Binocular Coordination of Saccades During Reading In Strabismic Children

Abstract

Purpose.: Toour knowledge, studies comparing binocular eye movements during reading task in strabismic children are scarce. The goal of our study was to examine binocular coordination of saccades during reading in strabismic children.

Methods.: Binocular eye movements were recorded by an infrared system (mobile EBT) in 18 strabismic children 6.8 to 16 years old (mean age 10.2 ± 3) who were asked to read a four-line text silently during binocular viewing. Data were compared to that of 18 age-matched non-strabismic control children.

Results.: Saccade amplitude was similar in strabismic and control children. In contrast, binocular coordination during and after the saccades was significantly poorer in strabismic children as opposed to control children. The duration of fixation also was significantly longer in strabismic children compared to controls. Binocular coordination in strabismic children who have binocular vision was better than those without binocular vision, but it still was worse than in control subjects.

Conclusions.: In strabismic children binocular saccade coordination is deficient and could be responsible for impaired reading capabilities. Binocular vision has an important role in improving binocular saccade yoking.

Introduction

Reading is a major task, and is essential for social and professional integration. While reading, eyes perform saccadic movements and fixations. Saccades are fast ballistic and preprogrammed movements that bring a target image to the fovea quickly; during a saccadic movement both eyes move in the same direction and with similar amplitude. 

Approximately 2% of children under 7 years old suffer strabismus. ¹ In many cases, it is responsible for abnormal alignment of the eyes and abnormal binocular vision. Treatment of strabismus must avoid amblyopia, and aims at restoring proper visual alignment with surgical treatment, if necessary, and if possible, at restoring normal binocular vision. 

Studies of binocular coordination during reading in children with or without strabismus are rather scarce. Bucci and Kapoula compared binocular coordination while reading isolated words, and when fixating single targets in eight 7-year-old children and in eight adults, both groups without strabismus. ² It was shown that the amplitude of the disconjugacy during and after the saccades was larger in children than in adults, regardless of the stimulus. The duration of fixation after the saccade also was longer in children than in adults, particularly when reading a single word. These results suggest that a large discoordination of saccades in children could explain a poorer identification of the word during the reading task and, therefore, a longer postsaccadic fixation than is required to identify the word properly while reading. Blythe et al. examined the binocular coordination of saccades during reading in 12 children (7–11 years old) and 12 young adults (18–21 years old). ³ They found that the disconjugacy at the beginning and end of the fixation was significantly larger in children than in adults. Reading skills mature with time, and it is hypothesized that cortical structures (e.g., frontal and parietal cortex) involved in eye movement control as well as those involved in linguistic processing (i.e., left temporal and parietal cortex) are developing during childhood and adolescence. ^4–6 Stifter et al. examined the reading performance in children with microstrabismic amblyopia. ⁷ They compared the reading performance (reading speed, reading acuity, and visual acuity) monocularly and binocularly in 20 children with microstrabismic amblyopia and in 20 normal non-strabismic children (aged 11.6 ± 1.4 years). In the monocular and binocular conditions, normal children display a higher reading speed than amblyopic children. This emphasizes the importance of testing reading as a key tool in evaluating visual capabilities in amblyopic/strabismic children. In adults, Kanonidou et al. examined reading performances under binocular and monocular vision in 20 subjects with strabismus and amblyopia. ⁸ They measured the reading speed, number of progressive and regressive saccades per line, saccade amplitudes, and fixation duration. Similarly to children, amblyopic adults' reading speed is slower than in normal subjects. This is true when testing the amblyopic eye (monocular viewing) or binocular viewing. Furthermore, Kanonidou et al. reported more regressive saccades and longer fixation durations, but similar saccadic amplitudes in strabismic subjects with respect to controls. ⁸ These findings, together with results from patients with central field loss ⁹ or from normal subjects with simulated central scotomas, ¹⁰ are in line with the hypothesis that a reduced or impaired visual span leads to impaired reading capabilities. 

