- Email: [email protected]
Why is walker-assisted gait metabolically expensive?
Gait & Posture 34 (2011) 265–269
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Why is walker-assisted gait metabolically expensive? Jonathon R. Priebe, Rodger Kram * Department of Integrative Physiology, University of Colorado, Boulder, CO 80309-0354, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Received 13 January 2011 Received in revised form 4 May 2011 Accepted 16 May 2011
Walker-assisted gait is reported to be 200% more metabolically expensive than normal bipedal walking. However, previous studies compared different walking speeds. Here, we compared the metabolic power consumption and basic stride temporal–spatial parameters for 10 young, healthy adults walking without assistance and using 2-wheeled (2W), 4-wheeled (4W) and 4-footed (4F) walker devices, all at the same speed, 0.30 m/s. We also measured the metabolic power demand for walking without any assistive device using a step-to gait at 0.30 m/s, walking normally at 1.25 m/s, and for repeated lifting of the 4F walker mimicking the lifting pattern used during 4F walker-assisted gait. Similar to previous studies, we found that the cost per distance walked was 217% greater with a 4F walker at 0.30 m/s compared to unassisted, bipedal walking at 1.25 m/s. Compared at the same speed, 0.30 m/s, using a 4F walker was still 82%, 74%, and 55% energetically more expensive than walking unassisted, with a 4W walker and a 2W walker respectively. The sum of the metabolic cost of step-to walking plus the cost of lifting itself was equivalent to the cost of walking with a 4F walker. Thus, we deduce that the high cost of 4F walker assisted gait is due to three factors: the slow walking speed, the step-to gait pattern and the repeated lifting of the walker. ß 2011 Elsevier B.V. All rights reserved.
Keywords: Walking Energetics Elderly
Approximately 2 million Americans use walkers to assist with their ambulation and more than three-quarters of walker users are elderly [1,2]. Many elderly people report that walkers are difficult to use [3,4]. A major difficulty is that the metabolic cost of ambulation with a walker is greatly elevated [5–9] combined with the diminished aerobic capacity associated with aging [10–14]. In this study, we quantified the major reasons why walker-assisted gait is so expensive. Three general types of walkers exist: 4-wheeled (4W), 2wheeled (2W) and 4-footed (4F). Wheeled designs are easier to use than 4F walkers, but 4F walkers are essential tools for many individuals with gait disabilities because they are more stable. 4F walkers decrease ambulation speed [5] involve an unnatural ‘‘stepto’’ gait and require lifting of the walker with each stride. In the step-to gait, the user must repeatedly step forward with one foot and then step-to the same position with the other foot [4], see Fig. 1. All three factors likely elevate the energetic cost of walking. Two independent studies [5,6] have reported that even in young, healthy persons, walking with a 4F walker increases the oxygen cost per unit distance walked (ml/kg/m) by 200% compared to unassisted bipedal walking! In contrast, wheeledwalkers incur relatively small increases in oxygen consumption
* Corresponding author. Tel.: +1 303 492 7984. E-mail address: [email protected] (R. Kram). 0966-6362/$ – see front matter ß 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2011.05.011
[5,6]. However, neither Holder et al. [5] nor Foley et al. [6] controlled for walking speed. The metabolic cost per distance (J/kg/m) for normal, unassisted walking strongly depends on speed. When humans walk either slower or faster than their normal, preferred walking speed (1.4 m/ s) the metabolic cost per distance increases, yielding a U-shaped relationship [15–21]. Walking energetics are also influenced by temporal–spatial factors such as stride length. Walking with either unnaturally short or long strides increases metabolic cost (reviewed by [22]). The effect of walker use on stride length is not known and might explain part of the greater metabolic cost of using walkers. The gait or pattern of leg movements also affects energetic cost. For example, at speeds slower than 2 m/s, walking is cheaper than running, but at speeds faster than 2 m/s, walking is more expensive e.g., [23]. The gaits chosen by quadrupedal animals (walk, trot or gallop) affect energetic cost in a similar way [24]. Because 4F walkers require the unusual step-to gait, it is important to examine the effects of gait alone. Healthy, bipedal human walking involves the smooth exchange of the kinetic and gravitational potential energy of the center of mass [25,26]. This mechanical energy exchange is thought to minimize the metabolic cost of walking. Because a step-to gait involves coming to a complete stop with each step, at which point kinetic energy is zero, the smooth exchange of mechanical energy is likely disrupted and thus has metabolic cost consequences. We investigated the energetic cost and basic stride temporal– spatial parameters involved in walking unassisted, with three
[(Fig._1)TD$IG]
266
J.R. Priebe, R. Kram / Gait & Posture 34 (2011) 265–269 1.4. Experimental procedures During eight energetics trials, we measured the subjects’ expired gases as they walked on a 5HP treadmill with a 50 cm wide belt. We measured preferred walking speeds over a 30 m walkway and stride frequency data on the treadmill. 1.5. Energetics
Fig. 1. The sequential footfall patterns for normal bipedal walking vs. the step-to gait used with a four-footed walker.
different walkers, and two different gait patterns. We hypothesized that: (1) at a fixed speed, walking with a 4F walker is metabolically more expensive that walking unassisted or with a 4W or 2W walker and (2) the greater cost of using a 4F walker is due to the slow walking speed, the step-to gait pattern, and the repeated lifting the walker. We tested these hypotheses on healthy, young subjects because people who are elderly or who have disabilities would be unable to sustain their locomotion long enough to obtain steady-state metabolic rate data [8]. 1. Methods 1.1. Subjects Ten young, healthy, adults volunteered [five males, five females, mean age 25.8 yrs (SD 4.4), mean body mass 70.9 kg (12.2)]. All subjects gave informed consent in accordance with the University of Colorado Human Research Committee policy. All subjects had prior experience with treadmill walking.
