| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Reports |
RP Di Fabio, PT, PhD, is Professor, Department of Physical Medicine and Rehabilitation, and a senior faculty member of the Graduate Programs in Neuroscience, University of Minnesota, Minneapolis, MN 55455 (USA)
C Zampieri, PT, PhD, is Postdoctoral Fellow, Neurological Sciences Institute, Oregon Health & Science University, Beaverton, Ore
P Tuite, MD, is Associate Professor, Department of Neurology, University of Minnesota
Address all correspondence to Dr Di Fabio at: difab001{at}umn.edu
Submitted May 25, 2007;
Accepted September 24, 2007
 | Abstract |
|---|
Subjects and Methods: Twelve subjects with gaze palsy (mild versus severe) secondary to progressive supranuclear palsy were evaluated during a stair-climbing task in a cross-sectional study of mechanisms influencing eye-foot coordination. Infrared oculography and electromagnetic tracking sensors measured eye and foot kinematics, respectively. The primary outcome measures were vertical gaze fixation scores, foot lift asymmetries, and sagittal-plane foot trajectories.
Results: The subjects with severe gaze palsy had significantly lower lag foot lift relative to lead foot lift than those with a mild form of gaze palsy. The lag foot trajectory for the subjects with severe gaze palsy tended to be low, with a heading toward contact with the edge of the stair. Subjects with severe gaze palsy were 28 times more likely to experience "fixation intrusion" (high vertical gaze fixation score) during an attempted shift of gaze downward than those with mild ocular motor deficits (odds ratio [OR]=28.3, 95% confidence interval [CI]=6.4–124.8). Subjects with severe gaze shift deficits also were 4 times more likely to have lower lag foot lift with respect to lead foot lift than those with mild ocular motor dysfunction (OR=4.0, 95% CI=1.7–9.7).
Discussion and Conclusion: The small number of subjects and the variation in symptom profiles make the generalization of findings preliminary. Deficits in gaze control may influence stepping behaviors and increase the risk of trips or falls during stair climbing. Neural and kinematic hypotheses are discussed as possible contributing mechanisms.
| Introduction |
|---|
 TopAbstract Introduction MethodResultsDiscussionConclusionReferences |
|---|
Lamont and Zehr5 suggested that stair climbing and other locomotor tasks, such as level walking and inclined walking, share a common "oscillatory core that sets up a basic rhythm" and symmetry of stepping. However, it is not clear how supraspinal inputs related to gaze control influence spinal interneurons to facilitate or inhibit rhythmic step symmetry.
Foot lift asymmetries were reported in community-dwelling older people with a history of previous falls or mobility deficits.6 These people had lower lag foot elevation relative to lead foot elevation, had lower lag foot sagittal-plane trajectories, and came in contact with obstacles in their walking path more frequently than people in an older cohort with no recent history of falls.6 Lower lag foot lift relative to lead foot lift during obstacle avoidance may be related to an interaction between gaze control deficits7,8 and the control of foot kinematics.1,2
Gaze saccade-foot stepping coordination is dependent on visual input during certain phases of gait. When visual input is eliminated during just the swing phase of gait, there is no decrement in the performance of a precision stepping task for people without impairments.9 It is possible to accurately step to a target that is made invisible throughout the swing phase of stepping. Decrements in the performance of precision stepping tasks, however, occur when visual fixation is prevented during the stance phase of the leg that is preparing to move to the next target.9 Accurate stepping behavior, therefore, relies on early visual input (prior to foot lift) for planning subsequent steps. Although stepping accuracy during some tasks appears to be influenced by eye movements in people who are healthy, the impact of disease that affects the ocular motor system has not been clearly delineated with respect to lower-limb coordination.
The purpose of this study was to determine whether foot kinematics during stair climbing are influenced by gaze shifts prior to step initiation. This study involved subjects with ophthalmoplegia secondary to a parkinsonian syndrome called "progressive supranuclear palsy" (PSP).10,11 Subjects with PSP were recruited to evaluate the influence of gaze control on foot lift kinematics, because people with this disorder have a range of ocular motor deficits and experience recurrent falls.12 Zampieri and Di Fabio12 recently summarized the clinical presentation of people with PSP. Slowness of vertical saccades is one of the main diagnostic criteria for PSP.11 The primary hypothesis was that people with severe gaze palsy would show lower lag foot clearance (relative to the lead foot) and lower lag foot trajectories during stair climbing than people with mild gaze palsy.
| Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
The extent of vertical gaze palsy can differ among people diagnosed with PSP. Some people have slowed but nearly full vertical saccades (mild deficit), whereas others have a limitation of vertical gaze amplitude (severe deficit).11 All subjects in this study showed some degree of gaze shift deficits. For the purposes of this study, subjects with a score of 3 or 4 on the downward saccade test item in the ocular motor section of the PSP Rating Scale14 (indicating 50% or less downward gaze from visual observation of the researcher) were classified as having severe deficits in gaze shift ability. Those with scores of 2 or less (more than 50% downward gaze range of motion) were classified as having mild deficits in gaze shift ability. The characteristics of the subjects participating in this study are outlined in the Table.14–16
|
Procedure
At the first visit, each subject was interviewed by a physical therapist and screened for basic cognitive and mobility functions. The PSP Rating Scale12,14 was administered to all subjects. This scale is a valid and reliable tool for the assessment of function in people with PSP.14 Higher scores reflect increasing disability (maximum score=100). The tool has 6 subscales; history, mentation, bulbar signs, ocular motor function, limb motor function, and gait and midline function. The test items that comprise the content of each subscale are described elsewhere.12,14 The PSP Rating Scale subscale scores for each subject in our study are illustrated in Figure 1. Within 1 week of the initial visit, the subjects returned to the testing facility for computerized ocular motor and kinematic analyses.
|
positioned laterally to avoid obstruction of the vision of the left eye. The infrared sensor has a sensitivity of 0.1 degree of eye movement, a horizontal range of ±30 degrees, and a vertical range of ±20 degrees.18 The eye-tracking camera allowed for the direct visualization of eye movement and assessment of eyelid and blink artifacts.
A 6-degree-of-freedom electromagnetic sensor
was attached to the most posterior point of the band at the level of the occiput to measure head rotations in pitch, yaw, and roll. Additional sensors were placed on the upper trunk at thoracic level 1 and on each foot midway between the lateral malleolus and the head of the metatarsal bone. Each trial was videotaped for subsequent descriptive analysis.
The kinematic motion analysis system included a "metal map" that was used to minimize the effects of metal in the environment surrounding the walkway. Static angular and position accuracies are rated by the manufacturer
,19 at 0.5 degree and 1.8 mm, respectively, over a sensor-transmitter range of 20 to 76 cm. Static angular and position resolutions are reported to be 0.1 degree and 0.5 mm, respectively, with the sensor 30.5 cm from the transmitter.19 Dynamic accuracy is reported to be 4 degrees root mean square.19
Eye infrared and kinematic data were collected simultaneously, and data sampling was multiplexed on a single Motion Monitor A/D board interfaced with a parsed serial collection mechanism for the 3-dimensional sensors. Digitized eye position signals were acquired through 2 separate channels on the A/D board: one for horizontal eye movement and the other for vertical eye movement. Only the vertical eye channel was analyzed in this study. Infrared oculography was sampled at 600 Hz. Head motion and foot motion were initially sampled at 100 Hz (the sampling rate is limited to this value by the sensor manufacturer). This means that for a raw data file sampled for 1 second, there would be 600 rows of eye data and 100 rows of kinematic data. The mismatch in data string length was resolved by use of a custom-made MATLAB software program|| to resample "up" with a polyphase implementation interpolation technique.
Once eye and kinematic data were interpolated to the same data string length (600 samples per second), the kinematic and eye data were filtered offline with a 10-Hz, low-pass, zero-phase-shift, second-order Butterworth filter. Saccadic eye movements are slowed in people with PSP.20 Smooth pursuit gain is depressed in people with PSP,21 but the efficacy of smooth pursuit excursion can vary with the target size.22 However, there is a relative sparing of the vestibulo-ocular reflex (VOR).21,23 It has been proposed that a precursor to gaze shift execution is suppression of the VOR.23 In order to measure gaze shift attempts in people with gaze palsy, therefore, the primary dependent measure of gaze control was not rapid eye movement but was the tracking relationship of eye motion and head motion during stepping (described below). In this paradigm, smooth eye movement components of gaze control and "slow" components of saccadic eye movements were retained with this level of filtering.
Gaze Shift Task
The start position for the subjects was standardized with the subjects’ feet 24 cm from the edge of the first of 2 stairs. Each stair was 10 cm in height (width of 57 cm and depth of 48 cm). Subjects began by looking straight ahead. They were cued with a combination of an auditory tone and a visual stimulus (right- or left-facing arrow projected at the end of the walkway) to initiate a step onto a platform. This procedure was used to standardize gaze position, minimize anticipation of the lead foot selected for stepping, and encourage gaze shifts toward the feet prior to step initiation.24 Once subjects heard the auditory tone, they could look wherever they wanted.
