Our study describes, for the first time to our knowledge, the prevalence of anisometropia and aniso-astigmatism in Northern Irish school children. While a prevalence of anisometropia (9.3%) similar to that found in our study has been reported by Shih et al. in a group of Taiwanese school children,
15 the high prevalence of anisometropia in Taiwan was attributed to the associated high prevalence of myopia in the study population, whereas in our study anisometropia at age 12 to 13 years was more common in children with moderate hyperopia than in those with myopia. A lower prevalence of anisometropia of 5.8% was reported in children in the United States aged 12 to 15 years where anisometropia was associated with myopic and hyperopic refractive errors.
12 However, the US study was not population-based and relied on non-cycloplegic measures of refractive error, so comparisons with our study are not straightforward. A further study of a UK adult population supports an association with ametropia (myopia and hyperopia).
23 It may be that our study participants were too young to demonstrate an association with myopia, which may become evident as the study population ages and more myopia emerges.
In our study aniso-astigmatism appeared to be associated with myopia in the younger age group and with moderate hyperopia in the older age group. However, the association with myopia must be treated cautiously as this was not found in the older age group and the prevalence of myopia was low in 6- to 7-year-olds. SER also has been used to classify refractive status and all four myopic 6- to 7-year-old subjects with aniso-astigmatism had been classified as myopic due to astigmatism in one eye.
White school children in Australia, where the prevalence of hyperopia and astigmatism is much lower,
17,22,24 have a much lower prevalence of anisometropia (1.6%) and aniso-astigmatism (0.8%) than similarly-aged white school children in our study. Support for an association between aniso-astigmatism and hyperopia comes from a study of American Indian children who have a high prevalence of aniso-astigmatism (15%) associated with hyperopia and astigmatism.
11
There was no significant difference in the distribution or prevalence of anisometropia or aniso-astigmatism between the two age groups. Previous reports on the prevalence of anisometropia with age have been equivocal, with some studies reporting no association
14,16 and others reporting an increased prevalence of anisometropia with age.
12,15,25,26 Many studies that show increasing prevalence with age have been done on populations with a high prevalence of myopia, the prevalence of which also is known to increase with age. Although the cross-sectional data from our study have shown that the prevalence of anisometropia did not vary statistically significantly between these two age groups, we cannot conclude that anisometropia remains stable throughout childhood or at an individual level. Much of the anisometropia seen in childhood is transitory.
8,27 Children at age 12 years with anisometropia may not necessarily have been anisometropic at age 6 years, with the converse also applying. Prospective follow-up data will help describe the change in anisometropia with age and its implication for visual function, including whether anisometropia and aniso-astigmatism precede hyperopia or occur as a result of it, and a review of our study's participants is underway.
The finding of an association of anisometropia with increased interocular differences in axial length, but not corneal curvature, is similar to other studies.
14,25 However, the Sydney Myopia Study also reports a statistically significant association between the prevalence of anisometropia and increased interocular variation in ACD measures (using the Zeiss IOLMaster) that was not found in our study. The reproducibility of using this instrument to measure ACD has been questioned previously,
28 as it relies on a subjective judgment by the operator and may explain the difference in findings. Aniso-astigmatism is associated not only with an increased interocular difference in corneal astigmatism, but also with axial length, possibly due to the relationship among aniso-astigmatism, hyperopia, and anisometropia.
It is difficult to explain how similar genetic and environmental influences cause an interocular difference in axial length growth. To date much of the research on anisometropia has concentrated primarily on the mechanism behind amblyopia development.
4 Although there is a clear association between strabismus and anisometropia,
9 with ocular dominance
29 and premature birth influencing anisometropia development,
30 no definitive biologic or environmental risk factors have been identified currently to our knowledge.
9 However it has been shown recently that spherical equivalent refractive error decreases to less hyperopia in the amblyopic eye of children who have had treatment for their amblyopia, and this emmetropization process is greatest in children with better ocular alignment.
31 It has been postulated that neural binocular coordination may be incomplete in anisometropia,
26 and it recently has been proposed that anisometropia may occur as a result of children adopting a reading position that results in an asymmetric working distance for the two eyes. This, in turn, would affect markedly the image the peripheral retina receives and the peripheral retina has been implicated in the development of refractive errors.
32 The association reported by this and other studies between anisometropia and higher spherical refractive errors may suggest that ametropic eyes are less efficient at controlling and coordinating ocular growth.
Anisometropia is an established amblyogenic factor,
4 and children with greater amounts of anisometropia have a higher prevalence and greater depth of amblyopia. It also is associated with esotropia even in the absence of amblyopia.
33 In our study, the prevalence of amblyopia was not ascertained, as neither best corrected acuities nor pinhole measures of acuity were made, so the relationship between anisometropia and amblyopia cannot be described. However, the visual acuity and binocular vision data give support to these associations: the prevalence of strabismus is higher in Northern Ireland than that reported in England, where the prevalence in 7-year-old children was 2.3% (95% CI, 2.0–2.7).
34 Although presenting interocular acuity differences were greater in anisometropia and aniso-astigmatism, it is possible that this was due to uncorrected or undercorrected refractive error. Further review of the participants will identify those with amblyopia and help to clarify any associations.
The opinions of professional groups and clinicians vary in terms of the amount of anisometropia that is considered a risk factor for the development of amblyopia.
4,35 Some investigators promote the use of a different cut-off point for differing types of anisometropia, for example a cut-off value of >1.00 DS for hyperopic anisometropia and >2.00 DS for myopic anisometropia, as amblyopia occurs more frequently and with lower amounts of anisometropia in hyperopic compared to myopic anisometropes.
5 The American Association for Pediatric Ophthalmology and Strabismus has introduced a set of standard risk factors that should be detected with preschool screening. These include anisometropia (either spherical or cylindrical) greater than 1.50 D as anisometropic amblyopia is rare, in the absence of strabismus, when anisometropia is less than 1.50 D. Although for the majority of patients there is a correlation between the severity of anisometropia and the degree of amblyopia,
36 Abrahamsson and Sjostrand suggested that the threshold for the development of anisometropic amblyopia may be as high as 3.00 DS
37 and further studies are required to discern the amount of anisometropia that is amblyogenic in the current population.