Abstract
The purpose of this study was to investigate the characteristic differences between the IP and C-Leg by making a comparative study of energy consumption and walking speeds in trans-femoral amputees. The subjects consisted of four persons with traumatic trans-femoral amputations aged 17 – 33 years who had been using the IP and were active in society. Fourteen able-bodied persons served as controls. First the energy consumption at walking speeds of 30, 50, 70, and 90 m/min was measured when using the IP. Then the knee joint was switched to the C-Leg. The same energy consumption measurement was taken once the subjects were accustomed to using the C-Leg. The most metabolically efficient walking speed was also determined. At a walking speed of 30 m/min using the IP and C-Leg, the oxygen rate (ml/kg/min) was, on average, 42.5% and 33.3% higher (P<0.05) than for the able-bodied group. At 50 m/min, the equivalent figures were 56.6% and 49.5% (P<0.05), while at 70 m/min the figures were 57.8% and 51.2% (P<0.05), and at 90 m/min the figures were 61.9% and 55.2% (P<0.05%). Comparing the oxygen rates for the subjects using the IP and C-Leg at walking speeds of 30 m/min and 90 m/min it was found that subjects who used C-Leg walked somewhat more efficiently than those who used IP. However, there was no significant difference between the two types at each walking speed. It was also determined that the most energy-efficient walking speed for subjects using the IP and C-Leg was the same as for the controls. Although the subjects in this study walked with comparable speed and efficiency whether they used the IP or C-Leg, the subjects' energy consumption while walking with the IP and C-Leg at normal speeds were much lower than previously reported. This study suggested that the microprocessor controlled knee joints appeared to be valid alternative for improving walking performance of trans-femoral amputees.
Introduction
It goes without saying that the knee joint is extremely important for walking (Waters et al. 1978). Trans-femoral amputees, who have lost a knee joint, must rely on a mechanical knee joint to provide its functions. Trans-femoral amputees using conventional knee joints walk more slowly than able-bodied people and expend more energy to do so (Gonzalez et al. 1974; Traugh et al. 1975; Waters et al. 1976) because of the limitations of the mechanical performance of such joints. More recently, the application of high technology has succeeded in mounting a micro-computer in the knee joint, making it theoretically possible for trans-femoral amputees to walk comfortably at their own pace. The Intelligent Knee Prosthesis (IP) has electronic control of the swing phase, and the C-Leg is able to electronically control both the swing phase and the stance phase.
A number of researchers have compared the IP and conventional knee joints (Buckley et al. 1997; Taylor et al. 1996). The use of an IP has been reported to reduce walking energy consumption for trans-femoral amputees. Chin et al. (2003) have reported that young IP users who have received appropriate training could walk with only about a 24% increase in energy expenditure at normal walking speeds compared to able-bodied people. Furthermore, Datta & Howitt (1998) tested subjects, who were already experienced in the use of Pneumatic Swing Phase Control knee joint (PSPC), when walking with an IP in order to investigate the subjective evaluation of users. This research found that all the amputees evaluated the IP as superior, and wished to switch to it even though they were accustomed to walking with the PSPC. Thus the IP is clearly demonstrated to be superior to conventional knee joints.
The performance of the C-Leg is expected to exceed the IP in the improvement it gives to the amputees' walking ability. However, the C-Leg is more recent development and, to the authors' knowledge, there has been very little reported on its clinical validity (Schmalz et al. 1999).
In this research the authors had the opportunity to test the C-Leg on trans-femoral amputees who were already accustomed to the IP and were socially active. The aim of this research was to examine the impact the characteristic differences between the IP and C-Leg have on energy expenditure during walking and the walking speed.
Subjects
The four test subjects were young, active, had unilateral trans-femoral amputations, had undergone prosthetic walking training using an IP, and were accustomed to using it and active in society (Chin et al. 2003). They were all male, aged 24.0±7.6 years, weighing 56.5±8.6 kg. The cause of amputation was trauma in all cases. Table I shows the characteristics of the subjects.
Characteristics of amputees.
