Amblyopia is a neurodevelopmental disorder characterized by reduced best-corrected visual acuity (BCVA) in one or both eyes without any detectable structural abnormality, affecting approximately 4.3% of children worldwide.1,2 It remains a leading cause of preventable unilateral visual impairment that can persist into adulthood. Based on etiology, amblyopia is classified as strabismic, refractive, or deprivation.3 Refractive amblyopia arises from uncorrected refractive error during the critical period of visual development, leading to chronic suppression and altered cortical processing.

Beyond reduced acuity, amblyopia is associated with widespread functional deficits. Children with amblyopia often exhibit impairments in contrast sensitivity, stereopsis, accommodative response, visuomotor integration, and oculomotor control.4–6 Abnormalities in saccadic and pursuit eye movements, including increased latency, reduced accuracy, fixation instability, and altered pursuit gain, are thought to be associated with alterations in cortical and subcortical networks that support visuomotor behaviour.4,7 These deficits negatively impact reading efficiency, visual tracking, spatial attention, and fine motor tasks, contributing to functional disability.8

Conventional amblyopia treatment focuses on refractive correction followed by occlusion therapy, both of which primarily target monocular acuity improvement.9 Although clinically effective, occlusion alone seldom normalizes binocularity or oculomotor function, and recurrence rates remain substantial.10,11 Active Vision Therapy (AVT) has gained attention as an adjunctive, neuroplasticity-based intervention that employs structured visual activities aimed at improving functional visual and visuomotor performance, including fixation stability, saccadic and pursuit accuracy, binocular coordination, and higher-order visual processing.12,13 Digital platforms such as the Vision Training System 4 (VTS4) incorporate interactive, multisensory tasks designed to stimulate cortical adaptation and improve functional performance.14 VTS4-based measurements represent task-based visuomotor performance metrics rather than direct eye-tracking–based physiological recordings.

Various therapeutic approaches have been explored to address amblyopia beyond conventional occlusion therapy, each targeting different aspects of visual dysfunction. Traditional occlusion primarily improves monocular visual acuity but has limited effects on binocular integration and oculomotor control. In contrast, dichoptic therapies and video game–based interventions aim to reduce interocular suppression and enhance binocular function through balanced stimulation of both eyes, with several studies demonstrating improvements in stereopsis and contrast sensitivity.15–17 Perceptual learning approaches, which involve repetitive practice of visual tasks, have also shown potential in refining visual processing and, to some extent, visuomotor performance.18

However, these approaches differ substantially in their mechanisms and therapeutic targets, and their effects on oculomotor function remain inconsistently characterized. While abnormalities in saccadic latency, accuracy, and smooth pursuit gain are well documented in anisometropic amblyopia,4,7 most intervention studies have focused primarily on sensory outcomes such as visual acuity and stereopsis, with relatively limited attention to objective changes in eye movement behaviour. Moreover, existing studies on binocular or dichoptic therapies have rarely incorporated structured oculomotor training components or standardized task-based assessments of saccadic and pursuit performance.

Active Vision Therapy (AVT) represents a multimodal approach that integrates monocular, binocular, vergence, and oculomotor training within a structured framework. Despite its increasing clinical use, there is a paucity of evidence specifically examining whether AVT can modify visuomotor task performance, particularly task-based saccadic response time, accuracy, and smooth pursuit performance, in children with anisometropic amblyopia, and whether such changes are associated with improvements in visual outcomes.

Therefore, the present study aims to address this gap by evaluating the effect of Active Vision Therapy combined with occlusion, compared with occlusion alone, on task-based saccadic and smooth pursuit performance in children with anisometropic amblyopia. In addition, the study examines the association between changes in oculomotor parameters and improvements in visual outcomes, to better understand the functional role of oculomotor rehabilitation in amblyopia management.

Data normality was assessed using the Shapiro–Wilk test. Continuous variables are presented as mean ± SD (standard deviation) or median (IQR, Interquartile range), and categorical variables as frequencies and percentages. Baseline characteristics were compared using independent t-tests or Mann–Whitney U tests for continuous variables and Chi-square or Fisher’s exact tests for categorical variables.

