Journal of Biomechanics 47 (2014) 1816–1821
Contents lists available at ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Patterns in the knee flexion-extension moment profile during stair ascent and descent in patients with total knee arthroplasty Jodie A. McClelland a,n, Julian A. Feller b, Hylton B. Menz a, Kate E. Webster a a b
School of Allied Health, Faculty of Health Sciences, La Trobe University, Melbourne, Australia OrthoSport Victoria, Epworth HealthCare, Melbourne, Australia
art ic l e i nf o
a b s t r a c t
Article history: Accepted 19 March 2014
The aim of this study was to investigate the prevalence of abnormal knee biomechanical patterns in 40 patients with a modern TKA prosthesis, compared to 40 matched control participants when ascending and descending stairs. Fewer patients were able to ascend (65%) or descend stairs (53%) unassisted than controls (83%). Of the participants who could ascend and descend, cluster analysis classified most patients (up to 77%) as demonstrating a similar knee moment pattern as all controls. A small subgroup of patients who completed the tasks did so with distinctly abnormal biomechanics compared to other patients and controls. These findings suggest that recovery of normal stair climbing is possible. However, rehabilitation might be more effective if it were tailored to account for these differences between patients. & 2014 Elsevier Ltd. All rights reserved.
Keywords: Knee arthroplasty Stairs Motion analysis Gait analysis Flexion
1. Introduction The ability to ascend and descend stairs is a task that is encountered frequently during daily living, particularly as stairs often provide the most convenient access point to buildings and services in the community. Unfortunately, it has been reported that stair climbing is the daily functional activity for which patients with total knee arthroplasty (TKA) are most dissatisfied with their outcome from surgery and there is evidence that stair climbing ability is not improved following TKA, and remains significantly impaired for up to 2 years following surgery (Bourne et al., 2010; Zeni and Snyder-Mackler, 2010b). Measuring knee biomechanics using motion analysis techniques has the benefit of providing information about the specific challenges faced by patients with TKA when negotiating stairs. Using this approach, the external knee joint flexion–extension moment pattern throughout stair ascent has been assessed in three previous studies (Andriacchi et al., 1982; Andriacchi et al., 1997; Catani et al., 2003). Each of these studies identified a subgroup of patients for whom the knee moment pattern was distinctly different from the other patients, and also from what is
n Correspondence to: Department of Physiotherapy, Faculty of Health Sciences, La Trobe University, Bundoora, Victoria 3086, Australia. Tel.: þ 61 3 9479 3254; fax: þ61 3 9479 5415. E-mail addresses: [email protected] (J.A. McClelland), [email protected] (J.A. Feller), [email protected] (H.B. Menz), [email protected] (K.E. Webster).
http://dx.doi.org/10.1016/j.jbiomech.2014.03.026 0021-9290/& 2014 Elsevier Ltd. All rights reserved.
typically expected of normal stair ascent (Andriacchi et al., 1982; Andriacchi et al., 1997; Catani et al., 2003). All authors of the previous studies agreed that design characteristics of some TKA prostheses may contribute to the development of an abnormal moment pattern during stair ascent, and particularly that the abnormal moment profile represented an attempt to avoid generating an external flexion moment (Andriacchi et al., 1982; Andriacchi et al., 1997; Catani et al., 2003). All three authors described a distinctly abnormal moment pattern in 40–60% of the patients. All authors agreed that characteristics of the sagittal moment in this proportion of patients suggested an attempt to avoid generating a knee flexion moment, which also avoided the need for quadriceps activity. Two of these studies were completed more than 20 years ago, and all of these studies assessed only patients with good or excellent outcome from TKA. Without recent studies investigating the profile of the sagittal plane moment it is not known what proportion of patients with modern prostheses may expect to demonstrate abnormal moment patterns, and how the biomechanical characteristics of these patients differ from normal. Therefore, an investigation of a larger cohort that is representative of typical contemporary patients would allow for more detailed description of the biomechanics of stair ascent and descent following TKA. The demands of descending stairs may be different from stair ascent, as the primary role of the quadriceps during descent is to absorb the force needed to lower bodyweight, rather than generate force for elevation, as in stair ascent. However, despite the potentially greater challenge that stair descent presents, there have been few studies that attempted to describe the biomechanics of patients during stair descent (Andriacchi et al., 1982; Bolanos et al., 1998;
J.A. McClelland et al. / Journal of Biomechanics 47 (2014) 1816–1821
Catani et al., 2003). None of these studies explored the prevalence of an abnormal pattern of knee moment during stair descent, therefore little is known about how the demand for quadriceps control changes throughout the activity. The aim of this study was therefore to investigate the prevalence of abnormal knee flexion–extension patterns during both stair ascent and descent in a representative group of patients following TKA with reference to a healthy control group. We hypothesised that most patients with TKA would ascend and descend stairs with knee moments that were similar to people without knee pathology. We also hypothesised that in a smaller proportion of patients the knee moment would be smaller in magnitude and delayed.
2. Materials and methods All patients who underwent TKA by a single experienced knee surgeon over 2 consecutive years were assessed for eligibility for study inclusion. Patients were invited to participate if they had undergone TKA for osteoarthritis at least twelve months and not more than 18 months earlier and were able to walk 10 metres without a gait aid. Patients with other documented orthopaedic, neurological or visual disturbances that may affect gait, including joint arthroplasties of the hips and ankles were excluded. Within 5 years, up to 30% of patients with primary TKA undergo TKA in the contralateral limb (Australian Orthopaedic Association, 2012; McMahon and Block, 2003; Shakoor et al., 2002). Therefore, to maximise the generalisability of the findings of this study, patients with bilateral knee arthroplasty were included, provided that the most recent procedure was undertaken at least 12 months prior to testing. From an initial list of 78 patients, 22 were excluded because of co-morbidities that may affect gait, nine declined and seven were unable to be contacted. The remaining 40 patients attended the gait laboratory for assessment. All participants had received a fully cemented Genesis-II posterior stabilised prosthesis (Smith and Nephew, Memphis, Tennessee, USA) and the patella was resurfaced in all knees. All participants received standard postoperative care, which emphasised early weightbearing mobility, quadriceps strengthening and maximisation of range of motion. In the 13 participants that had undergone bilateral knee arthroplasty surgery, only the most recent TKA was considered for analysis. The control group consisted of 40 participants who were assessed for eligibility to participate in the study using the same criteria as the TKA participants, except that control participants were excluded if they had undergone any joint arthroplasty surgery or reported any symptoms of lower limb osteoarthritis. All control participants were matched to the age (7 2 years) and gender of a member of the TKA group. A single limb of each control participant was used in the analysis and was determined by the operated limb of the matched TKA participant. Ethical approval was granted from the institution's Human Ethics Committee and written informed consent was obtained from all participants. An eight camera Vicon MX3 Motion Analysis System (Vicon, Oxford, UK) was used to collect kinematic data at a sampling rate of 100 Hz. The gait laboratory consisted of a 10 m walkway with two force plates (Kistler, Switzerland and AMTI, Watertwon, MA, USA) that were embedded into the floor in the middle of the walkway. A two-step staircase with dimensions based on the Building Code of Australia (step height of 180 mm and a step tread of 300 mm) was placed adjacent to these embedded forceplates (Office of the Australian Building and Construction Commissioner, 2006). A third force plate (AMTI, Watertown, MA, USA) was placed on the first step of the staircase so that the kinetic data of three consecutive steps could be collected (two level steps and one elevated step). Only data from this elevated force plate is included in this manuscript. All force plate data were collected at a sampling rate of 4000 Hz. Thirteen 14 mm diameter reflective markers were placed centrally over the sacrum, and bilaterally over the anterior superior iliac spines, lateral femoral epicondyles, lateral and medial malleoli, calcanei, and second metatarsals. Markers on 5 cm wands were placed on the lateral thighs and legs. A knee alignment device was used to determine the centre of the knee joint during a static data calibration. These marker placements were based on the modified Helen Hayes model (Davis et al., 1991; Kadaba et al., 1989). Additional markers were placed on the iliac crest to allow reconstruction of occluded ASIS markers (McClelland et al., 2010). During stair ascent, participants were instructed to take two steps on a level walkway before ascending the staircase in a reciprocal footfall pattern. During stair descent, the process was reversed so that participants commenced on the top step, descended onto the first elevated step which included a force plate and descended again to take 2 steps along the level walkway. There were no rails for support on the staircase. Participants were encouraged to practise ascending and descending the staircase as many times as necessary until comfortable with the process. Data collection did not proceed if patients were unable to ascend or descend without support, or were unable to ascend or descend with reciprocal footsteps.
