Voluntary upper-extremity movements in patients with unilateral peripheral vestibular hypofunction. (Research Report).

People with peripheral vestibular pathology have difficulty maintaining postural stability when visual and somatosensory cues are altered. (1-4) They demonstrate disturbances in the amplitude, selectivity, and adaptability of postural responses to platform perturbations, (5-10) to and they alter their motor performance during tasks of daily living such as locomotion and standing up. (11, 12) Among various locomotor tasks performed in lighted conditions (eg, walking, walking in place, running or hopping), individuals with vestibular loss have lower amplitudes and higher frequencies of vertical head oscillations and lower head angular velocities compared with control subjects. (11) During running without visual input, head oscillation amplitudes increase for people with vestibular loss. (11) In addition, these individuals demonstrate greater head oscillation variability during running in the light and dark compared with people without impairments. (11) Elderly people (67-90 years of age) with bilateral vestibular hypofunction (BVH) have been shown to alter momentum strategies for gait and the sit-to-stand transition. (12) Although subjects with BVH walked more slowly than subjects without impairments and restricted both sagittal- and vertical-plane linear momentum, their lateral linear and angular momentum was excessive. To rise from a chair, the subjects with BVH generally reduced the amount of linear and angular momentum used. (12)

The literature suggests to us that, when possible, people with vestibular pathology select strategies for functional tasks that maximize gaze stability (11) and postural stability. (11, 12) When task complexity is increased, as when asked to perform the same tasks in the dark, these individuals may have greater difficulty achieving the postural stability goal. (11) These findings support the notion that people with vestibular impairments are challenged not only by external postural perturbations but also by self-initiated perturbations associated with voluntary movement.

Independent function requires that a person perform many different voluntary movements (eg, standing up, stepping off a curb, picking up a child) in a variety of tasks or under different environmental constraints (self-paced versus externally paced movement, even versus uneven terrain, light versus dark). Understanding how people with vestibular dysfunction stand up, walk, or respond to external perturbations may or may not predict how they will perform other functional tasks. We have shown that people with unilateral peripheral vestibular loss demonstrate longer upper-extremity movement times (MTs) for reaching movements than people without impairments. (13) Such tasks require postural and voluntary muscle coordination, (14-16) and a coupling between head, eye, and hand movements. (17, 18) The primary purpose of our study was to determine whether people with peripheral vestibular impairment restricted or altered their body segment motion (head, trunk, lower extremity) during voluntary arm tasks compared with people without impairments.

A second purpose of this study was to determine whether the manipulation of task variables (location of target and task certainty) differentially influenced the movement of people with unilateral peripheral vestibular loss. To investigate these issues, we compared kinematic data from people with unilateral peripheral vestibular loss with that from an age-matched comparison group. We predicted that people with vestibular impairment would restrict motion, particularly at the head and trunk, compared with people without impairments. We also expected greater movement differences for more demanding tasks (those that required more head motion and included a target acquisition goal).

Postural and voluntary muscle response latencies have been reported to increase when the movement task is specified in advance by the experimenter (precued) versus when it is presented in a choice reaction time (CRT) design. (19-21) Because sensory cues for postural reference are reduced in people with peripheral vestibular loss, (1) an increase in task uncertainty should challenge these individuals to a greater extent than people without impairments. Thus, we expected greater movement differences between subjects with vestibular impairments and subjects without impairments when tasks were presented in a CRT versus precued paradigm.

Method

Subject selection, subject characteristics, experimental tasks, instrumentation, and procedures have been described in an earlier publication. (13) In this article, further details are provided regarding motion analysis instrumentation.

Subjects

Six adults with unilateral peripheral vestibular loss and 6 adults without such loss volunteered for participation in this study. Subjects with vestibular pathology were recruited from the Jordan Center for Balance Disorders, University of Pittsburgh Medical Center. We recruited the comparison group from the University of Pittsburgh, University of Pittsburgh Medical Center, and the Pittsburgh community. Experimental procedures were evaluated and approved by the University of Pittsburgh's Institutional Review Board, and all subjects provided informed consent.

The subjects with vestibular impairments and the comparison subjects were matched on age, sex, and activity level. The number of hours per week that subjects participated in physical activity (0, 1-2, 2-3, 3-4, 4-5, and >5) served as the criteria for matching the groups on activity level. The mean age of both groups was 49 years (vestibular group: SD=8.2 years; comparison group: SD=8.4 years). Each group contained 4 female subjects and 2 male subjects.

Potential subjects were excluded based on the following criteria: medical history of neurologic, orthopedic (past year), or compromising cardiac or respiratory problems; regular use of vestibular suppressant or psychoactive medicines (eg, Antivert, * Prozac, ([dagger]) Valium ([double dagger])); and alcohol or recreational substance use within 24 hours of testing. All subjects were screened for functional capabilities (including adequate muscle force), range of motion, tactile or proprioceptive sensory deficits, and the ability to stand and ambulate without upper-extremity support or physical assistance. If potential subjects showed impairments in any of these areas that could interfere with testing, they were excluded. All subjects with vestibular loss had balance difficulties and were currently under the care of a physician and a physical therapist. Comparison subjects did not have a past or present complaint of dizziness or imbalance.

