Prevalence of Color Vision Deficiency and its Correlation with Amblyopia and Refractive Errors among Primary School Children.

PURPOSE: To determine the prevalence of color vision deficiency (CVD) and its correlation with amblyopia and refractive errors among primary school children.
METHODS: In this population-based cross-sectional study, 2160 children were selected from 36 primary schools; 60 students were from each school (10 students in each grade), with equal sex distribution. A complete eye examination including refraction using a photorefractometer, determination of visual acuity (VA) and color vision using a Yang vision tester, and evaluation of ocular media opacity using a direct ophthalmoscope was performed. Children who could not answer at least 4 plates of the Ishihara color test were considered as color vision deficient subjects. Amblyopia was determined if pinhole VA was worse than 0.3 LogMAR (equal to 20/40).
RESULTS: The prevalence of CVD was 2.2% (95% CI: 1.5% to 3%) which was higher in male subjects (37 [3.5%] boys vs. 11 [1.0%] girls, P < 0.001). Mean VA was lower among students with CVD as compared to normal color vision children (P = 0.035) and amblyopia was observed in 8.3% (95% CI: 0.2% to 16.4%) of patients with CVD versus 2.1% (95% CI: 1.5% to 2.08%) of children with normal color vision perception (P = 0.005). A statistically significant correlation between lower VA and CVD was observed (P = 0.023).
CONCLUSION: Although CVD was correlated with lower VA and amblyopia, there was no relationship between CVD and the type of amblyopia, refractive error, anisometropia or strabismus.

INTRODUCTION

The prevalence of inherited red-green color vision deficiency (CVD) has been reported to be 8% and 0.4% in male and female individuals among European Caucasian populations[12345] and 4% to 6.5% among male subjects of Chinese origin.[567891011] In Saudi Arabia, the rate of CVD in female individuals has been 0.35%.[12] A marked difference between male and female ratios has usually been testified.[5]

Although decreased vision is the prominent sign of amblyopia,[13] there are certain associated microscopic anatomical and structural abnormalities in the retina,[1415] lateral geniculate body (LGB)[1516] and in area V1 of the visual cortex.[151718] Deficiency in other visual functions including contrast sensitivity, binocularity, the crowding phenomenon and visual evoked potential may be observed along with amblyopia.[19]

Patients affected with disorders such as optic neuropathies, macular diseases, media opacities and amblyopia show a higher prevalence of CVD; among these, amblyopic children demonstrate the lowest prevalence of CVD.[20] Color perception arises from signals generated by three cone photoreceptors with different spectral sensitivity functions. Signals from the retina which pass through the LGB are eventually transmitted to the cerebral cortex.[21] Transmission of color and motion information predominantly occurs by two major parallel pathways to the brain, where visual signals are reintegrated in the visual cortex. Retinal cells in the parvocellular pathway are responsible for fine and chromatic stimuli, while cells of the magnocellular pathway are responsible for moving and achromatic stimuli.[22] Some studies have revealed that monocular visual deprivation affects the size of parvocellular and magnocellular cells in the LGB which is more significant with long-term involvement of the eye,[2223] and may affect color vision. The color spectrum is perceived by different color wavelength sensitive cones which also have an effect on control of accommodation and refractive error (RE),[24] therefore problems with color perception may impact accommodation and refractive errors.

In the present study, we aimed to determine the prevalence of CVD and its correlation with amblyopia and refractive errors among primary school children in Tehran, Iran.

METHODS

In this population-based cross-sectional study, out of a total of 1781 primary schools in Tehran, 36 schools (including an equal number of public and private schools) were selected by random cluster sampling from different regions (North, South, West, East and Center) of Tehran, the capital city of Iran, from October 2013 to January 2014. There are six elementary levels in primary schools in Iran and in all selected schools, 10 students from each elementary grade were randomly selected for this study. The children were aged from 7 to 12 and the number of male and female subjects was equal.

A total of 2160 children were selected from 36 primary schools. These included 60 students from each school (10 students in each grade). Available data from 2150 children was used in the final analysis; incomplete data from 10 other children was discarded due to different reasons. Patients with mental retardation, ocular diseases or anomalies and disorders of fixation were excluded.

The study was approved by the Ethics Committee of the Ophthalmic Research Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran and Basir Eye Safety Research Center and adhered to the tenets of the Declaration of Helsinki. Formal written consent was obtained from the students’ parents prior to the study.

The participants were interviewed by school health instructors and a questionnaire concerning past health history and demographic status of the subjects was filled, then a complete eye examination including VA and color vision assessment, refractive error measurement, ocular deviation determination, anterior and posterior segments evaluation and red reflex observation was performed.

Visual Acuity and Color Vision Testing

We used the Yang vision tester (SIFI Diagnostic S.P.A-Via Castellana, 70/e-31100 Trevise, Italy) with constant luminance of 120 cd/m2 for VA examination using its Snellen E-chart with 5 letters on each line to include the effect of the crowding phenomenon. In addition, Ishihara color test on the Yang vision tester was used for CVD screening.

