Not only amblyopic but also dominant eye in subjects with strabismus show increased saccadic latency

Abstract

Amblyopia is a developmental disorder of vision usually associated with the presence of strabismus and/or anisometropia during early childhood. Subject literature has shown that both the amblyopic and fellow eyes (especially in strabismic subjects) may manifest a variety of perceptual and oculomotor deficits. Previous studies using simple saccadic responses (pro-saccades) showed an increased saccadic latency only for the amblyopic eye viewing conditions. So far, there have appeared no saccadic latency studies in strabismic amblyopia for more complex volitional saccades. In order to maximize the contribution of the central retina in the process of saccade initiation, we decided to use delayed saccadic responses in order to test the hypothesis about saccadic latency increase in both eyes in strabismic amblyopes. The results from our study have shown that saccadic latency is increased both in the dominant and amblyopic eyes. In addition, the amblyopic eye in the strabismic group showed greater increase in saccadic latency compared to an amblyopic eye in the anisometropic group from our previous study. The observed increase in saccadic reaction time for the dominant eye is novel and provides further evidence that the visual pathway associated with the dominant eye might be also impaired in strabismic amblyopia. Since an abnormal binocular input during visual system development may affect gaze stability in both eyes, we speculate that unsteady fixation accompanied with subtle perceptual deficits contribute to an increase in saccadic latency that is observed in the dominant eye. Moreover, it appears that the cortical processes related to saccade decisions are delayed both for amblyopic and fellow eyes in strabismic subjects.

Introduction

Amblyopia is a developmental disorder of vision caused by abnormal visual stimulation (usually strabismus, anisometropia) during infancy or early childhood (Ciuffreda, Levi, & Selenow, 1991; McKee, Levi, & Movshon, 2003). It was shown that an amblyopic eye in strabismic individuals manifests a variety of oculomotor deficits including directional impairment of smooth-pursuit eye movements (Schor, 1975), abnormal saccadic substitution during small-amplitude pursuit tracking (Ciuffreda, Kenyon, & Stark, 1979a), unsteadiness of eye position during fixation (Ciuffreda, Keynon, & Stark, 1979b; González, Wong, Niechwiej-Szwedo, Tarita-Nistor, & Steinbach, 2012; Shaikh, Otero-Millan, Kumar, & Ghasia, 2016; Subramanian, Jost, & Birch, 2013), an increased amplitude of the disconjugacy of saccades (Bucci, Kapoula, Yang, Roussat, & Brémond-Gignac, 2002), or the saccadic reaction time (latency) increase (Ciuffreda et al., 1978; McKee, Levi, Schor, & Movshon, 2016; Niechwiej-Szwedo, Chandrakumar, Goltz, & Wong, 2012). Furthermore, the functional imbalance between the two eyes (e.g., unilateral strabismus) during visual system development may affect not only visual processing associated with the amblyopic eye but also a visual pathway associated with the fellow nonamblyopic (dominant) eye and interactions between the two eyes (Harrad & Hess, 1992; Harwerth & Levi, 1983; Huang, Zhou, Lu, & Zhou, 2011). The subject literature provides evidence that the dominant eyes of unilateral amblyopes are not entirely normal and may show a variety of deficits including a decreased contrast sensitivity function (Chatzistefanou et al., 2005), reduced visual acuity (Kandel, Grattan, & Bedell, 1980; Varadharajan & Hussaindeen, 2012) and Vernier acuity (Levi & Klein, 1985), elevated blur discrimination (Simmers, Bex, & Hess, 2003), deficiency in motion perception (Ho et al., 2005; Simmers, Ledgeway, Hess, & McGraw, 2003), greater internal noise (Levi & Klein, 2003), attentive tracking deficits (Ho et al., 2006), degraded attentional modulation (Hou, Kim, Lai, & Verghese, 2016), or delayed visual decision response times (Farzin & Norcia, 2011). Moreover, several previous studies on eye movements in strabismic amblyopes reported abnormal smooth pursuit eye movements (Fukai, Tsutsui, & Nakamura, 1976), unsteady fixation (Bedell & Flom, 1985; Shaikh et al., 2016), or small fixational eccentricity (Bedell & Flom, 1985) for dominant eye viewing. 

