PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Vision Res. Author manuscript; available in PMC 2011 November 23.
Published in final edited form as:
PMCID: PMC3008343
NIHMSID: NIHMS236530

Visuospatial Perception and Navigation in Parkinson’s Disease

3. INTRODUCTION

Movement related symptoms including muscle rigidity, tremor, and bradykinesia are commonly associated with Parkinson’s disease (PD) (Berardelli et al. 2001), but non-motor symptoms, such as decrements in the sense of smell (Double et al., 2003), haptic acuity (Konczak et al., 2008) and visual perception (Lee and Harris, 1999) are also prevalent. Davidsdottir et al. (2005) and Lee and Harris (1999) administered questionnaires pertaining to various visual perceptual disorders, to individuals with PD. Many of the PD patients reported double vision, experienced changes in the detection of color and luminance contrast, misjudged spaces between objects, and perceived that vehicles and people appeared to move faster than they had in the past. These reported visual disorders may alter the perception of optic flow and as a result lead to changes in gait, such as decreased walking speed, increased stride frequency, and veering while walking.

Optic flow experienced by the observer, during locomotion, contains relevant information regarding heading direction and influences gait coordination and various walking parameters (Bruce et al., 1996; Dyre and Andersen, 1997; Prokop et al., 1997; Telford and Howard, 1996; Warren et al., 1988; Warren et al., 1991; Warren et al., 2001; Wilkie and Wann, 2006). Our research group explored the effects of systematic manipulations of optic flow speed on veering and the coordination of walking in healthy younger and older adults (Chou et al., 2009). Results supported previous findings (Himann et al., 1988; Konczak, 1994; Prokop et al., 1997) and showed that increases in optic flow speed were accompanied by decreases in walking speed, which was primarily adjusted via stride frequency. It was also demonstrated that humans veered away from the faster wall in an attempt to equalize the relative optic flow speed experienced in each hemi-field; the greater the difference between walls in terms of optic flow speed, the more veering occurred (Chatziastros et al., 1999; Chou et al., 2009; Duchon and Warren, 2002).

Visuospatial functional testing has shown differences in performance between PD patients for whom the left body-side was initially affected (LPD) as a result of the degeneration of the basal ganglia in the right hemisphere, compared to those initially affected on the right body-side (RPD) (Amick et al., 2006; Davidsdottir et al., 2008; Harris et al., 2003; Lee et al., 2001a; Schendan et al., 2009). During line bisection tasks, LPD consistently made judgments to the right of center, and RPD to the left (Lee et al., 2001b). Further research revealed that LPD underestimated the size of objects such as apertures (Lee et al., 2001b) and shapes (Harris et al., 2003) located on the left side of the visual space, their left visual hemi-field appeared compressed, whereas for RPD, the right visual field was more compressed than the left. To LPD, a unilateral compression of the visual field would cause the left border of the visual field to shift toward the right, resulting in a line bisection bias toward the right; for RPD, the opposite would be true. For a walker with PD, it is expected that veering occurs away from the perceived compressed visual field, such that LPD would walk toward the right and RPD toward the left.

Continued research revealed that PD patients perceive an egocentric reference point (ECRP) that has shifted toward the side of the brain with more extensive basal ganglia damage (Davidsdottir et al., 2008). The ECRP divides the perceived space into left and right hemi-fields with respect to the midline of the trunk (Karnath et al., 1991). A shifted ECRP among PD results in a perceived field of view that is shifted to the right for LPD and to the left for RPD and is in the same direction as the observed line bisection biases. Accordingly, a shifted ECRP may influence gait such that LPD would veer toward the right and RPD toward the left. Findings presented in the literature suggest that individuals with PD perceive a field of view that has shifted toward the side of the brain with initial basal ganglia damage. Because both predict a shift in the same direction, it is unclear whether the underlying mechanism of the shift is due to a unilateral compression of the visual field or a shifted ECRP. In either case, because of a shifted field of view it is expected that individuals with PD would walk in the same direction as the shift such that LPD would walk toward the right while RPD toward the left.

Veering during gait could also be attributed to asymmetries in the perception of optic flow speed. Research has shown that, during optic flow speed manipulations, individuals with LPD report the right visual field as moving faster than the left, while the opposite is true for RPD (Davidsdottir et al., 2008). A possible explanation for this asymmetric perception of optic flow is related to the unilateral compression of the visual field reported by Harris et al. (2003). During forward movement, texture elements within the compressed visual hemi-field travel a smaller perceived distance over the same period of time as compared to the uncompressed side. Texture elements on the compressed side would expand at a slower optic flow speed. As mentioned earlier, it has been shown that individuals walk away from the faster moving wall (Chatziastros et a., 1999; Chou et al., 2009; Duchon and Warren, 2002). Accordingly, individuals with LPD, who perceive slower optic flow in the left visual field compared to the right, should veer toward the left, away from the perceived faster moving visual field, whereas, RPD should move toward the right. These predictions are in the opposite direction of veering expected as a result of a shifted field of view due to unilateral compression of the visual field or a shifted ECRP.

In addition to a shifted field of view and/or asymmetric perceptions of optic flow speed, PD patients often show marked gait asymmetries (Johnsen et al., 2009; Lewek et al., 2009; Plotnik and Hausdorff, 2008) that may affect the coordination dynamics of walking and contribute to veering. The literature shows that individuals walk toward the side of the body with smaller step lengths (Courtine and Schiepatti, 2004). Accordingly, it is expected that individuals with PD veer toward their initially effected body-side; LPD would veer to the left and RPD right. The present study implemented virtual reality techniques to investigate the relative effects of a shifted field of view, asymmetric perceptions of optic flow speed, and gait asymmetries on heading direction, functional gait parameters, and the inter-limb coordination patterns of walking in patients with PD and healthy, age-matched control adults. Based on the outcomes of previous research it was hypothesized that:

  1. During the eyes open (EO), blindfolded (BF), and virtual reality (VR) walking trials of Experiment 1, findings reported in the literature would be replicated indicating that patients with PD have decreased walking speed (WS), increased stride frequency (SF), and smaller stride lengths (SL) than healthy age-matched adults.
  2. With symmetric and asymmetric optic flow speed manipulations, increases in optic flow speed would be accompanied by decreases in WS and SL and increases in SF for all participants. In contrast to the control group, PD patients would show an asymmetry between the initially-affected side and the secondarily-affected side in terms of SL as well as the phase and frequency relations between arm and leg movements.
  3. During symmetric optic flow speed manipulations, LPD patients would veer toward the right and RPD patients toward the left, indicating the presence of a shifted field of view among PD patients. When asymmetries in optic flow were presented, as in Experiment 3, all participants would use a navigation strategy that equalizes optic flow speed laterally and would walk away from the faster moving wall.

