Perceptual and Motor Skills, 1991, 72, 935-941. O Perceptual and Motor Skills 1991
ACCURACY O F WEIGHTBEARING ESTIMATION BY STROKE VERSUS HEALTHY SUBJECTS ' RICHARD W. BOHANNON Uniuersity of Connecticut Harrford (C7J Hospital
AND
DIANE TINTI-WALD Hartford (CT)Hospital
Summary.-This study was performed to describe and compare the accuracy of weightbearing at three target levels (25, 50, and 75% of body weight) of 14 ambulatory stroke subjects and 14 matched healthy subjects. Weightbeating through a designated lower extremity (stroke subjects-paretic, healthy subjects-randomly selected) was measured with digital scales. No significant difference was found in magnitude of weightbeating (%) between the stroke and healthy groups. Although making significantly greater errors in weightbearing than healthy subjects, stroke subjects did not consistently weightbear under target. Subjects with stroke tested in earlier studies were neither asked to stand symmetrically nor examined to judge whether they were capable of accepting the required weight through the paretic lower extremity. Before assuming that weightbearing asymmetry is a problem in patients with stroke, clinicians should examine weightbearing behavior more specifically.
Patients who have experienced cerebrovascular accidents (strokes) have a known tendency to weightbear asymmetrically during bilateral upright stance. More specifically, when asked to stand comfortably or straight the patients tend to place more weight through their nonparetic than through their paretic lower extremities (Bohannon & Larkin, 1985; Caldwell, MacDonald, MacNeil, McFarland, Turnball, & Wall, 1986; Dickstein, Nissan, Pillar, & Scheer, 1985; Mizrahi, Solzi, Ring, & Nisell, 1989). A reasonable, albeit untested explanation of the patients' weightbearing asymmetry is that they have an impaired ability to judge weightbearing in stance. The explanation is reasonable given what is already known about stroke patients' tendency to misjudge other sensory inputs, e.g., vertical and horizontal (Birch, Proctor, Bortner, & Lowenthal, 1960) and heaviness (Gandevia & McCloskey, 1977). A recent study of healthy individuals outlines a method by which the ability of stroke patients to judge weightbearing can be tested (Bohannon, Waters, & Cooper, 1989). That method requires subjects to adjust weightbearing through a designated lower extremity to a target level. The purpose of this study was to describe and compare the accuracy of weightbearing at three target levels in a small sample of stroke subjects and a matched sample of healthy subjects. Subjects
All subjects were volunteers who participated with informed consent. 'Address correspondence to Dr. R. W.Bohannon, School of Allied Health, U-101, University of Connecticut, Storrs, CT 06269-2101.
936
R. W. BOHANNON & D. TIFJTI-WALD
Fourteen subjects had experienced a cerebrovascular accident (stroke) prior to admission to the acute care hospital where the research was performed. The stroke subjects were consecutively available patients who met five entry criteria. The criteria were that the subjects could follow three-part verbal instructions, had intact appreciation of proprioception (great toe position) and light moving touch in the paretic lower extremity, could balance longer than 30 sec. with their feet apart, and were capable of bearing more than 75% of their body weight through their paretic lower extremities. All of the 14 stroke subjects were ambulatory (2 independently, 3 with a cane, 7 with a cane and supervision, 2 with a cane and minimal assistance). Ten were paretic on the left and 4 were paretic on the right; 6 were men and 8 were women. The time since onset of stroke (mean ? standard deviation, range) was 35.2 _+ 47.6 (6 to 195) days. Their age (mean? standard deviation, range) was 68.9 ? 14.4, (30 to 84) yrs. Their weight (mean _+ standard deviation, range) was 69.4 k 19.8 (46.4 to 107.3) kg. Fourteen individuals without known neurologic or orthopedic disorders also participated as healthy subjects. They were selected from a previously tested pool (Bohannon, et al., 1989) based on the match of their weight (primarily) and age and sex (secondarily) with the stroke subjects. All could balance for at least 30 sec. with their feet apart and were capable of bearing more than 75% of their weight through either lower extremity. All were independently ambulatory without an assistive device (i.e., cane); 7 were men and 7 were women. Their age (mean k standard deviation, range) was 71.5 8.0 (48 to 84) yr. Their weight (mean f standard deviation, range) was 69.1 +_ 9.4 (56.4 to 85.4) kg.