All subjects underwent vestibular testing, including caloric, rotational chair (sinusoidal harmonic acceleration at 0.02, 0.05, 0.1, and 0.5 Hz at 50 [degrees]/s peak velocity and at 0.1 Hz at 25 [degrees], 50 [degrees], 100 [degrees], and 150 [degrees]/s peak velocities), oculomotor, positional, and posturography testing. An Equitest platform ([section]) was used for posturography testing. A neurologist reviewed these results to clear subjects for further study participation. Any abnormal test finding excluded potential comparison subjects from testing. To obtain the comparison group, 11 individuals without known neurological impairments underwent the vestibular laboratory tests. Five of these individuals demonstrated abnormal test results and were excluded from further participation.

Potential subjects with vestibular impairments needed to demonstrate a 25% or greater reduced vestibular response (RVR) on caloric testing if the involved ear was irrigated first or a 30% or greater RVR if it was irrigated second. (22) Individuals with less than a 50% RVR were also required to have either a "vestibular pattern" (23) on posturography testing or an asymmetry on rotational chair testing. The reliability of caloric testing measurements in identifying people with vestibular abnormalities increases when a larger RVR value is used as a cutoff. (24) An asymmetric rotational response or a vestibular pattern on posturography increases the reliability of identifying a vestibular disorder because these patterns are seen in patients with unilateral vestibular deficits who have not yet compensated centrally. (24) Vestibular and posturography test results for the subjects with vestibular impairments are summarized in the Table.

The Dizziness Handicap Inventory (DHI), an assessment tool used to document a patient's symptoms and perceived disability, (25) was administered to the subjects with impaired vestibular function. Scores on the DHI range from 0 to 100, with higher scores indicating greater perceived disability. The mean DHI score of the subjects with vestibular impairments was 49.66 (SD=14.61).

All subjects with vestibular loss performed an initial home exercise program (custom designed by a physical therapist and aimed toward reducing vestibular-related symptoms and impairments) for 2 weeks before participating in the testing. For ethical reasons, we did not delay therapy in order for subjects to participate in this study. Exercising for 2 weeks was a potential threat to this study's internal validity. However, because this factor was consistent for all subjects with vestibular loss, we believe any potential effects were negligible.

Experimental Tasks

We obtained data from baseline reaction time (RT) trials and the arm movement tasks in sitting and standing positions. Kinematic data were collected only during standing trials. Because the RT data have been published previously, (13) seated trials and RT data are not discussed further.

Subjects performed 3 different voluntary arm movement tasks with their dominant upper extremity: an overhead reach to a target, a sideward reach to a target, and a forward flexion movement through 90 degrees. The overhead task required subjects to reach and compress a target switch that was suspended from the ceiling. The side task required subjects to reach to their side to compress a target switch. The 90-degree task required the subjects to flex their shoulder forward through 90 degrees of motion.

Subjects performed 2 blocks of 12 reaching trials in the standing position. In one trial block, tasks were precued. In the other trial block, the trials were presented in a CRT paradigm. Within a trial block, subjects performed 4 trials (in random order) of each reaching task (90-degree, side, overhead). Therefore, across both trial blocks, subjects completed 8 trials of each task.

Apparatus

An RT display, consisting of four 5-cm-diameter vertically arranged colored lights, signaled the upper-extremity tasks. The top (red) light warned subjects to prepare for the upcoming task signal (green, yellow, or blue light). A self-contained, microprocessor-based 8-function industrial process controller (Micromaster WP6200 Programmable Controller ([parallel])) drove the RT display.

The release of a momentary push-button fingerswitch (Radio Shack model 275-1566 (#)) strapped to the subject's leg signaled movement initiation. For the 90-degree task, movement completion was signaled when the subject's arm interrupted a motion sensor (Radio Shack model 49-551 (#)). The motion sensor was mounted on 2 adjustable, telescoping tripods. Compression of a 3.8-cm-diameter target (attached to the momentary push-button fingerswitch) signaled movement completion for the side and overhead tasks. The side and overhead targets were attached to telescoping rods.

The Peak Performance Analog Sampling Module ** was used to collect voltage signals from the fingerswitch, RT display, target switches, and motion sensor. Data sampling occurred at a rate of 20 Hz. Kinematic information was measured using an electromagnetic tracking system (Flock of Birds System ([dagger])([dagger])). The system consists of a transmitter that emits a pulsed DC magnetic field and receivers that detect and measure the magnetic signal. From the measured magnetic field characteristics, position and orientation of the body segments are computed independently by each receiver.

Subject preparation. Target location depended on the subject's height, arm length, and hand dominance. To position the subject under the overhead target, the experimenter (DBF) extended a plumb line from the target's center to the tip of the subject's acromial process. Next, the subject elevated his or her arm fully overhead, and the target was placed at the end of the subject's index finger. To adjust the side target height, the subject abducted the arm to 90 degrees, and the height was set accordingly. Side target distance was set at the subject's arm length plus 2.54 cm (1 in). Arm length was measured from the tip of the subject's acromial process to the end of the index finger. The motion sensor was set such that it became activated when the subject achieved 90 degrees of shoulder flexion. Once positioned, we marked the subject's foot placement on the floor to ensure consistency during the experiment.

