The effect of walking aids on balance and weight-bearing patterns of patients with hemiparesis in various stance positions. (Research Report).

Standing balance of individuals with hemiparesis is often compromised and is frequently characterized by increased postural sway as well as by a shift in the average position of the center of pressure (COP) toward the sound limb. (1-3) Walking aids have been shown to increase postural stability of patients with various neurological disorders, (4-6) and they are frequently used to increase postural stability, to augment muscle action, and to reduce the load on the involved weight-bearing lower extremity. (7) A goal of rehabilitation for many people following a stroke is the restoration of walking ability. (8) Thus, walking aids such as standard and quad canes are frequently recommended to assist patients in achieving this goal. (9,10) It is often assumed that the quad cane, with its wider base of support, may provide greater stability than the standard cane. (11) Therefore, patients frequently use a quad cane initially and progress to a standard cane as their overall stability and the weight-bearing capacity of the involved lower extremity improve. Further recovery may enable the patient to discard the cane completely. (12)

According to one of the most frequently used approaches to neurological rehabilitation, (13) the introduction of a cane for support and gait training is detrimental to the development of a normal gait pattern. Proponents of this treatment approach contend that the use of a walking aid encourages the patient to shift more weight away from the affected limb toward the uninvolved extremity, thus promoting the development of an asymmetrical gait pattern. (14) Furthermore, because the patient is more likely to fall either toward the hemiplegic side or backward, a cane is not considered as having a positive effect on balance and ambulation safety. (14) In the choice between a standard cane and a quad cane, the strong recommendation of those who support this approach is never to recommend the use of a quad cane, because it is believed to increase the stance and gait asymmetry even further. Data to support these claims, however, are lacking.

Despite the frequent use of walking aids and the controversy concerning the benefits of using a cane, few studies have focused on the effects of different walking aids on postural control, weight-bearing patterns, and gait of patients with hemiplegia. Milczarek et al (15) examined postural sway and average position of COP of patients with hemiparesis who used a standard cane or a quad cane. The researchers concluded that although both types of canes lead to a reduction in the extent of postural sway and to a shift in the mean position of the COP toward the cane side, the quad cane appears to offer no advantage over a standard cane.

Measurements of balance and weight-bearing patterns during quiet stance, however, when the feet are placed in a symmetrical stance position, may not represent the effect of walking aids during walking. Yet, very little attention has been directed to the effects of canes on gait. A cane can increase stability during the single-limb support phase. (9) Thus, patients with hemiparesis walking with either a standard cane or a quad cane showed increases in stride period, stride length, and step length of the affected limb, as well as decreased cadence and step width, in comparison with those walking without a cane. (9) Data from another study (16) also indicate that the use of a walking aid and the type of walking aid do not affect the asymmetrical trunk movements characteristic of patients with hemiparesis. Furthermore, no differences have been found between the peak and integrated forces applied to either a standard cane or a quad cane during ambulation. (12) Therefore, peak force applied to the cane may be a function of the capability of the affected lower extremity to bear weight during single-limb support and this capability cannot be improved by changing cane type. (12) Although these studies may provide some insight as to the effect of walking aids on different gait variables of patients with hemiparesis, the differential effects of standard and quad canes on postural sway and on weight-bearing patterns remain unclear.

The purpose of my study was to examine and compare the effects of a standard cane and a quad cane on postural sway and weight distribution in stance positions that I believe challenge balance more than the previously tested symmetrical parallel stance and thus that may more closely simulate gait. Therefore, I examined postural sway and weight bearing on the lower extremities with different canes in 3 positions (Fig. 1): with the heels aligned with each other (aligned position) and in staggered stepping positions with either the involved lower extremity placed forward (affected FW position) or with the uninvolved extremity placed forward (unaffected FW position). The objectives of the study were: (1) to compare postural sway and weight-bearing patterns of patients with hemiparesis and age-matched adults without hemiparesis in 3 stance positions, (2) to compare the effects of a standard cane and a quad cane on postural sway and weight-bearing patterns in each of the 3 positions, and (3) to compare postural sway and weight-bearing patterns among positions.

[FIGURE 1 OMITTED]

Three hypotheses concerning the effect of stance position and assistive device on postural sway and 4 hypotheses concerning their effect on weight distribution were postulated:

1. Subjects with hemiparesis (patient group) will demonstrate increased postural sway in comparison with an age-matched group of subjects without hemiparesis (comparison group) in all 3 positions when using corresponding support.

2. The quad cane will have a greater effect, decreasing postural sway more than the standard cane.

3. The position that forces more weight on the affected extremity (unaffected FW position) will demonstrate the highest postural sway.

