Knee extensor power predicts six-minute walk test performance in people with transfemoral amputations

Abstract

Introduction:

Lower limb amputees have increased metabolic costs during walking which may be mitigated by maintaining quadriceps strength and power following amputation. However, there are no current studies investigating the relationship between thigh strength and walking performance in individuals with transfemoral amputation.

Objective:

To quantify the relationship between intact limb quadriceps strength in transfemoral amputees and six-minute walk test (6MWT) performance.

Design:

Descriptive laboratory study

Setting:

Laboratory

Participants:

Eleven participants with unilateral transfemoral amputations from trauma or osteosarcoma (4F/7M, 46.21±12.68 years old, 28.24±20.57 years following amputation)

Interventions:

Strength and power testing on the intact limb followed by 6MWT with a flowmeter to measure oxygen uptake (VO₂).

Main Outcome Measures:

Strength included mass-normalized peak torque, average torque, and average power. 6MWT measures included total distance traveled and VO₂ normalized to distance and mass. Significant correlations (P≤.05) were retained for a regression analysis.

Results:

Peak isokinetic knee extensor torque was correlated with total VO₂ (r=−.60, P=.05) and distance traveled (r=.84, P=.001). Average isokinetic knee extensor torque was correlated with total VO₂ (r=−.61, P=.046) and distance traveled (r=.85, P=.001). Average knee extensor power was correlated with total VO₂ (r=−.67, P=.026) and distance traveled (r=.88, P<0.001). Peak isometric knee extensor torque was correlated with distance traveled (r=.69, P=.019). Average power explained 77.2% of the variance in distance traveled during the 6MWT (P<.001) and average power explained 44.2% of the variance in total VO₂ during the 6MWT (P=.026).

Conclusions:

Knee extensor strength was correlated with performance on the 6MWT in individuals with unilateral transfemoral amputation. The strongest relationship was between isokinetic quadriceps power and distance traveled, which suggests that developing quadriceps power in the intact limb following amputation may be an important factor to reduce metabolic cost of walking and support a return to an active lifestyle.

INTRODUCTION

Lower extremity amputation is common with over 150,000 amputations in the United States annually.¹ Although amputation is sometimes required in life-saving or prolonging procedures, the long-term consequences are daunting. Following amputation, many patients report knee and hip pain,² develop osteoarthritis in the intact limb,³ and up to 90% experience severe chronic low back pain.⁴ Chronic joint pain can lead to a sedentary lifestyle and reduced mobility, which is compounded by difficulty walking with a prosthesis. Most individuals demonstrate asymmetrical loading, favoring the amputated side by placing more weight and increased ground reaction forces through the intact limb.⁵ Increased weight on the intact limb leads to increased stance time and power absorption in the intact limb compared to the amputated side.⁶ This requires more eccentric work in the intact limb than in the affected limb. Lack of adequate muscular strength in the lower limb to attenuate these forces places increased stress on the joints, which may be displaced proximally,⁷ and may play a role in reported knee and hip pain in the intact limb.

Lower limb muscle weakness following amputation has been well documented, regardless of time since amputation.^8,9 Increasing quadriceps strength is important after an amputation because it is positively correlated with gait speed;^10,11 and increased gait speed is associated with reduced mortality.¹² Gait speed may also be associated with successful community mobility,¹² which leads to improved quality of life following amputation.¹³ Individuals with transfemoral amputation who resume an active lifestyle are able to maintain strength.^14,15 However, these individuals represent a minority of persons with lower limb transfemoral amputation; most individuals report more barriers than motivators to adopt an active lifestyle.¹⁶

Lower-limb amputees have increased metabolic costs during walking. This may be related to reduced strength after amputation.^9,17 A recent study suggests that increased strength reduces the metabolic cost of walking,¹⁸ likely by improving power generation during gait. Walking has a high metabolic cost in individuals with lower limb amputation,¹⁹ which means that they walk less, walk slower, and tire faster than able-bodied persons.²⁰ For individuals with amputation who are ambulatory, most adopt walking speeds that are 7–42% slower than able-bodied controls²¹ and slower than many stroke survivors or individuals with orthopedic injuries.¹² Although improved prosthetic technology may reduce metabolic cost, available prosthetic feet cannot fully restore plantarflexor power at push-off.²²

Restoring quadriceps strength is a common goal in rehabilitation programs following orthopedic injury, such as anterior cruciate ligament reconstruction or patellofemoral pain syndrome, to improve function and gait after injury, however there is little information regarding the relationship between quadriceps strength and gait following transfemoral amputation. Therefore, the purpose of this study was to quantify the relationship between intact limb quadriceps strength and performance on the six-minute walk test (6MWT). We hypothesized that those with stronger quadriceps would walk farther and have reduced metabolic cost on the 6MWT.