4. METHODS

4. 1. PARTICIPANTS

Subjects included in this study were initially recruited by the Psychology Department of Boston University and took part in a related bout of visuospatial testing prior to completing the current work (Davidsdottir et al., 2008). Mild to moderate PD patients (Hoehn and Yahr stages I – III) and healthy, age-matched, adult controls (HC) from the Boston area were recruited to the study. Informed consent approved by the Institute Review Board of Boston University was obtained according to the Declaration of Helsinki. Of the 31 PD and 18 HC individuals who participated in the visuospatial testing, 23 PD and 17 HC were included in the current experiment. Six individuals did not participate in the current experiment because they were not able to schedule the walking assessment within a short enough time frame of the visuospatial testing. The data for two participants was not included in these analyses because they were unable to complete the entire walking protocol. All participants were native speakers of English. Exclusion criteria included co-existing serious chronic medical illnesses (including psychiatric or neurological), use of psychoactive medication besides antidepressants and anxiolytics in the PD group, use of any psychoactive medications in the control group, history of intracranial surgery, traumatic brain injury, alcoholism or other drug abuse, or eye disease or abnormalities as noted on a neuro-ophthalmological examination. Individuals with a physical disability that prevented them from moving freely, such as past knee or hip surgeries or lower back pain, were excluded. All individuals were required and able to walk without assistive devices. Participants were not demented as indicated by scores of 26 or above on the Mini-Mental State Examination (Folstein et al., 1975) and 135 or above on the Mattis Dementia Rating Scale (Mattis, 1976; Mattis, 1988). In order to investigate the effects of the unilateral onset of PD on visual perception and gait coordination, the PD patients were divided into two groups: 1) those whose motor symptoms presented initially on the left side of the body (LPD) and 2) those initially affected on the right side of the body (RPD). This assessment was based on extensive review of neurology records and conducted by the Parkinson’s Disease and Movement Disorder’s Center at Boston University.

4. 2. MATERIALS

4. 2. 1. Virtual Reality System

A virtual hallway was created using World ToolKit Release 9 (Sense8, San Francisco, CA) on an Onyx2 Reality graphics work station (Silicon Graphics Inc., Mountainview, CA). The width of the hallway was 2.0 virtual meters, the height of the hallway was 2.55 virtual meters and the depth of the hallway was fixed at 15 virtual meters. The hallway’s two sidewalls were textured with randomly placed white dots, 12 virtual centimeters in diameter, with a frequency of ten dots per square meter. To simulate depth perception, dots at the far end of the hallway were smaller than dots nearer to the observer. The front wall, ceiling and floor were black (Figure 1). The visual scene was displayed on a ProView 60 Head Mounted Display (HMD) (Kaiser Electro-Optics Inc, Mountainview, CA) that weighed 1.4 kg. The HMD contains two, active, LCD panels (640 × 480 resolution, color, 60 Hz) and has a 60° field-of-view with true binocular overlap. HMD position and orientation were tracked and updated via an IS-900 LAT system (InterSense, Burlington, MA) with an accuracy of less than 0.3 mm for position data and 0.1º for orientation data. HMD eyepieces were adjusted so that participants could see comfortably and functionally with corrective lenses if necessary. All remaining ambient light from the lab room was eliminated with an additional mask that weighed 0.2 kg.

Figure 1
The Virtual Hallway

4. 2. 2. Three-Dimensional Kinematics

Dependent variables were calculated with kinematic data collected using an OptoTrak/3020 System (Northern Digital Inc., Waterloo, ON, Canada). An OptoTrak bank was placed on each side of the walkway and a third OptoTrak bank was located at the front end of the walkway in order to capture a full three-dimensional range of movement for at least eight strides. Cameras were calibrated to a mean error or 0.7 mm or less. A total of 14 active, light-emitting diodes (LEDs) were fixated to each ankle (lateral malleolus), knee (patella), wrist (radiocarpal joint), shoulder (humeral head), cheek (2 cm below zygomatic arch) and hip (anterior superior iliac spine). There were two additional LEDs, one attached to the chin and one to the HMD. Small deviations from anatomical landmarks were made to increase the LEDs’ visibility. Real time position of the 14 LEDs were sampled at a rate of 100 Hz and analyzed via MatLab (The MathWorks, Inc., Natick, MA). The position time-series were filtered using a zero-lag, fourth order, Butterworth, low-pass filter with a cut-off frequency of 5 Hz. Shoulder and wrist markers were used to determine arm angle relative to the vertical, while leg angle relative to vertical was calculated from hip and ankle position data. Forward wrist or foot movement resulted in a positive joint angular displacement. To account for the increase and decrease in acceleration during the speed-up and slow-down phase of each trial, only the middle six strides of each trial were included in the analyses.

4. 2. 3. Walking Speed Gates

During Experiment 1, two photoelectric gates (Safe House, Fort Worth, TX), placed 6 meters apart, were used in order to train participants to walk within the instructed walking speed range of 0.8 ± 0.05 m/s. Participants walked a total of 10 meters. The first two meters were prior to the photoelectric gates and allowed for the subject to reach steady walking pattern. The next six meters were timed. The last two meters were not timed and allowed the subject ample time to slow down and stop. The photoelectric gaits were used for training purposes only, and were not used to determine walking speed values reported in the Results section.