+
Instrumentation and Procedure Two digital Borg Electronic Scales were used to measure weightbearing. The scales registered weight in pounds to the nearest . I pound. They were calibrated prior to each testing session. They were placed side to side during testing. Prior to testing all subjects were oriented to the testing procedure using verbal introductions. They were told that they would be stepping onto the two scales, placing the left and right foot each on the middle of a scale. Once on the scales they were to look straight ahead. They were told that they would be requested to adjust their weight through the designated lower extremity (paretic for stroke subjects, randomly chosen for controls) to each of three randomly ordered target levels-25, 50, and 75% of their total weight They were informed that three attempts would be allowed at each target level. To give perspective regarding the weightbearing targets, the subjects were provided three guides. First, they were told that, when 50% of their total body weight was placed through the designated lower extremity, they would be standing symmetrically with even (hdf of their) weight through both feet. Second, they were told that 0% of their body weight
ACCURACY OF WEIGHT BEARING
937
through the designated lower extremity would mean that the lower extremity was off the scale and that 100% of their body weight through the designated lower extremity would mean that their other lower extremity was off the scale. Third, they were told that 25% weightbearing through the designated lower extremity would require a reduction of weight through the designated extremity from one-half (50%) of their total body weight and that 75% weightbearing would require an addition of weight through the designated lower extremity from one-half (50%) of their total body weight. For each weightbearing target subjects were told to step onto the scales (without shoes) as indicated previously and to adjust their weight through
FIG 1. Stroke subject adjusting her weight to a target level on digital scales while closely guarded by the examiner
938
R. W. BOHANNON
&
D. TINTI-WALD
the designated lower extremity. While so adjusting their weight, they were closely guarded by the investigator to assure their safety (Fig. 1). Once they were at what they perceived to be the appropriate target level of weightbearing, they notified the investigator who counted to four and recorded the weight from the digital scale. Subjects were then told to return to comfortable standing. The process then was repeated two more times with 10-sec. intervals of comfortable standing between trials at each target level. Between the target levels, subjects stepped off of the scales for 15 sec.
Analysis of Data Each weightbearing measurement was divided by total body weight to yield the actual percentage of weightbearing for each of the three trials at the three target levels. To provide an indication of reliability of measurement one-way analyses of variance for repeated measures were applied to the three trial measures at each target level for the stroke group and for the healthy group. The mean squared values were used to calculate intraclass correlation coefficients (Shrout & Fleiss, 1979, Formula 3.1). The intraclass correlation coefficients were consistent, with one exception, with fair to good reliability for both the stroke group (25% target = .801, 50% target = ,896, 75% target = ,753) and the healthy group (25% target = ,668, 50% target = ,882, 75% target = .866). The mean actual percent weightbearing for each weightbearing target level was calculated for each subject from the three trials. Errors in weightbearing were calculated for each subject by taking the absolute value of the difference between the mean actual percent weightbearing and the percent target-level weightbearing associated with it. For example, if a subject's mean actual percent weightbearing at the 50% target was 56.0%, then the error was 6.0%. The effects of group (stroke versus healthy) and target level (25%, 50%, and 75%) on actual weightbearing percent and error in weightbearing were assessed using two separate two-factor analyses of variance. In each analysis one of the factors (target level) was repeated. The level of significance selected was p < .05.