4. Aligned stance will be characterized by an asymmetrical weight distribution in the patient group only.

5. The use of canes will reduce weight borne by both the affected and unaffected lower extremities.

6. Both staggered foot positions will induce a weight shift to the extremity placed behind, with the shift being more pronounced in the comparison group and least pronounced in the patient group in the unaffected FW position.

7. More weight will be placed on the quad cane as compared with the standard cane in both groups.

Method

Subjects

Subjects with a diagnosis of unilateral hemiparesis following a stroke were recruited by means of convenience sampling from neurological inpatient and outpatient departments of a geriatric rehabilitation center. Included were patients who met the following criteria: (1) had a first stroke in anterior brain circulation, as evidenced by computed tomography or by clinical symptoms (17); (2) were independent in ambulation prior to stroke; (3) were able to stand independently without external support for at least 30 seconds with feet side by side; (4) were able to use a walking aid for indoor ambulation independently or with assistance; (5) were unable to walk independently outdoors; (6) had no medical contraindication to standing and exercising; (7) had no orthopedic or any other neurological condition that could limit standing ability; and (8) had no severe cognitive or communication impairment that could hamper the understanding of simple instructions.

Of the 42 patients who met these criteria, 7 were disqualified because they could not maintain the 2 asymmetrical stance positions. Five additional candidates were disqualified because even though they were able to stand independently with feet side by side as well as with the affected foot placed forward, they were unable to maintain independent stance with the unaffected foot placed forward. Therefore, a final sample of 12 female and 18 male subjects with a diagnosis of unilateral hemiparesis following a stroke (29 infarcts and 1 hemorrhage) participated in the study. The mean age of the subjects was 71.2 years (SD=7.0, range=59-89). Twelve subjects had left hemiparesis, and 18 subjects had right hemiparesis. Twenty-four participants were inpatients in the initial stages of rehabilitation (3 months or less since neurological insult), and 6 were outpatients (over 3 months since neurological insult). The mean number of days between neurological insult and testing was 77.7 (SD=61.7, range=18-210). Assistive devices used by the subjects, as determined by the treating therapist, were: standard canes (2 subjects); quad canes (24 subjects); and walkers (4 subjects). The subjects' mean Barthel Index (18) score was 73.7 (SD=12.5, range=50-100), and their mean Functional Ambulation Category (19) score was 3.4 (SD=0.9, range=2-5). Twenty (10 female, 10 male) age-matched, community-dwelling volunteers (mean age=72.1 years, SD-5.2, range=65-81) were recruited from among the visitors of the patients at the hospital and constituted the comparison group. All subjects gave informed consent to participate in the study.

Procedure

The study involved testing postural sway and weight distribution on the left and right lower extremities. The posturographic examinations were carried out with the Tetrax Portable Multiple System. * According to the manufacturer of the system, it includes 2 paired portable forceplates (12x30 cm each), each equipped with a separate pressure transducer (one for each heel and one for each forefoot). Vertical pressure fluctuations are sampled at a rate of 34 Hz with a low-pass filter at 1.5 Hz installed to control for nonphysiological vibrations. The Sway Index (SI), according to the manufacturer of the Tetrax system, was calculated as the resultant root mean square (RMS) of the COP amplitude normalized to body weight. The RMS represents the excursion of the path produced by the oscillations of the COP and has been used by several researchers (20) as a measure of postural steadiness. The SI has been shown to have high test-retest reliability (21) and to be highly correlated with the equilibrium score of the EquiTest system, ([dagger]) which is a well-established posturography system. (22) Sway Index validity, as a measure of postural steadiness, has been established in a variety of diagnoses such as diabetic neuropathy and work fatigue. (23,24) According to the manufacturer of the system, it also provides data on mean percentage of body weight (%BW) on each lower extremity. Mean %BW on an assistive device was extrapolated by calculating the differences between total body weight with and without support.

Tests were carried out in the following 3 positions (Fig. 1): (1) aligned position, in which the forceplates were placed at a 30-degree angle from each other, with the heel-end of the forceplates aligned with each other and separated by 3 cm; (2) affected FW position, in which the affected foot was placed forward and the unaffected foot was placed behind, with the position of forceplates such that the heel of the affected lower extremity was level with the toes of the unaffected lower extremity, and the horizontal distance between the forceplates was 3 cm; and (3) unaffected FW position, in which the previous position was reversed with the unaffected lower extremity placed forward and the affected lower extremity placed behind.