Subjects rested in a chair for 10 min before we collected data for quiet standing. We proceeded to the treadmill trials, each of which consisted of 7 min of walking followed by 3 min of seated rest. We collected metabolic data using an expired gas analysis system (Physio-Dyne Instruments, Quogue, NY) that measured the rates of oxygen consumption and carbon dioxide production. Subjects reached steady-state breathing during the first 4 min of every trial. We determined the average rates of oxygen consumption and carbon dioxide production for minutes 4–6 of each trial and divided by kilograms of body mass to calculate ml/kg/min. Using the Brockway equation [31], we converted ml/kg/s of O2 and CO2 into metabolic power in W/kg. We then subtracted the metabolic power for the standing trial to determine net metabolic power. We divided the gross metabolic power by the walking velocity to obtain the gross cost of transport (COT) measured in J/kg/m. To compare walking unassisted to using each walker device, we implemented four walking trials: bipedal (BP) unassisted, BP with a 4W walker, BP with a 2W walker, and step-to (ST) with a 4F walker. In order to partition the metabolic cost of walking associated with a 4F walker, the subjects performed two additional trials: walking with a ST gait pattern unassisted (i.e. without a walker), and repeated lifting of the 4F walker while standing in place. The order of the metabolic trials is shown in Table 1. The initial trial consisted of quiet standing, followed by six randomized trials, and ended with a normal bipedal walking trial at 1.25 m/s. The randomized walking trials were all at 0.30 m/s. We determined the temporal pattern of 4F walking from video analysis of the final 30 strides of familiarization with the 4F walker at 0.30 m/s. The unassisted ST and lift trials used a multi-beat computer-generated metronome to enforce the individual’s temporal movement pattern with the 4F walker. During the ST trial, subjects stepped with their lead leg when the first metronome beat occurred and then advanced their other leg when the second beat occurred. Subjects then paused when lifting of the 4F walker would normally take place. Conversely, for the lifting trial, subjects did not walk at all, but they lifted the 4F walker in front of them as though they were about to step forward. After placing the walker slightly in front of them, the subjects paused when normal ST walking would occur. The subjects then started the cycle again, but lifted the walker back to the starting position when the second metronome beat occurred, and then paused following the placement of the walker once again. The multi-beat metronome software was obtained from www.weirdmetronome.com. 1.6. Temporal–spatial parameters We determined the preferred walking speeds (PWS) using the following procedure for each walker device. Subjects practiced walking two lengths of the walkway, one quickly and one comfortably with a walker. We then instructed the subjects to ‘‘walk at whatever speed feels most comfortable’’ while we measured the time they took to walk the middle 24 m of the 30 m walkway. To permit the subjects to accelerate to and decelerate from their steady speed, 3 m of untimed walking was allowed at the start and end of each trial. The reported PWS is the average of three consecutive trials of walking the 24 m walkway. We also measured three normal, unassisted PWS trials. We measured the time for 10 strides using a stopwatch during steady state treadmill walking during each metabolic trial. We then calculated stride frequency by taking the inverse of that time. Knowing treadmill speed (V) and the measured stride frequency (SF), we calculated stride length (SL) as: SL = V/SF.
1.2. Walkers and gait patterns We studied three walking aids: a 4-wheeled walker (4W), a 2-wheeled walker (2W), and a 4-footed walker (4F). The 4W walker (Hugo X5, mass 8.65 kg) had two non-swivel rear wheels and two swivel caster front wheels (all 20 cm diameter). The 2W walker (Invacare Dual Blue-Release Wheeled Walker, 2.71 kg) had rubber feet on the back legs and two small non-swivel wheels (7.6 cm diameter) in the front. The 4F walker (Guardian Easy Care Double Button Folding Walker, mass 2.87 kg) had four rubber feet. We adjusted each walker so that each subject’s elbows were flexed between 20 and 30 degrees when both hands grasped the handles [27,28]. We studied two different gait patterns: bipedal (BP) and step-to (ST). 1.3. Training First, we demonstrated the bipedal and step-to gait patterns to the subjects. They then practiced over a 30 m walkway for a total of 25 min. We instructed the subjects to minimize incidental weight bearing on the walker by using a light grip [29]. The subjects practiced for 15 min with the 4F walker. Next, the subjects practiced with the 2W and 4W walkers for 5 min each. As recommended for familiarization [30], subjects practiced walking on the motorized treadmill at 0.30 m/s, the average speed of healthy, older adults using a 4F walker [6]. Subjects first walked on the treadmill BP for 3 min, then with the ST gait unassisted for 3 min, and finally ST with the 4F walker for 4 min.
1.7. Statistical analysis We used repeated-measures ANOVA with Tukey post hoc analysis to compare the means for each variable across trials. Our criterion for statistical significance was p < 0.05.
Table 1 Experimental trials. Trial order
Type of assistance
Speed
Gait
1 Random Random Random Random Random Random 8
None None 2-Wheeled walker 4-Wheeled walker 4-Footed walker None 4-Footed walker None
Standing 0.30 m/s 0.30 m/s 0.30 m/s 0.30 m/s 0.30 m/s Standing 1.25 m/s
None Bipedal Bipedal Bipedal Step-to Step-to Repeated lift Bipedal
[(Fig._3)TD$IG]
J.R. Priebe, R. Kram / Gait & Posture 34 (2011) 265–269
267
2.0
2. Results 2.1. Energetics
Table 2 Energetic and kinematic data.
[(Fig._2)TD$IG]
Mean SD Mean SD Mean SD Mean SD
9.73 3.72 2.26 0.38 2.95 0.36 1.36 0.08
4.17 2.01 0.94 0.17 7.91 1.08 0.60 0.09
4.33 2.12 0.97 0.21 8.01 1.00 0.60 0.07
4.88 2.39 1.11 0.28 8.48 1.30 0.54 0.10
7.55 3.05 1.72 0.25 10.52 0.82 0.57 0.05
5.52 2.94 1.22 0.28 8.83 1.09 0.57 0.05
1.84 0.64 0.43 0.11
n = 10; BP, bipedal; 4W, 4 wheeled walker; 2W, 2 wheeled walker; 4F, 4 footed ˙ 2 for standing was 6.11 ml/kg/min. walker; ST, step-to gait; VO
9 8
•
Net VO2 (ml/kg/min)
Lift
1.0
Bipedal
Step-to
4-footed walker
0.5
0.0
BP BP 4W 2W 4F ST Lift 1.25 m/s 0.30 m/s 0.30 m/s 0.30 m/s 0.30 m/s 0.30 m/s ˙ 2 Net VO (ml/kg/min) Net power (W/kg) Gross COT (J/kg/m) Stride length (m)
1.5
Net Metabolic Power (W/kg)
The energetic cost of walking was affected by three factors: speed, gait pattern, and walker lifting. Because there were no statistically significant differences between sexes for any of the energetic variables (p > 0.25), we pooled the data (Table 2). As expected, the slower walking speed increased the cost per distance. Net metabolic cost per distance of BP walking at 0.30 m/s was 73% greater than BP walking at 1.25 m/s (p < 0.001). To factor out speed, we made subsequent comparisons at the same speed (0.30 m/s) using different gaits and walkers. When subjects walked with the 4F walker and ST walking pattern at 0.30 m/s, the net metabolic cost per distance was ‘‘only’’ 82% greater than bipedal (BP) walking unassisted at the same speed (p < 0.001). The metabolic cost of walking with the 4W and 2W walkers was only 4% (p = 0.33) and 17% (p < 0.005) greater than BP walking at the same speed (Table 2 and Fig. 2). Calculated another way, walking with the 4F walker was 74% energetically more expensive that walking with a 4W walker and 55% more expensive than using a 2W walker (p < 0.001). At 0.30 m/s, the mean net metabolic power consumption while using the 4F walker (1.72 W/kg) was much greater than for bipedal
7 6
Fig. 3. At 0.30 m/s, the greater metabolic power demand of walking with a 4-footed walker compared to normal bipedal walking can be explained by the step-to gait and the cost of repeatedly lifting the walker. Graph depicts the net metabolic power consumption (mean SE, n = 10) for normal bipedal walking (left bar), step-to walking without any assistance (middle lower bar), and walking with a step-to gait using a 4-footed walker (right bar). Stacked on top of the middle bar is the net metabolic power demand for repeatedly lifting the walker while standing still. The sum of the net metabolic power for step-to walking and pure lifting is not significantly different from that for step-to walking with the 4-footed walker.