The RT display was set at a distance of 2 m from the subject. The display's center coincided with the subject's eye level. The fingerswitch was strapped to the lateral aspect of the subject's thigh. The subject's index finger rested comfortably on the switch.

Electromagnetic receivers were attached to the posterior aspect of the subject's head, trunk, thigh, and lower leg. The head receiver was attached to an adjustable headband. The trunk receiver was attached to a vest-like harness at the approximate level of the subject's fifth thoracic vertebra. Lower-extremity receivers were secured to the side of the body that corresponded to the subject's dominant upper extremity at the mid-thigh and at the mid-calf region.

The Flock of Birds transmitter sat on a nonmetallic platform 1.1 m above the floor and positioned behind the subject. The transmitter height was selected to reduce noise by minimizing the transmitter-to-receiver distance. The experimental apparatus and motion sensor placements are illustrated in Figure 1.

[FIGURE 1 OMITTED]

Testing. All testing took place in a well-lighted laboratory. The order of precued and CRT trials was randomly determined for each subject. The order of arm tasks (90 [degrees], overhead, side) within the task certainty condition was randomly determined but kept the same for all subjects.

Quiet standing trials were conducted before performance of the movement trials. For these trials, subjects stood as quietly as possible for 10 seconds. The purpose of these trials was to collect data for a baseline body position for use in the motion data analysis.

The procedure was similar for both precued and CRT trials. Each trial began with a 2.0-second warning signal from the top (red) light of the RT display. Following a random 1.0-, 1.5-, or 2.0-second interstimulus interval, one of the other 3 colored lights signaled the task to the subject. Each light signaled a specific arm task. The blue (second), yellow (third), and green (fourth) lights signaled the overhead, 90-degree, and side tasks, respectively. The 2.0-second task signal cued subjects to release their index finger from the fingerswitch. The trials were separated by a 20.0-second interval, and a 3-minute rest period separated the 2 trial blocks. For all trials, we sampled data for 10 seconds starting 4 seconds before the task signal.

Subjects performed 3 practice trials before starting the experiment. During practice, the experimenter instructed all subjects to release the fingerswitch quickly following the task signal and to move their arm as fast as possible to the appropriate target. The experimenter repeated these instructions at the beginning of each movement trial block. For each precued trial, the experimenter prompted the subjects to the ensuing task. For CRT trials, the task signal alone cued the subjects to the upcoming arm task.

Data Reduction

Body segment and joint motion data. From the quiet standing trial, a mean baseline measurement was obtained for each body segment and subtracted from the measurement obtained for the body segment's initial position during an arm movement trial. This calculation referenced the body at zero to ensure that all subjects began each movement trial from a neutral standing position.

The head and trunk segmental angles were determined directly from the head and trunk receivers. We calculated the neck angle by subtracting the corresponding trunk angle from the head angle. We determined the hip angle by subtracting the corresponding thigh angle from the trunk angle. Knee angle was determined by subtracting the corresponding thigh angle from the shank angle. The shank angle reflected the movement of the shank segment as the feet remained fixed to the ground during the tasks. We obtained head, neck, and trunk angles for the sagittal, frontal, and transverse planes. Hip and shank angles were obtained for the sagittal and frontal planes. For the knee, we determined the sagittal-plane angle only. The data reduction program determined the body segment or joint motion by calculating the difference in the degree of body segment or joint motion at the beginning of the task (illumination of the task signal) and at task completion (compression of a task target or motion sensor activation).

Head velocity. Preliminary examination of the motion data showed the greatest group differences in motion at the head and trunk. In addition, these differences appeared to be limited to transverse-plane motion. However, for the side task, group differences were less than that observed for the overhead and 90-degree tasks. We speculated that if the degree of motion was similar between groups for the side task, then perhaps the groups differed with respect to how they were executing the motion. Thus, for the side task, the question of a between-group difference in head velocity profile arose.

Angular head velocity measurements were calculated from the transverse-plane data obtained from the head segment. A simple first difference equation was used to calculate instantaneous angular head velocity. (26)

V[x.sub.i] = [x.sub.i + 1] - [x.sub.i - 1]/2 [DELTA]t

This equation was used to determine the instantaneous head velocity every 0.1 second.

Six velocity measurements (average velocity, peak velocity, timing of peak velocity, velocity at head movement initiation, timing of head movement initiation, and velocity at target acquisition) were determined from the instantaneous angular velocity values. Velocity at head movement initiation was selected from the instantaneous velocity values by the experimenter using the following criteria. This value was preceded by the task signal and stable, minimal head velocity values (near zero). It was followed by an increase in head velocity of at least 15 [degrees]/s. Timing of head movement initiation or the time period from the task signal to head movement initiation was also noted. Peak velocity was defined as the highest instantaneous velocity value. The length of time from the task signal to peak velocity or the timing of peak velocity was recorded. The instantaneous velocity value coincident with target acquisition was defined as velocity at target acquisition. Lastly, we calculated average velocity by averaging the instantaneous velocity values beginning at head movement initiation and ending at target acquisition.