Three trials were carried out in each position with the subjects using: (1) no external support (no cane), (2) a standard cane, and (3) a quad cane. Cane placement was determined by the subjects' preference. Testing time was 30 seconds for each trial. The order of positions was determined by block design, and the order of trials in each position was randomized. Because I believe the quality and support offered by the subjects' shoes varied considerably among the subjects, I decided to conduct all trials with subjects barefooted. Subjects were instructed not to move their feet until data collection for all 3 trials in one position was completed. All subjects were allowed a 4- to 5-minute rest period between positions.

The standard cane had a rubber tip measuring 3 cm in diameter. The quad cane had a trapezoid-shaped base of support, with the long anterior-posterior distance between rubber tips 32 cm, the short anterior-posterior distance 28 cm, and the medial-lateral distance 20 cm (area: 600 [cm.sup.2]). The rubber tips of the quad cane were 2 cm in diameter. Heights of canes were adjusted in length so that elbow was in 15 to 30 degrees of flexion when the subject stood on the balance plates and the canes were placed on the floor at the position of each subject's preference. (7) The subjects always used the assistive devices with the hand of the unaffected side; half of the comparison subjects (n=10) used them with their right hand, and half (n=10) used them with their left hand. For purposes of analysis, the side on which the assistive device was used was designated as "unaffected," and the opposite side was designated as "affected" in the comparison group as well.

Data Analysis

Four dependent variables were considered for analysis: (1) SI, with higher scores indicating more sway; (2) %BW borne by the unaffected lower extremity (%BW unaffected); (3) %BW borne by the affected lower extremity (%BW affected); and (4) %BW carried by the standard or quad cane (%BW support). A logarithmic transformation was applied to the SI, and an arcsine transformation was used for all proportion-based (percentage) analyses to meet normal distribution requirements for parametric statistical analysis. (25) The analysis consisted of a mixed-model analysis of variance (ANOVA), with specific individual condition comparisons based on adjusted Tukey-Kramer tests. Independent factors consisted of subjects (a random factor: nested within subject group); position (a fixed factor: aligned, affected FW, or unaffected FW); subject group (a fixed factor: patient or comparison); and type of support (a fixed factor: none, standard cane, or quad cane), as well as the interaction of group, position, and support. Some of the results concerning only the aligned position have been published previously. (26) Statistical analysis was performed using the SAS version 6.09 software program. ([double dagger]) Differences were regarded as significant at the P<.05 level.

Results

A summary of the results of the adjusted Tukey-Kramer tests comparing the effects of support in each position by group is presented in Table 1. A summary of the results of the adjusted Tukey-Kramer tests comparing the effects of stance position with same support by group is presented in Table 2.

Postural Sway

The mean and standard deviation of postural sway in each position by support and by group are presented in Figure 2. The ANOVA indicated an effect of group, support, and position and an interaction effect only between support and position (P<.0001).

[FIGURE 2 OMITTED]

Comparison between patient and comparison groups. Absence of an interaction effect between group and support and between group and position indicates that the postural sway of the patient group was invariably higher than that of the control group when the subjects were placed in the same position with the same assistive device (Fig. 2).

Comparison among effects of different assistive devices. As shown in Table 1, the use of support in each position had a parallel effect on postural sway in both groups. Comparison between stance with no cane and stance with a support indicated that the standard cane reduced postural sway only in the affected FW position, whereas the quad cane reduced sway in all 3 positions. Comparison between the standard cane and the quad cane indicated that the quad cane had a greater effect on reducing postural sway in both staggered positions.

Comparison among positions with same support. As shown in Table 2, subjects of both groups standing with no support were most stable in the aligned position. However, contrary to study hypothesis 3, while standing with no support, both subject groups were equally unstable in the affected FW and unaffected FW positions. When provided with a standard cane, there was a difference between the aligned and affected FW positions only in the comparison group, with the postural sway being greater in the affected FW position. The quad cane decreased postural sway in the affected FW position to a level equal to that of the aligned position in both groups. Yet, postural sway remained greater in the unaffected FW position in comparison with either the aligned position or the affected FW position with both the standard and quad canes in both groups.

Weight Distribution Between Lower Extremities

Mean %BW on each limb ([+ or -] SD) in each position by support and by group is presented in Figures 3 and 4. The ANOVA of %BW on the affected lower extremity indicated an effect of group and of stance position, as well as an interaction effect between group and stance position. The ANOVA of %BW on the unaffected lower extremity indicated an effect of group, support, and stance position, with an interaction effect between group and stance position.