The impact of strabismic amblyopia on saccadic reaction time has been investigated in only four studies so far. Schor (1975), with the use of a predictable square-wave stimulus, examined five strabismic amblyopes and did not report the saccadic latency increase either for the amblyopic or dominant eye. On the other hand, Ciuffreda et al. (1978), McKee et al. (2016), and Niechwiej-Szwedo et al. (2012), using the conventional and the simplest of all saccadic tasks—the step task, in which pro-saccades are generated (Carpenter, 2012), observed a significant increase in saccadic latency for amblyopic eye but not for dominant eye viewing. 

Until now, there have been no studies on more complex saccadic responses in strabismic amblyopia. Taking into consideration that oculomotor deficits related to amblyopic and dominant eyes of strabismic amblyopes were mainly associated with abnormal processing of the target from the central retina (Bedell & Flom, 1985; Ciuffreda et al., 1979a, 1979b; Fukai et al., 1976; Schor, 1975; Shaikh et al., 2016; Subramanian et al., 2013), we decided to study delayed saccadic responses in order to maximize the contribution of the central retina in the process of saccades initiation. A delayed saccade task required both good inhibitory control and visual fixation maintenance on the target in order to detect the GO signal given by the disappearance of the centrally fixated target. In this context, we predicted that not only the amblyopic but also the dominant eye may show a (subtle) saccadic latency increase during this specific oculomotor task. The main aim of the current study is to investigate how the saccadic reaction time for more complex, delayed saccadic responses is affected by strabismic amblyopia. 

Since strabismic and anisometropic amblyopes may show different patterns of visual loss (McKee et al., 2003), different temporospatial characteristics of calcarine activation (Choi et al., 2001), as well as different characteristics of fixational eye movements (Chung, Kumar, Li, & Levi, 2015), we decided to compare the results of the current study with our previous study (Perdziak, Witkowska, Gryncewicz, Przekoracka-Krawczyk, & Ober, 2014) performed on anisometropic amblyopes (where we used the same experimental paradigm and procedure) in order to explore the impact of these two subtypes of amblyopia on the initiation process of delayed saccadic responses. 

Methods

Participants

Ten subjects with strabismic amblyopia (seven females; mean age 39.4 ± 16.3 years) and 12 normal subjects (controls; six females; mean age 36.4 ± 13 years) took part in the study. Normal subjects were age matched and selected from the control group that participated in our previous study (Perdziak et al., 2014). Based on a medical interview, all subjects were healthy without any ophthalmological or neurological diseases and none were receiving medications known to affect attention or reaction time. All subjects underwent standard optometric examination (with special emphasis on binocular vision examination), which included the measurement of visual acuity (Snellen chart: decimal notation) with and without optical correction, ocular dominance (fixating via hole task), refractive error examination (static or Mohindra retinoscopy and subjective refraction), Worth four-dot test at far and near, stereopsis (Stereo Fly Test), and slit lamp (anterior segment) examination. Eye alignment was evaluated with the use of a prism cover test at far and near. The visual parameters of subjects with strabismic amblyopia are presented in Table 1. The strabismic group (SG) consisted of subjects with strabismic amblyopia according to the following criteria: 

•  
  - Constant strabismus during cover test procedure
•  
  - No history of deprivation
•  
  - Reduced best corrected visual acuity (BCVA) at least two lines for amblyopic eye (AE) compared to the dominant eye (DE) viewing and no clinically significant anisometropia (defined as interocular difference in refractive error above 1 diopter in both meridians).

This study was approved by the bioethic committee of the Medical University in Poznan, and all experiments were conducted in accordance with the ethical standards laid down in the Declaration of Helsinki. 

Table 1

 

Clinical characteristics of the participants.

 

Notes: RE = right eye; LE = left eye; BCVA = best corrected visual acuity; CXT = constant exotropia; CST = constant esotropia. Interocular suppression was evaluated with the Worth four-dot test.