Each subject was instructed to step onto the first stair and then stop. After a rest period of approximately 30 seconds, the subject was cued to step onto the second stair. Each step was considered one trial for analysis. Approximately 10 trials were analyzed for each subject.
Evaluating Foot Kinematics
Foot lift asymmetries, sagittal-plane foot trajectories, and maximum velocities were evaluated as the primary lower-extremity kinematic variables. The lead foot was the first foot lifted toward the stair, and the lag foot was the "trailing" foot. Foot lift asymmetries were assessed in 2 ways:
|
|
Evaluating Gaze Control
The vertical gaze fixation score (vGFS) was used as an index of gaze shift ability. The vGFS has been shown to be a valid measure of gaze control in people with PSP and discriminates between cohorts of people with good gaze shift abilities and people with poor gaze shift abilities.7 The vGFS is derived from the root-mean-square error of head pitch versus vertical eye position traces and is normalized to gaze amplitude.7 When log transformations are used to conform with normality assumptions of parametric statistics,25 larger negative values indicate larger downward gaze shift, whereas values moving in the positive direction indicate increasing visual fixation. Thus, as the log vGFS becomes more negative, down gaze shift becomes more prominent. Conversely, as the log vGFS approaches zero, gaze stabilization increases, as in the case of eyes moving opposite head motion.
Values of the log vGFS of greater than –0.55 were selected as a threshold to describe "fixation intrusion" (marked counterrotation of eye movement relative to the head during downward head pitch) that interfered with directing gaze downward. This threshold was selected on the basis of previous analyses of the vGFS in people with PSP7,23 (the values in those reports are generally comparable to the gaze scores reported here once a log transformation is applied). In the present study, the log vGFS was calculated for the interval between the stimulus onset (projected arrow and tone) and the onset of lead foot lift. Confining the measure of gaze control to a preparatory interval provides an index of attempted gaze shift that precedes the initiation of lead foot lift.
Data Analysis
One-way repeated-measures analyses of variance (across diagnostic groups) were used to evaluate each dependent variable: foot lift asymmetry, velocity, and log vGFS. Sagittal-plane foot trajectories were assessed descriptively. In order to determine how well gaze control and step asymmetry predicted the group origin of each trial (trial origin; that is, the accuracy of identifying trials belonging to a subject with severe ocular motor dysfunction versus a subject with mild ocular motor dysfunction), logistic regression was used with "severity group" as the dependent measure. The baseline demographic characteristics of the groups (Table) were evaluated with independent t tests.
The occurrence of a wrong lead foot landing on the stair (a foot not cued by the auditory and visual stimuli) was defined as a step error. During the stair-stepping task, subjects made step errors in 16 (12%) of 139 trials filtered for analysis. Only trials in which the correct lead foot was selected were used for subsequent analysis.
| Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
Foot Kinematics
The subjects with severe gaze shift dysfunction had significantly lower lag foot lift relative to lead foot lift than those with mild dysfunction (F=5.15; df=1,10; P=.046) (Fig. 2A). A display of individual foot lift asymmetry scores showed low lag foot asymmetry (positive scores) in 6 of 8 subjects with severe gaze shift dysfunction (75%), whereas only 1 subject with a mild gaze shift deficit (25%) demonstrated low lag foot asymmetry relative to lead foot asymmetry during stair stepping (Fig. 2B).
|
|
Gaze Control
Overall, subjects with severe dysfunction had a significantly greater preponderance of fixation (less negative log vGFS values) during the gaze preparation period prior to lead foot lift than subjects with mild dysfunction (F=7.90; df=1,10; P=.019) (Figs. 4A and 5). Four subjects with severe ocular motor dysfunction had log vGFS values more positive than –0.55 (50% of this group), indicating that half of this cohort experienced considerable fixation intrusion that interfered with downward gaze shifts. In contrast, only one subject with a mild ocular motor deficit (25% of this group) had log vGFS values more positive than –0.55 (Fig. 4B).
|
|
Accuracy in Predicting Trial Origin
The logistic regression model correctly classified 76% of the trials. The log vGFS and step asymmetry used as dichotomous variables were both significant predictors of trial origin. Subjects with severe gaze palsy were 28 times more likely to experience fixation intrusion during attempted downward gaze shift than those with mild ocular motor deficits (odds ratio=28.3, 95% confidence interval=6.4–124.8). Subjects with severe gaze shift dysfunction also were 4 times more likely to have lower lag foot lift with respect to lead foot lift than those with mild gaze shift deficits (odds ratio=4.0, 95% confidence interval=1.7–9.7).
| Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
A low trajectory of the lag foot during gait or stair stepping is not unique to PSP but may instead be a general marker of fall risk. Previous work showed that community-dwelling older people with a high risk for falling had a pattern of foot lift similar to that in people with PSP and severe ocular motor deficits when stepping over an obstacle in the walking path.6 It should be noted that older people with a high fall risk also generate fewer downward saccades when negotiating obstacles in their walking path.8 The underlying mechanisms accounting for low lag foot lift with respect to lead foot lift, therefore, may be similar in people with severe PSP and older people with a high fall risk.
Low lag foot lift increases the risk of foot contact with the edge of the stair during stepping up. There are many hypotheses that could explain this type of motor behavior, including poor visual localization of the next step location, disruption of the transfer of visual information to the leg motor systems, or the inability of the leg motor systems to effectively use visual inputs. A "neural" hypothesis centers on the disruption of neural mechanisms linking eye coordination and foot coordination. A "kinematic" hypothesis addresses lag foot kinematics mandated by foot position immediately prior to step initiation.
Neural Hypothesis for Eye-Foot Coordination in PSP
In theory, the function of eye saccades prior to foot lift is to provide visual guidance for the next step before the foot moves off the ground.26–28 Because the groups in the present study differed primarily in the ability to generate preparatory gaze shifts (Figs. 4 and 5), it is possible that disruption of the neural mechanisms linking eye and foot movements contributed to low lag foot lift.
People with severe ocular motor deficits are not able to generate sufficient downward gaze to localize a floor target during step initiation.7 Visualization of gaze position during platform stepping in a previous study7 showed that some subjects with PSP demonstrate fixation intrusion, which means that downward head pitch associated with attempted downward gaze is accompanied by upward counterrotation of the eyes, similar to a VOR (Fig. 5). Fixation intrusion may be caused by tonic activation of omnipause neurons unopposed by a healthy saccadic generator.20,29,30
Mohagheghi et al31 found that only lead limb elevation was influenced by online visual feedback. They proposed independent neural control mechanisms for the lead leg and the trailing leg during stepping because the absence of visual cues had no influence on the trailing limb trajectory.31 Different tasks in our study (stepping from a standing position versus stepping on approach while walking toward the obstacle) may account for the finding that gaze shift deficits were linked to foot lift asymmetries during platform stepping.
Kinematic Hypothesis for Foot Placement Prior to Foot Lift
Chou and Draganich2 reported that clearance of an obstacle by the toe of the trailing limb decreased significantly as toe-obstacle distance decreased. Decreases in toe-obstacle distance led to contact of the trailing (but not the leading) foot with the obstacle. In addition, when the lag foot was closer to the obstacle prior to stepping over, there were more contacts with the obstacle. Patla and Greig1 focused on the lead foot and observed that subjects without impairments failed to successfully step over an obstacle in the walking path when the lead foot was placed too close to the obstacle prior to stepping over. Placement of the foot too close to the obstacle results in obstacle contact even when the limb elevation is appropriate.1(p114)
The post hoc analysis of extra forward steps toward the stairs prior to stepping up onto a stair revealed a relatively high percentage of trials with extra steps for subjects with severe dysfunction (27%) compared with subjects with mild dysfunction (10%). The increased prevalence of extra steps in the former subjects could indicate an attempt to "parse" the stepping action as a way to compensate for impaired downward gaze control. However, people with severe ocular motor deficits did not alter foot lift kinematics and maintained low lag foot relative to lead foot lift regardless of the number of steps taken before negotiating the stair. These findings imply that the mechanism responsible for the step asymmetry pattern is independent of the foot-obstacle distance.
Axial rigidity may have prevented adequate hip extension prior to platform stepping in subjects with severe dysfunction. Once the step over the obstacle is initiated, the lag limb hip must extend before the lag foot lift. Limited hip extension could attenuate the swing trajectory of the lag foot over the obstacle. These findings are consistent with previous reports showing that older people with a history of falls had consistent reductions in hip extension during walking32 and low foot lift trajectories when stepping over obstacles.6
A kinematic hypothesis, however, cannot explain less gaze shift in trials with supplemental steps for subjects with severe dysfunction. If subjects moved closer to a stair by taking small forward steps prior to stepping onto the stair, more, not less, gaze shift would be required to visualize the stepping surface. Neural and kinematic hypotheses, therefore, probably interact to regulate stepping behaviors in people with PSP. The logistic regression model revealed that both gaze control and step kinematics were statistically significant predictors of trial origin. The findings from this predictive model might support the notion that a combination of neural and kinematic factors regulates stair-climbing activity.