[FIGURES 3-4 OMITTED]

Comparison between patient and comparison groups. As hypothesized, in the aligned position, the comparison subjects' weight while standing with no cane was evenly distributed, whereas the patients' mean %BW on the affected extremity in this position was lower (37.5% [+ or -] 8.8%). Weight on the patients' affected extremity remained lower than that of the comparison subjects in all positions regardless of support. Weight on the patients' unaffected extremity remained higher than that of the comparison subjects in all positions regardless of support, with the exception of the affected FW position with a standard cane and a quad cane where there were no differences between groups.

Comparison among effects of different assistive devices. As shown in Table 1, contrary to research hypothesis 5, use of an assistive device did not change weight-bearing patterns in either group, with only one exception in the patient group. In the affected FW position, use of the quad cane by the patients reduced weight bearing on the unaffected extremity from a mean of 64.8% [+ or -] 9.8% with no cane to 58.4% [+ or -] 11.4% with the quad cane.

Comparison among positions with same support. As shown in Table 2, changing position influenced weight distribution between extremities only in the comparison group, with unexpectedly no similar effect in the patient group. Although there were no differences in weight-bearing patterns in the comparison group between the aligned and affected FW positions regardless of support, the shift to the unaffected FW position created differences. There was, however, no affected extremity in this group. Yet, the unaffected FW position, where the assistive device was held close to the foot placed forward, induced a decrease in %BW on the "unaffected" foot placed forward (eg, with standard cane, %BW was reduced to 39.4% [+ or -] 11.1%) and an increase in %BW on "affected" foot placed behind (eg, with standard cane, %BW increased to 56.9% [+ or -] 10.0%). This pattern held true regardless of the type of assistive device used. However, with no cane, weight-bearing patterns were the same between the aligned position and both staggered positions.

Weight on Support

Means and standard deviations of %BW on support in each position by group are presented in Table 3. The ANOVA of %BW on support indicated an effect of support and stance position, as well as an interaction between group and stance position.

Comparison between patient and comparison groups. Mean %BW with an assistive device ranged between 3.6% [+ or -] 2.2% (unaffected FW position with standard cane) and 6.4% [+ or -] 3.4% (affected FW position with quad cane) in the patient group and between 3.1% [+ or -] 2.3% (aligned position with standard cane) and 4.3% [+ or -] 2.8% (affected FW position with quad cane) in the comparison group. There were no differences in mean %BW between groups.

Analysis of effect of support within each position. Table 1 shows that, for the comparison group, there was no difference in the %BW placed on the standard and quad canes in each of the positions. This was true only in the aligned position in the patient group. In both the affected FW and unaffected FW positions in the patient group, the %BW was higher on the quad cane than on the standard cane.

Comparison among positions with same support. Generally, as shown in Table 2, there was no difference in the %BW on support among positions. However, 2 exceptions were found in the patient group, where %BW on the quad cane was higher in the affected FW position both in comparison with that of the aligned position and in comparison with that of the unaffected FW position.

Discussion

Postural sway is primarily a measure of postural steadiness encountered while maintaining balance and has been shown to correlate with clinical measurements of balance in elderly people and in subjects with hemiparesis. (20,27) Although the patients with hemiparesis in my study demonstrated greater postural sway than the age-matched comparison subjects, external support in the form of standard and quad canes decreased postural sway in both subject groups. The lack of interaction between group and cane, as well as between group and stance position, indicates that the effect of standard and quad canes on postural sway followed a parallel pattern in both groups.

The results support the hypothesis that the quad cane is advantageous in comparison with the standard cane in reducing postural sway. In both staggered positions, the quad cane reduced postural sway more than the standard cane. Furthermore, in both the aligned and unaffected FW positions used in my study, only the quad cane had an effect on postural sway. These findings seem to conflict with those of a previous report, (15) which indicated that although both types of canes lead to a reduction in the extent of postural sway in the aligned position, the quad cane appears to offer no advantage over a standard cane. Only the results obtained in the affected FW position support the finding that the standard cane has any effect on postural sway when compared with stance with no cane. However, the consistently greater effect of the quad cane on postural sway measured in all 3 positions in both subject groups provides strong support for the conclusion that the quad cane increases postural steadiness more than a standard cane.

In my study, I could not determine the mechanism that enables the quad cane to reduce postural sway more than a standard cane. Although, as hypothesized, the %BW placed on the quad cane was higher than that placed on the standard cane in both staggered foot positions, the greatest differences in %BW between the 2 types of canes were on the average less than 1.8% of body weight (patients in the unaffected FW position). Such a difference seems an unlikely sole source for the differences in postural sway.