METHODS

This was a descriptive laboratory study that evaluated the relationship between quadriceps strength and power and walking performance in individuals with unilateral transfemoral amputation. Isokinetic and isometric quadriceps strength and power were measured on the intact limb and walking performance was measured using the 6MWT, which included distance walked and metabolic cost (VO₂).

Participants

Eleven participants with unilateral transfemoral amputations (4F/7M, 46.21±12.68 years old, 82.42±20.73kg, 179.19±13.46cm, 25.49±4.94 BMI, 28.24±20.57 years following amputation) volunteered for this study. Eleven participants were enrolled to identify a strong correlation (r=0.75) with 80% power and alpha of 0.05. All participants signed written informed consent approved by the university Institutional Review Board before study participation. Participants did not have any comorbidities that would have affected performance in this study and had transfemoral amputation due to trauma or osteosarcoma. Eight participants used a microprocessor knee and three used a mechanic knee for their prosthesis. Inclusion criteria included able to ambulate without the use of aids (e.g., walking cane) and walking without pain for at least ten minutes.

Procedures

All participants came to the laboratory for a single session. After providing informed consent, participants completed the Locomotor Capabilities Score (LCI),²³ the Orthotics and Prosthetics Users Survey Lower Extremity Functional Scale (LEFS),²⁴ and the Tampa Scale of Kinesiophobia (TSK-17).²⁵ A score ≤ 37 on the TSK-17 was considered low fear of movement.²⁶ Participants then completed isokinetic and isometric strength testing on the intact limb using a dynamometer (Biodex Systems 4). Participants were seated with 85° of hip flexion, with the ankle strapped to the dynamometer. The dynamometer axis was aligned to the lateral femoral epicondyle in line with the center of the sagittal plane knee joint. Participants practiced 3–5 repetitions at maximal output to familiarize with the testing protocol before completing six isokinetic concentric contractions at 120°/second. This angular velocity was chosen because it is the most comfortable for individuals with amputation.⁸ The range of motion was set from approximately 0° to 110°, based on each individual’s range of motion. After a two-minute rest interval, participants were set to 90° of knee flexion for isometric testing. Following at least two practice trials, participants completed three maximal isometric knee extension and flexion contractions for five seconds with 30 seconds of rest in between. Maximal voluntary contraction strength was defined as the peak torque generated across all trials. Verbal feedback and encouragement were provided throughout all tests to ensure maximal performance. Analog data from the isokinetic and isometric tasks were collected at a sampling rate of 125Hz and processed using the Biodex System software. Average torque, peak torque, and power were normalized to body mass. Body mass included the mass of the prosthetic limb.

After strength testing, participants were fitted with a silicone face mask connected to a turbine flowmeter (COSMED, K4b2) to measure breath-by-breath oxygen uptake during the 6MWT. Once fitted with the silicone mask, the participants remained seated for 5 minutes before stepping onto the treadmill. Participants completed the 6MWT on a treadmill, which has previously been shown to be valid and reliable;^27,28 oxygen uptake is similar in treadmill and floor tests for individuals with lower limb amputation.²⁹ Each participant was fitted with a gait belt and a licensed physical therapy assistant behind the participant on the treadmill ensured safety. The treadmill monitor was covered so that the participant was not able to see speed or distance, but had access to the speed buttons to control the treadmill speed.²⁷ Participants were asked to hold onto the front handles of the treadmill if needed for balance. The treadmill was slowly increased to 0.45m/s (1 mile/hr) and following a brief familiarization period (~1 minute), the six-minute walk test (6MWT) began. The familiarization period was short to ensure participants were not fatigued before beginning the 6MWT. Participants were instructed to walk as far as possible in six minutes and could increase or decrease the speed at any time. Participants were also permitted to stop the treadmill at any point if they needed to rest. Verbal encouragement throughout the 6MWT to support maximal effort was provided. Distance walked and treadmill speed were recorded each minute.