1817
Participants were asked to ascend and descend the staircase until a minimum of three and a maximum of five trials of data were collected from each limb. Patients were assessed using the American Knee Society Knee Score (KSS) at a routine 12 month post-operative follow-up visit with the surgeon (Insall et al., 1989). At the time of motion analysis testing, patients with TKA also completed the Total Knee Function Questionnaire (TKFQ), which collects information about patients' perceived functional abilities and their importance to daily life (Weiss et al., 2002). Responses relevant to overall satisfaction and stair climbing were included in this study. Vicon PlugIn Gait (Vicon, Oxford Metrics, Oxford, UK) software was used to estimate lower limb joint positions, based on the model proposed by Davis et al. (1991) and Kadaba et al. (1989). Joint kinematics were calculated using Euler angles (flexion/extension, abduction/adduction, and internal/external rotation). Joint moments were calculated using standard inverse dynamics from force vectors that were defined in a global coordinate system and projected to a local coordinate system of the rigid body segment distal to each joint. The moments referred to throughout this manuscript are external to the knee and normalised to bodyweight and height. A single step cycle from each trial was included for analysis and was time normalised to 100% of the step cycle. During stair ascent, initial contact onto the first elevated step was considered as the beginning of the step cycle (0%) and toe-off of the same foot as the end of the step cycle (100%). During stair descent, initial contact onto the first elevated step was considered as the beginning of the step cycle (0%) and toe-off of the same foot was defined as the end of the step cycle (100%). Knee flexion angles and moments in the sagittal plane were calculated for each trial of each participant. Key biomechanical variables were identified and averaged across all trials of each individual (Fig. 1). For stair ascent, these were the maximum knee flexion angle (degrees), the maximum knee flexion moment (%Bw‐Ht) and the time when the sagittal moment changed direction from flexion to extension (% of step cycle) (Fig. 1). For stair descent, these were the maximum knee flexion angle during loading phase (0–50% of the step cycle as bodyweight is accepted onto the limb) (deg), the time when the sagittal moment changed direction from extension to flexion (% of step cycle), the maximum knee flexion moment during loading (% Bw‐Ht), the maximum knee flexion moment during late stance phase (50–100% of the step cycle, %Bw‐Ht) and the time taken to complete each task (s). Hierarchical cluster analysis was used to identify subgroups of patients using characteristics of the sagittal moment profile. Two separate cluster analyses were performed in this study: (i) a single analysis to classify moment profiles of all control and TKA participants who completed the stair ascent task (n¼59) and (ii) a single analysis to classify moment profiles of all control and TKA participants who completed stair descent (n¼ 54). The hierarchical cluster analysis procedure (based on the method described by Ward (1963)) begins by assigning each case to its own cluster. There is then a stepwise process whereby moment profiles that are most similar are joined to form clusters. This process continues until a single cluster is formed in a hierarchical tree (dendogram). In this study, the similarity between gait profiles was determined by the squared Euclidean distance between profiles and by using a between-groups linkage (the average similarity within a cluster is calculated and the case with the distance most similar to the average is the next to join the cluster). Two distinct waveforms of the sagittal moment profile have been described in patients with TKA during a stair ascent task (Andriacchi et al., 1982; Andriacchi et al., 1997; Catani et al., 2003), therefore it was determined a priori that there would be two clusters in each analysis based on key characteristics of this moment. In these studies, abnormal stair ascent has been described as an apparent avoidance of quadriceps use, which was characterised by a reduction in the maximum external flexion moment, and reduced time during stance that the quadriceps were required (i.e. reduced time that the external moment was in a flexion direction). Therefore, in the cluster analysis of stair ascent the independent variables were determined a priori as (i) the magnitude of the maximum flexion moment and (ii) the percentage of the gait cycle where the knee flexion moment changed direction to become a knee extensor moment (Fig. 1C). Patterns of the sagittal moment have not been described during stair descent where the primary role of the quadriceps is to control lowering of bodyweight onto the descended step. An avoidance of demand for quadriceps control during this moment is likely to be reflected in a reduction of the maximum flexion moment and a delay in the need for quadriceps activity (i.e. delay in generation of a flexion moment). Therefore in cluster analysis of stair decent, the independent variables were defined a priori as (i) the magnitude of the maximum knee flexion moment and (ii) the percentage of the gait cycle where the initial knee extensor moment changed direction to become a knee flexor moment (Fig. 1D). Step-wise discriminant function analysis was used to evaluate the robustness of group allocation, and to evaluate the relative importance of each variable in determining the clusters. In this procedure the discriminant function is recalculated for all except one case. This model is then applied to the withheld case to assign group membership, and the process repeated until each case has been left out. Group assignment is considered robust if there are a high percentage of participants classified to the correct group (i.e. low misclassification rate). This process is similar to that described by Toro et al. (2007) and Mulroy et al. (2003). Key biomechanical characteristics during both stair ascent and descent were compared between groups, and between clusters using independent samples t-tests with po 0.05. For all biomechanical parameters reported in this manuscript,
1818
J.A. McClelland et al. / Journal of Biomechanics 47 (2014) 1816–1821
A
Ascent
B
Descent
*
*
C
D ** **
***
***
Fig. 1. Typical waveforms for the knee flexion angle throughout the step cycle of stair ascent (A) and stair descent (B); the knee flexion moment through stair ascent (C) and stair descent (D). The key variables used in data analysis were the maximum flexion during loading (*), the maximum knee flexion moment (**), and the time in moment direction change (***). patients with unilateral TKA were compared to participants with bilateral TKA using an independent t-test. As there were no differences between groups, these groups were collapsed to form a single TKA group for the cluster analysis.
3. Results The demographic characteristics of all participants are shown in Table 1. The TKA group was heavier than the control group, which is similar to other studies that compare TKA patients to participants of similar demographic characteristics without TKA (Lingard et al., 2004; McClelland et al., 2007; Zeni and SnyderMackler, 2010a). 3.1. Stair climbing All control participants and all but one TKA participant were able to ascend and descend. However, after several attempts, only 33 control participants (83%) and 26 TKA participants (65% of all patients) could complete stair ascent independently. The same 33 control participants (83%) and 21 of these 26 TKA participants (53% of all patients) were able to complete the stair descent task. 3.2. Knee biomechanics of stair ascent All of the control participants and 20 TKA participants (77% of patients that completed the task or 50% of all patients) were classified in Cluster 1 (grey traces in Fig. 2A), which was declared a normal knee moment pattern. The remaining six TKA participants (23% of patients that completed the task or 15% of all patients) were classified in Cluster 2 (black traces). All participants generated a knee flexion moment that changed direction to an extension moment during stair ascent. However, for patients in
Table 1 Demographic characteristics of TKA and control participants.
Age (years) Height (cm) Weight (kg) F:M Right limb affected:left limb affected Knee Society Score (/100) Knee Society Function Score (/100) Total Knee Function Questionaire Responses Extremely satisfied with TKA Stair ascent ‘not limited at all’ Stair descent ‘not limited at all’ n
Control group
TKA group
69.6 (8.3) 165.0 (9.5) 76.2 (17.9) 22:18 N/A N/A N/A
69.1 (8.0) 165.5 (11.5) 86.4 (17.5)n 22:18 24:16 79.1 (11.2) 86.7 (20.8)
N/A N/A N/A
78% 65% 27%
Denotes significant difference in comparison to control group p r 0.05.
