Assessment of macular choroidal and retinal thickness: a cohort study in Tibetan healthy children

This research investigates the distribution, progressive changes, and contributing factors of macular choroidal and retinal thickness in Tibetan children utilizing swept-source optical coherence tomography (SS-OCT). The Lhasa childhood study recruited 1632 students from seven primary schools in Lhasa. These participants underwent OCT and ophthalmological evaluations, encompassing retinal and choroidal thickness measurements, refractive error, axial length (AL), and systemic examinations. The median age of the scholars was 8.57 ± 0.50 years with a median spherical equivalent (SE) of 0.19 ± 1.28D. Multivariate regression analysis revealed that thinner macular choroid thickness was correlated with lower value of SE, worse best-corrected visual acuity, higher mean arterial blood pressure (MABP) and boys, while retinal thickness was associated with better image quality and lower value of SE. The choroid and retina were significantly thinner in myopic children. SE was positively related to the thickness of all choroidal and full retinal subregions. In comparison to baseline data from 20 months prior, most regions of the full retina had significantly thinned. Choroidal thickness of Tibetan children is thinner than that of same-age children from other regions. Thinning of retina, the outer-sector GCC and GCIPL may be specified as a follow-up and prognostic indicator for myopia.

study was conceived to assess the macular choroidal and retinal thickness in Tibetan children utilizing SS-OCT.We further investigated the impact of various associated factors on retinal and choroidal thickness through the detailed systematic and ophthalmic examination.These discoveries assist us in validating the developmental patterns of the retina and choroid in children.

General characteristics of the study
Owing to loss of follow-up, a total of 1632 children were incorporated in the OCT assessment, of whom 37 failed to complete the full examination; thus, a 90% response rate was achieved.After excluding images with low quality or significant abnormalities, 1555 children met the inclusion criteria.No significant gender or ethnic differences between the included and excluded samples were observed.The average age of the students involved in the assessment was 8.57 ± 0.50 years; 47.8% were female, and 95.1% were Tibetan.Table 1 showcases the systemic and ocular characteristics of all participating students, with an SE of 0.19 ± 1.28D.The percentages of children with myopia, emmetropia, and hyperopia were 22.3%, 27.8%, and 49.9%, respectively.

Distribution of macular full retina, GCC, GCIPL, RNFL, and choroid thickness
Figure 1 illustrates the distribution of thickness.The mean thickness of the macular choroid, full retina, GCC, GCIPL, and RNFL was 233.23 ± 44.40 μm, 272.72 ± 24.32 μm, 107.91 ± 11.37 μm, 70.82 ± 7.02 μm, and 37.21 ± 6.64 μm, respectively.In both the inner and outer annulus of the choroid, the nasal region was the thinnest, succeeded by the superior, while the temporal and inferior regions were comparatively thicker.Conversely, the distribution of all retinal layers was significantly thinner in the temporal and inferior regions but thickest in the nasal region.
Excluding the inferior outer annulus region, the ChT of all other regions was significantly thicker in girls than in boys (P < 0.05).In terms of mean thickness, full retina and RNFL were not significantly different between genders, but girls still had marginally thicker GCC and GCIPL layers.The detailed distribution of each sub-region and the comparison between genders are depicted in Supplementary Table 1.Analysis of covariance revealed no significant difference in choroidal thickness (F = 0.06, P = 0.80) between the genders after eliminating the confounding effect of AL and IOP, but still a difference in retinal thickness (F = 9.54, P = 0.002).
In the refractive sub-groups analysis of emmetropia, hyperopia, and myopia, the ChT was significantly different between groups (P < 0.001).The ChT and retinal thickness in the myopia group were significantly lower than in the hyperopic and emmetropic groups.Regarding full retinal thickness, only the foveal and inner-inferior regions were not significantly different between refractive groups.For the thickness of GCIPL and GCC, the outer-annulus region was significantly different between groups (P < 0.05) but not the inner-annulus.There were inter-group differences in RNFL in the nasal region but not in other regions or averaged thickness, as displayed in Supplementary Table 1.Under the condition of eliminating the effects of AL and IOP, there were still significant differences in choroidal (F = 34.49,P < 0.001) and retinal thickness (F = 6.36,P = 0.002) between different refractive groups.

