Adaptation of the Vestibulo-Ocular Reflex in Amblyopia

Transcription

1 Adaptation of the Vestibulo-Ocular Reflex in Amblyopia Carol A. Wesfall and Clifton M. Schor Adaptation of the vestibulo-ocular reflex (VOR) is demonstrated by changes in gain in response to discrepancies between visual and vestibular stimulation. The authors have investigated the effect of monocular asymmetries of OKN in amblyopia upon adapting the gain of the VOR. Adaptation was investigated by modifying the horizontal balance of the VOR. While monocularly fixating a head referenced spot for 2 min, subjects were sinusoidally oscillated on a chair inside an optokinetic drum that rotated in one direction (left or right) at the peak velocity of sinusoidal chair rotation. The VOR was then measured during continued sinusoidal body oscillation in darkness for 1 min. Imbalance of the horizontal VOR gain equalled the ratio of slow phase velocities in the rightward and leftward directions. After rotating the drum in the nasalward direction, an increase was observed in slow phase gain of the VOR in the nasalward direction for either of our amblyopes that was significantly greater than similar changes in gain for the normals. Increased VOR gains for the amblyopic group following temporalward stimulation were significantly less than the nasal aftereffect. Gain changes of the VOR in normals had similar magnitudes following nasal or temporal stimulation. These results suggest that disturbances of OKN in amblyopia are common to the pathways that modify the slow phase gain of the VOR. Invest Ophthalmol Vis Sci 26: , 1985 The optokinetic and vestibular systems function to minimize movements of the visual world on the retina. The vestibulo-ocular reflex (VOR) produces compensatory movements to head rotation with a latency that can be as short as 12 ms. 1 Optokinetic nystagmus (OKN) and pursuit movements minimize retinal image movements caused by visual field movement or body movement. The latency of the closed-loop optokinetic response is 80 ms, 1 making the short-latency open-loop vestibular reflex essential for clear vision during even the smallest of body movements in a normally sighted person. VOR gain can be modified by altering the magnitude or direction of the retinal slip associated with body movement. The VOR can be changed dramatically by wearing left-right reversing prisms. 2 It has also been changed in human after short term (1 min) adaptation to magnifying or reducing spectacles 3 and in monkey after wearing magnifying lenses or "world stabilizing" glasses. 4 The combination of sinusoidal body oscillation From the University of California, School of Optometry, Berkeley, California. Supported by National Eye Institute grant EYO-3532 (C.S.) and National Institutes of Health core grant # Submitted for publication: January 8, Reprint requests: Carol Westall, Department of Optometry, University of Wales Institute of Science and Technology (UWIST), Colum Drive, Cardiff, CF1 3EU, Wales. and a unidirectional horizontal visual stimulus increases VOR gain in the direction of the previous visual field movement and decreases it in the opposite direction. 5 Visual development and deprivation also influence the gain of the VOR. The VOR is present in newborn infants, 6 but visual experience is required to maintain an adequate vestibulo-ocular response to body movement, as is evidenced by the inadequate VOR in adults blind from an early age. 7 Several observations of movements in amblyopia and strabismus suggest that monocular deprivation and abnormal binocular vision might also be associated with disturbances of the VOR. In amblyopes and strabismics, the slow phase gain of monocular optokinetic nystagmus is reduced to temporally moving stimuli. 8 An aftereffect of visual field movement, motion after-nystagmus (MAN), is likewise impaired. MAN is weak or even reversed following adaptation to temporalward motion. Like optokinetic after-nystagmus (OKAN), MAN is measured in darkness after viewing an optokinetic stimulus. Unlike OKAN, in which the preceding optokinetic movements are unrestrained, MAN follows a period of viewing the stimulus while movements are suppressed by fixating a small stationary dot on the stimulus screen. It was suggested that an anomaly in the afferent pathways for OKN explained the MAN data. 9 In the present study we found that, following short term adaptation to a conflicting visual vestibular stim- 1724

