- Email: [email protected]
Comparison of Corneal Biomechanical Properties between Indian and Chinese Adults
Comparison of Corneal Biomechanical Properties between Indian and Chinese Adults Jacqueline Chua, BOptom, PhD,1,2 Monisha E. Nongpiur, MD, PhD,1,2 Wanting Zhao, PhD,1,2 Yih Chung Tham, PhD,1 Preeti Gupta, PhD,1,3 Charumathi Sabanayagam, PhD,1,2,3 Tin Aung, FRCS, PhD,1,2,3 Tien Yin Wong, FRCS, PhD,1,2,3 Ching-Yu Cheng, MD, PhD1,2,3 Purpose: To investigate the difference in corneal hysteresis (CH) and corneal resistance factor (CRF) between Indian and Chinese populations. Design: Population-based cross-sectional study. Participants: Three hundred eighty-two Singaporean Indian persons and 764 Singaporean Chinese 50 years of age or older were included from the Singapore Indian Eye Study and Singapore Chinese Eye Study, respectively. Methods: Participants underwent standardized systemic and ocular examinations and intervieweradministered questionnaires for risk factor assessment. The CH and CRF were measured with the Ocular Response Analyzer (Reichert Ophthalmic Instruments, Buffalo, NY). Information on genetic ancestry was derived using principal component analysis. Linear regression models were used to investigate the association of CH and CRF with potential risk factors. Main Outcome Measures: Corneal hysteresis and CRF. Results: After excluding participants with a history of intraocular surgery, a diagnosis of glaucoma suspect or glaucoma, refractive surgery, or presence of corneal abnormalities, CH and CRF readings were available for 382 Indian persons. For each Indian participant, 2 Chinese participants were selected and matched for age and gender (n ¼ 764). There were no differences in the clinical measurements of CH (10.61.6 mmHg; P ¼ 0.670) or CRF (10.31.7 mmHg; P ¼ 0.103) between the ethnic groups. However, after adjusting for covariates, Indian persons had, on average, 0.18-mmHg higher CH levels than in Chinese (95% confidence interval [CI], 0.02e0.38; P ¼ 0.031). Consistently, CH level was correlated significantly with genetic ancestry in the Southeast Asian population. Corneal resistance factor level was not associated independently with self-reported ethnicity (95% CI, 0.10 to 0.29; P ¼ 0.335). Conclusions: Chinese have lower CH than Indian persons, and this disparity may reflect biomechanical differences of the cornea. Ophthalmology 2017;-:1e9 ª 2017 by the American Academy of Ophthalmology
Glaucoma is the second leading cause of irreversible blindness worldwide.1,2 Increasing evidence suggests that conventional anatomic risk factors for primary open-angle glaucoma (POAG) do not explain adequately why many people with elevated intraocular pressure (IOP) and thin central corneal thickness (CCT) never demonstrate the disease. Clear evidence for the failure of IOP and CCT to explain the disease is the fact that Chinese persons have a reduced glaucoma-related risk profile (lower IOP and thicker cornea) than Indian persons,3 yet our populationbased data show no reduced POAG prevalence among a Chinese population compared with an Indian population.4,5 If Chinese persons have a reduced glaucoma-related risk profile (lower IOP and thicker cornea) than Indian persons, then something other than IOP and CCT must explain their similarly high glaucoma prevalence as that of Indian persons. Thus, deciphering why certain individuals are susceptible to glaucoma may reside in other properties of the eye, and not simply in the IOP or CCT. ª 2017 by the American Academy of Ophthalmology Published by Elsevier Inc.
Corneal hysteresis (CH) and corneal resistance factor (CRF) are clinical measurements of corneal biomechanics that can be measured from the Ocular Response Analyzer (ORA; Reichert Ophthalmic Instruments, Buffalo, NY).6e8 It has been speculated that corneal biomechanical properties may reflect that of the sclera and optic nerve head. Low CH, representing a less viscoelastic cornea, seems to be an indicator of which eyes will demonstrate POAG.9e11 In light of its relevance to glaucoma, a comparison of the CH and CRF values between Indian and Chinese persons, the 2 largest racial and ethnic groups globally, may provide useful clinical information for glaucoma assessment for different ethnic groups.4,5 However, such direct comparisons using similar examination protocols across ethnic groups have not been performed. The purpose of the current study was to examine variations of CH and CRF between Chinese and Indian persons living in Singapore. We hypothesized that despite having lower IOP and thicker CCT, Chinese people may have POAG prevalence as high as that of Indian persons as a result of lower CH. http://dx.doi.org/10.1016/j.ophtha.2017.03.055 ISSN 0161-6420/17
1
Ophthalmology Volume -, Number -, Month 2017 Methods Study Population Participants in this study were derived from Chinese and Indian persons who participated in a substudy under the Singapore Epidemiology of Eye Disease study. Chinese participants were from the Singapore Chinese Eye Study (2009e2011), whereas Indian persons were from the Singapore Indian Eye Study (2012e2015). In brief, an age-stratified (by 10-year age groups) random sampling in each ethnic group was used to select ethnic Indian and Chinese persons 40 to 80 years of age living across the southwestern part of Singapore during each stipulated study period. Study methodology and details were identical and have been described elsewhere.12 In brief, the Singapore Epidemiology of Eye Disease study is an ongoing, population-based study that was designed to evaluate the prevalence, risk factors, and impact of major eye diseases among Singaporeans 40 to 80 years of age residing in the southwestern part of Singapore. For this substudy, 551 consecutive Indian participants who attended the study clinic during their follow-up examination between April 1 and August 31, 2015, were included. Of the 551 recruited Indian participants, 169 participants were excluded because of a history of intraocular surgery (n ¼ 95) or because of being glaucoma suspects or being diagnosed with glaucoma (n ¼ 74). Finally, 382 Indian participants (69.3%) were included in the analysis. A total of 3353 Chinese persons participated in the study between 2009 and 2011. Among these, 2878 (86.8%) underwent corneal biomechanical measurement, after excluding those with missing corneal biomechanical measurements (n ¼ 43), history of intraocular surgery (n ¼ 341), and glaucoma suspects or those with glaucoma (n ¼ 91). We previously published the corneal biomechanical profile of the Chinese population showing that age and gender were determinants for CH and CRF.13 Therefore, we applied a frequency-matching technique based on age (by 5-year age groups) and gender matching, whereby 2 Chinese participants were selected for each Indian participant. Therefore, 764 ageand gender-matched Chinese participants were selected randomly from those Chinese participants with available corneal biomechanical measurements (n ¼ 2878). Ethics approval was obtained from the Singapore Eye Research Institute Institutional Review Board. All study participants provided written informed consent in adherence to the tenets of the Declaration of Helsinki.
Examination Procedures Both Indian and Chinese participants underwent similar clinical and ocular examinations in the same clinic (Singapore Eye Research Clinic).12 Each participant underwent a standardized interview and eye examination that included assessment of bestcorrected visual acuity, assessment of corneal biomechanical properties by the ORA, autorefraction, and keratometry (corneal curvature in millimeters; Canon RK-5, Autoref-Keratometer Japan]).14 Spherical equivalent was defined [Canon Inc., Ltd., Ota, as sphere plus half cylinder. Ocular biometry, including anterior chamber depth and axial length, was measured using noncontact partial coherence interferometry (IOLMaster; Carl Zeiss Meditec, Dublin, CA), CCT was measured by an ultrasound pachymeter (Advent; Mentor O & O, Inc., Norwell, MA), and IOP was measured by Goldmann applanation tonometry (Haag-Streit, Koenz, Switzerland). Slit-lamp examination of the anterior segment, evaluation of the retina, and dynamic gonioscopy (Sussman 4-mirror gonioscope; Ocular Instruments, Inc., Bellevue, WA) were conducted by the study ophthalmologists.15 Each participant underwent height and weight measurement, and these were used to determine the body mass index, which was
2
calculated as body weight (in kilograms) divided by body height (in meters) squared. Seated blood pressure and blood samples were collected during the examination procedures. Hypertension was defined as systolic blood pressure of 140 mmHg or more, diastolic blood pressure of 90 mmHg or more, physician diagnosed hypertension, or self-reported history of hypertension. Diabetes mellitus was defined as random glucose of 11.1 mmol/l or more, diabetic medication use, or a physician-diagnosed history of diabetes.
Ocular Response Analyzer Details regarding the ORA have been published previously.16 The ORA uses a noncontact rapid air pulse to generate a signal. The ORA signal depicts 2 IOP measurements. The difference between the first and second measurement is the CH and is an indicator of the viscous properties of the cornea. The ORA also provides measurement of CRF, which is an indicator of the overall resistance or elastic properties of the cornea.17 The ORA parameters were measured in the right eye of each eligible participant; when the right eye was ineligible, parameters were measured in the left eye. The measurement was repeated 3 times; differences between the ORA pressure readings of more than 3 mmHg warranted a fourth measurement. Exclusion criteria for this analysis were a history of intraocular surgery or refractive surgery, presence of corneal abnormalities such as keratoconus, corneal scarring that would preclude accurate ORA and IOP measurements, or a diagnosis of glaucoma suspect or glaucoma. Glaucoma suspects were defined according to prespecified criteria,15,18 which included IOP of more than 21 mmHg, gonioscopic findings of occludable angles, presence of peripheral anterior synechiae, cup-to-disc ratio of more than 0.6, disc asymmetry with cup-to-disc ratio difference of more than 0.2, pseudoexfoliation syndrome, pigment dispersion syndrome, and known glaucoma patients. Glaucoma was diagnosed according to the International Society of Geographical and Epidemiologic Ophthalmology classification.18
Ancestry Inference Using Genome-Wide Single Nucleotide Polymorphism Markers In addition to self-reported ethnicity, individual genetic ancestry were inferred using principal component (PC) analysis by the smartPCA program (EIGENSTRAT software version 4.2)19 to examine the relationship of race and its associations to CH and CRF.20 The method implemented in EIGENSTRAT works well if markers are independent21; hence, PC analysis was applied to a reduced single nucleotide polymorphism (SNP) set that had low linkage disequilibrium. The following steps were taken to arrive at a final set of 57 237 SNPs. Autosomal SNPs with minor allele frequency of 0.05 or more within each ethnic group in our study population were selected and used for analysis. Genotype data, which had undergone strict quality checks, were merged together and only SNPs shared by both ethnic groups were used for analysis. In addition, known high linkage disequilibrium regions were excluded (Chr5, 44e51.5 Mb; Chr6, 25e37 Mb; Chr8, 1e12.7 Mb; Chr11, 45e57 Mb; and Chr11, 84e86 Mb).22 Then, SNPs were pruned using the indep-pairwise option in PLINK (version 1.07; available at: http://pngu.mgh.harvard.edu/ wpurcell/plink/; accessed September 1, 2016.), with a window size of 50 SNPs, shifting by 5 SNPs at each step and removing 1 of a pair of SNPs if the linkage disequilibrium was more than 0.2. We then applied PC analysis to genotype data to infer continuous axes of genetic variation. Intuitively, the axes of variation reduce the data to a small number of dimensions,
Chua et al
Corneal Biomechanical Properties among Asians
Table 1. Characteristics of Participants by Ethnicity Characteristics
Indian (n [ 382)
Chinese (n [ 764)
P Value
Age (yrs) Gender (female) Intraocular pressure (mmHg) Central corneal thickness (mm) Corneal curvature (mm) Anterior chamber depth (mm) Refractive error (SE) Axial length (mm) Hypertension Diabetes HbA1c (%) Body mass index (kg/m2) Diastolic blood pressure (mmHg) Systolic blood pressure (mmHg) Corneal hysteresis Corneal resistance factor
60.06.9 204 (53.4) 15.32.5 544.133.5 7.60.2 3.20.3 0.32.0 23.51.0 245 (64.1) 148 (44.6) 6.61.5 26.65.0 77.59.2 137.218.8 10.61.6 10.41.8
59.87.2 408 (53.4) 14.23.1 552.733.3 7.70.3 3.20.3 0.42.6 23.91.4 450 (58.9) 102 (14.1) 6.11.0 23.63.5 77.69.6 136.019.1 10.61.6 10.21.7
0.804 1.000 <0.001 <0.001 0.144 0.848 <0.001 <0.001 0.087 <0.001 <0.001 <0.001 0.867 0.319 0.670 0.103
HbA1c ¼ glycosylated hemoglobin; SD ¼ standard deviation; SE ¼ spherical equivalence. Data are mean SD or number (%), as appropriate.