4. 3. TESTING

The experimental series was composed of three parts: 1) a manipulation of vision, 2) a symmetric manipulation of optic flow speed, and 3) an asymmetric manipulation of optic flow speed. Participants were trained to walk overground at 0.8 ± 0.05 m/s, first with eyes open, then while blindfolded, and finally while wearing the HMD. The instructed walking speed is lower than the average comfortable walking speed (≈1.2 m/s) of healthy adults (Morris et al., 1996); however, it was chosen because inter-limb and trunk coordination is relatively unstable at walking speeds between 0.6 and 0.9 m/s (Wagenaar and Van Emmerik, 2000). At walking speeds lower than 0.75 m/s, the arms and legs move in more or less a 2:1 frequency relation with an in-phase rotation between transverse pelvic and thoracic rotation, while at higher walking speeds the arms and legs move in a 1:1 frequency ratio with an out-of-phase transverse pelvic-thoracic rotation (Van Emmerik and Wagenaar, 1996). It was predicted that a systematic manipulation of optic flow speed around 0.8 m/s would allow for one of the two coordination patterns to emerge. During each vision condition, subjects practiced until five consecutive trials within the instructed walking speed range, as evaluated by the photoelectric gates, were achieved. A spotter remained at the side of the subject to insure safety throughout the entire experiment.

Experiment 1 – Vision Conditions

Participants walked with their eyes opened (EO), while blindfolded (BF), and while exposed to the virtual hallway (VR). During these conditions, in order to simulate a natural hallway environment, optic flow speed was the same as the subjects’ actual ongoing walking speed and the end of the hallway appeared to loom as subjects approached the end of the hallway.

Experiment 2 – Symmetric Optic Flow Conditions

A symmetric manipulation of optic flow speed was presented in which the dots that comprised the texture of the walls of the virtual hallway moved at equivalent speeds in each hemi-field. The optic flow speeds implemented were 0.0, 0.4, 0.8, 1.2 and 1.6 m/s (Chou et al., 2009). The speeds were presented to the participant in random order, which was accomplished using computer generated randomization procedures programmed in MatLab. Five trials were performed at each optic flow speed, yielding 25 symmetric trials in total. During Experiment 2 and 3, in order to highlight the effects of optic flow speed manipulation, the depth of the hallway was fixed at 15 virtual meters so that the hallway appeared endless and no looming of the front wall occurred.

Experiment 3 – Asymmetric Optic Flow Conditions

The speed of the right or left wall, selected at random, remained constant at 0.8 m/s, while the speed of the other wall was randomly manipulated through 0.0, 0.4, 0.8, 1.2 and 1.6 m/s. Five trials were conducted at each optic flow speed, which resulted in a total of 25 trials. This process was repeated with the other wall held at a constant optic flow speed of 0.8 m/s, yielding an additional 25 trials. In total, individuals completed 50 asymmetric optic flow speed trials (see also Chou et al., 2009).

4. 4. DEPENDENT VARIABLES

4. 4. 1. Stride Parameters

Walking Speed (WS), Stride Frequency (SF), and Stride Length (SL)

WS during the middle six strides was reported in meters per second (m/s) and calculated using the displacement of the chin marker in the X-axis divided by the total time it took to walk those strides. The SF was determined by dividing the time it took to travel the six strides by the number of strides taken and is presented in terms of strides per second (Hz). Estimates of left and right SL during the six strides are reported in meters (m) and calculated by dividing the total displacement in the X-axis of the respective ankle marker by the number of strides taken (in each case this value was 6).

4. 4. 2. Relative Power Index

The ratio of arm swing to leg swing was estimated via the calculated angular displacements of the arm and the leg. The respective time-series were transformed via a power spectral density (PSD) function that was estimated by a fast Fourier transform algorithm using the Welch method for power estimation and a Hanning window for smoothing (Wagenaar and Van Emmerik, 2000). Specific movement frequencies and corresponding power for the leg and arm movements were revealed for each time-series plot. The PSD was normalized by dividing the frequency power distribution by the mean power calculated over the 0.2 – 2.5 Hz frequency range for each trial separately. Judging from the PSD of the leg swing time-series, the frequency with the largest power was defined as the stride frequency. Step frequency is twice the stride frequency and was represented by the second peak in the PSD. The power in the PSD of the arm swing time-series at the stride and step frequencies was identified using the stride and step frequencies from the leg movements. In order to quantify the inter-limb coordination pattern, the relative power index (RPI) was calculated as follows:

equation M1
(1)

P1 is the power of the stride frequency in the arm swing time-series and P2 is the power of the step frequency in the arm swing time-series; by definition, both P1 and P2 are greater than or equal to zero. If P2 is equal to 0 and P1 is greater than 0, then RPI will be equal to positive 1, indicating a 1:1 frequency coupling between arm and leg. If P1 is equal to 0 and P2 is greater than 0, then RPI will be equal to −1, indicating a 2:1 coordination pattern. Thus, RPI ranges from −1 to 1, where positive values of the RPI signify that the arm movements are predominantly synchronized with stride frequency, and more negative values indicate that arm movements are predominantly locked onto step frequency (Wagenaar and Van Emmerik, 2000).

4. 4. 3. Generalized Relative Phase

The generalized relative phase (GRP) equations used in this experiment compute the relative phase between limbs locked into 1:1 as well as 1:2 inter-limb frequency ratios (Saltzman and Byrd, 2000; Sternad et al, 1999). The angular position of the arm and leg data was used to calculate the generalized relative phase between the following limb pairs: 1) left arm versus right arm (LARA), 2) left arm versus left leg (LALL), and 3) right arm versus right leg (RARL). When limbs are locked into a 1:1 frequency ratio, the equation for the GRP is as follows:

Ψ1:1(t) = [1 ∗ θ2(t)−1 ∗ θ1(t)] (mod360)
(2)

For limbs locked into a 1:2 frequency ratio the equation for GRP is as follows:

Ψ1:2(t) = [1 ∗ θ2(t)−2 ∗ θ1(t)] (mod360)
(3)

Where Ψ is the calculated GRP at time, t, and θ1 and θ2 are phase angles for the limbs compared. For the computation of the GRP for the LALL and RARL, the legs were the reference limbs; during the LARA analysis, the right arm was the reference limb. Values of GRP range from 0° to 360°, thus a GRP of 180° indicates an out-of-phase coordination (the limbs move in the opposite direction), where a GRP of 0° or 360° indicates an in-phase coordination (the limbs move in the same direction). In addition, relative to the reference limb, a GRP value between 0° and 180° denotes phase advance, and between 180° and 360° denotes phase delay (Saltzman et al., 1998). The GRP was calculated for each stride cycle where each stride cycle was identified by two consecutive maxima from the angular position data of the left and the right leg. The mean GRP of each trial was calculated by means of circular statistics (Sternad et al., 1999).