walking (p < 0.001; Table 2). The greater net metabolic power was due to both the ST gait and 4F walker lifting. The mean net metabolic power consumption for BP walking was 0.94 W/kg and increased by 30% to 1.22 W/kg when the subjects walked with the ST gait pattern unassisted. The mean net metabolic power required for just lifting the 4F walker was 0.43 W/kg. Thus, the hypothetical combined sum of the metabolic power required for ST walking (1.22 W/kg) plus 4F walker lifting (0.43 W/kg) was 1.65 W/kg which was not statistically different from the true 4F walker value (1.72 W/kg; p = 0.25; Fig. 3). Thus, when speed is controlled for, the high metabolic cost of using a 4F walker can be directly attributed to just two factors: the ST gait and the lifting of the walker.
5 2.2. Temporal–spatial parameters
4 3
Bipedal
4-wheeled walker
2-wheeled walker
4-footed walker
2 1 0 Fig. 2. Mean values (SE, n = 10) for the net rate of oxygen consumption during walking at 0.30 m/s using: a normal bipedal gait without assistance, a bipedal gait with a 4-wheeled walker, a bipedal gait with a 2-wheeled walker and a step-to gait with a 4footed walker. Walking with the 4-wheeled walker was not significantly greater than normal walking but the 2-wheeled walker increased the net rate of oxygen consumption by 17% (p < 0.005). Using the 4-footed walker increased the rate of oxygen consumption by 81% (p < 0.001).
At 0.30 m/s, there were no statistical differences in stride length between the trials using the different walkers (p > 0.05). Since there were no statistically significant differences between sexes for any of the temporal–spatial variables (p > 0.17), we pooled the data. Stride length with the 2W walker at 0.30 m/s (Table 2) only differed statistically from BP walking (p = 0.008). The type of walker significantly affected the PWS (p < 0.02). The slowest preferred walking speed occurred with the 4F walker (0.35 m/s), which was only one-fourth as fast as any other PWS (BP = 1.52 m/s, 4W = 1.37 m/s, and 2W = 1.23 m/s; p < 0.001). 3. Discussion Our results support both of our hypotheses. At the same speed, 0.30 m/s, walking with the 4F walker required much greater
268
J.R. Priebe, R. Kram / Gait & Posture 34 (2011) 265–269
metabolic power than bipedal walking unassisted. In accordance with our second hypothesis, the greater cost of the step-to gait combined with the cost of lifting the 4F walker could fully account for the greater metabolic cost of 4F walking compared to unassisted bipedal walking at the same speed. Previous studies [5,6] have reported that 4F walkers incur remarkably high metabolic penalties. In contrast, we compared the different gaits and walker types at the same speed. Although our findings showed a much greater metabolic cost of using walkers, the differences at matched speeds were not nearly as drastic. Foley et al. [6] reported that walking with a 4F walker at 0.30 m/s increased the net COT by 212% compared to unassisted bipedal walking at 1.4 m/s. We found a similar 217% greater net COT for the 4F walker condition at 0.30 m/s compared to normal, unassisted walking at 1.25 m/s (Table 2). However, when we compared BP walking and walking with the 4F walker at the same slow speed (0.30 m/s) there was an 82% greater net COT (Table 2). While the high metabolic cost of transport when using a 4F walker at slow speeds is of interest, the greater gross rate of oxygen ˙ 2 that occurs while using the 4F walker is a greater consumption VO ˙ 2 of 10 ml/kg/min is only a modest effort for young, concern. A VO ˙ 2 healthy people, but for an elderly user, it is a large fraction of VO ˙ 2 max declines max. Several cross-sectional studies suggest that VO approximately 0.58 ml O2/kg/min/year in older populations (people 55–85) [10–14]. Consider a typical 85-year old, commu˙ 2 max of approximately 15 ml/ nity dwelling woman that has a VO kg/min [14]. Even neglecting the fact that walking economy worsens with age [32], using a 4F walker at 0.30 m/s would demand a gross rate of oxygen consumption of 13.7 ml/kg/min or ˙ 2 max (Table 2, VO ˙ 2 standing plus net approximately 91% of her VO ˙ 2 4F). Only elite endurance athletes (e.g. marathoners) can VO sustain such high relative aerobic intensities [33]. Unlike the 4F walker, ambulating with a 4W walker at 0.30 m/s incurred only an insignificant 4% greater metabolic cost compared to walking unassisted at the same speed (p = 0.33, Fig. 2). However, 4W walkers do not have the inherent static stability of a 4F walker. When prescribing a walking device for a person who has trouble with their balance, stability and safety would take precedence over metabolic cost. An optimal solution might be a 4W walker that could instantly convert to a 4F walker via an intelligent electronic braking system. Such a device would remove most all of the energetic penalty of the 4F walker (ST gait and lifting) yet provide the stability of a 4F walker when needed. Alternatively, our data suggest that an ultra-light 4F walker (e.g. made of titanium or carbon fiber) could substantially reduce the cost of lifting the walker and thus minimize the energetic burden on the user. Before collecting these data, we expected that young subjects would prefer to walk faster with the 4F walker than older people. However, it appears that the 4F walker design intrinsically limits gait speed. On average, our young, healthy subjects chose to walk only slightly faster (0.35 m/s) with the 4F walker than healthy, older subjects (0.30 m/s) [6]. There also appears to be some influence of the 2W and 4W walkers. All but one subject walked slower with the wheeled walkers compared to their normal PWS. Rolling resistance or the inability to swing the arms may cause of the decrease in speed with wheeled walkers. 3.1. Limitations We did not investigate the effects of walkers on an elderly or disabled population who need the assistance of walkers to ambulate. It would be difficult to conduct such a study because the diminished aerobic capacity of such subjects would preclude sustainable, steady-state metabolic rates [8]. Nevertheless, the general mechanisms that we have quantified likely apply to elderly or disabled individuals.