Design and Analysis

For body segment or joint motion data, separate analyses were performed for each movement plane and task. We selected this approach because we expected body segment or joint motion to differ across planes of movement and for each task. For example, transverse-plane head motion should be greater for the side task than for the 90-degree task. Thus, we performed a total of 9 analyses for body segment or joint motion data. For each analysis, body segment or joint motion data were averaged across the 4 trials (by body segment or joint and task certainty condition) for each subject. The design of all analyses was a group X task certainty X angle analysis of variance (ANOVA) for repeated measures on task certainty and angle. For each plane, the angles included varied. The sagittal-plane analysis included the head, neck, trunk, hip, knee, and ankle angles. The frontal-plane analysis included the head, neck, trunk, hip, and ankle angles. The transverse-plane analysis included the head, neck, and trunk angles. For all analyses, the criterion alpha level for statistical significance was set at .05. Body segment or joint motion data were normalized for subject handedness. Thus, for data interpretation, all subjects used their right upper extremity for task performance.

Separate analyses were also performed for each dependent measure related to head angular velocity (average velocity, peak velocity, timing of peak velocity, velocity at head movement initiation, and velocity at target acquisition). For each dependent measure, values were averaged for the 4 side task trials for each subject. The averaged score was used in the statistical analysis. A 2 X 2 (group X task certainty) ANOVA for repeated measures on task certainty was performed for each velocity measure.

Results

Body Segment or Joint Motion

Subjects with vestibular loss were found to restrict body segment or joint motions while performing the 90-degree task compared with comparison subjects. A group X angle interaction effect was found for frontal-plane motion in the 90-degree task (F=5.0; df=4,90; P<.05). Means and standard deviations for the interaction effect are presented in Figure 2. For both groups, the head rotated slightly to the right, while the trunk rotated to the left. This created rotations of the neck that were opposite to those of the hips. A simple-effects post hoc test indicated that motion differences between groups were focused at the neck, trunk, and hip segments or joints. Although both groups demonstrated similar degrees of head and shank rotations, subjects with vestibular loss had smaller neck, trunk, and hip segment or joint rotations compared with subjects in the comparison group.

[FIGURE 2 OMITTED]

A group X angle interaction effect also occurred for transverse-plane motion in the 90-degree task (F=4.5; df=4,50; P<.05). Means and standard deviations for the interaction effect are presented in Figure 3. During this task, head and neck rotations occurred to the right, while the trunk rotated to the left. A simple-effects post hoc test revealed that subjects with vestibular loss minimized neck angle rotations but demonstrated similar head and trunk rotations compared with subjects in the comparison group.

[FIGURE 3 OMITTED]

Subjects with vestibular impairment also restricted body segment or joint rotations during the overhead task compared with comparison subjects. A group X angle interaction effect occurred for transverse-plane motion in the overhead task (F=5.0; df=4,50; P<.05). Means and standard deviations for the interaction effect are presented in Figure 4. During this task, head, neck, and trunk rotations occurred to the right. A simple-effects post hoc test identified that subjects with vestibular loss moved their trunk to a similar degree but restricted head and neck rotations compared with comparison subjects.

[FIGURE 4 OMITTED]

We did not observe any group main or interaction effects for sagittal plane motion. In addition, no group main or interaction effects were found for side task motion. As mentioned previously, we speculated that if the groups moved to a similar degree for the side task, then perhaps their velocity strategy differed. The following results verified this prediction.

Head Velocity

Results indicate that the subjects with vestibular loss used a different velocity strategy than the comparison subjects. Effects were observed only for average head velocity and head velocity at target acquisition.

A group main effect was observed for average head velocity (F=12.1; df=1,10; P<.05). The subjects with vestibular pathology moved their head an average of 15.2 [degrees]/s slower than the comparison subjects. The mean average velocity was 59.7 [degrees]/s (SD= 10.8) for the subjects with vestibular loss and 74.9 [degrees]/s (SD=13.9) for the comparison subjects.

A group X task certainty effect (F=5.2; df= 1,10; P < .05) was observed for head velocity at target acquisition. Means and standard deviations for the interaction are presented in Figure 5. A simple-effects post hoc test identified that differences in head velocity at target acquisition occurred during tasks performed under the CRT condition only. Under the CRT condition, the subjects with vestibular loss moved their head an average of 74.9 [degrees]/s slower at target acquisition than the comparison subjects.

[FIGURE 5 OMITTED]

An additional post hoc analysis was performed to determine whether laterality of vestibular impairment (left, right, or comparison) influenced head velocity at target acquisition. A 3 X 2 (laterality X task certainty) ANOVA for repeated measures on task certainty was performed. A main effect for laterality was found (F=8.55; df=2,14; P < .05). Mean head velocity at target acquisition was similar for subjects with left-sided vestibular impairment (ie, those with a reduction in function in the left ear) ([bar] X=19.2 [degrees]/s, SD=10.27) and subjects with right-sided vestibular impairment (ie, those with a reduction in function in the right ear) ([bar] X=32.7 [degrees]/s, SD=19.32), but different from the control group ([bar] X=82.1 [degrees]/s, SD= 36.6).