 

Apparatus

Details of the apparatus and experimental procedure have been described in our previous article (Perdziak et al., 2014). In brief, subjects' eye movements were recorded using a Saccadometer device (Ober Consulting, Poland; Ober, Dylak, Gryncewicz, & Przedpelska-Ober, 2009). The eye movement sensor utilizes the photoelectric method (based on the amplitude modulation of infrared reflection from the cornea; Ober, 1994), providing high temporal (1 ms) and spatial (5 arcmin) resolution for the detection of saccadic onsets. The onsets of saccadic movements were detected automatically on-line, based on local minimum eye velocity, followed by a movement exceeding 5°/s. Saccadic latencies, amplitudes, and spatial profiles were recorded and reviewed offline by an experienced operator. The stimulus was displayed using miniature laser projectors mounted on the sensor forehead plate (bound with the subject's head). 

Procedure

Each trial begun with a central fixation target, and after 200 ms peripheral stimuli for saccadic refixation were displayed, either 10° to the left, or 10° to the right, randomly. The visual stimuli for a saccade (red laser spot, ∼4.5 arcmin) were generated using miniature laser projectors mounted on the sensor saccadometer forehead plate. After randomized time (1000–2500 ms), the central target disappeared, giving a GO signal for saccade. The participants were instructed to look at the central target and make a saccade to the peripheral target as quickly as possible after the central target disappeared. Latency was considered as a time interval between the GO signal and the onset of saccadic eye movement. The participants were seated at the distance of 3 m in front of the white wall on which the targets were projected. The task was performed in a quiet and uniformly illuminated (∼250 lux) room. Before the experiment, all subjects underwent the practice trials in order to make them familiar with the experimental procedure. All subjects used their optimal refractive correction and performed the experiment in a silence in three viewing conditions: (a) binocular viewing; (b) monocular viewing with the dominant eye; and (c) monocular viewing with the nondominant/strabismic eye. All participants completed 100 trials for each viewing condition. The order of viewing conditions was randomized between subjects and there was a 1-min break between the trials. All data were collected in one session (1–1.5 hrs), which included optometric examination, practice trials, calibration, and experimental trials. 

Data analysis and statistics

Eye movements in the wrong direction affected by blinks as well as the trials without any responses were removed from the analysis. Moreover, we excluded saccades with latencies shorter than 80 ms and longer than 700 ms. All those criteria resulted in the elimination of average 12.5% of trials for each viewing condition in the strabismic group and 6.3% in the normal group. In the case of the strabismic group, the amblyopic eye viewing was associated with an increased number of excluded responses (19.2%) compared to the dominant eye viewing (9.7%) and binocular viewing condition (8.7%). In visually normal subjects, the percent of eliminated trials was quite similar between the viewing conditions (5.5%, 6.5%, and 6.9%). Additionally, the saccades with latencies above and under 2.5 SDs from the mean of individual latency distribution (for every viewing condition separately) were considered to take place not in response to the stimulus (despite meeting the previous criteria) and therefore were rejected (Irving, Steinbach, Lillakas, Babu, & Hutchings, 2006; Van der Stigchel, Mills, & Dodd, 2010). This operation resulted in further elimination of 2.8% of trials for each viewing condition in the strabismic group and 2.7% in the normal group. For every subject we calculated the mean and coefficient of variation (CV) of saccadic latency for three viewing conditions: monocular viewing with the dominant eye (DE); monocular viewing with the strabismic eye (SE)/nondominant eye (NDE), and binocular viewing (BV). We used a Pearson's chi-square test to analyze the differences in the size between the groups. The Kolmogorov–Smirnov test was used to evaluate the normality of data. The homogeneity of variance was tested by Levene's test, and sphericity was evaluated using Mauchly's test. A repeated-measures analysis of variance (ANOVA) with a between-subject factor (group [control subjects and subjects with strabismus]) and one within-subjects factors (viewing condition [DE, SE/NDE, and BV]) was used to analyze the mean values and CV of saccadic latency. Greenhouse–Geisser correction was used in the case of significant result of Mauchly's test. A post hoc Bonferroni test was used to analyze the significant effects of interaction. Moreover, we decided to compare the results obtained in patients with strabismic amblyopia with the results of 16 patients with anisometropic amblyopia obtained during the previous research (Perdziak et al., 2014). The procedure used in the previous research was the same, but unfortunately did not include the binocular viewing condition. We analyzed the difference between the strabismic amblyopia and anisometropic amblyopia for mean latency and CV latency in two conditions: amblyopic eye viewing and dominant eye viewing. Differences between the groups were analyzed using the Bonferroni method. 

