Energy cost, exercise intensity and gait efficiency of standard versus rocker-bottom axillary crutch walking.

D Nielsen, PhD, PT, is Associate Professor, Graduate Program in Physical Therapy, College of Medicine, The University of Iowa, 2600 Steindler Bldg, Iowa City, IA 52242 (USA). Address all correspondence to Dr Nielsen.

C Wadsworth, MS, PT, is Lecturer, Graduate Program in Physical Therapy, College of Medicine, The University of Iowa.

Ms Harris, Ms Minton, Ms Motley, and Ms Rowley were students in the Master of Physical Therapy Program, College of Medicine, The University of Iowa, when this study was completed in partial fulfillment of their degree requirements.

This study was approved by the Human Subjects Review Committee, College of Medicine, The University of Iowa.

This article was submitted August 29, 1989, and was accepted March 15, 1990.

To assess physiological variation during ambulation with assistive devices (eg, canes, walkers, crutches), energy cost, heart rate (HR), relative exercise intensity, and gait efficiency have commonly been used.[1-10] Blessey et al[11] determined values for these variables during unassisted ambulation. These values provide the basis of comparison for gait studies involving the use of assistive devices. Walking speed and step- or stride-length measurements are considered good clinical biomechanical indexes of overall walking performance.[12] Research indicates that energy cost and relative exercise intensity increase in a curvilinear manner with increases in walking velocity.[13-16] Step length and stride length also usually increase systematically with increases in walking speed.[12] Research on gait efficiency of normal and pathologic gaits has shown that individuals usually self-select the most efficient walking speed.[14-17] In this context, measurement of the energy cost per meter traveled (in milliliters of oxygen per kilogram-meter) has been used as a criterion measure of gait efficiency. Accordingly, the graphic plot of energy cost per meter traveled versus walking speed usually results in a minimum value at an individual's self-selected walking velocity (S-SWV). From an energy-conservation point of view, the S-SWV at this minimum value would be the most efficient walking speed. Devices designed to assist ambulation should keep energy expenditure and its associated exercise intensity to a minimum while still permitting as near normal walking speeds as possible.[12,15] In this context, speed has been considered a good clinical index of general walking ability.[12]

Several types of axillary crutches are available to patients. The standard single-tip axillary crutch is currently widely used in clinics. Recently, an aluminum rocker-bottom-type axillary crutch with two uprights and a curved base has been introduced. Sure-Gait[registered trademark] axillary crutch*). The rocker-bottom crutch design, however, is not entirely new. In 1917, Joll[18] reported on GE Healing's invention of a roller-bottom crutch made of wood, which purportedly allowed for a faster ambulation pace compared with the ordinary axillary crutch of that time. Although crutch designs throughout history have changed in an attempt to increase comfort, safety, and ease of use,[19,20] few data have been collected that objectively indicate the actual effectiveness of these design alterations. (* Lumex Inc, 100 Spence St, Bay Shore, NY 11706.)

Limited information is available comparing the rocker-bottom-type crutch with the standard single-tip crutch. Gillespie et al[21] assessed physiological differences between a wooden rolling-bottom crutch and a standard single-tip crutch. Significant between-crutch variation and walking speed, however, confounded their experimental design, invalidating the final energy cost comparison. A recent report by Annesley et al[1] is the only published study to date that specifically compares the Sure-Gait[8] crutch with the standard single-tip axillary crutch. Only 10 subjects participated in that study, however, and testing was performed during ambulation on a motor-driven treadmill at one arbitrarily specified walking speed. The results indicated no significant between-crutch differences in oxygen uptake ([VO.sub.2]) and HR. The authors recommended additional research with a larger subject sample and overground walking before any definitive conclusions could be made.

The purpose of our study was to investigate differences in ambulation with the standard axillary crutch versus the rocker-bottom crutch (Sure-Gait[registered trademark] crutch). The specific objectives were to 1) test for differences in S-SWV and stride length; 2) test for differences in energy cost, gait efficiency, and relative exercise intensity; and 3) evaluate subjective preferences for crutch type in ambulation during stair climbing and ramp walking. Because of the limited amount of and somewhat conflicting information available on this topic, no experimental hypotheses were formulated.

Method

Research Design

We used a 2-X-2, two-factor (crutch type and walking speed), repeated-measures research design in this study.[22] Four testing orders were used to randomize crutch type and walking speed.