Metabolic data were averaged every 60 seconds. Walking economy was calculated as gross oxygen consumption (total VO₂) and average VO₂ during the 6MWT. For Total VO₂, the total oxygen consumed during the 6MWT was summed and divided by body mass and meters walked (mL/min/kg*m). Average VO₂ was calculated as the average oxygen consumed during each minute of the 6MWT and divided by body mass and meters walked (mL/min/kg*m).³⁰

Statistical Analyses

Data were assessed for normality using the Shapiro-Wilk test. Bivariate Pearson correlations were calculated to identify the relationship between strength measures and 6MWT performance. Isokinetic strength measures included mass-normalized peak torque, average torque, and average power. Isometric strength measures included mass-normalized peak torque. Walking measures included total distance traveled during the 6MWT, average VO₂ consumed, and total VO₂ consumed. Both VO₂ measures were normalized to distance traveled and body mass (kgm). BMI was also correlated with all dependent variables. Significant correlations (p≤.05) were retained for a regression analysis.

A stepwise linear regression model was used to identify the proportion of variance explained by the distance traveled (model 1) and the total VO₂ (model 2).

RESULTS

All included participants had high subjective function on the LCI and LEFS and low fear of movement, as identified on the TSK-17 (Table 1). Mean strength and 6MWT measures are reported in Table 1.

Table 1.

Mean and standard deviation for patient demographics, strength, walking performance and subjective function.

Variable	Average	St. Dev.
Age (years)	46.21	12.68
Mass (kg)	82.42	20.73
Height (cm)	179.19	13.46
Body Mass Index	25.49	4.94
Years Following Amputation	28.24	20.57
Peak Isokinetic Knee Extensor Torque (Nm/kg)	1.49	0.39
Average Isokinetic Knee Extensor Torque (Nm/kg)	1.39	0.37
Average Isokinetic Knee Extensor Power (W/kg)	1.93	0.57
Peak Isometric Knee Extensor Torque (Nm/kg)	2.24	0.89
Peak Isokinetic Knee Flexor Torque (Nm/kg)	0.63	0.26
Peak Isometric Knee Flexor Torque (Nm/kg)	0.88	0.35
Walking Speed (m/s)	1.24	0.36
Distance (m)	413.67	122.25
Average VO2 (mL/min/kgm)	0.04	0.12
Total VO2 (mL/min/kgm)	6.51	3.26
Locomotor Capabilities Score (LCI)	55.60	0.97
Lower Extremity Functional Scale (LEFS)	66.90	7.36
Tampa Scale of Kinesiophobia (TSK-17)	30.80	2.86

BMI was not correlated with any variables (P =.21–.95). Mass-normalized peak isokinetic knee flexor torque was only significantly correlated with distance traveled (r=.70, P=.02) and mass-normalized peak isometric knee flexor torque was only correlated with distance traveled (r=.68, P=.02). There were no correlations between knee flexor strength and VO₂ variables (P=.06–.47).

Mass-normalized peak isokinetic knee extensor torque was significantly correlated with total VO₂ (r=−.60, P=.05) and distance traveled (r=.84, P<.01). Mass-normalized average isokinetic knee extensor torque was significantly correlated with total VO₂ (r=−.61, P=.05) and distance traveled (r=.85, P<.01). Mass-normalized average knee extensor power was significantly correlated with total VO₂ (r=−.67, P=.03) and distance traveled (r=.88, P<0.01). Mass-normalized peak isometric knee extensor torque was only significantly correlated with distance traveled (r=.69, P=.02). Average VO₂ was not correlated with any strength variables (P=.08–.84). All correlations for quadriceps strength and power are shown in Figure 1.

Figure 1.

Correlations between strength and body mass index and total VO₂ (mL/min/kgm), average VO₂ (mL/min/kgm), and distance traveled (m) on the 6MWT. Open circles indicate a significant correlation (P < 0.05).

Average knee extensor power had the strongest correlation with both distance and total VO₂ and was significantly correlated with all other strength variables, therefore was the only variable entered into the linear regression models. Average power explained 77.2% of the variance in distance traveled during the 6MWT (P<.01). In the second model, average power explained 44.2% of the variance in total VO₂ during the 6MWT (P=.03).

DISCUSSION

This is the first study to our knowledge to investigate the relationship between intact limb quadriceps strength and walking ability following transfemoral amputation. Similar research with transtibial amputees suggested greater thigh strength in the residual limb was associated with improved walking capacity.³¹ Although isometric quadriceps strength in the intact limb was correlated with distance traveled on the 6MWT, isokinetic knee extensor power had a stronger correlation with both distance and total oxygen consumed, supporting that a measure of isokinetic power is a better predictor of walking economy in those with unilateral transfemoral amputation. Transfemoral amputees have reduced aerobic capacity compared to able-bodied controls³² which may contribute to reduced daily steps following amputation. Following lower limb amputation, individuals walk approximately 3000 steps/day,³³ which is well below the thresholds for daily exercise. Sedentary behavior following amputation can lead to declines in mobility, increased cardiovascular disease, and increased mortality,³⁴ therefore, it is important to identify ways to improve activity in this population.