4. 4. 4. Lateral Drift

The average between left and right hip position data in the Z-axis was used in the calculation of lateral drift during walking. The measure of drift is the difference of this position data, in the Z direction, between the sixth (Z6) and the first stride (Z1), where Z6 is the maximum medio-lateral deviation in the Z-direction of the sixth stride cycle and Z1 is the maximum medio-lateral deviation of the first stride cycle. A positive value indicates rightward drift and a negative value indicates leftward drift. Lateral drift was recorded in millimeters. During the symmetric optic flow conditions of Experiment 2, if PD patients were guided by the equalization of optic flow speed, LPD would be expected to move towards the left, away from the perceived faster moving wall, while RPD would veer towards the right. Gait asymmetry would result in veering toward the body side with a smaller SL. In this case, LPD would walk toward the left and RPD to the right. By contrast, if individuals with PD experience a shifted field of view, LPD would veer towards the right and RPD toward the left.

4. 5. STATISTICS

A generalized linear model with a repeated measures design was used to evaluate whether there were significant effects of Vision Condition or Optic Flow Speed, and whether the PD and HC groups behaved differently. For Hypothesis 1, there was one between-group factor, Group (2 levels: PD and HC). There were two within-group factors: Vision Condition (3 levels: EO, BF and VR) and Trial (5 levels). With regard to Hypotheses 2 and 3 there was one between-group factor, Group (2 levels: PD and HC). There were two within-group factors: Optic Flow Speed (5 levels: 0.0, 0.4, 0.8, 1.2 and 1.6 m/s) and Trial (5 levels). For asymmetric trials, similar statistics were applied. Because the optic flow speed of one side of the hallway remained constant or fixed, however, a within-group factor, Fixed Wall (2 factors: Left Wall Fixed and Right Wall Fixed), was added. For all statistical analyses regarding stride length, relative power index, and generalized relative phase, an additional within-group factor, Side (2 levels: initially affected body-side and secondarily affected body-side) was included. These analyses were also conducted with the PD group separated into LPD and RPD in order to observe differences in gait asymmetry between the two groups; instances in which main effects or interaction effects were significant are reflected in the Results section. For these analyses the between-group factor, Group, had 3 levels (LPD, RPD and HC). Since all but one of the HC participants was right handed, the left body-side for this group was compared to the initially-affected body-side for LPD and RPD. For all tests, a two-tailed significance level of 0.05 was chosen. When significant interaction effects were found, post-hoc tests with a Bonferonni adjustment were performed. Only statistically significant results are reported.

A Mann-Whitney-Wilcoxon test was used to determine whether generalized relative phase values were significantly different than 180°. Similar statistics were used to determine whether the lateral drift values obtained for each group significantly differed from 0.

5. RESULTS

There were 17 HC (8 men and 9 women; 16 right handed and 1 left handed; mean age 60.57 ± 8.7 years, age range 46–73 years) and 23 PD patients, 14 of whom were LPD (7 men, 7 women; 13 right handed, 1 left handed; mean age 60.79 ± 9.15 years, age range 45–75 years) and 9 of whom were RPD (5 men, 4 women; 9 right handed; mean age 60.44 ± 7.55 years, age range 48–73 years), included in the present study. All groups were matched for age and education, F (2, 38) = 0.20, p = 0.82 and F (2, 38) = 0.38, p = 0.69, respectively. There were no significant differences between LPD and RPD in terms of disease duration, t (20) = 1.58, p = 0.13, or for extent of motor disability as measured by the Hoehn and Yahr scale (Kolmogorov-Smith, Z = 0.48, p = 0.97). The majority of PD participants used levodopa (19, or 82%) and/or dopamine agonists (16, or 70%). Five participants (22%) used monoamine oxidase inhibitors and catechol-O-methyltransferase inhibitors and eight (35%) used amantadine or anticholinergic agents. Six (26%) reported using antidepressant or antianxiety medications. LPD and RPD did not differ significantly in medication usage. All optic flow manipulations were conducted during peak medication hours.

5. 1. Experiment 1: Manipulation of Vision Condition

5. 1. 1. Stride Parameters

Walking Speed

A significant main effect for Vision Condition was observed, F (1, 34) = 20.74, p < 0.001, indicating that, pooled across PD and HC, individuals decreased WS during BF trials. There were significant differences in WS between EO and BF (p < 0.001) and between BF and VR (p < 0.001) (Table 1).

Table 1
Experiment 1: Stride Parameters (Mean ± Standard Deviation)

Stride Frequency

Significant main effects were found for Vision Condition, F (1, 34) = 5.29, p = 0.007, and Group, F (1, 34) = 5.62, p = 0.02. Pooled across PD and HC, subjects increased SF when blindfolded and, overall, PD had a significantly higher SF than HC (Table 1).

Stride Length

A significant main effect was found for Vision Condition, F (1, 38) = 39.57, p < 0.001, showing that all patients decreased SL during BF conditions. A significant main effect was found for Group, F (1, 38) = 6.20, p = 0.02, indicating that PD had a smaller SL than HC. Significant main effects were observed for Side, F (1, 38) = 4.13, p = 0.05, demonstrating that, for PD participants, the SL on the initially-affected side was smaller than on the secondarily-affected side; for HC, the left body-side SL was smaller than the right (Table 1). When this analysis was repeated with the PD group divided into LPD and RPD, there was a significant interaction between Side and Group, F (2, 37) = 4.69 p = 0.02. A significant main effect for Side was observed for RPD only, F (1, 8) = 7.03, p = 0.03. The main effect for Side was not significant for LPD, F (1, 13) = 0.14, p = 0.72, nor for HC, F (1, 16) = 1.66, p = 0.22.

5. 1. 2. Relative Power Index

Significant main effects were observed for Vision Condition, F (1, 38) = 3.43, p = 0.04, indicating that RPI values increased during BF and VR conditions. There was a significant interaction between Side and Group, F (1, 38) = 8.18, p = 0.007. When analyzed separately, there was a significant main effect for Side for PD, F (1, 22) = 10.77, p = 0.003, but not for HC, F (1, 16) = 0.49, p = 0.50. The initially-affected side had a significantly lower RPI than the secondarily-affected side for PD only (Table 2). When the RPI analysis was repeated with the PD group divided into LPD and RPD, there was a significant interaction between Side and Group, F (2, 38) = 4.36 p = 0.007. A significant main effect for Side was observed for LPD only, F (1, 13) = 6.98, p = 0.02; RPD approached significance at F (1, 8) = 5.05, p = 0.06 with HC at F (1, 16) = 0.49, p = 0.50.

