Cornea
Biometry of the Cornea and Anterior Chamber in
Chinese Eyes: An Anterior Segment Optical Coherence
Tomography Study
Leonard H. Yuen,1 Mingguang He,2 Tin Aung,1 Hla M. Htoon,3 Donald T. Tan,1,4 and
Jodhbir S. Mehta1
PURPOSE. To investigate the normative data of corneal and
anterior segment biometric parameters and their associations
in Chinese adults, for use in preoperative assessment for cor-
neal and anterior segment surgery.
METHODS. This cross-sectional, population-based study in-
cluded 750 subjects aged ⱖ50 years. The subjects underwent
an ophthalmic examination including imaging with anterior
segment optical coherence tomography (AS-OCT). Dimensions
were subsequently measured with the Zhongshan Assessment
Program (ZAP). Also measured was posterior corneal arc
length (PCAL), a novel parameter defined as the arc distance
between scleral spurs on the posterior border of the cornea.
Correlations with age, sex, height, weight, body mass index
(BMI), refractive sphere and cylinder, and intraocular pressure
were also measured.
RESULTS. The subjects’ mean age was 63.3 ⫾ 7.9 years and 349
(46.5%) were men. Corneal parameters (mean ⫾ SD) included
PCAL (12.924 ⫾ 0.544 mm), anterior chamber depth (ACD;
2.684 ⫾ 0.309 mm), central corneal thickness (562.39 ⫾ 31.85
␮m), anterior chamber curvature (7.35 ⫾ 0.37 mm), and pos-
terior corneal curvature (6.65 ⫾ 0.34 mm). A moderate corre-
lation was observed between PCAL and ACD (R ⫽ 0.55, P ⬍
0.001) and a poor correlation between PCAL and age, height,
weight, and BMI. Multivariate analysis showed a significant
association between PCAL and ACD, ACC, PCC, and cylinder
(P ⬍ 0.05).
CONCLUSIONS. In this Chinese population, PCAL was found to
correlate moderately with ACD. The data may be useful for
corneal and anterior segment procedures such as Descemet’s
stripping automated endothelial keratoplasty (DSAEK). (Invest
Ophthalmol Vis Sci. 2010;51:3433–3440) DOI:10.1167/iovs.09-
4307
A ccurate qualitative and quantitative information on the
morphology and anatomy of the cornea and anterior seg-
ment enables the surgeon to make informed preoperative de-
cisions in anterior segment surgery. Qualitative information
From the 1Singapore National Eye Centre (SNEC), Singapore; the
2
Zhongshan Ophthalmic Centre, Guangzhou, China; the 3Singapore
Eye Research Institute (SERI), Singapore; and the 4Yong Loo Lin School
of Medicine, National University of Singapore, Singapore.
Supported by National Research Foundation Translational and
Clinical Research Programme Grant TCR R621/42/2008.
Submitted for publication July 13, 2009; revised December 9,
2009, and January 6, 2010; accepted January 24, 2010.
Disclosure: L.H. Yuen, None; M. He, None; T. Aung, Carl Zeiss
Meditec (F); H.M. Htoon, None; D.T. Tan, None; J.S. Mehta, None
Corresponding author: Jodhbir S. Mehta, Singapore National Eye
Centre, 11 Third Hospital Avenue, Singapore 168751;
jodmehta@gmail.com.
Investigative Ophthalmology & Visual Science, July 2010, Vol. 51, No. 7
Copyright © Association for Research in Vision and Ophthalmology
Downloaded From: https://iovs.arvojournals.org/ on 09/09/2018
can affect surgical outcome, for example, in selecting the
appropriate-sized phakic intraocular lens (pIOL) or posterior
chamber intraocular lens (PCIOL),1,2 the sizing of corneal im-
plant rings (Intacs; Addition Technology, Des Plaines, IL) to
correct keratoconus, the assessment of LASIK flap thickness
after keratorefractive surgery,3 and the optimization of graft
sizes for Descemet’s stripping automated endothelial kerato-
plasty (DSAEK) surgery. Such information provides enhanced
safety in surgery and reduced postoperative complications.
The development of the anterior segment optical coher-
ence tomography (AS-OCT, Visante; Carl Zeiss Meditec, Dub-
lin, CA) has allowed a fast, noncontact method of imaging the
cornea and anterior chamber4 – 6 for objective quantification
and evaluation.7 With a single anterior segment image, the
cornea, both angles (in one meridian including the scleral
spur5,6), the anterior portion of the lens, and the iris surface are
imaged. It has been shown to be highly reproducible with
respect to repeatable image acquisition and thus has excellent
interobserver and intraobserver variability.8
A novel algorithm software that uses AS-OCT images, the
Zhongshan Assessment Program (ZAP), offers a simple, user-
friendly quantification of anterior segment anatomy9 with high
inter- and intraobserver agreement and repeatability.9,10 A
novel parameter that is measured and evaluated in this study is
the posterior corneal arc length (PCAL), defined as the arc
distance of the posterior corneal border between scleral spurs.
Imaging of the anterior corneal surface has been described
extensively in the literature; however, accurate data on the
shape of the posterior surface of the cornea are limited,11
perhaps due to the complexities of imaging this surface. To
obtain the images, acquisition must be performed through the
anterior surface of the cornea, which can act as a magnifying
lens and distort the perceived shape of the posterior cornea.
This study serves to describe the corneal and anterior seg-
ment dimensions of an ethnic population in Singapore and to
document for the first time PCAL data in this population,
which may be a predictive parameter for anterior chamber
surgery. The information from the study will form baseline
normative data for the assessment of any surgical intervention
in this population group.
METHODS
Study Population and Recruitment
The study was approved by the Institutional Review Board of the
Singapore Eye Research Institute and was performed in accordance
with the tenets of the Declaration of Helsinki. In this population-based,
cross-sectional study, subjects older than 50 years who did not have
any ophthalmic complaints were recruited from a government-run
community polyclinic providing primary health care services. This
polyclinic serves more than 10,000 people per month, mainly of lower
3433
 3434 Yuen et al.
FIGURE 1. Appearance of the ZAP software using AS-OCT images: for
each image, the image file is opened, and the two scleral spurs are
identified and marked. Calculate OCT then measures all parameters
including the PCAL (in millimeters), as shown by the arrow in the
right image.
to middle socioeconomic status, with a high proportion requiring
chronic disease management. The subjects were identified by system-