4. Conclusion Like previous studies, we found that using a 4F walker at a typical slow speed (0.30 m/s) is 200% more expensive per distance than bipedal walking at a normal speed (1.25 m/s). However, when 4F and bipedal walking are compared at the same speed (0.30 m/s), 4F is still 82% more expensive metabolically. Two major reasons account for the greater metabolic cost: the step-to walking pattern required by the 4F walker, and the repeated lifting of the 4F walker for ambulation. Development of a lighter 4F walker or a dynamically stable 4W walker might mitigate the high metabolic cost of using a 4F walker. Conflicts of interest There are no conflicts of interest. References [1] Jones ML, Sanford JA. People with mobility impairment in the United States today and in 2010. Assist Tech 1996;8:43–53. [2] Kaye HS, Kang T, LaPlante MP.In: Mobility device use in the United States, Disability statistics report, 14. Washington, DC: U.S. Department of Education, National Institute on Disability and Rehabilitation Research; 2000. [3] Mann WC, Hurren D, Tomita M, Charvat B. An analysis of problems with walkers encountered by elderly persons. Phys Occup Ther Geriatr 1995;13:1– 23. [4] Bateni H, Maki BE. Assistive devices for balance and mobility: benefits, demands and adverse consequences. Arch Phys Med Rehabil 2005;86:134–45. [5] Holder CG, Haskvitz EM, Weltman A. The effect of assistive devices on the oxygen cost, cardiovascular stress, and perception of nonweight-bearing ambulation. J Orthop Sports Phys Ther 1993;18:537–42. [6] Foley MP, Prax B, Crowell R, Boone T. Effects of assistive devices on cardiorespiratory demands in older adults. Phys Ther 1996;76:1313–9. [7] Ulkar B, Yavuzer G, Guner R, Ergin S. Energy expenditure of the parapalegic gait: comparison between different walking aids and normal subjects. Int J Rehabil Res 2003;26:213–7. [8] Eggermont LH, van Heuvelen MJ, van Keeken BL, Hollander AP, Scherder EJ. Walking with a rollator and the level of physical intensity in adults 75 years of age or older. Arch Phys Med Rehabil 2006;87:733–6. [9] Cetin E, Muzembo J, Pardessus V, Prusieux F, Thevenon A. Impact of different types of walking aids on the physiological energy cost during gait for elderly individuals with several pathologies and dependent on a technical aid for walking. Ann Phys Med Rehabil 2010;53:399–405. [10] Astrand I, Astrand PO, Hallback I, Kilbom A. Reduction in maximal oxygen uptake with age. J Appl Physiol 1973;35:649–54. [11] Pollock M, Miller H, Ribisl P. Effect of fitness on aging. Phys Sports Med 1978;1:45–8. [12] Pollock ML, Foster C, Knapp D, Rod JL, Schmidt DH. Effect of age and training on aerobic-capacity and body composition of master athletes. J Appl Physiol 1987;62:625–31. [13] Robinson S, Dill DB, Robinson RD, Tzankoff SP, Wagner JA. Physiological aging of champion runners. J Appl Physiol 1976;41:46–51. [14] Puggaard L, Larsen JB, Stovring H, Jeune B. Maximal oxygen uptake, muscle strength and walking speed in 85-year-old women: effects of increased physical activity. Aging (Milano) 2000;12:180–9. [15] Browning RC, Kram R. Energetic cost and preferred speed of walking in obese vs. normal weight woman. Obes Res 2005;13:891–9. [16] Cotes JE, Meade F. The energy expenditure and mechanical energy demand in walking. Ergonomics 1960;3:97–119. [17] Den Otter AR, Geurts ACH, Mulder T, Duysens J. Speed related changes in muscle activity from normal to very slow walking speeds. Gait Posture 2004;19:270–8. [18] Franklin BA, Pamatmat A, Johnson S, Scherf J, Mitchell M, Rubenfire M. Metabolic cost of extremely slow walking in cardiac patients: implications for exercise testing and training. Arch Phys Med Rehabil 1983;64:564–5. [19] Margaria R. Positive and negative work performances and their efficiencies in human locomotion. Int Z Angew Physiol Einschl Arbeitsphysiol 1968;25:339– 51. [20] Ralston HJ. Energy-speed relation and optimal speed during level walking. Int Z Angew Physiol 1958;17:277–83. [21] Waters RL, Lunsford BR, Perry J, Byrd R. Energy-speed relationship of walking: standard tables. J Orthop Res 1988;6:215–22. [22] Umberger BR, Martin PE. Mechanical power and efficiency of level walking with different stride rates. J Exp Biol 2007;210:3255–65. [23] Mercier J, Le Gallais D, Durrand M, Goudal C, Micallef J, Prefaut C. Energy expenditure and cardiorespiratory responses at the transition between walking and running. Eur J Appl Physiol 1994;69:525–9. [24] Hoyt DF, Taylor CR. Gait and the energetics of locomotion in horses. Nature 1981;292:239–40.
J.R. Priebe, R. Kram / Gait & Posture 34 (2011) 265–269 [25] Cavagna GA, Margaria R. Mechanics of walking. J Appl Physiol 1966;21:271–8. [26] Cavagna GA, Saibene F, Margaria R. External work in walking. J Appl Physiol 1963;18:1–9. [27] Fast A, Wang FS, Adrezin RS, Cordaro MA, Ramis J, Sosner J. The instrumented walker: usage patterns and forces. Arch Phys Med Rehabil 2005;76:484–91. [28] Nabizadeh SA, Hardee TB, Towler MA, Chen VT, Edlich RF. Technical considerations in the selection and performance of walkers. J Burn Care Rehabil 1993;14:182–8. [29] Youdas JW, Kotajarvi BJ, Padgett DJ, Kaufman KR. Partial weight-bearing gait using conventional assistive devices. Arch Phys Med Rehabil 2005;86:394–8.
269
[30] Wall JC, Charteris J. A kinematic study of long-term habituation to treadmill walking. Ergonomics 1981;24:531–42. [31] Brockway JM. Derivation of formulae used to calculate energy expenditure in man. Human Nutr Clin Nutr 1987;41:463–71. [32] Martin PE, Rothstein DE, Larish DD. Effects of age and physical activity status on the speed-aerobic demand relationship of walking. J Appl Physiol 1992;73:200–6. [33] Brooks GA, Fahey T, Baldwin K. Exercise physiology—human bioenergetics and its applications. McGraw-Hill; 2004.