Task Certainty

Minimal task certainty effects were found in this study. We did not observe any group X task certainty effects in the body segment or joint motion analyses. The only group X task certainty effect was previously indicated in the "Head Velocity" section.

Discussion

The prediction that subjects with vestibular loss would restrict motion to a greater extent than the comparison subjects was partially supported by our results. During the 90-degree task, subjects with vestibular loss demonstrated less frontal- and transverse-plane motion than subjects in the comparison group. In the frontal plane, head and shank segment motion was negligible and did not differ between the groups. Frontal body segment or joint motion occurred primarily at the neck, trunk, and hip. In these areas, less motion was observed in the subjects with vestibular loss. In the transverse plane, both groups moved their head and trunk segments during the 90-degree task. Resulting neck angle rotations were found to be smaller for the subjects with vestibular loss than for the comparison subjects. No differences in sagittal-plane body segment or joint motion were found between groups for the 90-degree task.

The 90-degree task clearly differed from the overhead and side tasks. The overhead and side tasks required subjects to move their head to afford target visualization and acquisition. In the 90-degree task, movements of the body segments probably served to compensate for center-of-gravity shifts caused by arm motion. We previously reported that arm MTs for this task were longer for subjects with vestibular impairments than for comparison subjects. (13) By slowing arm movement, less compensatory body segment motion would be needed to maintain postural stability.

Subjects with vestibular loss also restricted motion during the overhead task. Specifically, they reduced transverse-plane head and neck rotations compared with the comparison subjects. Body segment and joint rotations did not differ between groups for sagittal- and frontal-plane motions. For this task, target visualization and acquisition required sagittal-plane body segment or joint motion. We believe that this task constraint prevented subjects with vestibular loss from restricting their sagittal-plane motion. An explanation for the indifference in frontal-plane motion between groups is more difficult to ascertain. Both groups performed considerable frontal-plane head segment motion. The comparison subjects had a mean head angle in the frontal plane of 49.2 degrees (SD= 14.8). The subjects with vestibular loss had a mean head angle in the frontal plane of 45.9 degrees (SD=10). Subjects with vestibular loss had longer arm MTs for this task than did comparison subjects. (13) Therefore, it is possible that by slowing the arm, the threat of postural stability was minimized. This would allow subjects to move their head and trunk to the extent needed for target visualization. If this were true, then the failure to find group differences in frontal- and sagittal-plane motion might be explained. However, it does not explain why subjects with vestibular loss restricted transverse-plane motion.

Although subjects with vestibular pathology restricted transverse-plane motion for the 90-degree and overhead tasks, they did not restrict this motion during performance of the side task. Like the overhead task, this task required target visualization. Minimizing transverse-plane motion could have led to target acquisition errors. We did not observe group differences in frontal- or sagittal-plane motion for this task either. A moderate degree of motion occurred in the sagittal plane for this task. For example, the means for sagittal-plane head segment motion were 30.9 degrees (SD=9.3) for the comparison subjects and 29.0 degrees (SD=12) for the subjects with vestibular loss. Much less motion occurred in the frontal plane. The means for frontal-plane head segment motion were 9.8 degrees (SD=12.6) for the comparison subjects and 9.0 degrees (SD=5.2) for the subjects with vestibular loss. Again, subjects with vestibular loss had increased arm MTs for this task compared with the comparison subjects. (13) The slowness in arm motion coupled with target acquisition demands, in our view, might explain the lack of group motion differences for this task. However, the head velocity analyses revealed that subjects with vestibular pathology did indeed alter their movement strategy for the side task. In contrast to the comparison group, subjects with vestibular loss move their head more slowly.

Our results suggest that across tasks, subjects with vestibular loss consistently altered their transverse-plane motion. These subjects were selected based on clinical diagnostic tests that evaluated horizontal semicircular canal function. Thus, our findings confirm the existence of a relationship between vestibular test results and task performance. Additional relationships between pathology and task performance might have been observed if we had been capable of testing for other vestibular problems. The exploration of other kinematic data also may have uncovered different results, we determined the degree of head motion for each task by calculating the difference between the initial segment position and its position at task completion. Future research should utilize kinematic analyses, including study of velocity variables for all tasks.