atic sampling (every fifth patient registered at the polyclinic) and asked
to participate in the study after they provided written informed con-
sent. A detailed questionnaire including demographic, socioeconomic
details, and medical and ocular history was administered. The exclu-
sion criteria were a history of glaucoma, prior intraocular surgery or
penetrating eye injury, and corneal disorders, such as corneal dystro-
phy, corneal opacity, or pterygium that would prevent anterior cham-
ber depth (ACD) measurement.
Imaging
Images were collected by using AS-OCT at the Singapore National Eye
Centre after informed consent was obtained from all participants. The
details of the AS-OCT imaging technology have been described previ-
ously.9,12 Briefly, a 1.3-␮m infrared light is used to obtain high-resolu-
tion, cross-sectional tomographic images of the anterior segment.5–7
The image is horizontally composed of 256 A-scans in 16 mm with
1024 points per A-scan at 8 mm of depth. Each image has a maximum
transverse and axial resolution of 60 and 18 ␮m, respectively. Scanning
at 2000 axial scans per second, the machine needs approximately 1⁄8
second to scan an eye. Images were taken directly from the machine’s
output function as 816 ⫻ 636 pixel JPEG (lossless compression) files.
All selected images were temporal–nasal (i.e., horizontal) scans, to
maximize visibility of anatomic location and repeatability.9,10 Consec-
utive images of only the right eye were used to ensure uniformity with
other studies9,13 and to reduce bias.
Image Processing
All AS-OCT images were assessed by one ophthalmologist (LY) and
were processed with inbuilt software that dewarps the images (adjust-
ing for distortions arising from corneal optical properties). The scleral
spurs, defined as the anatomic junction between the inner wall of the
trabecular meshwork and the sclera, were identified (Fig. 1).13 There is
a prominent inner extension of the sclera at its thickest part,9,13 and in
this study, it was defined as a change in curvature of the inner surface
of the angle wall, often appearing as an inward protrusion of the sclera.
For each image the image file was opened, and the two scleral spurs
were identified on an individual scan. The algorithm then calculated all
parameters and the information was recorded (Fig. 1).
The ZAP (Guangzhou, China) software automatically extracted the
300 ⫻ 600 8-bit gray scale (intensities from 0 to 255) image portion of
the output file and performed noise and contrast conditioning.9 A
binary copy of the image was then produced in which pixels were
either 1’s (tissue) or 0’s (open space), depending on whether they
were brighter or darker than a calculated threshold. Algorithms de-
fined the borders of the corneal epithelium and endothelium and the
anterior surface of the iris. The algorithms used basic edge arguments
(five consecutive 0’s above and five consecutive 1’s below indicated an
anterior surface point) to describe the borders. The corneal border
Downloaded From: https://iovs.arvojournals.org/ on 09/09/2018
IOVS, July 2010, Vol. 51, No. 7
data were fitted with polynomial curves and a line-smoothing algo-
rithm that was explicitly defined by the edge-finding algorithms and
used the derivative data to repair steplike portions of the border.
Anatomic distances are calculated for anterior segment and corneal
parameters: ACD, central corneal thickness (CCT), anterior and poste-
rior corneal curvatures (ACC and PCC), and PCAL. All measurements
are in millimeters except for CCT, which is in micrometers.
Statistical Methods
Parametric and nonparametric tests were used to compare continuous
variables according to data distribution. Spearman’s rho correlation,
linear regression, and logistic regression analyses were used to assess
factors relating to the PCAL. P ⬍ 0.01 was considered statistically
significant (Stata; StataCorp LP, College Station, TX, and Excel; Mi-
crosoft, Redmond, WA).
RESULTS
A total of 750 consecutive eyes from Chinese Singaporeans
were included in the study. The mean age of the patients was
63.3 ⫾ 7.9 (range, 50 –90.0) years; 401 (53.4%) were women.
The mean ⫾ SD PCAL was 12.92 ⫾ 0.54 mm, ACD 2.68 ⫾ 0.31
mm, CCT 562.39 ⫾ 31.85 ␮m, ACC 7.36 ⫾ 0.37 mm, and PCC
6.65 ⫾ 0.34 mm. Table 1 summarizes the demographics of the
patients’ corneal parameters, stratified by age group and sex.
Posterior Corneal Arc Length
There was poor (r ⬍ 0.3) to moderately strong correlation (r ⫽
0.6 – 0.8),14 between PCAL and the other corneal parameters:
ACD (r ⫽ 0.55, r2 ⫽ 0.31, P ⬍ 0.001), CCT (r ⫽ ⫺0.071, r2 ⫽
0.005, P ⫽ 0.051), ACC (r ⫽ 0.114, r2 ⫽ 0.013, P ⫽ 0.002), and
PCC (r ⫽ 0.307, r2 ⫽ 0.026, P ⬍ 0.001; Fig. 2). Overall, there
was a poor correlation between PCAL and age (r ⫽ ⫺0.095, r2
⫽ 0.009, P ⬍ 0.01), height (R ⫽ 0.199, R2 ⫽ 0.04, P ⬍ 0.001),
weight (R ⫽ 0.151, R2 ⫽ 0.023, P ⬍ 0.001), and BMI (R ⫽
0.048, R2 ⫽ 0.002, P ⫽ 0.186; Fig. 3). PCAL also showed poor
correlation with asphericity, cylinder, and IOP (Fig. 4). Tables
2 and 3 summarize the relationship of PCAL in quartiles with
measured corneal and anterior segment parameters.
Univariate and Multivariate Analyses
Table 4 shows univariate and multivariate analyses for PCAL
with measured ocular, systemic, and refractive parameters. In
the univariate analysis, only CCT, BMI, and IOP were not
significantly associated with PCAL. (BMI was used for the
analysis instead of height and weight because of collinearity.)
A multivariate analysis with PCAL as the dependent param-
eter showed statistical significance in comparison with ACD,
ACC, PCC, CCT, and cylinder.
Sex. Logistic regression analyses showed that all corneal
parameters were significantly smaller in the women, except for
CCT, which was not significantly different between the men
and the women (OR ⫽ 0.998, P ⫽ 0.492; Table 5).
Age, Height, and Weight. The mean age of the patients
was 63.3 ⫾ 7.9 (range, 50 –90.0) years, and the mean height
and weight of the subjects was 160.2 ⫾ 7.9 cm and 61.3 ⫾ 10.7
kg, respectively. In the measured corneal and anterior segment
parameters, there was a decreasing trend with increasing age,
and an increasing trend with increasing height and weight,
which was somewhat expected although not always statisti-
cally significant (Table 6).
DISCUSSION
The biometric analysis of the anterior segment of the eye has
become increasingly important in the decision-making and
 IOVS, July 2010, Vol. 51, No. 7 Biometry of the Cornea and Anterior Chamber in Chinese Eyes 3435