Table 2
Experiment 1: Relative Power Index (Mean ± Standard Deviation)

5. 1. 3. Generalized Relative Phase

LALL and RARL

The main effects for Vision Condition and Group approached significance at, F (1, 37) = 2.88, p = 0.06, and F (1, 37) = 3.47, p = 0.07 respectively.

LARA

There were no significant main effects for Vision Condition or Group. The mean generalized relative phase for LARA was significantly smaller than 180° for LPD, p = 0.003, and was significantly larger than 180° for RPD, p = 0.003, indicating that the secondarily-affected right arm led the initially-affected left arm in the LPD patients, whereas the secondarily-affected left arm led the initially-affected right arm in the RPD patients. The results obtained from the LALL, RARL, and LARA analysis of Experiment 1 are very similar to those of Experiment 2.

5. 1. 4. Lateral Drift

The analysis of lateral drift yielded a significant main effect for Vision Condition, F (1, 38) = 8.28, p = 0.001. However, post-hoc analyses were not significant and did not reveal any significant differences between EO, BF, and VR. When lateral drift values for each vision condition was compared to 0, significant differences were obtained during BF, for LPD and RPD, p = 0.005 (Figure 2).

Figure 2
Experiment 1, lateral drift values for PD initially affected on the left body-side (LPD; n = 14; black bars), PD initially affected on the right body side (RPD; n = 9; grey bars), and healthy, age-matched, control adults (HC; n = 17; white bars). Participants ...

5. 2. Experiment 2: Symmetric Optic Flow Conditions

5. 2. 1. Stride Parameters

Walking Speed

A significant main effect was observed for Optic Flow Speed, F (4, 37) = 10.91, p < 0.001. Pooled across PD and HC groups, subjects decreased WS when optic flow increased, and increased WS when optic flow decreased (Table 3).

Table 3
Experiment 2: Stride Parameters (Mean ± Standard Deviation)

Stride Frequency

A significant main effect was observed for Optic Flow Speed, F (4, 37) = 9.30, p < 0.001, demonstrating that increasing optic flow speed resulted in a decreased SF, while decreasing optic flow speed was accompanied by an increase in SF. The main effect for Group approached significance at F (4, 37) = 3.38, p = 0.07, with PD showing a tendency to walk at a higher SF than HC (Table 3).

Stride Length

There was a significant interaction between Optic Flow Speed and Group, F (4, 38) = 3.07, p = 0.02. When PD and HC were analyzed separately, only PD maintained a significant main effect for Optic Flow Speed, F (4, 22) = 4.32, p = 0.003, showing that only PD significantly decreased SL as optic flow speed increased. A significant main effect was observed for Side, F (1, 38) = 4.15, p = 0.05, indicating that for PD, the initially-affected side SL was shorter than the secondarily-affected side, and for HC, the left side SL was shorter than the right side (Table 4). When the SL analysis was repeated with the PD group divided into LPD and RPD, there was a significant interaction between Side and Group, F (2, 37) = 6.46, p = 0.004. A significant main effect was observed for RPD, F (1, 8) = 5.51, p = 0.05 and for HC, F (1, 22) = 6.84, p = 0.02; there was no significant main effect for LPD, F (1, 13) = 2.14, p = 0.17.

Table 4
Experiment 2: Relative Power Index (Mean ± Standard Deviation)

5. 2. 2. Relative Power Index

A significant main effect was found for Optic Flow Speed, F (4, 38) = 2.63, p = 0.04, indicating lower RPI values between the arms and legs as the optic flow speed increased and higher RPI values as the optic flow speed decreased. There was a significant interaction between Side and Group, F (1, 38) = 9.73, p = 0.003. When analyzed separately, PD maintained a significant effect due to Side, F (1, 22) = 11.33, p = 0.003, whereas HC did not, F (1, 16) = 1.09, p = 0.31. These results demonstrate that the initially-affected side RPI was significantly lower than the secondarily-affected side for PD (Table 4). When the RPI analysis was repeated with the PD group divided into LPD and RPD, there was a significant interaction between Side and Group, F (2, 38) = 3.56, p = 0.004. A significant main effect for Side was observed for LPD only, F (1, 13) = 6.85, p = 0.02. For RPD, the main effect for Side approached significance, F (1, 8) = 4.80, p = 0.06, whereas no significant main effect was observed for HC at F (1, 16) = 1.09, p = 0.31.

5. 2. 3. Generalized Relative Phase

LALL and RARL

No significant main or interaction effects were observed for Optic Flow Speed, Group or Side. For all participants, values were less than 180° indicating that the legs were in slight phase advance of the arms (Table 5).

Table 5
Experiment 2: Generalized Relative Phase (Mean ± Standard Deviation)

LARA

No significant main or interaction effects for Optic Flow Speed, Group, or Side were observed. The mean values for LPD were significantly lower than 180°, p < 0.001, while HC and RPD were significantly higher than 180°, p < 0.001, indicating that for LPD and RPD, the secondarily-affected arm led the initially-affected arm (Table 5).

5. 2. 4. Lateral Drift

The main effect for Optic Flow Speed approached significance, F (4, 38) = 2.24, p = 0.07. RPD lateral drift values were significantly different from 0, p < 0.001, and toward the left, which suggests a shift in the field of view toward the left for these subjects. By contrast, LPD and HC did not show drift (p = 0.44, and p = 0.20, respectively) (Figure 3).

Figure 3
Experiment 2, lateral drift values for PD initially affected on the left body-side (LPD; n = 14; black bars), PD patients initially affected on the right body-side (RPD; n = 9; grey bars), and healthy, age-matched, control adults (HC; n= 17; white bars). ...

5. 3. Experiment 3: Asymmetric Optic Flow Conditions

5. 3. 1. Stride Parameters

Walking Speed

A significant main effect was observed for Optic Flow Speed, F (4, 38) = 3.23, p = 0.01, revealing that when the optic flow speed of either wall increased, all groups reduced WS, and when optic flow speed decreased, all groups increased WS.