Factors influencing macular thickness
Univariate linear regression was used to analyze the effects of systemic and ocular factors on the thickness of each layer of the choroid and retina, and those with P > 0.1 were included in the multivariate regression analysis.Considering the problem of covariance among variables, we included SE (excluding AL), BCVA (excluding stereoacuity), height (excluding weight) in the multivariate regression model.Multivariate linear regression analysis revealed that macular choroid thickness was associated with SE (P < 0.001, r = 0.27), BCVA (P < 0.001, r = 0.08), MABP (P = 0.008, r = 0.05) and gender (P = 0.011, r = 0.06).Conversely, macular retinal and stratified thickness were primarily correlated with image quality and SE (P < 0.01 for all).Thickness of GCIPL, GCC and RNFL layers were all correlated with BCVA, while GCC was also positively correlated with height (P < 0.05 for all) (Table 2).The correlation between image quality and retinal thickness may be due to the clarity of the image, which affects the accuracy of the segmentation the measured thickness.Upon adjusting for age, sex, height, and weight, Pearson's correlation analysis portrayed that the value of SE was positively correlated with the thickness of most retinal regions, excluding the central and inferior regions of the inner annulus.Moreover, all regions of choroidal thickness were significantly correlated with SE (P < 0.001, r = 0.2-0.3),with the most robust correlation in the inner and outer sector of nasal region (r = 0.264, 0.257, respectively), succeeded by the inferior, superior, and temporal regions.For the thickness of GCIPL and GCC, only the outer annulus was more correlated with SE (P < 0.01), but the inner annulus had no apparent correlation.RNFL thicknesses in the average, inner-inferior, and nasal regions were negatively related to SE (P < 0.05).(Table 3).

Progression and change in macular thickness after 20 months
The alterations in the retinal thickness of Lhasa children after nearly 2 years (20 months), compared to the baseline examination, are illustrated in Table 4. Excluding the inner-nasal, outer-inferior, and central regions, thickness of the full retina decreased significantly (P < 0.05).The average thickness of GCIPL also decreased, but GCC exhibited no significant change.Changes in retinal thickness between the two years were compared with refractive groups at baseline (Table 5).The myopic group demonstrated larger changes than the hyperopic group in regions of a thinning retina and fewer changes in regions of a thickening retina, but without a statistically significant difference.Correlation analysis depicted that the change in SE was positively correlated solely with the change in GCIPL thickness (r = 0.10, P = 0.002) but not with the thickness of other areas of the retina.