Results

Analysis between the strabismic and control group

Mean saccadic latency

The main effect of viewing condition, F(1.379, 27.582) = 13.230, P < 0.001, η_p² = 0.398, as well as the effect of the group was statistically significant, F(1, 20) = 14.375, P = 0.001, η_p² = 0.418. The strabismic group (SG) also showed the significant increase in the mean saccadic latency as compared to the control group (CG). Moreover, we found the significant interaction between the group and viewing conditions, F(1.379, 27.582) = 9.499, P = 0.002, η_p² = 0.322. A post hoc Bonferroni test revealed statistically significant difference in saccadic latency between the SE and DE (difference in the means = 57.6 ms, P < 0.001, SE = 320 ± 52 ms, DE = 263 ± 45 ms) as well as between the SE and binocular viewing condition (difference of the means = 52.1 ms, P = 0.001, SE = 320 ± 52 ms, BV = 268 ± 66 ms). In the control group, mean saccadic latency for binocular viewing condition was insignificantly decreased when compared to the DE and NDE viewing conditions (BV = 213 ± 28 ms, DE = 225 ± 34 ms, NDE = 225 ± 26 ms). Moreover, in the control group, there were no statistically significant differences between all viewing conditions. For each viewing condition, strabismic amblyopes showed significant increase in the mean saccadic latency as compared to normal subjects (for DE: difference in the means = 37.4 ms, P = 0.038, SG = 263 ± 45 ms, CG = 225 ± 34 ms; for SE/NDE: difference in the means = 95.6 ms, P < 0.001, SG = 320 ± 52 ms, CG = 225 ± 26 ms; for BV: difference in the means = 55.5 ms, P = 0.015, SG = 268 ± 66 ms, CG = 213 ± 28 ms; Figure 1). 

Figure 1

 

Mean latency and latency's coefficient of variation (CV) in the strabismic and control groups as a function of viewing condition. Notes: The vertical bars indicate the standard error of the mean (SEM); *P < 0.05; **P < 0.01; ***P < 0.001.

Variability of saccadic latency

The CV of the saccadic reaction time for strabismic and control groups in the three viewing conditions is presented in Figure 1. Among strabismic amblyopes, the variability of saccadic latency was increased as compared to the normal group for all viewing conditions (DE: SG = 24% ± 6%, CG = 18% ± 7%; SE/NDE: SG = 28% ± 7%, CG = 18% ± 7%; BV: SG = 25% ± 6%, CG = 17% ± 7%), which was confirmed by a significant effect of the group, F(1, 20) = 13.714, P = 0.001, η_p² = 0.407. The main effect of viewing condition, F(2, 40) = 0.964, P = 0.390, η_p² = 0.046, as well as the group × viewing condition interaction, F(2, 40) = 0.461, P = 0.634, η_p² = 0.023, were nonsignificant. 

Strabismus versus anisometropia

Figure 2 presents all measured parameters (mean latency and CV) for the strabismic (SG) and anisometropic group (AG) in two viewing conditions (dominant eye [DE] and amblyopic eye [AE]). For the dominant eye viewing condition statistical analysis did not reveal any significant differences, although it can be observed that the values of parameters were different: mean saccadic latency (SG = 263 ± 45 ms, AG: 237 ± 45 ms) and its CV (SG: 24% ± 6, AG: 20% ± 7). However, an amblyopic eye in the strabismic group revealed the increased saccadic latency as compared to anisometropic group (mean latency: 320 ± 52 vs. 262 ± 48 ms), which was confirmed by a significant result from a pairwise comparison (difference of the means = 57.7; P = 0.008). The CV did not differ between the groups (AE: SG = 28% ± 7%; AG = 22% ± 8%). 

