CHOROIDAL THICKNESS IN HEALTHY CHINESE CHILDREN AGED 6 to 12

The Shanghai Children Eye Study

Abstract

Purpose: 

To explore the characteristics of choroidal thickness (ChT) in Chinese children.

Methods: 

A total of 144 healthy children, aged 6 years to 12 years old, were enrolled in the study. The ChT of subfovea and peripheral locations 0.5, 1.5, and 2.5 mm away from the fovea were evaluated by enhanced depth imaging optical coherence tomography. The association between subfoveal ChT and systemic, as well as ocular factors, including age, sex, height, weight, body mass index, axial length, refractive error, intraocular pressure, preterm history, and the refractive status of parents were studied.

Results: 

The mean subfoveal ChT was 302 ± 63 μm. In the nasal, superior, and inferior areas, the ChT of locations closer to the fovea was thicker than those farther away from the fovea (all P < 0.05); however, ChT was not significantly different among different locations in the temporal area (P = 0.16). The ChT of the nasal quadrant was significantly thinner than that of other areas (P < 0.01). Subfoveal ChT decreased with age, axial length, preterm history, and increased with height. Sex was not statistically associated with subfoveal ChT.

Conclusion: 

In Chinese children, the ChT is thinnest in the nasal quadrant and thicker in central regions than in peripheral areas. The subfoveal ChT independently decreases with age, axial length, preterm history, and increases with height.

In Brief

The choroidal thickness of Chinese children decreases with age, axial length, preterm history, and increases with height. The choroid is thinnest in the nasal quadrant, and thicker in central regions.

The choroid has many vital functions, such as supplying the outer retina with oxygen and nutrients, regulating ocular temperature, regulating intraocular pressure, and absorbing light.^1,2 Dysfunction of the choroid can cause vision-threatening diseases, such as central serous chorioretinopathy and choroidal neovascularization.³ Despite the physiological and pathological importance of the choroid, research into choroidal characteristics had been limited because of the difficulty in visualization and measurement. However, the usage of enhanced depth imaging technology in spectral domain optical coherence tomography enabled accurate visualization of the choroid, and led to a growing interest in this research area.⁴

Decreases in choroidal thickness (ChT) are consistently observed in adults. Risk factors for reductions in ChT include increasing age, myopic refractive error, and increasing axial length.^5–9 Thus far, there has been a limited number of studies researching choroidal characteristics and development in children and the conclusions of these studies are conflicting. For instance, several studies indicated a decrease in subfoveal ChT with age,^10,11 whereas some other studies found that ChT increases with age.^12–15 In addition, results regarding the relationship between ChT and sex in children are not consistent among published studies. A study conducted in Italy indicated that girls have a larger choroidal volume than boys¹³; however, some other studies have reported that sex is not related to ChT in children.^15,16 Moreover, although some researchers have concluded that ChT is related to axial length and myopia,^11,14 other studies have not found this relationship.¹⁰

There has been one study examining the ChT in Chinese children; however, it focused on unilateral amblyopic participants.¹⁷ Large differences in choroidal characteristics can exist in different populations, and given the current inconsistent results regarding ChT in children, it is necessary to study ChT in normal Chinese children. This study explored the characteristics of ChT, and its relationship to systemic and ocular factors in normal Chinese children, aged 6 years to 12 years old.

Methods

Setting and Participants

This research was a part of the Shanghai Children Eye Study, which is a school-based survey of eye health in a large sample of children in Shanghai, China. The aim of Shanghai Children Eye Study is to provide scientific evidence for the prevention and control of eye disease in children and adolescents. The Shanghai Children Eye Study study group has previously published several papers.^18–20 This study was conducted in school in June 2014. A total of 154 healthy children aged 6 years to 12 years old from one randomly selected primary school participated in this study. The inclusion criteria were as follows: 1) best-corrected visual acuity more than 20/25; 2) no previous or current severe eye disease, such as congenital glaucoma, cataract, or retinopathy; 3) voluntary acceptance of ocular examinations. The research team consisted of one ophthalmologist, five optometrists, and two public health physicians. Preliminary research was conducted before the study to test the practicality of the protocol and the repeatability of the measurements.

The study was conducted according to the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of Shanghai General Hospital, Shanghai Jiao Tong University. All the students understood the study protocol. Parents provided written informed consent before enrollment.

