Purpose
To describe the distribution of central corneal thickness (CCT), intraocular pressure (IOP), and their determinants and association with glaucoma in Chinese adults.
Design
Population-based cross-sectional study.
Methods
Chinese adults aged 50 years and older were identified using cluster random sampling in Liwan District, Guangzhou. CCT (both optical [OCCT] and ultrasound [UCCT]), intraocular pressure (by Tonopen, IOP), refractive error (by autorefractor, RE), radius of corneal curvature (RCC), axial length (AL), and body mass index (BMI) were measured, and history of hypertension and diabetes (DM) was collected by questionnaire. Right eye data were analyzed.
Results
The mean values of OCCT, UCCT, and IOP were 512 ± 29.0 μm, 542 ± 31.4 μm, and 15.2 ± 3.1 mm Hg, respectively. In multiple regression models, CCT declined with age ( P < .001) and increased with greater RCC ( P < .001) and DM ( P = .037). IOP was positively associated with greater CCT ( P < .001), BMI ( P < .001), and hypertension ( P < .001). All 25 persons with open-angle glaucoma had IOP <21 mm Hg. CCT did not differ significantly between persons with and without open- or closed-angle glaucoma. Among 65 persons with ocular hypertension (IOP >97.5th percentile), CCT (555 ± 29 μm) was significantly ( P = .01) higher than for normal persons.
Conclusions
The distributions of CCT and IOP in this study are similar to that for other Chinese populations, though IOP was lower than for European populations, possibly due to lower BMI and blood pressure. Glaucoma with IOP <21 mm Hg is common in this population. We found no association between glaucoma and CCT, though power (0.3) for this analysis was low.
Intraocular pressure (IOP) is widely recognized as the most important modifiable risk factor for the development of glaucoma. Mean IOP in the white populations of Europe and North America is higher than that of East Asian populations by approximately 2 to 5 mm Hg. Some, but not all, epidemiologic studies in whites have identified a positive association between IOP and age, but a similar age pattern has not been reported for East Asian populations.
Thinner central corneal thickness (CCT) is known to be associated with lower measured IOP, and may also be an independent risk factor for open-angle glaucoma. CCT varies between racial groups, with values being lower in persons of African and possibly Mongolian descent as compared to Europeans. Other clinic-based studies have not identified significant differences between Asians, Hispanics, and whites. As different measurement methods were used in the studies described, caution is required in making comparisons between them.
Other determinants of IOP have also been investigated in cross-sectional studies. Diabetes mellitus (DM) and hypertension have been found to be associated with thicker corneas as well as higher IOP. Several studies have reported higher IOP in the setting of myopia, though 1 non-population-based study in young Singaporean adults reported the opposite, and other studies have found no association.
There are limited population-based data available on IOP and CCT among adult Chinese dwelling in the People’s Republic of China. The Beijing Eye Study reported the IOP distribution measured by noncontact tonometry in a defined population identified without random sampling in northern China. Another study has reported the IOP distribution by Perkins tonometry for a northern Chinese population. The IOP distribution differed substantially between these 2 studies, which may be attributable to different measurement and sampling methods.
The Liwan Eye Study is a population-based study conducted in urban southern China. The purpose of this paper is to report the IOP and CCT distribution, systemic and ocular determinants of IOP and CCT, and associations of IOP and CCT with glaucoma in this cohort. In order to enhance comparability with other studies, both optical and ultrasound pachymetry were used to measure CCT in the present report.
Materials and Methods
The detailed study protocol has been reported elsewhere. In brief, subjects were enrolled from a population-based study conducted among residents of Liwan District, Guangzhou City, aged 50 years and above using cluster random sampling. Fieldwork was carried out between September 10, 2003 and February 14, 2004.
A standard questionnaire was administered by a trained interviewer. DM and hypertension were defined based on self-reported history of a diagnosis or previous medication use. Height and weight were measured with the subject standing and without shoes, using a standard calibrated scale. Body mass index (BMI) was calculated as the weight in kilograms divided by the square of the height in centimeters.
A handheld autorefractor (ARK-30; Nidek Corp, Gamagori, Japan) was used to measure noncycloplegic refraction and radius of corneal curvature (RCC). The refraction data were converted to spherical equivalent (SE, sphere + 1/2 of cylinder). An average of the horizontal and perpendicular RCC was used in data analysis.