Cluster 2, the moment pattern was distinctly different from normal and was characterised by a reduction in the magnitude of the flexion moment and a change in moment direction from flexion to extension that occurred earlier in the step cycle. In confirmation of this cluster analysis, all participants in Cluster 1 (0% misclassification rate) and all but one case (17% misclassification rate) were correctly classified in the leave-one-out procedure. The Standardised Canonical Coefficients were 0.771 for the timing of the change in moment direction from flexion to extension and 0.375 for the magnitude of the maximum flexion moment. Peak knee flexion angle was significantly reduced in Cluster 2 patients compared to controls (Table 2). The peak knee flexion moment was significantly reduced in patients in both clusters compared to controls, and was significantly reduced in patients in Cluster 2 compared to patients in Cluster 1. There were no significant differences between patients in Cluster 1 and those in Cluster 2 during stair ascent in terms of age, body mass index, gender
J.A. McClelland et al. / Journal of Biomechanics 47 (2014) 1816–1821
Cluster 1
Cluster 2
TKR group average
1819
Control group average
Fig. 2. Knee flexion moment throughout stair ascent (A) and stair descent (B) for each TKA participant.
Table 2 Comparison of peak knee biomechanics between clusters of patients with TKA and controls for both stair ascent and stair descent. Control group
Cluster 1
Cluster 2
57.7 (5.5) 2.7 (1.2)n
53.4 (8.5)n 0.3 (0.2)n,nn
21.6 (5.8)
18.1 (7.1)
3.7 (5.4)n,nn
3.2 (1.1)
2.6 (1.5)
0.3 (0.7)n,nn
6.1 (1.2)
5.6 (1.4)
4.1 (1.4)n,nn
Stair ascent Maximum knee flexion (deg) 60.8 (5.4) Maximum flexion moment (% Bw-Ht) 3.8 (1.2) Stair descent Maximum flexion angle during loading (deg) Maximum flexion moment during loading (% Bw-Ht) Maximum flexion moment at end stance (% Bw-Ht)
TKA group
n
Denotes significant difference in comparison to control group p r 0.05. Denotes significant difference between patients in Cluster 1 and patients in Cluster 2 p r0.05. nn
Table 3 Comparison of peak knee biomechanics between patients with unilateral TKA and patients with bilateral TKA for both stair ascent and stair descent.
Stair ascent Stride time (s) Time of change in moment direction (% step cycle) Maximum knee flexion (degrees) Maximum flexion moment (% Bw-Ht)
Unilateral TKA
Bilateral TKA
1.2 (0.1) 49.5 (21.7)
1.2 (0.2) p¼ 0.26 41.8 (30.5) p¼ 0.47
55.2 (6.3) 2.1 (1.3)
58.9 (7.5) 2.5 (1.8)
Stair descent Stride time (s) 1.1 (0.3) Time of change in moment direction 20.0 (14.3) (% step cycle) Maximum flexion angle during loading 12.6 (8.4) (deg) Maximum flexion moment during 1.4 (1.7) loading (% Bw-Ht) Maximum flexion moment at end stance 5.1 (1.7) (% Bw-Ht)
Level of significance
p¼ 0.21 p¼ 0.47
1.3 (0.2) p¼ 0.11 18.7 (12.1) p¼ 0.84 13.8 (12.2) p¼ 0.79 1.8 (1.6)
p¼ 0.59
5.1 (1.4)
p¼ 0.95
distribution, pain in the contralateral limb, KSS, or satisfaction with TKA from the TKFQ. Peak knee flexion angles and moments were not significantly different between patients with bilateral TKA and patients with unilateral TKA (Table 3).
3.3. Knee biomechanics of stair descent All of the control participants and 14 of the 21 patients with TKA (67% of patients that completed the task or 35% of all patients) were classified in Cluster 1 (Fig. 2B). The remaining seven patients with TKA (33% of patients that completed the task or 18% of all patients) were classified in Cluster 2. For all participants, an initial knee extension changed direction to become a flexion moment. However, for patients in Cluster 2, the moment pattern was distinctly different from normal and was characterised by a delay in this change of direction and the absence of a flexion moment in early stance. Confirming these clusters, all participants in Cluster 1 (0% misclassification rate) and all but one case (14% misclassification rate) were correctly classified in the leave-one-out procedure. The Standardised Canonical Coefficients were 1.012 for the timing of the change in moment direction from extension to flexion and 0.068 for the magnitude of the maximum flexion moment. Patients with an abnormal knee moment pattern (Cluster 2) descended stairs with less knee flexion and lower peak knee flexion moments than both the control group and the patients in Cluster 1 (Table 2). Peak knee flexion angles and moments were not significantly different between patients with bilateral TKA and patients with unilateral TKA (Table 3). There were no significant differences between patients in Cluster 1 and those in Cluster 2 for stair descent in terms of age, body mass index, gender distribution, pain in the contralateral limb, KSS, or satisfaction with TKA from the TKFQ.
4. Discussion Fewer patients than controls were able to complete both the ascent and descent tasks. In previous studies where patients with TKA were asked to complete a similar activity, there were no reports of patients failing to complete the task (Andriacchi et al., 1982, 1997; Berti et al., 2006; Bolanos et al., 1998; Fantozzi et al., 2003; Mandeville et al., 2007, 2008; Wilson et al., 1996). However, these studies were designed specifically to evaluate prosthesis design characteristics, and were restrictive in their inclusion of only patients who had achieved good or excellent outcome from surgery. In the current study, all consecutive patients were invited to participate and there was no minimal level of function required. The current results may therefore be a more accurate representation of post-operative stair climbing ability in the broader TKA clinical population. Collectively, these findings add to information from the growing body of literature that suggest that, as a group, patients
1820
J.A. McClelland et al. / Journal of Biomechanics 47 (2014) 1816–1821
with TKA do not regain functional abilities that are similar to people of similar age and sex that do not have known knee pathology. Using cluster analysis, we found that each patient ascended stairs with a sagittal plane knee moment that changed magnitude and direction in one of two distinct patterns. Although a distinction between ‘normal’ and ‘abnormal’ biomechanical patterns during stair ascent has been previously described (Andriacchi et al., 1982, 1997; Catani et al., 2003), investigation by cluster analysis allows for a more definitive description of the different patterns. This is also the first study to report that two distinct patterns of the sagittal plane knee moment may also be identified during stair descent. In both activities, most of the patients who could ascend and descend stairs did so with a moment that changed direction in a similar pattern to all of the control participants, and may be considered representative of normal movement. Most of the peak knee biomechanics of these patients were also not different from normal. Therefore, half of all patients can expect to ascend stairs with knee biomechanics that are similar to normal 12 months after TKA, and up to 33% can also expect a similar outcome when descending stairs. Despite these more optimistic outcomes, one in five patients who were able to complete the stair tasks, but did so in a way that was distinctly abnormal. The defining characteristic of this subgroup of patients was the apparent avoidance of generating a knee flexion moment, as evidenced by a reduction in magnitude of the knee flexion moment and a premature change to a knee extension moment. The absence of large knee flexion moments has been interpreted as a preference for patients to either avoid using the quadriceps, or to create a secondary stabilising co-contraction between the quadriceps and the hamstrings (Andriacchi et al., 1982; Rudolph et al., 2001). In patients with TKA, the presence of this pattern has been explained as a compensatory reaction to alleged instability of specific prosthetic design characteristics. In the current study, all patients were implanted the same TKA prosthesis, yet only a relatively small proportion of patients who completed the activity adopted this abnormal pattern. The proportion of patients who demonstrated an abnormal moment pattern in the current study was too small to allow meaningful statistical analysis of the potential factors contributing to its development. There is evidence that post-operative rehabilitation programs are effective in improving function and biomechanics for patients after TKA (Kumar et al., 1996; McClelland et al., 2012; Petterson et al., 2009; Piva et al., 2010). It is possible that rehabilitative strategies that specifically target retraining stair climbing in patients with TKA may improve biomechanical outcome. Stair descent appears to provide a greater challenge for patients than stair ascent. Fewer patients were able to descend the stairs than ascend, and an abnormal moment pattern was present in a greater proportion of patients who could complete the task during descent than during ascent. These findings were reflected in the patients' self-report questionnaires, in which almost threequarters of the patients reported that stair descent was limited compared to one-third for stair ascent. Therefore investigation of stair descent as a high level outcome measure of function after TKA may be warranted. There are some limitations of this study that should be considered. The patients with TKA were significantly heavier than the control participants. It is possible that this disparity could account for the differences in biomechanics between groups. However, given the high incidence of obesity in this patient population, selecting participants to match for body mass would limit the generalisability of the findings of this study (Zeni and Snyder-Mackler, 2010a). A further limitation is that the staircase used in this study was restricted to two elevated steps. A two-step ascent or descent is commonly required in community ambulation
and these findings may represent these situations. However it is not known whether similar biomechanics to those described are used when patients ascend or descend a flight of stairs. Furthermore, a pre-operative assessment of stair ascent and descent was not possible in this study. Therefore, we were unable to investigate the role that pre-operative stair-climbing may have in determining outcome from TKA. In summary, almost half of all patients with TKA in this study were unable to ascend and descend a two-step staircase without assistance compared to 80% of people without TKA. Although the majority of these patients climbed stairs in a way that was similar to normal, a distinct subgroup of patients exists who adopted abnormal strategies to ascend or descend the stairs. Rehabilitation strategies that specifically address these characteristics may improve stair climbing ability in patients with TKA.