The size, weight, and stability of the quad cane may provide for greater psychological confidence than the standard cane, which has been shown to favorably influence posture control. (28) It also has been claimed that the quad cane offers more stability by providing a greater medial-lateral base of support than the standard cane. (11)

Another possibility relates to the inherent stability of the quad cane. Research suggests that when a cane is held still relative to the body, the somatosensory input obtained from the hand, arm, or fingertips concerning body sway in reference to the still cane may assist in the interpretation of body sway and invoke the muscular responses that inhibit further sway. (29) The wider base of support of the quad cane and its inherent stability may provide a more stable reference to body sway than the standard cane, thereby enhancing postural adaptation.

The base of support was narrower in the 2 "mid-stance" positions than in the aligned position; therefore, the increased postural sway observed in these positions in both subject groups is not surprising. Because both feet of the comparison subjects were not involved due to any central nervous system lesions, their postural sway in stance with no external support was, as expected, unaffected by which foot was placed forward. I found it somewhat surprising that, in the patient group, stance with no support was as unstable with the involved foot placed forward as in the reversed foot placement. As hypothesized, placing one foot forward induced in the comparison group a slight shift from symmetrical weight distribution between both feet. Thus, although the %BW of the comparison group standing in the aligned position with no external support was equally distributed between both feet, in the unaffected FW position, the %BW was lower on the foot placed forward than on the foot placed behind. With a similar shift in weight distribution in the patients with hemiparesis, stance that forces more weight on the affected limb (ie, the unaffected FW position) might be expected to be less stable than the stance in which more weight is carried by the unaffected limb (ie, the affected FW position).

Postural sway was not altered, however, when the patients standing with no external support moved from the affected FW position to the unaffected FW position. This finding may have been because, in this group, standing in a staggered foot position did not induce a shift in weight distribution between the lower extremities parallel to the shift induced in the comparison group. The %BW on the affected and unaffected extremities in the aligned position with no external support was 37.5% [+ or -] 8.8% and 62.5% [+ or -] 28.8%, respectively, and this asymmetrical weight distribution did not change in the patient group across positions even with external support. Thus, it seems as if the weight borne on the affected extremity in the aligned position is actually the maximally tolerated weight that this limb can comfortably support. When put in a position that would naturally induce more weight bearing on the affected limb (eg, the unaffected FW position), the patients still shifted their weight to the unaffected extremity placed forward and were thereby able to avoid further compromise to postural steadiness. This finding, is consistent with those of a previous study (30) in which patients tended to limit weight shifting to the affected limb even when their sound extremity was placed on a 17-cm high step.

The equal postural sway and weight distribution in the 2 staggered foot positions may have resulted from an inadvertent bias in subject selection. Inclusion was limited to patients who were able to stand independently for 30 seconds in each of the positions. During the screening process, 5 individuals who were able to maintain the aligned and affected FW positions were excluded from the study because they were unable to maintain stance in the unaffected FW position. At the same time, all of the subjects who able to maintain the unaffected FW position also were able to maintain the aligned and affected FW positions. Although excluding these subjects was unavoidable, this may have led to biased results. It could well be that for individuals who are more disabled, postural sway is greatest in the unaffected FW position where more weight is forced on the affected limb and that only the patients with greater stability who were able to maintain their stance in the unaffected FW position by compensating and shifting their weight forward to the unaffected limb were included in this study.

Although the postural sway of both groups did not differ between the 2 staggered foot positions when no external support was used, a different pattern emerged with the use of a walking aid. As postulated, postural sway was greater in the unaffected FW position than in the affected FW position for both groups when subjects used the same walking aid. Moreover, use of the quad cane in the affected FW position even decreased sway to the level of the aligned position with the same support. However, because contrary to the research hypothesis the unaffected FW position did not induce in the patient group a weight shift toward the affected extremity, the reason for the greater postural sway in this position must be sought elsewhere. It has been shown that the use of a cane contralateral to the affected side reduces the total forces across the hip joint. (31) Because the cane was always held on the unaffected side, the decreased postural sway in the affected FW position can be attributed to the greater base of support obtained in this position, in contrast to the base of support obtained in the unaffected FW position (Fig. 1). Thus, the findings support the use of a cane in the hand contralateral to the more affected lower extremity, which simulates a reciprocal gait pattern. (32)

Although there were differences between the weight placed on the straight cane and the weight placed on the quad cane in both the unaffected FW and affected FW positions, the hypothesized reductions in weight bearing from both extremities were not observed. Thus, although weight on the support will shift the center of pressure toward the cane side, (15) this shift in center of pressure does not necessarily induce a reduction in weight bearing on the affected limb, because higher reductions were observed in both subject groups in the unaffected extremity that was closer to the cane. This finding is important because it contradicts the claim made by some authors (14) that the use of a cane encourages a weight shift away from the affected limb and thus increases stance asymmetry.