The sockets used were quadrilateral suction sockets in all cases, while the feet were Sure Flex III in three cases and C-Walk in one case. The fitting and alignment of the IP for all subjects was carried out by a certified prosthetist. The programming of the IP for all subjects was done by a physiotherapist according to the manufacturer's instructions. The change from the IP to C-Leg, alignment and fitting, programming of the C-Leg to suit each user were carried out by a prosthetist licensed by Otto Bock. After changing from the IP to C-Leg, the subjects were allowed to practice walking to familiarize themselves with it. A group of able-bodied people served as the control group which was composed of 14 healthy people (10 male, 4 female) with characteristics similar to the amputees; the age range was 25.2±4.0 years and the weight range was 62.5±15.3 kg. The subjects were well informed of the purpose of this study and possible risks, and an agreed consent was obtained from each subject.
Method
Knee joints used
IP (NABTESCO, Japan)
This is a single axis knee joint using a pneumatic cylinder, with only the swing phase controlled electronically. The degree of cylinder valve opening is automatically varied by a microcomputer in response to the amputee's walking speed, thereby varying the reactive force of the cylinder to control the optimum lower leg swing.
C-leg (Otto Bock, Germany)
This is a single axis knee joint using a hydraulic cylinder, with swing and stance phases controlled by an electro-mechanical system. The built-in sensors can detect the floor reaction force moment and angle of knee joint in 1/50 of a second. To provide optimum control, the microcomputer can calculate the hydraulic resistance for the swing and stance phases, which is based on data from the sensors.
Prosthetic walking training programme using the IP
The period required for the programme, from starting to fit a prosthesis to the completion of rehabilitation is 8 weeks.
The conventional training programme in walking with a prosthesis emphasized acclimatization to its use, bearing an adequate amount of body weight on the prosthesis, and the correction of abnormal walking patterns. When amputees could overcome these problems and achieve a walking speed within a certain range that suited them, their training was completed. Therefore, to walk faster they had to increase the length of stride taken with the sound leg, but that approach cannot be expected to yield a great speed increase.
Under the prosthetic walking training programme the authors' proposed, the kind of walking training provided by a conventional programme is augmented by training aimed at increasing the walking speed. In order to do this, it is necessary to increase both stride and cadence, in an appropriate balance. At the authors' centre, guidelines for stride and cadence for each walking speed were set: at walking speeds of 50 – 70 m/min, the stride should be approximately 60 cm and the cadence should be 80 – 100 steps/min; at walking speeds of 70 – 90 m/min, the stride should be approximately 70 cm and the cadence should be 100 – 120 steps/min; and at walking speeds of 90 – 110 m/min, the strides should be approximately 80 cm with the cadence of at least 120 steps/min. To make strides more even, guide marks for each speed were made on the walking path so that the amputees could always check their paces against them. To achieve uniform cadence, amputees carried portable metronomes and adjusted their pace to fit the constant rhythm. In this way the amputees' walking speeds were gradually raised and they continued their training until they were able to walk continuously for at least five minutes at each speed. During the course of training the programming of the IP was adjusted if the swing phase did not keep up with the walking speed.
Measurement of energy consumption during ambulation
Firstly the energy consumption of the subjects was measured as they walked with the IP. Then the knee joint was changed from the IP to the C-Leg. The same energy consumption measurements were taken once the subjects had become accustomed to using the C-Leg. None of the subjects used walking aids while walking.
The subjects first sat for 15 min before testing, and then they were instructed to walk at speeds of 30, 50, 70 and 90 m/min. Each test walk lasted 5 min with 15 min of rest between each test. Each subject walked around a rectangular track that had a perimeter of 100 m. During the test walks, measurements of the subject's respiratory gas were carried out by means of a portable telemetric system (K4 system; COSMED, Italy) (Figure 1). This system measured the following cardiorespiratory parameters: oxygen uptake (V
˙O2), carbon dioxide production, heart rate, and minute ventilation.
Measurement of energy consumption. Continuous measurement of the subject's respiratory gas during walking was carried out by means of a portable telemetric system (K4 system). Physiotherapists walked along with the amputees to check whether the target walking speed was being reached.