Conflict of interest statement The authors have no conflicts of interest. References Andriacchi, T.P., Galante, J.O., Fermier, R.W., 1982. The influence of total kneereplacement design on walking and stair-climbing. J. Bone Joint Surg. (Am.) 64A (9), 1328–1335. Andriacchi, T.P., Yoder, D., Conley, A., Rosenberg, A., Sum, J., Galante, J.O., 1997. Patellofemoral design influences function following total knee arthroplasty. J. Arthroplast. 12 (3), 243–249. Australian Orthopaedic Association, 2012. Annual Report. National Joint Replacement Registry. Retrieved on 8th April, 2013 from 〈http://www.aoa.org.au/docs/ miscellaneous/click-here-to-view-the-report-.pdf?sfvrsn=0〉. Berti, L., Benedetti, M.G., Ensini, A., Catani, F., Giannini, S., 2006. Clinical and biomechanical assessment of patella resurfacing in total knee arthroplasty. Clin. Biomech. 21 (6), 610–616. Bolanos, A.A., Colizza, W.A., McCann, P.D., Gotlin, R.S., Wootten, M.E., Kahn, B.A., Insall, J.N., 1998. A comparison of isokinetic strength testing and gait analysis in patients with posterior cruciate-retaining and substituting knee arthroplasties. J. Arthroplast. 13 (8), 906–915. Bourne, B.B., Chesworth, B.M., Davis, A.M., Mahomed, N.N., Charron, K.D.J., 2010. Patient satisfaction after total knee arthroplasty: who is satisfied and who is not? Clin. Orthop. Relat. Res. 468 (1), 57–63. Catani, F., Benedetti, M.G., De Felice, R., Buzzi, R., Giannini, S., Aglietti, P., 2003. Mobile and fixed bearing total knee prosthesis functional comparison during stair climbing. Clin. Biomech. 18 (5), 410–418. Davis 3rd, R.B., Ounpuu, S., Tyburski, D., Gage, J.R., 1991. A gait analysis data collection and reduction technique. Hum. Mov. Sci. 10, 575–587. Fantozzi, S., Benedetti, M.G., Leardini, A., Banks, S.A., Cappello, A., Assirelli, D., Catani, F., 2003. Fluoroscopic and gait analysis of the functional performance in stair ascent of two total knee replacement designs. Gait Posture 17, 225–234. Insall, J.N., Dorr, L.D., Scott, R.D., Scott, W.N., 1989. Rationale of the knee society clinical rating system. Clin. Orthop. 248, 13–14. Kadaba, M.P., Ramakrishnan, H.K., Wootten, M.E., Gainey, J., Gorton, G., Cochran, G.V., 1989. Repeatability of kinematic, kinetic, and electromyographic data in normal adult gait. J. Orthop. Res. 7 (6), 849–860. Kumar, P.J., McPherson, E.J., Dorr, L.D., Wan, Z., Baldwin, K., 1996. Rehabilitation after total knee arthroplasty: a comparison of 2 rehabilitation techniques. Clin. Orthop. Relat. Res. 331, 93–101. Lingard, E.A., Katz, J.N., Wright, E.A., Sledge, C.B., Kinemax Outcomes, G., 2004. Predicting the outcome of total knee arthroplasty. J. Bone Joint Surg. (Am) 86-A (10), 2179–2186. Mandeville, D., Osternig, L.R., Chou, L.-S., 2007. The effect of total knee replacement on dynamic support of the body during walking and stair ascent. Clin. Biomech. 22, 787–794. Mandeville, D., Osternig, L.R., Lantz, B.A., Mohler, C.G., Chou, L.S., 2008. The effect of total knee replacement on the knee varus angle and moment during walking and stair ascent. Clin. Biomech. 23 (8), 1053–1058. McClelland, J.A., Webster, K.E., Feller, J.A., 2007. Gait analysis of patients following total knee replacement: a systematic review. Knee 14 (4), 253–263. McClelland, J.A., Webster, K.E., Grant, C., Feller, J., 2010. Alternative modelling procedures for pelvic marker occlusion during motion analysis. Gait Posture 31 (4), 415–419. McClelland, J.A., Zeni, J.A., Haley, R., Snyder-Mackler, L., 2012. Functional and biomechanical improvements using biofeedback for retraining symmetrical movement patterns after total knee arthroplasty: a case report. J. Orthop. Sports Phys. Ther. 42 (2), 135–144. McMahon, M., Block, J.A., 2003. The risk of contralateral total knee arthroplasty after knee replacement for osteoarthritis. J. Rheumatol. 30 (8), 1822–1824.
J.A. McClelland et al. / Journal of Biomechanics 47 (2014) 1816–1821
Mulroy, S., Gronley, J., Weiss, W., Newsam, C., Perry, J., 2003. Use of cluster analysis for gait pattern classification of patients in the early and late recovery phases following stroke. Gait Posture 18 (1), 114–125. Office of the Australian Building and Construction Commissioner, 2006. National Code, August 19th, 2008. Retrieved on June 25th, 2008 from 〈http://www.abcc. gov.au/abcc/NationalCode/?gclid=CN_H7-mbmZUCFRlRagodbRzZPA〉. Petterson, S.C., Mizner, R.L., Stevens, J.E., Raisis, L., Bodenstab, A., Newcomb, W., Snyder-Mackler, L., 2009. Improved function from progressive strengthening interventions after total knee arthroplasty: a randomized clinical trial with an imbedded prospective cohort. Arthritis Rheum. 61 (2), 174–183. Piva, S.R., Gil, A.B., Almeida, G.J., DiGioia 3rd, A.M., Levison, T.J., Fitzgerald, G.K., 2010. A balance exercise program appears to improve function for patients with total knee arthroplasty: a randomized clinical trial. Phys. Ther. 90 (6), 880–894. Rudolph, K.S., Axe, M.J., Buchanan, T.S., Scholz, J.P., Snyder-Mackler, L., 2001. Dynamic stability in the anterior cruciate ligament deficient knee. Knee Surg. Sports Traumatol. Arthrosc. 9 (2), 62–71. Shakoor, N., Block, J.A., Shott, S., Case, J.P., 2002. Nonrandom evolution of end-stage osteoarthritis of the lower limbs. Arthritis Rheum. 46 (12), 3185–3189.
1821
Toro, B., Nester, C.J., Farren, P.C., 2007. Cluster analysis for the extraction of sagittal gait patterns in children with cerebral palsy. Gait Posture 25 (2), 157–165. Ward, J.H.J., 1963. Hierarchical grouping to optimize an objective function. J. Am. Stat. Assoc. 58, 236–244. Weiss, J.M., Noble, P.C., Conditt, M.A., Kohl, H.W., Roberts, S., Cook, K.F., Mathis, K.B., 2002. What functional activities are important to patients with knee replacements? Clin. Orthop. Relat. Res. 404, 172–188. Wilson, S.A., McCann, P.D., Gotlin, R.S., Ramakrishnan, H.K., Wootten, M.E., Insall, J.N., 1996. Comprehensive gait analysis in posterior-stabilized knee arthroplasty. J. Arthroplast. 11 (4), 359–367. Zeni Jr, J.A., Snyder-Mackler, L., 2010a. Most patients gain weight in the 2 years after total knee arthroplasty: comparison to a healthy control group. Osteoarthr. Cartil. 18 (4), 510–514. Zeni Jr, J.A., Snyder-Mackler, L., 2010b. Preoperative predictors of persistent impairments during stair ascent and descent after total knee arthroplasty. J. Bone Joint Surg. (Am.) 92-A (5), 1130–1136.