Both optical and ultrasound pachymetry were used to evaluate CCT before pharmacologic dilation of the pupils. In the optical pachymetry method, CCT was measured with a device (Device I; Haag-Streit, Bern, Switzerland) mounted on a slit lamp (Model 900; Haag-Streit, Bern, Switzerland), at 1.6x objective magnification with a +2.5 diopter eyepiece, and read to the nearest 0.01 mm. Three measurements were carried out and the median of the 3 readings was recorded. Ultrasound pachymetry was carried out with A-mode ultrasound (Echoscan US1800; Nidek, Corp) with the pachymetry probe perpendicular to the cornea at the center of the pupil. Five separate measurements were automatically taken and the average value recorded. Axial length, anterior chamber depth, and lens thickness were measured using the same machine, with the biometric probe held perpendicular to the pupil center. The best trace figure of 10 separate measurements for each parameter was taken. If the standard deviation exceeded 0.13 mm for the anterior chamber depth, the reading deviating the furthest from the median was deleted and additional measurements obtained. An average of 10 readings was used in data analysis.
IOP was measured with a handheld tonometer (Tonopen; Mentor, Norwell, Massachusetts, USA) after instilling topical anesthesia (0.4% Benoxil; Oxybuprocaine, Santen, Osaka, Japan) and before pupil dilation. The device’s internal calibration program was run at the beginning of each day. The measurements were repeated if the standard error (SE) exceeded 5%. If 3 consecutive measurements could not achieve SE <5% or if the subject could not cooperate, the IOP was considered not measurable. One measurement was taken and recorded for each eye. The decision to employ the Tonopen was made in view of a previous manometric study suggesting the Tonopen provides a more accurate measure of true IOP in Chinese eyes.
The diagnosis of glaucoma was made according to ISGEO (International Society of Geographical Epidemiological Ophthalmology) criteria. The highest level of evidence requires optic disc and visual field findings consistent with glaucoma (vertical cup-to-disc ratio [VCDR] or between-eyes asymmetry >97.5th percentile for the population, and a reproducible glaucomatous field defect). If the visual field test is not reliable, a severely damaged disc (VCDR >99.5th percentile) alone is considered compatible with a diagnosis of glaucoma. If the optic disc cannot be examined because of severe media opacity, subjects who are blind and have undergone previous glaucoma surgery or who have an IOP greater than the 99.5 th percentile for the population are classified as having glaucoma. If the pigmented trabecular meshwork is not visible in 3 or more quadrants without dynamic gonioscopy in the setting of primary glaucoma as defined above, primary angle-closure glaucoma (PACG) is diagnosed. Primary glaucoma without such narrowing of the angle is considered to be primary open-angle glaucoma (POAG). Ocular hypertension is defined as IOP >97.5th percentile for the population (≥20 mm Hg in this case).
Statistical Analysis
Given that the IOP and CCT readings were similar in both eyes of subjects (correlation coefficient 0.75 for IOP and 0.85 for optical CCT [OCCT]), measurements from the right eye were selected for analysis. Data from eyes with previous intraocular surgery (including laser), corneal disease, and trauma involving the cornea were excluded. The normality of the distribution of IOP and CCT was confirmed and a multiple linear regression model constructed to assess the effects of potential determinants on IOP and CCT, and of IOP and CCT on glaucoma. Standardized regression coefficients (SRC) were estimated in order to convert all independent variables to a standard distribution (mean = 0, standard deviation = 1), thereby equalizing the scale of the independent variables in the model. The proportion of variation in the dependent variable explained by the model was calculated (R 2 ). Proportions and mean values were compared using the χ 2 and t test respectively. P values <.05 were considered statistically significant. Data obtained by optical and ultrasound pachymetry were analyzed separately. Data analysis was performed using the Stata Package (Stata 8.0; Stata Corp, College Station, Texas, USA).
Results
In total, 1865 people were identified as eligible in the sampling district. Among them, 1405 (75.4%) were successfully examined. Reasons for nonparticipation included 168 refusals (9%), 66 (3.5%) severely ill or physically disabled, and 226 (12.1%) who could not be contacted after at least 3 home visits. In general, people aged 80 years and over were less likely to participate than other age groups. Among the 1405 subjects, 12 did not have IOP data available. Optical CCT data were missing for 130 eyes, and ultrasonic CCT data were not available in an additional 13 eyes. Forty-five subjects were further excluded from analyses because of previous ocular surgery (24 subjects), significant corneal disease (16 subjects), ocular trauma (2 subjects), exotropia preventing accurate measurement (2 subjects), and phthisis bulbi (1 subject). Therefore, data from 1348, 1218, and 1205 eyes were available for IOP, optical CCT, and ultrasonic CCT (UCCT) analysis respectively ( Table 1 ).