The first objective of our study was to examine the binocular coordination of saccadic eye movements during reading in strabismic children and to compare these data with that recorded in a group of non-strabismic control children of similar reading age. Our driving hypothesis, based on previous studies from our research group, is that impaired binocular sensory capabilities could cause poorer binocular motor coordination during and after the saccades leading to delayed reading capability. ^11,12  

Materials and Methods

Subjects

A total of 18 strabismic children between 6.8 and 16 years old (mean age 10.2 ± 3 years) participated in the study. Strabismic children were recruited from the Department of Ophthalmology, Robert Debré Children Hospital in Paris. We also tested 18 age-matched control children (mean age 10.1 ± 2.9 years). All children were native French speakers and had no known reading difficulties. 

The investigation adhered to the principles of the Declaration of Helsinki and was approved by our institutional Human Experimentation Committee. Informed parental consent was obtained for each subject after the nature of the procedure had been explained. 

Table. 

 

Clinical Characteristic of Strabismic Children

Children, y	Glasses Correction	Corrected Visual Acuity	Angle of Strabismus (Prism Diopters)	Stereoacuity (TNO)	Type of Strabismus
C1 (7.6)	RE : −1.25 (−0.75)155°	RE : 20/20	30 X'−X'T	40″	Intermittent exotropia
	LE : −1.50 (−1.00)150°	LE : 20/20	30 X−XT
C2 (8.6)	RE : −1.25 (−1.5)140°	RE : 20/20	30 X'−X'T	60″	Intermittent exotropia
	LE : −0.50 (−2.00)0°	LE : 20/20	18−20 X−XT
C3 (9.3)	RE : +1.00 (−0.25)75°	RE : 20/20	12 X'−X'T	60″	Intermittent exotropia
	LE : +1.00 (−0.25)90°	LE : 20/20	25 X−XT
C4 (9.4)	RE : +0.75 (−0.25)175°	RE : 20/20	25 X'−X'T	40″	Intermittent exotropia
	LE : +0.75	LE : 20/20	25 XT
C5 (9.7)	RE : +6.50	RE : 20/20	10 X'−XT	40″	Intermittent exotropia
	LE : +7.50	LE : 20/20	25 X−XT
C6 (10.5)	RE : −1.00	RE : 20/20	35 X'−X'T	60″	Intermittent exotropia
	LE : −1.00	LE : 20/20	35 X−XT
C7 (11.3)	RE : +5.75(−2.25)170°	RE : 20/20	12 E'T	400″	Accommodative esotropia
	LE : +6.00 (−2.50)0°	LE : 20/20	10−12 ET
C8 (11.4)	RE : 0.00	RE : 20/20	18 X'−X'T	60″	Intermittent exotropia
	LE : 0.00	LE : 20/20	6 X−XT
C9 (13.8)	RE : +0.25	RE : 20/20	30 E'−E'T	240″	Accommodative esotropia
	LE : +0.25 (−0.25)0°	LE : 20/20	8 E−ET
C10 (15.9)	RE : +4.25(−0.25)45°	RE : 20/20	10 E'−E'T	60″	Accommodative esotropia
	LE : +3.75(−0.50)135°	LE : 20/20	1 E
C11 (6.8)	RE : −1.75 (−2.25)15°	RE : 20/20	55 E'T + 10 HTG	−	Congenital esotropia
	LE : −1.25 (−1.75)30°	LE : 20/20	55 ET + 10 HTG
C12 (7)	RE : +1.75(−1.00)165°	RE : 20/20	25 E'T	−	Congenital esotropia
	LE : +0.75(−1.00)165°	LE : 20/20	25 ET
C13 (7)	RE : +2.25	RE : 20/20	35 E'T	−	Accommodative esotropia
	LE : +5.75	LE : 20/20	30 ET
C14 (7.3)	RE : +6.75 (−2.00)15°	RE : 20/20	45 E'T	−	Accommodative esotropia
	LE : +7.25 (−2.00)10°	LE : 20/20	35 ET
C15 (8.9)	RE : +1.25	RE : 20/20	40 E'T	−	Congenital esotropia
	LE : +1.00 (−1.00)75°	LE : 20/20	30 ET
C16 (9.4)	RE : +6.00(−0.50)140°	RE : 20/20	30 E'T	−	Accommodative esotropia
	LE : +5.00 (−0.50)20°	LE : 20/20	30 ET
C17 (13.9)	RE : +2.25(−0.50)145°	RE : 20/20	45 E'T	−	Accommodative esotropia
	LE : +2.00	LE : 20/20	45 ET
C18 (16)	RE : +5.50 (−2.00)0°	RE : 20/20	6 E'T	−	Accommodative esotropia
	LE : +5.75 (−2.25)0°	LE : 20/20	2 ET