Within-group longitudinal changes across baseline, three weeks, six weeks, and three months were analyzed using repeated-measures ANOVA for normally distributed variables, with Greenhouse–Geisser correction applied when required. Non-normally distributed outcomes (stereopsis) were analyzed using the Friedman test with Bonferroni-adjusted post-hoc comparisons. Effect sizes were reported as partial eta squared (η²p). Given the non-randomized allocation, analyses were interpreted as assessing associations between intervention and outcomes rather than definitive causal effects.

Between-group longitudinal effects were examined using two-way repeated-measures ANOVA, with time as the within-subject factor and therapy group as the between-subject factor. Given the quasi-experimental design and limited sample size, multivariable adjustment for covariates (e.g., age, baseline amblyopia severity, and adherence) was not performed to avoid model overfitting and unstable parameter estimates. Baseline comparability was assessed to minimize confounding risk. However, residual confounding cannot be excluded, and between-group effects should be interpreted cautiously as associative rather than fully adjusted causal estimates. Between-group differences in three-month change scores were assessed using independent t-tests or Mann–Whitney U tests, as appropriate, with mean differences and 95% confidence intervals reported.

Partial correlations between changes in oculomotor parameters and visual outcomes adjusted for therapy group were assessed using Spearman coefficients, based on data distribution, using baseline-to–three-month change scores.

Statistical significance was set at α = 0.05. All analyses were performed using SPSS version 22.0 (IBM Corp., Armonk, NY, USA).

This hospital-based longitudinal study demonstrated that AVT combined with occlusion was associated with greater improvements in visual function and performance on oculomotor tasks, indicating enhanced functional visuomotor control, compared with occlusion therapy alone in children with anisometropic amblyopia. It is important to note that our measures reflect behavioural task performance rather than direct recordings of eye movement latency or gain. Importantly, these findings address a key gap in the literature by providing objective evidence that a structured, multimodal vision therapy program can improve performance on visuomotor tasks, including task-based saccadic response time, accuracy, and smooth pursuit performance, in this population.

Previous research has shown that amblyopic individuals exhibit delayed and imprecise saccades and other visuomotor dysfunctions, reflecting disrupted cortical processing and impaired binocular integration.4,20 While conventional occlusion therapy remains effective in improving best-corrected visual acuity, it primarily targets monocular sensory input and may not adequately address oculomotor deficits such as abnormal saccadic and pursuit eye movement, which are increasingly recognised as integral features of amblyopia-related visual dysfunction.1,21

In contrast to conventional approaches, emerging therapies such as dichoptic training, video game–based interventions, and perceptual learning have been developed to target binocular function and cortical plasticity. These approaches have demonstrated improvements in stereopsis and contrast sensitivity15–17, but their effects on oculomotor control remain less consistently reported. Notably, most of these interventions do not explicitly incorporate structured saccadic or pursuit training components, nor do they routinely quantify changes in eye movement behaviour using task-based metrics. The present findings therefore extend the existing literature by demonstrating that a therapy paradigm explicitly integrating task-based oculomotor training can yield measurable improvements in both task-based saccadic and smooth pursuit performance.

In the present study, both AVT–occlusion and occlusion-only groups experienced significant within-group gains in BCVA, stereopsis, contrast sensitivity, and parameters of oculomotor task performance over three months (Table 2). The AVT–occlusion group demonstrated faster and larger improvements in task-based saccadic accuracy, saccadic response time, and smooth pursuit performance, with significant time × therapy interactions (Table 3). These findings suggest that while spontaneous or occlusion-driven improvements in oculomotor task performance may occur, targeted visuomotor training accelerates and enhances these adaptations. While the AVT–occlusion group demonstrated faster and larger improvements across several outcomes, the use of an alternating (non-random) allocation procedure limits the ability to draw causal inferences from these findings. As allocation was predictable and not concealed, the possibility of selection bias cannot be excluded, even though measured baseline characteristics were comparable between groups. Accordingly, the observed between-group differences should be interpreted as associations rather than definitive evidence of treatment efficacy. These findings are hypothesis-generating and support the need for well-designed randomized controlled trials with concealed allocation to confirm the causal effects of AVT on oculomotor and visual performances.