Given that in my study I did not simulate single-limb stance during gait, it is possible that during ambulation more weight would be borne by the cane, leading to a reduction in weight bearing on the involved extremity as well. The average peak vertical cane force applied by subjects with stroke during gait has been reported to be in the range of 12.7% to 16% of body weight. (10) Although these forces are higher than those obtained in my study, they are much lower than those reported for people with orthopedic impairments. (33) The low %BW on the support suggests to me that the primary effect of the cane on postural sway even during a staggered foot position where postural control is challenged by a narrow base of support is achieved through sensory input or by improved confidence, rather than through physical support.

I studied a convenience sample of people with moderate impairment secondary to hemiparesis in the early stages of rehabilitation, and no data are available concerning the patients who did not meet the study's inclusion criteria. Generalization, therefore, to patients with more severe impairment is limited. People with severe deficits in postural control may use a walking aid more for weight bearing than for postural control and will thus demonstrate a different pattern of weight distribution between the lower extremities. Furthermore, although tests in staggered foot positions require greater postural control than the standard aligned position, they do not replicate the dynamic and rhythmic weight shifting that must occur during gait. Further research is needed to determine the effect of various canes on postural stability and weight bearing during ambulation.

Conclusion

The results indicate that a quad cane is more effective than a standard cane in reducing postural sway of patients with hemiparesis. Postural sway increases as the base of support decreases, and the effect of a walking aid on postural sway is greatest when the aid is held contralateral to the foot placed forward, thus simulating a reciprocal gait pattern. The results also suggest that the transfer of body weight to the cane during stance does not increase the asymmetrical weight distribution between the lower extremities even in staggered foot positions, because the majority of weight is transferred to the cane from the unaffected extremity. Although these results indicate differential effects of various canes during gait as well, further research is needed to determine the short- and long-term effects of various canes on postural control and weight distribution during gait of patients following stroke.

Table 1.
Results of Adjusted Tukey-Kramer Tests Comparing the Effects of
Support in Each Position (a) by Group

Comparison                      Patients Group (n=30)
Among Types
of Support                      Aligned   Affected FW   Unaffected FW

Sway Index
  No cane vs standard cane      NS (b)    <.0001        NS
  No cane vs quad cane          .013      <.0001        <.0001
  Standard cane vs quad cane    NS        <.0001         .007

% on affected extremity
  No cane vs standard cane      NS        NS            NS
  No cane vs quad cane          NS        NS            NS
  Standard cane vs quad cane    NS        NS            NS

% on unaffected extremity
  No cane vs standard cane      NS        NS            NS
  No cane vs quad cane          NS         .044         NS
  Standard cane vs quad cane    NS        NS            NS

% on support
  Standard cane vs quad cane    NS         .034          .002

Comparison                      Comparison Group (n=20)
Among Types
of Support                      Aligned   Affected FW   Unaffected FW

Sway Index
  No cane vs standard cane      NS         .004         NS
  No cane vs quad cane          .014      <.0001        .038
  Standard cane vs quad cane    NS        <.0001        .001

% on affected extremity
  No cane vs standard cane      NS        NS            NS
  No cane vs quad cane          NS        NS            NS
  Standard cane vs quad cane    NS        NS            NS

% on unaffected extremity
  No cane vs standard cane      NS        NS            NS
  No cane vs quad cane          NS        NS            NS
  Standard cane vs quad cane    NS        NS            NS

% on support
  Standard cane vs quad cane    NS        NS            NS

(a) Positions: with the heels aligned with each other (aligned
position) and in staggered foot positions with either the affected
or unaffected extremity placed forward (affected FW and unaffected
FW positions).

(b) NS = not significant.