The actual percent weightbearing demonstrated by subjects of each group at each target is summarized in Fig. 2 and Table 1. There was c o significant difference (F,,2, = 1.19, p = .29) in the actual percent weightbearing between the two groups (stroke versus healthy). There was, however, a significant difference (F,,,, = 145.04, p
ACCURACY O F WEIGHTBEARING ESTIMATION BY STROKE VERSUS HEALTHY SUBJECTS ' RICHARD W. BOHANNON Uniuersity of Connecticut Harrford (C7J Hospital
AND
DIANE TINTI-WALD Hartford (CT)Hospital
Summary.-This study was performed to describe and compare the accuracy of weightbearing at three target levels (25, 50, and 75% of body weight) of 14 ambulatory stroke subjects and 14 matched healthy subjects. Weightbeating through a designated lower extremity (stroke subjects-paretic, healthy subjects-randomly selected) was measured with digital scales. No significant difference was found in magnitude of weightbeating (%) between the stroke and healthy groups. Although making significantly greater errors in weightbearing than healthy subjects, stroke subjects did not consistently weightbear under target. Subjects with stroke tested in earlier studies were neither asked to stand symmetrically nor examined to judge whether they were capable of accepting the required weight through the paretic lower extremity. Before assuming that weightbearing asymmetry is a problem in patients with stroke, clinicians should examine weightbearing behavior more specifically.
Patients who have experienced cerebrovascular accidents (strokes) have a known tendency to weightbear asymmetrically during bilateral upright stance. More specifically, when asked to stand comfortably or straight the patients tend to place more weight through their nonparetic than through their paretic lower extremities (Bohannon & Larkin, 1985; Caldwell, MacDonald, MacNeil, McFarland, Turnball, & Wall, 1986; Dickstein, Nissan, Pillar, & Scheer, 1985; Mizrahi, Solzi, Ring, & Nisell, 1989). A reasonable, albeit untested explanation of the patients' weightbearing asymmetry is that they have an impaired ability to judge weightbearing in stance. The explanation is reasonable given what is already known about stroke patients' tendency to misjudge other sensory inputs, e.g., vertical and horizontal (Birch, Proctor, Bortner, & Lowenthal, 1960) and heaviness (Gandevia & McCloskey, 1977). A recent study of healthy individuals outlines a method by which the ability of stroke patients to judge weightbearing can be tested (Bohannon, Waters, & Cooper, 1989). That method requires subjects to adjust weightbearing through a designated lower extremity to a target level. The purpose of this study was to describe and compare the accuracy of weightbearing at three target levels in a small sample of stroke subjects and a matched sample of healthy subjects. Subjects
All subjects were volunteers who participated with informed consent. 'Address correspondence to Dr. R. W.Bohannon, School of Allied Health, U-101, University of Connecticut, Storrs, CT 06269-2101.
936
R. W. BOHANNON & D. TIFJTI-WALD
Fourteen subjects had experienced a cerebrovascular accident (stroke) prior to admission to the acute care hospital where the research was performed. The stroke subjects were consecutively available patients who met five entry criteria. The criteria were that the subjects could follow three-part verbal instructions, had intact appreciation of proprioception (great toe position) and light moving touch in the paretic lower extremity, could balance longer than 30 sec. with their feet apart, and were capable of bearing more than 75% of their body weight through their paretic lower extremities. All of the 14 stroke subjects were ambulatory (2 independently, 3 with a cane, 7 with a cane and supervision, 2 with a cane and minimal assistance). Ten were paretic on the left and 4 were paretic on the right; 6 were men and 8 were women. The time since onset of stroke (mean ? standard deviation, range) was 35.2 _+ 47.6 (6 to 195) days. Their age (mean? standard deviation, range) was 68.9 ? 14.4, (30 to 84) yrs. Their weight (mean _+ standard deviation, range) was 69.4 k 19.8 (46.4 to 107.3) kg. Fourteen individuals without known neurologic or orthopedic disorders also participated as healthy subjects. They were selected from a previously tested pool (Bohannon, et al., 1989) based on the match of their weight (primarily) and age and sex (secondarily) with the stroke subjects. All could balance for at least 30 sec. with their feet apart and were capable of bearing more than 75% of their weight through either lower extremity. All were independently ambulatory without an assistive device (i.e., cane); 7 were men and 7 were women. Their age (mean k standard deviation, range) was 71.5 8.0 (48 to 84) yr. Their weight (mean f standard deviation, range) was 69.1 +_ 9.4 (56.4 to 85.4) kg.