TABLE 1. Measured Corneal and Anterior Segment Parameters Stratified by Age and Sex

ACD CCT ACC PCC PCAL

Age (mm) (␮m) (mm) (mm) (mm) n

All Persons
Mean 2.680 562.39 7.36 6.67 12.92 750
SD 0.310 31.85 0.37 0.34 0.54
Men
All Ages
Mean 2.72 563.25 7.39 6.70 12.99 349
SD 0.30 32.97 0.40 0.36 0.54
50–59 years
Mean 2.78 569.09 7.44 6.72 13.06 102
SD 0.32 31.79 0.39 0.33 0.53
60–69 years
Mean 2.74 567.18 7.36 6.69 12.96 144
SD 0.29 30.59 0.39 0.37 0.57
ⱖ70 years
Mean 2.65 551.96 7.37 6.69 12.97 103
SD 0.28 34.80 0.42 0.38 0.52
P for trend P ⫽ 0.007 P ⫽ 0.000 P ⫽ 0.249 P ⫽ 0.731 P ⫽ 0.35
Women
All Ages
Mean 2.65 561.64 7.32 6.61 12.86 401
SD 0.31 30.87 0.35 0.31 0.54
50–59 years
Mean 2.73 564.91 7.33 6.64 12.97 191
SD 0.31 31.65 0.34 0.31 0.54
60–69 years
Mean 2.60 561.39 7.32 6.60 12.80 146
SD 0.32 31.89 0.35 0.31 0.54
ⱖ70 years
Mean 2.53 552.44 7.33 6.58 12.70 64
SD 0.23 23.92 0.39 0.31 0.46
P for trend P ⫽ 0.000 P ⫽ 0.019 P ⫽ 0.96 P ⫽ 0.325 P ⫽ 0.000

management of corneal and anterior segment procedures, mizing the size of grafts for endothelial keratoplasty proce-
ranging from pre- and postoperative evaluation of AC IOLs dures such as DSAEK, where the donor cornea, adherent to
and phakic intraocular lenses (pIOLs),15 to potentially opti- the posterior stroma, protrudes into the AC and may also

FIGURE 2. Scatterplots with best fit

regression line and confidence inter-
vals between PCAL and corneal pa-
rameters: ACD (r ⫽ 0.55, P ⬍ 0.001),
CCT (r ⫽ ⫺0.071, P ⫽ 0.051), ACC
(r ⫽ 0.114, P ⫽ 0.002), and PCC (r ⫽
0.307, P ⬍ 0.001).

