To assess the effects of glasses for anisometropia on stereopsis and to determine the factors that affect the level of stereopsis.
Retrospective observational case series.
One hundred six nonamblyopic patients who were wearing glasses for anisometropia and 56 who were wearing glasses for isoametropia were enrolled. The levels of stereopsis in the anisometropic patients were divided into normal (≤40 seconds of arc), equivocal (40 < – ≤ 100), and subnormal (100 < – ≤ 400) and compared with those in the isoametropic patients. It was evaluated whether the amount of interocular difference in the lens power of the glasses, the type of anisometropia, a history of amblyopia, and the age at the time of the prescription of the first glasses were related to the stereopsis.
In the anisometropia, the mean stereopsis (seconds of arc) was 77.52 (40-200) in the Titmus-fly test and 52.78 (40-100) in the Randot stereotest. The rate of normal and equivocal stereopsis was 87.7% in the Titmus-fly test and 96.9% in the Randot stereotest. The isoametropic patients demonstrated better stereopsis (52.86 and 39.20 in either test) than did the anisometropic patients ( P < .05). The stereopsis was worse in the spherical hyperopic type of anisometropia than in the spherical myopic type ( P < .05). The level of stereopsis was not related to the other factors that were investigated.
The level of stereopsis with the wearing of anisometropic glasses was clinically near normal and the glasses did not seriously affect the binocular vision regardless of the severity of the anisometropia.
The prevalence of anisometropia in children has been estimated as 2%∼3.8%. Anisometropia is one of the basic defects in childhood amblyopia and strabismus. The conventional therapy is correction of the refractive error using anisometropic glasses in children with amblyopia or strabismus to enhance their optimal visual development. However, the principled approach to anisometropic refractive error in children who do not have strabismus or amblyopia remains to be determined. It is perplexing because they can see well with their better eye without depending on glasses, but they are still at risk for abnormal binocular function even though they do not have strabismus or amblyopia yet. Physicians often hesitate to prescribe anisometropic glasses because they worry that it would be hard to adapt to the glasses and that the glasses would cause aniseikonia.
Aniseikonia, a perceived retinal image size difference, could lower the compliance with glasses wear and disrupt binocular functions such as fusion and stereopsis. Katsumi and associates suggested that the binocular system can compensate for up to a 3.0% difference in the perceived retinal image size, but in higher aniseikonia the binocular system can no longer compensate for the difference and binocular inhibition takes place. Campos and Enoch reported that a larger than 5% aniseikonia resulted in loss of stereopsis. Most of the previous studies on the relationship of aniseikonia to binocularity, however, were experimental. Indeed, it has not been clearly revealed that patients with natural anisometropia have aneiseikonia. In the report by Kramer and associates, 70% of the axial anisometropia patients had aniseikonia, 30% had none, and all the patients with combined refractive and axial anisometropia had aniseikonia. Lubkin and associates reported that natural anisometropia is correlated with an increase in aniseikonia, and that anisometropia has a major causal role in amblyopia, a role that is augmented by aniseikonia. Winn and associates examined aniseikonia in anisometropic patients when corrected with glasses and contact lenses, and they insisted that contact lens correction keeps the aniseikonia at a minimum level, and that contact lenses can thus be a better alternative than glasses.
There are insufficient empirical and pragmatic data, however, on natural anisometropia in children with stereopsis in the actual clinical setting. Lubkin and associates reported the retention of stereopsis in adults with monocular aphakia and good visual acuity in glasses correction for anisometropia. In their report, large groups of aphakic patients rarely had subjective complaints clinically. Lubkin and associates commented that although over 5% of aniseikonia cases are generally considered incompatible with binocularity, it still remains to be seen if tolerance to this defect may be higher than physicians have suspected. Hwang and associates recently reported a good visual outcome and a good final stereopsis with anisometropic glasses after part-time occlusion therapy in children with anisometropic amblyopia. However, that study was on the outcome of amblyopia treatment and not a discrete analysis of the effect of anisometropic glasses on stereopsis.
This study was conducted to assess the effects of glasses for anisometropia on stereopsis in actual patients with no amblyopia or successful cessation of their amblyopia therapy. The factors that might affect the level of stereopsis were also investigated.