Subjects

The subjects in this study were 24 healthy female volunteers with a mean age of 23.8 years (SD = 2.2) and a mean weight of 62.5 kg (SD = 6.4). Participants were sedentary to moderately active. None were highly trained athletes. Written informed consent was obtained from each subject prior to participation in the study. From a clinical perspective, the subjects were considered fairly representative of individuals seen for acute foot, ankle, or leg injuries requiring temporary use of crutches for ambulation. As often is the case in the clinic, none of the subjects had prior experience with crutch walking. For practical reasons, specifically the availability of subject volunteers, only female subjects were used. From a physiological point of view, we would expect the results to be comparable with those of male subjects of similar ages, except that the HR measurements could be slightly higher depending on sex differences in levels of cardio-respiratory fitness.

Procedure

We collected data during two 1-hour test sessions scheduled at least 48 hours apart. Testing was conducted on a 60-m-long, flat, tiled hallway and an adjacent walking ramp and stairwell. Self-selected walking velocity and stride length were determined for each crutch type. Steady-state HR and [VO.sub.2] measurements were obtained for each of four 5-minute walking tests identified according to crutch-type ambulation and S-SWV. Biomechanical (S-SWV and stride length) and physiological (HR and [VO.sub.2]) measurements were not taken during ramp walking or stair climbing.

During the initial test session, all necessary paperwork was completed and the subjects were fitted for crutches. After a demonstration and adequate time to practice walking with each crutch type, S-SWV and stride length were determined for one crutch type. These measurements were followed by HR and 902 measurements during ambulation with both crutch types at the S-SWV of the first crutch type. Ten-minute rest periods were provided between each test trial. The second test session consisted of determining S-SWV and stride length of the second crutch type, followed by HR and [VO.sub.2] measurements for both crutches at the second determined S-SWV. At the conclusion of this test session, the subjects ambulated with each crutch type up and down stairs and a ramp. They then completed a subjective questionnaire concerning crutch preferences.

Crutch fitting was performed according to generally accepted guidelines.[23] Care was taken to standardize crutch height and degree of elbow flexion between the two crutch types to ensure biomechanical equality. Both the standard single-tip and rocker-bottom crutches weighed 1.98 kg per pair.

Each subject was instructed in the performance of a two-point, non-weight-bearing, swing-through gait pattern. This gait pattern was selected for standardization purposes because it eliminated the need to control a partial weight-bearing gait. Further standardization was attained by having the subjects hold up via partial knee flexion) their dominant leg during ambulation. Limb dominance was determined by the leg the subject preferred to use when kicking.

The S-SWV tests were conducted based on a previously established protocol.[17] The subjects were instructed to walk at a comfortable speed that they could maintain for at least 5 minutes. The hallway in which they walked was marked off into four consecutive 10-m segments. During the fifth minute of ambulation, the investigators used stopwatches to measure the subject's walking time for each of the four segments and counted the strides taken over the 40-m walkway. Self-selected walking velocity and stride length were calculated according to the following equations:

  S-SWV = 10 m/mean time for four 10-m
            walking segments              (1)
    Stride length = 40 m/total number
                 of strides               (2)

Various techniques have been used to determine S-SWV. Multiple timed trials (ie, three-five repeated time measurements for back-and-forth walking over a calibrated distance) is probably the most common approach.[24] However, we have found this technique to be less reliable than a 5-minute steady-state test protocol.[17] Test-retest analysis of the data collected on 17 cerebral-palsied children for the 5-minute test protocol produced a nonsignificant 2.4 sec/min between-day difference in S-SWV and a significant Pearson product-moment correlation coefficient of .84. Based on these findings, we adopted the 5-minute steady-state method as our standard protocol for this as well as other gait studies. [15,25]

Walking velocity during HR and [VO.sub.2] tests was controlled with an instrumented speedometer cane,[24] which one of the investigators used while walking in front of the subject. Heart rate was measured with a hand-held, digital pulse-rate monitor* immediately following each walking test. Oxygen uptake was determined through indirect calorimetry by the open-circuit method. Timed samples (40-60 L) of the subject's expired air were collected in plastic Douglas bags during the last minute of each walking test. Subsequent air volume measurements and gas analyses were performed on a semi-automated on-line computer system. The system consisted of a 120-L-capacity Collins gasometer and mixing chamber** and Beckman [O.sub.2] and [CO.sub.2] electronic gas analyzers,*** all connected on-line to a preprogrammed XT-IBM**** compatible laboratory computer.

* 1-2-3 Heart Rate Monitor, Heart Rate Inc, 3188-E Airway Ave, Costa Mesa, CA 92626.