Downloaded From: https://iovs.arvojournals.org/ on 09/09/2018

 3436 Yuen et al.
crowd the angle. This study was conducted to document
and present the findings of corneal characteristics in Chi-
nese eyes from Singapore and to record associations and
correlations of ocular and systemic parameters with a novel
parameter, PCAL.
From our study, the mean PCAL was 12.92 mm (95% CI,
12.88 –12.96 mm), with a median of 12.91 mm. There was a
moderate correlation with ACD (R ⫽ 0.55, R2 ⫽ 0.3185, P ⬍
Downloaded From: https://iovs.arvojournals.org/ on 09/09/2018
IOVS, July 2010, Vol. 51, No. 7
FIGURE 3. Scatterplots with best fit
regression line with confidence inter-
vals between PCAL against age (r ⫽
⫺0.095, P ⫽ 0.01), height (r ⫽
0.199, P ⬍ 0.001), weight (r ⫽ 0151,
P ⬍ 0.001), and BMI (r ⫽ 0.048, P ⫽
0.186), showing no strong correla-
tions between the parameters.
0.001). Table 7 shows the relationship between ACD in quar-
tiles and other corneal parameters. Univariate and multivariate
analyses also showed a positive association between the two
parameters. In our study, a shallow ACD was found to be
associated with older age, hyperopic refractive error, female
sex, shorter body stature, which confirms previous population-
based studies on various ethnic populations (including Chinese
eyes) and hospital-based investigations.16 –19
FIGURE 4. Scatterplots with best fit
regression line with confidence inter-
vals between PCAL against sphere (D)
(r ⫽ ⫺0.156, P ⬍ 0.001), cylinder (D)
(r ⫽ ⫺0.103, P ⫽ 0.006), and IOP
(mm Hg) (r ⫽ ⫺0.063, P ⫽ 0.002).
 IOVS, July 2010, Vol. 51, No. 7 Biometry of the Cornea and Anterior Chamber in Chinese Eyes 3437

TABLE 2. Relationship of PCAL (in quartiles) with Measured Corneal TABLE 4. Univariate and Multivariate Analyses
and Anterior Segment Parameters
Univariate Multivariate
ACD CCT ACC PCC
PCAL (mm) n (mm) (␮m) (mm) (mm) PCAL Coefficient P Coefficient P

1st Quartile (ⱕ 12.13) ACD 0.564 ⬍0.001 0.541 ⬍0.001

Mean 190 2.46 566.20 7.31 6.53 ACC 0.123 0.001 ⫺0.291 ⬍0.001
SD 0.24 31.68 0.37 0.33 PCC 0.315 ⬍0.001 0.497 ⬍0.001
2nd Quartile (12.13–12.92) CCT ⫺0.066 0.073 ⫺0.103 ⬍0.001
Mean 185 2.64 561.90 7.31 6.60 Sex ⫺0.118 0.001 ⫺0.023 0.457
SD 0.26 32.24 0.38 0.33 Age ⫺0.089 0.015 0.001 0.973
3rd Quartile (12.92–13.68) BMI 0.066 0.071 0.028 0.337
Mean 188 2.73 560.50 7.39 6.68 Sphere ⫺0.156 ⬍0.001 0.018 0.556
SD 0.27 31.30 0.41 0.32 Cylinder ⫺0.103 0.006 ⫺0.085 0.004
4th Quartile (ⱖ 13.68) IOP ⫺0.063 0.085 ⫺0.021 0.467
Mean 187 2.92 560.90 7.40 6.80
SD 0.28 32.13 0.33 0.31 PCAL as the independent variable with measured ocular, systemic,
P for trend 0.000 0.277 0.010 0.000 and refractive parameters.