| Participants a (n = 1360) | Refusals a (n = 45) | P Value | |
|---|---|---|---|
| Age (years) | 65.1 (9.9) | 70.0 (8.5) | .001 |
| Men (%) | 43.3% | 56.7% | .235 |
| Body mass index (kg/m 2 ) | 23.5 (3.4) | 23.0 (3.7) | .427 |
| Hypertension (%) | 32.4% | 35.3% | .721 |
| Diabetes (%) | 6.9% | 8.3% | .441 |
| Axial length (mm) | 23.3 (1.5) | 23.5 (2.0) | .572 |
| Spherical equivalent refractive error (diopters) | −0.4 (2.7) | −0.8 (2.8) | .369 |
| Radius of corneal curvature (mm) | 7.8 (0.3) | 7.7 (0.3) | .480 |
The mean age of the 1348 subjects with IOP data was 64.8 ± 9.8 years, and 576 (42.7%) were men. No differences were found in age between men and women ( P = .84, t test).
Figures 1 and 2 illustrate the distributions of IOP and optical and ultrasonic CCT in this population. IOP (S-K test for normality, P for skewness = 0.530, P for kurtosis = 0.567) and ultrasonic CCT (P for skewness = 0.236, P for kurtosis = 0.100) data were normally distributed. UCCT readings were 30 μm greater (95% limit of agreement: -16.4 to 76.6 μm) than OCCT, although the 2 measurements were highly correlated (r = 0.69, P < .001).
Table 2 shows the distribution of CCT by age and gender. Both optical and ultrasonic CCT tended to decrease with age in an approximately linear fashion. Optical and ultrasonic CCT decreased by 3.5 μm ( P < .001) and 4.0 μm ( P < .001) per decade of age in this population. Similarly, IOP was lower among older persons, with a decline of 0.3 mm Hg per decade for all subjects ( P = .003) ( Table 3 ). No gender differences were detected for CCT and IOP.
| OCCT | UCCT | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Male | Female | Total | Male | Female | Total | |||||||
| Age (y) | n | Mean (SD) | n | Mean (SD) | n | Mean (SD) | n | Mean (SD) | n | Mean (SD) | n | Mean (SD) |
| 50-59 | 188 | 516.2 (29.2)3 | 239 | 515.7 (27.7) | 427 | 515.6 (28.3) | 186 | 549.8 (31.8) | 236 | 545.2 (29.3) | 422 | 547.2 (30.5) |
| 60-69 | 158 | 511.6 (28.2)2 | 211 | 510.9 (26.1) | 369 | 511.2 (27.0) | 158 | 542.5 (30.6) | 210 | 540.0 (29.5) | 368 | 541.1 (29.9) |
| 70-79 | 147 | 506.6 (30.4)1 | 203 | 510.8 (32.7) | 350 | 509.0 (31.8) | 145 | 531.5 (31.7) | 199 | 539.9 (32.7) | 344 | 537.4 (32.6) |
| 80-93 | 27 | 505.0 (25.1)1 | 45 | 500.9 (23.4) | 72 | 502.6 (24.1) | 27 | 531.2 (29.0) | 44 | 530.8 (34.1) | 71 | 530.9 (31.9) |
| Total | 520 | 511.4 (29.2) | 698 | 511.6 (28.7) | 1218 | 511.6 (29.0) | 516 | 541.3 (32.1) | 689 | 541.7 (30.8) | 1205 | 541.5 (31.4) |
| Beta a | −0.49 | −0.25 | −0.35 | −0.55 | −0.29 | −0.40 | ||||||
| P value b | <.001 | .026 | <.001 | <.001 | .017 | <.001 | ||||||
a Regression coefficient of the linear regression between CCT and age.
| Male | Female | Total | ||||
|---|---|---|---|---|---|---|
| Age (years) | n | Mean (SD) | n | Mean (SD) | n | Mean (SD) |
| 50–59 | 208 | 15.3 (2.9) | 266 | 15.5 (2.8) | 474 | 15.4 (2.9) |
| 60–69 | 169 | 15.3 (3.7) | 226 | 15.3 (3.3) | 395 | 15.3 (3.4) |
| 70–79 | 165 | 14.7 (2.8) | 223 | 15.2 (3.0) | 388 | 15.0 (3.0) |
| 80–93 | 34 | 13.5 (2.9) | 57 | 14.7 (3.4) | 91 | 14.2 (3.2) |
| Total | 576 | 15.0 (3.2) | 772 | 15.4 (3.1) | 1348 | 15.2 (3.1) |
| Beta a | −0.04 | −0.02 | −0.03 | |||
| P value b | .004 | .040 | .003 | |||
a Regression coefficient of the linear regression between intraocular pressure and age.
b P value of the linear regression between intraocular pressure and age.
