Prevalence of anisometropia in children and adolescents

Introduction

Anisometropia is an ocular disorder characterized by an interocular difference (IOD) in refractive error. It represents a specific refractive condition where the two eyes of an individual, can have asymmetric eye growth.¹ This condition can occur in situations of myopic, hyperopic or astigmatic asymmetry and is strongly associated with the development of other eye changes such as aniseikonia, amblyopia, diplopia and strabismus.^2,3

Although there is no uniformly defined dioptric value for its clinical classification, an IOD in the spherical equivalent (SE) of 1 diopter or more is accepted as the threshold, for most authors.^1,4–8 However, even using this limit, the scientific literature presents a significant variation in the prevalence values of anisometropia in terms of age, gender and ethnicity.^4,5,8,9 Factors associated with lifestyle and educational level have also been referred to as risk factors for anisometropia.^7,8,10

Early detection and early treatment is crucial to prevent permanent visual loss. Although it is not clear what is the ideal age to perform the correction, in order to guarantee an ideal visual development and maturation, the early correction of anisometropia is important.¹¹ This prevents the development of other changes such as aniseikonia, amblyopia and strabismus,^3,12,13 and even in small degrees (<1D) facilitates emmetropization.¹¹ It also improves quality of life, reducing or eliminating symptoms of visual discomfort. In this way, visual screening at a young age is useful in identifying who is most likely to benefit from early optical correction or preventive treatment.^11,14,15

The clinical methods used to characterize anisometropia are refractive techniques, with autorefraction, using Plusoptix, one of the most recommended techniques for screening activities.^16,17 This instrument allows quantifying the refractive error in which it is possible to obtain a very similar value to that obtained by cycloplegic refraction^17–19 and with excellent precision in anisometropia signalling.¹⁹

Studies on the prevalence of anisometropia focus on children or adults, with less research being found in adolescence.

The aim is to estimate the prevalence of anisometropia (spherical and astigmatic) and to analyse its pattern of variation in a sample of children and adolescents, from preschool education (from three to six years old) to the various cycles of basic education (from the 1^st to the 9^th school year in Portugal, from six to fifteen years old).

Methods

This is an observational cross sectional study to evaluate the prevalence of anisometropia in children and adolescents. Data from the Portuguese Census 2011, showed a population of 319284 from 0 to 14 years old in the central region.²⁰ For a confidence level of 95% and a 5% margin of error, taking into account that the prevalence of the studied condition is unknown, it was fixed at 50% to obtain a large enough sample size. The result was a minimum of 384 subjects.

The data collection took place in five schools of the central region of Portugal, including all students that were authorized to participate by their legal guardians, between October 2018 and February 2019. Participants were 749 children and adolescents aged between three and sixteen years. Students data for which it was not possible to obtain refraction were excluded, due to technical issues associated with the performance of the instrument (presence of strabismus, opacities, retinal anomalies or when the refractive error exceeded the instrument's measurement limit - spherical measurement range or cylindrical from −7.00 to +5.00D) or due to lack of cooperation from the participant. Only two participants were excluded.

The refractive error was measured with the paediatric auto refractometer model A09 by PlusOptix, Nuremberg, Germany, by the average of three consecutive measurements, binocular and without the use of a cycloplegic. The PlusOptix allows measuring the refractive error in open field simultaneously in both eyes and under the same conditions, in a fast, easy, safe, non-invasive way, at a distance of one meter from the subject's eyes.

Results

Refractive state

According to the classification criteria previously defined for the classification of refractive state, it was concluded that in the study sample 71.16% (n = 533) were emmetropic. Among subjects with significant refractive error (n = 216), it was found that hyperopia was the most prevalent refractive error (n = 109, corresponding to 14.6% in the studied population) followed by myopia (n = 49, corresponding to 6.5% in the population), anisometropia (n = 46, corresponding to 6.1% in the population) and astigmatism (n = 25 where 12 were cases of simple astigmatism and 13 compound astigmatism). Figure 1 shows graphically the distribution of the different refractive states in the study sample. The representation of astigmatism, refers only to the occurrence of simple astigmatism, without a significant SE. The mean values of refractive errors, according to the previous classification, are shown in Table 3.

Figure 1. Refractive error distribution.

Table 3. Mean values and standard deviation of refractive errors of subjects with significant ametropia (right eye data).