The linear relationship between the two variables, PCAL
and ACD, can be calculated as y ⫽ 10.257 ⫹ 0.9939x, where
y ⫽ PCAL and x ⫽ ACD. This positive relationship is visualized
in Figure 5; the arc’s positive relationship with the vertical
distance between the posterior cornea and the anterior lens
capsule is self-explanatory. We postulate that the perpendicu-
lar height between the posterior cornea and a horizontal line
from scleral spur to scleral spur would provide a stronger and
more robust correlation in this instance, as the PCAL also uses
the scleral spurs as landmarks (Fig. 5). As this measurement
was not part of the study, we believe the best estimator of
PCAL at present would be the linear equation as just stated.
There was a significant difference in PCAL between the
sexes, with the men exhibiting larger dimensions than the
women, 12.99 ⫾ 0.54 and 12.86 ⫾ 0.54 mm, respectively (P ⫽
0.001). This difference highlights the importance of customi-
zation of corneal surgery in men and women. Furthermore,
there was a significant inverse correlation with age and in the
univariate and multivariate analyses (P ⬍ 0.01). Hence, the
preoperative management of a younger man would be signifi-
cantly different from that of an elderly female patient. This
highlights a further need to customize surgical decisions to
achieve optimal outcomes.
Currently, there is no optimal selection of donor graft size
for DSAEK. This variable is left for the individual surgeon to
decide, based purely empirically on a peripheral 1- to 2-mm
clearance of the donor margins from the corneal limbus with-
out taking into account vertical clearance of the donor margins
from the iris and chamber angle. One of the major advantages
of DSAEK is to be able to transplant a larger surface area of
endothelial cells compared with standard penetrating kerato-
plasty (PK). A 9.0-mm graft, conventionally used in
DSAEK,20 –22 transfers 26% more surface area of healthy donor
endothelial cells than does the standard 8.0-mm graft more
commonly used in PK.23–25 However, because of the meniscal
configuration of the donor graft after ALTK or manual dissec-
tion, the thickened peripheral portions of the graft are at
increased risk of touching and adhering to the adjacent iris/
drainage angle, especially in Chinese eyes with shallower an-
terior chambers, and the resultant peripheral anterior syn-
echiae (PAS) may contribute to raised IOP after surgery and
also increase the risk of allograft rejection, similar to PAS
formation at the graft– host junction in PK. A solution may be
to design an algorithm using PCAL dimensions, subtracting a
constant to attain the maximum size graft that may be safely
implanted without angle compromise with respect to both
diameter and thickness of graft.
The mean radius of the ACC in our study was 7.36 mm (95%
CI, 7.33–7.38), with a median of 7.33 mm. There was a signif-
icant difference between the sexes in our study: The mean ⫾
SD for the men was 7.39 ⫾ 0.40 mm and for the women,
7.32 ⫾ 0.35 mm (P ⫽ 0.018). Dubbleman et al.11 also showed
a statistically significant difference between the sexes in a
predominantly Caucasian population.
The mean radius of the PCC in our study was 6.65 mm (95%
CI, 6.63– 6.68), with a median of 6.67 mm, which is compara-
ble to the range published in the literature. The radius of the
schematic Gullstrand eye is 6.8 mm,26 whereas in the sche-
matic eye of Le Grand and El Hage27 and Liou and Brennan,28
it is 6.5 and 6.4 mm, respectively.
Although the dimensions are both of the posterior cornea,
there was only a fair correlation between PCC and PCAL (R ⫽
0.307, R2 ⫽ 0.094, P ⬍ 0.001). As corneal curvatures are
measured for the central 3 mm of the cornea, this measure-
ment only provides the tangential point at 3 mm and not the
entire arc length. This result also suggests that the deviation
from the mean PCAL is accounted for by the variation beyond
the central 3 mm of the cornea to the scleral spurs.
There was a difference between the sexes in our study for
posterior corneal curvatures, the mean ⫾ SD for the men was
6.70 ⫾ 0.36 mm and for the women, 6.61 ⫾ 0.31 mm (P ⫽
0.001). Dubbleman et al.11 also showed a statistically signifi-
cant difference between the sexes, the men measuring 6.60 ⫾
0.03 mm and the women 6.456 ⫾ 0.03 mm (P ⬍ 0.01). The
mean age in their study was 39 ⫾ 14 years (cf. the mean age in
our study, 63.3 ⫾ 7.9 years). As far as we know, our results

TABLE 3. Spearman Correlation (r) with PCAL as the Independent Variable

PCAL ACD CCT ACC PCC Age Height Weight BMI Sphere Cylinder IOP

r 1.000 0.550** ⫺0.071 0.114* 0.307* ⫺0.095* 0.199* 0.151* 0.048 ⫺0.156* ⫺0.103* ⫺0.063*
(P) (⬍0.001) (0.051) (0.002) (⬍0.001) ⫺(0.01) (⬍0.001) (⬍0.001) (0.186) (⬍0.001) (0.006) (0.002)

The results show weak to moderately strong correlations between the measured corneal and anterior segment parameters, and systemic and
refractive parameters.
* Correlation is significant at the 0.01 level (two-tailed).

Downloaded From: https://iovs.arvojournals.org/ on 09/09/2018

 3438 Yuen et al. IOVS, July 2010, Vol. 51, No. 7

TABLE 5. Differences in Parameters According to Sex TABLE 7. Relationship of ACD (in Quartiles) with Measured Corneal
and Anterior Segment Parameters
Parameter/
Sex Mean SD Max Min ␤ OR P PCAL CCT ACC PCC
ACD (mm) n (mm) (␮m) (mm) (mm)
ACD, mm
Male 2.72 0.30 3.54 1.99 ⫺0.782 0.457 0.001* 1st Quartile ⱕ2.28
Female 2.65 0.32 3.65 1.88 Mean 189 12.55 559.2 7.33 6.59
CCT, ␮m SD 0.48 30.6 0.39 0.36
Male 563.25 32.97 659.2 474 ⫺0.002 0.998 0.492 2nd Quartile 2.28–2.68
Female 561.64 30.87 663.9 478 Mean 191 12.79 561.2 7.36 6.66
ACC, mm SD 0.45 32.5 0.34 0.33
Male 7.39 0.40 8.61 5.36 ⫺0.475 0.622 0.018† 3rd Quartile 2.68–3.17
Female 7.32 0.35 8.51 6.19 Mean 184 13.05 565.04 7.38 6.67
PCC, mm SD 0.48 30.90 0.42 0.36
Male 6.7 0.36 7.95 4.8 ⫺0.788 0.455 0.001* 4th Quartile ⱖ3.17
Female 6.61 0.31 7.95 5.52 Mean 186 13.32 564.30 7.34 6.68
PCAL, mm SD 0.44 33.20 0.34 0.36
Male 12.99 0.54 14.36 11.33 ⫺0.440 0.644 0.001* P for trend 0.000 0.251 0.533 0.034
Female 12.86 0.54 14.44 11.41