Table 2.
Results of Adjusted Tukey-Kramer Tests Comparing the Effects of
Stance Position (a) With Same Support by Group

                                 Patient Group (n=30)

Comparison Among Positions       No Cane  Standard Cane  Quad Cane

Sway Index
  Aligned vs affected FW         <.001     .056          NS (b)
  Aligned vs unaffected FW       <.0001   <.0001         <.0001
  Affected FW vs unaffected FW   NS       <.0001         <.0001

% on affected extremity
  Aligned vs affected FW         NS       NS             NS
  Aligned vs unaffected FW       NS       NS             NS
  Affected FW vs unaffected FW   NS       NS             NS

% on unaffected extremity
  Aligned vs affected FW         NS       NS             NS
  Aligned vs unaffected FW       NS       NS             NS
  Affected FW vs unaffected FW   NS       NS             NS

% on support
  Aligned vs affected FW         --       NS             0.021
  Aligned vs unaffected FW       --       NS             NS
  Affected FW vs unaffected FW   --       .023           NS

                                 Comparison Group (n=20)

Comparison Among Positions       No Cane  Standard Cane  Quad Cane

Sway Index
  Aligned vs affected FW         <.0001    .003          NS
  Aligned vs unaffected FW       <.0001   <.0001         <.0001
  Affected FW vs unaffected FW   NS        .002          <.0001

% on affected extremity
  Aligned vs affected FW         NS       NS             NS
  Aligned vs unaffected FW       NS        .050           .045
  Affected FW vs unaffected FW   <.0001   <.0001         <.0001

% on unaffected extremity
  Aligned vs affected FW         NS       NS             NS
  Aligned vs unaffected FW       NS        .035           .052
  Affected FW vs unaffected FW    .002    <.0001         <.0001

% on support
  Aligned vs affected FW         --       NS             NS
  Aligned vs unaffected FW       --       NS             NS
  Affected FW vs unaffected FW   --       NS             NS

(a) Positions: with the heels aligned with each other (aligned
position) and in staggered foot positions with either the affected
or unaffected extremity placed forward (affected FW and unaffected
FW positions).

(b) NS=not significant.

Table 3.
Mean ([+ or -] SD) of Percentage of Body Weight on Support in Each
Position (a) by Group

                Patient Group (n=30)

Support         Aligned            Affected FW        Unaffected FW

Standard cane   4.0 [+ or -] 2.3   4.9 [+ or -] 2.6   3.6 [+ or -] 2.2
Quad cane       5.0 [+ or -] 2.8   6.4 [+ or -] 3.4   5.4 [+ or -] 2.9

                Comparison Group (n=20)

Support         Aligned            Affected FW        Unaffected FW

Standard cane   3.1 [+ or -] 2.3   3.9 [+ or -] 2.9   3.8 [+ or -] 2.1
Quad cane       3.9 [+ or -] 2.5   4.3 [+ or -] 2.8   4.2 [+ or -] 2.1

(a) Positions: with the heels aligned with each other (aligned
position) and in staggered foot positions with either the affected or
unaffected extremity placed forward (affected FW and unaffected FW
positions).

* Tetrax Ltd, 56 Miryam St, Ramat Gan, Israel.

([dagger]) NeuroCom International, 9570 SE Lawnfield Rd, Clackamas, OR 97015.

([double dagger]) SAS Institute Inc, PO Box 8000, Cary, NC 27511.

References

(1) Dettman M, Linder M, Sepic S. Relationships among walking performance, postural stability, and functional assessments of the hemiplegic patient. Am J Phys Med. 1987;66:77-90.

(2) Bernhardt J, Ellis P, Denisenko S. Changes in balance and locomotion measures during rehabilitation following stroke. Physiother Res Int. 1998;3:109-122.

(3) Pai YC, Rogers MW, Hedman LD, Hanke T. Alterations in weight transfer capabilities in adults with hemiparesis. Phys Ther. 1994;74: 647-657.

(4) Lu CL, Yu B, Basford JR, et al. Influences of cane length on the stability of stroke patients. J Rehabil Res Dev. 1997;34:91-100.

(5) Ashton-Miller JA, Yeh MW, Ricahrdson JK, Galloway T. A cane reduces loss of balance in patients with peripheral neuropathy: results from a challenging unipedal balance test. Arch Phys Med Rehabil. 1996;77:446-452.

(6) Nandapalan V, Smith CA, Jones AS, Lesser TH. Objective measurement of the benefit of walking sticks in peripheral vestibular balance disorders, using the sway weight balance platform. J Laryngol Otol. 1995;109:836-840.

(7) Joyce BM, Kirby RL. Canes, crutches and walkers. Am Fam Physician. 1991;43:535-542.

(8) Bohannon RW, Andrews AW, Smith MB. Rehabilitation goals of patients with hemiplegia. Int J Rehabil Res. 1988;11:181-183.

(9) Kuan TS, Tsou JY, Su FC. Hemiplegic gait of stroke patients: the effect of using a cane. Arch Phys Med Rehabil. 1999;80:777-784.