The mean value of the last 2 min of each test walk was taken as the measurement value at that speed. Measurement of energy consumption of the control group was done in the same fashion as the amputees. A walking meter with a speed-meter function (manufactured by Minato, Japan) was used to maintain the specified walking speeds, and the subject walked alongside a therapist who guided their walking speed.
Statistical analysis
The ages, body weights and oxygen rates of the able-bodied group and the subjects were compared using a non-paired t-test. The comparison of oxygen rates when using the IP and C-Leg were examined using a paired t-test. The comparison was judged statistically significant with a significance level of less than 5%.
Results
No significant difference was observed between the subject group and the able-bodied group in age and body weight. Table II shows oxygen cost values (milliliters of oxygen consumed per kilogram of body weight per meter walked, ml/kg/m) for the control group and the subjects using the IP and C-Leg at each walking speed. Table III shows oxygen rate values (milliliters of oxygen consumed per kilogram of body weight per minute, ml/kg/min) for the control group and the subjects using the IP and C-Leg at each walking speed.
Oxygen cost values for the control group and when using IP and C-Leg at each walking speed.
Data are mean ± SD.
Oxygen uptake values for the control group and when using IP and C-Leg at each walking speed.
Data are mean ± SD.
∗Comparison between when using IP and C-Leg.
Figure 2 is a graph of average oxygen cost values for the subjects using the IP and C-Leg and the able-bodied group at each walking speed. Judging from this graph, the minimum oxygen cost for the able-bodied group was while walking at 70 m/min. For the subjects using the IP and C-Leg, the minimum oxygen cost was also while walking at 70 m/min. The walking speed which produces the lowest oxygen cost is the most efficient speed, in energy terms.
Average oxygen cost value curves for the control group and when using IP and C-Leg.
At a walking speed of 30 m/min using the IP and C-Leg, energy consumption (ml/kg/min) was, on average 42.5% and 33.3% higher (P<0.05) than that for the able-bodied group. At 50 m/min, the equivalent figures were 56.6% and 49.5% (P<0.05), while at 70 m/min the figures were 57.8% and 51.2% (P<0.05), and at 90 m/min the figures were 61.9% and 55.2% (P<0.05%).
Comparing the oxygen rates for the subjects using the IP and C-Leg at each walking speed, when the subjects walked at slower (30 m/min) and faster (90 m/min) speeds, subjects who used the C-Leg walked somewhat more efficiently than those who used the IP. However, there was no significant difference between the two types at each walking speed.
Discussion
For trans-femoral amputees using conventional knee joints, walking speed is slow and walking energy consumption is high. Gonzalez et al. (1974) and Traugh et al. (1975) have measured walking energy consumption when walking with older types of trans-femoral prosthesis. According to their reports, even at around 50% of the comfortable walking speed of able-bodied people, trans-femoral amputees used 65% more energy in walking. Waters et al. (1976) also measured walking energy consumption when using trans-femoral prostheses. According to their report, traumatic trans-femoral amputees used 56% more energy to walk at approximately 60% of the able-bodied comfortable walking speed. McClenaghan et al. (1989) conducted a metabolic study of gait in trans-femoral amputees. They reported that amputees had a 52% higher energy consumption compared to a control group during free-paced walking speed (approximately 66% of the free-paced walking speed of control group). Cammisa et al. (1990) reported that the net energy cost for trans-femoral amputees was 75% greater than that for the normal subjects at 88% of the free walking speed of normal subjects. Therefore, an important issue in prosthetic rehabilitation for trans-femoral amputees is to find ways to raise their walking speed and reduce their walking energy consumption.