2 No. 12 ADAPTATION OF VESTIBULO-OCULAR REFLEX IN AMDLYOPIA / Wesroll ond Schor 1725 ulus, the VOR could be modified in all subjects tested. Adaptation of the VOR was asymmetric in amblyopic subjects. Their VOR gain increased more following adaptation to nasalward than temporalward field motion. s Materials and Methods Eye movements often strabismic amblyopes, some of whom had anisometropia, and seven normal control observers were examined. Visual acuity of amblyopic s ranged from 20/40 to 20/400, and that of nonamblyopic s was 20/20 or better. Each subject was informed of the nature of the procedures, and their informed consent was obtained prior to their participation in this study. chair position position Baseline VOR bias DS 10* Eye Movement Recording Eye movements were monitored with a corneal reflection infrared sensing device (Gulf Western SGVH/ 2; Waltham MA). Eye movements were recorded on FM tape and on a strip chart recorder. The accuracy of the movement recording was 10 arc min. The Stimulus With head restrained by a head rest, subjects were seated on a chair that could be mechanically oscillated about a vertical axis inside a surrounding 4-ft diameter white translucent optokinetic drum. The constant angular velocity of the drum (15 deg/sec) could be matched to the peak velocity of sinusoidal chair rotation (.3 Hz). In a prior investigation of stimulus frequency and velocity dependence of the VOR, 9 we found that this stimulus combination was very effective in revealing VOR imbalance which is best seen with low frequency and high velocity combinations. Random black dots (2 to 10 deg) lined the surface of the drum. The room lights were on during the optokinetic stimulation, and the aftereffects of the stimulus were measured in total darkness. Procedure The bias of the VOR was quantified from the movements recorded when observers were horizontally oscillated in total darkness for 30 sec. Observers were seated in a chair that rotated sinusoidally (20 deg peak to peak) at 0.3 Hz with a peak velocity of 15 deg/sec about the vertical axis of the chair. The resulting VOR is shown in Figure 1, along with a record of chair position and a cumulative plot of the slow components of the VOR. The cumulative plot was derived manually by subtracting the saccadic movements from the orig- slow phase position A/V 5 seconds Fig. 1. Baseline VOR bias: VOR bias for amblyopic (right). Top trace shows chair position record; middle trace shows position record; lower trace shows a continuous plot of the VOR slow phase position. Leftward and rightward directions for movements are indicated by the L and R at the ends of the scale. inal recording of the VOR. The peak velocities of the left and right slow phase movements were measured from slopes fit to the steepest portion of the cumulative plots. Means of these peak velocities had errors of less than 10%. The imbalance or bias of the VOR was defined as the ratio of mean rightward to leftward slow phase velocities. Adaptation of the VOR s were sinusoidahy oscmaieu as aoove wiin the room lights on, and with one occluded. During this adaptation period, the optokinetic drum was rotated about the chair at the same peak velocity as the chair in one direction. In order to avoid contamination of the results with asymmetric nasal and temporal optokinetic nystagmus, movements were suppressed during adaptation with a head referenced fixation spot. Thus the retinal slip was continually changing from as little as zero when the chair rotated with the drum to

3 1726 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / December 1985 Vol. 26 a maximum of twice that, due to the chair alone, when the chair reached its peak velocity in the opposite direction to the drum rotation. Maximum movement of the amblyopic s during this fixation suppression procedure was less than one deg. 9 The adaptation period was two min. This was long enough to observe an effect of the adaptation, but not too long to cause the subject to become drowsy. This short duration of adaptation was used to minimize secondary afternystagmus which is fully manifest in darkness after 4 min of optokinetic stimulation. 