Measured corneal and anterior segment parameters are smaller in

the women with statistical significance for ACD, ACC, PCC, and PCAL.
* Significant at the 0.01 level.
† Significant at the 0.05 level.
show the first record of a difference between the sexes in the
corneal curvatures of Chinese eyes.
Dubbleman et al.11 found that ACC and PCC were not
age-dependent (P ⫽ 0.97 and 0.26, respectively). Our results
concur that there was no correlation of ACC and PCC with age
(P ⫽ 0.483 and 0.609, respectively).
There was no correlation or statistically significant associa-
tion between CCT and PCAL, suggesting parameter indepen-
dence. There was, however, a slight decrease in corneal thick-
ness with increasing age, which was not observed in a previous
study. Dubbelman et al.11 also showed that the men had a
slightly greater (⬃3 ␮m) CCTs (581 ␮m) than did the women
(578 ␮m); in our study the mean CCT in the men was 563.25 ⫾
32.97 ␮m and in the women, 561.64 ⫾ 30.87 ␮m, which not
a statistically significant difference.
There was a weak correlation between age and the corneal
parameters, with a tendency toward a decrease in corneal
dimensions with increasing age that was statistically significant
with PCAL (P ⫽ 0.01), ACD (P ⬍ 0.001), and CCT (P ⬍ 0.001;
Tables 2, 3, 8). Univariate regression also showed a significant
inverse association between age and PCAL, ACD, and CCT
(Table 4). Regression analyses showed a significantly positive
association between height and corneal parameters, except for
CCT (P ⫽ 0.402). A significantly positive association was also
found between weight and corneal parameters, except for CCT
(P ⫽ 0.054).
In addition to the aforementioned parameters, univariate
analysis showed a significant association between PCAL and
asphericity and cylinder (Table 4). There was no significant
association with CCT, IOP, and BMI. A multivariate analysis,
with PCAL as the dependent parameter, showed that only
cylinder was significantly associated. Further analysis on the
horizontality of the cylindrical axes compared with the hori-
zontality of the AS-OCT scans showed no statistical signifi-
cance. Cylinders were also analyzed at the 30° and 60° planes
(with corresponding opposite axes) and also showed no sig-
nificant difference (P ⬎ 0.05).
General systemic disorders may affect ocular surface and
corneal physiology; however, the anatomic measurements in
our cohort seemed to remain the same, with or without dis-
ease. Subgroup analyses showed no difference between the
eyes of patients with and without diabetes, those with and
without ischemic heart disease, and those with and without
hypertension (Table 9).
The inclusion of patients with higher myopia in this study is
important for the normative database in Singapore, and in the
Chinese diaspora, as myopia is prevalent. For completeness,
we performed exclusion analysis, removing high myopes
(⬎ ⫺5.0 D) and hyperopes (⬎5.0 D), and found that there was
no significant difference between this group and the original
750 eyes (P ⬎ 0.05). When myopes and hyperopes were then
divided into two groups, except for PCAL, all parameters be-
tween the two groups were similar and showed no significant
difference (Table 10). The sample for refractive sphere fol-
lowed a normal distribution.

TABLE 6. Correlation of Age, Height, and Weight with Measured

Corneal and Anterior Segment Parameters

Age Height Weight

␤ P ␤ P ␤ P

ACD ⫺0.006 ⬍0.001* 0.178 ⬍0.001* 0.169 ⬍0.001*

ACC ⫺0.54 0.483 0.132 ⬍0.001* 0.003 0.010 FIGURE 5. The vertical distance (solid line) between the posterior
PCC ⫺0.44 0.609 0.007 ⬍0.001* 0.096 0.008 cornea and the lens (ACD). The ACD varies with the vertical location
CCT ⫺0.04 ⬍0.001* 0.031 0.402 0.070 0.054 of the lens. Vertical dashed line: the perpendicular height between the
PCAL ⫺0.089 0.015† 0.195 ⬍0.001* 0.162 ⬍0.001* posterior cornea and an imaginary horizontal dashed line from scleral
spur to scleral spur (circles). It is postulated that this vertical height
* Significant at the 0.01 level. remains constant in relation to PCAL, which also uses the scleral spurs
† Significant at the 0.05 level. as landmarks.