Research Methods

Personal information from each student, including age, sex, height, weight, birth, and family disease history, was recorded. Height and weight were measured with each student in light clothing and with shoes off. The refractive status of parents and the birth condition of the children were obtained from questionnaires. A series of ophthalmic examinations, including visual acuity, noncycloplegic refraction examination, sensorimotor examination, slit-lamp biomicroscopy, fundus examination, and axial length measurement were performed in both eyes of each student. Visual acuity was measured at 5 m using a standard logarithmic tumbling E chart (wh01; Wehen, Guangzhou, China). Refraction examination was performed with a desktop auto-refractor (version KR-8800; Topcon Corporation, Tokyo, Japan) with a measurement range of −20 D ∼ +20 D. Spherical equivalent refraction was used to classify refractive status. Axial length was measured using IOL Master (version 5.02; Carl Zeiss Meditec, Oberkochen, Germany).

Choroidal thickness measurements were conducted from 9 AM to 2 PM. According to previous studies,^21,22 although diurnal variations exist in ChT, there is no significant variation in ChT during this period. Horizontal and vertical crosshair scans were conducted using spectral domain optical coherence tomography (RTVue 100; Optovue Inc., Fremont, CA) with the enhanced depth imaging system and 6-mm line scan. The wavelength of the optical coherence tomography (OCT) instrument was 840 ± 10 nm and the scanning speed was 26,000 A-scan/second. A total of 32 B-Scans were averaged for each measurement and eye tracking was used during the examination. The axial resolution was 5 μm. Signal Strength Index was used to monitor and control the quality of the image. Images with a Signal Strength Index over 40 were considered acceptable. Choroidal thickness was determined as the distance from the outer border of the retinal pigment epithelium (RPE) and the inner border of the chorioscleral interface, and was measured manually by two OCT technicians independently using the software provided by the OCT system. Figure 1 provides examples of horizontal and vertical OCT images, and an illustration of the different locations where ChT was measured for each participant. A total of 14 locations were measured in each participant, including the fovea (horizontal and vertical), 0.5 mm inferior to fovea (I0.5), 1.5 mm inferior to fovea (I1.5), 2.5 mm inferior to fovea (I2.5), 0.5 mm superior to fovea (S0.5), 1.5 mm superior to fovea (S1.5), 2.5 mm superior to fovea (S2.5), 0.5 mm nasal to fovea (N0.5), 1.5 mm nasal to fovea (N1.5), 2.5 mm nasal to fovea (N2.5), 0.5 mm temporal to fovea (T0.5), 1.5 mm temporal to fovea (T1.5), and 2.5 mm temporal to fovea (T2.5). The subfoveal ChT was calculated as the mean of the horizontal and vertical measured subfoveal ChT.

Fig. 1: Representative horizontal (H) and vertical (V) OCT images showing the method of ChT measurement. Choroidal thickness (double-headed arrows) was defined as the distance from the outer border of the retinal pigment epithelium (RPE) and inner border of the chorioscleral interface. Choroidal thickness was measured at the fovea, 0.5, 1, and 2.5 mm nasal, temporal, superior, and inferior to the fovea.

The mean of the two values taken by the OCT technicians was recorded as final data and used for statistical analyses. To ensure the reproducibility of the ChT measurements by the two observers, both technicians read randomly selected images from 20 eyes in advance. The inner-observer correlation of average measures for the subfoveal ChT was 0.98, and the correlation of the peripheral locations ranged from 0.96 ∼ 0.99. A Bland–Altman analysis was used to calculate interobserver agreement. The mean difference of subfoveal ChT measurements in 144 right eyes between the 2 observers was 1 μm, and the 95% limits of agreement was −15 ∼ 17 μm (Figure 2). The interobserver correlation was 99.24% (P < 0.001).

Fig. 2: The Bland–Altman plot represents the agreement between the choroid thickness measurements obtained by the two readers.

Statistical Analyses

A database was established using Epidata3.1 and two staff members independently input the data. The three-dimensional scatter plot of the subfoveal ChT, axial length, and age was performed with Matlab (version2011b; MathWorks). SPSS (version 20.0; IBM) was used for all the other statistical analyses.

Because the correlation between the 2 eyes was tight (the correlation index of the ChT under subfovea between 2 eyes was 0.79 and ranged from 0.58 to 0.82 for peripheral locations), only the right eye data were used for statistical analysis. The characteristics are presented as the means ± SD for continuous variables and as rates (proportions) for the categorical data. Statistical significance was defined as P < 0.05 (2 tailed).