Quadriceps strength has been associated with mobility³⁵ and development of knee osteoarthritis.³⁶ Osteoarthritis in the intact knee and hip following unilateral transfemoral amputation is particularly prevalent,³ however there is limited information regarding modalities to improve function following lower limb amputation to slow down disease progression. Increasing quadriceps power in this population may modulate increased stress on the knee joint.³⁷ Isokinetic knee extensor power is also a strong determinant of mobility in those with osteoarthritis,³⁸ suggesting that developing and maintaining isokinetic knee extensor power may be an important factor for individuals with lower limb amputation to remain active. Future studies in this area should continue investigating what isokinetic knee extensor power threshold is required to increase mobility and daily steps in those with amputation without increasing long-term consequences.

Quadriceps strength and power are also associated with gait speed,^10,11 which is supported in the current study. There was a strong positive correlation (r=.88) between isokinetic knee extensor power and distance traveled on the 6MWT meaning those with more powerful quadriceps musculature walked faster and farther in six minutes. Powerful quadriceps are a main factor in 6MWT performance, explaining 77.2% of the variance in distance traveled. Individuals with amputation who are motivated to participate in sport and return to an active lifestyle following amputation maintain quadriceps strength,^14,15 and in turn, more powerful quadriceps. Rehabilitation following amputation may benefit from increasing quadriceps power before discharge to increase likelihood of maintaining an active lifestyle.

While isokinetic power was a strong predictor of distance traveled on the 6MWT, it was moderately correlated with total VO₂ and explained 44.2% of the variance. The negative correlation supports the hypothesis that those with stronger quadriceps have better walking economy, as evidenced by consuming less VO₂ to walk a farther distance in six minutes. Metabolic cost of walking, even in young and fit individuals with amputation, is 44–47% greater compared to healthy, able-bodied controls.³⁹ Changes to prosthetic design offer some reduction in metabolic cost, however prosthetic design may also contribute to poor gait; passive prosthetic knees lead to more gait asymmetry than microprocessor-controlled knees,⁴⁰ but gait asymmetries also exist with more sophisticated prostheses.⁴¹ Further, sophisticated protheses can be costly and may not be affordable for many patients. In contrast, isokinetic quadriceps power can be improved through ongoing physical therapy and at-home exercises for a relatively low cost and may be a more cost-effective approach to reduce the metabolic cost of walking.

It is important to note that the individuals with transfemoral amputation in this study demonstrated strong quadriceps muscles on the isometric test. When comparing isometric knee extensor strength to young healthy controls using the same methods,⁴² transfemoral amputees in the current study were as strong as college-aged healthy individuals. In comparison, amputees had significantly decreased isokinetic knee extensor power, which suggests that a measure of isokinetic power may be a more appropriate indicator of quadriceps weakness in this population.

Study Limitations

The 6MWT was completed on a treadmill to minimize hallway turns. Turning during walking can be difficult for individuals with amputation due to increased balance and neuromuscular demands,^43,44 and the treadmill allowed for six minutes of walking without turning interruptions. There were also more men than women with amputation in the current study, however all participants had amputation from trauma or osteosarcoma, which tend to occur more often in men.⁴⁵ Gait mechanics can differ based on sex, which in turn, can affect metabolic efficiency.⁴⁶ Future studies in this area should consider comparing sexes with amputation. We also measured isokinetic knee extensor power using a dynamometer as opposed to measurement of power using a camera-based system to measure knee angular velocity, which may have inflated the power metric.⁴⁷ We reported average power as opposed to peak power, which would reduce the inflated power and used a similar method that has been used with many orthopaedic populations to be able to draw larger comparisons. Future studies evaluating knee extensor power may benefit from the use of a camera-based system to accurately measure angular velocity. Lastly, no participants in the current study had dysvascular amputation. Individuals with dysvascular amputation may have a different relationship between strength and 6MWT performance and future research is needed in this area.

CONCLUSIONS

Although quadriceps strength was correlated with distance traveled, the strongest relationship was between isokinetic quadriceps power and distance traveled as well as total VO₂. The results of this study suggest that developing quadriceps power in the intact limb following amputation may be an important factor to reduce metabolic cost of walking and support a return to an active lifestyle.