The BCVA gains observed in the study (0.30 logMAR for AVT–occlusion vs 0.18 logMAR for occlusion-only, mean difference −0.13, 95% CI −0.19 to −0.07, p = 0.01) aligned with previous reports. Suwal et al. (2024) similarly reported gains of 0.32 ± 0.11 logMAR with AVT and 0.27 ± 0.19 logMAR with patching, with AVT showing particular benefit in severe amblyopia (p = 0.031) .5 Differences between studies likely reflect variations in baseline severity, therapy intensity, adherence, sample size, and participant characteristics.

The relatively short six-week optical adaptation period, shorter than that used in major amblyopia clinical trials, may have influenced baseline stabilization and partially confounded interpretation of between-group differences. The definition of persistent amblyopia after six weeks of optical adaptation warrants careful consideration. Previous studies, including those from the Pediatric Eye Disease Investigator Group (PEDIG), have demonstrated that visual acuity can continue to improve with refractive correction alone for up to 12–16 weeks or longer. Accordingly, it is possible that some participants in the present study may have experienced additional gains with continued optical adaptation alone.22,23 However, the decision to initiate adjunctive therapy after six weeks was guided by the absence of clinically meaningful improvement during this period and reflects a pragmatic, clinic-based approach aimed at avoiding delays in treatment escalation. Importantly, both study groups were subjected to the same adaptation duration, which helps preserve internal comparability, although it does not eliminate the possibility that longer adaptation could have attenuated between-group differences. Future studies incorporating longer optical adaptation periods are warranted to better isolate treatment effects.24

In the study, stereopsis improved in both groups (40 ′ for AVT–occlusion vs 20 ′ for occlusion-only), although the between-group difference was not significant (p = 0.80). Early studies reported minimal or inconsistent stereo gains, but recent approaches using dichoptic stimulation, perceptual learning, and virtual reality have demonstrated significant and sustained improvements.15,16 For example, Vedamurthy et al. (2015) reported dichoptic video game training improved stereoacuity by 0.27 log arcsec (46%) immediately post-training and 0.24 log arcsec (42%) at follow-up in adults with anisometropic amblyopia.17 Compared with these approaches, the stereopsis gains observed with AVT in the present study were modest, which may reflect differences in therapeutic emphasis, as AVT in this protocol prioritized task-based oculomotor and visuomotor training alongside binocular engagement rather than exclusively targeting interocular suppression. Similarly, Suwal et al. observed significant stereoacuity gains over three months with both active vision therapy (0.32 ± 0.11 logMAR) and occlusion (0.27 ± 0.19 logMAR), with greater improvement in the AVT group with severe amblyopia (p = 0.03) .5 Variations across studies likely reflect differences in age, baseline stereoacuity, amblyopia severity, therapy type, intensity, duration, adherence, and outcome measures. Compared with dichoptic or perceptual learning therapies, AVT appears comparable for BCVA and gain in oculomotor task performance, while stereopsis improvements were modest, likely reflecting ceiling effects in children with mild-to-moderate amblyopia and limited statistical power.18

Although the AVT–occlusion group demonstrated a statistically significant improvement in contrast sensitivity at three months (mean difference = 0.06 log units, p = 0.005). This change is modest and may approach the known test–retest variability of the Pelli-Robson chart, typically reported as 0.05–0.15 log units, indicating that changes smaller than this range may reflect measurement variability rather than true clinical change.25 Therefore, while the improvement suggests a positive trend, its clinical relevance should be interpreted cautiously.

Functional associations between gain in oculomotor task performance and sensory improvements further emphasize the interdependence of sensory and motor recovery.26 Improvements in task-based saccadic accuracy (ρ = −0.43, p = 0.006) and smooth pursuit (ρ = −0.53, p = 0.001) performances were significantly correlated with BCVA gains, and faster task-based saccadic response time (ρ = 0.35, p = 0.026) was associated with better stereopsis. These findings support the hypothesis that improvements in oculomotor task performance are not merely secondary effects but may contribute functionally to visual recovery, reinforcing the rationale for integrating training on visuomotor task performance into amblyopia rehabilitation. This supports the view that amblyopia is not solely a monocular acuity deficit but involves binocular and visuomotor dysfunction that benefits from multi-component rehabilitation.27 Age and gender were not included as covariates in these association analyses because baseline distributions were comparable between groups, and the study was not powered to support age- or sex-specific subgroup analyses.