 The deviation of the eyes was assessed with the cover–uncover test and prism. The binocular vision was evaluated with the TNO test for stereoscopic depth discrimination. LE, left eye; RE, right eye; X−XT, intermittent exotropia measured at far distance (5 m); X'−X'T, intermittent exotropia measured at near distance (30 cm); E−ET, accommodative esotropia measured at far distance (5 m); E'−E'T, accommodative esotropia measured at near distance (30 cm); E'T and ET, esotropia measured at far (5 m) and at near (30 cm) distance, respectively; O and O', orthophoria measured at far (5 m) and at near (30 cm) distance, respectively.

Ophthalmologic and Orthoptic Examination

All strabismic children underwent ophthalmologic and orthoptic examination to evaluate their visual function. Clinical data of each strabismic child are shown in the Table. The monocular visual acuity was normal (≥20/20) for all children. Ten children (C1–C10) had binocular vision. For the majority of them it was in the normal range (≤60 seconds of arc with the TNO test); C7 and C9 only had reduced binocular vision of approximately 400 and 240 seconds of arc. The majority of these children (C1, C2, C3, C4, C5, C6, and C8) had intermittent divergent strabismus, while the other three children (C7, C9, and C10) had partially accommodative esotropia. For the other 8 children (C11–C18) binocular vision was not present at all; three of them (C11, C12, and C15) had congenital esotropia, while the other children (C13, C14, C16, C17, and C18) had accommodative esotropia. 

Visual functions also were evaluated in the control group. All control children had normal monocular visual acuity (≥20/20), and normal binocular vision (≤60 seconds of arc with the TNO test). None of the control children had strabismus. 

Reading Task

The reading paradigm used has been described previously, ¹³ and consisted of reading a four line paragraph containing 40 words and 174 characters. The size of the text was 29° wide and 6.4° high; mean character width was 0.5°, and the text was written in black “Courier” font on a white background. Children were asked to read the text silently. (7–9 years: extract of “Jojo Lapin fait des farces,” Gnid Bulton ed. Hachette, Fig. 1A; 10–12 years: “Bagarres à l'école,” Marc Cantin et Eric Gasté, ed. Castro Cadet, Fig. 1B: and 13–18 years: “La guerre des boutons,” Louis Pergaud, ed. Folio, Fig. 1C). All texts come from three different books that usually are used by French teachers in different class levels (7–9, 10–12, and after 13 years old). We chose these age-specific texts to ensure that all words were well-known and easily understood by the children. Indeed, our focus being on monitoring oculomotor behavior, we did not want the text to present any additional word understanding or processing difficulty. 

Figure 1. 

 

Reading task used for children with a reading age of 7 to 9 years (A), 10 to 12 years (B), and 13 to 15 years (C), respectively.

Children were asked to read the text silently at their own speed and raise one finger when they finished. 