describing as much variability as possible. The first 2 PCs captured 5.45% of variation, whereas the third PC captured only 0.09% of variation.
Table 2. Mean (Standard Deviation) of Corneal Hysteresis and Corneal Resistance Factor by Age Group, Gender, and Ethnicity Indian (n [ 382) Corneal hysteresis Age group (yrs) 50e54 55e59 60e64 65e69 70e74 75 P value for trend Gender Male Female P value Corneal resistance factor Age group (yrs) 50e54 55e59 60e64 65e69 70e74 75 P value for trend Gender Male Female P value
Chinese (n [ 764)
No.
Mean SD
No.
Mean SD
104 108 72 61 22 15
11.11.4 10.71.5 10.41.5 10.01.7 10.71.6 10.01.1 <0.001*
208 216 144 122 44 30
11.01.7 10.71.6 10.71.5 10.21.4 10.11.9 9.71.5 <0.001*
178 204
10.61.6 10.61.6 0.957y
356 408
10.31.5 10.91.6 0.001z
104 108 72 61 22 15
10.91.8 10.41.7 10.11.6 9.91.9 10.62.0 9.80.8 0.002*
208 216 144 122 44 30
10.51.9 10.31.6 10.21.5 10.01.6 9.91.8 9.21.8 <0.001*
178 204
10.31.8 10.51.7 0.268z
356 408
9.91.6 10.51.7 <0.001z
SD ¼ standard deviation. *Test for differences in trend between age groups. y Test for differences between gender, based on independent t test. z Test for differences between gender, based on Kruskal-Wallis test.
Statistical Analysis The mean values of CH and CRF and their ocular and systemic associations were studied. The ShapiroeWilk test was used to assess the normality of the distribution of the corneal biomechanical parameters. To compare continuous variables among groups, an independent t test was performed for normally distributed variables, whereas the KruskaleWallis test was used for nonnormally variables. The chi-square test was used for categorical variables. The strengths of the linear relationship between variables were assessed by Pearson’s correlation coefficients. Associations between ocular and systemic factors with CH and CRF were assessed using univariate and multivariate linear regression models. In the fully adjusted multivariate model, the differences in CH and CRF between Indian and Chinese persons were adjusted for age, gender, ethnicity, IOP, CCT, corneal curvature, axial length, and factors with P < 0.05 in the univariate model simultaneously. In the reduced adjusted multivariate model, we adjusted only for age, gender, ethnicity, and individual ocular parameter separately, based on the results from the univariate model. Furthermore, genetic ancestry inferred from SNPs was used in place of self-reported ethnicity to determine its association with CH and CRF. Analyses were performed using STATA software version 13 (StataCorp, College Station, TX).
Results Two Chinese participants were matched for each of the 382 Indian participants included in the analysis, based on age (by 5-year age groups) and gender (n ¼ 764). Table 1 compares the mean values of the ocular and systemic variables between the Indian and Chinese populations. The mean ages standard deviation of the participants were 60.06.9 years for Indian persons and 59.87.2 years for Chinese persons. Of these participants, 53.4% were women. Indian persons were more likely to have a higher IOP, thinner CCT, more hyperopic refractive error, shorter axial length, higher prevalence of diabetes, higher hemoglobin A1c levels, and greater body mass index than Chinese persons (P < 0.001). There were no differences in the overall CH (10.61.6 mmHg; P ¼ 0.670) or the CRF (10.31.7 mmHg; P ¼ 0.103) values between ethnic groups (Table 1).
3
4 Table 3. Association with Corneal Hysteresis by Ethnicity Assessed Using Multivariate Linear Regression Models All 0.43 0.32 0.13 0.19 0.81 0.18 0.01 0.11 0.03 0.26 0.09 0.03 0.03 0.01
(0.56 to 0.31)z (0.14e0.50)z (0.18 to 0.44) (0.16e0.21)z (1.16 to 0.46)z (0.45 to 0.10) (0.04 to 0.03) (0.19 to 0.04)x (0.23 to 0.16) (0.04 to 0.51) (0.01e0.17)x (0.05 to 0.01)x (0.13 to 0.07) (0.05 to 0.05)
Reference 0.01 (0.19 to 0.19)
(0.55 to 0.29)z (0.05e0.41)x (0.56 to 0.06) (0.17e0.22)z (1.21 to 0.47)z d d 0.06 (0.13 to 0.02) d d 0.07 (0.01 to 0.15) 0.03 (0.05 to 0.01)x d d 0.42 0.23 0.25 0.19 0.84
Reference 0.18 (0.02e0.38)x
0.47 0.10 0.53 0.17 0.37 0.03 0.05 0.11 0.18 0.20 0.06 0.04 0.13 0.03
(0.69 to 0.24)z (0.41 to 0.21) (0.09 to 1.14) (0.12e0.21)z (1.01 to 0.26) (0.50 to 0.44) (0.03 to 0.13) (0.27 to 0.04) (0.15 to 0.51) (0.14 to 0.55) (0.05 to 0.17) (0.07 to 0.01)x (0.04 to 0.03) (0.05 to 0.011) d d
Chinese Persons Model 2y
Model 1*
(0.68 to 0.22)z (0.42 to 0.24) (0.44 to 0.84) (0.12e0.22)z (1.30 to 0.14) d d 0.04 (0.22 to 0.13) d d 0.03 (0.07 to 0.13) 0.03 (0.06 to 0.01) d d 0.45 0.09 0.20 0.17 0.58
d d
Model 2y
Model 1* 0.42 0.53 0.01 0.20 1.00 0.24 0.02 0.11 0.12 0.27 0.13 0.01 0.09 0.02
(0.57 to 0.27)z (0.31e0.75)z (0.36 to 0.36) (0.16e0.22)z (1.42 to 0.59)z (0.58 to 0.11) (0.06 to 0.03) (0.20 to 0.03)x (0.37 to 0.12) (0.07 to 0.61) (0.02e0.25)x (0.05 to 0.02) (0.21 to 0.03) (0.08 to 0.05) d d
(0.56 to 0.26)z (0.17e0.61)z (0.79 to 0.08)x (0.17e0.23)z (1.41 to 0.54)z d d 0.06 (0.14 to 0.02) d d 0.13 (0.02e0.24)x 0.03 (0.06 to 0.01) d d 0.41 0.39 0.43 0.20 0.97
d d
HbA1c ¼ glycosylated hemoglobin; SE ¼ spherical equivalent; d ¼ data were intentionally omitted. Data are b (95% confidence interval). *Model 1 adjusted for age and gender and additionally adjusted for ethnicity in the analysis of the 2 ethnic groups combined. y Model 2 adjusted for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, HbA1c, body mass index, and additionally adjusted for ethnicity in the analysis of the 2 ethnic groups combined. z P < 0.001. x P < 0.05.
Ophthalmology Volume -, Number -, Month 2017
Age (per 10 yrs) Gender (female) Intraocular pressure (per 10 mmHg) Central corneal thickness (per 10 mm) Corneal curvature (mm) Anterior chamber depth (mm) Refractive error (SE) Axial length (mm) Hypertension Diabetes HbA1c (%) Body mass index (kg/m2) Diastolic blood pressure (per 10 mmHg) Systolic blood pressure (per 10 mmHg) Self-reported ethnicity Chinese Indian
Indian Persons Model 2y
Model 1*
Characteristics
Chua et al
Corneal Biomechanical Properties among Asians
Figure 1. The distribution of (A) corneal hysteresis and (B) corneal resistance factor between Indian and Chinese participants. *Data and P value shown for corneal hysteresis are after adjustment for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, body mass index, hemoglobin A1c, and ethnicity. yData and P value shown for corneal resistance factor are after adjustment for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, diabetes, systolic blood pressure, and ethnicity.
Corneal Hysteresis and Corneal Resistance Factor with Age and Gender Table 2 shows the variation of CH and CRF with age and gender. Negative correlations between age and CH were seen for both ethnic groups (P < 0.001 for trend; Table 2). A similar association was noted between age and CRF for Indian persons (P ¼ 0.002 for trend) as well as Chinese persons (P < 0.001 for trend; Table 2). There were no gender differences in the CH (P ¼ 0.957) or CRF (P ¼ 0.268) values for Indian persons (Table 2). However, Chinese women had higher mean CH (10.9 vs. 10.3 mmHg; P < 0.001) and CRF (10.5 vs. 9.9 mmHg; P < 0.001) than Chinese men (Table 2).