Across tasks, subjects with vestibular loss altered their task performance primarily by slowing arm movement (13) and restricting or slowing transverse-plane head motion. We believe these findings suggest that our tasks posed more of an eye-head coordination than postural stability challenge to the subjects with vestibular loss. Bizzi and colleagues (27-29) originally discussed the linear summation hypothesis of eye-head coordination, which proposed that the vestibular ocular reflex (VOR) operated continuously throughout the head movement. This hypothesis has been shown to be limited regarding its application to a wide range of gaze shifts. In humans, linear summation has been shown to occur for small eye-head gaze shifts of about 10 degrees. (30, 31) As the size of the gaze shift increases, the VOR enters a region where it is switched off but is subsequently reactivated as gaze approaches the target. (32-34) Dynamic disturbances of the VOR have been reported in both monkeys and humans following unilateral labyrinthine loss. (35,36) During performance of the side task, subjects with vestibular dysfunction moved their head more slowly than comparison subjects. We propose that this strategy was selected to compensate for an impaired VOR in order to minimize target acquisition errors. Because we did not record eye movements, this hypothesis cannot be verified. However, we did examine the effect of lesion laterality on head velocity at target acquisition. Head velocity at target acquisition was not influenced by lesion laterality. This finding was surprising to us and contrasts with clinical reports of horizontal VOR asymmetry detected during rapid passive head impulses toward the affected ear. (37) The Halmagi head-impulse test (a clinical test of horizontal semicircular canal function) depends on Ewald's second law, which states that rotations that excite a canal are greater than rotations that inhibit a canal. (38) Each horizontal canal responds to head motion in either direction, resulting in an excitation of its vestibular nerve during ipsilateral head motion and inhibition of its vestibular nerve during contralateral head motion. Subjects in our study performed active head rotations with much lower accelerations and velocities than those used in the Halmagyi head-impulse test. Although a positive Halmagyi impulse sign is a clinical indicator of horizontal semicircular canal function, it may not predict whether a person with unilateral vestibular loss will alter his or her head motion during a voluntary task that requires coordination between the head, eyes, and other body segments. Our data suggest that despite the laterality of lesion, individuals with vestibular loss may alter head motion in both directions.

Our second prediction regarding task certainty was partially supported by our results. Task certainty influenced only some of the movement variables and did not influence between-group differences in body segment or joint motion. Task certainty did differentially influence groups with respect to velocity at target acquisition. The groups behaved similarly under the precued condition.

Under task uncertainty (CRT condition), the subjects with vestibular dysfunction moved their head more slowly at target acquisition than the comparison subjects. Comparison subjects actually had greater head velocities at target acquisition under the CRT condition compared with the precued condition. Across groups, we found that head movement was initiated later tot the CRT ([bar] X = 0.62 second, SD = 0.12) as compared with the precued condition ([bar] X = 0.45 second, SD = 0.05). Therefore, both groups selected the strategy of delaying head motion during the CRT condition. Our data suggest that the comparison subjects were able to offset this delay by increasing their head velocity later in the movement. Subjects with vestibular loss did not show this ability. Therefore, increasing task complexity influenced the movement behavior of subjects with vestibular loss.

The results of our study have implications for rehabilitation of individuals with vestibular loss. Our data add to the existing data that suggest that individuals with vestibular loss have difficulty with the execution of functional voluntary movement. (11,12) Much previous research has centered on the ability of people with vestibular impairment to regain balance following an externally imposed perturbation. Based on our findings, we suggest that vestibular rehabilitation should include the practice of a variety of tasks, including those in which the postural challenge is both self-generated and externally generated.

The need for varied task practice is further suggested by the finding that group movement strategy differences in this study were primarily restricted to the head and neck. We observed few differences in trunk and hip motion, with no differences in distal segment motion. We suggest that individuals with vestibular loss demonstrate a mix of motor coordination impairments. Within the context of voluntary upper-extremity tasks, the impairment manifests itself in the upper body segments. For a different task, such as stepping, a different motor coordination impairment could arise. Thus, vestibular rehabilitation assessments should include a thorough examination of the underlying impairments associated with a particular functional complaint.

The finding that task certainty influenced the movement strategies of individuals with vestibular impairment also has clinical implications. Vestibular rehabilitation, in our opinion, should provide opportunities for the patient to practice tasks under conditions of reduced certainty. For example, if using obstacle course negotiation to train a patient's balance, task demands could be changed from one trial to the next. Consider an obstacle course that includes a ball placed in the patient's path. In the first trial, the patient is directed to step over the ball. In the second trial, the patient is told to pick up the ball and throw it to the therapist. On the third attempt, the patient is directed to kick the ball across the room. In each trial, the task demand (step over, pick up, or kick) is not given until the patient approaches the ball. This is important for "real-life" situations where environmental challenges cannot always be predicted.

Summary

Based on our present and previously reported (13) Findings, we have shown that people with unilateral peripheral vestibular loss alter both head and arm motion during voluntary reaching. These alterations may be used to offset the effect of a gaze stability impairment. Increasing arm MT and slowing head velocity would make it more likely that the hand, eyes, and head are aligned at the time of target acquisition. We also found that subjects with unilateral peripheral vestibular loss consistently reduced transverse head or neck motion for experimental tasks in which this motion was not essential for target visualization. We believe that this finding confirms a relationship between horizontal canal pathology and functional movement. Lastly, we found that task uncertainty affected head velocity strategies differently for subjects with vestibular loss than for comparison subjects. This strategy may have improved gaze stabilization and accuracy during target acquisition.

The results of our study have direct implications for rehabilitation. Disability following vestibular loss may reflect varying degrees of impairments from the processes underlying postural control. The interaction between vestibulospinal and vestibulo-ocular systems is complex and may very well depend on the task and environment. Successful rehabilitation depends on the clinician's ability to critically examine each patient's specific functional complaint. Thus, in our opinion, rehabilitation programs for patients with vestibular dysfunction should be individualized and aim to provide the patient with the skills to overcome multiple task and environmental challenges.