Contents lists available at ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Patterns in the knee flexion-extension moment profile during stair ascent and descent in patients with total knee arthroplasty Jodie A. McClelland a,n, Julian A. Feller b, Hylton B. Menz a, Kate E. Webster a a b
School of Allied Health, Faculty of Health Sciences, La Trobe University, Melbourne, Australia OrthoSport Victoria, Epworth HealthCare, Melbourne, Australia
art ic l e i nf o
a b s t r a c t
Article history: Accepted 19 March 2014
The aim of this study was to investigate the prevalence of abnormal knee biomechanical patterns in 40 patients with a modern TKA prosthesis, compared to 40 matched control participants when ascending and descending stairs. Fewer patients were able to ascend (65%) or descend stairs (53%) unassisted than controls (83%). Of the participants who could ascend and descend, cluster analysis classified most patients (up to 77%) as demonstrating a similar knee moment pattern as all controls. A small subgroup of patients who completed the tasks did so with distinctly abnormal biomechanics compared to other patients and controls. These findings suggest that recovery of normal stair climbing is possible. However, rehabilitation might be more effective if it were tailored to account for these differences between patients. & 2014 Elsevier Ltd. All rights reserved.
Keywords: Knee arthroplasty Stairs Motion analysis Gait analysis Flexion
1. Introduction The ability to ascend and descend stairs is a task that is encountered frequently during daily living, particularly as stairs often provide the most convenient access point to buildings and services in the community. Unfortunately, it has been reported that stair climbing is the daily functional activity for which patients with total knee arthroplasty (TKA) are most dissatisfied with their outcome from surgery and there is evidence that stair climbing ability is not improved following TKA, and remains significantly impaired for up to 2 years following surgery (Bourne et al., 2010; Zeni and Snyder-Mackler, 2010b). Measuring knee biomechanics using motion analysis techniques has the benefit of providing information about the specific challenges faced by patients with TKA when negotiating stairs. Using this approach, the external knee joint flexion–extension moment pattern throughout stair ascent has been assessed in three previous studies (Andriacchi et al., 1982; Andriacchi et al., 1997; Catani et al., 2003). Each of these studies identified a subgroup of patients for whom the knee moment pattern was distinctly different from the other patients, and also from what is
n Correspondence to: Department of Physiotherapy, Faculty of Health Sciences, La Trobe University, Bundoora, Victoria 3086, Australia. Tel.: þ 61 3 9479 3254; fax: þ61 3 9479 5415. E-mail addresses: [email protected] (J.A. McClelland), [email protected] (J.A. Feller), [email protected] (H.B. Menz), [email protected] (K.E. Webster).
http://dx.doi.org/10.1016/j.jbiomech.2014.03.026 0021-9290/& 2014 Elsevier Ltd. All rights reserved.
typically expected of normal stair ascent (Andriacchi et al., 1982; Andriacchi et al., 1997; Catani et al., 2003). All authors of the previous studies agreed that design characteristics of some TKA prostheses may contribute to the development of an abnormal moment pattern during stair ascent, and particularly that the abnormal moment profile represented an attempt to avoid generating an external flexion moment (Andriacchi et al., 1982; Andriacchi et al., 1997; Catani et al., 2003). All three authors described a distinctly abnormal moment pattern in 40–60% of the patients. All authors agreed that characteristics of the sagittal moment in this proportion of patients suggested an attempt to avoid generating a knee flexion moment, which also avoided the need for quadriceps activity. Two of these studies were completed more than 20 years ago, and all of these studies assessed only patients with good or excellent outcome from TKA. Without recent studies investigating the profile of the sagittal plane moment it is not known what proportion of patients with modern prostheses may expect to demonstrate abnormal moment patterns, and how the biomechanical characteristics of these patients differ from normal. Therefore, an investigation of a larger cohort that is representative of typical contemporary patients would allow for more detailed description of the biomechanics of stair ascent and descent following TKA. The demands of descending stairs may be different from stair ascent, as the primary role of the quadriceps during descent is to absorb the force needed to lower bodyweight, rather than generate force for elevation, as in stair ascent. However, despite the potentially greater challenge that stair descent presents, there have been few studies that attempted to describe the biomechanics of patients during stair descent (Andriacchi et al., 1982; Bolanos et al., 1998;
J.A. McClelland et al. / Journal of Biomechanics 47 (2014) 1816–1821
Catani et al., 2003). None of these studies explored the prevalence of an abnormal pattern of knee moment during stair descent, therefore little is known about how the demand for quadriceps control changes throughout the activity. The aim of this study was therefore to investigate the prevalence of abnormal knee flexion–extension patterns during both stair ascent and descent in a representative group of patients following TKA with reference to a healthy control group. We hypothesised that most patients with TKA would ascend and descend stairs with knee moments that were similar to people without knee pathology. We also hypothesised that in a smaller proportion of patients the knee moment would be smaller in magnitude and delayed.
2. Materials and methods All patients who underwent TKA by a single experienced knee surgeon over 2 consecutive years were assessed for eligibility for study inclusion. Patients were invited to participate if they had undergone TKA for osteoarthritis at least twelve months and not more than 18 months earlier and were able to walk 10 metres without a gait aid. Patients with other documented orthopaedic, neurological or visual disturbances that may affect gait, including joint arthroplasties of the hips and ankles were excluded. Within 5 years, up to 30% of patients with primary TKA undergo TKA in the contralateral limb (Australian Orthopaedic Association, 2012; McMahon and Block, 2003; Shakoor et al., 2002). Therefore, to maximise the generalisability of the findings of this study, patients with bilateral knee arthroplasty were included, provided that the most recent procedure was undertaken at least 12 months prior to testing. From an initial list of 78 patients, 22 were excluded because of co-morbidities that may affect gait, nine declined and seven were unable to be contacted. The remaining 40 patients attended the gait laboratory for assessment. All participants had received a fully cemented Genesis-II posterior stabilised prosthesis (Smith and Nephew, Memphis, Tennessee, USA) and the patella was resurfaced in all knees. All participants received standard postoperative care, which emphasised early weightbearing mobility, quadriceps strengthening and maximisation of range of motion. In the 13 participants that had undergone bilateral knee arthroplasty surgery, only the most recent TKA was considered for analysis. The control group consisted of 40 participants who were assessed for eligibility to participate in the study using the same criteria as the TKA participants, except that control participants were excluded if they had undergone any joint arthroplasty surgery or reported any symptoms of lower limb osteoarthritis. All control participants were matched to the age (7 2 years) and gender of a member of the TKA group. A single limb of each control participant was used in the analysis and was determined by the operated limb of the matched TKA participant. Ethical approval was granted from the institution's Human Ethics Committee and written informed consent was obtained from all participants. An eight camera Vicon MX3 Motion Analysis System (Vicon, Oxford, UK) was used to collect kinematic data at a sampling rate of 100 Hz. The gait laboratory consisted of a 10 m walkway with two force plates (Kistler, Switzerland and AMTI, Watertwon, MA, USA) that were embedded into the floor in the middle of the walkway. A two-step staircase with dimensions based on the Building Code of Australia (step height of 180 mm and a step tread of 300 mm) was placed adjacent to these embedded forceplates (Office of the Australian Building and Construction Commissioner, 2006). A third force plate (AMTI, Watertown, MA, USA) was placed on the first step of the staircase so that the kinetic data of three consecutive steps could be collected (two level steps and one elevated step). Only data from this elevated force plate is included in this manuscript. All force plate data were collected at a sampling rate of 4000 Hz. Thirteen 14 mm diameter reflective markers were placed centrally over the sacrum, and bilaterally over the anterior superior iliac spines, lateral femoral epicondyles, lateral and medial malleoli, calcanei, and second metatarsals. Markers on 5 cm wands were placed on the lateral thighs and legs. A knee alignment device was used to determine the centre of the knee joint during a static data calibration. These marker placements were based on the modified Helen Hayes model (Davis et al., 1991; Kadaba et al., 1989). Additional markers were placed on the iliac crest to allow reconstruction of occluded ASIS markers (McClelland et al., 2010). During stair ascent, participants were instructed to take two steps on a level walkway before ascending the staircase in a reciprocal footfall pattern. During stair descent, the process was reversed so that participants commenced on the top step, descended onto the first elevated step which included a force plate and descended again to take 2 steps along the level walkway. There were no rails for support on the staircase. Participants were encouraged to practise ascending and descending the staircase as many times as necessary until comfortable with the process. Data collection did not proceed if patients were unable to ascend or descend without support, or were unable to ascend or descend with reciprocal footsteps.
