Comparison of Spatiotemporal and Energy Cost of the Use of 3 Different Walkers and Unassisted Walking in Older Adults

Methods 

Participants 

Thirteen healthy, community-dwelling elders between the ages of 70 and 90 years who could walk unassisted agreed to participate in the study (table 1). Subjects were recruited through the subject registry from the University of Texas Medical Branch, Claude D. Pepper Older Americans Independence Center and through advertisements in the local newspaper. An investigator reviewed with each subject his/her medical history to rule out any major acute or chronic health condition that would affect their walking, such as a history of stroke with residual hemiparesis or severe osteoarthritis of the lower extremities. Seven subjects completed both the gait analysis comparisons and the oxygen consumption (V̇o2) comparisons. Three subjects completed only the gait analysis and another 3 completed only the V̇o2 comparisons; therefore, we had 10 subjects in the gait comparisons and 10 subjects in the V̇o2 comparisons. Technical problems with the portable gas analyzer prevented 6 subjects from completing the V̇o2 comparisons. Subjects provided written informed consent according to procedures approved by the institutional review board at the University of Texas Medical Branch in Galveston.

Table 1. Subject Characteristics

	Test	n	Age (y)	Sex (men/women)	Height (m)	Weight (kg)
	Gait parameters	10	73.9±3.9(65–78)	2/8	1.65±0.08	74.5±9.3
	V̇o2	10	74.1±3.7(65–79)	3/7	1.65±0.06	70.1±8.1

NOTE. Values are n, mean ± standard deviation (SD) (range), or mean ± SD.

WalkAbout Description 

The user is completely surrounded in a frame when using the Walkabout (Fig 1, Fig 2). One side opens to let a user enter and then close the device securely. The device moves by body contact as the user walks. The WalkAbout is available in 3 sizes, small, medium, and large, to accommodate people of different heights. The top rail is positioned approximately waist high to provide stability. The footprint of the base of the device is larger in circumference than the top rail, so that its legs are attached to the top rail at an outward angle toward the base to provide maximum stability. There are brakes on 2 wheels to provide stability while a user enters, exits, or sits. The WalkAbout is made of 14-gauge carbon steel and is designed to support full body weight. Its caster wheels let it roll easily over the ground. The device moves as the user walks as a result of the attachment of a safety seat to the WalkAbout (see Fig 1, Fig 2). Consequently, users do not need upper-body strength to walk and can rest their arms on the top rail, which is padded for comfort.

View Large Image Download to PowerPoint

Fig 1. Subject walking in the WalkAbout. Reprinted with permission of the American Congress of Rehabilitation Medicine and American Academy of Physical Medicine and Rehabilitation.10

View Large Image Download to PowerPoint

Fig 2. Subject seated in the WalkAbout. Reprinted with permission of the American Congress of Rehabilitation Medicine and American Academy of Physical Medicine and Rehabilitation.10

Procedures 

Gait analysis 

We used an instrumented walkway (GAITRite)b to measure temporal and spatial parameters of gait. To normalize the spatial parameters to a subject’s height, we measured his/her leg length from the greater trochanter to the lateral malleolus and entered that data into the system, along with the subject’s age, sex, height, and weight. The subject stood at the end of the 3.6-m (12-ft) walkway and then walked at his/her fastest, but safest, speed to the other end of the walkway. This test was repeated twice for each condition and the 2 trials were averaged for definitive data. Subjects performed the test under 4 conditions: fast walking without an assistive device, walking with a WalkAbout, walking with a standard, wheeled walker, and walking with a Merry Walker.c We selected the Merry Walker because it is a walker-chair combination that surrounds the user and is somewhat similar in design to the WalkAbout. The frame of the Merry Walker is made of 16-gauge steel that is bottom weighted to avoid tipping. The height of the wheeled walker was adjusted for each subject’s standing position, with the walker approximately 30cm in front of the subject. The walker height was adjusted so that one’s elbows were in approximately 30° of flexion when the walker was being held. The WalkAbout height was adjusted so that the top bar was approximately waist high for each subject. The Merry Walker is not adjustable but is available in small (1.45–1.55m [58–62in]), medium (1.58–1.67m [53–57in), and large sizes (1.70–1.80m [68–72in]). We used the appropriate size for each subject. Each condition was performed in random order to avoid a learning effect.