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Why is walker-assisted gait metabolically expensive? Jonathon R. Priebe, Rodger Kram * Department of Integrative Physiology, University of Colorado, Boulder, CO 80309-0354, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Received 13 January 2011 Received in revised form 4 May 2011 Accepted 16 May 2011
Walker-assisted gait is reported to be 200% more metabolically expensive than normal bipedal walking. However, previous studies compared different walking speeds. Here, we compared the metabolic power consumption and basic stride temporal–spatial parameters for 10 young, healthy adults walking without assistance and using 2-wheeled (2W), 4-wheeled (4W) and 4-footed (4F) walker devices, all at the same speed, 0.30 m/s. We also measured the metabolic power demand for walking without any assistive device using a step-to gait at 0.30 m/s, walking normally at 1.25 m/s, and for repeated lifting of the 4F walker mimicking the lifting pattern used during 4F walker-assisted gait. Similar to previous studies, we found that the cost per distance walked was 217% greater with a 4F walker at 0.30 m/s compared to unassisted, bipedal walking at 1.25 m/s. Compared at the same speed, 0.30 m/s, using a 4F walker was still 82%, 74%, and 55% energetically more expensive than walking unassisted, with a 4W walker and a 2W walker respectively. The sum of the metabolic cost of step-to walking plus the cost of lifting itself was equivalent to the cost of walking with a 4F walker. Thus, we deduce that the high cost of 4F walker assisted gait is due to three factors: the slow walking speed, the step-to gait pattern and the repeated lifting of the walker. ß 2011 Elsevier B.V. All rights reserved.
Keywords: Walking Energetics Elderly
Approximately 2 million Americans use walkers to assist with their ambulation and more than three-quarters of walker users are elderly [1,2]. Many elderly people report that walkers are difficult to use [3,4]. A major difficulty is that the metabolic cost of ambulation with a walker is greatly elevated [5–9] combined with the diminished aerobic capacity associated with aging [10–14]. In this study, we quantified the major reasons why walker-assisted gait is so expensive. Three general types of walkers exist: 4-wheeled (4W), 2wheeled (2W) and 4-footed (4F). Wheeled designs are easier to use than 4F walkers, but 4F walkers are essential tools for many individuals with gait disabilities because they are more stable. 4F walkers decrease ambulation speed [5] involve an unnatural ‘‘stepto’’ gait and require lifting of the walker with each stride. In the step-to gait, the user must repeatedly step forward with one foot and then step-to the same position with the other foot [4], see Fig. 1. All three factors likely elevate the energetic cost of walking. Two independent studies [5,6] have reported that even in young, healthy persons, walking with a 4F walker increases the oxygen cost per unit distance walked (ml/kg/m) by 200% compared to unassisted bipedal walking! In contrast, wheeledwalkers incur relatively small increases in oxygen consumption
* Corresponding author. Tel.: +1 303 492 7984. E-mail address: [email protected] (R. Kram). 0966-6362/$ – see front matter ß 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2011.05.011
[5,6]. However, neither Holder et al. [5] nor Foley et al. [6] controlled for walking speed. The metabolic cost per distance (J/kg/m) for normal, unassisted walking strongly depends on speed. When humans walk either slower or faster than their normal, preferred walking speed (1.4 m/ s) the metabolic cost per distance increases, yielding a U-shaped relationship [15–21]. Walking energetics are also influenced by temporal–spatial factors such as stride length. Walking with either unnaturally short or long strides increases metabolic cost (reviewed by [22]). The effect of walker use on stride length is not known and might explain part of the greater metabolic cost of using walkers. The gait or pattern of leg movements also affects energetic cost. For example, at speeds slower than 2 m/s, walking is cheaper than running, but at speeds faster than 2 m/s, walking is more expensive e.g., [23]. The gaits chosen by quadrupedal animals (walk, trot or gallop) affect energetic cost in a similar way [24]. Because 4F walkers require the unusual step-to gait, it is important to examine the effects of gait alone. Healthy, bipedal human walking involves the smooth exchange of the kinetic and gravitational potential energy of the center of mass [25,26]. This mechanical energy exchange is thought to minimize the metabolic cost of walking. Because a step-to gait involves coming to a complete stop with each step, at which point kinetic energy is zero, the smooth exchange of mechanical energy is likely disrupted and thus has metabolic cost consequences. We investigated the energetic cost and basic stride temporal– spatial parameters involved in walking unassisted, with three
[(Fig._1)TD$IG]
266
J.R. Priebe, R. Kram / Gait & Posture 34 (2011) 265–269 1.4. Experimental procedures During eight energetics trials, we measured the subjects’ expired gases as they walked on a 5HP treadmill with a 50 cm wide belt. We measured preferred walking speeds over a 30 m walkway and stride frequency data on the treadmill. 1.5. Energetics
Fig. 1. The sequential footfall patterns for normal bipedal walking vs. the step-to gait used with a four-footed walker.
different walkers, and two different gait patterns. We hypothesized that: (1) at a fixed speed, walking with a 4F walker is metabolically more expensive that walking unassisted or with a 4W or 2W walker and (2) the greater cost of using a 4F walker is due to the slow walking speed, the step-to gait pattern, and the repeated lifting the walker. We tested these hypotheses on healthy, young subjects because people who are elderly or who have disabilities would be unable to sustain their locomotion long enough to obtain steady-state metabolic rate data [8]. 1. Methods 1.1. Subjects Ten young, healthy, adults volunteered [five males, five females, mean age 25.8 yrs (SD 4.4), mean body mass 70.9 kg (12.2)]. All subjects gave informed consent in accordance with the University of Colorado Human Research Committee policy. All subjects had prior experience with treadmill walking.