Clinical Relevance
Questions related to clinical practice focus on the feasibility of training people to improve gaze control and on the potential to reduce the risk of falling. It is important to note that subjects with mild gaze shift deficits and subjects with severe gaze shift deficits had similar levels of disability when tested for gross functional ability, regardless of their level of ocular motor control (Fig. 1B). However, the PSP Rating Scale may not be a sensitive tool for determining fall risk. Although all subjects demonstrated some postural instability, people with severe gaze shift deficits also showed step asymmetry, with low lag foot relative to lead foot clearance over the edge of the stair. This finding suggests that people with severe gaze shift deficits are at greater risk for trips and falls than those who have only mild gaze shift dysfunction.
It is possible that gaze shift ability can be enhanced with practice. Auditory feedback (linking sound pitch with eye excursion) has shown some promise as a training tool for improving gaze shift ability in people with PSP,33 but changes in eye movement capacity have not yet been correlated with improved mobility. Di Fabio et al7 recently reported that some people with PSP show fixation intrusion during platform stepping, as described earlier. Fixation intrusion is the inability to suppress the VOR in preparation for a downward gaze shift. This means that when the head tilts downward in an attempt to look at the stair, the eyes counterrotate upward and interfere with a gaze shift toward the floor. Therefore, training people to suppress the VOR with techniques such as those illustrated by Duvoisin et al34 may have some positive short-term effect. People with PSP who could suppress the VOR were found to have better attention and visuospatial abilities,7,23 and it can be inferred that those with better eye and cognitive functions would navigate the environment more effectively than those with gaze shift and cognitive deficits.
The study of VOR suppression therapy to facilitate gaze shift behaviors in people with PSP is the subject of future investigations. However, there is some support for "eye-foot" training in the literature related to other neurological deficits. Crowdy et al35 trained people with cerebellar ataxia and deficits in gaze control. They practiced eye movements needed to fixate on irregularly spaced floor targets placed on a walkway. Crowdy et al35 found that eye movement rehearsal was effective for improving saccade control and selected gait parameters. When saccades were hypometric (with gaze falling short of the floor-based target), multiple small saccades occurred in sequence to achieve visual fixation on the target before the foot moved to that target27; these findings confirm a link between eye and foot movements coordinated through the central nervous system.
Limitations
People with severe ocular motor deficits may have a more advanced stage of PSP than those with mild gaze shift dysfunction. It could be argued that overall deterioration of the central nervous system produces differences in foot kinematics independent of gaze control. Measures of overall function, such as Unified Parkinson's Disease Rating Scale (UPDRS) motor subscale scores, gross cognitive function, and self-selected gait speed, however, were not statistically different between the groups (Table). The PSP Rating Scale subscale scores differed only for ocular motor function (Fig. 1B). However, logistic regression provided a predictive model that incorporated both gaze shift ability and lower-extremity motor control in the determination of whether trials were likely associated with subjects who had mild gaze shift deficits versus those with severe gaze shift deficits. The limited sample size may introduce some error of inference, reduce the power of the analysis, and limit the generalization of the results. Thus, the results of the present study should be viewed as preliminary.
| Conclusion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
| Footnotes |
|---|
The protocol for this study was approved by the University of Minnesota Human Subjects in Research Committee.
This study was supported by the National Institute of Disability and Rehabilitation Research (grant H133G030159 to Dr Di Fabio).
* Microguide Inc, 1635 Plum Ct, Downers Grove, IL 60515. 
Arrington Research Inc, 27237 N 71st Place, Scottsdale, AZ 85266
Innovative Sports Training Inc, 3711 N Ravenswood, Suite 150, Chicago, IL 60613.
Ascension Technology Corp, PO Box 527, Burlington, VT 05402.
|| The Mathworks Inc, 3 Apple Hill Dr, Natick, MA 01760-2098.
| References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Zampieri and R. P Di Fabio Balance and Eye Movement Training to Improve Gait in People With Progressive Supranuclear Palsy: Quasi-Randomized Clinical Trial Physical Therapy, December 1, 2008; 88(12): 1460 - 1473. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