Stride Frequency

The main effect for Group approached the level of significance, F (1, 38) = 3.67, p = 0.06, showing that the mean SF for PD was slightly higher than for HC.

Stride Length

There was a significant main effect for Group, F (1, 37) = 8.67, p = 0.006 indicating that the mean SL for PD was significantly shorter than for HC. When the PD group was divided into LPD and RPD, there was a significant interaction effect between Side and Group, F (2, 36) = 3.01, p = 0.06. A significant main effect was observed for RPD only, F (1, 8) = 10.93, p = 0.01. There was no significant main effect for Side for LPD, F (1, 13) = 0.17, p = 0.70, nor for HC at F (1, 15) = 1.14, p = 0.30.

5. 3. 2. Relative Power Index

A significant interaction effect between Group and Side was observed, F (1, 38) = 13.72, p = 0.001. When analyzed separately, only PD showed a significant main effect due to Side, F (1, 22) = 14.32, p = 0.001, indicating that the RPI of the initially-affected side was significantly lower than the secondarily-affected side for a. When the RPI analysis was repeated with the PD group divided into LPD and RPD, there was a significant interaction effect between Side and Group, F (2, 37) = 7.78, p = 0.002. A significant main effect for Side was observed for LPD, F (1, 13) = 8.68, p = 0.1, and for RPD, F (1, 8) = 6.36, p = 0.04. For HC, the main effect for Side was not significant, F (1, 16) = 2.35, p = 0.15.

5. 3. 3. Generalized Relative Phase

No significant main effects for Optic Flow Speed or Group and no interaction effects were observed for LALL, RARL and LARA. The mean values for LARA were significantly lower than 180° for LPD, p = 0.003, and significantly higher than 180° for RPD, p = 0.003, showing that, for LPD and RPD, the secondarily-affected arm was in slight phase advance of the initially-affected arm. The results from Experiment 3, obtained from the analysis of WS, SL, SF, RPI, and GRP, are similar to those of Experiment 2.

5. 3. 4. Lateral Drift

The analysis of lateral drift revealed a significant main effect for Optic Flow Speed, F (4, 38) = 5.33, p < 0.001, indicating that, overall, participants veered away from the faster moving wall and the amount of veering increased as the difference in optic flow speed between the walls increased. At the same time, LPD and HC showed significant veering toward the right of the hallway (p < 0.001, p = 0.03, respectively), and RPD showed significant veering toward the left of the hallway, p < 0.001 (Figures 4a and 4b).

Figure 4Figure 4
Figure 4a: Experiment 3 lateral drift values for PD patients initially affected on the left body-side (LPD; n = 14; black bars), PD patients initially affected on the right body-side (RPD; n = 9; grey bars), and healthy, age-matched, control adults (HC; ...

6. DISCUSSION

The aim of the present study was to investigate the effects of symmetric and asymmetric manipulations of optic flow speed on the coordination of walking and heading direction among patients with PD as compared to healthy, age-matched, control adults. Our results support the first hypothesis in that PD walked with a lower walking speed, a higher stride frequency and utilized smaller stride lengths than did HC. In support of hypothesis 2, both symmetric and asymmetric increases in optic flow speed were accompanied by decreases in walking speed and stride frequency, while decreases in optic flow speed were accompanied by increases in walking speed and increased stride frequency. The observed effect of optic flow speed on walking speed supports previous findings in the literature showing that modulations of optic flow speed are accompanied by changes in walking performance (Chou et al., 2009; Prokop et al., 1997; Schubert et a., 2005). During Experiment 3, RPD veered toward the left and LPD veered toward the right of the hallway. This result supports the third hypothesis and signifies that PD patients perceive a field of view that has shifted toward the side of space associated with the side of the brain with more basal ganglia damage. At the same time, during Experiment 3, participants veered away from the faster moving wall; a significant effect that increased as the difference in optic flow speed between the two walls increased. This result supports the third hypothesis and is consistent with the literature regarding the equalization of optic flow during navigation (Chatziastros et al., 1991; Chou et al., 2009; Davidsdottir et al., 2008; Duchon and Warren, 2002) which shows that when there is a discrepancy between the optic flow speeds of two walls, patients tend to move away from the faster moving wall.

The observed effects of the optic flow speed manipulations on lateral drift are consistent and support the literature (Chatziastros et al., 1999; Chou et al., 2009; Duchon and Warren, 2002), in spite of limitations to the study. Small effect sizes may be attributed to small group sizes, which reduce the power of the study design. Effect sizes may also be small because subjects were specifically instructed to walk at 0.8 m/s. Furthermore, the experiment lasted about 2.5 hours, which may have led to fatigue; however, subjects were allowed to take breaks. The weight of the HMD could have influenced the coordination of walking, but previous work with the HMD (Chou et al. 2009; Giphart et al., 2007) and the results of Experiment 1 in the present study show that the VR trials were not statistically different than those of the EO conditions. In addition, lateral drift values obtained during the EO and VR conditions and throughout Experiment 2 coincide with previous works (Chatziastros et al., 1999; Chou et al., 2009; Lee et al., 2001) and show a tendency for healthy subjects to walk, steer and point slightly toward the left. Furthermore, during blindfolded conditions subjects walked significantly toward the left, a result that, although not fully understood, is consistent with previous work (Chou et al., 2009).

The PD participants in the present study showed clear gait asymmetries. For all PD the relative power index of the secondarily-affected side was closer to a value of positive 1 than the initially-affected side, which indicates a more stable 1:1 relation between stride frequency and arm swing on the secondarily-affected arm as compared to the initially-affected arm. The variability of the relative power index for the secondarily-affected side was lower than for the initially-affected side. Further analysis of the generalized relative phase shows that the secondarily-affected arm is in slight phase advance of the initially-affected arm (Table 4). For all PD the initially-affected side stride length was significantly lower than the secondarily-affected side stride length.