(10) Chen CL, Chen HC, Wong MK, et al. Temporal stride and force analysis of cane-assisted gait in people with hemiplegic stroke. Arch Phys Med Rehabil. 2001;82:43-48.

(11) Olsson EC, Smidt GL. Assistive devices. Clin Phys Ther. 1990;23: 141-157.

(12) Dickstein R, Abulaffio N, Pillar T. Vertical force loaded on walking canes in hemiparetic patients. Gait Posture. 1993;1:113-118.

(13) Sackley C, Lincoln NB. Physiotherapy treatment for stroke patients: a survey of current practice. Physiotherapy Theory and Practice. 1996;12: 87-96.

(14) Davies PM. Steps to Follow: The Comprehensive Treatment of Patients With Hemiplegia. 2nd ed. Berlin, Germany: Springer; 2000:274-275.

(15) Milczarek JJ, Kirby RL, Harrison ER, MacLeod DA. Standard and four-footed canes: their effect on the standing balance of patients with hemiparesis. Arch Phys Med Rehabil. 1993;74:281-285.

(16) Tyson SF. Trunk kinematics in hemiplegic gait and the effect of walking aids. Clin Rehabil. 1999;13:295-300.

(17) Bamford J, Sundercock P, Dennis M, et al. Classification and natural history of clinically identifiable subtypes of cerebral infarction. Lancet. 1991;337:1521-1526.

(18) Wade DT, Collin C. The Barthel ADL index: a standard measure of physical disability? Int Disabil Studies. 1988;10:64-67.

(19) Holden MK, Gill KM, Magliozzi MR, et al. Clinical gait assessment in the neurologically impaired: reliability and meaningfulness. Phys Ther. 1984;64:35-40.

(20) Berg KO, Maki BE, Williams U, et al. Clinical and laboratory measures of postural balance in elderly population. Arch Phys Med Rehabil. 1992;73:1073-1080.

(21) Kohen-Raz R. Application of tetra-ataziametric posturography in clinical and developmental diagnosis. Percept Mot Skills. 1991;73: 635-656.

(22) Turner DM. Evaluation of the Tetras Interactive Balance System and Equitest Using Normal Subjects [Thesis]. Southampton, United Kingdom: University of Southampton; 1998.

(23) Kohen-Raz R, Himmelfarb M, Tsur S, et al. An initial evaluation of work fatigue and circadian changes as assessed by multiplate posturography. Percept Mot Skills. 1996;82:547-557.

(24) Oppenheim U, Kohen-Raz A, Daitz A, Azarya M. Postural characteristics of diabetic neuropathy. Diabetes Care. 1999;22:328-332.

(25) Zar JH. Biostatistical Analysis. 2nd ed. Englewood Cliffs, NJ: Prentice-Hall Inc; 1984.

(26) Laufer Y. Effect of 1-point and 4-point canes on balance and weight distribution in patients with hemiparesis. Clin Rehabil. 2002;16: 141-148.

(27) Stevenson TJ, Garland SJ. Standing balance during internally produced perturbations in subjects with hemiplegia: validation of the balance scale. Arch Phys Med Rehabil. 1996;77:656-662.

(28) Alexander NB. Postural control in older adults. J Am Geriatr Soc. 1994;42:93-108.

(29) Jeka JJ. Light touch contact as a balance aid. Phys Ther. 1997;77: 476-487.

(30) Laufer Y, Dickstein R, Resnik S, Marcovitz E. Weight-bearing shifts of hemiparetic and healthy adults upon stepping on stairs of various heights. Clin Rehabil. 2000;14:125-129.

(31) Norkin CC, Levangie PK. Joint Structure and Function: A Comprehensive Analysis. 2nd ed. Philadelphia, Pa: FA Davis Co; 1992.

(32) O'Sullivan SB, Schmitz TJ. Physical Rehabilitation: Assessment and Treatment. Philadelphia, Pa: FA Davis Co; 1994.

(33) Engel J, Amir A, Messer, Caspi I. Walking cane designed to assist partial weight bearing. Arch Phys Med Rehabil. 1983;64:386-388.

Y Laufer, PT, PhD, is Head, Physical Therapy Department, Faculty of Social Welfare and Health, University of Haifa, Mount Carmel, Israel 31905 ([e-mail]). At the time of data collection, she was also on the staff of the Physical Therapy Department, Flieman Geriatric Rehabilitation Hospital, Ramot Remez, Haifa, Israel.

This study was approved by the Institutional Review Board of Flieman Geriatric Rehabilitation Hospital.

This article was submitted October 24, 2001, and was accepted August 21, 2002.