The examination was performed at 2 meters distance in day light. The Ishihara test consists of 16 plates and we asked children to read the colored number at the middle of each plate monocularly. If the child was able to read 13 plates or more correctly, the child was considered as normal color vision, otherwise she/he was considered to be color vision deficient. The order of presentation of the plates was changed from one eye to the other eye to prevent cheating.

If the child had glasses, VA and color vision were assessed with his/her own correction. The eye was suspected to be amblyopic if VA was 20/40 or less, then the measurement was repeated by a 2 mm pinhole and amblyopia was confirmed if VA did not improve to better than 20/40 with the pinhole. VA was classified according to pinhole VA into 20/20-20/30, 20/40-20/100, and less than 20/100.

Refractive Errors

The refractive status of all children was measured using a photorefractometer (PlusoptiX SO4 GmbH, Nürnberg, Germany) with no cycloplegia by a trained technician; this device has a reported sensitivity of 63%-94% and specificity of 62%-99%.[25] Photorefraction was repeated three times for each subject and the average result was used for statistical analysis. The measurement range of the photorefraction device is from −7.00 diopter (D) of myopia to +5.00 D of hyperopia, therefore cases with refractive errors out of this range were considered to have “high” refractive errors. Anisometropia was defined as spherical equivalent (SE) difference of at least one diopter between the fellow eyes. Myopia, hyperopia and astigmatism were considered as SE ≤−0.50 D, SE ≥+2.00 D and cylindrical power of 0.75 D or more, respectively.

Ocular Deviation

Ocular deviation was checked using the alternate cover test if VA was more than 20/200 and the Krimsky method in subjects with VA less than 20/200. The deviation was measured for near (33 cm) and far (6 m). In order to reveal any extraocular muscle dysfunction, the extraocular muscles were evaluated in nine different gazes.

Ocular Structure

We assessed the ocular media by examining the red reflex using a direct ophthalmoscope (HEINE BETA® 200; Herrsching, Germany). If the size of central lens opacity was more than 3 mm, the child was suspected of having cataract. The macula and optic nerve head were examined by direct ophthalmoscopy to exclude children with retinal lesions.

Statistical Analysis

To present data, we used mean values, standard deviation, median, range, frequency and percentage data. To evaluate the association of different factors with CVD, we used logistic regression analysis and odds ratio with 95% confidence intervals. The multilevel method was used to consider the intra-cluster correlation of observations. All statistical analyses were performed by SPSS software (version 21.0, IBM Co., Chicago, IL, USA). P values less than 0.05 were considered as statistically significant.

RESULTS

A total of 2150 children aged 7-12 (mean, 9.4 ± 1.7) years were studied. Details of baseline demographics and ocular findings are presented in Table 1. CVD was detected in 48 children indicating a prevalence of 2.2% (95% CI: 1.5% to 3%). The prevalence of CVD in male subjects was significantly higher as compared to female students [3.5 vs. 1.0%, P < 0.001, Table 1].

Table 1

Demographic and ocular characteristics of children with normal and deficient color vision

Mean and median pinhole VA were lower in children with CVD as compared to subjects with normal color vision [P = 0.035, Figure 1]. There was a significant difference between children with CVD and normal color vision subjects in terms of amblyopia; the prevalence of which was 8.3% (CI: 0.2-16.4%) in children with CVD versus 2.1% (CI: 1.5-2.08%) in those without CVD [P = 0.005, Tables 1 and 2]. However, there was no significant correlation between CVD and refractive errors, anisometropia, or strabismus [Table 1].

Figure 1

Mean visual acuity (LogMAR) in children with and without color vision deficiency. VA, visual acuity; logMAR, logarithm of the minimum angle of resolution

Table 2

Basic characteristics of 48 children with color vision deficiency

The prevalence of CVD among different VA categories is presented in Figure 2. The increased positive line slope indicates more reduction of VA; however this association should be considered with caution due to small sample size (3 children with CVD among 5 children with significant reduction of VA). Figure 3 presents a comprehensive view of the normal (no amblyopia and no CVD) examined children and those with amblyopia, CVD or combination thereof in the studied population.

Figure 2

The percent of color vision deficiency in different categories of visual acuity. VA, visual acuity

Figure 3

Comperhensive view of the examined population based on amblyopia and color vision deficiency. CVD, color vision deficiency.

Out of 48 children with CVD, 77.1% were male, 16.6% had hyperopia >+2.00 D, 25% had astigmatism >0.75 D, 6.3% had exotropia, 8.3% had amblyopia and 4 subjects showed coexisting CVD and amblyopia. The deficiency in color perception occurred in children with VA of 0.3 LogMAR or less. The majority of children were high hyperopic (spherical equivalent ≥+4.00 D) and astigmatic (>1.00 D), half of the cases were exotropic and one had anisometropia [Table 2]. Clinical data of the 49 amblyopic children are presented in Table 3. A comparable proportion of male (51.02%) and female (48.97%) children were amblyopic. There were more cases with high refractive errors among amblyopic children and most strabismic amblyopic children had esotropia (8 out of 13 cases, 61.54%). The total sum in some columns in Tables 2 and 3 are not equal to actual numbers of studied students, as photorefraction was not possible in certain children with high refractive errors due to the limited range of the device (−7.00 to +5.00 D), pupillary abnormalities, media opacities or strabismus.