Table.
Summary of Caloric, Rotational, and Posturography Results for
Subjects With Vestibular Impairments

                                                   Abnormal
Subject                          Rotational        Posturography
No.      Caloric Test            Test              Conditions

1        Absent left response    Right asymmetry   1-6
2        58% left RVR (a)        Normal            2-6
3        53% left RVR            Normal            2-6
4        Absent right response   Normal            3-6
5        40% right RVR           Left asymmetry    Normal
6        Absent right response   Not tested        6

(a) RVR = reduced vestibular response.

* Pfizer Inc, 235 E 42nd St, New York, NY 10017-5755.

([dagger]) Dista Products Co, Div of Eli Lilly and Co, General Offices, Lilly Corporate Center, Indianapolis, IN 46285.

([double dagger]) Roche Pharmaceuticals, Roche Laboratories Inc, 340 Kingsland St, Nutley, NJ 07110-1199.

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

([parallel]) Lafayette Instrument, PO Box 5729, 3700 Sagamore Pkwy N, Lafayette, IN 47903.

(#) Radio Shack, 200 Taylor St, Ft Worth, TX 76102.

** Peak Performance Technologies Inc, 7388 S Revere Pkwy, Suite 603, Englewood, CO 80112.

([dagger])([dagger]) Ascension Technology Carp, PO Box 527, Burlington, VT 05402.

References

(1) Nashner LM. Adaptation of human movement to altered environments. Trends Neurosci. 1982;5:352-365.

(2) Black FO, Nashner LM. Vestibulo-spinal control differs in patients with reduced versus distorted vestibular function. Acta Otolaryngol Suppl. 1984;406:110-114.

(3) Barin K, Seitz CM, Welling DB. Effect of head orientation on the diagnostic sensitivity of posturography in patients with compensated unilateral lesions. Otolaryngol Head Neck Surg. 1992;106:355-362.

(4) Fox CR, Paige GD. Effect of head orientation on human postural stability following unilateral vestibular ablation. J Vestib Res. 1990-1991; 1:153-160.

(5) Allum JH, Pfaltz CR. Influence of bilateral and acute unilateral peripheral vestibular deficits on early sway stabilizing responses in human tibialis anterior muscles. Acta Otolaryngol Suppl. 1984;406:115-119.

(6) Allum JH, Keshner EA. Vestibular and proprioceptive control of sway stabilization. I n: Bles W, Brandt T, eds. Disorders of Posture and Gait. Amsterdam, the Netherlands: Elsevier Science; 1986:19-40.

(7) Keshner EA, Allum JH. Plasticity in pitch sway stabilization: normal habituation and compensation for peripheral vestibular deficits. In: Bles W, Brandt T, eds. Disorders of Posture and Gait. Amsterdam, the Netherlands: Elsevier Science; 1986:289-314.

(8) Allum JH, Honegger F, Schicks H. The influence of a bilateral peripheral vestibular deficit on postural synergies. J Vestib Res. 1994;4:49-70.

(9) Horak FB, Nashner LM, Diener HC. Postural strategies associated with somatosensory and vestibular loss. Exp Brain Res. 1990;82:167-177.

(10) Nashner LM, Diener HC, Horak FB. Selection of human postural synergies differ with peripheral somatosensory vs vestibular loss. Soc Neurosci Abstracts. 1985;11:704.

(11) Pozzo T, Berthoz A, Lefort L, Vitte E. Head stabilization during various locomotor tasks in humans, II: patients with bilateral peripheral vestibular deficits. Exp Brain Res. 1991;85:208-217.

(12) Kaya BK, Krebs DE, Riley PO. Dynamic stability in elders. Momentum control in locomotor ADL. J Gerontol A Biol Sci Med Sci. 1998;53: M126-M134.

(13) Borello-France DF, Gallagher JD, Redfern M, et al. Voluntary movement strategies of individuals with unilateral peripheral vestibular hypofunction. J Vestib Res. 1999;9:265-275.

(14) Belen'kii VY, Gurfinkel VS, Pal'tsev YI. Elements of voluntary movements. Biophysics. 1967;12:134-141.

(15) Lee WA. Anticipatory control of posture and task muscles during rapid arm flexion. J Motor Behav. 1980;12:185-196.

(16) Lee WA, Buchanan TS, Rogers MW. Effects of arm acceleration and behavioral conditions on the organization of postural adjustments during arm flexion. Exp Brain Res. 1987;66:257-270.

(17) Helsen WF, Elliot D, Starkes JL, Ricker KL. Temporal and spatial coupling of point of gaze and hand movements in aiming. J Motor Behav. 1998;30:249-259.

(18) Carnahan H, Marteniuk RG. Hand, eye, and head coordination while pointing at perturbed targets. J Motor Behav. 1994;24:135-146.

(19) Brown JE, Frank JS. Influence of event anticipation on postural actions accompanying voluntary movement. Exp Brain Res. 1987;67: 645-650.