The primary spatial parameter measured by the GAITRite was stride length measured on the line of progression between the heel points of 2 consecutive footfalls of the same foot (fig 3).

View Large Image Download to PowerPoint

Fig 3. Spatial parameters of gait. Reprinted with permission of CIR Systems Inc.

The temporal parameters of gait measured were: (1) gait speed obtained after dividing the distance (measured on the horizontal axis from the heel point of the first footfall to the heel point of the last footfall) by the ambulation time (time elapsed between the first contacts of the first and the last footfalls), expressed in meters per second; (2) cadence, expressed as the number of steps per ambulation time (in steps/min); and (3) normalized mean gait speed, expressed as gait velocity divided by leg length (LL) (in m/s by LL/s).

We used stride length, gait speed, cadence, and normalized gait speed as our outcome variables for both unassisted gait and gait with the 3 assistive devices. The GAITRite has been reported to be valid and reliable in repeated testing with healthy subjects11, 12 and subjects using an assisted gait.13

Five-minute walk test 

Each subject was tested under the 4 previously described conditions while walking for 5 minutes in a rectangular hallway that was approximately (9.6×17.4m [32×58ft]). Each subject was familiarized with each assistive device before beginning the test. Subjects were told to rest their elbows on the upper bar of the WalkAbout, but to not bear weight on the upper extremities, while walking normally. For the Merry Walker, subjects were told to hold onto the front bar of the Merry Walker in order to move the device while walking, but to walk as normally as possible and not put weight on the front bar. For the standard wheeled walker, subjects were shown how to do a step-through gait while supporting some weight on the upper extremities when taking a step. The walker was rolled forward before the next step was taken. Subjects practiced using each device, then rested before beginning the test. They were instructed to walk as far as possible at their self-selected speed for 5 minutes. An investigator walked behind each subject, informing him/her of the elapsed time and measuring the distance with a measuring wheel.d

Subjects were fitted with a portable gas analyzer (V̇O2000)e that measures oxygen consumed and heart rate during functional activities. The V̇O2000 is a lightweight, portable device that has a mouthpiece for collecting expired air, sensors for analyzing oxygen and carbon dioxide content of expired air, a heart rate monitor, a battery pack, and a transmitter worn by the subject. A receiving unit receives and stores the transmitted data. These data are then downloaded into a computer to calculate values for V̇o2, minute ventilation, and respiratory exchange ratio (RER).

Measurements taken with portable gas analyzers are reported to be as valid as those taken with a standard metabolic cart, with a Pearson product-moment correlation of .90,13 and to demonstrate test-retest reliability during submaximal and maximal workloads with intraclass correlation coefficients ranging from .78 to .86.14 We tested the test-retest reliability of a portable gas analyzer on 40 healthy subjects in our lab while the subjects walked back and forth over a 5-m walkway. The device was reliable with a coefficient of .83. Although walking on ground over a short distance with frequent turns is slightly different than walking on a rectangular path, as in this procedure, it does demonstrate that the protocol is reliable. We selected a portable device rather than a standard metabolic cart to give the subject greater freedom of movement when using the assistive devices.

The subjects’ age, sex, height, and weight were entered into the system. The V̇O2000 was calibrated according to the manufacturer’s directions with gases of known content and volume before each subject was tested. A heart rate monitor was placed around the thorax. A mouthpiece was placed in the subject’s mouth to collect expired air, and a clip was placed on the nose to eliminate any breathing through the nose. The analyzer was placed in a small backpack worn by the subjects, who stood for several minutes to become familiar with breathing with the device before the test was started.