10 Since the deficit of the temporalward slow phase response of OKN in amblyopia is most apparent with small field sizes, 8 we limited the field size to 40 deg horizontally and 30 deg vertically. We used a head stationary rectangular mask to block out the peripheral field. The room lights were extinguished immediately following the 2 min of combined vestibular and optokinetic stimulation, and the VOR was measured for 60 sec in total darkness. Data was analyzed from the last 30 sec in darkness during which there were no motion or optokinetic effects on afternystagmus. The adaptation was always monocular. We first tested the amblyopic (or non-dominant of normals) with the drum moving in the temporalward direction. ^This was to ensure that the strong aftereffect of nasalward adaptation did not influence the results. We then examined the effect of nasalward adaptation to the same. The subject rested for 3 min between trials looking at stationary objects to discharge any remaining adaptation." The effects of unidirectional adaptation were then examined in the second, first in the temporalward direction and then in the nasalward direction. The Reappearance of the VOR Bias following the Decay of Afternystagmus It was important to differentiate between the possibilities that (1) the VOR had undergone short term modification, and (2) the apparent modification of the VOR was simply the addition of a spontaneous afternystagmus to the VOR in the dark. In a previous investigation 9 the same amblyopes studied here viewed a visual stimulus rotating horizontally in one direction. Lights were extinguished and movements were measured in darkness, with no body oscillation. The velocity of the afternystagmus decreased to the subjects baseline drift bias velocity in less than 30 sec. The previous study used a different apparatus and stimulus conditions than were used in the current investigation. A control experiment was conducted in the current study, using the same subjects, to determine if there were significant differences between durations of afternystagmus in darkness (measured with and without vestibular stimulation) following 2 min of adaptation to combined unidirectional visual motion and sinusoidal vestibular stimulation. In darkness with body stationary there was an afternystagmus that decayed as in the previous study. The duration of the aftereffect (measured in the current study) was quantified in four amblyopes and four normal subjects. Durations without body movement in darkness were 18, 22, 12, and 16 sec for four amblyopic subjects. If after adaptation, the subjects were sinusoidally oscillated in darkness, at the same amplitude and frequency as during adaptation, there was a bias of the VOR in the same direction as the previous drum rotation, that lasted longer than aftereffects without body rotation. Durations with body movement, in darkness were 55, 125, 72, and 60 sec for the same amblyopic subjects. Similar results for this control study were found for non-amblyopic subjects. Durations for the first dark condition for 4 normal subjects were 10, 15,20, and 16 sec; durations increased for the second dark condition to 50, 57, 65, and 77 sec for the same normal subjects. Accordingly, results of adaptation were analyzed after the first 30 sec of VOR in darkness in order to avoid their contamination by afternystagmus. Baseline Bias of the VOR Results The ratio of mean rightward velocity to mean leftward velocity of the sinusoidal VOR was found for each subject in darkness before any adaptation. This ratio identified the direction and magnitude of the VOR bias. A ratio that was greater than one demonstrated a rightward bias, and a ratio that was less than one demonstrated a leftward bias. The ratios are shown in the first numerical column in Table 1. Five amblyopes and three normals had rightward biases of their VOR. The rest had leftward biases. Effect of Short-term Visual Adaptation on the VOR Figure 2 shows the movement traces for an amblyope ( 8). The top trace represents the chair position. The next trace shows the baseline VOR bias shown in Figure 1. The third trace shows the modified VOR after adaptation to temporalward (rightward) motion. Following adaptation the velocity increased to the right and decreased to the left. This is seen clearly in Figure 3, where the cumulative plots of the movement records from Figure 2 are plotted. The lower cumulative trace of Figure 3 shows the VOR modified in the temporalward direction. The lower trace in Figure 2 shows the pronounced aftereffect of adaptation to nasalward motion. The slow phase

