Downhill Walking Discriminates Between Types of Knee Arthroplasty: Normal Stride Length

Downhill walking discriminates between types of knee arthroplasty: normal stride length requires an intact ACL.

ABSTRACT

PurposeTotal knee arthroplasty (TKA) advocates report no functional advantage of unicompartmental knee arthroplasty (UKA) despite it being an anatomically sparing procedure. ACL insufficiency is known to cause difficulty when walking downhill. The aim of the study was to determine if walking downhill would be a useful way of reporting any functional deficit following differing kinds of knee arthroplasty.

Methods19 UKA and 14 TKA patients with high Oxford knee scores (OKS) were evaluated at a minimum 1 year after their operation with downhill gait analysis. 19 healthy young subjects were used as controls.Downhill gait analysis was done on an instrumented treadmill that was ramped at the rear to produce a declination. All subjects after a period of habituation were tested for preferred and top downhill walking speed with associated ground reaction and temporospatial measurements.

ResultsThe UKA group which were well matched demographically with the TKA group, however had higher mean OKS (44.8 vs 41.9 p=0.03). UKA group had a faster mean downhill top walking speed than TKA group (1.75m/sec vs 1.52m/sec p=0.000) despite having the same cadence (135 vs 134 steps/min). This 12% difference in speed appeared largely due to increase in stride length (173 vs 150 cm p=0.000) and normal weight acceptance which were similar to the controls.

ConclusionUsing an instrumented treadmill to test a commonly performed task, downhill walking,is a useful way of reporting functional differences between arthroplasty groups. The UKA group was able to walk 12% faster downhill than the TKA group with a similar mean cadence.

Level of EvidenceLevel III, retrospective comparative study.

Introduction:

In the last decade unicompartmental knee arthroplasty (UKA) has gained increasing popularity due to advantageous surgical and medical factors including; minimal invasive surgery, anatomical and biomechanicalpreservation, ease of revision, less blood transfusions, less DVTs, decreased length of stay and 30 day mortality[1-9]. However proponents of total knee arthroplasty (TKA) cite survivorship in UKA is less satisfactory and functional outcomes are comparative when comparing pre and post function results[10].

Instrumented treadmills have successfully demonstrated functional differences between various arthroplasty procedures and approaches as it has the advantage of allowing subjects to be tested at a range of speeds and inclinations[11, 12]. However there is a concern cardiovascular fitness would create a potential selection bias when testing different type of patients on treadmills as seen in different studies[13]. As walking may be the single most important activity of daily living and ACL insufficiency causing difficulty in this activity when walking downhill, we surmised that using a treadmill to walk downhill would report functional advantages which would not be traditionally reportable. Downhill walking would help minimise the selection bias of cardiovascular fitness as it is an eccentric and gravity dependent exercise with documented less oxygen consumption[14] and more importantly depends on a stable construct of the knee implant.

Participants and Methods

Participants

Ethical approval was sought and gained prior to commencement of the study. 52 subjects were tested on the instrumented treadmill, 3 groups (UKA, TKA, and young healthy control) of 19, 14, and 19 respectively. The two high performing arthroplasty groups were recruited from a database of patient related outcome measures (PROMs) and were chosen based on high Oxford knee scores (OKS) with a minimum 12 months post hip arthroplasty. They had unilateral or bilateral knee disease and werecardiovascularly fit, with no further joint or lower limb disease. Patients with non-knee arthroplasty, neurological, medical or any other lower limb disease that might affect the top walking performance and speed were excluded. All subjects were recruited by a research assistant who kept the treadmill assessor blinded to prevent any testing bias. A young healthy control group as well was recruited to ensure normal knee with an intact ACL. They consisted of hospital and university members, who were cardiovascularly fit with no neurological or lower limb/joint disease and had active lifestyles.

Surgical Intervention and Rehabilitation

All 33 arthroplasty subjects were operated through a subvastus approach by senior surgical authors who had extensive experience in knee arthroplasty. The UKA group group had cemented or uncementedOxford implant type inserted by one author whoprimarily undertakes this procedure, while the total knee replacements were predominantly inserted by the other senior author who primarily performs TKA. The TKA group had cemented Smith & Nephew Genesis Knee System. All subjects had undergone standard rehabilitation programmes with full weight bearing allowed from day one postoperatively. There were no documented postoperative complications and all subjects had been discharged from routine follow up.

