Abstract
Purpose:
Most individuals with infantile nystagmus (IN) have an idiosyncratic gaze angle at which their nystagmus intensity is minimized. Some adopt an abnormal head posture to use this “null zone,” and it has therefore long been assumed that this provides people with nystagmus with improved visual acuity (VA). However, recent studies suggest that improving the nystagmus waveform could have little, if any, influence on VA; that is, VA is fundamentally limited in IN. Here, we examined the impact of the null zone on VA.
Methods:
Visual acuity was measured in eight adults with IN using a psychophysical staircase procedure with reversals at three horizontal gaze angles, including the null zone.
Results:
As expected, changes in gaze angle affected nystagmus amplitude, frequency, foveation duration, and variability of intercycle foveation position. Across participants, each parameter (except frequency) was significantly correlated with VA. Within any given individual, there was a small but significant improvement in VA (0.08 logMAR) at the null zone as compared with the other gaze angles tested. Despite this, no change in any of the nystagmus waveform parameters was significantly associated with changes in VA within individuals.
Conclusions:
A strong relationship between VA and nystagmus characteristics exists between individuals with IN. Although significant, the improvement in VA observed within individuals at the null zone is much smaller than might be expected from the occasionally large variations in intensity and foveation dynamics (and anecdotal patient reports of improved vision), suggesting that improvement of other aspects of visual performance may also encourage use of the null zone.
Infantile nystagmus (IN) is a regular, repetitive, predominantly horizontal involuntary movement of the eyes. It usually develops within the first 6 months of life, resulting in ocular oscillations that are constantly present and persist throughout life. Even in the absence of any other detectable pathology, cases of IN are typically associated with a moderate reduction in visual acuity (VA).
1
For reasons that are not fully understood, the orientation of the eye in the orbit (i.e., gaze angle) affects one or more of the characteristics of the involuntary oscillations, including the amplitude, frequency, and/or waveform type.
2,3 This results in a direction of gaze in which the intensity of the oscillations is at a minimum, termed the ‘null position’ or ‘null zone.’
4 Individuals with IN whose null zone is not straight ahead will often adopt an abnormal head posture in order to place the eyes at this gaze angle,
1 thus dampening the nystagmus and often increasing the duration of foveations (the period in each cycle of the waveform during which the eyes move most slowly). This null zone may be used preferentially in many situations.
1 One might therefore presume that utilizing the null zone would cause VA to increase. Indeed, when plotted between individuals with IN, foveation duration is positively associated with VA.
5 Moreover, a study by Costa et al.
6 demonstrated that the clinical VA of children with IN (as measured using the Lea Grating Acuity Test) was significantly improved by using the null zone. A recent study by Proudlock et al. (
IOVS 2016;57:ARVO E-Abstract 980) has found similar results, reporting that changes in gaze angle (through use of the nystagmus null zone) cause significant changes in clinically measured VA.
In contrast to these findings, recent work has suggested that VA may be fundamentally limited in adults with IN,
7 meaning that treatments aiming to reduce (or even eliminate) retinal image motion associated with the eye movements are unlikely to yield large improvements to VA. This is at direct odds with the conventional view that reducing nystagmus intensity and/or increasing foveation duration will lead to improved VA. It should be remembered that, due to the retinal image motion resulting from the incessant eye movements, there is likely to be a dynamic component to the visual input in the presence of nystagmus, unlike most visual pathologies, which are static. As a result, VA (an exclusively spatial measure of the resolving power of the visual system) cannot provide a complete account of the visual experience in those with IN. Temporal factors, such as cycle-to-cycle variability in foveation position (which is known to be correlated with clinical VA between individuals
8,9), are also likely to have an impact on visual performance. In the clinic, the time taken to make a measurement of VA is not standardized. Factors specific to IN may affect how long it takes to achieve a VA threshold. This may explain why some clinical studies report a link between nystagmus characteristics and VA, whereas others do not. In studies that have measured VA using a psychophysical protocol, such as a forced choice staircase in which the participants have unlimited time to achieve their threshold resolution, modifications to the nystagmus waveform have repeatedly failed to elicit significant changes in VA.
10–12 On the other hand, therapeutic studies that measure VA using clinical letter charts frequently report changes in acuity.
13,14
Between individuals, VA is known to correlate with characteristics of the nystagmus waveform, such as foveation duration and accuracy.
5,15–17 Furthermore, several studies have investigated, in normally sighted individuals, the relationship between VA and foveation duration in simulated nystagmus waveforms (i.e., the test stimulus is moved in such a way as to mimic nystagmus).
18–21 The data from each of these studies are presented in
Figure 1, and clearly show an exponential relationship between simulated foveation duration and VA
across individuals; that is, VA improves with foveation duration.
In the present study, we aimed to determine the extent to which use of the null zone (as opposed to other gaze angles) affects VA in adults with IN, using a staircase protocol. Although lengthy in duration, these psychophysical techniques provide a more accurate visual resolution threshold than standard clinical testing, due to repeated measurement and the explicit lack of time constraints. In order to achieve this, we displayed visual targets at three horizontal gaze angles (null zone and two positions away from the null, including straight-ahead) to provoke changes in the participants' eye movements, and measured the threshold VA at each position while simultaneously recording eye movements.