Downloaded From: https://iovs.arvojournals.org/ on 09/09/2018

 IOVS, July 2010, Vol. 51, No. 7 Biometry of the Cornea and Anterior Chamber in Chinese Eyes 3439

TABLE 8. Spearman Correlation (r) of Measured Corneal and Anterior Segment Parameters (Excluding PCAL) with Systemic
and Refractive Parameters

ACD CCT ACC PCC Age Height Weight BMI Sphere Cylinder IOP

ACD 1.000 0.050 ⫺0.017 0.081* ⫺0.173† 0.168† 0.144† 0.075* ⫺0.355† 0.011 ⫺0.079*
(P) (0.1730) (0.640) (0.026) (⬍0.001) (⬍0.001) (⬍0.001) (0.039) (⬍0.001) (0.780) (0.031)
CCT 0.05 1.000 ⫺0.025 ⫺0.002 ⫺0.145† 0.028 0.052 0.046 ⫺0.146† 0.31 ⫺0.10
(P) (0.173) (0.491) (0.963) (⬍0.001) (0.451) (0.152) (0.209) (⬍0.001) (0.415) (0.78)
ACC ⫺0.017 ⫺0.025 1.000 0.802† ⫺0.036 0.179† 0.105† 0.026 0.188 ⫺0.021 ⫺0.015
(P) (0.64) (0.491) (⬍0.001) (0.33) (⬍0.001) (0.004) (0.474) (0.245) (0.571) (0.69)
PCC 0.081* ⫺0.002 0.802† 1.000 ⫺0.017 0.181† 0.101† 0.026 0.037 ⫺0.071 ⫺0.1
(P) (0.026) (0.963) (⬍0.001) (0.648) (⬍0.001) (0.006) (0.475) (0.323) (0.06) (0.791)

* Correlation is significant at the 0.05 level (two-tailed).

† Correlation is significant at the 0.01 level (two-tailed).

TABLE 9. Subgroup Analyses of Eyes with or without Systemic Diseases

Diabetes vs. No Hypertension vs. No Ischemic Heart Disease Sphericity (Excluding High
Diabetes Hypertension vs. No IHD Refractive Errors)

PCAL 0.108 0.217 0.322 0.799

ACD 0.438 0.909 0.696 0.493
CCT 0.082 0.884 0.615 0.995
ACC 0.302 0.795 0.166 0.731
PCC 0.457 0.342 0.678 0.635

No significant difference was found.

As a community-based epidemiologic study, our analyses
targeted AS-OCT images collected from one ethnicity, the Chi-
nese. Although not necessarily fully representative of all Chi-
nese populations globally, our cohort did include native Singa-
porean Chinese, immigrant mainland Chinese, and expatriate
Chinese (e.g., first- or second-generation Singaporean Chinese)
who had immigrated from abroad to Singapore, and our sample
size of 750 subjects was reasonably large, enabling statistical
comparison of other parameters, using one eye of each subject,
to reduce bias.
Other potential limitations of this study include the inability
to detect the scleral spur, which has previously been report-
ed.7 The visibility of the scleral spur was not detected in
images where the internal surface of the sclera formed a
smooth continuous line (with no inward protrusion of the
sclera or change in its curvature) or in images with suboptimal
quality. The study was also limited in that the measurements
were restricted to horizontal nasal–temporal AS-OCT scans, as
these have been shown to be the most consistent with respect
to obtaining high-quality images for the ZAP program to ana-
lyze.13 AS-OCT imaging has been shown to be restricted by
eyelid anatomy for vertical scans. The mean age of the study’s
population (63.3 ⫾ 7.9 years) is relatively older than that of
Dubbelman et al.11 (39 ⫾ 14 years) and of Pinero et al.15
(32.82 ⫾ 7.91 years) and therefore comparisons with these
patient populations may not be entirely valid. Furthermore, our
population targeted those who were 50 years old and older;
TABLE 10. Subgroup Analyses between Myopes and Hyperopes
Myopes Hyperopes
PCAL 13.035 12.835
ACD 2.664 2.696
CCT 562.775 562.236
ACC 7.369 7.352
PCC 6.675 6.647
Results showed no difference between myopes and hyperopes
when higher refractive errors (⬎ ⫺5.0 D, ⬎5.0 D) and emmetropes
were removed (n ⫽ 608).
hence, our results are not generalizable to younger patients. It
is reassuring, however, to find similar analytic outcomes of
parameters between the sexes and associations with systemic
measurements.
This article serves to provide normal values of corneal
dimensions and a novel parameter, PCAL, for Chinese eyes. We
believe that these data will be clinically applicable for the
surgical management of posterior corneal procedures (e.g.,
DSAEK) and in the assessment of patients requiring other forms
of anterior segment or corneal surgery, including anterior
chamber or iris-supported phakic IOL surgery, IOL implanta-
tion in cataract surgery, or refractive intrastromal corneal im-
plantation.
References
1. Alio JL. Advances in phakic intraocular lenses: indications, efficacy,
safety and new designs. Curr Opin Ophthalmol. 2004;15:350 –
357.
2. Elgohary MA, Chauhan DS, Dowler JG. Optical coherence tomog-
raphy of intraocular lens implants and their relationship to the
posterior capsule: a pilot study comparing a hydrophobic acrylic
to a plate haptic silicone type. Ophthalmic Res. 2006;38:116 –124.
3. Avila M, Li Y, Song JC, Huang D. High speed optical coherence
tomography for management after laser in situ keratomileusis. J
Cataract Refract Surg. 2006;32:1836 –1842.
4. Nolan WP, Aung T, Machin Det al. Detection of narrow angles and
established angle closure in Chinese residents of Singapore: poten-
tial screening tests. Am J Ophthalmol. 2006;141:896 –901.
5. Radhakrishnan S, Goldsmith J, Huang D et al. Comparison of
optical coherence tomography and ultrasound biomicroscopy for
P detection of narrow anterior chamber angles. Arch Ophthalmol.
2005;123:1052–1059.
⬍0.001 6. Radhakrishnan S, Huang D, Smith SD. Optical coherence tomog-
0.236 raphy imaging of the anterior chamber angle. Ophthalmol Clinic
0.849 North Am. 2005;18:375–381, vi.
0.599 7. Dorairaj S, Liebmann J, Ritch R. Quantitative evaluation of anterior
0.366 segment parameters in the era of imaging. Trans Am Ophthalmol
Soc. 2007;105:99 –110.
8. Muler M, Dahmen G, Porksen E, et al. Anterior chamber angle mea-
surement with optical coherence tomography: intraobserver and in-
terobserver variability. J Refract Cataract Surg. 2006;32:1803–1808.