Patient Related Outcome Measures and Anthropometric Measures

Before gait performance was analyzed, OKS, UCLAand EuroQol 5 part questionnaire (EQ-5D) with the EuroQol visual analogue scale (EQ-VAS) scores were obtained[15-18]. After the psychometric score assessment, the subject’s height, weight and leg length were recorded; each was taken at least twice to minimize error. The leg length was measured by taking the distance between the anterior superior iliac spine (ASIS) and the medial malleolus on the ipsilateral side and was taken for both sides[19].

Gait Analysis

Gait performance was tested on a validated instrumented treadmill (KistlerGaitway®, Kistler Instrument Corporation, Amherst NY)[20, 21]. The treadmill was manufactured with tandempiezo-electric force plates beneath the treadmill belt, which measured vertical ground reaction forces (GRF) and associated temporo-spatial variables. The monitoring software included a patented algorithm which distinguished left and right foot strikes. The treadmill had a speed range of 0.1 to 22.0 km/hr and adjustable in 0.1km/hr increments. The rear of the treadmill was ramped with 30 cm axle stands in order to create a 7 degree decline (-12%) for downhill walking (figure 1)

Before treadmill testing began, all participants had to give informed consent to the researcher who was blinded to the type of kneearthroplasty. After a 6 minute acclimatization period as suggested by Matsas[22], the patients downhill preferred walking speed (PWS) was assessed by either increasing or decreasing at 0.1km/hr increments. After the downhill PWS was attained, the speed was increased incrementally until either the subject felt uncomfortable, or downhill top down walking speed (TWS) performance had deteriorated. Downhill top walking speed was documented as the fastest the subject could walk without running. All walking measurements were collected without the aid of handrails or other props. The procedure generally took 12 minutes of continuous walking and was completed without difficulty by all subjects. At all incremental intervals of speed, the vertical component of the ground reaction forces, center of pressure (COP) and temporal measurements were collected for both limbs with a sampling frequency of 100Hz over 10 seconds. Hof scaling and body weight normalising was also applied to the outputted mechanical data to correct for leg length and mass differences, respectively[23]. All treadmill outputted mechanical data was subject to averaging by custom written MATLAB software script to handle the large amount of continuous data being exported, as a 10 second interval normally entailed a minimum of 6 steps for each limb. The data was further broken into implant and non-implanted limbs for the arthroplasty groups and dominant and non-dominant limbs for the healthy control group. Figure 2 is given to demonstrate a graphical depiction of the variables collected

Figure 1 showing a subject walking downhill on the instrumented treadmill.

Figure 2

Statistical Analysis

Power calculation was done prior to the study based on a previous study on flat top walking speed, with power set at 0.90 and α =0.05 with a standard deviation of 0.19 m/sec and mean difference of 0.22m/sec, a sample size of 34 was anticipated.

Statistical analysis was performed with SPSS (IBM SPSS Statistics, version 21).The variables for each of the subject group were compared to each other using an one-way analysis of variance (ANOVA) with Tukey post hoc test. For continuous variables between the arthroplasty groups an independent t-test was used and for categorical variable (gender), a chi squared test was used. A significance level of α=0.05 was used throughout.

Results

Subject Demographics

Patient demographic data were reasonably matched for the UKA and TKA groups. The gender, age, body mass index (BMI) and height were similar in each group (Table 1). The follow-up periods of gait analysis for the arthroplasty groups were also similar. The healthy control group consisted of 13 males and 6females; their BMI, height and age were significantly different than the arthroplasty groups.

.