Subjects walked for 5 minutes (the 5-minute walk test [5MWT]) and were allowed to rest until they indicated they were ready to begin the next test. The rests generally lasted from 3 to 5 minutes. The mouthpiece and nose clip were removed between tests. All tests were conducted on the same day. The data were downloaded to the computer in order to calculate the V̇o2 per millimeter per body weight per minute (in mL·kg·−1min−1). Samples were taken every 20 seconds and averaged to obtain the oxygen consumed in each minute. The values obtained during minutes 4 and 5 of the task were the definitive data and were averaged. This ensured that a metabolic steady state was established in response to the activity. The distance walked during minute 4 and during minute 5 was recorded and used to calculate the V̇o2 per millimeter per kilogram body weight per meter walked (in mL·kg·−1m−1). This accommodated for the subjects’ differing walking speeds.

Results 

Ten subjects aged 65 years and older were tested for the spatial and temporal parameters of gait, and 10 subjects were tested with the portable oxygen analyzer while walking (see table 1). Seven subjects did both the gait and the 5MWTs. Three subjects completed the gait tests only and another 3 completed the 5MWT.

Table 2 shows the means and SDs for the gait tests. The results of the repeated-measures ANOVA for the gait parameters indicate there were significant differences between conditions for each of the dependent variables. Follow-up comparisons yielded significant differences when unassisted walking was compared with the Merry Walker for stride length (t9=7.04, P<.001), speed (t9=3.53, P<.006), and normalized gait speed (t9=3.62, P<.006), and when unassisted walking was compared with the WalkAbout for cadence (t9=−2.97, P<.016).

Table 2. Gait Parameters and 5MWT Under 4 Conditions

	Parameter	Unassisted Walking	Wheeled Walker	WalkAbout	Merry Walker
	Gait speed (m/s)	1.53±0.44	1.58±0.23	1.47±0.30	1.18±0.24⁎
	Normalized gait speed (m/s per LL/s)	1.85±0.50	1.91±0.33	1.79±0.36	1.43±1.29⁎
	Cadence (steps/min)	141.6±20.3	146.3±21.7	162.0±29.7⁎	159.2±27.9
	Average stride length (cm)	129.6±24.2	130.4±16.0	111.8±23.4	90.5±18.8⁎

NOTE. Values are mean ± SD.

⁎ P<.016.

Table 3 shows the means and SDs for the 5MWT. The results of the repeated-measures ANOVA for the 5-minute walk variables indicate there were significant differences between conditions for each of the dependent variables. Follow-up comparisons between the Merry Walker and unassisted walking yielded significantly lower values for distance walked (t9=−4.49, P<.002) and gait speed (t9=−3.93, P<.003), and higher values for RER (t9=−3.63, P<.005) and minute ventilation (t9=−4.07, P<.003) when using the Merry Walker. V̇o2 was higher when subjects used a wheeled walker and the Merry Walker, compared with unassisted walking.

Table 3. Five-Minute Walk Values Under 4 Conditions

	Parameter	Normal	Wheeled Walker	WalkAbout	Merry Walker
	Distance (m)	353.6±69.8	344.2±63.8	333.4±71.3	289.0±63.6⁎
	Gait speed (m/s)	1.18±0.23	1.15±0.21	1.11±0.24	.96±0.21⁎
	V̇o2 (mL·kg−1·m−1)	0.16±0.04	0.21±0.05⁎	0.18±0.04	0.22±0.05⁎
	RER/m	0.003±0.000	0.003±0.000	0.003±0.000	0.003±0.000⁎
	Minute ventilation (L·min−1·m−1)	0.35±0.09	0.41±0.07	0.39±0.09	0.44±0.13⁎

NOTE. Values are mean ± SD.

⁎ P<.016.