Remarkable were the findings that during all optic flow speed manipulations, RPD had a tendency to walk toward the left of the hallway while LPD tended to walk toward the right of the hallway. Previous research has shown that, because of the mechanical consequences of stride length asymmetry, veering occurs toward the side with smaller step lengths (Courtine and Schiepatti, 2004). This would predict that if gait asymmetry were the main contributor to veering in PD, LPD would veer left and RPD would veer right, towards their initially-affected side. However, participants veered toward the secondarily-affected side, which suggests that, along with asymmetries in stride length and other gait parameters, PD patients perceive a shifted field of view, that not only biases line bisection tasks (Lee et al. 2001b), but also alters heading direction such that patients veer toward the side of the brain with more basal ganglia damage. Although, it is unclear whether this shifted field of view is due to a unilateral compression of the visual field or a shifted ECRP, results from previous research show that veering during walking is significantly correlated with ECRP shifts during various optic flow conditions (Davidsdottir et al., 2008). Our research group is currently designing virtual environments to investigate unilateral compression of the visual field and shifts in ECRP.

With regard to the optic flow speed equalization findings in which the participant walked away from the faster moving wall, it could be argued that, rather than optic flow speed, the perceived difference in spatial frequency between the two walls during the asymmetric optic flow speed manipulation caused the veering (Chen et al., 1998). For example, Chatziastros and colleagues (1999) defined optic flow as a temporal change in the optic array that is the product of optic flow speed and the spatial frequency of texture that makes up the visual scene. Their participants steered via joystick down a virtual hallway with unequal spatial frequencies in the two visual hemi-fields; the result showed steering away from the wall with the lower spatial frequency. According to these results, it could be hypothesized that individuals would walk away from the wall with the most number of dots or the faster moving wall. Aside from optic flow speed and spatial frequency, various other visual perceptual cues, such as motion parallax, looming, and horizontal frequency can influence gait coordination and should be investigated using the VR environment.

From the results of the present study it can be concluded that equalizing optic flow speed between the left and the right visual hemi-fields is a strategy that humans implement to control heading direction during walking. Furthermore, for individuals with PD, a shifted field of view influences heading direction such that LPD walk toward the right and RPD toward the left, toward the side of the brain with the most basal ganglia damage. It is open to further investigation as to whether a shifted egocentric reference point results in the perception of a compressed visual hemi-field contra-lateral to the shift or vice versa. Support of this hypothesis would explain the results previously obtained by Harris et al. (2003) and Lee et al. (2001b) in terms of the underestimation of object size in the visual hemi-field opposite from the side of the brain with more extensive basal ganglia damage among individuals with PD.

RESEARCH HIGHLIGHTS

  • Parkinson’s Disease subjects show clear gait asymmetry.
  • All participants veered away from the faster moving wall.
  • Parkinson’s Disease subjects veer due to a shifted field of view, rather than gait asymmetry.

Acknowledgments

This project was funded by a grant from the National Institute of Health (R01 NS050446-01 to A.C.G.) and from a Grant-in-Aid of Research, Sigma Xi (to S.D.). We gratefully thank Erik Giphart from the Steadman-Hawkins Sports Medicine Foundation, Chris Lawlor from the Electronics Design Facility at Boston University, and Tami Rork DeAngelis from the Center for Neurorehabilitation at Boston University for their invaluable support. Our efforts to recruit patients with PD were generously supported by Marie Saint-Hilaire, MD, and Cathi Thomas, RN, MS, of Boston University Medical Center Neurology Associates and by Boston area Parkinson Disease support groups. Most importantly, we thank all of the individuals who participated in this study.

2. ABBREVIATIONS

PD
Parkinson’s Disease
LPD
PD subjects whose movement related symptoms initiated on the left body-side
RPD
PD subjects whose movement related symptoms initiated on the right body-side
HC
Healthy Control
ECRP
Egocentric Reference Point
HMD
Head Mounted Display
EO
Eyes Open
BF
Blindfolded
VR
Virtual Reality
WS
Walking Speed
SF
Stride Frequency
SL
Stride Length
RPI
Relative Power Index
GRP
Generalized Relative Phase
LARA
A comparison between the Left Arm and the Right Arm
LALL
A comparison between the Left Arm and the Left Leg
RARL
A comparison between the Right Arm and the Right Leg
LED
Light Emitting Diode
PSD
Power Spectral Density