Table 1: Subject Demographics

Subject / Control / UKA / TKA
Sex M:F / 13:6† / 9:10 / 6:8
Age (yrs) / 28.1± 10.6† / 64.8 ± 5.0 / 67.5 ± 9.2
BMI / 24.2 ± 3.2† / 29.3 ± 4.1 / 29.1 ± 4.5
Leg Length (cm) / 93.2 ± 5.1 / 89.9 ± 3.6 / 90.5 ± 6.4
Height (cm) / 175.1 ±8.2† / 168.6 ± 6.5 / 166.2 ± 11.2
Follow-up (months) / NA / 25.8 ± 13.1 / 27.3 ± 14.7

The values are indicated as means ± standard deviation; †significant difference between patient groups versus control (p <0.05); ‡significant difference between patient groups (p <0.05).

Patient Related Outcome Measures

Both arthroplasty groups demonstrated predictably good marks however the OKS (p=.03) and EQ-5D (p=.02)was significantly higher in the UKA group. The UCLA and EQ-VAS results were not significantly different in thearthroplasty groups.

Table 2: Patient Reported Outcome Measures

PROM / UKA / TKA / p
Oxford / 44.8 ± 2.9 / 41.9 ± 4.7 / .03
UCLA / 7.6 ± 1.3 / 7.0 ± 1.4 / .21
EQ-5D / 0.93 ± 0.10 / 0.82 ± 0.13 / .02
EQ-VAS / 84.9 ± 14.1 / 77.1 ± 15.4 / .14

Treadmill Gait Analysis

At a slow downhill preferred walking speed, all three groups (UKA, TKA, Control) demonstrated similar temporal-spatial and ground reaction force results (Table 3,4 and Figure 3-5). Only variable of significance at PWS was the widened gait seen in both UKA and TKAgroups (p<0.003) when compared to the healthy controls. Only at downhill TWS the differences became more apparent (Figure 4). The UKA patients walked significantly faster than the TKA patients (1.75 vs 1.52 m/sec p=0.000) (Table 3) despite having the same cadence (134 vs 135step/min). The 12% difference in speed was largely due to the reduced mean step (p=0.001) and stride lengths (p=0.000) seen in the TKA patients. These finding persisted after Hof scaling, which normalised speed for leg length (p=0.000). Both arthroplasty groups when compared to the control group were significantly (p<0.05) inferior with regards to speed, step length and gait width but this was more considerable in the TKA group.

The vertical ground reaction force data (Table 4 and Figures 6-8) of the UKA patients were nearer to normal than were the TKA patients at downhill TWS; however they both had significantly less mean weight acceptance force when compared to the young healthy controls. Both arthroplasty groups appeared to load symmetrically when compared to loading to the contralateral limbs (Figure 7,8). The UKA however had a noticeable reduction in pushoffforce in the implanted limb when compared to the contralateral limb (figure 8), this was not seen in the TKA group.

Table 3: Temporal-Spatial Results at PWSand TWS

Variable / PWS / TWS
Control / UKA / TKA / Control / UKA / TKA
(n=19) / (n=19) / (n=14) / (n=19) / (n=19) / (n=14)
Mean / SD / Mean / SD / Mean / SD / Mean / SD / Mean / SD / Mean / SD
Speed (m/s) / 1.32 / 0.08 / 1.28 / 0.07 / 1.28 / 0.08 / 1.91 / 0.07 / 1.75† / 0.14 / 1.52†‡ / 0.13
Hof Speed (H) / 0.44 / 0.03 / 0.43 / 0.02 / 0.43 / 0.02 / 0.63 / 0.03 / 0.59† / 0.05 / 0.51†‡ / 0.04
Cadence (step/min) / 118.6 / 5.8 / 122.7 / 6.2 / 127.5† / 11.3 / 132.7 / 4.8 / 134.9 / 8.0 / 133.9 / 9.6
Step Length (cm) / 71.2 / 6.1 / 69.5 / 4.1 / 66.8 / 6.5 / 92.2 / 6.7 / 85.6† / 7.4 / 75.2†‡ / 8.3
Stride Length (cm) / 142.6 / 12.2 / 138.6 / 8.0 / 133.7 / 12.9 / 184.6 / 13.3 / 173.2 / 14.4 / 150.2†‡ / 16.6
Contact Time (s) / 0.69 / 0.05 / 0.66 / 0.04 / 0.65 / 0.08 / 0.59 / 0.04 / 0.58 / 0.05 / 0.60 / 0.06
Step Time (s) / 0.51 / 0.03 / 0.49 / 0.03 / 0.48 / 0.05 / 0.46 / 0.03 / 0.44 / 0.02 / 0.45 / 0.04
SingleLimbStance (s) / 0.36 / 0.02 / 0.32 / 0.05 / 0.33 / 0.04 / 0.33 / 0.02 / 0.31 / 0.04 / 0.33 / 0.05
Gait Width (cm) / 11.0 / 2.7 / 13.4† / 2.9 / 14.0† / 2.5 / 10.6 / 2.0 / 13.2† / 2.5 / 13.6† / 2.3