Discussion
To the best of our knowledge, this is the sole cohort study to broadly discuss the thickness of the choroid and various retina layers in the detailed macular sub-regions of children in Lhasa, Tibet.In the baseline examination of LCES, we meticulously described the macular retinal and circumpapillary RNFL thickness 8 .In the second year of follow-up, we measured and compared the ChT and retinal thickness of Lhasa children.The observations include the following: (1) The thickness of the choroid and retina was significantly reduced in myopic children.
The myopic group demonstrated more pronounced changes than the hyperopic children in regions of retinal  In previous studies, Nagasawa et al. 24 reported that the central ChT of eight-year-old children in Japan with the SE of − 0.04 ± 0.96D was approximately 260.4 ± 57.2 μm.Also, Peiyao Jin et al. discovered that the average ChT of students in Shanghai was 251 ± 62 mm (SE = 0.20D, AL = 23.18mm); also in studies of children in Australia (ChT = 350 μm) 25 , Queensland (ChT = 337 ± 65 μm) 26 , Shandong in China (ChT = 283 ± 67 μm) 27 , ChT were thicker than that in the current study.It is generally accepted that the choroid progressively becomes thinner as myopia advances 2,28 .Despite higher values of SE and shorter AL, the ChT in this study remains thinner than those observed in Shandong 27 and Inner Mongolia 29 (SE = − 1.41D, − 1.20D respectively) in China.We hypothesize that this difference may be attributed to ethnic variations.This research revealed that, after adjusting for AL, the choroidal thickness of the myopic group was significantly less than that of the hyperopic and emmetropic groups, with the difference being most noticeable in the central and nasal regions of both the inner and outer annulus.Jin et al. 4 reported similar findings, but they proposed that choroidal thinning was more evident in myopic eyes in the superior and inferior perifoveal regions.We also found a significant positive correlation between choroidal thickness and the value of SE, with the most significant correlation in the nasal region.The choroidal thinning of myopic group was more pronounced in the nasal area where the choroid was originally thinner in distribution, indicating that the nasal area may be a more sensitive region for predicting the onset of early myopia.It appears that choroidal thinning occurs early in myopic progression, possibly before axial elongation, implying that ChT thinning may act as an earlier biomarker of myopia progression 30,26 .www.nature.com/scientificreports/While we did not observe significant changes in the choroidal perifoveal area in myopic eyes, we did identify such results in the thinning of the retina, particularly the GCC and GCIPL layers.Previous studies have produced inconsistent evidence on the relationship between retinal thickness and myopia.Some studies 31,32 have reported that the retinal thickness of myopic patients is significantly less.Some found that the retinal thickness is centrally thickened and peripherally thinned 28,33 .In this study, the macular retinal thickness of myopic patients was less than that of emmetropic and hyperopic patients in all sub-regions except the central region.Moreover, the average retinal thickness also demonstrated a significant positive correlation with SE, which is consistent with the conclusions of our baseline study 8 .The mean and outer-sector thickness of GCC and GCIPL in this study were thinner in myopic students, and both were positively correlated with SE.Jin et al. 34 and Yoo et al. 35 also found that outer-sector GCC thickness was positively correlated with SE, whereas inner-sector was not.This implies that the thickness of the outer-sector GCC and GCIPL is more sensitive to SE changes and may be considered potential risk factors for myopia progression, requiring long-term monitoring and follow-up.The difference in thickness of the RNFL layer was insignificant between refractive groups, but SE was negatively correlated with the nasal and mean thickness of RNFL in this study.
Serving as the source of retinal blood supply and oxygen provision, choroidal thickness is also significantly correlated with retinal thickness.This study discovered that the thickness of each choroidal region was significantly correlated with retinal thickness, excluding the central region.The correlation is slightly higher at the outer annulus.Jin et al. 4 reported that retinal thinning is observed later during the progression of myopia.Considering the systemic factors influencing the choroid, changes in retinal thickness may be estimated more reliably.This suggests that changes in retinal thickness may also serve as a monitoring factor for the onset or progression of myopia 2 .However, changes in retinal thickness may occur later than changes in choroidal blood flow and thickness; thus, it may be designated as a follow-up and prognostic indicator for myopia, despite its low sensitivity.
A clear correlation between choroidal thickness and oxygen saturation was not found in this study, but a positive correlation was found with MABP.The choroid, as a vascular layer in the eye, is susceptible to blood flow-related factors such as blood pressure 36,37 .Zhu et al. 29 also reported a negative correlation between subfoveal choroidal thickness and SBP and DBP in children.Thus the association between choroidal thickness and blood pressure may require further follow-up and demonstration.ChT was not significantly associated with systemic factors such as age, height, and weight in our study.While retinal thickness of different layers was positively correlated with weight.Additionally, we discovered that stereoacuity was negatively associated with GCIPL and GCC thickness.Stereoacuity gradually develops to the adult level in childhood, which was correlated to binocular vision and amblyopia 38,39 .The strong correlation between BCVA and stereoacuity was also found in the present study.Therefore, in the multivariate analysis we included only BCVA and excluded stereoacuity.Thinning of GCIPL and GCC is generally associated with glaucoma progression [40][41][42] , but its relevence to amblyopia is inconclusive [43][44][45] .Thinning of GCC and GCIPL may induce stereopsis disorder due to affecting visual field or potential visual function, but this deduction requires further research for confirmation.
Comparing the changes in retinal thickness between the baseline and the second year of follow-up, it was found that the retinal thickness of Lhasa children was significantly reduced, except for the central area, which had thickened.GCC and GCIPL were also significantly thinner.Correlation analysis demonstrated that the change in GCIPL was significantly positively correlated with the change in SE, but no clear correlation was found between full retinal and GCC thickness.Similar to the above conclusions, Jin 34 reported that the 1-year change in SE was correlated only with outer retinal layer (ORL) change but not with full retina and GCC.Although retinal thickness reduces with decreasing SE value, this change does not exhibit a clear dose-response relationship.
Despite the undeniable advantages of this study-for instance, a large-scale sample, minority subjects, standardized examination, integrated OCT data, and intensive analysis-it also has some limitations.Due to the lack of local medical equipment, we used a different machine to measure AL at baseline.Equipment changes may have affected our data comparisons at follow-up.Therefore, statistical analysis of baseline AL data was not included in this study.Due to the small sample proportion of children of Han and other ethnic groups, differences in the thickness and development of the retina and choroid between ethnic groups are not statistically comparable.Children's collaboration with OCT testing has not been greatly enhanced due to language barrier.These problems may be resolved by increasing the sample in further follow-up.
In conclusion, this study illustrates that SS-OCT can effectively obtain extensive ophthalmic biometry data on Tibetan children.The choroidal thickness of Tibetan children is thinner than that of same-age children from other regions.A thicker choroid was associated with higher values of SE, best-corrected vision acuity, and mean arterial blood pressure in multivariate linear regression.Choroid thinning may serve as a preliminary biomarker of myopia progression.Thinning of the retina, the outer-sector GCC and GCIPL may be designated as a followup and prognostic indicator for myopia, despite their lower sensitivity.The findings may aid in establishing a standardized Tibetan children's ophthalmology database, particularly to illustrate the relationship between the retinal thickness of different layers and refractive status.Our current study further suggests that routine OCT screening may be effective for the long-term monitoring of ocular diseases in the plateau of China.