The data distribution was examined using Kolmogorov–Smirnov test. Normally distributed continuous variables were analyzed by t-Test. Mann–Whitney U tests were used for nonnormally distributed continuous variables. The categorical variables were compared using the Chi-square test. All the ChT values were normally distributed. The differences in systematic/ocular factors between different sexes were tested using an independent Samples t-Test. Comparisons of subfoveal ChT and ChT at different measurement locations were performed with the paired t-test. The comparisons among different peripheral locations were analyzed by repeated-measures analysis of variance and contrast analysis. Simple correlation analysis was performed to test the relationship between subfoveal ChT and systematic and ocular factors, including age, sex, height, weight, body mass index, axial length, refractive error, intraocular pressure, preterm history, and parents refractive status. Stepwise multiple regression analysis was performed to determine independent factors associated with subfoveal ChT.

Result

Among the 154 students enrolled in the study, 10 were excluded because of unclear OCT images (Signal Strength Index < 40). Therefore, a final total of 144 children, 70 boys (48.61%) and 74 girls (51.39%), were included in the study. There was no significant difference in sex or age between included and excluded students (all P > 0.05).

The General Characteristics of Participants

The age of participants ranged from 6 years to 12 years old, with a mean age of 9.25 ± 1.62 years. The spherical equivalent refraction of participants ranged from −6.63 to 4.75 diopters (D), with a mean of −1.04 ± 1.80 D. All children had 20/20 best-corrected visual acuity. The shortest axial length was 20.70 mm (in a 9-year-old girl) and the longest was 26.73 mm (in an 11-year-old boy). The mean axial length was 23.59 ± 1.09 mm. The characteristics of the 144 participants are listed in Table 1. Boys had significantly longer axial length (P = 0.041) and lower intraocular pressure (P = 0.004) than girls. No other differences between sexes were observed.

Table 1: General Characteristics Characteristic of the 144 Study Participants

The Choroidal Thickness and Topographic Variation

The subfoveal ChT of all participants ranged from 152 μm to 447 μm, with a mean value of 302 ± 63 μm. The mean value and the SD of the ChT of peripheral locations are shown in Table 2. The ChTs of peripheral locations T0.5 (P = 0.02) and T1.5 (P < 0.01) were thicker, however, the ChT of S0.5, I0.5, and T2.5 were not significantly different (all P > 0.1) compared with the subfoveal ChT. The ChTs of all the other locations were significantly thinner than the subfoveal ChT (all P < 0.05). The comparisons among different peripheral locations are shown in Table 2. Generally speaking, the nasal quadrant ChT was significantly thinner than the other areas (all P < 0.01) and the temporal quadrant ChT was statistically thicker than the other areas (P < 0.05). There was no difference between the superior and the inferior quadrant ChTs (P > 0.05). In the nasal, superior, and inferior quadrants, the ChT of locations closer to the fovea was thicker compared with locations further away from the fovea (all P < 0.05), but there was no difference in ChT among different locations in the temporal quadrant (P = 0.16).

Table 2: Comparison Among Choroidal Thickness of Different Peripheral Spots Mean (SD)

The Relationship Between Subfoveal Choroid Thickness and Other Factors

The correlation between subfoveal ChT and systematic and ocular factors is listed in Table 3. Age, axial length, refractive error, weight, and height were related to subfoveal ChT, whereas sex was not statistically associated with subfoveal ChT. Stepwise multiple regression analysis suggested that age, axial length, height, and preterm history are independent factors associated with subfoveal ChT. Choroidal thickness decreases with increasing age, axial length, preterm history, and increases with height. These 4 factors could explain 34.2% of the ChT value (R² = 0.342), within which age and axial length accounts for 81.58% of the influence (R² = 0.279). Choroidal thickness decreased by 23 μm for every 1 mm increase in axial length and subfoveal ChT decreased by 17 μm for each year increase age (Table 4). Figure 3 shows the effect of age and axial length on subfoveal ChT by a three-dimensional scatter plot. Although there is a tight relation between age and axial length (r = 0.42, P < 0.01), no statistical significant interaction (P = 0.36) or collinearity (variance inflation factor = 1.23) was found between age and axial length on the variations in ChT.

Table 3: Correlate Coefficients of Subfoveal Choroidal Thickness With Related Factors

Table 4: Results From the Stepwise Multiple Regression Analysis Evaluating Systematic/Ocular Variables Associated With Subfoveal ChT (Right Eyes)

Fig. 3: Three-dimensional scatter plot of age (x-axis), subfoveal ChT (y-axis), and axial length (z-axis).