Subjects rested in a chair for 10 min before we collected data for quiet standing. We proceeded to the treadmill trials, each of which consisted of 7 min of walking followed by 3 min of seated rest. We collected metabolic data using an expired gas analysis system (Physio-Dyne Instruments, Quogue, NY) that measured the rates of oxygen consumption and carbon dioxide production. Subjects reached steady-state breathing during the first 4 min of every trial. We determined the average rates of oxygen consumption and carbon dioxide production for minutes 4–6 of each trial and divided by kilograms of body mass to calculate ml/kg/min. Using the Brockway equation [31], we converted ml/kg/s of O2 and CO2 into metabolic power in W/kg. We then subtracted the metabolic power for the standing trial to determine net metabolic power. We divided the gross metabolic power by the walking velocity to obtain the gross cost of transport (COT) measured in J/kg/m. To compare walking unassisted to using each walker device, we implemented four walking trials: bipedal (BP) unassisted, BP with a 4W walker, BP with a 2W walker, and step-to (ST) with a 4F walker. In order to partition the metabolic cost of walking associated with a 4F walker, the subjects performed two additional trials: walking with a ST gait pattern unassisted (i.e. without a walker), and repeated lifting of the 4F walker while standing in place. The order of the metabolic trials is shown in Table 1. The initial trial consisted of quiet standing, followed by six randomized trials, and ended with a normal bipedal walking trial at 1.25 m/s. The randomized walking trials were all at 0.30 m/s. We determined the temporal pattern of 4F walking from video analysis of the final 30 strides of familiarization with the 4F walker at 0.30 m/s. The unassisted ST and lift trials used a multi-beat computer-generated metronome to enforce the individual’s temporal movement pattern with the 4F walker. During the ST trial, subjects stepped with their lead leg when the first metronome beat occurred and then advanced their other leg when the second beat occurred. Subjects then paused when lifting of the 4F walker would normally take place. Conversely, for the lifting trial, subjects did not walk at all, but they lifted the 4F walker in front of them as though they were about to step forward. After placing the walker slightly in front of them, the subjects paused when normal ST walking would occur. The subjects then started the cycle again, but lifted the walker back to the starting position when the second metronome beat occurred, and then paused following the placement of the walker once again. The multi-beat metronome software was obtained from www.weirdmetronome.com. 1.6. Temporal–spatial parameters We determined the preferred walking speeds (PWS) using the following procedure for each walker device. Subjects practiced walking two lengths of the walkway, one quickly and one comfortably with a walker. We then instructed the subjects to ‘‘walk at whatever speed feels most comfortable’’ while we measured the time they took to walk the middle 24 m of the 30 m walkway. To permit the subjects to accelerate to and decelerate from their steady speed, 3 m of untimed walking was allowed at the start and end of each trial. The reported PWS is the average of three consecutive trials of walking the 24 m walkway. We also measured three normal, unassisted PWS trials. We measured the time for 10 strides using a stopwatch during steady state treadmill walking during each metabolic trial. We then calculated stride frequency by taking the inverse of that time. Knowing treadmill speed (V) and the measured stride frequency (SF), we calculated stride length (SL) as: SL = V/SF.
1.2. Walkers and gait patterns We studied three walking aids: a 4-wheeled walker (4W), a 2-wheeled walker (2W), and a 4-footed walker (4F). The 4W walker (Hugo X5, mass 8.65 kg) had two non-swivel rear wheels and two swivel caster front wheels (all 20 cm diameter). The 2W walker (Invacare Dual Blue-Release Wheeled Walker, 2.71 kg) had rubber feet on the back legs and two small non-swivel wheels (7.6 cm diameter) in the front. The 4F walker (Guardian Easy Care Double Button Folding Walker, mass 2.87 kg) had four rubber feet. We adjusted each walker so that each subject’s elbows were flexed between 20 and 30 degrees when both hands grasped the handles [27,28]. We studied two different gait patterns: bipedal (BP) and step-to (ST). 1.3. Training First, we demonstrated the bipedal and step-to gait patterns to the subjects. They then practiced over a 30 m walkway for a total of 25 min. We instructed the subjects to minimize incidental weight bearing on the walker by using a light grip [29]. The subjects practiced for 15 min with the 4F walker. Next, the subjects practiced with the 2W and 4W walkers for 5 min each. As recommended for familiarization [30], subjects practiced walking on the motorized treadmill at 0.30 m/s, the average speed of healthy, older adults using a 4F walker [6]. Subjects first walked on the treadmill BP for 3 min, then with the ST gait unassisted for 3 min, and finally ST with the 4F walker for 4 min.
1.7. Statistical analysis We used repeated-measures ANOVA with Tukey post hoc analysis to compare the means for each variable across trials. Our criterion for statistical significance was p < 0.05.
Table 1 Experimental trials. Trial order
Type of assistance
Speed
Gait
1 Random Random Random Random Random Random 8
None None 2-Wheeled walker 4-Wheeled walker 4-Footed walker None 4-Footed walker None
Standing 0.30 m/s 0.30 m/s 0.30 m/s 0.30 m/s 0.30 m/s Standing 1.25 m/s
None Bipedal Bipedal Bipedal Step-to Step-to Repeated lift Bipedal
[(Fig._3)TD$IG]
J.R. Priebe, R. Kram / Gait & Posture 34 (2011) 265–269
267
2.0
2. Results 2.1. Energetics
Table 2 Energetic and kinematic data.
[(Fig._2)TD$IG]
Mean SD Mean SD Mean SD Mean SD
9.73 3.72 2.26 0.38 2.95 0.36 1.36 0.08
4.17 2.01 0.94 0.17 7.91 1.08 0.60 0.09
4.33 2.12 0.97 0.21 8.01 1.00 0.60 0.07
4.88 2.39 1.11 0.28 8.48 1.30 0.54 0.10
7.55 3.05 1.72 0.25 10.52 0.82 0.57 0.05
5.52 2.94 1.22 0.28 8.83 1.09 0.57 0.05
1.84 0.64 0.43 0.11
n = 10; BP, bipedal; 4W, 4 wheeled walker; 2W, 2 wheeled walker; 4F, 4 footed ˙ 2 for standing was 6.11 ml/kg/min. walker; ST, step-to gait; VO
9 8
•
Net VO2 (ml/kg/min)
Lift
1.0
Bipedal
Step-to
4-footed walker
0.5
0.0
BP BP 4W 2W 4F ST Lift 1.25 m/s 0.30 m/s 0.30 m/s 0.30 m/s 0.30 m/s 0.30 m/s ˙ 2 Net VO (ml/kg/min) Net power (W/kg) Gross COT (J/kg/m) Stride length (m)
1.5
Net Metabolic Power (W/kg)
The energetic cost of walking was affected by three factors: speed, gait pattern, and walker lifting. Because there were no statistically significant differences between sexes for any of the energetic variables (p > 0.25), we pooled the data (Table 2). As expected, the slower walking speed increased the cost per distance. Net metabolic cost per distance of BP walking at 0.30 m/s was 73% greater than BP walking at 1.25 m/s (p < 0.001). To factor out speed, we made subsequent comparisons at the same speed (0.30 m/s) using different gaits and walkers. When subjects walked with the 4F walker and ST walking pattern at 0.30 m/s, the net metabolic cost per distance was ‘‘only’’ 82% greater than bipedal (BP) walking unassisted at the same speed (p < 0.001). The metabolic cost of walking with the 4W and 2W walkers was only 4% (p = 0.33) and 17% (p < 0.005) greater than BP walking at the same speed (Table 2 and Fig. 2). Calculated another way, walking with the 4F walker was 74% energetically more expensive that walking with a 4W walker and 55% more expensive than using a 2W walker (p < 0.001). At 0.30 m/s, the mean net metabolic power consumption while using the 4F walker (1.72 W/kg) was much greater than for bipedal
7 6
Fig. 3. At 0.30 m/s, the greater metabolic power demand of walking with a 4-footed walker compared to normal bipedal walking can be explained by the step-to gait and the cost of repeatedly lifting the walker. Graph depicts the net metabolic power consumption (mean SE, n = 10) for normal bipedal walking (left bar), step-to walking without any assistance (middle lower bar), and walking with a step-to gait using a 4-footed walker (right bar). Stacked on top of the middle bar is the net metabolic power demand for repeatedly lifting the walker while standing still. The sum of the net metabolic power for step-to walking and pure lifting is not significantly different from that for step-to walking with the 4-footed walker.