** International Medical Equipment Co, 11950 Riverwood Dr, Burnsville, MN 55378.

*** Beckman Instruments Inc, 3900 River Rd, Schiller Park, IL 60176.

**** International Business Machines Corp, Old Orchard Rd, Armonk, NY 10504.

Oxygen uptake was used as the criterion measure of the energy cost of walking. Gait efficiency was calculated from the ratio of [VO.sub.2] and S-SWV (in meters per minute):

Gait efficiency = [VO.sub.2]/S-SWC =

mL [O.sub.2]/kg[times]m[sup.16] (3)

Percentage of age-predicted maximum heart rate (%APMHR) was used as an index of the relative exercise intensity of walking:

%APMHR = (exercise HR/220 - age)

X 100[sup.26] (4)

The reliability of our [VO.sub.2] and HR measurements has been established previously. In an earlier study,[27] we investigated the changes in [VO.sub.2] and HR under standardized walking conditions (treadmill and continuous overground walking) for a functional range of walking velocities from 26.8 m/min (1.0 mph) to 107.3 m/min (4.0 mph) with 13.4-m/min (0.5 mph) increments. Between-day measurement analysis showed no significant differences in V02 or HR. The range in mean differences was 0.03 to 0.43 mL [O.sub.2][times] [kg.sup.-1] [times] [min.sup.-1] for [VO.sub.2] and 0 to 3 bpm for HR. The standardized walking conditions enhanced the measurement precision but also reduced the total measurement variability, resulting in small within- and between-subject variances. Between-test correlation coefficients were subsequently not computed.

Data Analysis

The statistical analysis was performed with the general linear model of the Statistical Analysis System (SAS) library program at the Weeg Computer Center of The University of Iowa (Iowa City, Iowa). The.05 level was adopted as the level of significance for this study.

Descriptive statistics (means, standard deviations, and within-subject standard errors) were calculated on all variables. Student's paired t tests were used to test for differences in S-SWV and stride length. Oxygen uptake, gait efficiency, and %APMHR were tested by an analysis of variance (ANOVA). A frequency analysis and a Wilcoxon matched-pairs signed-rank test were performed on the questionnaire data.[28]

Results

Figure 3 shows the mean S-SWV and stride-length values. Compared with unassisted ambulation,[11] the S-SWVs for crutch walking were appreciably slower. Visual inspection suggested negligible between-crutch differences, which were verified by nonsignificant t-test values (P [greater than] .05). As shown in Figure 4, [VO.sub.2] and %APMHR were elevated compared with unassisted ambulation.[11] The ANOVA revealed nonsignificant F-test results (P [greater than] .05) for crutch-x-speed interaction and for the main effects for between-crutch differences. Similarly, the crutch-x-speed interaction and between-crutch differences in gait efficiency were also nonsignificant (P [greater than] .05). Analysis of the questionnaire data showed statistically significant (P [less than] .05) preferences for the standard single-tip axillary crutch for stair-climbing, overall safety, and predicted long-term use. The rocker-bottom crutch was preferred for walking up a ramp. There were no preferences for walking down a ramp or for general assessment of efficiency.

Discussion

The research literature suggests that people spontaneously self-select an optimally efficient walking speed, referred to as the free-paced walking velocity or S-SWV.[15] This finding has been observed in individuals with normal gait as well as in selected patient groups with pathological gaits. Pathological gait usually results in shorter step lengths or stride lengths and slower S-SWVs, with the magnitude of the decreases being directly related to the degree of impairment. Self-selected walking velocity has subsequently been considered an acceptable clinical index of general walking ability. In our study, S-SWVs for both crutches were slower than S-SWV for unassisted ambulation reported in healthy women.[11] Ambulation with either crutch type decreased S-SWV by a similar magnitude, approximately 23% as previously reported.[4,5,14] Of particular concern to our experimental questions, S-SWV for the rocker-bottom crutch was not different from S-SWV for the standard axillary crutch, indicating that neither crutch afforded any advantage in allowing faster absolute S-SWVs. Because we found no significant difference in S-SWV, we did not expect to find any appreciable between-crutch differences regarding the other variables investigated in this study.

Energy cost for both types of crutches was 60% greater than that reported for S-SWV during unassisted ambulation.[11] Our subjects' [VO.sub.2] values are similar to those reported by Dounis et al[4] and by Fisher and Patterson[5] for the standard axillary crutch using a two-point, non-weight-bearing, swing-through gait. Direct comparison of our subjects'[VO.sub.2] values with those reported by Annesley et all for Sure-Gait[registered trademark] crutch walking is difficult because of differences in gait pattern (three-point, swing-to vs swing-through), walking speed (1.5 vs 2.2 mph), and mode of walking (treadmill vs overground). From an experimental point of view, however, the results corroborate each other because no between-crutch differences were found in either study.