Table 4 summarizes the regression model for potential determinants of optical and ultrasonic CCT. In the final multiple regression model, younger age ( P < .001), higher radius of corneal curvature (eg, a flatter cornea, P < .001), and diabetes ( P = .037) were significantly associated with greater ultrasonic CCT. For optical CCT however, only younger age ( P = .001) and flatter cornea ( P = .011) were significantly associated with a thicker cornea in the final multiple regression model, while diabetes was not ( P = .168). Among all the significant predictors, age and corneal curvature contributed equally to variation in CCT (standardized regression coefficient = −0.12 and 0.08 for age and corneal curvature for OCCT; −0.15 and 0.13 for age and corneal curvature for UCCT).
| OCCT | UCCT | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Univariate | Age- and Gender-Adjusted | Final Model a | Univariate | Age- and Gender-Adjusted | Final Model b | |||||||||
| β | P | β | P | β | P | Standardized Regression Coefficient | β | P | β | P | β | P | Standardized Regression Coefficient | |
| Age | −0.35 | <.001 | −0.34 § | <.001 | −0.37 | <.001 | −0.12 | −0.48 | <.001 | −0.48 § | <.001 | −0.47 | <.001 | −0.15 |
| BMI | 0.02 | .946 | 0.08 | .776 | — | — | — | 0.12 | .693 | 0.04 | .888 | — | — | — |
| Hypertension | −3.90 | .052 | −2.96 | .146 | — | — | — | −4.11 | .056 | −2.64 | .222 | — | — | — |
| Diabetes | 4.32 | .057 | 5.61 | .049 | 5.18 | .168 | 0.04 | 6.41 | .100 | 8.34 | .032 | 7.90 | .037 | 0.07 |
| Axial length | 1.28 | .251 | 1.43 | .310 | — | — | — | 1.07 | .082 | 1.21 | .087 | — | — | — |
| Corneal curvature | 7.52 | .049 | 7.22 | .033 | 9.47 | .011 | 0.08 | 12.78 | <.001 | 12.56 | <.001 | 14.21 | <.001 | 0.13 |
| Refractive error | −0.35 | .276 | −0.32 | .315 | — | — | — | −0.05 | .876 | −0.002 | .995 | — | — | — |
| R 2 = 0.02 c | R 2 = 0.04 c | |||||||||||||
a Multiple regression model with OCCT as dependent variable, age, diabetes mellitus, and corneal curvature as independent variables.
b Multiple regression model with UCCT as dependent variable, age, diabetes mellitus, and corneal curvature as independent variables.
Systemic and ocular determinants of IOP are shown in Table 5 . The final multiple regression model displayed better fitness when using UCCT (R 2 = 0.11) as compared to OCCT (R 2 = 0.06), and thus only this model is given (results for both were essentially the same). After adjusting for age and gender, IOP decreased with age ( P = .003) and was positively correlated with BMI ( P < .001), CCT ( P < .001), and self-reported hypertension ( P < .001). In the final model, higher IOP was significantly associated with higher BMI ( P < .001), self-reported hypertension ( P < .001), and thicker CCT ( P < .001), whereas age was no longer significantly associated with IOP. The mean IOP increased by approximately 2.3 mm Hg for every 100-μm increment in UCCT. UCCT was the most important contributor to IOP variation, with the greatest absolute standardized regression coefficient (SRC = 0.25, 0.15, and 0.12 for UCCT, BMI, and presence of hypertension, respectively).
| Univariate | Adjusting for Age and Gender | Final Model a | |||||
|---|---|---|---|---|---|---|---|
| β | P | β | P | β | P | Standardized Regression Coefficient | |
| Age | −0.02 | .003 | −0.02 | .003 | −0.002 | .823 | −0.01 |
| BMI | 0.17 | <.001 | 0.16 | <.001 | 0.14 | <.001 | 0.15 |
| Hypertension | 0.81 | <.001 | 0.84 | <.001 | 0.82 | <.001 | 0.12 |
| Diabetes | 0.39 | .314 | 0.37 | .335 | — | — | — |
| Optical CCT | 0.02 | <.001 | 0.02 | <.001 | — | — | — |
| Ultrasonic CCT | 0.02 | <.001 | 0.02 | <.001 | 0.03 | <.001 | 0.25 |
| Axial length | −0.02 | .716 | 0.002 | .977 | — | — | — |
| Radius of corneal curvature | −0.44 | .113 | −0.33 | .277 | — | — | — |
| Refractive error | −0.04 | .212 | −0.04 | .227 | — | — | — |
| R 2 = 0.11 b | |||||||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