Ocular and Systemic Factors with Corneal Hysteresis The association between CH and ocular and systemic factors by ethnic groups is summarized in Table 3. Multivariate linear regression analysis showed that younger age (b ¼ 0.42 per 10 years; 95% confidence interval [CI], 0.55 to 0.29; P < 0.001), female gender (b ¼ 0.23; 95% CI, 0.05e0.41; P ¼ 0.011), thicker CCT (b ¼ 0.19 per 10 mm; 95% CI, 0.17e0.22; P < 0.001), steeper corneal curvature (b ¼ 0.84; 95% CI, 1.21 to 0.47; P < 0.001), lesser body mass index levels (b ¼ 0.03; 95% CI, 0.05 to 0.01; P ¼ 0.009), and Indian ethnicity compared with Chinese ethnicity (b ¼ 0.18; 95% CI, 0.02e0.38; P ¼ 0.031; Fig 1A) were associated independently with higher CH levels (model 2 in Table 3). Further analysis stratified by ethnic group showed that the higher CH levels with female gender, steeper corneal curvature, and higher hemoglobin A1c levels were seen for Chinese persons, but not for Indian persons. We used another multivariate regression model, adjusting only for age, gender, ethnicity, and individual ocular parameters separately, to identify which ocular factor was the negative confounder that accounted for the observed null association between ethnicity and corneal hysteresis in the univariate analysis (model 1 in Table 3). The CH levels in Indian persons were, on average,
0.25 mmHg higher than those in Chinese persons, after controlling for CCT (95% CI, 0.06e0.43; P ¼ 0.010). There was a positive correlation between CCT and CH for Indian persons (r ¼ 0.38; P < 0.001) and Chinese persons (r ¼ 0.54; P < 0.001), with no statistically significant difference between the 2 groups (P ¼ 0.349). The CH levels remained similar between Indian and Chinese persons after adjusting for corneal curvature and axial length (all P > 0.05).
Ocular and Systemic Factors with Corneal Resistance Factor The association of CRF with ocular and systemic factors by ethnic groups is shown in Table 4. Multivariate linear regression analysis showed that younger age per 10 years (b ¼ 0.34; 95% CI, 0.47 to 0.20; P < 0.001), female gender (b ¼ 0.28; 95% CI, 0.10e0.45; P ¼ 0.002), higher IOP per 10 mmHg (b ¼ 1.57; 95% CI, 1.27e1.87; P < 0.001), thicker CCT per 10 mm (b ¼ 0.24; 95% CI, 0.21e0.26; P < 0.001), steeper corneal curvature (b ¼ 0.76; 95% CI, 1.13 to 0.40; P < 0.001), and presence of diabetes (b ¼ 0.28; 95% CI, 0.07e0.49; P ¼ 0.010) were associated independently with higher CRF levels (model 2 in Table 5). The CRF levels in Indian persons were similar to those in Chinese persons in both the full model (Table 5; b ¼ 0.09; 95% CI, 0.10 to 0.29; P ¼ 0.355; Fig 1B) and reduced model (P > 0.05). Further analysis stratified by ethnic group showed that the higher CRF levels with female gender, steeper corneal curvature, and presence of diabetes were seen for Chinese persons, but not for Indian persons.
Genetic Ancestry and Corneal Biomechanical Parameters The Indian and Chinese ancestries were clearly distinguishable by the first PC of genetic ancestry (Fig 2).3 Briefly, the first and second PCs sequentially (cumulatively) accounted for 5.26% and 0.19% (5.45%) of the total genetic variation. Each standard deviation increase in the first PC (standard deviation, 0.01) was associated significantly with an 11.72-mmHg lower CH level
5
6 Table 4. Association with Corneal Resistance Factor by Ethnicity Assessed Using Multivariate Linear Regression Models All
Age (per 10 yrs) Gender (female) Intraocular pressure (per 10 mmHg) Central corneal thickness (per 10 mm) Corneal curvature (mm) Anterior chamber depth (mm) Refractive error (SE) Axial length (mm) Hypertension Diabetes HbA1c (%) Body mass index (kg/m2) Diastolic blood pressure (per 10 mmHg) Systolic blood pressure (per 10 mmHg) Self-reported ethnicity Chinese Indian
0.35 0.40 2.15 0.26 0.75 0.13 0.03 0.06 0.15 0.51 0.13 0.01 0.16 0.11
Indian Persons Model 2y
(0.49 to 0.22)z 0.34 (0.47 to 0.20)z (0.20e0.60)z 0.28 (0.10 to 0.45)x z (1.83e2.46) 1.57 (1.27 to 1.87)z (0.24e0.29)z 0.24 (0.21 to 0.26)z (1.12 to 0.35)z 0.76 (1.13 to 0.40)z (0.43 to 0.18) d (0.07 to 0.10) d (0.14 to 0.20) 0.01 (0.08 to 0.06) (0.06 to 0.36) d 0.28 (0.07 to 0.49)x (0.25e0.77)z (0.04e0.22)x d (0.03 to 0.02) d x d (0.05e0.27) 0.04 (0.01 to 0.09) (0.05e0.16)z
Reference 0.17 (0.04 to 0.38)
Reference 0.09 (0.10 to 0.29)
Model 1* 0.41 0.10 2.65 0.25 0.15 0.06 0.04 0.08 0.30 0.46 0.08 0.01 0.31 0.13
Chinese Persons Model 2y
(0.67 to 0.16)x 0.38 (0.62 to 0.13)x (0.26 to 0.45) 0.03 (0.31 to 0.37) (2.00e3.30)z 2.17 (1.51e2.84)z (0.20e0.29)z 0.21 (0.16e0.26)z (0.88 to 0.58) 0.46 (1.22 to 0.30) (0.48 to 0.59) d (0.05 to 0.13) d (0.25 to 0.10) 0.02 (0.19 to 0.16) (0.08 to 0.67) e (0.07e0.85)x 0.27 (0.06 to 0.60) (0.04 to 0.21) (0.05 to 0.02) d (0.12e0.51)x d (0.04e0.23)z 0.02 (0.07 to 0.11) d d
d d
Model 1* 0.33 0.55 1.98 0.27 0.99 0.21 0.05 0.05 0.09 0.57 0.18 0.01 0.10 0.10
Model 2y
(0.49 to 0.16)z 0.33 (0.48 to 0.17)z (0.31e0.79)z 0.40 (0.19e0.60)z z (1.62e2.34) 1.36 (1.03e1.69)z (0.24e0.30)z 0.24 (0.21e0.27)z (1.44 to 0.55)z 0.90 (1.31 to 0.49)z (0.57 to 0.16) d d (0.10 to 0.01)x (0.14 to 0.03) 0.01 (0.08 to 0.73) (0.18 to 0.35) d (0.22e0.92)x 0.32 (0.04e0.61)x (0.06e0.31)x (0.03 to 0.02) d (0.03 to 0.23) d (0.03e0.16)x 0.06 (0.01e0.11)x d d
d d
HbA1c ¼ glycosylated hemoglobin; SE ¼ spherical equivalent; d ¼ data were intentionally omitted. Data are b (95% confidence interval). *Model 1 adjusted for age and gender, and additionally adjusted for ethnicity in the analysis of the 2 ethnic groups combined. y Model 2 adjusted for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, diabetes, and systolic blood pressure and additionally adjusted for ethnicity in the analysis of the 2 ethnic groups combined. z P < 0.001. x P < 0.05.
Ophthalmology Volume -, Number -, Month 2017
Model 1*
Characteristics
Chua et al
Corneal Biomechanical Properties among Asians
Table 5. Associations of Genetic Ancestry with Corneal Hysteresis and Corneal Resistance Factor Corneal Hysteresis (mmHg) Model 1* Ancestry PC 1 PC 2
Corneal Resistance Factor (mmHg) y
Model 2
Model 3z
Model 1*
b (95% Confidence Interval)
P Value
b (95% Confidence Interval)
P Value
b (95% Confidence Interval)
P Value
b (95% Confidence Interval)
P Value
1.14 (9.16 to 6.88) 5.27 (15.34 to 4.79)
0.780 0.304
11.72 (19.74 to 3.70) 0.43 (9.51 to 8.65)
0.004 0.930
7.40 (15.92 to 1.13) 4.83 (15.54 to 5.88)
0.089 0.377
7.62 (15.66 to 0.37) 3.04 (6.24 to 12.32)
0.062 0.520
PC ¼ principal component. b represents the change in corneal hysteresis or corneal resistance factor for each standard deviation unit increase in PCs. *Model 1 adjusted for age and gender. y Model 2 adjusted for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, hemoglobin A1c, and body mass index. z Model 3 adjusted for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, diabetes, and systolic blood pressure.
(95% CI, 19.74 to 3.70; P ¼ 0.004), after adjusting for significant factors in the multivariate regression model (model 2 in Table 5), as well as in a reduced multivariate model adjusting only for age, gender, and CCT (b ¼ 8.53; 95% CI, 15.94 to 1.11; P ¼ 0.024). These analyses of genetic PC, which was estimated on the basis of genome-wide genotypes, consistently indicated that Chinese ancestry was associated with lower levels of CH than Indian ancestry. That is, there was a significant trend of decreasing CH as the percentage of Chinese ancestry increased. Genetic ancestry was not associated significantly with CRF, both before and after adjustment for relevant confounders (all P > 0.05; models 1 and 3 in Table 5).
Discussion In this population-based study in Singapore, after adjusting for ocular biometric parameters, namely IOP, CCT, corneal curvature, and axial length, Chinese persons had significantly lower CH and similar CRF values when compared with Indian persons. Comparing the differences in CH value between Asian subgroups may be relevant when
Figure 2. The principal components of genetic ancestry in the Singapore Indian Eye Study and Singapore Chinese Eye Study: principal component 1 (PC1) versus principal component 2 (PC2) by self-reported ethnicity. There are 2 distinct clusters, corresponding to Indian (blue) and Chinese (red) study participants. First principal component (PC1) perfectly distinguishes these 2 populations, and PC2 is able to capture the ancestry variations within the Indian group.