Eye Movements Recording

Eye movements were recorded by a noninvasive system using infrared camera and mirror to record horizontal and vertical eye position independently and simultaneously for each eye: the Mobile EyeBrain Tracker (Mobile EBT; provided in the public domain by www.eye-brain.com; e(ye)BRAIN, Ivry-sur-Seine, France). This eye tracker is a medical device CE marked for medical purposes. Recording frequency was set up to 300 Hz. The mobile eBT is linear and the precision of the system is 0.25° during static acquisition. Calibration was done at the beginning of eye movement recordings. During the calibration procedure, children were asked to fixate a grid of 13 points (diameter 0.5°) mapping the screen. Point positions in degree in horizontal/vertical plan were −20.9°/12.2°, 0°/12.2°, 20.9°/12.2°, −10.8°/6.2°, 10.8°/6.2°, −20.9°/0°, 0°/0°, 20.9°/0°, −10.8°/-6.2°, 10.8/-6.2°, −20.9°/-12.2°, 0°/-12.2°, and 20.9°/-12.2°. Each calibration point required a fixation of 250 ms to be validated. A polynomial function with five parameters was used to fit the calibration data and to determine the visual angles. Calibration factors for each eye were determined from the eye positions during the calibration procedure (see the reports of Bucci et al. ^13,14 ). There was no obstruction of the visual field with the recording system and the calibrated zone covers a visual angle of ±22°. 

Procedure

Children were seated in a chair in a dark room, in front of a flat screen displaying the text at a fixed distance of 58 cm. The head of the child was held straight with a head-rest; viewing was binocular. 

After the calibration procedure, the reading task was presented to the child. The duration of the task was kept short (lasting a couple of minutes) to avoid head movements. 

Data Analysis

Calibration factors for each eye were determined from the eye positions during the calibration procedure (see the studies of Bucci et al. ^13,14 ) The software MeyeAnalysis (provided with the eye tracker; e(ye)BRAIN),was used to extract saccadic eye movements from the data. It determines automatically the start and end of each saccade by using a built-in saccade detection algorithm. The algorithm used to detect saccades is adapted from the report of Nyström and Holmqvist. ¹⁵ The algorithm searches for velocity peaks by identifying samples where the velocity is larger than a velocity threshold (θ > θ_PT). An iterative data-driven approach is proposed to finding a suitable threshold. The iterative algorithm is given an initial peak velocity detection threshold PT₁, which could be in the range of 100° to 300°/sec, but the choice is not critical as long as there are saccades, with peak velocities reaching this threshold. For all samples with velocities lower than PT₁ , the average (μ) and SD (σ) are calculated. The threshold is updated as $P{T}_n={\mu }_{n-1}+6{\sigma }_{n-1}$for each iteration. For each detected saccade peak (hose detected after the last iteration), the algorithm searches backward (from the leftmost peak saccade sample) and forward (from the rightmost peak saccade sample) in time for the saccade onset and offset. Saccade onset is defined as the first sample that goes below the saccade onset threshold and where ${\theta }_i-{\theta }_{i+1}\ge 0.$ Saccadic offset is defined as the first sample that goes below the saccade offset threshold and where ${\theta }_i-{\theta }_{i+1}\le 0.$ All detected saccades were verified by the investigator and corrected/discarded if necessary.  

We analyzed all forward and backward saccades starting from or finishing on a word. Oblique saccades made to start a new line were excluded from the analysis given that the quality of disconjugacy of these saccades is not well known either in children or in adult normal populations. For each recorded saccade, we examined the amplitude of the conjugate ([left eye + right eye]/2) and the disconjugate components (left eye – right eye) during the saccade. The disconjugacy was measured as the change in vergence between the beginning and end of each saccade. We also examined the disconjugate component of the postsaccadic drift and the fixation duration over the period between two saccades. Figure 2 shows an example of binocular recordings from a strabismic child (C13) and a non-strabismic age-matched child. 

Figure 2. 

 

Data analysis. Binocular recordings of saccades during reading from a strabismic child (C13) and non-strabismic child (9.5 years old). LE, left eye; RE, right eye.

Data were entered in an ANOVA using the four groups of children (strabismic with and without binocular vision, and non-strabismic age-matched children) as intersubject factor and as a fixed factor, the individual means of the saccade amplitude, of the disconjugacy measured during and after the saccades, and of the duration of fixation. As we showed previously that the quality of binocular coordination during and after the saccades is age-dependent, ^13,14 two groups of age matched non-strabismic children were used to perform the analysis. The effect was considered significant when the P value was below 0.05. 