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

LITERATURE CITED

1. Amick MM, Schendan HE, Ganis G, Cronin-Golomb A. Frontostriatal circuits are necessary for visuomotor transformation: Mental rotation in Parkinson’s disease. Neuropsychologia. 2006;44:339–349. [PubMed]
2. Berardelli A, Rothwell JC, Thompson PD, Hallett M. Pathophysiology of bradykinesia in Parkinson’s disease. Brain. 2001;124 (2):131–46. [PubMed]
3. Bruce V, Georgeson MA, Green PR. Visual Perception: Physiology, Psychology, and Ecology. 3. Psychology Press; Hove, East Sussex, UK: 1996.
4. Chatziastros A, Wallis G, Bülthoff H. Vision in Vehicles VIII. Boston, MA, USA: 1999. Aug, The use of optical flow and splay angle in steering a central path.
5. Chen Y, Beddell HE, Frishman LJ. The precision of velocity discrimination across spatial frequency. Percept Psychophys. 1998;60 (8):1329–36. [PubMed]
6. Chou YH, Wagenaar RC, Saltzman E, Giphart JE, Young D, Davidsdottir S, Cronin-Golumb A. Effects of symmetric and asymmetric optic flow speed manipulations on locomotion in younger and older adults. J Gerontol. 2009;64B (2):222–31. [PMC free article] [PubMed]
7. Courtine G, Schieppati M. Tuning of a basic coordination pattern constructs straight-ahead and curved walking in humans. J Neurophysiol. 2004;91:1524–1535. [PubMed]
8. Davidsdottir S, Cronin-Golomb A, Lee A. Visual and spatial symptoms in Parkinson’s disease. Vision Res. 2005;45:1285–96. [PubMed]
9. Davidsdottir S, Wagenaar R, Young D, Cronin-Golomb A. Impact of optic flow perception and egocentric coordinates on veering in Parkinson’s disease. Brain. 2008;131:2882–93. [PubMed]
10. Double KL, Rowe DB, Hayes M, Chan DKY, Blackie J, Corbett A, Joffe R, Fung VS, Morris J, Halliday GM. Identifying the pattern of olfactory deficits in Parkinson’s disease using the brief smell identification test. Arch Neurol. 2003;60:545–9. [PubMed]
11. Duchon AP, Warren WH. A visual equalization strategy for locomotor control. Psychol Sci. 2002;13 (3):272–8. [PubMed]
12. Dyre BP, Andersen GJ. Image velocity magnitudes and perception of heading. J Exp Psychol Hum Percept Perform. 1997;23 (2):546–65. [PubMed]
13. Folstein MF, Folstein SE, McHugh PR. ‘Mini-mental state’. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189–98. [PubMed]
14. Giphart JE, Chou YH, Kim DH, Bortnyk CT, Wagenaar RC. Effects of virtual reality immersion and walking speed on coordination of arm and leg movements. Presence. 2007;16:399–413.
15. Harris JP, Atkinson EA, Lee AC, Nithi K, Fowler MS. Hemispace differences in the visual perception of size in left hemiParkinson’s disease. Neuropsychologia. 2003;41:795–807. [PubMed]
16. Himann JE, Cunningham DA, Rechnitzer PA, Paterson DH. Age related changes in speed of walking. Med Sci Sports Exerc. 1988;20 (2):161–8. [PubMed]
17. Johnsen EL, Mogensen PH, Sunde NA, Østergaard K. Improved asymmetry of gait in Parkinson’s disease with DBS. Mov Disord. 2009;24 (4):590–7. [PubMed]
18. Karnath HO, Schenkel P, Fischer B. Trunk orientation as the determining factor of the ‘contralateral’ deficit in the neglect syndrome and as the physical anchor of the internal represenatation of the body in space. Brain. 1991;114 (4):1997–2014. [PubMed]
19. Konczak J. Effects of optic flow on the kinematics of human gait: A comparison of young and older adults. J Mot Behav. 1994;26 (3):225–36. [PubMed]
20. Konczak J, Li K-Y, Tuite PJ, Poizner H. Haptic perception of object curvature in Parkinson’s disease. PloS ONE. 2008;3(7):e2625. doi: 10.1371/journal.pone.000265. [PMC free article] [PubMed] [Cross Ref]
21. Lee A, Harris J. Problems with perception of space in Parkinson’s disease: a questionnaire study. Neuro-opthalmology. 1999;22 (1):1–15.
22. Lee AC, Harris JP, Atkinson EA, Fowler MS. Evidence from a line bisection task for visuospatial neglect in Left Hemiparkinson’s disease. Vision Res. 2001a;41:2677–86. [PubMed]
23. Lee AC, Harris JP, Atkinson EA, Fowler MS. Disruption of estimation of body-scaled aperture width in Hemiparkinson’s disease. Neuropsychologia. 2001b;39:1097–1104. [PubMed]
24. Lewek MD, Poole R, Johnson J, Halawa O, Huang X. Arm swing magnitude and asymmetry during gait in the early stages of Parkinson’s disease. Gait Posture. 2010;31 (2):256–60. [PMC free article] [PubMed]
25. Mattis S. Mental status examination for organic mental syndromes in the elderly patient. In: Bellak LK, Toksoz BE, editors. Geriatric Psychiatry. Grune and Stratton; New York: 1976.
26. Mattis S. Dementia Rating Scale: professional manual. Psychological Assessment Resources; Odessa, FL: 1988.
27. Morris ME, Iansek R, Matayas TA, Summers JJ. Stride length regulation in Parkinson’s disease: Normalization strategies and underlying mechanisms. Brain. 1996;119:551–68. [PubMed]
28. Plotnik M, Hausdorff JM. The role of gait rhythmicity and bilateral coordination of stepping in the patholphysiology of freezing of gait in Parkinson’s disease. Mov Disord. 2008;23 (Supp 2):S444–50. [PubMed]
29. Prokop T, Schubert M, Berger W. Visual influence on human locomotion. Modulation to changes in optic flow. Exp Brain Res. 1997;114 (1):63–70. [PubMed]
30. Saltzman E, Byrd D. Task-dynamics of gestural timing: Phase windows and multifrequency rhythms. Hum Mov Sci. 2000;19:499–526.
31. Saltzman E, Lofqvist A, Kay B, Kinsella-Shaw J, Rubin P. Dynamics of intergestural timing: a perturbation study of lip-larynx coordination. Exp Brain Res. 1998;123:412–424. [PubMed]
32. Schendan HE, Amick M, Cronin-Golomb A. Role of a lateralized parietal-basal ganglia circuit in hierarchical pattern perception: Evidence from Parkinson’s disease. Behav Neurosci. 2009;123:125–136. [PMC free article] [PubMed]
33. Schubert M, Prokop T, Brocke F, Berger W. Visual kinesthesia and locomotion in Parkinson’s disease. Mov Disord. 2005;20:141–50. [PubMed]
34. Srinivasan MV, Lehrer M, Kirchner W, Zhang W. Range perception through apparent image speed in freely flying honeybees. Vis Neurosci. 1991;6 (5):519–35. [PubMed]
35. Sternad D, Turvey MT, Saltzman E. Dynamics of 1:2 coordination: generalizing relative phase to n:m rhythms. J Mot Behav. 1999;31:207–223. [PubMed]
36. Telford L, Howard IP. Role of optical flow field asymmetry in the perception of heading during linear motion. Percept Psychophys. 1996;58 (2):283–8. [PubMed]
37. Van Emmerik REA, Wagenaar RC. Effects of walking velocity on relative phase dynamics in the trunk in human walking. J Biomech. 1996;29:1175–1184. [PubMed]
38. Wagenaar RC, van Emmerik REA. Resonant frequencies of arms and legs identify different walking patterns. J Biomech. 2000;33 (7):853–61. [PubMed]
39. Warren WH, Kay BA, Zosh WD, Duchon AP, Sahuc S. Optic flow is used to control human walking. Nat Neurosci. 2001;4:213–6. [PubMed]
40. Warren WH, Mestre DR, Blackwell AW, Morris MW. Perception of circular heading from optical flow. J Exp Psychol Hum Percept Perform. 1991;17 (1):28–43. [PubMed]
41. Warren W, Morris M, Kalish M. Perception of translational heading from optical flow. J Exp Psychol Hum Percept Perform. 1988;14(4):646–60. 37. [PubMed]
42. Wilkie RM, Wann JP. Judgements of path, not heading, guide locomotion. J Exp Psychol Hum Percept Perform. 2006;31 (1):88–96. [PubMed]