In recent years the use of microprocessor controlled knee joints such as the IP and C-Leg, has made it theoretically possible for trans-femoral amputees to walk at a wide range of speeds. The results of this study have shown that the most energy-efficient walking speed for subjects using the IP and C-Leg was 70 m/min. This walking speed is close to the comfortable walking speed for an able-bodied person, which is put at 70 ∼ 80 m/min (Waters et al. 1978; Gonzalez et al. 1974; Chin et al. 2003; Gailey et al. 1994). Furthermore, subjects were able to walk at a wide range of speeds, from 30 – 90 m/min, using either type of knee joint. These facts suggest that the use of either the IP or C-Leg in addition to appropriate prosthetic training may be one of the solutions to raise the walking speed of trans-femoral amputees.
There are various possible methods for reducing walking energy consumption, but the possibility which appears most likely is to reduce the weight of the prosthesis. However, Czerniecki et al. (1994) report that their research on eight young trans-femoral amputees found that prosthesis weight had no influence on energy consumption. The next most likely possibility is to build the amputee's physical fitness to make a relative reduction in the energy consumption. Chin et al. (2001) have reported that an endurance training programme based on heart rate at the anaerobic threshold can improve the physical fitness of trans-femoral amputees. However, there have been very few reports on physical fitness evaluation methods for amputees, or on the effects of whole body endurance training, and those which exist are limited (Crouse et al. 1990; Kurdibaylo 1991; Chin et al. 1997; Chin et al. 2002). The last possibility is to improve the ambulatory performance of the amputees by using the microprocessor controlled knee joints with appropriate prosthetic training.
There have been scattered reports in recent years of the use of IP to reduce the walking energy consumption of trans-femoral amputees (Buckley et al. 1997; Taylor et al. 1996; Chin et al. 2003). The IP only improves the swing phase dynamic of the gait. Control of the stance phase requires either some mechanical means, such as a weight activated locking knee, or voluntary control by the user. Voluntary control requires muscle work to extend the hip joint on the amputated side. The C-Leg uses electronic control for both the swing phase and the stance phase, avoiding the need for voluntary control. The reduced muscle work means that the C-Leg is expected to provide a corresponding reduction in energy consumption, relative to the IP. The results of this research showed that when the amputees walked at slower (30 m/min) and faster (90m/min) speeds, amputees who used the C-Leg walked somewhat more efficiently than those who used the IP. However, there were no significant differences between the IP and C-Leg. The authors would like to suggest the following points regarding the reason for this. Firstly, gait evaluation of test subjects was conducted in a research setting and may not have necessarily reflected their chosen way of walking in the day-to-day situation. Therefore, the results in this study may have been affected by the lack of walking evaluation under conditions such as steps, slopes and uneven ground, which may have made greater use of the stance phase control function of the C-Leg. Secondly, the improvement in gait efficiency close to that of the control group may have been influenced by the training protocols that were proposed. Further study is required to evaluate walking ability under conditions which exploit the full functions of the IP and C-Leg.
Although the subjects in this study walked with comparable speed and efficiency whether they had used the IP or C-Leg, the subjects' energy consumption while walking with IP and C-Leg at speeds of 70 m/min and 90 m/min were 57.8%, 51.2% and 61.9%, 55.2% greater, respectively, than for an able-bodied person. These energy consumption increase rates are much lower than previously reported (Gonzalez et al. 1974; Traugh et al. 1975; Waters et al. 1976; McClenaghan et al. 1989; Cammisa et al. 1990), when walking at the same speeds as able-bodied people is considered. From the above results, it is reasonable to state that both the IP and C-Leg enable walking at a wide range of speeds, and the walking energy consumption is lower than previously reported with conventional knee joints. It is not possible to draw firm conclusions from the results obtained with only the test subjects using microprocessor controlled knee joints, since similar training was not provided to amputees using non-microprocessor controlled knee joints. However, this research demonstrated that young, fit amputees can be expected to exploit the high level of function of microprocessor controlled knee joints with appropriate training protocols and can improve in gait efficiency.
Conclusion
The subjects in this study walked with comparable speed and efficiency whether they used the IP or C-Leg and there were no significant differences between the IP and C-Leg. The results may have been influenced by the research setting and training protocols provided. Further detailed studies of gait in amputees using the IP and C-Leg are necessary including comparisons to amputees using non-microprocessor controlled knee joints.