Results

Saccade Amplitude

Figures 3A and 3B show the mean amplitude of the saccades in strabismic children with and without binocular vision and in control age-matched children. There was no significant group effect (F_{(3, 32)} = 1.71, P = 0.18). This suggested that the mean amplitude of the saccades of strabismic children with and without binocular vision (mean 4.9 ± 0.44° and 4.9 ± 0.50°, respectively) was not different from what was observed in control age-matched children (mean 4.5 ± 0.3° and 3.8 ± 0.20°, respectively). 

Figure 3. 

 

Saccade amplitude. Individual mean of saccade amplitude for strabismic children with and without binocular vision (A), and for non-strabismic age-matched children (B). Vertical bars indicate the SE.

Saccade Disconjugacy

Figures 4A and 4B show the mean disconjugacy of the saccades in strabismic children with and without binocular vision, and in control age-matched children. The ANOVA test showed a significant group effect (F_{(3, 32)} = 70, P = 3.6 × 10⁻¹⁴). Post hoc comparison showed that the saccade disconjugacy was significantly higher in strabismic children with and without binocular vision (mean 0.4 ± 0.04° and 1.01 ± 0.06°, respectively) than in control age-matched children (mean 0.2 ± 0.01° and 0.3 ± 0.04°). Saccades disconjugacy also was significantly higher in strabismic children without binocular vision than in strabismic children with binocular vision (P = 7.66 × 10⁻¹²). 

Figure 4. 

 

Saccade disconjugacy. Individuals mean of saccade disconjugacy for strabismic children with and without binocular vision (A), and for non-strabismic age-matched children (B). Vertical bars indicate the SE.

Postsaccadic Drift Disconjugacy

Figures 5A and 5B show the mean disconjugacy of the postsaccadic drift in strabismic children with and without binocular vision, and in control age-matched children. The ANOVA test revealed a significant group effect (F_{(3, 32)} = 41, P = 4.3 × 10⁻¹¹). Post hoc comparison shows that the postsaccadic drift disconjugacy was significantly higher in strabismic children with and without binocular vision (mean 0.5 ± 0.05° and 1.15 ± 0.1°, respectively) than in control age-matched children (mean 0.26 ± 0.01° and 0.32 ± 0.02°). Postsaccadic drift disconjugacy also was significantly higher in strabismic children without binocular vision than in strabismic children with binocular vision (P = 7.4 × 10⁻⁹). 

Figure 5. 

 

Postsaccadic drift disconjugacy. Individual mean of the postsaccadic fixation disconjugacy for strabismic children with and without binocular vision (A), and for non-strabismic age-matched children (B). Vertical bars indicate the SE.

Duration of Fixation

Figures 6A and 6B shows the duration of fixation in strabismic children with and without binocular vision and in control age-matched children. The ANOVA test showed a significant group effect (F_{(3, 32)} = 6.36, P = 0.001). Post hoc comparison showed that the duration of fixation was significantly longer in strabismic children with and without binocular vision (mean 428.5 ± 20.2 ms and 428.4 ± 32.9 ms, respectively) than in control age-matched children (mean 292.5 ± 28 ms and 340.9 ± 28.93 ms). 

Figure 6. 

 

Fixation duration. Individual mean of fixation duration for strabismic children with and without binocular vision (A), and for non-strabismic age-matched children (B). Vertical bars indicate the SE.

Discussion

The main findings from our study were as follows: While reading a text, saccade amplitudes were similar in strabismic children and in control children. Strabismic children had poorer binocular coordination during and after the saccades than normal age-matched children. Saccadic disconjugacy and postsaccadic drift disconjugacy were worse in strabismic children without binocular vision. The duration of fixation was longer in strabismic children than in age-matched control children. 