investigating the role of altered corneal biomechanics in ocular diseases such as glaucoma and keratoconus. We also assessed the ethnic differences of CH and CRF on the basis of genetic ancestry using the same methodology to collect data between ethnic groups. Comparing the mean CH among different populationbased studies would help to clarify the interethnic variation of corneal biomechanical properties. The clinical measurement of CH in our study sample was 10.61.6 mmHg, which was comparable between Chinese and Indian persons and was similar to that of healthy Japanese23 and Brazilian24 participants. However, interethnic variations in ocular parameters may account for the similarities observed among these studies. To compare the association between CH and ethnicity more accurately, we incorporated the effect of relevant ocular parameters, namely CCT, corneal curvature, and axial length, and computed the adjusted mean values of CH between the 2 ethnic groups (Table 3; Fig 1A). We found that CH is 0.18 mmHg lower for Chinese adults 50 years of age and older compared with similarly aged Indian persons, after adjusting for all other variables in the multiple linear regression model. This suggests that similarities in the measured CH between Chinese and Indian persons are explained largely by the differences in ocular biometry parameters, specifically CCT found between these 2 groups, signifying a distinct biomechanical difference between their corneas. Ocular parameters may influence clinical measurements of the ORA. We found that eyes with thicker cornea, steeper cornea, and shorter axial length exhibit higher CH levels. Similarly, other studies have shown that the ORA is influenced by ocular parameters such as CCT,24e28 corneal curvature,13 and axial length.13,29,30 Recently, the UK Biobank study showed that diabetics essentially have similar IOP measurements as nondiabetics after compensating for their corneal biomechanics.31 Leite et al27 further demonstrated that the interethnic differences of clinical CH measurement between black and white Americans were nullified because CH was correlated significantly with CCT. Similarly, we found that CCT was the major ocular parameter that influenced the hysteresis of the tissue, adding another layer of complexity to determining the socalled true rigidity of the cornea. The variety of errors encountered during clinical measurements is similar to the
7
Ophthalmology Volume -, Number -, Month 2017 analogy for IOP measurement using tonometry; the clinical parameter of CH obtained from the ORA in relation to measuring the cornea’s viscoelasticity (varying stiffness with rate of loading) clearly falls in this domain. The modest difference in CH suggests that there could be other factors that may explain the ethnic differences in glaucoma prevalence and individual susceptibility to disease. Unfortunately, our understanding of these ocular parameters that result in CH measurement errors or bias is far from complete. The cornea itself is a relatively complex tissue that exhibits various mechanical characteristics such as viscoelasticity, anisotropy, nonlinearity, and heterogeneity. The ORA is the first commercially available technology that allowed measurement of some biomechanical properties of the cornea. Improvements in technology and subsequent availability of newer instruments such as the Corneal Visualization Scheimpflug Technology may help to characterize better the complex biomechanical characteristics of the cornea in vivo. Study Strengths and Limitations Self-reported race and ethnicity are used commonly in health research to assess an individual’s background origin. The use of self-reported race and ethnicity in research studies can be particularly challenging for individuals who may not be fully aware of their own complex ancestry mixture. In the current study, using objective and accurate estimates of genetic ancestry, we further showed true genetic ancestry-based differences in CH between the 2 Asian subgroups. Other strengths of the study include its population-based, relatively large sample size and the availability of 2 major Asian ethnic groups with similar standardized study methodology. For example, we have identical methods of measurement for ocular parameters, such as ORA, autokeratometer, IOLMaster, ultrasound pachymeter, and Goldmann applanation tonometry, in each ethnic group, which enabled the direct and robust comparison of data. However, our study has some limitations. Subclinical corneal cases may be overlooked because our participants underwent only slitlamp examination, which rules out corneal edema or corneal dystrophy, and did not undergo assessment using corneal topographers such as Orbscan (Bausch & Lomb, Rochester, NY) or Pentacam (Oculus, Inc., Arlington, WA) By including only healthy participants, we were not able to evaluate differences in corneal biomechanics between races in glaucomatous patients. Future longitudinal studies are needed to elucidate the relationship between CH and susceptibility to glaucoma developing in Chinese and Indian persons. In conclusion, our population-based study of Indian and Chinese adults 50 years of age or older in Singapore confirms the presence of ethnic variation in CH, but not CRF, between Asian subgroups. Overall, after adjusting for ocular parameters, Chinese participants have a lower CH when compared with Indian persons. In view that corneal hysteresis is lower in eyes with POAG, this may contribute to the POAG prevalence in Chinese compared with Indian persons.
8
References 1. Bourne RR, Stevens GA, White RA, et al. Causes of vision loss worldwide, 1990e2010: a systematic analysis. Lancet Glob Health. 2013;1:e339-e349. 2. Tham YC, Li X, Wong TY, et al. Global prevalence of glaucoma and projections of glaucoma burden through 2040: a systematic review and meta-analysis. Ophthalmology. 2014;121:2081-2090. 3. Chua J, Tham YC, Liao J, et al. Ethnic differences of intraocular pressure and central corneal thickness: the Singapore Epidemiology of Eye Diseases study. Ophthalmology. 2014;121:2013-2022. 4. Baskaran M, Foo RC, Cheng CY, et al. The prevalence and types of glaucoma in an urban Chinese population: the Singapore Chinese Eye Study. JAMA Ophthalmol. 2015;133: 874-880. 5. Narayanaswamy A, Baskaran M, Zheng Y, et al. The prevalence and types of glaucoma in an urban Indian population: the Singapore Indian Eye Study. Invest Ophthalmol Vis Sci. 2013;54:4621-4627. 6. Deol M, Taylor DA, Radcliffe NM. Corneal hysteresis and its relevance to glaucoma. Curr Opin Ophthalmol. 2015;26:96102. 7. Narayanaswamy A, Su DH, Baskaran M, et al. Comparison of ocular response analyzer parameters in chinese subjects with primary angle-closure and primary open-angle glaucoma. Arch Ophthalmol. 2011;129:429-434. 8. Grise-Dulac A, Saad A, Abitbol O, et al. Assessment of corneal biomechanical properties in normal tension glaucoma and comparison with open-angle glaucoma, ocular hypertension, and normal eyes. J Glaucoma. 2012;21:486-489. 9. Congdon NG, Broman AT, Bandeen-Roche K, et al. Central corneal thickness and corneal hysteresis associated with glaucoma damage. Am J Ophthalmol. 2006;141:868-875. 10. Medeiros FA, Meira-Freitas D, Lisboa R, et al. Corneal hysteresis as a risk factor for glaucoma progression: a prospective longitudinal study. Ophthalmology. 2013;120:1533-1540. 11. De Moraes CV, Hill V, Tello C, et al. Lower corneal hysteresis is associated with more rapid glaucomatous visual field progression. J Glaucoma. 2012;21:209-213. 12. Lavanya R, Jeganathan VS, Zheng Y, et al. Methodology of the Singapore Indian Chinese Cohort (SICC) eye study: quantifying ethnic variations in the epidemiology of eye diseases in Asians. Ophthalmic Epidemiol. 2009;16:325-336. 13. Narayanaswamy A, Chung RS, Wu RY, et al. Determinants of corneal biomechanical properties in an adult Chinese population. Ophthalmology. 2011;118:1253-1259. 14. Chua J, Lim B, Fenwick EK, et al. Prevalence, risk factors, and impact of undiagnosed visually significant cataract: the Singapore Epidemiology of Eye Diseases Study. PLoS One. 2017;12:e0170804. 15. Chua J, Baskaran M, Ong PG, et al. Prevalence, risk factors, and visual features of undiagnosed glaucoma: the Singapore Epidemiology of Eye Diseases Study. JAMA Ophthalmol. 2015;133(8):938-946. 16. Luce DA. Determining in vivo biomechanical properties of the cornea with an ocular response analyzer. J Cataract Refract Surg. 2005;31:156-162. 17. Kotecha A. What biomechanical properties of the cornea are relevant for the clinician? Surv Ophthalmol. 2007;52(Suppl 2): S109-S114. 18. Foster PJ, Buhrmann R, Quigley HA, et al. The definition and classification of glaucoma in prevalence surveys. Br J Ophthalmol. 2002;86:238-242.
Chua et al
Corneal Biomechanical Properties among Asians
19. Price AL, Patterson NJ, Plenge RM, et al. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 2006;38:904-909. 20. Chua J, Koh JY, Tan AG, et al. Ancestry, socioeconomic status, and age-related cataract in Asians: the Singapore Epidemiology of Eye Diseases Study. Ophthalmology. 2015;122:2169-2178. 21. Patterson N, Price AL, Reich D. Population structure and eigenanalysis. PLoS Genet. 2006;2:e190. 22. Chen J, Zheng H, Bei JX, et al. Genetic structure of the Han Chinese population revealed by genome-wide SNP variation. Am J Hum Genet. 2009;85:775-785. 23. Kamiya K, Hagishima M, Fujimura F, et al. Factors affecting corneal hysteresis in normal eyes. Graefes Arch Clin Exp Ophthalmol. 2008;246:1491-1494. 24. Fontes BM, Ambrosio Jr R, Alonso RS, et al. Corneal biomechanical metrics in eyes with refraction of -19.00 to þ9.00 D in healthy Brazilian patients. J Refract Surg. 2008;24:941-945. 25. Kotecha A, Elsheikh A, Roberts CR, et al. Corneal thicknessand age-related biomechanical properties of the cornea measured with the ocular response analyzer. Invest Ophthalmol Vis Sci. 2006;47:5337-5347.
26. Shah S, Laiquzzaman M, Cunliffe I, et al. The use of the Reichert ocular response analyser to establish the relationship between ocular hysteresis, corneal resistance factor and central corneal thickness in normal eyes. Cont Lens Anterior Eye. 2006;29:257-262. 27. Leite MT, Alencar LM, Gore C, et al. Comparison of corneal biomechanical properties between healthy blacks and whites using the Ocular Response Analyzer. Am J Ophthalmol. 2010;150:163-168.e1. 28. Mangouritsas G, Morphis G, Mourtzoukos S, et al. Association between corneal hysteresis and central corneal thickness in glaucomatous and non-glaucomatous eyes. Acta Ophthalmol. 2009;87:901-905. 29. Shen M, Fan F, Xue A, et al. Biomechanical properties of the cornea in high myopia. Vision Res. 2008;48:2167-2171. 30. Chang PY, Chang SW, Wang JY. Assessment of corneal biomechanical properties and intraocular pressure with the Ocular Response Analyzer in childhood myopia. Br J Ophthalmol. 2010;94:877-881. 31. Chan MP, Grossi CM, Khawaja AP, et al. Associations with intraocular pressure in a large cohort: results from the UK Biobank. Ophthalmology. 2016;123:771-782.
Footnotes and Financial Disclosures Originally received: December 17, 2016. Final revision: March 29, 2017. Accepted: March 29, 2017. Available online: ---.
or funding organization had no role in the design or conduct of this research. Author Contributions: Manuscript no. 2016-1066.
1
Singapore Eye Research Institute, Singapore National Eye Centre, Singapore, Republic of Singapore.
2
Conception and design: Chua, Tham, Gupta, Sabanayagam, Aung, Wong, Cheng Analysis and interpretation: Chua, Zhao, Sabanayagam, Cheng
Ophthalmology and Visual Sciences Academic Clinical Program, DukeNUS Medical School, Singapore, Republic of Singapore.