Research subjects and design
The Lhasa Childhood Eye Study (LCES) is an epidemiological cohort study that included grade-1 students from seven elementary schools in Lhasa, Tibet, over a 5-year follow-up period until they entered junior high school.At baseline in 2019, 1856 Tibetan students, with an average age of 6.83 years, were included in the cohort and examined yearly.The research protocol complied with the Declaration of Helsinki and was approved by the Ethics Committee of Beijing Tongren Hospital, Capital Medical University (No. TRECKY2019-058).An exhaustive ophthalmic examination was performed according to the standard operating procedures and supervised (C) Full retinal thickness (distance between the red and green boundaries); ganglion cell + inner plexiform layers (distance between the yellow and blue boundaries); ganglion cell complex (distance between the red and blue boundaries, vitreal to inner nuclear layer); retinal nerve fiber layer (distance between the red and yellow boundaries), choroid (distance between the green and purple boundaries).

Figure 1 .
Figure 1.Thickness distribution maps of macular layers.Topographic distributions of macular choroid, full retina, ganglion cell-inner plexiform layer, ganglion cell complex and retinal nerve fiber layer in different sectors are shown through the picture.The mean ± SD values (lm) are presented.

( 5 )
Compared to baseline data from 20 months earlier, most regions of the full retina had significantly decreased.The change in SE was positively associated solely with the change in GCIPL thickness.

Figure 2 .
Figure 2. The Radial 9 × 9 mm scan region (A) and macula ETDRS grid, (B) overlaid with example projected images.(C)Full retinal thickness (distance between the red and green boundaries); ganglion cell + inner plexiform layers (distance between the yellow and blue boundaries); ganglion cell complex (distance between the red and blue boundaries, vitreal to inner nuclear layer); retinal nerve fiber layer (distance between the red and yellow boundaries), choroid (distance between the green and purple boundaries).

Table 1 .
Systemic and ocular parameters in Tibetan children between sexes.All data were listed as mean ± standard deviation for continuous variables and as median (quartile) for category variables.BMI body mass index, SBP systolic blood pressure, DBP diastolic blood pressure, MABP mean arterial blood pressure, VA visual acuity, BCVA best-corrected visual acuity.*P < 0.05, statistical differences between sexes.

Table 2 .
Multiple linear regression analysis between macular thickness and systemic parameters.*P < 0.05, significantly correlated parameters.

Table 3 .
Correlation between retinal, choroidal sub-region thickness and SE.T temporal, S superior, N nasal, I inferior.*Thickness of regions significantly associated with spherical equivalent refraction in the Pearson correlation analysis.thinningandfewerchanges in regions of retinal thickening.(2)Thevalue of SE exhibited a positive correlation with all regions of choroidal and retinal thickness, excluding the central and inner inferior regions of the full retina.(3) Every choroidal subregion was significantly correlated with the thickness of the full retina, excluding the central region.(4) The thickness of the choroid, GCIPL, and GCC of girls were significantly thicker than those of boys.The average and central thickness of the full retina and choroid showed no significant correlation with age, height, weight, or BMI.

Table 4 .
Paired t-test for changes in retinal thickness between 2 years.*P < 0.05, Statistical differences between the thickness in the baseline and second-year follow-up examination.

Table 5 .
Comparison of retinal thickness changes between different refractive errors (grouped by baseline examination).