Children who were prescribed glasses at a single institution were retrospectively identified. Approval by Hallym University Sacred-heart Hospital Institutional Review Board/Ethics Committee was obtained for this retrospective case series, and all the study procedures were conducted following the tenets of the Declaration of Helsinki.
The anisometropic subjects were selected using clinical records from January 1, 2002 to December 31, 2010. The eligibility criteria were: (1) the presence of anisometropia and the wearing of anisometropic glasses; (2) absence of amblyopia or complete resolution of previous amblyopia (interocular difference ≤1 Snellen line for the best-corrected visual acuity [BCVA]); and (3) measures of near stereoacuity with glasses. The stereoacuity should be that which was obtained after the cessation of the patient’s amblyopia treatment and when the patient was older than 6 years, to ensure the reliability of the data. If the child was tested during multiple visits after age 6 and after the cessation of the amblyopia treatment, the most recent result was chosen. The exclusion criteria were remaining amblyopia (interocular difference ≥2 Snellen lines for the BCVA) and the presence of any strabismus in the alternate cover-uncover test. Patients with a history of retinal or optic nerve disease, glaucoma, cataract, nystagmus, media opacities, ocular trauma, neurologic disorder, and mental retardation were also excluded.
For all the patients, a detailed ophthalmic examination, including measurement of BCVA, slit-lamp examination, manifest and cycloplegic refractions, tests for ocular alignment, pupillary examination, and fundus examination, was conducted. Cycloplegic refractions were performed after the instillation of 3 drops of cyclopentolate (1%) and phenylephrine (0.5%) eye drops. All the refraction procedures were performed with a handheld retinoscope by the same pediatric ophthalmologist (the corresponding author). Each lens power value for optimum refractive correction was determined by the same pediatric ophthalmologist (the corresponding author). Especially, the power over the more ametropic eye was re-evaluated to determine the BCVA at each visit, and new glasses were prescribed for better visual acuity, whenever they were needed.
In this study, anisometropia was defined as an interocular difference between the lens power of glasses in the spherical equivalent of ≥1.0 diopter (D). The types of anisometropia were classified into spherical, astigmatic, and mixed. The spherical type was defined as that with a ≤1.5 D interocular difference in the cylinder. It was subdivided into the spherical myopic and spherical hyperopic types, depending on the spherical refractive error in the bad eye. The patients with a >1.5 D interocular difference in the cylinder were assigned to the astigmatic type if they had a <1.0 D interocular difference in the sphere, or to the mixed type if they had a ≥1.0 D interocular difference in the sphere. For example, a patient with a lens power of +1.00 D = +0.75 D × 90 degrees in his right eye and +4.00 D = +1.75 D × 90 degrees in his left eye was classified into the spherical hyperopic type, and another patient with a lens power of +0.50 D = +0.50 D × 90 degrees in his right eye and +4.50 D = +2.50 D × 90 degrees in his left eye was classified into the mixed type. The severity of the anisometropia was defined as the amount of interocular difference in the spherical equivalent between the lens powers of the glasses.
To set up a control group for the comparison of anisometropia, the clinical records of the patients who visited the authors’ clinic during a period of 4 months, from May to August 2010, were reviewed and the patients who wore glasses for isoametropic refractive errors were selected. The inclusion and exclusion criteria were the same as those for the anisometropic subjects, except for the criteria for the interocular difference for isoametropic glasses (difference in the spherical equivalent of <1.0 D between the lens powers of the 2 eyes).
Finally, 106 subjects in the anisometropia group and 56 subjects in the control group were enrolled. The following information was obtained from each medical record (items 1-4 were obtained from both groups and items 5-7 were obtained from the amisometropia group): (1) the age at the time of performance of the stereopsis test from which the stereoacuity that was analyzed in this study was obtained; (2) the subject’s sex; (3) the stereoacuity; (4) the refractive error; (5) the severity of the anisometropia; (6) history of amblyopia treatment (none, occlusion therapy, or atropine penalization); and (7) the age at the time of prescription of the first glasses. Stereoacuity was measured using the Titmus-fly stereotest (Stereo Optical Co, Inc, Chicago, Illinois, USA) and the Randot stereotest (Stereo Optical Co, Inc). All the patients underwent a stereopsis test with their own glasses. There have been no clear definitions of normal stereoacuity that stands for bifixation. Forty to 60 seconds of arc is usually accepted as normal stereopsis. The range of 60-100 seconds of arc is considered equivocal (uncertain but probably normal) and 100-400 seconds of arc, subnormal binocularity. Therefore, in this study, the level of stereopsis (seconds of arc) of ≤40 was considered normal stereopsis; of 40 < – ≤100, equivocal; and of 100 < – ≤ 400, subnormal.