Data collection: Chua, Tham, Gupta
3
Overall responsibility: Chua, Nongpiur, Sabanayagam, Aung, Wong, Cheng
Department of Ophthalmology, Yong Loo Lin School of Medicine, National University of Singapore and National University Health System, Singapore, Republic of Singapore. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Supported by the SingHealth Foundation, Singapore, Republic of Singapore (grant no.: SHF/FG563P/2014); the National Medical Research Council, Singapore, Republic of Singapore (grant nos.: 0796/2003, IRG07nov013, IRG09nov014, STaR/0003/2008, CG/SERI/2010, and NMRC/CSA/033/ 2012 [C.-Y.C]); and the Biomedical Research Council, Singapore, Republic of Singapore (grant nos.: 08/1/35/19/550 and 09/1/35/19/616). The sponsor
Obtained funding: Not applicable
Abbreviations and Acronyms: CCT ¼ central corneal thickness; CH ¼ corneal hysteresis; CI ¼ confidence interval; CRF ¼ corneal resistance factor; IOP ¼ intraocular pressure; ORA ¼ Ocular Response Analyzer; PC ¼ principal component; POAG ¼ primary open-angle glaucoma; SNP ¼ single nucleotide polymorphism. Correspondence: Ching-Yu Cheng, MD, PhD, 20 College Road, The Academia, Level 6, Discovery Tower, Singapore 169856, Republic of Singapore. E-mail: [email protected].
9
Glaucoma is the second leading cause of irreversible blindness worldwide.1,2 Increasing evidence suggests that conventional anatomic risk factors for primary open-angle glaucoma (POAG) do not explain adequately why many people with elevated intraocular pressure (IOP) and thin central corneal thickness (CCT) never demonstrate the disease. Clear evidence for the failure of IOP and CCT to explain the disease is the fact that Chinese persons have a reduced glaucoma-related risk profile (lower IOP and thicker cornea) than Indian persons,3 yet our populationbased data show no reduced POAG prevalence among a Chinese population compared with an Indian population.4,5 If Chinese persons have a reduced glaucoma-related risk profile (lower IOP and thicker cornea) than Indian persons, then something other than IOP and CCT must explain their similarly high glaucoma prevalence as that of Indian persons. Thus, deciphering why certain individuals are susceptible to glaucoma may reside in other properties of the eye, and not simply in the IOP or CCT. ª 2017 by the American Academy of Ophthalmology Published by Elsevier Inc.
Corneal hysteresis (CH) and corneal resistance factor (CRF) are clinical measurements of corneal biomechanics that can be measured from the Ocular Response Analyzer (ORA; Reichert Ophthalmic Instruments, Buffalo, NY).6e8 It has been speculated that corneal biomechanical properties may reflect that of the sclera and optic nerve head. Low CH, representing a less viscoelastic cornea, seems to be an indicator of which eyes will demonstrate POAG.9e11 In light of its relevance to glaucoma, a comparison of the CH and CRF values between Indian and Chinese persons, the 2 largest racial and ethnic groups globally, may provide useful clinical information for glaucoma assessment for different ethnic groups.4,5 However, such direct comparisons using similar examination protocols across ethnic groups have not been performed. The purpose of the current study was to examine variations of CH and CRF between Chinese and Indian persons living in Singapore. We hypothesized that despite having lower IOP and thicker CCT, Chinese people may have POAG prevalence as high as that of Indian persons as a result of lower CH. http://dx.doi.org/10.1016/j.ophtha.2017.03.055 ISSN 0161-6420/17
1
Ophthalmology Volume -, Number -, Month 2017 Methods Study Population Participants in this study were derived from Chinese and Indian persons who participated in a substudy under the Singapore Epidemiology of Eye Disease study. Chinese participants were from the Singapore Chinese Eye Study (2009e2011), whereas Indian persons were from the Singapore Indian Eye Study (2012e2015). In brief, an age-stratified (by 10-year age groups) random sampling in each ethnic group was used to select ethnic Indian and Chinese persons 40 to 80 years of age living across the southwestern part of Singapore during each stipulated study period. Study methodology and details were identical and have been described elsewhere.12 In brief, the Singapore Epidemiology of Eye Disease study is an ongoing, population-based study that was designed to evaluate the prevalence, risk factors, and impact of major eye diseases among Singaporeans 40 to 80 years of age residing in the southwestern part of Singapore. For this substudy, 551 consecutive Indian participants who attended the study clinic during their follow-up examination between April 1 and August 31, 2015, were included. Of the 551 recruited Indian participants, 169 participants were excluded because of a history of intraocular surgery (n ¼ 95) or because of being glaucoma suspects or being diagnosed with glaucoma (n ¼ 74). Finally, 382 Indian participants (69.3%) were included in the analysis. A total of 3353 Chinese persons participated in the study between 2009 and 2011. Among these, 2878 (86.8%) underwent corneal biomechanical measurement, after excluding those with missing corneal biomechanical measurements (n ¼ 43), history of intraocular surgery (n ¼ 341), and glaucoma suspects or those with glaucoma (n ¼ 91). We previously published the corneal biomechanical profile of the Chinese population showing that age and gender were determinants for CH and CRF.13 Therefore, we applied a frequency-matching technique based on age (by 5-year age groups) and gender matching, whereby 2 Chinese participants were selected for each Indian participant. Therefore, 764 ageand gender-matched Chinese participants were selected randomly from those Chinese participants with available corneal biomechanical measurements (n ¼ 2878). Ethics approval was obtained from the Singapore Eye Research Institute Institutional Review Board. All study participants provided written informed consent in adherence to the tenets of the Declaration of Helsinki.
Examination Procedures Both Indian and Chinese participants underwent similar clinical and ocular examinations in the same clinic (Singapore Eye Research Clinic).12 Each participant underwent a standardized interview and eye examination that included assessment of bestcorrected visual acuity, assessment of corneal biomechanical properties by the ORA, autorefraction, and keratometry (corneal curvature in millimeters; Canon RK-5, Autoref-Keratometer Japan]).14 Spherical equivalent was defined [Canon Inc., Ltd., Ota, as sphere plus half cylinder. Ocular biometry, including anterior chamber depth and axial length, was measured using noncontact partial coherence interferometry (IOLMaster; Carl Zeiss Meditec, Dublin, CA), CCT was measured by an ultrasound pachymeter (Advent; Mentor O & O, Inc., Norwell, MA), and IOP was measured by Goldmann applanation tonometry (Haag-Streit, Koenz, Switzerland). Slit-lamp examination of the anterior segment, evaluation of the retina, and dynamic gonioscopy (Sussman 4-mirror gonioscope; Ocular Instruments, Inc., Bellevue, WA) were conducted by the study ophthalmologists.15 Each participant underwent height and weight measurement, and these were used to determine the body mass index, which was
2
calculated as body weight (in kilograms) divided by body height (in meters) squared. Seated blood pressure and blood samples were collected during the examination procedures. Hypertension was defined as systolic blood pressure of 140 mmHg or more, diastolic blood pressure of 90 mmHg or more, physician diagnosed hypertension, or self-reported history of hypertension. Diabetes mellitus was defined as random glucose of 11.1 mmol/l or more, diabetic medication use, or a physician-diagnosed history of diabetes.
Ocular Response Analyzer Details regarding the ORA have been published previously.16 The ORA uses a noncontact rapid air pulse to generate a signal. The ORA signal depicts 2 IOP measurements. The difference between the first and second measurement is the CH and is an indicator of the viscous properties of the cornea. The ORA also provides measurement of CRF, which is an indicator of the overall resistance or elastic properties of the cornea.17 The ORA parameters were measured in the right eye of each eligible participant; when the right eye was ineligible, parameters were measured in the left eye. The measurement was repeated 3 times; differences between the ORA pressure readings of more than 3 mmHg warranted a fourth measurement. Exclusion criteria for this analysis were a history of intraocular surgery or refractive surgery, presence of corneal abnormalities such as keratoconus, corneal scarring that would preclude accurate ORA and IOP measurements, or a diagnosis of glaucoma suspect or glaucoma. Glaucoma suspects were defined according to prespecified criteria,15,18 which included IOP of more than 21 mmHg, gonioscopic findings of occludable angles, presence of peripheral anterior synechiae, cup-to-disc ratio of more than 0.6, disc asymmetry with cup-to-disc ratio difference of more than 0.2, pseudoexfoliation syndrome, pigment dispersion syndrome, and known glaucoma patients. Glaucoma was diagnosed according to the International Society of Geographical and Epidemiologic Ophthalmology classification.18
Ancestry Inference Using Genome-Wide Single Nucleotide Polymorphism Markers In addition to self-reported ethnicity, individual genetic ancestry were inferred using principal component (PC) analysis by the smartPCA program (EIGENSTRAT software version 4.2)19 to examine the relationship of race and its associations to CH and CRF.20 The method implemented in EIGENSTRAT works well if markers are independent21; hence, PC analysis was applied to a reduced single nucleotide polymorphism (SNP) set that had low linkage disequilibrium. The following steps were taken to arrive at a final set of 57 237 SNPs. Autosomal SNPs with minor allele frequency of 0.05 or more within each ethnic group in our study population were selected and used for analysis. Genotype data, which had undergone strict quality checks, were merged together and only SNPs shared by both ethnic groups were used for analysis. In addition, known high linkage disequilibrium regions were excluded (Chr5, 44e51.5 Mb; Chr6, 25e37 Mb; Chr8, 1e12.7 Mb; Chr11, 45e57 Mb; and Chr11, 84e86 Mb).22 Then, SNPs were pruned using the indep-pairwise option in PLINK (version 1.07; available at: http://pngu.mgh.harvard.edu/ wpurcell/plink/; accessed September 1, 2016.), with a window size of 50 SNPs, shifting by 5 SNPs at each step and removing 1 of a pair of SNPs if the linkage disequilibrium was more than 0.2. We then applied PC analysis to genotype data to infer continuous axes of genetic variation. Intuitively, the axes of variation reduce the data to a small number of dimensions,
Chua et al
Corneal Biomechanical Properties among Asians
Table 1. Characteristics of Participants by Ethnicity Characteristics
Indian (n [ 382)
Chinese (n [ 764)
P Value
Age (yrs) Gender (female) Intraocular pressure (mmHg) Central corneal thickness (mm) Corneal curvature (mm) Anterior chamber depth (mm) Refractive error (SE) Axial length (mm) Hypertension Diabetes HbA1c (%) Body mass index (kg/m2) Diastolic blood pressure (mmHg) Systolic blood pressure (mmHg) Corneal hysteresis Corneal resistance factor
60.06.9 204 (53.4) 15.32.5 544.133.5 7.60.2 3.20.3 0.32.0 23.51.0 245 (64.1) 148 (44.6) 6.61.5 26.65.0 77.59.2 137.218.8 10.61.6 10.41.8
59.87.2 408 (53.4) 14.23.1 552.733.3 7.70.3 3.20.3 0.42.6 23.91.4 450 (58.9) 102 (14.1) 6.11.0 23.63.5 77.69.6 136.019.1 10.61.6 10.21.7
0.804 1.000 <0.001 <0.001 0.144 0.848 <0.001 <0.001 0.087 <0.001 <0.001 <0.001 0.867 0.319 0.670 0.103
HbA1c ¼ glycosylated hemoglobin; SD ¼ standard deviation; SE ¼ spherical equivalence. Data are mean SD or number (%), as appropriate.
describing as much variability as possible. The first 2 PCs captured 5.45% of variation, whereas the third PC captured only 0.09% of variation.