Refractive group	Wearing glasses (n)		Spherical equivalent (D)		Cylindrical component (D)
	Yes	No	Mean ± SD	Range [min; max]	Mean ± SD	Range [min; max]
Myopia	44	5	−2.67 ± 1.32	−5.79; −0.92	−0.39 ± 0.33	−1.29; 0.00
Hyperopia§	22	85	+1.46 ± 0.59	+1.00; +5.25	−0.41 ± 0.47	−2.50; 0.00
Simple astigmatism	6	6	+0.29 ± 0.55	−0.83; +1.21	−1.50 ± 0.33	−2.07; −1.14
Anisometropiaϯ	37	8	Interocular difference
			1.53 ± 0.82	0.00; 3.79	0.78 ± 0.85	0.00; 3.9

§ Two cases without data regarding the use of glasses.

ϯ One case without data regarding the use of glasses.

Influence of sociodemographic variables

The study of the influence of sociodemographic variables in the occurrence of anisometropia is presented in Table 4. The Chi-square test indicates that there was no association between MA and any of the factors under analysis.

Table 4. Relationship of gender, area of residence and study cycle, in anisometropia.

	Gender				Chi-square test
	Female (350)		Male (399)
	N	%	N	%	P
TA	25	7.1	21	5.3	0.29
SA	21	6	16	4	0.21
MA	8	2.3	7	1.8	0.60

	Area of residence				Chi-square test
	Rural (320)		Urban (423)
	N	%	N	%	p
TA	22	6.8	23	5.4	0.42
MA	7	2.2	8	1.9	0.78
SA	18	5.6	18	4.3	0.39

	School stage								Chi-square test
	Preschool (103)		1st cycle (231)		2nd cycle (181)		3rd cycle (234)
	N	%	N	%	N	%	N	%	p
TA	3	2.9	10	4.3	11	6.1	22	9.4	0.06
MA	2	1.9	3	1.3	6	3.3	4	1.7	0.52
SA	1	1.0	9	3.9	7	3.9	20	8.6	0.012*
p(adjust)	0.36		<0.001*		<0.001*		0.015*

* Significant at the 0.05 level.

No significantly different occurrence of anisometropia was found according to gender or area of residence (p > 0.05). In relation to the school, there was a pattern of variation that increased with the cycle of studies, ranging from 2.9% in preschool education to 9.4% in the 3^rd cycle of studies, however this association was only statistically significant for SA (χ²₍₃₎ = 10.918; p = 0.012), where there was a rate of 1% in preschool education and 8.6% in the 3^rd cycle. It should be noted that the study cycle is dependent on age.

Figure 2 illustrates the distribution of anisometropia according to the type of refractive error, for each school cycle. It is observed that myopic anisometropia was present in all study cycles, registering a considerable increase from the beginning of school, that is, from the 1^st cycle to the 3^rd cycle. On the other hand, hyperopic anisometropia was manifested on a larger scale in children of the 1^st cycle, with a lower occurrence in the following cycles. As for the simple MA, it was present in all study cycles and there was no specific pattern in its variation with the study cycle progress.

Figure 2. Relative frequency of the variation of the various types of anisometropia, with the school cycle.

Discussion

An anisometropia rate of 6.1% was found, considering the spherical and meridional anisometropia. It was found that this rate varies with the cycle of studies, showing an increase as the level of education advances, ranging from 2.9% in preschool education to 9.4% in the 3^rd cycle of basic education. No statistically significant differences were found in the distribution of anisometropia either between genders or between areas of residence. It was also found that myopic anisometropia was the most prevalent (46%), with a considerable increase in the 3^rd cycle of studies.

Table 5 summarises worldwide epidemiological data regarding studies in children and adolescents, which used the same criterion for anisometropia definition (IOD ≥ 1D).

Comparing with others studies, some authors point to a frequency similar to the one found on this research,^4,7,8,11,25 others point to a lower frequency²⁴ and others still refer to a higher frequency.^22,23 Studies that included only children aged five and six years report rates of 1.3% and 1.6% (SA).^5,26 The frequency of SA, found in the present study, in preschool education (children from three to six years old) was 1%, this value being closer to those studies. Other studies included participants from 15 to 19 years old and report rates of 11,2% for SA.²³ For the 3^rd cycle of studies (average between 12 and 15 years of age), the present study indicated a frequency of 8.6% (SA) and according to the observed variation pattern, at a more advanced age and school stage, a higher rate is expected.

Geographical and methodological issues make it difficult to compare prevalence studies. There is a pattern of greater variability in studies on the Asian continent, where for an identical age group, there are records ranging from 2.5%²⁴ to more than 10%;⁷ however this is the continent where more studies on the subject are found. In Europe, more similar results are found, 4,6% and 6,9%.^4,11 In Portugal, a study in children from 6 to 11 years old found a 0.7% rate (5 in 672) of uncorrected anisometropia,²⁷ similar to the 1.1% (8 in 749) found in the present study.