To determine the factors affecting the stereopsis in anisometropia, it was evaluated whether or not the amount of interocular difference in the lens power of the glasses, the type of anisometropia, a history of amblyopia, and the age at the time of the prescription of the first glasses were related to the stereopsis.
For the statistical analysis, SPSS software (v.17.0; SPSS Inc, Chicago, Illinois, USA) was used. A χ 2 test, a paired t test, an independent t test, ANOVA, and a Pearson correlation analysis were performed. The post hoc tests for ANOVA were performed with the least significant difference (LSD) and Tukey honestly significant difference (HSD) tests. The null hypothesis was rejected at the .05 level of significance.
The level of stereopsis and the baseline characteristics of the 106 patients in the anisometropia group and the 56 patients in the control group are provided in Table 1 . The age and sex distributions did not significantly differ between the 2 groups. Among the 106 anisometropic patients, 97 underwent both tests and 9 underwent only the Titmus-fly stereotest based on the medical records. All 56 patients in the control group underwent both tests. There was a high correlation between the stereopsis that was obtained via the 2 tests of each patient (r = 0.608 and P = <.01 for the anisometropia group and r = 0.675 and P = <.01 for the control group, based on the Pearson correlation analysis). The stereopsis was worse in the anisometropia group, and the proportion of patients with normal stereopsis was also smaller in the anisometropia group ( Table 1 ).
|Sex (male:female)||49:57||29:27||.51 b|
|Age a (range)||9.42 ± 3.06 (6-21)||9.04 ± 2.32(6-16)||.40 c|
|Refractive error (SE)||3.38 ± 2.16 d||3.10 ± 1.88 e|
|Severity of anisometropia||4.83 ± 2.70|
|Stereopsis (range) by Titmus test||77.52 ± 37.79 (40-200)||52.78 ± 46.59 (40-100)||<.01 c|
|normal stereopsis||18/106 (16.9%)||26/56 (46.4%)||<.01 b|
|equivocal stereopsis||75/106 (70.8%)||30/56 (53.6%)|
|subnormal stereopsis||13/106 (12.3%)||0/56 (0%)|
|Stereopsis (range) by Randot stereotest||52.86 ± 29.38 (25-200)||39.20 ± 17.34 (25-100)||<.01 c|
|normal stereopsis||45/97 (46.4%)||39/56 (69.6%)||<.01 b|
|equivocal stereopsis||49/97 (50.5%)||17/56 (30.4%)|
|subnormal stereopsis||3/97 (3.1%)||0/56 (0%)|
On the types of anisometropia, 31 cases were of the spherical myopic type, 51 were of the spherical hyperopic type, 16 were of the astigmatic type, and 8 were of mixed type. The stereopsis did not significantly differ among the 4 types of anisometropia ( Table 2 ). However, the spherical hyperopic type seemingly reduced the stereopsis with marginal significance ( P = .07 for the Titmus test and P = .06 for the Randot stereotest, from ANOVA). The post hoc tests for ANOVA showed a significant difference between the spherical hyperopic type and the spherical myopic type in the LSD test ( P = .02 for the Titmus test and P = .02 for the Randot stereotest). In the Tukey HSD test, there was a marginally significant difference between the spherical hyperopic type and the spherical myopic type ( P = .08 for the Titmus test and P = .07 for the Randot stereotest). The other comparisons between types in the post hoc tests did not yield statistically significant results ( P = .25-.99). Therefore, the comparison was confined to spherical anisometropia, and the stereopsis was worse in the spherical hyperopic type than in the spherical myopic type ( P = .03 for the Titmus test and P = .02 for the Randot stereotest, from the independent t test).