Abstract
The
aim of this study was to develop quantitative measures of postural stability
and apply these to the investigation of postural control requirements during
crutch-assisted ambulation. Subjects with paraplegia (PARA) and able-bodied
(AB) subjects walked in two different designs of orthosis. Outcome measures
were: duration in various support phases, duration inside/outside base of
support (BoS), distances of the body’s projected
centre of gravity (PCoG) to the BoS edges,
and normalised sway distances. PARA spent significantly more time in
statically stable postures (p < .001), with greater time in a quadripod
phase (p = .001) and less time in a bipod phase (p = .021), compared to AB. A
significantly larger ‘stability margin’ was seen in PARA during both the
quadripod (p = .004) and tripod (p = .013) phases than in AB. At higher gait
speeds in AB increasing periods were spent in statically unstable posture.
Future attempts to improve gait speed and efficiency for individuals with
paraplegia will need to address the challenge to dynamic postural stability of
increasing periods in statically unstable postures.
1.
Introduction
Health professionals regard paraplegic ambulation in bilateral knee-ankle-foot orthoses (KAFOs) with walking aids using either a swing-through or reciprocal gait pattern as impractical. Several different types of hip-knee-ankle-foot orthoses are in clinical use, including the Isocentric Reciprocating Gait Orthosis (IRGO) and the Walkabout Orthosis (WO). The WO employs an entirely different design concept to the HGO or RGO with a single-axis joint medially linking two KAFOs. A prototype device linking bilateral KAFOs, called the Moorong Medial Linkage Orthosis (MMLO), was subsequently developed to minimise the distance between the hip joint centres and the hinge axis of the MMLO. The purpose of the present study was to examine postural stability, balance and support requirements in subjects with paraplegia (PARA) compared to able-bodied subjects (AB) ambulating in the IRGO and MMLO orthoses by analysing motion of the body’s PCoG in relation to the BoS during gait.
Quadrupedal locomotion is statically stable when the projected centre of gravity of the body (PCoG) remains inside the base of support (BoS). Static stability throughout the entire gait cycle is only possible at very slow speeds of locomotion [1]. Normal human bipedal gait involves periods in both statically stable and unstable postures, with stability of the body maintained dynamically by the rate of movement and inertia [2].
Specific hypotheses that were tested included: (1) that there would be significant differences in stability between PARA and AB during orthosis-assisted gait using crutches; and (2) that periods of dynamic instability would increase with increasing gait speed in AB.
2.
Methods
Three male individuals (height =
1.73 ±0.04 m, weight = 74.0 ±11.0, age = 30.7 ±6.1) with
paraplegia (PARA) and three able-bodied individuals (AB) (height = 1.76 ±0.05 m, weight = 81.0 ±7.8, age
= 32.7 ±5.9) matched for stature, age and
gender gave their informed consent to participate in the study. Approval for
the project was gained from the Human Ethics Committee, University of Sydney.
PARA underwent biomechanical assessment after twelve weeks of supervised gait training by an experienced physiotherapist. Three-dimensional video, crutch and ground reaction force data were recorded at 60 Hz as subjects ambulated in a gait laboratory using a motion analysis system (Expert Vision HiRes [EVa], Motion Analysis Corp., Santa Rosa, U.S.A.). Video data were recorded from a neutral trial for all subjects. PARA performed five trials at a self-selected pace, whilst AB performed three trials walking at three speeds (preferred pace, slow and fast pace) in the orthoses. A 7-segment model (left and right lower limb, pelvis/trunk/head, left and right upper arm, left and right forearm/hand/crutch segments) with six degrees of freedom for each segment was employed. Co-ordinates for the centre of gravity (CoG) of the body were determined.
The edges of the base of support (BoS) were defined by lines joining the centroids of markers over the lateral malleoli and lower crutches. Periods of limb and crutch contact with the ground were identified graphically using a cursor in plots of the ground reaction forces superimposed with velocities of the relevant markers. The timing data thus obtained were used to define the support phases: quadripod, tripod, bipod and unipod. Within each defined gait phase, distances of the projected centre of gravity (PCoG) to all active edges of the BoS were calculated using in-house software (Matlab, Mathworks, USA) and the minimum detected. A statically stable posture was defined by a positive value for the minimum distance. In addition, the PCoG path relative to active BoS edges were plotted and used to define whether PCoG was located inside or outside the BoS. The duration of time spent in each of the four gait phases, namely quadripod, tripod, bipod and unipod, were also calculated.
Normalised sway distance (m) for CoG, representing the distance traveled minus the distance progressed in one minute, was calculated as the sum of displacements between samples of CoG minus its net displacements in anterior-posterior (AP) and medio-lateral (ML) directions respectively over gait period and subsequently time normalised.
A mixed analysis of variance design of the factors spinal cord-injured status and orthosis and trial number was employed (SPSS for Windows 9.0) to investigate differences within and between paraplegic and able-bodied groups. An alpha level of p < .05 was set a priori to determine criterion statistical significance, but in the text exact probabilities have been displayed.