Because the energy cost of crutch walking was increased, our criterion measure of gait efficiency-energy cost per meter traveled-correspondingly increased. The overall mean value (0.34 mL [O.sub.2]/kg [times] m) is similar to the value reported by Fisher and Patterson.[5] This value is approximately two times greater than for unassisted ambulation,[11] which translates into an approximate 100% reduction in general gait efficiency.

From a clinical perspective, the relative exercise intensity of ambulation with assistive devices is often a primary concern. Based on the degree of walking impairment and the specific cardiorespiratory fitness status of the patient, the exercise stress of walking may be intolerable. In this context, the %APMHR has been used as a criterion measure of relative exercise intensity. Extended walking at values greater than 85% of APMHR is usually intolerable to most patients. Heart rate values during crutch walking may be slightly elevated because of the upper extremity component of exercise. The %APMHR, however, is still an acceptable index of relative exercise intensity. Our observed range of 69% to 72% of APMHR was approximately 45% greater than normal, unassisted ambulation as documented by Blessey et al.[11] Based on the average age of 50 years for the subjects in the study of Annesley et al,[1] the computed 83% of APMHR for treadmill walking exceeded our values, but the reported 47% increase over unassisted ambulation was comparable. As expected, our values better agreed with those reported by Fisher and Patterson[5] and Pagliarulo et al[8] (ie, 70%-75%) for similar crutch-walking conditions. Our values would also appear to be within the exercise tolerance of most healthy individuals. The lack of any significant difference in HR reported by Annesley et all and no difference in %APMHR in our study suggest no difference in relative exercise intensity between standard axillary and rocker-bottom crutch walking.

Studies indicate that stride length mirrors walking speed and is directly related to general gait function.[12] We found no significant difference in stride length between the two crutch types. This finding suggests little bio-mechanical difference in walking performance between the crutches.

Prior to the recent study of Annesley and colleagues,[1] only one study[21] had investigated similar type crutches. Contrary to the results of our study, Gillespie et al[21] found the energy cost for the rolling-bottom crutch to be significantly lower than for the standard axillary crutch. However, subjects ambulated at a significantly slower velocity with the rolling-bottom crutch as opposed to the standard crutch. Consequently, the measured [VO.sub.2] values would be expected to be lower.

The manufacturer's product literature suggests that the rocker-bottom crutch is more comfortable, more convenient, and safer than conventional crutches. Our questionnaire data, however, suggest a preference for the standard single-tip crutch for stair-climbing and for safety and as the overall crutch of choice. A hypothesis is that the standard single-tip crutch may provide a sense of increased stability because of its fixed one-point contact and the cushioning and suctioning effects of the cup-shaped rubber crutch tip. Interestingly, the rocker-bottom crutch was preferred for ascending ramps. An explanation may be that the rocker-bottom crutch's curvature decreases the perception of "vaulting" uphill on the ramped surface. From a fiscal standpoint, there is no appreciable difference in cost between the two types of crutches. In our area, the retail price for the Sure-Gait[registered trademark] crutch and for the standard axillary crutch is $43.50 and $44, respectively.

Clinical Implications and Recommended Research

Based on the results of our study, there appears to be no apparent advantage in terms of energy consumption to using the rocker-bottom crutch over the standard single-tip crutch. Additional research considering various walking conditions, however, may be helpful. Specifically, we suggest similar investigations but testing during ambulation at different walking speeds, with different gait patterns on different grades and walking surfaces. The ability to negotiate different architectural barriers would also be of interest. The inclusion of male subjects and subjects of different ages would be advantageous.

Summary

Within the limitations of this investigation, crutch walking with a two-point, non-weight-bearing, swing-through gait pattern produced a 20% decrease in S-SWV, a 60% increase in energy cost, a 45% increase in relative exercise intensity, and a 100% reduction in gait efficiency compared with values reported for unassisted ambulation.[11] We found no difference in walking performance, however, between the standard single-tip and rocker-bottom axillary crutches. A subjective preference for the standard single-tip crutch was evident for ascending and descending stairs, overall safety, and long-term use.