4 No. 12 ADAPTATION OF VESTIBULO-OCULAR REFLEX IN AMDLYOPIA / Wesroll ond Schor 1727 Table 1. Slow phase velocity bias Amblyopes Normals Preadaptation A-T B-T A-N B-N Preadaptation C-T id-t i C-N i D-N * 6* 7 8* 9* 10* lit.12t 13f 14f 15f 16* 17f A = Amblyopic. B = Non-amblyopic. * Amblyopic = right. N = drum in nasal direction. T = temporal direction. C = non-dominant. D = Dominant. t Dominant = right. t Neither dominant, right classified as dominant. movements produced from this trace are shown in the middle of Figure 3. The ratio of the mean rightward velocity to the mean leftward velocity of the VOR was calculated after adaptation to nasaiward and temporalward drum rotations. After adaptation of either to rightward motion, the rightward and leftward slow phase velocities of the VOR increased and decreased respectively relative to the preadaptation level. The ratio of the right mean velocity to left mean velocity was thus larger. After adaptation to leftward drum rotation the ratio was less than the preadaptation ratio in all but one case ( 9 non-amblyopic to temporalward motion). The amount of change in the VOR was calculated by dividing the postadaptation bias of the VOR (rightward/leftward slow phase velocity) by the preadaptation bias after rightward adaptation. The inverse division was used to calculate the amount of change in the VOR for leftward adaptation. That is, when the modification was in the same direction as the preceding stimulus motion, the amount of change from preadaptation ratio to postadaptation ratio was calculated by dividing the largest of the two ratios by the smallest. Thus modified VOR biases after adaptation to rightward and leftward stimulus movement are on equivalent scales around 1. Table 2 shows these ratios which represent the degree of modification of the VOR. Values greater than 1 indicate a modification of the VOR in the appropriate direction for the preceding stimulus. Our results demonstrated short-term modification of the VOR after visual adaptation of normal and amblyopic observers. Statistical Analysis Data variability was greater for amblyopes than for normals. We used a non-parametric statistical analysis because parametric analysis would have been invalid for the unequal variance of our populations and small sample size. The Wilcoxon test for independent means chair position vor after temporal adaptation vor after nasai adaptation 5 seconds DS-OD Fig. 2. Modified VOR: Eye position trace of amblyopic subject before visual adaptation (second trace), after adaptation to temporalward motion (third trace), and after adaptation to nasaiward motion (lower trace). The modification of the VOR is seen clearly by the direction of the predominant slow phase movements.

5 1728 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / December 1985 Vol. 26 baseline bias VOR slow phase position DS-OD was the statistic used to compare our two populations. The Wilcoxon matched pairs statistic was used to compare results within subjects. Comparison of Dominant and Non-dominant Eyes Table 3 compares the aftereffects of a specific direction of adaptation for (1) the non-dominant of a normal subject versus the amblyopic of an amblyopic subject; (2) the dominant of a normal subject versus the non-amblyopic of an amblyopic subject. The amount of modification was actually greater (P < 0.01) after nasalward adaptation of the amblyopic compared to the non-dominant s of normal observers. After temporalward adaptation, there was no significant difference in the amount of modification of the VOR between the two groups. The comparison of the dominant s of normal subjects to the non-amblyopic of amblyopes revealed marginally more (P < 0.075) modification of the VOR after nasalward stimulus movement. Again there was no significant difference between the two groups after temporalward stimulus movement. Table 3 compares the amount of modification of the VOR between the amblyopic and non-amblyopic s of amblyopic subjects. There was no difference in VOR modification following nasalward or temporalward adaptation between the two s of these subjects. 5 seconds Fig. 3. VOR slow phase position: Modification of VOR demonstrated by comparing slow phase position of the baseline VOR (upper trace) to the modified VOR biases. These traces are derived from those in Figure 2. Comparison of Normal and Amblyopic Eyes of Amblyopes Asymmetry of the VOR was evaluated by comparing nasalward to temporalward adaptation (Table 4). In amblyopic s, the modification of the VOR was compared in response to nasalward and temporalward Table 2. Ratios of vor modification to nasalward Amblyopes and temporalward visual motion Normals A-T B-T xt A-N B-N xn C-T D-T xt C-N D-N xn : ' ; : > >.58 :' A = Amblyopic. B = Non-amblyopic. C = Non-dominant. D = Dominant. N = Drum in nasal direction. T = Drum in temporal direction.