Figure 2

 

Mean latency and latency's coefficient of variation (CV) in strabismic and anisometropic amblyopia groups. Notes: The vertical bars indicate the standard error of the mean (SEM); *P < 0.05; **P < 0.01; ***P < 0.001.

Discussion

In the present study the delayed saccade latencies in subjects with strabismic amblyopia were examined and compared to normal subjects as well as to anisometropic amblyopes from our previous study (Perdziak et al., 2014). It is important to stress, based on our findings from the study, that the dominant eyes in subjects with strabismic amblyopia have shown an increased saccadic latency as compared to the dominant eyes in subjects with normal vision. In addition, we observed an increase in saccadic latency for the amblyopic eye, no binocular advantage for saccadic latency (both for strabismic amblyopes and normal subjects), and an increased variability in saccadic reaction time for all viewing conditions in strabismic amblyopes. With respect to the anisometropic group, an amblyopic eye in the strabismic group revealed the increased saccadic latency. 

Increased saccadic latency

Although the neural mechanisms responsible for the increase in saccadic latency in amblyopia are still not fully understood, it is widely accepted that this effect is rather sensory than motor in origin (Ciuffreda et al., 1978; Niechwiej-Szwedo et al., 2012; Niechwiej-Szwedo, Goltz, Chandrakumar, Hirji, & Wong, 2010; Perdziak et al., 2014; Schor, 1975). The potential mechanisms related to the saccadic latency increase for the amblyopic eye viewing conditions were broadly discussed in the previous research (McKee et al., 2016; Niechwiej-Szwedo et al., 2012; Niechwiej-Szwedo et al., 2010; Perdziak et al., 2014). In general, slower processing of visual information (sensory acquisition of the target; Ciuffreda et al., 1978; Niechwiej-Szwedo et al., 2012; Niechwiej-Szwedo et al., 2010; Perdziak et al., 2014) or slower supply of afferent visual information to the brain (Perdziak et al., 2014) are considered to delay a saccadic response (or decision about saccade initiation) in amblyopic eyes. In this context, our results related to the saccadic latency increase for amblyopic eye stimulation are not surprising and are in agreement with previous studies performed on strabismic amblyopes (Ciuffreda et al., 1978; McKee et al., 2016; Niechwiej-Szwedo et al., 2012). Niechwiej-Szwedo et al. (2012) studied visuomotor behavior in 14 subjects with strabismic amblyopia and found a saccadic latency increase in the amblyopic eyes (this effect was more pronounced for more central targets), but only in subjects with severe acuity deficits (i.e., 20/200 in the amblyopic eye) and no stereopsis. In addition, reduced amplitude precision for the amblyopic eye was observed in subjects with severe or mild acuity loss. Recently, McKee et al. (2016) published data gathered on a large group of amblyopic subjects and also reported the increased saccadic latency for the amblyopic eye. However, none of the previous studies reported the increase in saccadic latency for a fellow nonamblyopic eye. 