walking (p < 0.001; Table 2). The greater net metabolic power was due to both the ST gait and 4F walker lifting. The mean net metabolic power consumption for BP walking was 0.94 W/kg and increased by 30% to 1.22 W/kg when the subjects walked with the ST gait pattern unassisted. The mean net metabolic power required for just lifting the 4F walker was 0.43 W/kg. Thus, the hypothetical combined sum of the metabolic power required for ST walking (1.22 W/kg) plus 4F walker lifting (0.43 W/kg) was 1.65 W/kg which was not statistically different from the true 4F walker value (1.72 W/kg; p = 0.25; Fig. 3). Thus, when speed is controlled for, the high metabolic cost of using a 4F walker can be directly attributed to just two factors: the ST gait and the lifting of the walker.
5 2.2. Temporal–spatial parameters
4 3
Bipedal
4-wheeled walker
2-wheeled walker
4-footed walker
2 1 0 Fig. 2. Mean values (SE, n = 10) for the net rate of oxygen consumption during walking at 0.30 m/s using: a normal bipedal gait without assistance, a bipedal gait with a 4-wheeled walker, a bipedal gait with a 2-wheeled walker and a step-to gait with a 4footed walker. Walking with the 4-wheeled walker was not significantly greater than normal walking but the 2-wheeled walker increased the net rate of oxygen consumption by 17% (p < 0.005). Using the 4-footed walker increased the rate of oxygen consumption by 81% (p < 0.001).
At 0.30 m/s, there were no statistical differences in stride length between the trials using the different walkers (p > 0.05). Since there were no statistically significant differences between sexes for any of the temporal–spatial variables (p > 0.17), we pooled the data. Stride length with the 2W walker at 0.30 m/s (Table 2) only differed statistically from BP walking (p = 0.008). The type of walker significantly affected the PWS (p < 0.02). The slowest preferred walking speed occurred with the 4F walker (0.35 m/s), which was only one-fourth as fast as any other PWS (BP = 1.52 m/s, 4W = 1.37 m/s, and 2W = 1.23 m/s; p < 0.001). 3. Discussion Our results support both of our hypotheses. At the same speed, 0.30 m/s, walking with the 4F walker required much greater
268
J.R. Priebe, R. Kram / Gait & Posture 34 (2011) 265–269
metabolic power than bipedal walking unassisted. In accordance with our second hypothesis, the greater cost of the step-to gait combined with the cost of lifting the 4F walker could fully account for the greater metabolic cost of 4F walking compared to unassisted bipedal walking at the same speed. Previous studies [5,6] have reported that 4F walkers incur remarkably high metabolic penalties. In contrast, we compared the different gaits and walker types at the same speed. Although our findings showed a much greater metabolic cost of using walkers, the differences at matched speeds were not nearly as drastic. Foley et al. [6] reported that walking with a 4F walker at 0.30 m/s increased the net COT by 212% compared to unassisted bipedal walking at 1.4 m/s. We found a similar 217% greater net COT for the 4F walker condition at 0.30 m/s compared to normal, unassisted walking at 1.25 m/s (Table 2). However, when we compared BP walking and walking with the 4F walker at the same slow speed (0.30 m/s) there was an 82% greater net COT (Table 2). While the high metabolic cost of transport when using a 4F walker at slow speeds is of interest, the greater gross rate of oxygen ˙ 2 that occurs while using the 4F walker is a greater consumption VO ˙ 2 of 10 ml/kg/min is only a modest effort for young, concern. A VO ˙ 2 healthy people, but for an elderly user, it is a large fraction of VO ˙ 2 max declines max. Several cross-sectional studies suggest that VO approximately 0.58 ml O2/kg/min/year in older populations (people 55–85) [10–14]. Consider a typical 85-year old, commu˙ 2 max of approximately 15 ml/ nity dwelling woman that has a VO kg/min [14]. Even neglecting the fact that walking economy worsens with age [32], using a 4F walker at 0.30 m/s would demand a gross rate of oxygen consumption of 13.7 ml/kg/min or ˙ 2 max (Table 2, VO ˙ 2 standing plus net approximately 91% of her VO ˙ 2 4F). Only elite endurance athletes (e.g. marathoners) can VO sustain such high relative aerobic intensities [33]. Unlike the 4F walker, ambulating with a 4W walker at 0.30 m/s incurred only an insignificant 4% greater metabolic cost compared to walking unassisted at the same speed (p = 0.33, Fig. 2). However, 4W walkers do not have the inherent static stability of a 4F walker. When prescribing a walking device for a person who has trouble with their balance, stability and safety would take precedence over metabolic cost. An optimal solution might be a 4W walker that could instantly convert to a 4F walker via an intelligent electronic braking system. Such a device would remove most all of the energetic penalty of the 4F walker (ST gait and lifting) yet provide the stability of a 4F walker when needed. Alternatively, our data suggest that an ultra-light 4F walker (e.g. made of titanium or carbon fiber) could substantially reduce the cost of lifting the walker and thus minimize the energetic burden on the user. Before collecting these data, we expected that young subjects would prefer to walk faster with the 4F walker than older people. However, it appears that the 4F walker design intrinsically limits gait speed. On average, our young, healthy subjects chose to walk only slightly faster (0.35 m/s) with the 4F walker than healthy, older subjects (0.30 m/s) [6]. There also appears to be some influence of the 2W and 4W walkers. All but one subject walked slower with the wheeled walkers compared to their normal PWS. Rolling resistance or the inability to swing the arms may cause of the decrease in speed with wheeled walkers. 3.1. Limitations We did not investigate the effects of walkers on an elderly or disabled population who need the assistance of walkers to ambulate. It would be difficult to conduct such a study because the diminished aerobic capacity of such subjects would preclude sustainable, steady-state metabolic rates [8]. Nevertheless, the general mechanisms that we have quantified likely apply to elderly or disabled individuals.