The observed improvements in visuomotor task performance likely reflect enhanced functional task execution rather than direct evidence of neural changes. AVT tasks, including saccadic jumps, smooth pursuit tracking, vergence, and binocular coordination, are designed to engage visuomotor and attention-related processes. Previous studies have implicated regions such as the middle temporal visual area, parietal cortex, and frontal eye fields in motion processing and visuospatial control28–30, which are relevant to the types of tasks used in the present study. However, the current findings should be interpreted as improvements in behavioural task performance, as no direct neurophysiological or eye-tracking measurements were performed. In contrast, occlusion therapy primarily targets monocular form vision, which may partly explain the comparatively smaller improvements observed in visuomotor task performance in the occlusion-only group.1

Clinically, structured task-based saccadic, pursuit, and binocular coordination exercises can complement traditional occlusion therapy, offering a more comprehensive approach by targeting both sensory and motor components of visual function.31,32 This is particularly relevant given that many contemporary amblyopia therapies emphasize sensory recovery, whereas functional visual performance in real-world tasks depends heavily on efficient oculomotor task performance. Although the observed changes were statistically significant, further research is needed to determine the extent to which these improvements translate into real-world functional benefits such as reading efficiency, academic performance, and visuomotor coordination.

Several limitations should be considered when interpreting these results. First, the use of an alternating allocation strategy rather than true randomization represents a key methodological limitation. This quasi-experimental design introduces potential selection bias and limits internal validity, thereby weakening the strength of causal inferences. Although baseline characteristics were comparable, unmeasured confounding cannot be excluded.

Second, although sample size estimation was informed by prior stereopsis outcomes, the study was not originally powered based on a predefined primary endpoint, and therefore findings should still be interpreted with consideration of potential type II error for smaller effects. In addition, the absence of allocation concealment and masking (blinding) of outcome assessment may have introduced performance and assessor bias.

Third, an important consideration is the potential influence of practice (learning) effects. AVT participants were repeatedly exposed to the same computerized tasks during training and assessment without prior familiarization. Consequently, some improvements in task-based saccadic and smooth pursuit performance may reflect increased task familiarity, enhanced hand–eye coordination, or strategy optimization rather than purely physiological changes in oculomotor control. While the occlusion-only group followed the same assessment schedule, additional exposure during AVT sessions may have amplified these learning effects. Therefore, between-group differences should be interpreted cautiously, as they may partially reflect practice-related gains rather than solely treatment-specific effects. Finally, follow-up was limited to three months, precluding long-term outcome assessment.

Given the non-randomized allocation and modest sample size, adjusted analyses including multiple covariates (e.g., age, baseline amblyopia severity, and treatment adherence) were not performed to avoid overfitting and unstable estimates. Baseline comparability for key demographic and clinical variables was confirmed, and occlusion adherence was broadly similar across participants. Nonetheless, residual confounding related to age, baseline severity, or adherence cannot be excluded and may have influenced between-group comparisons and effect estimates. Taken together, the absence of predefined covariate adjustment, non-random allocation, and lack of task familiarization introduce potential sources of bias that may have influenced between-group estimates and should be considered when interpreting the magnitude of observed effects.

In summary, AVT combined with occlusion is associated with accelerated and improvements in task-based saccadic and smooth pursuit performance, alongside visual acuity gains, in children with anisometropic amblyopia. By directly addressing oculomotor task performance, a relatively underexplored domain in amblyopia intervention studies, these findings support a function-oriented, multimodal approach to amblyopia rehabilitation. Larger, randomized trials with long-term follow-up and functional outcome measures are warranted to confirm these findings and establish optimal AVT protocols in pediatric amblyopia.

Active Vision Therapy combined with occlusion was associated with greater and faster improvements in both oculomotor task performance and visual function in children with anisometropic amblyopia compared with occlusion alone. Specifically, AVT–occlusion demonstrated greater improvements in task-based saccadic accuracy, task-based saccadic response time, and task-based smooth pursuit performance, along with earlier improvements in BCVA and stereopsis. These results highlight a potential association between sensory and visuomotor functional recovery and support the use of function-oriented, binocularly based vision therapy protocols as adjuncts to conventional occlusion therapy. However, these findings should be interpreted with consideration of potential practice effects related to repeated task exposure, and studies incorporating gold-standard eye-tracking, longer optical adaptation periods, and randomised control designs are needed to confirm these preliminary findings.

None.