What then accounts for the increase in saccadic latency for dominant eyes in the examined strabismic amblyopes? Although our findings will not provide a conclusive answer to this question, we would like to propose a possible explanations and routes for further research. In all previous studies on strabismic amblyopia (Ciuffreda et al., 1978; McKee et al., 2016; Niechwiej-Szwedo et al., 2012), the saccadic reaction time was tested with the use of pro-saccades generated during the step task—without temporal delay between the offset of the fixation target and the presentation of the peripheral target. Pro-saccades, the so-called reflexive (or exogenous) saccades, are made in response to a novel peripheral stimulus (Walker, Walker, Husain, & Kennard, 2000) and in contrast to volitional delayed saccades, there is no cognitive judgment involved, such as the response to a GO signal coming from the central retina. Thus, the important difference between the previous studies and the present one is that, in our paradigm, more complex volitional saccadic responses were required and more visual processing received from the central retina was engaged in the saccade initiation process. At the beginning, the saccade towards the peripheral target must be suppressed, which requires good inhibitory control from frontal and prefrontal cortex to withhold the execution of planned eye motor program (Munoz & Everling, 2004). Next, during the preparatory phase, visual fixation must be maintained and the spatial characteristic of a saccade is planned (computed), but the saccade initiation process stays inactivated until the permission “GO” is provided via central retina. In the literature, it has been proved that not only spatial vision deficits (Chatzistefanou et al., 2005; Kandel et al., 1980; Levi & Klein, 1985; Simmers, Bex, & Hess, 2003; Varadharajan & Hussaindeen, 2012) but also abnormalities associated with motion perception (Ho et al., 2005; Simmers, Ledgeway, Hess, & McGraw, 2003), attention (Ho et al., 2006; Hou et al., 2016), visual decision-making (Farzin & Norcia, 2011), or visual evoked potentials (De Mendonça et al., 2013; Shawkat, Kriss, Timms, & Taylor, 1998; Watts, Neveu, Holder, & Sloper, 2002) may be present in both the fellow and the amblyopic eye. Moreover, it was found that both fellow and amblyopic eyes may show (especially in strabismic amblyopia) a variety of oculomotor deficits. Fukai et al. (1976) reported an abnormal smooth pursuit movement of the fellow eye in strabismic amblyopia. Bedell and Flom (1985) investigated nine strabismic amblyopes and found that both amblyopic and fellow eyes manifest a high velocity of nasal drifts under monocular viewing conditions. They concluded that a centrally generated nasal drift bias is responsible for anomalous oculomotor behavior of both eyes in strabismic amblyopes (Bedell & Flom, 1985). Moreover, González et al. (2012) reported a significant decrease in fixation stability in the amblyopic eyes during both binocular and monocular viewing. With respect to the fellow eye, fixation stability depended on viewing condition such that was normal for binocular and monocular fellow eye viewing, but was abnormal behind the occluder (when the viewing eye was amblyopic). Moreover, they found that amblyopic subjects and the controls did not differ in terms of the rate or magnitude of intrusive microsaccades. Recently, Shaikh et al. (2016) studied fixational eye movements in amblyopic children and found an increase in amplitude with decreased frequency of microsaccades for amblyopic eye viewing. Furthermore, they also reported an increase in variance and velocity of ocular drifts during both dominant and amblyopic eye viewing conditions. On the other hand, Chung et al. (2015), using the retinal imaging technique (without head stabilization), reported comparable drift velocities in the amblyopic and fellow eyes as well as in the control subjects. In general, the above findings indicate that the loss of binocular visual cues due to monocular impairment of vision, may affect gaze stability in both eyes (Bedell & Flom, 1985; González et al., 2012; Schneider et al., 2013), which suggests that fixation stability of each eye is governed by a common neural network (Schneider et al., 2013). It is worth to mention that previous studies (McKee et al., 2016; Perdziak et al., 2014) considered the potential influence of refractory period between microsaccades on the latency of subsequent saccade, but in relation to above findings, it seems that it should not contribute to the saccadic latency increase rather, since the frequency of microsaccades was found to be comparable between amblyopic and control group (González et al., 2012) or even decreased for amblyopic eye viewing (Shaikh et al., 2016). Although our eye movements system (Saccadometer) does not allow for effective fixational eye movements record, it is generally accepted that the fixation instability in amblyopic subjects is mainly due to abnormal ocular drift (Bedell & Flom, 1985; Ciuffreda et al., 1979b; Ciuffreda et al., 1980; González et al., 2012) and in the context of the present study, it deserves special attention. Although drifts in nonamblyopic eyes are generally not rapid enough to degrade visual acuity, we suppose that the saccadic latency increase in dominant eyes is related to the increased retinal movement of the fixation target during the preparatory phase, which causes more uncertainty and may delay the cortical decision about saccade initiation. This interpretation is supported by the fact that the interocular difference in visual acuity (acuity deficit) was found to correlate with fixation instability of the fellow eye during binocular and monocular viewing (González et al., 2012). Note, that in our amblyopic group, up to five subjects manifested severe amblyopia (visual acuity equal or worse than 0.2). In the light of these findings, it is likely that the centrally presented cue for saccade initiation (GO signal) connected with subtle visual and fixational eye movements deficits, contributed to the observed saccadic latency increase in fellow nonamblyopic eyes. Moreover, Di Russo et al. (2003) showed that visually normal subjects with better fixation stability (professional shooters) tend to have shorter saccadic latency as compared to the controls. In this context, a saccadic latency increase in the dominant eye is not entirely surprising and may represent another subtle deficit in visual system whose proper development was limited by strabismic amblyopia. To date, the relationship between the saccadic latency and fixation stability is not entirely clear and future research should attempt to verify the effects of abnormal fixation pattern on saccades initiation in amblyopic subjects. 