Table 2. Mean (Standard Deviation) of Corneal Hysteresis and Corneal Resistance Factor by Age Group, Gender, and Ethnicity Indian (n [ 382) Corneal hysteresis Age group (yrs) 50e54 55e59 60e64 65e69 70e74 75 P value for trend Gender Male Female P value Corneal resistance factor Age group (yrs) 50e54 55e59 60e64 65e69 70e74 75 P value for trend Gender Male Female P value
Chinese (n [ 764)
No.
Mean SD
No.
Mean SD
104 108 72 61 22 15
11.11.4 10.71.5 10.41.5 10.01.7 10.71.6 10.01.1 <0.001*
208 216 144 122 44 30
11.01.7 10.71.6 10.71.5 10.21.4 10.11.9 9.71.5 <0.001*
178 204
10.61.6 10.61.6 0.957y
356 408
10.31.5 10.91.6 0.001z
104 108 72 61 22 15
10.91.8 10.41.7 10.11.6 9.91.9 10.62.0 9.80.8 0.002*
208 216 144 122 44 30
10.51.9 10.31.6 10.21.5 10.01.6 9.91.8 9.21.8 <0.001*
178 204
10.31.8 10.51.7 0.268z
356 408
9.91.6 10.51.7 <0.001z
SD ¼ standard deviation. *Test for differences in trend between age groups. y Test for differences between gender, based on independent t test. z Test for differences between gender, based on Kruskal-Wallis test.
Statistical Analysis The mean values of CH and CRF and their ocular and systemic associations were studied. The ShapiroeWilk test was used to assess the normality of the distribution of the corneal biomechanical parameters. To compare continuous variables among groups, an independent t test was performed for normally distributed variables, whereas the KruskaleWallis test was used for nonnormally variables. The chi-square test was used for categorical variables. The strengths of the linear relationship between variables were assessed by Pearson’s correlation coefficients. Associations between ocular and systemic factors with CH and CRF were assessed using univariate and multivariate linear regression models. In the fully adjusted multivariate model, the differences in CH and CRF between Indian and Chinese persons were adjusted for age, gender, ethnicity, IOP, CCT, corneal curvature, axial length, and factors with P < 0.05 in the univariate model simultaneously. In the reduced adjusted multivariate model, we adjusted only for age, gender, ethnicity, and individual ocular parameter separately, based on the results from the univariate model. Furthermore, genetic ancestry inferred from SNPs was used in place of self-reported ethnicity to determine its association with CH and CRF. Analyses were performed using STATA software version 13 (StataCorp, College Station, TX).
Results Two Chinese participants were matched for each of the 382 Indian participants included in the analysis, based on age (by 5-year age groups) and gender (n ¼ 764). Table 1 compares the mean values of the ocular and systemic variables between the Indian and Chinese populations. The mean ages standard deviation of the participants were 60.06.9 years for Indian persons and 59.87.2 years for Chinese persons. Of these participants, 53.4% were women. Indian persons were more likely to have a higher IOP, thinner CCT, more hyperopic refractive error, shorter axial length, higher prevalence of diabetes, higher hemoglobin A1c levels, and greater body mass index than Chinese persons (P < 0.001). There were no differences in the overall CH (10.61.6 mmHg; P ¼ 0.670) or the CRF (10.31.7 mmHg; P ¼ 0.103) values between ethnic groups (Table 1).
3
4 Table 3. Association with Corneal Hysteresis by Ethnicity Assessed Using Multivariate Linear Regression Models All 0.43 0.32 0.13 0.19 0.81 0.18 0.01 0.11 0.03 0.26 0.09 0.03 0.03 0.01
(0.56 to 0.31)z (0.14e0.50)z (0.18 to 0.44) (0.16e0.21)z (1.16 to 0.46)z (0.45 to 0.10) (0.04 to 0.03) (0.19 to 0.04)x (0.23 to 0.16) (0.04 to 0.51) (0.01e0.17)x (0.05 to 0.01)x (0.13 to 0.07) (0.05 to 0.05)
Reference 0.01 (0.19 to 0.19)
(0.55 to 0.29)z (0.05e0.41)x (0.56 to 0.06) (0.17e0.22)z (1.21 to 0.47)z d d 0.06 (0.13 to 0.02) d d 0.07 (0.01 to 0.15) 0.03 (0.05 to 0.01)x d d 0.42 0.23 0.25 0.19 0.84
Reference 0.18 (0.02e0.38)x
0.47 0.10 0.53 0.17 0.37 0.03 0.05 0.11 0.18 0.20 0.06 0.04 0.13 0.03
(0.69 to 0.24)z (0.41 to 0.21) (0.09 to 1.14) (0.12e0.21)z (1.01 to 0.26) (0.50 to 0.44) (0.03 to 0.13) (0.27 to 0.04) (0.15 to 0.51) (0.14 to 0.55) (0.05 to 0.17) (0.07 to 0.01)x (0.04 to 0.03) (0.05 to 0.011) d d
Chinese Persons Model 2y
Model 1*
(0.68 to 0.22)z (0.42 to 0.24) (0.44 to 0.84) (0.12e0.22)z (1.30 to 0.14) d d 0.04 (0.22 to 0.13) d d 0.03 (0.07 to 0.13) 0.03 (0.06 to 0.01) d d 0.45 0.09 0.20 0.17 0.58
d d
Model 2y
Model 1* 0.42 0.53 0.01 0.20 1.00 0.24 0.02 0.11 0.12 0.27 0.13 0.01 0.09 0.02
(0.57 to 0.27)z (0.31e0.75)z (0.36 to 0.36) (0.16e0.22)z (1.42 to 0.59)z (0.58 to 0.11) (0.06 to 0.03) (0.20 to 0.03)x (0.37 to 0.12) (0.07 to 0.61) (0.02e0.25)x (0.05 to 0.02) (0.21 to 0.03) (0.08 to 0.05) d d
(0.56 to 0.26)z (0.17e0.61)z (0.79 to 0.08)x (0.17e0.23)z (1.41 to 0.54)z d d 0.06 (0.14 to 0.02) d d 0.13 (0.02e0.24)x 0.03 (0.06 to 0.01) d d 0.41 0.39 0.43 0.20 0.97
d d
HbA1c ¼ glycosylated hemoglobin; SE ¼ spherical equivalent; d ¼ data were intentionally omitted. Data are b (95% confidence interval). *Model 1 adjusted for age and gender and additionally adjusted for ethnicity in the analysis of the 2 ethnic groups combined. y Model 2 adjusted for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, HbA1c, body mass index, and additionally adjusted for ethnicity in the analysis of the 2 ethnic groups combined. z P < 0.001. x P < 0.05.
Ophthalmology Volume -, Number -, Month 2017
Age (per 10 yrs) Gender (female) Intraocular pressure (per 10 mmHg) Central corneal thickness (per 10 mm) Corneal curvature (mm) Anterior chamber depth (mm) Refractive error (SE) Axial length (mm) Hypertension Diabetes HbA1c (%) Body mass index (kg/m2) Diastolic blood pressure (per 10 mmHg) Systolic blood pressure (per 10 mmHg) Self-reported ethnicity Chinese Indian
Indian Persons Model 2y
Model 1*
Characteristics
Chua et al
Corneal Biomechanical Properties among Asians
Figure 1. The distribution of (A) corneal hysteresis and (B) corneal resistance factor between Indian and Chinese participants. *Data and P value shown for corneal hysteresis are after adjustment for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, body mass index, hemoglobin A1c, and ethnicity. yData and P value shown for corneal resistance factor are after adjustment for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, diabetes, systolic blood pressure, and ethnicity.
Corneal Hysteresis and Corneal Resistance Factor with Age and Gender Table 2 shows the variation of CH and CRF with age and gender. Negative correlations between age and CH were seen for both ethnic groups (P < 0.001 for trend; Table 2). A similar association was noted between age and CRF for Indian persons (P ¼ 0.002 for trend) as well as Chinese persons (P < 0.001 for trend; Table 2). There were no gender differences in the CH (P ¼ 0.957) or CRF (P ¼ 0.268) values for Indian persons (Table 2). However, Chinese women had higher mean CH (10.9 vs. 10.3 mmHg; P < 0.001) and CRF (10.5 vs. 9.9 mmHg; P < 0.001) than Chinese men (Table 2).
Ocular and Systemic Factors with Corneal Hysteresis The association between CH and ocular and systemic factors by ethnic groups is summarized in Table 3. Multivariate linear regression analysis showed that younger age (b ¼ 0.42 per 10 years; 95% confidence interval [CI], 0.55 to 0.29; P < 0.001), female gender (b ¼ 0.23; 95% CI, 0.05e0.41; P ¼ 0.011), thicker CCT (b ¼ 0.19 per 10 mm; 95% CI, 0.17e0.22; P < 0.001), steeper corneal curvature (b ¼ 0.84; 95% CI, 1.21 to 0.47; P < 0.001), lesser body mass index levels (b ¼ 0.03; 95% CI, 0.05 to 0.01; P ¼ 0.009), and Indian ethnicity compared with Chinese ethnicity (b ¼ 0.18; 95% CI, 0.02e0.38; P ¼ 0.031; Fig 1A) were associated independently with higher CH levels (model 2 in Table 3). Further analysis stratified by ethnic group showed that the higher CH levels with female gender, steeper corneal curvature, and higher hemoglobin A1c levels were seen for Chinese persons, but not for Indian persons. We used another multivariate regression model, adjusting only for age, gender, ethnicity, and individual ocular parameters separately, to identify which ocular factor was the negative confounder that accounted for the observed null association between ethnicity and corneal hysteresis in the univariate analysis (model 1 in Table 3). The CH levels in Indian persons were, on average,
0.25 mmHg higher than those in Chinese persons, after controlling for CCT (95% CI, 0.06e0.43; P ¼ 0.010). There was a positive correlation between CCT and CH for Indian persons (r ¼ 0.38; P < 0.001) and Chinese persons (r ¼ 0.54; P < 0.001), with no statistically significant difference between the 2 groups (P ¼ 0.349). The CH levels remained similar between Indian and Chinese persons after adjusting for corneal curvature and axial length (all P > 0.05).