6 No. 12 ADAPTATION OF VESTIBULO-OCULAR REFLEX IN AMDLYOPIA / Wesroll and Schor 1729 visual motion. There was significantly more modification to nasalward than temporalward stimuli (P < 0.01). No such difference was found for the nonamblyopic s of amblyopic subjects. When the mean nasalward, (R nasal + L nasal)/2, and mean temporalward, (R temporal + L temporal)/2, modification of both s of our amblyopes were considered, then the difference in mean modification to nasalward and temporalward motion became significant (P < 0.01). This was because those amblyopic s that adapted less to nasalward than to temporalward motion were associated with a reverse effect for the non-amblyopic s. There was more adaptation to nasalward than temporalward motion for these non-amblyopic s. Thus, although the amblyopic was mainly responsible for the significant nasal-temporal asymmetry, the non-amblyopic s of other amblyopes contributed to the highly significant difference between the adaptation of the VOR to the two directions of motion. For normal s there was no significant difference between the after effects of nasalward and temporalward motion upon the VOR. Amblyopic s were not consistently more asymmetric than non-amblyopic s, indicating that the acuity loss of the amblyopic is not the cause for the asymmetry. The breakdown in binocular connections in strabismic amblyopia must be associated with an inability to modify the VOR equally in response to nasalward and temporalward stimulus movement. Discussion In this study we modified the baseline VOR bias by horizontal visual stimulation during sinusoidal body oscillation in normal and amblyopic humans. Our experiment was to assess the association between known asymmetries of the optokinetic system to the adaptability of the vestibular system. Amblyopes show a deficit in the monocular optokinetic response to temporally moving targets in either. 8 Motion after nystagmus is likewise affected. The asymmetry appears to result from a disturbance of a motor control mechanism that processes temporally moving stimuli. 9 In this experiment we demonstrated that amblyopes showed significantly more adaptation in the nasalward direction than in the temporalward direction. There was no significant difference between the s of our amblyopes. These aftereffects to nasalward motion were greater in amblyopes than in normals. Although humans blind from an early age show a severe reduction in the gain of their VOR 7 and monocularly deprived cats are unable to modify their VOR, 5 amblyopes showed modification after both nasalward and tern- Table 3. Statistical testing for differences between normal and amblyopic subjects and between amblyopic and non-amblyopic s of amblyopic subjects Direction of adaptation Normal nondominant Normal dominant Amblyope amblyopic Variables Amblyope amblyopic Amblyope nonamblyopic Amblyope nonamblyopic Significance <0.01 <O.O75 poralward adaptation of either, although of unequal magnitude. The nature of the asymmetry does not appear to be the addition of a constant error or drift bias such as MAN. The short-term modification of the VOR was found to be dependent on semicircular canal stimulation in the dark and not on the static aftereffect of visual-vestibular stimulation. Of all our subjects, the longest duration for the static aftereffect (motion afternystagmus) was 22 sec. Following the decay of this static aftereffect, the vestibular bias reappeared with body oscillation. Other subjects had even shorter decays of MAN. In no case did MAN persist longer than 22 sec, and with this short adaptation period (2 min) no subject demonstrated inverted or negative MAN after the positive phase of the aftereffect. Consequently, the measures of VOR imbalance after the first 30 sec in darkness were taken after MAN had decayed to zero. Table 4. Statistical testing for differences between nasalward and temporalward adaptation Eyes Amplyopic Non-amblyopic Amblyope mean both s Normal mean both s Variables Significance <0.01 <0.01