Eight individuals with idiopathic IN participated in the study (three females; 20–50 years [mean age, 33]). The diagnosis of IN as reported by the participant or their ophthalmologist was investigated by an optometrist using high-speed eye movement recording, ophthalmoscopy, color vision testing, slit-lamp examination, and a detailed family history. No participants reported being under medical treatment or having undergone previous surgery for nystagmus. Clinical VA was measured using a self-illuminated Bailey-Lovie chart; participants were given as long as they wished to view the chart and were encouraged to continue reading until at least four letters on a line were incorrectly identified. Participants with any comorbid visual pathology besides nystagmus were excluded (one participant from an original total of nine was excluded due to previous retinal detachment). The investigation was carried out in accordance with the Declaration of Helsinki; informed consent was obtained from the participants after explanation of the nature and possible consequences of the study. The Cardiff School of Optometry and Vision Sciences Research Ethics Audit Committee granted approval for this study.
Participants were fitted with a head-mounted 1000 Hz eye tracker (IRIS; Skalar Medical BV, Delft, The Netherlands) and seated at a table with a chin/headrest. The head was comfortably restrained with foam inserts placed beside the temples. A computer-controlled rotational mirror system was used to calibrate the eye tracker. The experimental equipment and calibration method have been described previously.
22 Following calibration, the foam inserts were removed, and the null position (rounded to the nearest 5°) for each participant was determined by asking participants to view a Landolt C target presented in the center of a 17″ monitor at an optical distance of 7 m, using the head posture with which they could most easily view the target. This gave a reading from the IRIS system of orbital eye position, indicating the amount of head turn required to view the target most comfortably.
All participants were made familiar with the psychophysical staircase procedure before recording began. The foam inserts were returned to the headrest to stabilize the head, and participants were asked to locate the gap in a single Landolt C, using a two-alternative forced choice paradigm (gap left or gap right). The starting size optotype was 0.40 logMAR above each participant's best clinical VA. The presentation of subsequent Landolt C targets followed a staircase procedure using a fixed step size of 0.075 logMAR and a three-up/one-down criterion. The staircase terminated after the criteria of 80 presentations and eight reversals had been satisfied. Visual acuity was estimated as the mean of the final six reversals.
23 Participants performed the task at three gaze positions: their null position, primary gaze, and one other eccentric gaze position, chosen to represent a wide range of viewing angles. In the one participant whose null position coincided with straight-ahead, two eccentric gaze positions were used. Eye movements were recorded throughout. Gaze angles were achieved by using the computer-controlled rotational mirror system to present the stimulus at specific angles of gaze (see
Fig. 2).
Regression analyses of the resulting data set were performed using SPSS for Windows (SPSS, Inc., Chicago, IL, USA). The changes to waveform characteristics (amplitude, frequency, foveation duration, and variability of foveation position) elicited by varying gaze angle were compared to the change in VA obtained both across and within participants.
Grouping data from all participants, amplitude exhibited a significant linear relationship with VA (R2 = 0.33, F1,22 = 10.82, P = 0.003). Approximately 33% of the variance in VA can be accounted for by nystagmus amplitude. No significant correlation (linear or exponential) between VA and nystagmus frequency was evident in this group of participants.
Again, grouping data from all participants, standard deviation of foveation position showed a significant linear relationship with VA (
R2 = 0.27,
F1,22 = 8.24,
P = 0.009;
Fig. 4C). The relationship between foveation duration and VA (
Fig. 4D) can be described by an exponential function with the following equation:
The time constant of this function is 30 ms, which is within the range of time constants previously reported by Chung and Bedell
19 and others in studies in which normally sighted individuals were exposed to stimuli with motion simulating nystagmus waveforms
18–21 (see
Fig. 1). Thus, 95% of the total VA change occurred after three times the exponential time constant. Data across participants in our study indicate that maximal VA should be achieved with foveation durations of 90 ms or longer.
Conducting a regression ANOVA revealed a significant relationship between foveation duration and VA across individuals (R2 = 0.58, F1,22 = 30.72, P < 0.0001). Indeed, nearly 60% of the variation in VA can be accounted for by foveation duration.
In order to determine whether there was a within-participant effect of gaze angle on VA, the change in VA was plotted against the change in each parameter of the nystagmus waveform at and farthest away from the null zone. These are shown in
Figure 5.
Using a linear mixed model analysis, none of the five nystagmus parameters (amplitude, frequency, intensity, foveation duration, or foveation position variability) showed a significant relationship with VA in the eight participants. Nonetheless, paired samples t-tests examining VA in the null zone and at the two other recorded gaze angles (i.e., away from null and then farther from the null zone) showed statistically significant improvements in VA (0.05 logMAR: P = 0.046, and 0.08 logMAR: P = 0.015, respectively).