. But to increase frictional force amputees must increase perpendicular GRF. Amputees’ instability in stance hinders their efforts to apply this GRF. McVay and Redfern [21] demonstrated that RCOF increases as slope increases; therefore, the risk of slipping will increase on steeper slopes. The limited range of prosthetic ankle motion and inadequate proprioception increases this risk. In able-bodied walking the RCOF can be increased, as the area of contact is increased, through plantar-flexion. The prosthetic ankle cannot do this without sufficient force; therefore, in order to achieve foot-flat, they must roll over the heel, relying on the correct relationship between shear and friction forces. If the RCOF, can be increased by other means, e.g. foot to floor contact, then the point at which frictional force is greater than shear force can be reached with less perpendicular GRF and the amputee can increase GRF more confidently. By increasing GRF they will increase ankle moments and deflection, which will provide better foot–floor contact. 5. Conclusions This study identified the different gait characteristics between transtibial amputees and able-bodied participants, as well as the compensations required by the sound and residual limbs, when walking on an incline. The main issue for unilateral transtibial amputees is instability in stance, caused by limited ankle range of motion and lack of ankle power generation of the prosthetic limb and inadequate proprioception and force tolerance of the residual limb. The main mechanical problem in conventional foot prostheses is the lack of active plantar and dorsiflexion. This was true for both walking conditions. Without ankle motion, the body has to compensate for reduced push-off power and inadequate heel-contact to foot-flat movement. These compensations may be difficult for the elderly, particularly the increased muscle activation. Prosthetic foot designs must incorporate a full range of active ankle motion. Conventional prosthetic feet do not generate the power necessary to provide push-off. This requires the sound and residual limbs to lift using their own strength. The normal pattern of power absorption and generation should be used in designing or programming prosthetic feet, to suit the sound limb gait cycle. Until an amputee can balance comfortably on their residual limb, they are not able to attain a normal gait pattern. To improve stability, areas of investigation should include foot/ ground contact regarding RCOF, tolerance of weight force and moments through the residual limb, as well as rehabilitation training to improve balance on the prosthesis.

D.R. Vickers et al. / Gait & Posture 27 (2008) 518–529

Dealing with the issues identified with conventional prosthetic feet, and the limitations of the residual limb should allow the joints of the residual and sound limbs to return to their normal range of motion and thereby achieve a more normal gait pattern on inclines.

Conflicts of interest There are no conflicts of interest to report in regards to this article.

Acknowledgement The work presented in this paper was undertaken for C. Palk’s MBiomedE project.

References [1] Bowker JH, Michael JW. Atlas of limb prosthetics: surgical, prosthetic, and rehabilitation principles. In: Mosby year book2nd ed., St. Louis: American academy of orthopaedic surgeons; 1992. [2] Marks LJ, Michael JW. Science, medicine, and the future: artificial limbs. BMJ 2001;323:732–5. [3] Meier 3rd RH. Disability evaluation and impairment rating for the person with an amputation. Phys Med Rehabil Clin N Am 2001;12:621–36. [4] Marinakis GN. Interlimb symmetry of traumatic unilateral transtibial amputees wearing two different prosthetic feet in the early rehabilitation stage. J Rehabil Res Dev 2004;41:581–90. [5] Bagley AM, Skinner HB. Progress in gait analysis in amputees: a special review. Crit Rev Phys Rehabil Med 1991;3:101–20. [6] Hsu MJ, Nielsen DH, Yack HJ, Shurr DG. Physiological measurements of walking and running in people with transtibial amputations with 3 different prostheses. J Orthop Sports Phys Ther 1999;29:526–33. [7] Nyska M, Shabat S, Arya A, McCabe C, Linge K, Klenerman L. A comparative study of different below-knee prostheses by dynamic foot pressure analysis. Int J Rehabil Res 2002;25:341–4. [8] Bateni H, Olney SJ. Kinematic and kinetic variations of below-knee amputee gait. J Prosthet Orthot 2002;14:2–11. [9] Isakov E, Burger H, Gregoric M, Marincek C. Isokinetic and isometric strength of the thigh muscles in below-knee amputees. Clin Biomech 1996;11:233–5. [10] Barth DG, Schumacher L, Sienko Thomas S. Gait analysis and energy cost of below-knee amputees wearing six different prosthetic feet. J Prosthet Orthot 1992;4:63–75. [11] Isakov E, Keren O, Benjuya N. Trans-tibial amputee gait: time– distance parameters and EMG activity. Prosthet Orthot Int 2000; 24:216–20.

