This study evaluated the distribution of ECD in Chinese school children. Previous studies have reported the ECD distribution in different populations, but most of the subjects were adults. For reference, we list the ECD distribution in subgroups closest in age from previous studies (Table 2)14–27. It is worth noting that most of these previous studies were not population-based and the sample size was relatively small. Large-scale studies are still needed to establish normal ECD values among different populations by age. The mean (± SD) WTW in this study was 12.14 ± 0.38mm. Reports of normal WTW distribution were reviewed. The mean (± SD) WTW measured with Orbscan II in a German study34 was reported as 11.71 ± 0.42 mm. An Iranian study35 showed that the mean (± SD) WTW measured with Orbscan II in people aged 18-45 years was 11.65 ± 0.36mm, while another Iranian study36 found that the mean WTW measured with the LENSTAR/BioGraph in a 40- to 64-year-old population was 11.80 mm and that in the 40- to 44-year-old subgroup was 11.91mm. A Chinese study37 examined a large number of Chinese cataractous eyes and reported the mean WTW (± SD) measured with the ZEISS IOLMaster 700 as 12.00 ± 0.45 mm in males and 11.80 ± 0.50 mm in females aged 30- 40 years, close to our result. All of these studies suggested a decreasing trend of WTW with age. Thus, the relatively greater WTW in this study among Chinese school children is reasonable. The AL distribution in this study is similar to that in the population-based multicenter studies for Australian children38 and European children39, both measured with the Zeiss IOLmaster and that in the Anyang Childhood Eye Study for Chinese children40 using Lenstar LS900.
Table 2
Studies of corneal endothelial cell density.
Author | Year | Location | Age | Design | Sample Size | ECD (cells/mm2) |
Rao SK14 | 2000 | Chennai India | 20-30 y | | 104 eyes | 2782 ± 250 |
Müller A15 | 2002 | Glasgow area and central Scotland UK | 5-15 y | | 119 cases | 3542 ± 510 range 2576-5316 |
Padilla MD16 | 2004 | Makati Philippines | 20-30 y | | 114 eyes/ 57 cases | 2949 ± 270 |
Hashemian MN17 | 2006 | Iran | 20-30 y | | 102 eyes/ 102 cases | 2407 ± 399 |
Yunliang S18 | 2007 | Shantou China | 11-20 y | | 100 eyes | 3308 ± 356 |
Higa A19 | 2010 | Kumejima Japan | 40-49 y | population-based | 827 eyes/ 827 cases | 3031 ± 359 |
Mohammad-Salih PA20 | 2011 | Kuantan Malaysia | 20-30 y | | 49 eyes/ 49 cases | 2783 ± 286 |
Galgauska S21 | 2013 | Vilnius Lithuania | 20-29 y | | 55 eyes/ 28 cases | 2931 ± 371 range 2232-3610 |
Arici C22 | 2014 | Istanbul Turkey | 20-30 y | | 42 eyes/ 21 cases | 2910.2 ± 365.9 |
Ewete T23 | 2016 | Nigeria | 20-30 y | | 81 cases | 2860.70 ± 227.06 |
Islam QU24 | 2017 | Karachi Pakistan | 12-20 y | | 84 eyes/ 42 cases | 3021.24 ± 312.24 |
Elbaz U25 | 2017 | Toronto, Ontario, Canada | 4-5y | | 24 eyes/ 24 cases | 3746 ± 370 range 3145-5013 |
Abdellah MM26 | 2019 | Egypt | 20-30 y | | 89 eyes/ 89 cases | 2933.75 ± 345.92 range 2843.5-2983.7 |
Tananuvat N27 | 2020 | Chiang Mai Tailand | 11-20 y | | 72 eyes/ 36 cases | 2944.65 ± 231.95 |
This study | 2021 | Nanjing China | 6.83-7.83 y | population-based | 1171 eye/ 1171 cases | 2875.34 ±195.00 range 2066-3546 |
Age-related ECD decrease has been reported for several times. In this study, even within the narrow age range (6.83-7.83 years), ECD presented a decreasing trend towards age increase. However, age didn’t seem to be the only determinant of ECD. Within a relative narrow age range, we explored the correlation between ECD and BW, general condition (age, gender and BMI) and ocular biometric parameters.
The results in this study presented positive correlation between ECD and BW and revealed that children born with low BW had significantly lower ECD than those born with normal BW and high BW. It is known that environments in utero such as maternal nutrition could affect fetal gene expression and developmental plasticity by epigenetic pathways41. BW may partly reflect antenatal development of the fetus and status of the newborn. To our knowledge, none of previous studies have reported the correlation between ECD and BW, though several studies demonstrated associations between BW and other corneal parameters. One study found an association between a lower BW and the following changes — steeper CR, smaller WTW, thinner CCT, and shorter AL42. Another study revealed that low BW is associated with thinner corneas in adolescence43. This study suggested that ECD had positive correlations with BW and CCT, which may explain the consistency to some extent. We speculate that prenatal growth development might affect corneal morphology in childhood, adolescence and adults.
This study showed that girls had higher mean ECD than boys. Another population-based Japanese study among adults also demonstrated that the mean ECD in women was significantly greater than that in men after adjusting for age19. Likewise, in the Scottish study, the ECD was slightly higher in the girls than that in the boys, although the difference was not statistically significant15. The Filipino study also showed that women had a mean ECD 7.8% higher than men16. Nevertheless, a previous Chinese study reported significantly higher mean ECD in men than that in women19. Other studies such as the Egyptian study26 and the Iranian study17 presented no significant differences in mean ECD between men and women.
BMI is a statistical index using a person's weight and height to provide an estimate of body fat. This study proved no significant correlation between ECD and BMI. However, a previous Chinese study found that ECD in children aged 3 years with dysplasia and obesity was lower than that in normal children. But this difference did not apply to other age groups (4-7 years, 7-10 years and 10-12 years)44. Another previous study reported that CCT had no correlation with BMI in adults45. Whether malnutrition or overnutrition during specific life stage exerts great impact on ECD needs further study.
ECD decreases at a relatively rapid rate during early childhood, which is considered mainly due to normal eye growth and increase in corneal size7. Whether corneal size influences ECD of school children was explored in this study. We even calculated SAC according to the area formula for spherical cap, but no correlation was found. A previous study accorded to our result in the matter of corneal diameter46. Other studies found ECD to be negatively correlated with corneal diameter in children14,15,27. However, one of these studies admitted that the decline in ECD after the age of 2 years is presumably caused by cell loss rather than corneal growth15.
Controversial results towards the correlation between ECD and CCT have been reported by previous studies, including insignificant correlation26,27, negative correlation47, and positive correlation19,21. This study demonstrated the correlation to be positive, though CCT increase by 10mm accompany ECD increase by 4 cells/mm2.
The strengths of this study include its population-based design, large sample size, and standardized examination protocols performed by an expert team. The age range of the subjects is relatively narrow, decreasing the impact of large age span on ECD distribution. Our analyses of associations with ECD are different from previous studies, by taking BW and anthropometric parameters into consideration. The exploration among normal children may help us understand the development of ECD and provide references for developing the prediction model for ECD decrease in the future. The limitation of this study is that BW collected through questionnaire may be biased. In addition, the correlation obtained from the cross-sectional study needs to be validated and deepened by further studies.