Table 3

Basic characteristics of 49 amblyopic children

Table 4 compares basic characteristics among amblyopic children, those with CVD, and subjects with coexisting amblyopia and CVD, versus normal (no CVD and no amblyopia) children. Although mean VA among children with no CVD and no amblyopia was lower than color vision deficient children, the standard deviation was wider among children with CVD and there was no statistically significant difference. Mean age of amblyopic children were significantly lower than normal (no CVD and no amblyopia) children (P = 0.033).

Table 4

Comparison of all the basic characteristics among examined children

DISCUSSION

In the current study, in order to improve diagnostic accuracy, the Yang vision tester and PlusoptiX SO4 photorefractometer were applied. The former test can be calibrated for desired distances ranging from 30 cm to 9 m and testing conditions are adjusted automatically; thus, accurate color vision testing is possible at different distances. Its luminance can also be set from 80 to 320 candela (cd)/m2 and the commonly used value is 120 cd/m2 which is constant for all measurements. Therefore, we tested color perception in all children under standard conditions of our system with more stability as compared with the conventional Ishihara test.

The Ishihara test was also applied in our study as it is commonly used to screen for red-green CVD and can be learned easily and performed rapidly in children. It contains of ten characters differing in size (thinnest portions being under 0.5 cm), parallel to VA results.[26]

We performed photorefraction with no cycloplegia according to manufacturer recommendations, since induced peripheral aberrations due to a dilated pupil makes measurement of astigmatism more difficult and there is a possibility of off-axis refraction. In addition, photorefraction was performed at a longer working distance (1 m or more) as compared to other refractive error measurement tools such as a retinoscope or autorefractometer, therefore accommodation might be more relaxed.

In our study, 48 out of 2150 children had CVD [Table 1]. The male to female ratio was 3.5 with higher prevalence in male subjects which is consistent with other studies.[512] CVD is usually inherited by an X-linked recessive pattern[5] and studies on European Caucasians have revealed that CVD is approximately 20 times more prevalent among male as compared to female subjects.[12345] The difference in male to female ratio in our population, as compared to European studies, may be the lower age or a different genetic basis for CVD in our population.

We observed that all four patients with combined CVD and amblyopia had a VA of 20/40 or less [Table 2] which is comparable to the study by Bradley et al[27] reporting that all patients with combined CVD and amblyopia had VA of less than 20/50. We also found a statistically significant correlation between CVD and VA which means CVD was more prevalent in children with lower VA [Figures 1 and 2] as von Noorden[28] and other researchers have stated.[202729] These findings are consistent with studies showing that color perception is affected when VA decreases to less than 20/50.[2] Furthermore, it has been reported that color vision loss occurs more frequently when the VA is less than 20/400[20] or 20/200.[29]

McCulley et al[26] reported that the Ishihara test was the color test most dependent on good VA and it was not affected up to VA of 0.72 LogMAR (20/106). In our study, 6 of 8 eyes (75%) with VA worse than 0.7 LogMAR did not show CVD which is not consistent with this assumption. It seems that determination of VA as an accuracy criterion for the Ishihara test should be considered with caution [Table 2].

Although there was no difference between normal and color vision deficient children considering strabismus and refractive errors, amblyopia was significantly more common among children with CVD [P = 0.005, Table 1]. Qian et al[30] found a lower incidence of myopia among children with abnormal color perception as compared to normal (no CVD and no amblyopia) children while in the present study, there was no difference between normal color vision and CVD children in terms of myopia.

Ocular deviation especially exotropia was seen in 50% of children with coexisting amblyopia and CVD whereas 25% of them showed anisometropia [Table 2]. Bradley et al[27] also observed strabismus in 66% of children with coexisting amblyopia and CVD with no difference among certain deviation types but the majority had unsteady foveal or eccentric fixation. Koçak-Altintas et al[19] found no relationship between CVD and any types of amblyopia such as anisometropia or strabismus which is comparable to our results.

Bradley et al[27] reported that most amblyopic children with CVD were hyperopic (sphere ≥+2.50 D). In the present study, 4 children had both amblyopia and CVD, and 5 out of 6 amblyopic eyes had hyperopia more than + 4.00 D which is comparable to Bradley's findings [Table 2].

One of the limitations of the current study is that examined children were screened only for red-green CVD since the Ishihara test is not able to detect blue-yellow deficiency. Children who were diagnosed with CVD were not retested using another color vision test with more sensitivity to validate our observations. Moreover, lack of cycloplegic refraction and determination of best corrected VA among all children was another limitation of our study.

In summary, although CVD was correlated with amblyopia and decreased vision, there was no relationship between CVD and the type of amblyopia, refractive error, anisometropia and ocular deviation. Exotropia and hyperopia were the most common form of ocular deviation and refractive error among children with coexisting CVD and amblyopia.

Financial Support and Sponsorship

Nil.

Conflicts of Interest

There are no conflicts of interest.