Postural stability and visual impairment: Assessing balance in children with strabismus and amblyopia

Abstract

Background: Vision plays an important role in controlling posture and balance in children. Reduced postural control has been reported in children with strabismus, but little has been reported specifically in amblyopia.

Objective: To investigate whether children with amblyopia have reduced balance compared to both children with strabismus without amblyopia and healthy controls.

Study design and methods: In this cross-sectional study, a total of 56 patients and healthy controls were recruited from the Ophthalmology and Otolaryngology Clinics at The Hospital for Sick Children, Toronto. Participants were divided into three groups: (1) 18 with unilateral amblyopia (strabismic amblyopia or mixed mechanism); (2) 16 with strabismus only without amblyopia; and (3) 22 visually-normal controls. The primary outcome was the balance performance as measured by the balance subtest of the Bruininks-Oseretsky Test of Motor Proficiency 2 [BOT2].

Results: The age and gender-adjusted BOT2 balance scores were significantly reduced in the amblyopia group (mean score 9.0 ± 3.1 SD) and the strabismus without amblyopia group (mean score 8.6 ± 2.4 SD) compared to visually normal controls (mean score 18.9 ± 4.2) (p<0.0001), but no statistical difference was demonstrated between the two patient groups (p = 0.907). Further subgroup analysis of the strabismus only group did not reveal a statistically significant difference in performance on BOT2 balance score between strabismus only patients with good stereopsis 60 sec or better (BOT2 mean score 9.8±3.0 SD) to patients with 3000 sec or no stereopsis (BOT2 mean score 7.9±1.7) (p = 0.144).

Conclusion: Our findings suggest that normal vision plays an important role in the development and maintenance of balance control. When normal binocular vision is disrupted in childhood in strabismus and/or amblyopia, not only is the vision affected, but balance is also reduced. Our results indicate that the presence of even mild binocular discordance/dysfunction (patients with intermittent strabismus and good stereopsis) may lead to postural instability.

Fig 1. Age and gender adjusted BOT2 score among the different groups.

Diagram A is a boxplot illustrating the distribution of the BOT2 corrected point score for age and gender (range 1–35) among the three groups: normal controls, strabismus only without amblyopia group and the amblyopia group. The definition of the corrected point score is included in the methods section. The inter-quartile area of the strabismus only group and the amblyopia are much lower (7.0–9.25 score and 7.0–10.25 score, respectfully) compared to the control group 15.75–22.25 score. The blue area in the graph is the average corrected score level (11–19) according to descriptive categories of the BOT2 test. The strabismus only group (mean score 8.6 ± 2.4 SD) and the amblyopia group (mean score 9.0 ± 3.1 SD) had much lower scores compared to the normal controls group (mean score 18.9 ± 4.2 SD) [F (2, 53) = 60.62, px10^-3], but the mean corrected BOT2 score of these two groups was not statistically different. Diagram B shows the distribution of the BOT2 corrected scale score according to five descriptive categories of performance based on the BOT2 test’s nomogram (Well-Above Average, Above Average, Average, Below Average and Well-Below Average). The graph demonstrates the number of subjects distributed in each descriptive category among the three study groups (strabismus only group, amblyopia group and normal controls). The performance of all of the subjects in the control group was categorized as average and above, while most of the strabismus group and the amblyopia group performed below average. ^o Signifies an outlier that is outside of the 95 percentile. * Signifies statistically significant difference from normal controls using one-way ANOVA.

Fig 2. Means of individual static balance tasks scores among the groups.

The diagram illustrates the difference in means of the individual task time in seconds (time to loss of balance) among the different groups. Included in this analysis are the static tasks: 1- Tandem stand with eyes open (EO), 3- One-foot stand EO, 4- Tandem with eyes closed (EC), 6- One-foot stand EC, 7- One-foot stand on a balance beam (BB) EO, 8- Tandem stand on a beam EO, 9- One-foot stand on a beam EC. Task 6–9 demonstrated a statistically significant difference between the two patient groups and the visually normal controls. Both groups were not statistically different from each other. * Signifies a statistically significant difference among the groups using one-way ANCOVA with age and gender as co-variant. EO = eyes open. EC = eyes closed. BB = balance beam.

Fig 3. Age influences the difference in performance between viewing conditions.

This scatter plot demonstrates the effect age has on the performance between viewing conditions (eye open vs eyes closed). The y-axis is the difference between the performance measured in seconds (time to loss of balance), while the eyes are open (EO), versus the time to loss of balance while the eyes are closed (EC). The BOT-2 scores have age-adjusted standards for the balance subtest, however the data presented in these graphs are not corrected for age, as this information is not available for specific tasks. Fig 3(A) shows that on the tandem task this difference in performance between viewing conditions is markedly reduced in patients older than 10 years of age. Fig 3(B) shows that on the one-foot stand this difference is reduced in participants older than 12.5 years of age. The red line signifies the approximate age, beyond which the difference between viewing conditions is markedly reduced and no longer variable. The older the kids are the difference in their performance is less influenced by the viewing conditions, and they perform with their eyes closed almost as well as with their eyes open.

Fig 4. Mean score of three static tasks among the groups in different viewing conditions.

The diagram illustrates the difference in means of three different tasks the tandem task (A), the one-foot stand task (B) and the one-foot stand on a beam (C). The performance was measured as time to loss of balance in seconds with the eyes open (EO) and with the eyes closed (EC) among the different groups. Using the one-way repeated measures ANCOVA the amblyopia and the strabismus only group performed worse with their eyes closed compared to normal controls on the tandem and the one-foot stand (p<0.001), but this was not statistically significant on the one-foot stand on a beam (p = 0.25). The results presented in this figure are for males. Similar reduction in performance while the eyes were closed were also demonstrated in females. *Covariates appearing in the model are evaluated with the age value at 10.03 years. EO = eyes open. EC = eyes closed.