+
Instrumentation and Procedure Two digital Borg Electronic Scales were used to measure weightbearing. The scales registered weight in pounds to the nearest . I pound. They were calibrated prior to each testing session. They were placed side to side during testing. Prior to testing all subjects were oriented to the testing procedure using verbal introductions. They were told that they would be stepping onto the two scales, placing the left and right foot each on the middle of a scale. Once on the scales they were to look straight ahead. They were told that they would be requested to adjust their weight through the designated lower extremity (paretic for stroke subjects, randomly chosen for controls) to each of three randomly ordered target levels-25, 50, and 75% of their total weight They were informed that three attempts would be allowed at each target level. To give perspective regarding the weightbearing targets, the subjects were provided three guides. First, they were told that, when 50% of their total body weight was placed through the designated lower extremity, they would be standing symmetrically with even (hdf of their) weight through both feet. Second, they were told that 0% of their body weight
ACCURACY OF WEIGHT BEARING
937
through the designated lower extremity would mean that the lower extremity was off the scale and that 100% of their body weight through the designated lower extremity would mean that their other lower extremity was off the scale. Third, they were told that 25% weightbearing through the designated lower extremity would require a reduction of weight through the designated extremity from one-half (50%) of their total body weight and that 75% weightbearing would require an addition of weight through the designated lower extremity from one-half (50%) of their total body weight. For each weightbearing target subjects were told to step onto the scales (without shoes) as indicated previously and to adjust their weight through
FIG 1. Stroke subject adjusting her weight to a target level on digital scales while closely guarded by the examiner
938
R. W. BOHANNON
&
D. TINTI-WALD
the designated lower extremity. While so adjusting their weight, they were closely guarded by the investigator to assure their safety (Fig. 1). Once they were at what they perceived to be the appropriate target level of weightbearing, they notified the investigator who counted to four and recorded the weight from the digital scale. Subjects were then told to return to comfortable standing. The process then was repeated two more times with 10-sec. intervals of comfortable standing between trials at each target level. Between the target levels, subjects stepped off of the scales for 15 sec.
Analysis of Data Each weightbearing measurement was divided by total body weight to yield the actual percentage of weightbearing for each of the three trials at the three target levels. To provide an indication of reliability of measurement one-way analyses of variance for repeated measures were applied to the three trial measures at each target level for the stroke group and for the healthy group. The mean squared values were used to calculate intraclass correlation coefficients (Shrout & Fleiss, 1979, Formula 3.1). The intraclass correlation coefficients were consistent, with one exception, with fair to good reliability for both the stroke group (25% target = .801, 50% target = ,896, 75% target = ,753) and the healthy group (25% target = ,668, 50% target = ,882, 75% target = .866). The mean actual percent weightbearing for each weightbearing target level was calculated for each subject from the three trials. Errors in weightbearing were calculated for each subject by taking the absolute value of the difference between the mean actual percent weightbearing and the percent target-level weightbearing associated with it. For example, if a subject's mean actual percent weightbearing at the 50% target was 56.0%, then the error was 6.0%. The effects of group (stroke versus healthy) and target level (25%, 50%, and 75%) on actual weightbearing percent and error in weightbearing were assessed using two separate two-factor analyses of variance. In each analysis one of the factors (target level) was repeated. The level of significance selected was p < .05.
The actual percent weightbearing demonstrated by subjects of each group at each target is summarized in Fig. 2 and Table 1. There was c o significant difference (F,,2, = 1.19, p = .29) in the actual percent weightbearing between the two groups (stroke versus healthy). There was, however, a significant difference (F,,,, = 145.04, p