Discussion 

The spatial and temporal characteristics of elderly subjects while using the WalkAbout are comparable to unassisted, fastest walking except for faster cadence. The wheeled walker also had comparable gait values compared with unassisted gait. With the Merry Walker, subjects consistently had reduced gait speed and normalized gait speed and decreased stride lengths compared with unassisted walking or when walking with the other assistive devices. A recent comparison15 of gait when using a wheeled walker and unassisted gait in young adults performing a partial weight-bearing gait demonstrated significantly lower gait speed, cadence and step, and stride lengths with the wheeled walker. These results were accompanied by reduced vertical peak ground reaction forces for both the partial weight-bearing lower extremity and the other lower extremity, indicating that the additional force was placed on the upper extremities while walking with the wheeled walker. Our elderly subjects were encouraged to walk as normally as possible at their fastest pace while using the devices and they were able to accomplish this using the wheeled walker and the WalkAbout. Another study16 compared results with a standard walker, wheeled walker, and unassisted ambulation in elderly patients with Parkinson’s disease who had motor blocks or freezing while walking. Walking speed was significantly reduced for subjects using either assistive device compared with the unassisted condition. Our subjects, in contrast, did not have any neurologic conditions, and did not have a reduction in walking speed with 2 of the 3 devices. Several subjects noted that the seat of the Merry Walker limited their step lengths while walking, but this did not occur with either the wheeled walker or the WalkAbout.

It should be noted that we instructed our subjects to walk at their fastest, but safest, speed during this short gait test. Consequently, the gait speeds are slightly higher than those reported in a study of subjects in a similar age range.17 This allowed us to compare walking with and without the devices at a faster speed than the self-selected speed during the 5MWT in our group of healthy, older adults. This may limit the generalizability of these comparisons with clinical situations in which people often walk much slower, because the spatial (stride length) and temporal (cadence) measures we observed are speed dependent. As a follow-up to this study, we tested older people who displayed mobility disabilities and compared their walking with the usual assistive device and the WalkAbout. We found comparable gait speed, step length, and distance walked in 5 minutes when subjects were walking in the WalkAbout or the usual assistive device.

The results of the 5MWT were consistent with what we observed during the shorter walking task in that distance walked and gait speed were less when the subjects walked with the Merry Walker than when walking unassisted. Although these variables were reduced when the subjects walked with the wheeled walker and the WalkAbout, these differences were not significant. The V̇o2 values were higher than normal as subjects walked with the wheeled walker and the Merry Walker. Values for the RER and minute ventilation per meter walked were also higher with the Merry Walker. Previous research18 has reported that V̇o2 per meter walked while using a wheeled walker and non-weight-bearing ambulation with young adults increased the V̇o2 by 50%, but that older adults had no difference in V̇o2 during unassisted ambulation and ambulation with a wheeled walker.19 Our results suggested a 12.5% increase in V̇o2 per meter walked with the WalkAbout, a 31% increase with the wheeled walker, and a 37.5% increase for the Merry Walker. Higher cardiovascular responses to use of a standard walker and wheeled walker during non-weight-bearing gait have also been reported.20

Differences in our results compared with those of previous research were consistent with the fact that our subjects were bearing weight equally on both lower extremities during walking. These subjects were also healthy, older adults who did not use assistive devices. Higher V̇o2 with the assistive devices could be attributed to a subject’s lack of habitual use of these devices. The higher oxygen costs for the wheeled walker and the Merry Walker may reflect greater upper-extremity support when using these devices. We did not attempt to control the amount of upper-extremity support during walking with the devices; this was a limitation of our study.

Walkers also present other limitations. A wheeled walker is difficult to use on uneven surfaces such as walking outside on grass. Both the Merry Walker and the WalkAbout share this limitation. In addition, the Merry Walker and WalkAbout have relatively wide bases that limit their utility in confined spaces.

Conclusions 

With the new assistive device, the WalkAbout, users demonstrated spatial and temporal parameters of gait that are comparable to walking without an assistive device. Another assistive device comparable to the WalkAbout impeded normal gait. The oxygen demands of walking were similar to those in unassisted walking for the WalkAbout, but were higher for the wheeled walker and the Merry Walker. These results may help when prescribing assistive devices for older adults are prescribed.