The values are indicated as means ± standard deviation; †significant difference between implant versus control (p <0.05); ‡significant difference between implant groups (p <0.05);H=normalized to leg length; BW=normalized to body weight.

Table 4: Vertical Ground Reaction Forces at PWS

Variable / PWS / TWS
Control / UKA / TKA / Control / UKA / TKA
(n=19) / (n=19) / (n=14) / (n=19) / (n=19) / (n=14)
Mean / SD / Mean / SD / Mean / SD / Mean / SD / Mean / SD / Mean / SD
Weight Acceptance (BW) / 1.40 / 0.11 / 1.37 / 0.11 / 1.37 / 0.17 / 1.70 / 0.11 / 1.63† / 0.15 / 1.47†‡ / 0.16
Mid Stance (BW) / 0.72 / 0.11 / 0.77 / 0.09 / 0.77 / 0.10 / 0.54 / 0.10 / 0.72† / 0.14 / 0.76† / 0.12
Push-Off (BW) / 0.91 / 0.09 / 0.90 / 0.05 / 0.91 / 0.06 / 0.84 / 0.09 / 0.83 / 0.08 / 0.88 / 0.09
Loading Rate (BW/s) / 14.7 / 3.4 / 16.9 / 3.5 / 18.9† / 5.3 / 27.5 / 6.5 / 25.1 / 6.0 / 25.1 / 5.9
PushOff Rate (BW/s) / 6.7 / 1.9 / 6.2 / 2.0 / 6.1 / 2.0 / 6.6 / 2.4 / 4.9 / 2.2 / 5.5 / 2.1
Impulse (BW/s) / 0.52 / 0.03 / 0.50 / 0.03 / 0.49 / 0.05 / 0.46 / 0.02 / 0.45 / 0.02 / 0.46 / 0.04

The values are indicated as means ± standard deviation; †significant difference between implant versus control (p <0.05); ‡significant difference between implant groups (p <0.05);BW=normalized to body weight.

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Discussion

This small study used an instrumented treadmill to test arthroplasty patients walking downhill to detect significant differences in gait. The principle limitation is lack of randomisation and a small sample size. However, despite lack of randomisation both arthroplasty groups were decently matched and were operated by surgeons with large patient numbers who have a preferenceto a certain operation. The potential selection bias will always exist unless randomisation occurs but we feel the metric did a reasonable job of not relying on cardiovascular function but rather a stable knee construct. Although the study was small, it did confirm our primary hypothesis and our suspicions that downhill walking would differentiate between arthroplasty groups.

Recent studies on downhill walking with the elderly have demonstrated reduced oxygen and energy consumption in excess of 25% when compared to flat walking[14, 24]. Furthermore the study proposes that this form of exercise may benefit a population of people who may have reduced exercise tolerance like in the elderly as it relies on an eccentric activity. These conclusions would reduce the potential cardiovascular selection bias seen in differing patient groups and help objectively determine if there was a functional difference between arthroplasty groups. Furthermore as osteoarthritis is a primarily a degenerative condition caused by high mileage, activity in these patients should not be an issue and this was reflected in the high UCLA scores in both groups which were not significantly different.

This study is the first to report comparing different arthroplasty patients walking downhill at slow and high speeds. The declination was 7 degrees (ie -12%) and did not cause any difficulty for the subjects when tested. The UKA group walked 12% faster than the TKA and this was primarily due to increased step and stride length as the cadence were almost identical. Both arthroplasty groups however were significantly slower than the healthy control at top downhill walking speed and this may be due the significant difference in mean age (66vs28yrs). Both arthroplasty groups exhibited symmetrically loading when compared to the contralateral normal limb demonstrating a stable construct with both implant groups. Some of the ground reaction force parameters were significantly less than the controls, but recent studies comparing joint disease free young and elderly walking downhill exhibited similar results and it was hypothesized to a difference in muscle mass[24].