(20) Laussau-Wray ER, Parker AW. Neuromuscular responses of elderly women to tasks of increasing complexity imposed during walking. Eur J Appl Physiol. 1993;67:476-480.

(21) Inglin B, Woollacott M. Age-related changes in anticipatory postural adjustments associated with arm movements. J Gerontol. 1988;43: M105-M113.

(22) Furman JM, Jacob RG. Jongkees' formula re-evaluated: order effects in the response to alternate binaural bithermal caloric stimulation using closed-loop irrigation. Acta Otolaryngol. 1993;113:3-10.

(23) Furman JM. Role of posturography in the management of vestibular patients. Otolaryngol Head Neck Surg. 1995;112:8-15.

(24) Jacobson GP, Newman CW, Kartush JM. Handbook of Balance Function Testing. St Louis, Mo: Mosby Year Book; 1993.

(25) Jacobson GP, Newman CW. The development of the Dizziness Handicap Inventory. Arch Otolaryngol Head Neck Surg. 1990;116:424-427.

(26) Winter DA. Biomechanics and Motor Control of Human Movement. 2nd ed. New York, NY: John Wiley & Sons Inc; 1990.

(27) Bizzi E, Kalil RE, Tagliasco V. Eye-head coordination in monkeys: evidence for centrally patterned organization. Science. 1971;171:452-454.

(28) Dichgans J, Bizzi E, Morasso P, Tagliasco V. Mechanisms underlying recovery of eye-head coordination following bilateral labyrinthectomy in monkeys. Exp Brain Res. 1973;18:548-562.

(29) Morasso P, Bizzi E, Dichgans J. Adjustment of saccade characteristics during head movements. Exp Brain Res. 1973;16:492-500.

(30) Guitton D, Volle M. Gaze control in humans: eye-head coordination during orienting movements to targets within and beyond the oculomotor range. J Neurophysiol. 1987;58:427-459.

(31) Laurutis VP, Robinson DA. The vestibulo-ocular reflex during human sacaddic eye movements. J Physiol (Lond). 1986;373:209-233.

(32) Tomlinson RD. Combined eye-head gaze shifts in the primate, III: contributions to the accuracy of gaze saccades. J Neurophysiol. 1990;64: 1873-1891.

(33) Tomlinson RD, Bahara PS. Combined eye-head gaze shifts in the primate, II: interactions between saccades and the vestibuloocular reflex. J Neurophysiol. 1986;56:1558-1570.

(34) Pelisson D, Prablanc C, Urquizar C. Vestibuloocular reflex inhibition and gaze saccade control characteristics during eye-head orientation in humans. J Neurophysiol. 1988;59:1007-1013.

(35) Fetter M, Zee DS. Recovery from unilateral labyrinthectomy in rhesus monkey. J Neurophysiol. 1988;59:370-393.

(36) Honrubia V, Jenkins HA, Baloh RW, et al. Vestibulo-ocular reflexes in peripheral labyrinthine lesions, I: unilateral dysfunction. Am J Otolaryngol. 1984;5:15-26.

(37) Halmagyi GM, Curthoys I. A clinical sign of canal paresis. Arch Neurol. 1988;45:737-739.

(38) Baloh RW, Honrubia V, Konrad HR. Ewald's second law re-evaluated. Acta Otolargyngol. 1977;83:474-479.

All authors provided concept/idea/research design. Dr Borello-France, Dr Gallagher, and Dr Furman provided writing. Dr Borello-France, Dr Furman, and Dr Redfern provided data collection, and Dr Borello-France and Dr Gallagher provided data analysis. The Human Movement and Balance Laboratory (Director, Dr Redfern) and the Department of Otolaryngology at the University of Pittsburgh provided facilities/equipment used in this study. Dr Furman provided patient subjects and performed neurological and vestibular evaluations. Dr Gallagher, Dr Furman, Dr Redfern, and Dr Carvell provided consultation (including review of manuscript before submission).

This study was completed in partial fulfillment of the requirements for Dr Borello-France's Doctor of Philosophy degree in education at the University of Pittsburgh.

This study was approved by the Institutional Review Board at the University of Pittsburgh.

This work was funded, in part, by a Doctoral Research Award from the Foundation for Physical Therapy and by a grant from the National Institute on Aging (NIH-NIA AG 10009).

This article was submitted July 31, 2000, and was accepted September 4, 2001.

DF Borello-France, PT, PhD, is Assistant Professor, Department of Physical Therapy, Duquesne University, 111 Health Sciences Bldg, Pittsburgh, PA 15282 (USA) ([e-mail]). Address all correspondence to Dr Borello-France.

JD Gallagher, PhD, is Associate Professor, Department of Health, Physical, and Recreation Education, School of Education, University of Pittsburgh, Pittsburgh, Pa.

JM Furman, MD, PhD, is Professor, Department of Otolaryngology, University of Pittsburgh, Pittsburgh, Pa.

MS Redfern, PhD, is Associate Professor, Departments of Bioengineering and Otolaryngology, University of Pittsburgh, Pittsburgh, Pa.

GE Carvell, PT, PhD, is Professor, Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh, Pittsburgh, Pa.