Moreover, abnormal binocular inputs during visual system development may also affect visual pathway associated with the nonamblyopic eye as well as cortical connections to downstream areas of the brain, including the parietal and frontal cortexes that are believed to be responsible for decision-making and response selection processes (Farzin & Norcia, 2011; Wang, Crewther, & Yin, 2015). Farzin and Norcia (2011) found that the visual decision response times were delayed in both the amblyopic and nonamblyopic eyes. Thus, cortical networks underlying decision processing possibly depend on binocular visual experience during early childhood (Farzin & Norcia, 2011). In our previous work related to anisometropic amblyopia (Perdziak et al., 2014), we performed a saccadic latency analysis using the Linear Approach to Threshold with Ergodic Rate (LATER) decision model (see Carpenter, 1999; Carpenter & Williams, 1995; Noorani & Carpenter, 2011, 2016). Our results showed that the rate of rise of the decision signal is decreased in the amblyopic eye. Although in the current study we did not include the LATER analysis, it is also possible that the observed increase in saccadic latency in both eyes in our strabismic amblyopes result from deficient cortical processing associated with decision-making about saccade initiation. 

An amblyopic eye in the strabismic group revealed increased saccadic latency as compared to anisometropic group (320 ± 52 ms vs. 262 ± 48 ms) from our previous study (Perdziak et al., 2014). Although two previous studies did not support our observation (Niechwiej-Szwedo et al., 2012; Niechwiej-Szwedo et al., 2010), more recent work (McKee et al., 2016) performed on a large group of amblyopic subjects also showed that saccadic latency is increased in amblyopic eyes in strabismic subjects as compared to amblyopic eyes in anisometropic individuals. With respect to the dominant eye, statistical analysis did not show any significant differences in saccadic latency between the strabismic and anisometropic groups. However, it can be observed that in the strabismic group, the mean saccadic latency for dominant eye (263 ± 45 ms) was increased as compared to anisometropic group (237 ± 45 ms). Due to the small sample available, we did not distinguish between binocular and nonbinocular subjects, but it is important to note that most of our anisometropic amblyopes had stereopsis, and most of our strabismic amblyopes were nonbinocular observers and manifested deeper amblyopia as compared to the anisometropic group from our previous research (Perdziak et al., 2014). It is commonly known that the lack of binocular vision affects not only visual but also a variety of other functions related to visuomotor coordination (Grant & Moseley, 2011; Suttle, Melmoth, Finlay, Sloper, & Grant, 2011) and motor control abilities (Przekoracka-Krawczyk, Nawrot, Czaińska, & Michalak, 2014; Przekoracka-Krawczyk, Nawrot, Kopyciuk, & Naskręcki, 2015). It was found that nonbinocular subjects (usually strabismic amblyopes) tend to have poorer optotype and Vernier acuity than those with residual binocular function (usually anisometropic amblyopes; McKee et al., 2003). Moreover, with the loss of binocular vision, saccadic latency tends to increase (McKee et al., 2016). Several recent studies have attempted to verify the existence of correlation between visual acuity and fixation stability in amblyopic subjects. Subramanian et al. (2013) studied fixation stability in 89 amblyopic children and reported positive correlation between visual acuity and fixation stability for amblyopic eyes in strabismic and mixed (aniso-strabismic) groups but not for amblyopic eyes in the anisometropic group. Thus, not only visual but also motor and oculomotor deficits are in general manifested much more in strabismic amblyopes (which are usually a nonbinocular observers), and therefore, it is not surprising that saccade initiation is more delayed in the strabismic group compared to the anisomtropic group. 