7 1700 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / December 1985 Vol. 26 The change in velocity caused by VOR adaptation was also greater than the velocity of MAN. VOR changes ranged between subjects from 2 to 4 deg/sec while MAN ranged from 0.13 to 2.04 deg/sec. It is possible that the aftereffects causing MAN and OKAN are still operating even after MAN decays to zero and that they interact with the VOR. MAN is classified as a static aftereffect of visual motion stimulation. Here we demonstrated a dynamic aftereffect because it was only manifested during body oscillation. Perhaps a common process underlies both the static and dynamic aftereffects. During the adaptation period, persistent retinal slip during head turns provided afference to the optokinetic system and the vestibulo-cerebellum. It has been suggested that modification of the VOR is brought about by a change in the variable gain element in the brainstem, initiated by feedback from the vestibulo-cerebellum The aftereffect of persistent retinal slip in one direction showed up in darkness; the gain of the vestibular movements increased in that direction while the movement velocity in the opposite direction tended to be decreased. Because the optokinetic response was also modified, as was evidenced by optokinetic after nystagmus, 9 it can be assumed that the modification took place in the later stages of the vestibular pathway, beyond the vestibular nuclei where it is thought that the optokinetic and vestibular pathways converge. 13 The effectiveness of short-term modification of the gain of the VOR in amblyopes reflected the monocular asymmetry found in the OKN response. The results were paradoxical in that we showed more adaptation after nasalward stimulation, rather than less adaption after temporalward stimulation. Perhaps the decreased afference to temporal pathways of the optokinetic system in amblyopia caused an overall compensatory enhancement in VOR plasticity, resulting in increased nasal adaptation and normal temporal adaptation. Key words: vestibulo-ocular reflex, visual-vestibular interactions, adaptation. of vestibulo-ocular reflex, amblyopia, optokinetic nystagmus References 1. Cohen B, Henn V, Raphen Th, and Dennett D: Velocity storage, nystagmus, and visual-vestibular interactions in humans. Ann New York Acad Sci 374:421, Gonshor A and Melvill Jones G: Extreme vestibulo-ocular adaptation induced by prolonged optical reversal of vision. J Physiol 256:381, Collewijn H, Martins AJ, and Steinman RM: Compensatory movements during active and passive head movements: fast adaptation to changes in visual magnification. J Physiol 340:259, Miles FA and Eighmy BB: Longterm adaptive changes in primate vestibulo-ocular reflex 1. Behavioral observations. J Neurophysiol 43:1406, Harris LR and Cynader M: Modification of the balance and gain of the vestibulo-ocular reflex in the cat. Exp Brain Res 44:57, Ornitz EM, Atwell CW, Walter DO, Hartman EE, and Kaplan AR: The maturation of vestibular nystagmus in infancy and childhood. Acta Otolaryngol 88:244, Sherman KR and Keller EL: Vestibulo-ocular reflexes in blind subjects. ARVO Abstracts. Invest Ophthalmol Vis Sci 22(Suppl): 271, Schor CM and Levi DM: Disturbances of small-field horizontal and vertical optokinetic nystagmus in amblyopia. Invest Ophthalmol Vis Sci 19:668, Schor CM and Westall CA: Visual and vestibular sources of fixation instability in amblyopes. Invest Ophthalmol Vis Sci 25: 728, Brandt T, Dichgans J, and Buchle W: Motion habituation: inverted self-motion perception and optokinetic afternystagmus. Exp Brain Res 21:337, Lisberger SG, Miles FA, Optican LM, and Eighmy BB: Optokinetic response in monkey: underlying mechanisms and their sensitivity to long-term adaptive changes in the vestibulo-ocular reflex. J Neurophysiol 45:869, Precht W: Visual-vestibular interactions in vestibular neurones: functional pathway organization. Ann New York Acad Sci 374: 230, Miles FA and Lisberger SG: Plasticity in the vestibulo-ocular reflex: a new hypothesis. Ann Rev Neurosci 4:273, 1981.