Phillip Kramer1, Elliot M. Frohman2, Daniele Nuti3, David S. Zee4, From: 1The New Jersey Neuroscience Institute, Edison, NJ. 2Dept of Neurology and Ophthalmology, UT Southwestern Medical Center, Dallas, TX. 3Dept of Otolaryngology, University of Siena. 4Dept of Neurology, Neurosciences, Otolaryngology and Ophthalmology, The Johns Hopkins Hospital, Baltimore, MD.
We looked at two data sets to determine the effects of horizontal head position and step velocity on the Vestibulo-Ocular Reflex.
Experiment 1. We studied six normal subjects to investigate if horizontal head-on-body position and the velocity of step rotations influence the vestibulo-ocular reflex (VOR). Seated subjects were rotated using left and right velocity steps of 60, 120 and 180 d/s. For each velocity step the head was stabilized with a bite bar at 60 deg left of center, centered and 60 deg right of center for a total of 18 rotations done in a Latin square order. Eye movements were recorded at 500 Hz using scleral search coils. The eye movements during rotation were analyzed for slow phase velocity, gain, time constant (Tc), mean beating span (MBS - the mean of the number of degrees the eyes travel during each slow phase), mean beating frequency (MBFq - the mean of the inverse of time between fast phases) and mean beating field (MBF - the mean of the eye position at the center of each slow phase). Leftward and rightward rotations were averaged and head position was considered either ipsi- or contra-lateral to the direction of rotation. Analysis of variance indicated that head position did not influence any of these parameters, but that step velocity affected all except gain. As step velocity increased from 60 to 180 d/s; Tc decreased (11.5 to 9.0 sec, p=0.02), MBS increased (5.8 to 10.7 deg, p=0.001) and MBFq increased (2.8 to 3.2 Hz, p=0.007). While the change in MBF was statistically significant (p=0.04) it was not proportional to step velocity (nearly midline at 60 and 180 d/s and 5 deg contralateral to the direction of chair rotation at 120 d/s).
Experiment 2. We retrospectively examined the effect of step velocity
on the vestibulo-ocular reflex (VOR) during rotary chair testing of 896
"dizzy" patients. Patients were exposed to four velocity step
profiles (60 d/s left, 60 d/s right, 240 d/s left and 240 d/s right) with
their head maintained in the midline position. The 60 d/s rotation always
preceded the 240 d/s rotations. Eye movements were recorded during both
the per and post portions of the rotation for a total of eight accelerations.
At The Johns Hopkins Hospital (JHH) eye movements were recorded using
electrooculography at 100 Hz and slow phases were identified automatically
but verified manually. At The New Jersey Neuroscience Institute (NJNI)
eye movements were recorded using infrared videooculography at 60 Hz and
slow phase velocities were calculated with the use of a heuristic filter
and by determining the modal velocity every 0.25 sec. The resulting slow
phase velocity (SPV) data was fitted to the form
SPV=A*exp(-t/Tc).
Where t is time and Tc is the time constant of the exponential decay. The gain of the VOR was calculated by dividing the calculated peak SPV (A) by the peak velocity of chair rotation.
The records of 985 patients were examined (658 from JHH and 327 from NJNI) and 89 were eliminated due to incomplete testing, noise or excessive spontaneous nystagmus leaving 896 patients.
We made seven comparisons and found them all to be statistically significant p << 0.0001. Gains were higher and time constant were longer for rotations at 60 d/s than for rotations at 240 d/s. For both 60 and 240 d/s the gain during the per portion of the rotation exceeded the gain during the post portion of the rotation. For both 60 and 240 d/s the Tc during the post portion of the rotation exceeded the Tc during the per portion of the rotation. Finally directional preponderance was higher for rotations at 60 d/s than at 240 d/s. These results are shown in Table 1. For instance the first line of Table 1 is read as - the gain of the VOR for step rotations to 60 d/s exceeded the gain of the VOR for step rotations to 240 d/s 71% of the time (p much less than 0.0001). The mean gain for 60 d/s steps was 0.58 and for 240 d/s steps was 0.47.
Table 1 - Step rotations to 60 d/s preceded
step rotations to 240 d/s - 896 patients
|
Parameter
|
How Often
|
p
|
Mean
|
Mean
|
|
Gain - 60 d/s > 240 d/s
|
71%
|
<<0.0001
|
60 d/s - 0.58
|
240 d/s - 0.47
|
|
Tc - 60 d/s > 240 d/s
|
71%
|
<<0.0001
|
60 d/s - 16.5s
|
240 d/s - 13.2s
|
|
Gain - 60 d/s Per>Post
|
63%
|
<<0.0001
|
Per - 0.61
|
Post - 0.55
|
|
Gain - 240 d/s Per>Post
|
60%
|
<<0.0001
|
Per - 0.50
|
Post - 0.47
|
|
Tc - 60 d/s Per<Post
|
60%
|
<<0.0001
|
Per - 15.9s
|
Post - 17.0s
|
|
Tc - 240 d/s Per<Post
|
66%
|
<<0.0001
|
Per - 12.5s
|
Post - 13.9s
|
|
DP - 60 d/s > 240 d/s
|
59%
|
<<0.0001
|
60 d/s - 10.5%
|
240 d/s - 7.8%
|
Since the 60 d/s rotations were performed before the 240 d/s rotation it is possible that the 60 vs 240 d/s results are influence by the order of testing. To examine this an additional 15 patients were tested at NJNI using the same protocol except the 240 d/s rotations were performed before 60 d/s rotations. For these subjects the results were similar (Table 2). Gains were higher and time constant were longer for rotations at 60 d/s than for rotations at 240 d/s. Therefore the order of testing does not seem to be a factor.
Table 2 - Step rotations to 240 d/s preceded step rotations to 60 d/s - 15 patients
|
Parameter
|
How Often
|
p
|
Mean
|
Mean
|
|
Gain - 60 d/s > 240 d/s
|
72%
|
<<0.002
|
60 d/s - 0.49
|
240 d/s - 0.43
|
|
Tc - 60 d/s > 240 d/s
|
77%
|
<<0.0001
|
60 d/s - 17.6s
|
240 d/s - 14.2s
|
Discussion
From this data it appears that horizontal head position during step rotations
does not affect the VOR. Increasing velocity of rotation affects the VOR
by decreasing the Tc, increasing the MBS and the MBF. Varying step velocities
between 60 and 180 d/s does not seem to affect the gain of the VOR but
a step velocity of 240 d/s results in a lower gain. A possible explanation
that the gain of the VOR is decreased at 240 d/s is that this velocity
may exceed the saturation velocity of the VOR in humans.