Saccade Amplitude during Reading

Our study showed that saccade amplitude observed in strabismic and in control children of comparable age was similar during the reading of a text. Kanonidou et al. showed similar results in strabismic amblyopic adults compared to control adults. ⁸ In other words, strabismus does not seem to influence the amplitude of saccades. Also, our results confirmed and extended our previous work dealing with saccades in strabismic children, ^11,12 in which it was shown that strabismic children were able to localize a target properly. Taken together, all these findings suggested that the cortical and subcortical structures responsible for the computation and execution of the saccades are functional in these subjects. 

Disconjugacy during and after the Saccades

Our study showed that strabismic children showed poorer binocular coordination during and after the saccades when reading than age-matched control children. These results extended the previous study of our group ¹¹ showing large disconjugacy during and after the saccades in strabismic children while saccading to targets. We hypothesized that the fine control of binocular saccade coordination is based on learning mechanisms allowing an efficient relationship between the motor command of the saccades and the vergence sub-systems. These new results suggested that this relationship is deficient in strabismic children. Furthermore, in our study we also showed that binocular vision has an important role in controlling binocular saccades given that strabismic children with binocular vision showed smaller disconjugacy during and after the saccades. This is in line with several previous studies of our group showing the importance of binocular vision and proper vergence disparity capabilities for a good binocular coordination of saccades. Indeed, in children with dyslexia whose vergence abilities are deficient ¹³ and in children with vergence abnormalities (Gaertner et al., personal communication) the disconjugacy during and after the saccades while reading a text has been found to be abnormally large. 

In conclusion, we believe that during a cognitive task, such as reading a text, binocular vision is necessary to bring both eyes onto the word, which in turns allows rapid and efficient reading. To understand better the role of binocular vision in reading, we may need further studies examining strabismic children's reading skills and comprehension. Although some results show that smaller disparities during reading are present and well tolerated in normal populations, ^3,16 the large disconjugacy of saccades we observed in strabismic children could prevent proper identification and understanding of the words. 

Duration of Fixation

Our findings showed significantly longer fixations in strabismic children than in control children. This could be the consequence of a lower quality of vision caused by the large disconjugacy reported during the postsaccadic fixation period, delaying a proper linguistic processing. Kanonidou et al. observed similar results in strabismic amblyopic adults compared to control adults, and suggested that such abnormal reading pattern is a strategy used by strabismic subjects to override their abnormal sensory visual input. ⁸ Our data also are in line with the study of Jainta et al. ¹⁷ showing that during reading of a blurred text, normal adults lengthen the duration of fixation. 

Our data showed that binocular vision does not influence the duration of fixation given that both groups of strabismic children showed similar durations of fixation. We suggest that duration of fixation depends more on the deviation of ocular axis than on an abnormal sensory visual input. Such eye deviation could lead to a longer fixation because of the difficulty in identifying each word. 

It could be questioned whether strabismic children read better with only one eye. Stein et al. showed that dyslexic children with poor vergence control and binocular vision impairment could improve their reading capabilities after monocular occlusion. ¹⁸ However, Barrett et al. recently examined a new technique using the suppression mechanism in adults with strabismus and amblyopia, and pointed out the role of the input from the deviated eye in increasing the binocular cooperation. ¹⁹ In other words, they observed that the visual system is working better with two eyes open than with the best eye alone. Further studies comparing monocular and binocular reading abilities in larger groups of strabismic children (with and without binocular vision) will be needed to improve our understanding of the role of binocular vision capabilities during reading. 

Finally, we suggest that further studies dealing with reading in a large group of strabismics with esotropia and exotropia also are necessary. It is well known that the underlying suppression mechanisms are quite different in those two types of strabismus and their reading performances also could be different. 

Acknowledgments

We thank the children who participated in the study, Florence Daumas for the management of children's appointments, and Savita Bernal for revising the English version of the manuscript. 

Footnotes

 The authors alone are responsible for the content and writing of this paper.

Footnotes

 Disclosure: C. Lions, None; E. Bui-Quoc, None; M. Seassau, e(ye)BRAIN (E); M.P. Bucci, None