For us, the studies that have tried to differentiate anisometropic amblyopia from strabismic amblyopia on the basis of evoked potential latency or reaction time, appear especially interesting. McKerral, Polomeno, Leporé, and Lachapelle (1999) showed that while the peak time of the pattern visual evoked potential in anisometropic and strabismic amblyopia was significantly delayed, two types of amblyopia could not be distinguished. However, McKerral et al. (1999) showed that it is possible to distinguish strabismic from the anisometropic amblyopia using interocular differences in reaction time measurements—in strabismic amblyopes the interocular reaction time difference is significantly higher than in anisometropic amblyopes. Although McKerral et al. (1999) used reaction time measurement during a different task (subjects were instructed to signal the detection of the stimuli by pressing a manual switch), our results seem to be in agreement with the observation by McKerral et al. (1999), supporting a functional distinction between anisometropic and strabismic amblyopes. 

An increase in variability of saccadic latency was also reported by Niechwiej-Szwedo et al. (2010) and Schor (1975). However, in our previous study (Perdziak et al., 2014) performed on anisometropic amblyopes we did not find increased variability for amblyopic eye viewing. Thus, more research on a larger group of amblyopic subjects is needed to understand this effect. 

Lack of binocular advantage

Binocular summation is an example of excitatory interaction in the visual system and requires good and similar inputs from both eyes. To the best of our knowledge, the effect of binocular advantage on the saccadic reaction time in strabismic amblyopia (expressed as latency decrease for binocular viewing) has been investigated only in one previous study. Niechwiej-Szwedo et al. (2012), with the use of pro-saccades, reported decreased saccadic latency during the binocular viewing condition for normal observers and the lack of this effect during binocular viewing for strabismic amblyopes. Our findings confirmed the lack of binocular advantage in the strabismic amblyopes, but failed to be in agreement with findings by Niechwiej-Szwedo et al. (2012) with respect to binocular advantage for visually normal subjects: In the current study, saccadic latency was only insignificantly decreased in the BV condition. It is known that in normal subjects the simultaneous use of both eyes is associated with reduced reaction time (Woodman, Young, Kelly, Simoens, & Yolton, 1990). Therefore, our finding is rather surprising and may result from different saccadic task and relatively small sample compared to Niechwiej-Szwedo et al. (2012). Hence, future works on a larger group of subjects are needed to explore the binocular advantage in saccadic latency both for pro-saccades and voluntary saccades. We are going to study this topic in our laboratory in the near future. 

Conclusions

Based on the results obtained from the current study, we found that with respect to delayed saccades, strabismic amblyopes reveal saccadic latency increase both for amblyopic eye and also to a lesser degree, for dominant eye viewing. In addition, the amblyopic eye in the strabismic group showed an increased saccadic latency as compared to the anisometropic group from our previous study. Our finding suggests that strabismic amblyopia affects saccade initiation process not only for the amblyopic eye but also for the dominant eye, probably due to oculomotor deficits related to visual fixation and subtle perceptual deficits that are commonly observed in both eyes among strabismic amblyopes. 

Acknowledgments

The authors thank Agata Gryc (MA in English philology) for proofreading of the revised version of the manuscript. 

Commercial relationships: Jan Ober is CEO of Ober Consulting Ltd. Company, which provides eye movement measurement equipment (Saccadometer), used in this study (I). Dagmara Witkowska and Wojciech Gryncewicz are employees of Ober Consulting Ltd., which provides eye movement measurement equipment (Saccadometer), used in this study (E). Jan Ober is a patent holder of Saccadometer measurement device (P). 

Corresponding author: Maciej Perdziak. 

Email: [e-mail]. 

Address: Laboratory for Oculomotor Research, Polish Academy of Sciences, Poznan, Poland.