529

[12] Culham EG, Peat M, Newell E. Below-knee amputation: a comparison of the effect of the SACH foot and single axis foot on electromyographic patterns during locomotion. Prosthet Orthot Int 1986;10:15– 22. [13] McIntosh AS, Beatty KT, Dwan LN, Vickers DR. Gait dynamics on an inclined walkway. J Biomech 2006;39:2491–502. [14] Macfarlane PA, Nielsen DH, Shurr DG, Meier K. Gait comparisons for below-knee amputees using a flex-foot (TM) versus a conventional prosthetic foot. J Prosthet Orthot 1991;3:150–61. [15] Sin SW, Chow DHK, Cheng JCY. Significance of non-level walking on transtibial prosthesis fitting with particular reference to the effects of anterior–posterior alignment. J Rehabil Res Dev 2001;38:1–6. [16] Pandian G, Kowalske K. Daily functioning of patients with an amputated lower extremity. Clin Orthop 1999;1:91–7. [17] Buczek FL, Cavanagh PR, Kulakowski BT, Pradhan P. Slip resistance needs of the mobility disabled during level and grade walking. In: Gray BE, editor. Slips, stumbles & falls: pedestrian footwear & surfaces, ASTM STP 1103. Philadelphia: American Society for Testing and Materials; 1990. p. 39–54. [18] Davis R, Ounpuu S, Tyburski D, Gage J. A gait analysis data collection and reduction technique. Hum Mov Sci 1991;10:575–87. [19] Kadaba MP, Ramakrishnan HK, Wooten ME. Measurement of lower extremity kinematics during level walking. J Orthop Res 1990;8:383– 92. [20] Johanson ME, Radtka SA. Amplitude threshold criteria improve surface electrode specificity during walking and functional movements. Gait Posture 2006;4:429–34. [21] McVay EJ, Redfern MS. Rampway safety: foot force as a function of rampway angle. Am Ind Hyg Assoc J 1994;55:626–34. [22] Kadaba MP, Ramakrishnan HK, Wootten ME, Gainey J, Gorton G, Cochran GV. Repeatability of kinematic, kinetic, and electromyographic data in normal adult gait. J Orthop Res 1989;7:849–60. [23] Armitage P, Berry G. Statistical methods in medical research, 2nd ed., Oxford: Blackwell Scientific; 1987. [24] Hubbard WA, McElroy GK. Benchmark data for elderly, vascular transtibial amputees after rehabilitation. Prosthet Orthot Int 1994;18:142–9. [25] Lehmann JF, Price R, Boswell-Bessette S, Dralle A, Questad K, deLateur BJ. Comprehensive analysis of energy storing prosthetic feet: flex foot and Seattle foot versus standard SACH foot. Arch Phys Med Rehabil 1993;74:1225–31. [26] Klute GK, Berge JS, Segal AD. Heel-region properties of prosthetic feet and shoes. J Rehabil Res Dev 2004;41:535–46. [27] Lange GW, Hintermeister RA, Schlegel T, Dillman CJ, Steadman JR. Electromyographic and kinematic analysis of graded treadmill walking and the implications for knee rehabilitation. JOSPT 1996;23:294– 301. [28] Jones ME, Bashford GM, Mann JM. Weight bearing and velocity in trans-tibial and transfemoral amputees. Prosthet Orthot Int 1997;21:183–6. [29] Jones ME, Steel JR, Bashford GM, Davidson IR. Static versus dynamic prosthetic weight bearing in elderly transtibial amputees. Prosthet Orthot Int 1997;21:100–6. [30] Gottschall JS, Kram R. Ground reaction forces during downhill and uphill running. J Biomech 2005;38:445–52.