This is not randomized data, so conclusions need to be made with caution, but comparing the arthroplasty groups with a metric that relies on eccentric muscle activity and less oxygen it appears that UKA group has a functional advantage. We have described downhill temporospatial and ground reaction force data on arthroplastypatients which up to this point has not been described. Both knee groups demonstrate symmetrical loading, but the UKA group closer resembled the healthy young group in most parameters. The study was not designed to determine the cause for the difference in stride length in both group, but intact cruciate ligaments is hypothesized to cause this functional advantage seen. The simple metric of a treadmill in any gym with an inclination and reversing the belt direction could be a means of helping rehabilitate patients particularly when exercise tolerance is an issue.

Acknowlegements: We would like to thank the Wellcome Trust and the EPSRC for funding this project

1.Akizuki, S., et al., In vivo determination of kinematics for subjects having a Zimmer Unicompartmental High Flex Knee System. J Arthroplasty, 2009. 24(6): p. 963-71.

2.Amin, A.K., et al., Unicompartmental or total knee arthroplasty?: Results from a matched study. Clin Orthop Relat Res, 2006. 451: p. 101-6.

3.Jeer, P.J., A.J. Cossey, and G.C. Keene, Haemoglobin levels following unicompartmental knee arthroplasty: influence of transfusion practice and surgical approach. Knee, 2005. 12(5): p. 358-61.

4.Hollinghurst, D., et al., No deterioration of kinematics and cruciate function 10 years after medial unicompartmental arthroplasty. Knee, 2006. 13(6): p. 440-4.

5.Hopper, G.P. and W.J. Leach, Participation in sporting activities following knee replacement: total versus unicompartmental. Knee Surg Sports Traumatol Arthrosc, 2008. 16(10): p. 973-9.

6.Isaac, S.M., et al., Does arthroplasty type influence knee joint proprioception? A longitudinal prospective study comparing total and unicompartmental arthroplasty. Knee, 2007. 14(3): p. 212-7.

7.Saldanha, K.A., et al., Revision of Oxford medial unicompartmental knee arthroplasty to total knee arthroplasty - results of a multicentre study. Knee, 2007. 14(4): p. 275-9.

8.Newman, J., R.V. Pydisetty, and C. Ackroyd, Unicompartmental or total knee replacement: the 15-year results of a prospective randomised controlled trial. J Bone Joint Surg Br, 2009. 91(1): p. 52-7.

9.Becker, R. and J.N. Argenson, Unicondylar knee arthroplasty: what's new? Knee Surg Sports Traumatol Arthrosc, 2013. 21(11): p. 2419-20.

10.Lyons, M.C., et al., Unicompartmental versus total knee arthroplasty database analysis: is there a winner? Clin Orthop Relat Res, 2012. 470(1): p. 84-90.

11.Aqil, A., et al., The gait of patients with one resurfacing and one replacement hip: a single blinded controlled study. Int Orthop, 2013. 37(5): p. 795-801.

12.Wiik, A.V., et al., Unicompartmental knee arthroplasty enables near normal gait at higher speeds, unlike total knee arthroplasty. J Arthroplasty, 2013. 28(9 Suppl): p. 176-8.

13.Mont, M.A., et al., Gait analysis of patients with resurfacing hip arthroplasty compared with hip osteoarthritis and standard total hip arthroplasty. J Arthroplasty, 2007. 22(1): p. 100-8.

14.Gault, M.L., R.E. Clements, and M.E. Willems, Cardiovascular responses during downhill treadmill walking at self-selected intensity in older adults. J Aging Phys Act, 2013. 21(3): p. 335-47.

15.Bellamy, N., et al., Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol, 1988. 15(12): p. 1833-40.

16.Beaule, P.E., et al., The value of patient activity level in the outcome of total hip arthroplasty. J Arthroplasty, 2006. 21(4): p. 547-52.