1817
Participants were asked to ascend and descend the staircase until a minimum of three and a maximum of five trials of data were collected from each limb. Patients were assessed using the American Knee Society Knee Score (KSS) at a routine 12 month post-operative follow-up visit with the surgeon (Insall et al., 1989). At the time of motion analysis testing, patients with TKA also completed the Total Knee Function Questionnaire (TKFQ), which collects information about patients' perceived functional abilities and their importance to daily life (Weiss et al., 2002). Responses relevant to overall satisfaction and stair climbing were included in this study. Vicon PlugIn Gait (Vicon, Oxford Metrics, Oxford, UK) software was used to estimate lower limb joint positions, based on the model proposed by Davis et al. (1991) and Kadaba et al. (1989). Joint kinematics were calculated using Euler angles (flexion/extension, abduction/adduction, and internal/external rotation). Joint moments were calculated using standard inverse dynamics from force vectors that were defined in a global coordinate system and projected to a local coordinate system of the rigid body segment distal to each joint. The moments referred to throughout this manuscript are external to the knee and normalised to bodyweight and height. A single step cycle from each trial was included for analysis and was time normalised to 100% of the step cycle. During stair ascent, initial contact onto the first elevated step was considered as the beginning of the step cycle (0%) and toe-off of the same foot as the end of the step cycle (100%). During stair descent, initial contact onto the first elevated step was considered as the beginning of the step cycle (0%) and toe-off of the same foot was defined as the end of the step cycle (100%). Knee flexion angles and moments in the sagittal plane were calculated for each trial of each participant. Key biomechanical variables were identified and averaged across all trials of each individual (Fig. 1). For stair ascent, these were the maximum knee flexion angle (degrees), the maximum knee flexion moment (%Bw‐Ht) and the time when the sagittal moment changed direction from flexion to extension (% of step cycle) (Fig. 1). For stair descent, these were the maximum knee flexion angle during loading phase (0–50% of the step cycle as bodyweight is accepted onto the limb) (deg), the time when the sagittal moment changed direction from extension to flexion (% of step cycle), the maximum knee flexion moment during loading (% Bw‐Ht), the maximum knee flexion moment during late stance phase (50–100% of the step cycle, %Bw‐Ht) and the time taken to complete each task (s). Hierarchical cluster analysis was used to identify subgroups of patients using characteristics of the sagittal moment profile. Two separate cluster analyses were performed in this study: (i) a single analysis to classify moment profiles of all control and TKA participants who completed the stair ascent task (n¼59) and (ii) a single analysis to classify moment profiles of all control and TKA participants who completed stair descent (n¼ 54). The hierarchical cluster analysis procedure (based on the method described by Ward (1963)) begins by assigning each case to its own cluster. There is then a stepwise process whereby moment profiles that are most similar are joined to form clusters. This process continues until a single cluster is formed in a hierarchical tree (dendogram). In this study, the similarity between gait profiles was determined by the squared Euclidean distance between profiles and by using a between-groups linkage (the average similarity within a cluster is calculated and the case with the distance most similar to the average is the next to join the cluster). Two distinct waveforms of the sagittal moment profile have been described in patients with TKA during a stair ascent task (Andriacchi et al., 1982; Andriacchi et al., 1997; Catani et al., 2003), therefore it was determined a priori that there would be two clusters in each analysis based on key characteristics of this moment. In these studies, abnormal stair ascent has been described as an apparent avoidance of quadriceps use, which was characterised by a reduction in the maximum external flexion moment, and reduced time during stance that the quadriceps were required (i.e. reduced time that the external moment was in a flexion direction). Therefore, in the cluster analysis of stair ascent the independent variables were determined a priori as (i) the magnitude of the maximum flexion moment and (ii) the percentage of the gait cycle where the knee flexion moment changed direction to become a knee extensor moment (Fig. 1C). Patterns of the sagittal moment have not been described during stair descent where the primary role of the quadriceps is to control lowering of bodyweight onto the descended step. An avoidance of demand for quadriceps control during this moment is likely to be reflected in a reduction of the maximum flexion moment and a delay in the need for quadriceps activity (i.e. delay in generation of a flexion moment). Therefore in cluster analysis of stair decent, the independent variables were defined a priori as (i) the magnitude of the maximum knee flexion moment and (ii) the percentage of the gait cycle where the initial knee extensor moment changed direction to become a knee flexor moment (Fig. 1D). Step-wise discriminant function analysis was used to evaluate the robustness of group allocation, and to evaluate the relative importance of each variable in determining the clusters. In this procedure the discriminant function is recalculated for all except one case. This model is then applied to the withheld case to assign group membership, and the process repeated until each case has been left out. Group assignment is considered robust if there are a high percentage of participants classified to the correct group (i.e. low misclassification rate). This process is similar to that described by Toro et al. (2007) and Mulroy et al. (2003). Key biomechanical characteristics during both stair ascent and descent were compared between groups, and between clusters using independent samples t-tests with po 0.05. For all biomechanical parameters reported in this manuscript,
1818
J.A. McClelland et al. / Journal of Biomechanics 47 (2014) 1816–1821
A
Ascent
B
Descent
*
*
C
D ** **
***
***
Fig. 1. Typical waveforms for the knee flexion angle throughout the step cycle of stair ascent (A) and stair descent (B); the knee flexion moment through stair ascent (C) and stair descent (D). The key variables used in data analysis were the maximum flexion during loading (*), the maximum knee flexion moment (**), and the time in moment direction change (***). patients with unilateral TKA were compared to participants with bilateral TKA using an independent t-test. As there were no differences between groups, these groups were collapsed to form a single TKA group for the cluster analysis.
3. Results The demographic characteristics of all participants are shown in Table 1. The TKA group was heavier than the control group, which is similar to other studies that compare TKA patients to participants of similar demographic characteristics without TKA (Lingard et al., 2004; McClelland et al., 2007; Zeni and SnyderMackler, 2010a). 3.1. Stair climbing All control participants and all but one TKA participant were able to ascend and descend. However, after several attempts, only 33 control participants (83%) and 26 TKA participants (65% of all patients) could complete stair ascent independently. The same 33 control participants (83%) and 21 of these 26 TKA participants (53% of all patients) were able to complete the stair descent task. 3.2. Knee biomechanics of stair ascent All of the control participants and 20 TKA participants (77% of patients that completed the task or 50% of all patients) were classified in Cluster 1 (grey traces in Fig. 2A), which was declared a normal knee moment pattern. The remaining six TKA participants (23% of patients that completed the task or 15% of all patients) were classified in Cluster 2 (black traces). All participants generated a knee flexion moment that changed direction to an extension moment during stair ascent. However, for patients in
Table 1 Demographic characteristics of TKA and control participants.
Age (years) Height (cm) Weight (kg) F:M Right limb affected:left limb affected Knee Society Score (/100) Knee Society Function Score (/100) Total Knee Function Questionaire Responses Extremely satisfied with TKA Stair ascent ‘not limited at all’ Stair descent ‘not limited at all’ n
Control group
TKA group
69.6 (8.3) 165.0 (9.5) 76.2 (17.9) 22:18 N/A N/A N/A
69.1 (8.0) 165.5 (11.5) 86.4 (17.5)n 22:18 24:16 79.1 (11.2) 86.7 (20.8)
N/A N/A N/A
78% 65% 27%
Denotes significant difference in comparison to control group p r 0.05.