Ocular and Systemic Factors with Corneal Resistance Factor The association of CRF with ocular and systemic factors by ethnic groups is shown in Table 4. Multivariate linear regression analysis showed that younger age per 10 years (b ¼ 0.34; 95% CI, 0.47 to 0.20; P < 0.001), female gender (b ¼ 0.28; 95% CI, 0.10e0.45; P ¼ 0.002), higher IOP per 10 mmHg (b ¼ 1.57; 95% CI, 1.27e1.87; P < 0.001), thicker CCT per 10 mm (b ¼ 0.24; 95% CI, 0.21e0.26; P < 0.001), steeper corneal curvature (b ¼ 0.76; 95% CI, 1.13 to 0.40; P < 0.001), and presence of diabetes (b ¼ 0.28; 95% CI, 0.07e0.49; P ¼ 0.010) were associated independently with higher CRF levels (model 2 in Table 5). The CRF levels in Indian persons were similar to those in Chinese persons in both the full model (Table 5; b ¼ 0.09; 95% CI, 0.10 to 0.29; P ¼ 0.355; Fig 1B) and reduced model (P > 0.05). Further analysis stratified by ethnic group showed that the higher CRF levels with female gender, steeper corneal curvature, and presence of diabetes were seen for Chinese persons, but not for Indian persons.
Genetic Ancestry and Corneal Biomechanical Parameters The Indian and Chinese ancestries were clearly distinguishable by the first PC of genetic ancestry (Fig 2).3 Briefly, the first and second PCs sequentially (cumulatively) accounted for 5.26% and 0.19% (5.45%) of the total genetic variation. Each standard deviation increase in the first PC (standard deviation, 0.01) was associated significantly with an 11.72-mmHg lower CH level
5
6 Table 4. Association with Corneal Resistance Factor by Ethnicity Assessed Using Multivariate Linear Regression Models All
Age (per 10 yrs) Gender (female) Intraocular pressure (per 10 mmHg) Central corneal thickness (per 10 mm) Corneal curvature (mm) Anterior chamber depth (mm) Refractive error (SE) Axial length (mm) Hypertension Diabetes HbA1c (%) Body mass index (kg/m2) Diastolic blood pressure (per 10 mmHg) Systolic blood pressure (per 10 mmHg) Self-reported ethnicity Chinese Indian
0.35 0.40 2.15 0.26 0.75 0.13 0.03 0.06 0.15 0.51 0.13 0.01 0.16 0.11
Indian Persons Model 2y
(0.49 to 0.22)z 0.34 (0.47 to 0.20)z (0.20e0.60)z 0.28 (0.10 to 0.45)x z (1.83e2.46) 1.57 (1.27 to 1.87)z (0.24e0.29)z 0.24 (0.21 to 0.26)z (1.12 to 0.35)z 0.76 (1.13 to 0.40)z (0.43 to 0.18) d (0.07 to 0.10) d (0.14 to 0.20) 0.01 (0.08 to 0.06) (0.06 to 0.36) d 0.28 (0.07 to 0.49)x (0.25e0.77)z (0.04e0.22)x d (0.03 to 0.02) d x d (0.05e0.27) 0.04 (0.01 to 0.09) (0.05e0.16)z
Reference 0.17 (0.04 to 0.38)
Reference 0.09 (0.10 to 0.29)
Model 1* 0.41 0.10 2.65 0.25 0.15 0.06 0.04 0.08 0.30 0.46 0.08 0.01 0.31 0.13
Chinese Persons Model 2y
(0.67 to 0.16)x 0.38 (0.62 to 0.13)x (0.26 to 0.45) 0.03 (0.31 to 0.37) (2.00e3.30)z 2.17 (1.51e2.84)z (0.20e0.29)z 0.21 (0.16e0.26)z (0.88 to 0.58) 0.46 (1.22 to 0.30) (0.48 to 0.59) d (0.05 to 0.13) d (0.25 to 0.10) 0.02 (0.19 to 0.16) (0.08 to 0.67) e (0.07e0.85)x 0.27 (0.06 to 0.60) (0.04 to 0.21) (0.05 to 0.02) d (0.12e0.51)x d (0.04e0.23)z 0.02 (0.07 to 0.11) d d
d d
Model 1* 0.33 0.55 1.98 0.27 0.99 0.21 0.05 0.05 0.09 0.57 0.18 0.01 0.10 0.10
Model 2y
(0.49 to 0.16)z 0.33 (0.48 to 0.17)z (0.31e0.79)z 0.40 (0.19e0.60)z z (1.62e2.34) 1.36 (1.03e1.69)z (0.24e0.30)z 0.24 (0.21e0.27)z (1.44 to 0.55)z 0.90 (1.31 to 0.49)z (0.57 to 0.16) d d (0.10 to 0.01)x (0.14 to 0.03) 0.01 (0.08 to 0.73) (0.18 to 0.35) d (0.22e0.92)x 0.32 (0.04e0.61)x (0.06e0.31)x (0.03 to 0.02) d (0.03 to 0.23) d (0.03e0.16)x 0.06 (0.01e0.11)x d d
d d
HbA1c ¼ glycosylated hemoglobin; SE ¼ spherical equivalent; d ¼ data were intentionally omitted. Data are b (95% confidence interval). *Model 1 adjusted for age and gender, and additionally adjusted for ethnicity in the analysis of the 2 ethnic groups combined. y Model 2 adjusted for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, diabetes, and systolic blood pressure and additionally adjusted for ethnicity in the analysis of the 2 ethnic groups combined. z P < 0.001. x P < 0.05.
Ophthalmology Volume -, Number -, Month 2017
Model 1*
Characteristics
Chua et al
Corneal Biomechanical Properties among Asians
Table 5. Associations of Genetic Ancestry with Corneal Hysteresis and Corneal Resistance Factor Corneal Hysteresis (mmHg) Model 1* Ancestry PC 1 PC 2
Corneal Resistance Factor (mmHg) y
Model 2
Model 3z
Model 1*
b (95% Confidence Interval)
P Value
b (95% Confidence Interval)
P Value
b (95% Confidence Interval)
P Value
b (95% Confidence Interval)
P Value
1.14 (9.16 to 6.88) 5.27 (15.34 to 4.79)
0.780 0.304
11.72 (19.74 to 3.70) 0.43 (9.51 to 8.65)
0.004 0.930
7.40 (15.92 to 1.13) 4.83 (15.54 to 5.88)
0.089 0.377
7.62 (15.66 to 0.37) 3.04 (6.24 to 12.32)
0.062 0.520
PC ¼ principal component. b represents the change in corneal hysteresis or corneal resistance factor for each standard deviation unit increase in PCs. *Model 1 adjusted for age and gender. y Model 2 adjusted for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, hemoglobin A1c, and body mass index. z Model 3 adjusted for age, gender, intraocular pressure, central corneal thickness, corneal curvature, axial length, diabetes, and systolic blood pressure.
(95% CI, 19.74 to 3.70; P ¼ 0.004), after adjusting for significant factors in the multivariate regression model (model 2 in Table 5), as well as in a reduced multivariate model adjusting only for age, gender, and CCT (b ¼ 8.53; 95% CI, 15.94 to 1.11; P ¼ 0.024). These analyses of genetic PC, which was estimated on the basis of genome-wide genotypes, consistently indicated that Chinese ancestry was associated with lower levels of CH than Indian ancestry. That is, there was a significant trend of decreasing CH as the percentage of Chinese ancestry increased. Genetic ancestry was not associated significantly with CRF, both before and after adjustment for relevant confounders (all P > 0.05; models 1 and 3 in Table 5).
Discussion In this population-based study in Singapore, after adjusting for ocular biometric parameters, namely IOP, CCT, corneal curvature, and axial length, Chinese persons had significantly lower CH and similar CRF values when compared with Indian persons. Comparing the differences in CH value between Asian subgroups may be relevant when
Figure 2. The principal components of genetic ancestry in the Singapore Indian Eye Study and Singapore Chinese Eye Study: principal component 1 (PC1) versus principal component 2 (PC2) by self-reported ethnicity. There are 2 distinct clusters, corresponding to Indian (blue) and Chinese (red) study participants. First principal component (PC1) perfectly distinguishes these 2 populations, and PC2 is able to capture the ancestry variations within the Indian group.