Downloaded From: https://iovs.arvojournals.org/ on 09/09/2018

 3440 Yuen et al.
9. Console J, Sakata L, Aung T, Friedman D, He M. Quantitative
analysis of anterior segment optical coherence tomography
images: the Zhongshan Angle Assessment Program. Br J Ophthal-
mol. 2008;92:1612–1616.
10. Chan JB, Huang EH, Yuen LH, et al. Reproducibility of cornea
measurements in anterior segment OCT images analyzed with the
Zhongshan Assessment Program (ZAP): poster presented at Euro-
pean Society of Cataract and Refractive Surgery (ESCRS), Barce-
lona, Spain, 2009.
11. Dubbelman M, Sicam V, van der Heijde G. The shape of the
anterior and posterior surface of the aging human cornea. Vision
Res. 2006;46:993–1001.
12. Su DH, Friedman DS, See JL, et al. Degree of angle closure and
extent of peripheral anterior synechiae: an anterior segment OCT
study. Br J Ophthalmol. 2008;92:103–107.
13. Sakata L, Lavanya R, Friedman D et al. Assessment of the scleral
spur in anterior segment optical coherence tomography images.
Arch Ophthalmol. 2008;126:181–185.
14. Chan YH. Correlational Analysis. Singapore Med J. 2003;44(12):
614 – 619.
15. Pinero D, Plaza A, Alio J. Anterior segment biometry with 2 imaging
technologies: very-high-frequency ultrasound scanning versus optical
coherence tomography. J Cataract Refract Surg. 2008;34:95–102.
16. Xu L, Cao WF, Wang YX, et al. Anterior chamber depth and
chamber angle and their associations with ocular and general
parameters: The Beijing Eye Study. Am J Ophthalmol. 2008;145:
929 –936.
17. Congdon NG, Youlin Q, Quigley H, et al. Biometry and primary
angle glaucoma among Chinese, white, and black populations.
Ophthalmology. 1997;104:1489 –1495.
Downloaded From: https://iovs.arvojournals.org/ on 09/09/2018
IOVS, July 2010, Vol. 51, No. 7
18. Devereux JG, Foster PJ, Baasanhu J, et al. Anterior chamber depth
measurement as a screening tool for primary angle closure glau-
coma in an East Asian population. Arch Ophthalmol. 2000;118:
257–263.
19. Nolan WP, See JL, Chew PT et al. Detection of primary angle
closure using anterior segment optical coherence tomography in
Asian eyes. Ophthalmology. 2007;114:33–39.
20. O’Brien P, Lake D, Saw V et al. Endothelial keratoplasty: case
selection in the learning curve. Cornea. 2008;27:1114 –1118.
21. Gorovoy MS. Descemet-stripping automated endothelial kerato-
plasty. Cornea. 2006;25:886 – 889.
22. Price MO, Price FW Jr. Descemet’s stripping with endothelial
keratoplasty: comparative outcomes with microkeratome-dis-
sected and manually dissected donor tissue. Ophthalmology.
2006;113:1936 –1942.
23. Thompson RW Jr, Price MO, Bowers PJ, et al. Long term graft
survival after penetrating keratoplasty. Ophthalmology. 2003;110:
1396 –1402.
24. Bertelmann E, Pleyer U, Reick P. Risk factors for endothelial cell
loss post-keratoplasty. Acta Ophthalmol Scand. 2006;84:766 –770.
25. Tan DTH, Janardhanan P, Zhou H, et al. Penetrating keratoplasty in
Asian Eyes; the Singapore Corneal Transplant Study. Ophthalmol-
ogy. 2008;115:975–982.
26. Atchison DA, Smith G. Optics of the Human Eye. Oxford, UK:
Butterworth-Heinemann; 2000;34 –35, 166 –167, 251–256.
27. Le Grand Y, El Hage SG. Physiological Optics. Berlin: Springer-
Verlag; 1980:65– 67.
28. Liou HL, Brennan NA. Anatomically accurate, finite model eye for
optical modelling. J Opt Soc Am Opt Image Sci. 1997;14:1684 –
1695.