Cluster 2, the moment pattern was distinctly different from normal and was characterised by a reduction in the magnitude of the flexion moment and a change in moment direction from flexion to extension that occurred earlier in the step cycle. In confirmation of this cluster analysis, all participants in Cluster 1 (0% misclassification rate) and all but one case (17% misclassification rate) were correctly classified in the leave-one-out procedure. The Standardised Canonical Coefficients were 0.771 for the timing of the change in moment direction from flexion to extension and 0.375 for the magnitude of the maximum flexion moment. Peak knee flexion angle was significantly reduced in Cluster 2 patients compared to controls (Table 2). The peak knee flexion moment was significantly reduced in patients in both clusters compared to controls, and was significantly reduced in patients in Cluster 2 compared to patients in Cluster 1. There were no significant differences between patients in Cluster 1 and those in Cluster 2 during stair ascent in terms of age, body mass index, gender
J.A. McClelland et al. / Journal of Biomechanics 47 (2014) 1816–1821
Cluster 1
Cluster 2
TKR group average
1819
Control group average
Fig. 2. Knee flexion moment throughout stair ascent (A) and stair descent (B) for each TKA participant.
Table 2 Comparison of peak knee biomechanics between clusters of patients with TKA and controls for both stair ascent and stair descent. Control group
Cluster 1
Cluster 2
57.7 (5.5) 2.7 (1.2)n
53.4 (8.5)n 0.3 (0.2)n,nn
21.6 (5.8)
18.1 (7.1)
3.7 (5.4)n,nn
3.2 (1.1)
2.6 (1.5)
0.3 (0.7)n,nn
6.1 (1.2)
5.6 (1.4)
4.1 (1.4)n,nn
Stair ascent Maximum knee flexion (deg) 60.8 (5.4) Maximum flexion moment (% Bw-Ht) 3.8 (1.2) Stair descent Maximum flexion angle during loading (deg) Maximum flexion moment during loading (% Bw-Ht) Maximum flexion moment at end stance (% Bw-Ht)
TKA group
n
Denotes significant difference in comparison to control group p r 0.05. Denotes significant difference between patients in Cluster 1 and patients in Cluster 2 p r0.05. nn
Table 3 Comparison of peak knee biomechanics between patients with unilateral TKA and patients with bilateral TKA for both stair ascent and stair descent.
Stair ascent Stride time (s) Time of change in moment direction (% step cycle) Maximum knee flexion (degrees) Maximum flexion moment (% Bw-Ht)
Unilateral TKA
Bilateral TKA
1.2 (0.1) 49.5 (21.7)
1.2 (0.2) p¼ 0.26 41.8 (30.5) p¼ 0.47
55.2 (6.3) 2.1 (1.3)
58.9 (7.5) 2.5 (1.8)
Stair descent Stride time (s) 1.1 (0.3) Time of change in moment direction 20.0 (14.3) (% step cycle) Maximum flexion angle during loading 12.6 (8.4) (deg) Maximum flexion moment during 1.4 (1.7) loading (% Bw-Ht) Maximum flexion moment at end stance 5.1 (1.7) (% Bw-Ht)
Level of significance
p¼ 0.21 p¼ 0.47
1.3 (0.2) p¼ 0.11 18.7 (12.1) p¼ 0.84 13.8 (12.2) p¼ 0.79 1.8 (1.6)
p¼ 0.59
5.1 (1.4)
p¼ 0.95
distribution, pain in the contralateral limb, KSS, or satisfaction with TKA from the TKFQ. Peak knee flexion angles and moments were not significantly different between patients with bilateral TKA and patients with unilateral TKA (Table 3).
3.3. Knee biomechanics of stair descent All of the control participants and 14 of the 21 patients with TKA (67% of patients that completed the task or 35% of all patients) were classified in Cluster 1 (Fig. 2B). The remaining seven patients with TKA (33% of patients that completed the task or 18% of all patients) were classified in Cluster 2. For all participants, an initial knee extension changed direction to become a flexion moment. However, for patients in Cluster 2, the moment pattern was distinctly different from normal and was characterised by a delay in this change of direction and the absence of a flexion moment in early stance. Confirming these clusters, all participants in Cluster 1 (0% misclassification rate) and all but one case (14% misclassification rate) were correctly classified in the leave-one-out procedure. The Standardised Canonical Coefficients were 1.012 for the timing of the change in moment direction from extension to flexion and 0.068 for the magnitude of the maximum flexion moment. Patients with an abnormal knee moment pattern (Cluster 2) descended stairs with less knee flexion and lower peak knee flexion moments than both the control group and the patients in Cluster 1 (Table 2). Peak knee flexion angles and moments were not significantly different between patients with bilateral TKA and patients with unilateral TKA (Table 3). There were no significant differences between patients in Cluster 1 and those in Cluster 2 for stair descent in terms of age, body mass index, gender distribution, pain in the contralateral limb, KSS, or satisfaction with TKA from the TKFQ.
4. Discussion Fewer patients than controls were able to complete both the ascent and descent tasks. In previous studies where patients with TKA were asked to complete a similar activity, there were no reports of patients failing to complete the task (Andriacchi et al., 1982, 1997; Berti et al., 2006; Bolanos et al., 1998; Fantozzi et al., 2003; Mandeville et al., 2007, 2008; Wilson et al., 1996). However, these studies were designed specifically to evaluate prosthesis design characteristics, and were restrictive in their inclusion of only patients who had achieved good or excellent outcome from surgery. In the current study, all consecutive patients were invited to participate and there was no minimal level of function required. The current results may therefore be a more accurate representation of post-operative stair climbing ability in the broader TKA clinical population. Collectively, these findings add to information from the growing body of literature that suggest that, as a group, patients
1820
J.A. McClelland et al. / Journal of Biomechanics 47 (2014) 1816–1821
with TKA do not regain functional abilities that are similar to people of similar age and sex that do not have known knee pathology. Using cluster analysis, we found that each patient ascended stairs with a sagittal plane knee moment that changed magnitude and direction in one of two distinct patterns. Although a distinction between ‘normal’ and ‘abnormal’ biomechanical patterns during stair ascent has been previously described (Andriacchi et al., 1982, 1997; Catani et al., 2003), investigation by cluster analysis allows for a more definitive description of the different patterns. This is also the first study to report that two distinct patterns of the sagittal plane knee moment may also be identified during stair descent. In both activities, most of the patients who could ascend and descend stairs did so with a moment that changed direction in a similar pattern to all of the control participants, and may be considered representative of normal movement. Most of the peak knee biomechanics of these patients were also not different from normal. Therefore, half of all patients can expect to ascend stairs with knee biomechanics that are similar to normal 12 months after TKA, and up to 33% can also expect a similar outcome when descending stairs. Despite these more optimistic outcomes, one in five patients who were able to complete the stair tasks, but did so in a way that was distinctly abnormal. The defining characteristic of this subgroup of patients was the apparent avoidance of generating a knee flexion moment, as evidenced by a reduction in magnitude of the knee flexion moment and a premature change to a knee extension moment. The absence of large knee flexion moments has been interpreted as a preference for patients to either avoid using the quadriceps, or to create a secondary stabilising co-contraction between the quadriceps and the hamstrings (Andriacchi et al., 1982; Rudolph et al., 2001). In patients with TKA, the presence of this pattern has been explained as a compensatory reaction to alleged instability of specific prosthetic design characteristics. In the current study, all patients were implanted the same TKA prosthesis, yet only a relatively small proportion of patients who completed the activity adopted this abnormal pattern. The proportion of patients who demonstrated an abnormal moment pattern in the current study was too small to allow meaningful statistical analysis of the potential factors contributing to its development. There is evidence that post-operative rehabilitation programs are effective in improving function and biomechanics for patients after TKA (Kumar et al., 1996; McClelland et al., 2012; Petterson et al., 2009; Piva et al., 2010). It is possible that rehabilitative strategies that specifically target retraining stair climbing in patients with TKA may improve biomechanical outcome. Stair descent appears to provide a greater challenge for patients than stair ascent. Fewer patients were able to descend the stairs than ascend, and an abnormal moment pattern was present in a greater proportion of patients who could complete the task during descent than during ascent. These findings were reflected in the patients' self-report questionnaires, in which almost threequarters of the patients reported that stair descent was limited compared to one-third for stair ascent. Therefore investigation of stair descent as a high level outcome measure of function after TKA may be warranted. There are some limitations of this study that should be considered. The patients with TKA were significantly heavier than the control participants. It is possible that this disparity could account for the differences in biomechanics between groups. However, given the high incidence of obesity in this patient population, selecting participants to match for body mass would limit the generalisability of the findings of this study (Zeni and Snyder-Mackler, 2010a). A further limitation is that the staircase used in this study was restricted to two elevated steps. A two-step ascent or descent is commonly required in community ambulation
and these findings may represent these situations. However it is not known whether similar biomechanics to those described are used when patients ascend or descend a flight of stairs. Furthermore, a pre-operative assessment of stair ascent and descent was not possible in this study. Therefore, we were unable to investigate the role that pre-operative stair-climbing may have in determining outcome from TKA. In summary, almost half of all patients with TKA in this study were unable to ascend and descend a two-step staircase without assistance compared to 80% of people without TKA. Although the majority of these patients climbed stairs in a way that was similar to normal, a distinct subgroup of patients exists who adopted abnormal strategies to ascend or descend the stairs. Rehabilitation strategies that specifically address these characteristics may improve stair climbing ability in patients with TKA.