3.
Results
For the relative proportions of gait time spent in the different support phases and minimum distances of PCoG to BoS edges in each phase for PARA compared to AB walking at various speeds no significant within-group differences were found for orthosis (Figures 1 and 2).
(b) (a)
Figure
1 Relative gait time (a) and
minimum distances of PcoG to BoS (b) in different support phases for PARA and
AB at various speeds with all orthoses combined.
PARA spent a significantly greater proportion of gait time in a quadripod stance in comparison to AB walking at an equivalent gait speed (slow pace) (p = .001), with the PCoG significantly further from the BoS edge in PARA (p = .004). The proportion of gait time spent in the tripod phase was not significantly different between PARA and AB at any pace (p = .204). However, while the PCoG path for PARA in tripod stance remained well within the BoS, it extended beyond the edge of the BoS in AB at all paces, with the difference in CoG location between groups ranging between 5.5-7cm (p = .013). PARA spent on average less than 1.5% of gait time in a dynamically unstable bipod posture, unlike AB (p = .021). Similarly, PARA only spent a very small proportion of gait time with the body’s CoG located outside the BoS in comparison to AB (p = .002).
In AB, with increasing speed of ambulation in orthoses using crutches, there was a significant shift towards greater periods of time spent in more unstable postures (p = .010), such that approximately 60% of gait time at the faster pace where the CoG was lying outside BoS. Dynamic instability arose not only from increased periods of time in the less stable bipod and unipod postures, but from the larger minimum distances of CoG outside BoS which were associated with these phases. The least stable postures, where the CoG was lying furthest outside the BoS, occurred during a bipod phase when supports were one-foot and the unilateral crutch (mean minimum distance -0.184±0.045m, absolute minimum distance -0.300m) or unipedal phase (mean minimum distance -0.209±0.021m, absolute minimum distance -0.280m).
Overall, sway distances for CoG were significantly greater in PARA than AB, both in the AP (p = .004) and ML (p = .025) direction. In PARA, normalised sway distance in the AP direction for CoG (NSDAPCoG) was significantly greater in the IRGO than in the MMLO (p = .025). Interestingly, as well as an orthosis effect for NSDAPCoG, a significant interaction between group and orthosis was seen, whereby in AB the NSDAPCoG was less in the IRGO than in the MMLO (p = .017), possibly indicating greater trunk restriction in the IRGO as well as different mechanisms between the groups. Also, whilst no significant differences were found between orthoses, NSDMLCoG tended to be greater in the medial-linkage devices than IRGO in both groups.
4.
DISCUSSION
The subjects
with paraplegia in this study ambulated very slowly with mean velocities
between 0.08-0.10m∙s-1, consistent with other studies. This
may reflect the study design in which novice clients were recruited and
provided with equivalent amounts of training by an expert physiotherapist to
ensure satisfactory technique and similar proficiency was achieved in each
device. Whilst the period of training was sufficient for the purposes of this
pilot study, the amount of training in each orthosis (9 sessions) was only
approximately half the dosage described in other studies.
Results of the current study
support theoretical models of stability described by McGhee and Frank [1] and
findings by Karcnik and Kralj [2] for forced quadripod gait in an able-bodied subject
and crutch-supported below-knee amputee gait. As speed of ambulation increased in AB, a progression towards greater
amounts of time spent in statically unstable postures was seen with more
simultaneous and overlapping periods of foot/crutch and/or crutch/crutch
lifting and advancement. These results highlight, in particular, the increasing
control challenge that individuals with paraplegia would face at higher speeds
of ambulation. Results from this present pilot study suggest that a person with
paraplegia may adopt a gait with movement patterns and support duty cycles that
optimise static stability, avoiding postures with large negative stability
margins, such as demonstrated during unilateral crutch and foot swing in AB.
5.
Summary and
Conclusions
For individuals with paraplegia ambulating using an orthosis and
crutches, static stability is maintained throughout the gait cycle by using a
crawling quadripod pattern at very slow gait speeds, although this is quite
inefficient. Ambulating at faster speeds whilst potentially leading to improved
gait efficiency, also involves far greater control challenge, with increasing
periods spent in statically unstable postures. It may be possible in the future
with a hybrid FES-orthosis to achieve higher speeds of ambulation with improved
gait efficiency, better utilising the effects of gravity and inertia to
maintain dynamic postural stability.
Acknowledgment
This study was financially supported by grants from the Northern Sydney Area Health Service (No. 94/21) and the Motor Accidents Authority of N.S.W. (“WalkBack to the Future”). The authors thank Ms. Uangthip Rattanaprasert for her assistance with laboratory setup, data tracking and initial processing.
References
[1] McGhee RB, Frank AA. On the stability properties of quadruped creeping gaits. Math Biosci 1968; 3: 331-351.
[2] Karcnik T, Kralj A. Stability and energy criteria in healthy and paraplegic subject gait. Artif Organs 1997; 21: 191-194.