investigating the role of altered corneal biomechanics in ocular diseases such as glaucoma and keratoconus. We also assessed the ethnic differences of CH and CRF on the basis of genetic ancestry using the same methodology to collect data between ethnic groups. Comparing the mean CH among different populationbased studies would help to clarify the interethnic variation of corneal biomechanical properties. The clinical measurement of CH in our study sample was 10.61.6 mmHg, which was comparable between Chinese and Indian persons and was similar to that of healthy Japanese23 and Brazilian24 participants. However, interethnic variations in ocular parameters may account for the similarities observed among these studies. To compare the association between CH and ethnicity more accurately, we incorporated the effect of relevant ocular parameters, namely CCT, corneal curvature, and axial length, and computed the adjusted mean values of CH between the 2 ethnic groups (Table 3; Fig 1A). We found that CH is 0.18 mmHg lower for Chinese adults 50 years of age and older compared with similarly aged Indian persons, after adjusting for all other variables in the multiple linear regression model. This suggests that similarities in the measured CH between Chinese and Indian persons are explained largely by the differences in ocular biometry parameters, specifically CCT found between these 2 groups, signifying a distinct biomechanical difference between their corneas. Ocular parameters may influence clinical measurements of the ORA. We found that eyes with thicker cornea, steeper cornea, and shorter axial length exhibit higher CH levels. Similarly, other studies have shown that the ORA is influenced by ocular parameters such as CCT,24e28 corneal curvature,13 and axial length.13,29,30 Recently, the UK Biobank study showed that diabetics essentially have similar IOP measurements as nondiabetics after compensating for their corneal biomechanics.31 Leite et al27 further demonstrated that the interethnic differences of clinical CH measurement between black and white Americans were nullified because CH was correlated significantly with CCT. Similarly, we found that CCT was the major ocular parameter that influenced the hysteresis of the tissue, adding another layer of complexity to determining the socalled true rigidity of the cornea. The variety of errors encountered during clinical measurements is similar to the
7
Ophthalmology Volume -, Number -, Month 2017 analogy for IOP measurement using tonometry; the clinical parameter of CH obtained from the ORA in relation to measuring the cornea’s viscoelasticity (varying stiffness with rate of loading) clearly falls in this domain. The modest difference in CH suggests that there could be other factors that may explain the ethnic differences in glaucoma prevalence and individual susceptibility to disease. Unfortunately, our understanding of these ocular parameters that result in CH measurement errors or bias is far from complete. The cornea itself is a relatively complex tissue that exhibits various mechanical characteristics such as viscoelasticity, anisotropy, nonlinearity, and heterogeneity. The ORA is the first commercially available technology that allowed measurement of some biomechanical properties of the cornea. Improvements in technology and subsequent availability of newer instruments such as the Corneal Visualization Scheimpflug Technology may help to characterize better the complex biomechanical characteristics of the cornea in vivo. Study Strengths and Limitations Self-reported race and ethnicity are used commonly in health research to assess an individual’s background origin. The use of self-reported race and ethnicity in research studies can be particularly challenging for individuals who may not be fully aware of their own complex ancestry mixture. In the current study, using objective and accurate estimates of genetic ancestry, we further showed true genetic ancestry-based differences in CH between the 2 Asian subgroups. Other strengths of the study include its population-based, relatively large sample size and the availability of 2 major Asian ethnic groups with similar standardized study methodology. For example, we have identical methods of measurement for ocular parameters, such as ORA, autokeratometer, IOLMaster, ultrasound pachymeter, and Goldmann applanation tonometry, in each ethnic group, which enabled the direct and robust comparison of data. However, our study has some limitations. Subclinical corneal cases may be overlooked because our participants underwent only slitlamp examination, which rules out corneal edema or corneal dystrophy, and did not undergo assessment using corneal topographers such as Orbscan (Bausch & Lomb, Rochester, NY) or Pentacam (Oculus, Inc., Arlington, WA) By including only healthy participants, we were not able to evaluate differences in corneal biomechanics between races in glaucomatous patients. Future longitudinal studies are needed to elucidate the relationship between CH and susceptibility to glaucoma developing in Chinese and Indian persons. In conclusion, our population-based study of Indian and Chinese adults 50 years of age or older in Singapore confirms the presence of ethnic variation in CH, but not CRF, between Asian subgroups. Overall, after adjusting for ocular parameters, Chinese participants have a lower CH when compared with Indian persons. In view that corneal hysteresis is lower in eyes with POAG, this may contribute to the POAG prevalence in Chinese compared with Indian persons.
8
References 1. Bourne RR, Stevens GA, White RA, et al. Causes of vision loss worldwide, 1990e2010: a systematic analysis. Lancet Glob Health. 2013;1:e339-e349. 2. Tham YC, Li X, Wong TY, et al. Global prevalence of glaucoma and projections of glaucoma burden through 2040: a systematic review and meta-analysis. Ophthalmology. 2014;121:2081-2090. 3. Chua J, Tham YC, Liao J, et al. Ethnic differences of intraocular pressure and central corneal thickness: the Singapore Epidemiology of Eye Diseases study. Ophthalmology. 2014;121:2013-2022. 4. Baskaran M, Foo RC, Cheng CY, et al. The prevalence and types of glaucoma in an urban Chinese population: the Singapore Chinese Eye Study. JAMA Ophthalmol. 2015;133: 874-880. 5. Narayanaswamy A, Baskaran M, Zheng Y, et al. The prevalence and types of glaucoma in an urban Indian population: the Singapore Indian Eye Study. Invest Ophthalmol Vis Sci. 2013;54:4621-4627. 6. Deol M, Taylor DA, Radcliffe NM. Corneal hysteresis and its relevance to glaucoma. Curr Opin Ophthalmol. 2015;26:96102. 7. Narayanaswamy A, Su DH, Baskaran M, et al. Comparison of ocular response analyzer parameters in chinese subjects with primary angle-closure and primary open-angle glaucoma. Arch Ophthalmol. 2011;129:429-434. 8. Grise-Dulac A, Saad A, Abitbol O, et al. Assessment of corneal biomechanical properties in normal tension glaucoma and comparison with open-angle glaucoma, ocular hypertension, and normal eyes. J Glaucoma. 2012;21:486-489. 9. Congdon NG, Broman AT, Bandeen-Roche K, et al. Central corneal thickness and corneal hysteresis associated with glaucoma damage. Am J Ophthalmol. 2006;141:868-875. 10. Medeiros FA, Meira-Freitas D, Lisboa R, et al. Corneal hysteresis as a risk factor for glaucoma progression: a prospective longitudinal study. Ophthalmology. 2013;120:1533-1540. 11. De Moraes CV, Hill V, Tello C, et al. Lower corneal hysteresis is associated with more rapid glaucomatous visual field progression. J Glaucoma. 2012;21:209-213. 12. Lavanya R, Jeganathan VS, Zheng Y, et al. Methodology of the Singapore Indian Chinese Cohort (SICC) eye study: quantifying ethnic variations in the epidemiology of eye diseases in Asians. Ophthalmic Epidemiol. 2009;16:325-336. 13. Narayanaswamy A, Chung RS, Wu RY, et al. Determinants of corneal biomechanical properties in an adult Chinese population. Ophthalmology. 2011;118:1253-1259. 14. Chua J, Lim B, Fenwick EK, et al. Prevalence, risk factors, and impact of undiagnosed visually significant cataract: the Singapore Epidemiology of Eye Diseases Study. PLoS One. 2017;12:e0170804. 15. Chua J, Baskaran M, Ong PG, et al. Prevalence, risk factors, and visual features of undiagnosed glaucoma: the Singapore Epidemiology of Eye Diseases Study. JAMA Ophthalmol. 2015;133(8):938-946. 16. Luce DA. Determining in vivo biomechanical properties of the cornea with an ocular response analyzer. J Cataract Refract Surg. 2005;31:156-162. 17. Kotecha A. What biomechanical properties of the cornea are relevant for the clinician? Surv Ophthalmol. 2007;52(Suppl 2): S109-S114. 18. Foster PJ, Buhrmann R, Quigley HA, et al. The definition and classification of glaucoma in prevalence surveys. Br J Ophthalmol. 2002;86:238-242.
Chua et al
Corneal Biomechanical Properties among Asians
19. Price AL, Patterson NJ, Plenge RM, et al. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 2006;38:904-909. 20. Chua J, Koh JY, Tan AG, et al. Ancestry, socioeconomic status, and age-related cataract in Asians: the Singapore Epidemiology of Eye Diseases Study. Ophthalmology. 2015;122:2169-2178. 21. Patterson N, Price AL, Reich D. Population structure and eigenanalysis. PLoS Genet. 2006;2:e190. 22. Chen J, Zheng H, Bei JX, et al. Genetic structure of the Han Chinese population revealed by genome-wide SNP variation. Am J Hum Genet. 2009;85:775-785. 23. Kamiya K, Hagishima M, Fujimura F, et al. Factors affecting corneal hysteresis in normal eyes. Graefes Arch Clin Exp Ophthalmol. 2008;246:1491-1494. 24. Fontes BM, Ambrosio Jr R, Alonso RS, et al. Corneal biomechanical metrics in eyes with refraction of -19.00 to þ9.00 D in healthy Brazilian patients. J Refract Surg. 2008;24:941-945. 25. Kotecha A, Elsheikh A, Roberts CR, et al. Corneal thicknessand age-related biomechanical properties of the cornea measured with the ocular response analyzer. Invest Ophthalmol Vis Sci. 2006;47:5337-5347.
26. Shah S, Laiquzzaman M, Cunliffe I, et al. The use of the Reichert ocular response analyser to establish the relationship between ocular hysteresis, corneal resistance factor and central corneal thickness in normal eyes. Cont Lens Anterior Eye. 2006;29:257-262. 27. Leite MT, Alencar LM, Gore C, et al. Comparison of corneal biomechanical properties between healthy blacks and whites using the Ocular Response Analyzer. Am J Ophthalmol. 2010;150:163-168.e1. 28. Mangouritsas G, Morphis G, Mourtzoukos S, et al. Association between corneal hysteresis and central corneal thickness in glaucomatous and non-glaucomatous eyes. Acta Ophthalmol. 2009;87:901-905. 29. Shen M, Fan F, Xue A, et al. Biomechanical properties of the cornea in high myopia. Vision Res. 2008;48:2167-2171. 30. Chang PY, Chang SW, Wang JY. Assessment of corneal biomechanical properties and intraocular pressure with the Ocular Response Analyzer in childhood myopia. Br J Ophthalmol. 2010;94:877-881. 31. Chan MP, Grossi CM, Khawaja AP, et al. Associations with intraocular pressure in a large cohort: results from the UK Biobank. Ophthalmology. 2016;123:771-782.
Footnotes and Financial Disclosures Originally received: December 17, 2016. Final revision: March 29, 2017. Accepted: March 29, 2017. Available online: ---.
or funding organization had no role in the design or conduct of this research. Author Contributions: Manuscript no. 2016-1066.
1
Singapore Eye Research Institute, Singapore National Eye Centre, Singapore, Republic of Singapore.
2
Conception and design: Chua, Tham, Gupta, Sabanayagam, Aung, Wong, Cheng Analysis and interpretation: Chua, Zhao, Sabanayagam, Cheng
Ophthalmology and Visual Sciences Academic Clinical Program, DukeNUS Medical School, Singapore, Republic of Singapore.
Data collection: Chua, Tham, Gupta
3
Overall responsibility: Chua, Nongpiur, Sabanayagam, Aung, Wong, Cheng
Department of Ophthalmology, Yong Loo Lin School of Medicine, National University of Singapore and National University Health System, Singapore, Republic of Singapore. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Supported by the SingHealth Foundation, Singapore, Republic of Singapore (grant no.: SHF/FG563P/2014); the National Medical Research Council, Singapore, Republic of Singapore (grant nos.: 0796/2003, IRG07nov013, IRG09nov014, STaR/0003/2008, CG/SERI/2010, and NMRC/CSA/033/ 2012 [C.-Y.C]); and the Biomedical Research Council, Singapore, Republic of Singapore (grant nos.: 08/1/35/19/550 and 09/1/35/19/616). The sponsor
Obtained funding: Not applicable
Abbreviations and Acronyms: CCT ¼ central corneal thickness; CH ¼ corneal hysteresis; CI ¼ confidence interval; CRF ¼ corneal resistance factor; IOP ¼ intraocular pressure; ORA ¼ Ocular Response Analyzer; PC ¼ principal component; POAG ¼ primary open-angle glaucoma; SNP ¼ single nucleotide polymorphism. Correspondence: Ching-Yu Cheng, MD, PhD, 20 College Road, The Academia, Level 6, Discovery Tower, Singapore 169856, Republic of Singapore. E-mail: [email protected].
9