Discussion

The present cross-sectional study explored choroidal characteristics and development in children during normal eye growth in the Chinese population. Until now, there have been limited reports and many unsettled arguments regarding the choroidal profile in children. The present study found that the ChT is thinnest in the nasal quadrant and thicker in central regions than in peripheral areas. In addition, we found that subfoveal ChT independently decreases with age, axial length, preterm history, and increases with height.

In this study, the mean value of the subfoveal ChT in Chinese children aged 6 years to 12 years old was 302 ± 63 μm, which is higher than that reported in Japanese children aged 3 years to 15 years old (260.4 ± 57.2 μm),¹¹ and lower than children, aged 10 years to 15 years old, from white populations (330 ∼ 369 ± 65 ∼ 81 μm).^12,14–16 Apart from the population differences, the high prevalence of myopia and associated long axial length in Asian children most likely results in thinner choroids.²³ The choroid of Chinese children was significantly thicker than that reported in Chinese adults (254 ∼ 261 ± 88 ∼ 107 μm).^9,24 Although a Spanish study reported no significant difference in ChT between 50 adults and 43 children,²⁵ our finding was consistent with a Japanese study, in which 100 children and 83 adults were examined.¹¹ This result is also consistent with the finding that, in adults, subfoveal ChT decreases with age.²⁶

The topographic distribution of choroid observed in our study indicated that ChT is thickest in the temporal area and thinnest in the nasal area, which is consistent with previous reports of children and adults of other populations.^{8,11,12,14,15,25} The entrance of the optic nerve in the nasal quadrant might result in a thicker retina and thinner choroid. Some previous studies of adults and children have indicated that the ChT of the inferior region is thinner than that of the superior region,^7,12,14 however, no significant difference was found between the ChT of the superior and inferior quadrants in this study. Previous research has indicated that the choroid is thicker in more central regions.^12,14 Our results showed that the ChT of locations closer to the fovea is thicker compared with locations farther away from the fovea in the nasal, superior, and inferior areas, but not in the temporal area. This might be related to the finding that choroidal vessels fill up more rapidly in the macular region.²⁷ The data showed that the ChT was not thickest in the subfoveal, but rather at the spot 0.5 mm temporally from the fovea, which is consistent with a Korean study.¹⁰

The strong relationship between axial length and ChT has been found in several previous pediatric and adult studies.^5,11–15,17 It has also been reported that subfoveal ChT is significantly thinner in myopes compared with nonmyopes.¹⁴ Recently, Li et al observed 1,323 children aged 11 years to 12 years old with no significant refractive error, and found a negative relation between ChT and axial length.¹⁶ However, Park and Oh¹⁰ found no relationship between ChT and axial length or refractive status. Our results indicate that axial length is inversely related to ChT and that every 1 mm increase in axial length results in a 23-μm decrease in subfoveal ChT. This is similar to an Australian study involving 194 children aged 4 years to 12 years, which found that every 1 mm increase in axial length resulted in a 26.1-μm decrease in subfoveal ChT.¹² Although both refraction and axial length were significantly associated with ChT in the univariate analyses, the lack of association between ChT and refractive error in the multivariate analysis was most likely due to the multicollinearity between ChT, refractive error and axial length.

In this study, there was a 17-μm decrease in subfoveal ChT with each year of increasing age, suggesting that age is another important factor that is inversely related to ChT. This finding is consistent with several previous studies in Asian children. For example, a recent Korean study¹⁰ of 48 children aged 4 years to 10 years old and a Japanese study of 100 children 3 years to 15 years old also found decreases in ChT with age.¹¹ In addition to Asian population, a Spanish study including 83 children aged 3 years to 15 years old, without refractive error, found that ChT in subfoveal and temporal areas were negatively correlated with age.²⁵ On the contrary, some studies in white populations drew opposite conclusions. Read et al¹² investigated 194 children aged 4 years to 12 years, Mapelli et al¹³ studied 52 children aged 2 years to 17 years, and Bidau-Garnier et al¹⁵ observed 174 children aged 3.5 years to 14 years, and all concluded that ChT increases with age. According to an 18-month longitudinal study,²⁸ ChT increases with age in most children from 10 years to 15 years old, however, in those children who are undergoing faster axial eye growth, a thinning of the ChT is observed.