In the literature, the pattern of variation of anisometropia as a function of age and during the school period is not clear, presenting discordant results. Although many studies focus on children, it is common to have relatively wide age ranges. Most authors conclude that anisometropia varies with age,^4,5,7,28 although there are also studies where this relationship has not been found.^10,13 Studies on the subject, at school age, which showed an increasing prevalence with age were carried out in populations with a high frequency of myopia, a condition whose prevalence increases in adolescence.²⁸ On the other hand, longitudinal studies show that the prevalence of anisometropia increases after children start attending school.^5,7,25 Given these two lines that justify the variation of anisometropia during school age, it appears that they are related to each other, since the progress in the school path is accompanied by increasing age. In the present study, the anisometropia prevalence was also found to increase with advancement in the level of education. Consequently the age factor is also contributing to this situation, with myopic anisometropia being the one that most contributes to this variation pattern.

Regarding the influence of gender on anisometropia, contradictory data are found in the scientific literature. While in one study it is reported that the prevalence found was higher in males²⁸ in others it is reported that the prevalence rates are higher in females.^10,23 The results of the present study revealed a higher frequency in females (7.1%) than in males (5.3%), however these differences were not statistically significant. This finding is in line with the results of other authors.^9,13,25,26

The influence of living in rural or urban areas has also been the object of study by several researchers, considering the development of myopia and, consequently, myopic anisometropia, which is more pronounced in urban areas.^7,8,10 The present study did not prove whether the area of residence and the frequency of anisometropia were related. This parameter is highly dependent on the way in which each author classifies the area of residence, as rural or urban, and the limits of these regions are sometimes difficult to define. Also, living in a rural area and working in an urban area means that the daily experience of some populations turns out to be more urban, in both environments. Some studies show that lifestyle parameters, such as reading and writing habits and time spent indoors, contribute to the prevalence of anisometropia variation.^7,8,10,11

This study has several strengths. The study was carried out in children and adolescents in a school environment. This allows minimizing the potential bias that occurs in the sampling process and avoids overestimation of the problem, when the investigation is carried out in a clinical environment. For this study, spherical anisometropia and astigmatic anisometropia were considered. The latter being disregarded in several studies and there is a record that this anisometropia is the one that most varies in terms of ethnicity.⁹

Certain limitations can be pointed out in this study. Firstly, the use of autorefraction without cycloplegia is highlighted, which is not the gold standard method of refraction. This fact limits the analysis of the distribution of the different types of refractive errors found in the population studied. Despite the use of an open field auto refractometer, recognized as an instrument with good agreement with cycloplegic retinoscopy and which eliminates the need for cycloplegia in children^17,18 tends to underestimate hyperopia.^17,19 However, for the present study, this situation does not weaken the conclusions, as the refraction was evaluated in an open field and binocularly, both eyes are exposed to the same conditions. In addition to that studies by other authors with the same methodology, point out a sensitivity of 100% for the diagnosis of anisometropia.¹⁹ Secondly, the choice of the cut-off point for the classification of anisometropia is pointed out. The literature recommends considering an IOD of at least 1.00D, however the sensitivity studies of the auto refractometer used, recommend considering 1.25D.¹⁶ Thirdly, the signalling of the area of residence of the participants as rural or urban is reported, since the limits of these regions were at times difficult to define. Noting also that being a study carried out in an area of the inner country, whose territorial classification is of low density, it is expectable that habits and behaviours are more uniform between rural and urban areas. This wouldn’t be the case if the same study had been carried out in an area of greater population density.

For future studies, data regarding reading habits and time spent outdoors will be collected, since these can be risk factors for the development of refractive errors.

Conclusions

The present study estimated the prevalence of anisometropia in Portuguese children from preschool to the 3^rd cycle of basic education finding an occurrence rate of 6.1%. The results of this work also showed that the level of the study cycle and the spherical anisometropia are related, verifying that it is low in preschool education and higher in the 3^rd cycle of basic education. It was also found that, hyperopic anisometropia was more prevalent in younger children and that with the progress of the school path, myopic anisometropia predominated. Regarding anisometropia, 17% of the subjects were uncorrected, which can be associated to an absence of visual related symptoms due to a possible amblyopia. Taking into account the cases of uncorrected anisometropia, in our opinion the implementation of visual screening programs is essential for the timely detection and correction of possible eye problems. This course of action will lead to better development, learning and school outcomes.

Data availability