Conflict of interest statement The authors have no conflicts of interest. References Andriacchi, T.P., Galante, J.O., Fermier, R.W., 1982. The influence of total kneereplacement design on walking and stair-climbing. J. Bone Joint Surg. (Am.) 64A (9), 1328–1335. Andriacchi, T.P., Yoder, D., Conley, A., Rosenberg, A., Sum, J., Galante, J.O., 1997. Patellofemoral design influences function following total knee arthroplasty. J. Arthroplast. 12 (3), 243–249. Australian Orthopaedic Association, 2012. Annual Report. National Joint Replacement Registry. Retrieved on 8th April, 2013 from 〈http://www.aoa.org.au/docs/ miscellaneous/click-here-to-view-the-report-.pdf?sfvrsn=0〉. Berti, L., Benedetti, M.G., Ensini, A., Catani, F., Giannini, S., 2006. Clinical and biomechanical assessment of patella resurfacing in total knee arthroplasty. Clin. Biomech. 21 (6), 610–616. Bolanos, A.A., Colizza, W.A., McCann, P.D., Gotlin, R.S., Wootten, M.E., Kahn, B.A., Insall, J.N., 1998. A comparison of isokinetic strength testing and gait analysis in patients with posterior cruciate-retaining and substituting knee arthroplasties. J. Arthroplast. 13 (8), 906–915. Bourne, B.B., Chesworth, B.M., Davis, A.M., Mahomed, N.N., Charron, K.D.J., 2010. Patient satisfaction after total knee arthroplasty: who is satisfied and who is not? Clin. Orthop. Relat. Res. 468 (1), 57–63. Catani, F., Benedetti, M.G., De Felice, R., Buzzi, R., Giannini, S., Aglietti, P., 2003. Mobile and fixed bearing total knee prosthesis functional comparison during stair climbing. Clin. Biomech. 18 (5), 410–418. Davis 3rd, R.B., Ounpuu, S., Tyburski, D., Gage, J.R., 1991. A gait analysis data collection and reduction technique. Hum. Mov. Sci. 10, 575–587. Fantozzi, S., Benedetti, M.G., Leardini, A., Banks, S.A., Cappello, A., Assirelli, D., Catani, F., 2003. Fluoroscopic and gait analysis of the functional performance in stair ascent of two total knee replacement designs. Gait Posture 17, 225–234. Insall, J.N., Dorr, L.D., Scott, R.D., Scott, W.N., 1989. Rationale of the knee society clinical rating system. Clin. Orthop. 248, 13–14. Kadaba, M.P., Ramakrishnan, H.K., Wootten, M.E., Gainey, J., Gorton, G., Cochran, G.V., 1989. Repeatability of kinematic, kinetic, and electromyographic data in normal adult gait. J. Orthop. Res. 7 (6), 849–860. Kumar, P.J., McPherson, E.J., Dorr, L.D., Wan, Z., Baldwin, K., 1996. Rehabilitation after total knee arthroplasty: a comparison of 2 rehabilitation techniques. Clin. Orthop. Relat. Res. 331, 93–101. Lingard, E.A., Katz, J.N., Wright, E.A., Sledge, C.B., Kinemax Outcomes, G., 2004. Predicting the outcome of total knee arthroplasty. J. Bone Joint Surg. (Am) 86-A (10), 2179–2186. Mandeville, D., Osternig, L.R., Chou, L.-S., 2007. The effect of total knee replacement on dynamic support of the body during walking and stair ascent. Clin. Biomech. 22, 787–794. Mandeville, D., Osternig, L.R., Lantz, B.A., Mohler, C.G., Chou, L.S., 2008. The effect of total knee replacement on the knee varus angle and moment during walking and stair ascent. Clin. Biomech. 23 (8), 1053–1058. McClelland, J.A., Webster, K.E., Feller, J.A., 2007. Gait analysis of patients following total knee replacement: a systematic review. Knee 14 (4), 253–263. McClelland, J.A., Webster, K.E., Grant, C., Feller, J., 2010. Alternative modelling procedures for pelvic marker occlusion during motion analysis. Gait Posture 31 (4), 415–419. McClelland, J.A., Zeni, J.A., Haley, R., Snyder-Mackler, L., 2012. Functional and biomechanical improvements using biofeedback for retraining symmetrical movement patterns after total knee arthroplasty: a case report. J. Orthop. Sports Phys. Ther. 42 (2), 135–144. McMahon, M., Block, J.A., 2003. The risk of contralateral total knee arthroplasty after knee replacement for osteoarthritis. J. Rheumatol. 30 (8), 1822–1824.
J.A. McClelland et al. / Journal of Biomechanics 47 (2014) 1816–1821
Mulroy, S., Gronley, J., Weiss, W., Newsam, C., Perry, J., 2003. Use of cluster analysis for gait pattern classification of patients in the early and late recovery phases following stroke. Gait Posture 18 (1), 114–125. Office of the Australian Building and Construction Commissioner, 2006. National Code, August 19th, 2008. Retrieved on June 25th, 2008 from 〈http://www.abcc. gov.au/abcc/NationalCode/?gclid=CN_H7-mbmZUCFRlRagodbRzZPA〉. Petterson, S.C., Mizner, R.L., Stevens, J.E., Raisis, L., Bodenstab, A., Newcomb, W., Snyder-Mackler, L., 2009. Improved function from progressive strengthening interventions after total knee arthroplasty: a randomized clinical trial with an imbedded prospective cohort. Arthritis Rheum. 61 (2), 174–183. Piva, S.R., Gil, A.B., Almeida, G.J., DiGioia 3rd, A.M., Levison, T.J., Fitzgerald, G.K., 2010. A balance exercise program appears to improve function for patients with total knee arthroplasty: a randomized clinical trial. Phys. Ther. 90 (6), 880–894. Rudolph, K.S., Axe, M.J., Buchanan, T.S., Scholz, J.P., Snyder-Mackler, L., 2001. Dynamic stability in the anterior cruciate ligament deficient knee. Knee Surg. Sports Traumatol. Arthrosc. 9 (2), 62–71. Shakoor, N., Block, J.A., Shott, S., Case, J.P., 2002. Nonrandom evolution of end-stage osteoarthritis of the lower limbs. Arthritis Rheum. 46 (12), 3185–3189.
1821
Toro, B., Nester, C.J., Farren, P.C., 2007. Cluster analysis for the extraction of sagittal gait patterns in children with cerebral palsy. Gait Posture 25 (2), 157–165. Ward, J.H.J., 1963. Hierarchical grouping to optimize an objective function. J. Am. Stat. Assoc. 58, 236–244. Weiss, J.M., Noble, P.C., Conditt, M.A., Kohl, H.W., Roberts, S., Cook, K.F., Mathis, K.B., 2002. What functional activities are important to patients with knee replacements? Clin. Orthop. Relat. Res. 404, 172–188. Wilson, S.A., McCann, P.D., Gotlin, R.S., Ramakrishnan, H.K., Wootten, M.E., Insall, J.N., 1996. Comprehensive gait analysis in posterior-stabilized knee arthroplasty. J. Arthroplast. 11 (4), 359–367. Zeni Jr, J.A., Snyder-Mackler, L., 2010a. Most patients gain weight in the 2 years after total knee arthroplasty: comparison to a healthy control group. Osteoarthr. Cartil. 18 (4), 510–514. Zeni Jr, J.A., Snyder-Mackler, L., 2010b. Preoperative predictors of persistent impairments during stair ascent and descent after total knee arthroplasty. J. Bone Joint Surg. (Am.) 92-A (5), 1130–1136.