4. Conclusion Like previous studies, we found that using a 4F walker at a typical slow speed (0.30 m/s) is 200% more expensive per distance than bipedal walking at a normal speed (1.25 m/s). However, when 4F and bipedal walking are compared at the same speed (0.30 m/s), 4F is still 82% more expensive metabolically. Two major reasons account for the greater metabolic cost: the step-to walking pattern required by the 4F walker, and the repeated lifting of the 4F walker for ambulation. Development of a lighter 4F walker or a dynamically stable 4W walker might mitigate the high metabolic cost of using a 4F walker. Conflicts of interest There are no conflicts of interest. References [1] Jones ML, Sanford JA. People with mobility impairment in the United States today and in 2010. Assist Tech 1996;8:43–53. [2] Kaye HS, Kang T, LaPlante MP.In: Mobility device use in the United States, Disability statistics report, 14. Washington, DC: U.S. Department of Education, National Institute on Disability and Rehabilitation Research; 2000. [3] Mann WC, Hurren D, Tomita M, Charvat B. An analysis of problems with walkers encountered by elderly persons. Phys Occup Ther Geriatr 1995;13:1– 23. [4] Bateni H, Maki BE. Assistive devices for balance and mobility: benefits, demands and adverse consequences. Arch Phys Med Rehabil 2005;86:134–45. [5] Holder CG, Haskvitz EM, Weltman A. The effect of assistive devices on the oxygen cost, cardiovascular stress, and perception of nonweight-bearing ambulation. J Orthop Sports Phys Ther 1993;18:537–42. [6] Foley MP, Prax B, Crowell R, Boone T. Effects of assistive devices on cardiorespiratory demands in older adults. Phys Ther 1996;76:1313–9. [7] Ulkar B, Yavuzer G, Guner R, Ergin S. Energy expenditure of the parapalegic gait: comparison between different walking aids and normal subjects. Int J Rehabil Res 2003;26:213–7. [8] Eggermont LH, van Heuvelen MJ, van Keeken BL, Hollander AP, Scherder EJ. Walking with a rollator and the level of physical intensity in adults 75 years of age or older. Arch Phys Med Rehabil 2006;87:733–6. [9] Cetin E, Muzembo J, Pardessus V, Prusieux F, Thevenon A. Impact of different types of walking aids on the physiological energy cost during gait for elderly individuals with several pathologies and dependent on a technical aid for walking. Ann Phys Med Rehabil 2010;53:399–405. [10] Astrand I, Astrand PO, Hallback I, Kilbom A. Reduction in maximal oxygen uptake with age. J Appl Physiol 1973;35:649–54. [11] Pollock M, Miller H, Ribisl P. Effect of fitness on aging. Phys Sports Med 1978;1:45–8. [12] Pollock ML, Foster C, Knapp D, Rod JL, Schmidt DH. Effect of age and training on aerobic-capacity and body composition of master athletes. J Appl Physiol 1987;62:625–31. [13] Robinson S, Dill DB, Robinson RD, Tzankoff SP, Wagner JA. Physiological aging of champion runners. J Appl Physiol 1976;41:46–51. [14] Puggaard L, Larsen JB, Stovring H, Jeune B. Maximal oxygen uptake, muscle strength and walking speed in 85-year-old women: effects of increased physical activity. Aging (Milano) 2000;12:180–9. [15] Browning RC, Kram R. Energetic cost and preferred speed of walking in obese vs. normal weight woman. Obes Res 2005;13:891–9. [16] Cotes JE, Meade F. The energy expenditure and mechanical energy demand in walking. Ergonomics 1960;3:97–119. [17] Den Otter AR, Geurts ACH, Mulder T, Duysens J. Speed related changes in muscle activity from normal to very slow walking speeds. Gait Posture 2004;19:270–8. [18] Franklin BA, Pamatmat A, Johnson S, Scherf J, Mitchell M, Rubenfire M. Metabolic cost of extremely slow walking in cardiac patients: implications for exercise testing and training. Arch Phys Med Rehabil 1983;64:564–5. [19] Margaria R. Positive and negative work performances and their efficiencies in human locomotion. Int Z Angew Physiol Einschl Arbeitsphysiol 1968;25:339– 51. [20] Ralston HJ. Energy-speed relation and optimal speed during level walking. Int Z Angew Physiol 1958;17:277–83. [21] Waters RL, Lunsford BR, Perry J, Byrd R. Energy-speed relationship of walking: standard tables. J Orthop Res 1988;6:215–22. [22] Umberger BR, Martin PE. Mechanical power and efficiency of level walking with different stride rates. J Exp Biol 2007;210:3255–65. [23] Mercier J, Le Gallais D, Durrand M, Goudal C, Micallef J, Prefaut C. Energy expenditure and cardiorespiratory responses at the transition between walking and running. Eur J Appl Physiol 1994;69:525–9. [24] Hoyt DF, Taylor CR. Gait and the energetics of locomotion in horses. Nature 1981;292:239–40.
J.R. Priebe, R. Kram / Gait & Posture 34 (2011) 265–269 [25] Cavagna GA, Margaria R. Mechanics of walking. J Appl Physiol 1966;21:271–8. [26] Cavagna GA, Saibene F, Margaria R. External work in walking. J Appl Physiol 1963;18:1–9. [27] Fast A, Wang FS, Adrezin RS, Cordaro MA, Ramis J, Sosner J. The instrumented walker: usage patterns and forces. Arch Phys Med Rehabil 2005;76:484–91. [28] Nabizadeh SA, Hardee TB, Towler MA, Chen VT, Edlich RF. Technical considerations in the selection and performance of walkers. J Burn Care Rehabil 1993;14:182–8. [29] Youdas JW, Kotajarvi BJ, Padgett DJ, Kaufman KR. Partial weight-bearing gait using conventional assistive devices. Arch Phys Med Rehabil 2005;86:394–8.
269
[30] Wall JC, Charteris J. A kinematic study of long-term habituation to treadmill walking. Ergonomics 1981;24:531–42. [31] Brockway JM. Derivation of formulae used to calculate energy expenditure in man. Human Nutr Clin Nutr 1987;41:463–71. [32] Martin PE, Rothstein DE, Larish DD. Effects of age and physical activity status on the speed-aerobic demand relationship of walking. J Appl Physiol 1992;73:200–6. [33] Brooks GA, Fahey T, Baldwin K. Exercise physiology—human bioenergetics and its applications. McGraw-Hill; 2004.