Bidau-Garnier et al¹⁵ suggested that choroid growth in childhood is followed by a progressive decrease in thickness in adulthood, whereas another study indicated that puberty promotes choroidal thickening.¹⁶ Combining their conclusions and results of the current, we suspect that there are two stages of choroid growth before adulthood, during early childhood and again during puberty, and that the growth of the choroid slows down or stops between these two stages. Similar bimodal growth patterns can be seen in height and weight in childhood and adolescence. During the plateau period between the two stages of growth, the choroid begins to attenuate with the elongation of the axial length, and thinning occurs, which is more significant in children who develop myopia. This explains the inverse relationship between ChT and age, which is often observed in studies with myopic participants,^10,11 whereas studies in which participants have no refractive error often find that ChT increases with age.¹² Thus, the inverse relationship between age and ChT reported in Asian children could be related to the higher prevalence of myopia and later start of puberty observed in Asian compared with white children.^23,29 However, to verify this hypothesis, future large sample longitudinal studies are necessary.

Apart from age and axial length, the results of this study indicated two other independent factors associated with subfoveal ChT: height and preterm birth. Choroidal thickening with increasing height has been reported previously,^15,16 and suggests that the choroid is growing along with the body. Because height is also a good representative of puberty status, the correlation between height and ChT could be related to the second growth stage of the choroid during puberty. Our finding that ChT is thin in children with preterm history is consistent with the Copenhagen Child Cohort 2000 Eye Study,³⁰ which indicated that thinner choroids were associated with lower birth weight, lower birth length, and being small for the gestational age in children aged 11 years to 12 years old. Another study reported that, in preterm infants, ChT increased with age but was thinner than in full-term infants, suggesting delayed development due to retinopathy of prematurity.³¹ Children with a preterm history, with or without retinopathy, might have deficient vascularity in the choroid, which presents as ChT thinning.

In this study, although boys were found to have longer axial length than girls, sex was not associated with ChT in Chinese children after adjusting for age and/or refractive status. This is consistent with the results of studies performed by Bidaut-Garnier et al¹⁵ and Park and Oh.¹⁰ However, Mapelli et al¹³ found that the volume of the choroid was higher in girls than in boys. Differences in measurement methods used in each study might be one reason for this inconsistency.

There were some limitations in this study. First, the sample was relatively small. To confirm the results and conclusions, a multicentered, random sampling study, using larger samples should be performed in the future. Second, because this was a cross-sectional study, it is unclear whether change in ChT started before or after changes in age and axial length, thus further longitudinal investigations should be conducted. Third, the noncycloplegic refraction examination used in our study is not an accurate way to measure refractive error, especially in those with hyperopia.³² This could be why no correlation between refractive error and subfoveal ChT was found in the multivariate analysis. Fourth, only horizontal and vertical scan lines were collected in our study; applying a denser scanning protocol could provide a clearer understanding of the regional thickness variations of the choroid and should be used in the future studies. Last but not the least, the influence of ocular magnification on the transverse scale should be considered to obtain accurate peripheral choroidal measurements. In this study, the differences of peripheral ChT with and without magnification amendment were measured in 12 OCT images with extreme axial length, including 4 with axial length longer than 26 mm, and 8 with axial length less than 22 mm. In one image with axial length of 20.94 mm, the ChT of the location 2.5 mm away from the fovea was 396 μm before the magnification, and 375 μm after it. In another image with axial length of 26.09 mm, the ChT of location 2.5 mm away from the fovea was 304 μm before the magnification, and 285 μm after it. For the rest of the images, the differences of the ChT measurements before and after the magnification were all less than 10 μm.

Former epidemiological research and animal experiments have indicated that the choroid has important roles in the modulation of the refractive state and may be involved in the development of refractive error.^{14,26,33–37} The prevalence of myopia in Chinese children, aged 13 years to 15 years old has reached 80%.²³ This high prevalence of myopia makes Chinese children an ideal population in which to study the relationship between choroid and refractive error. Thus, the next step of the Shanghai Children Eye Study will be to conduct a prospective study to explore the changes of choroid in the development of refractive error.

In conclusion, ChT is thicker in more central regions except in the temporal area and the nasal quadrant ChT is thinnest among areas. There is no sex difference in ChT in Chinese children. Choroidal thickness independently decreases with age, axial length, preterm history, and increases with height.

Acknowledgments

The authors would like to express our sincerely gratitude to Dr. Brien Holden and Dr. Ravi C. Bakaraju from Brien Holden Vision institution for their inspiration and guidance.