References

1. Annesley AL, Almada-Norfleet M, Arnall DA, Cornwall MW. Energy expenditure of ambulation using the Sure-Gait[registered trademark] crutch and the standard axillary crutch. Phys Ther. 1990;70:18-23.

2. Bard G, Ralston HJ. Measurement of energy expenditure during ambulation, with special reference to evaluation of assistive devices. Arch Phys Med Rehabil. 1959;40:415-420.

3. Cordrey LJ, Ford AB, Ferrer MT. Energy expenditure in assisted ambulation. J Chronic Dis. 1958;7:228-233.

4. Dounis E, Rose GK, Wilson RSE, et al. A comparison of efficiency of three types of crutches using oxygen consumption. Rheumatol Rehabil. 1980;19:252-255.

5. Fisher SV, Patterson RP. Energy cost of ambulation with crutches. Arch Phys Med Rehabil. 1981;62:250-256.

6. Ganguli S, Bose KS, Datta SR, et al. Biomechanical approach to the functional assessment of the use of crutches for ambulation. Ergonomics. 1974;17:365-374.

7. Hinton CA, Cullen KE. Energy expenditure during ambulation with ortho crutches and axillary crutches. Phys Ther, 1982;62:813-819.

8. Pagliarulo MA, Waters RL, Hislop HJ. Energy cost of walking of below-knee amputees having no vascular disease. Phys Ther 1979;59:538-542.

9. Sankarankutty M, Stallard J, Rose GK. The relative efficiency of "swing-through" gait on axillary, elbow and Canadian crutches compared to normal walking. J Biomed Eng. 1979;1:55-57.

10. Stallard J, Sankarankutty M, Rose GK. A comparison of axillary, elbow and Canadian crutches. Rheumatol Rehabil. 1978;17:237-239.

11. Blessey RL, Hislop HJ, Waters RL, Antonelli D. Metabolic energy cost of unrestrained walking. Phys Ther. 1976;56:1019-1024.

12. Skinner HB, Effeney DJ. Special review: gait analysis in amputees. Am J Phys Med. 1985;64:82-89.

13. Corcoran PJ, Brengelmann GL. Oxygen uptake in normal and handicapped subjects in relation to speed of walking beside a velocity-controlled cart. Arch Phys Med Rehabil. 1970;51:78-87.

14. McBeath AA, Bahrke M, Balke B. Efficiency of assisted ambulation determined by oxygen consumption measurement. J Bone Joint Surg Am. 1974;56:994-1000.

15. Nielsen DH, Shurr DG, Golden JC, et al. Comparison of energy cost and gait efficiency during ambulation in below-knee amputees using different prosthetic feet: a preliminary report. Journal of Prosthetics and Orthotics. 1988;1:23-30.

16. Ralston HJ. Energy speed relation and optimal speed during level walking. Int Z Angew Physiol. 1958;17:277-283.

17. Smyntek Dj. Efficiency and Assement of Gait in Cerebral-Palsied Children. Iowa City, Iowa: The University of Iowa; 1978. Master's thesis.

18. Joll CA. An improved crutch. Lancet. 1917;1:583.

19. Epstein S. Art, history and the crutch. Ann Med Hist. 1937;9:304-313.

20. Hall RG. A rolling crutch. JAMA 1918;70:666-668.

21. Gillespie FC, Fisher J, Williams CS, et al. A physiological assessment of the rolling crutch. Ergonomics. 1983;26:341-347.

22. Lundquist EF. Design and Analysis of Experiments in Psychology and Education. Boston, Mass: Houghton Mifflin Co; 1953:17.

23. Minor MAD, Minor S. Patient Care Skills. Reston, Va: Reston Publishing Co Inc; 1984:162.

24. Nielsen DH, Gerleman DG, Amundsen LR, Hoeper DA. Clinical determination of energy cost and walking velocity via stopwatch or speedometer cane and conversion graphs. Phys Ther. 1982;62:591-596.

25. Lough LK, Nielsen DH. Ambulation of children with myelomeningocele: parapodium versus parapodium with ORLAU swivel modification. Dev Med Child Neurol. 1986;28:489-497.

26. Fox SM, Naughton JP, Haskell WL. Physical activity and the prevention of coronary heart disease. Ann Clin Res. 1971;3:404-432.

27. Rohrig W. Submaximal Exercise Testing: Treadmill and Floor Walking. Iowa City, Iowa: The University of Iowa; 1978. Master's thesis.

28. Sokal RR, Rohlf FJ. Biometry. San Francisco, Calif: WH Freeman & Co Publishers; 1969:400-403.