Perera SA, Wong TY, Tay W, Foster PJ, Saw S, Aung T. Refractive Error, Axial Dimensions, and Primary Open-Angle GlaucomaThe Singapore Malay Eye Study. Arch Ophthalmol. 2010;128(7):900-905. doi:10.1001/archophthalmol.2010.125
To explore the relationship of refractive error and ocular biometry with primary open-angle glaucoma (POAG) in the Malay population.
The Singapore Malay Eye Study is a population-based cross-sectional survey that examined 3280 persons (78.7% response) aged 40 to 80 years. Participants underwent a standardized clinical examination including slitlamp biomicroscopy, Goldmann applanation tonometry, refraction, dilated optic disc assessment, and measurement of axial length (AL) and central corneal thickness (CCT). Glaucoma was defined according to International Society for Geographical and Epidemiological Ophthalmology criteria.
After adjusting for possible confounders, persons with moderate or high myopia (less than −4.0 diopters, right eyes) were more likely to have POAG (odds ratio [OR], 2.87; 95% confidence interval, 1.09-7.53); this association remained significant after controlling for CCT (2.80; 1.07-7.37). Longer AL was associated with POAG (ORs, 2.49, 3.61, and 2.88, comparing quartiles 2, 3, and 4, respectively, with quartile 1 of AL; P = .03 for trend). If CCT was controlled for, persons with AL in quartile 4 were 3 times more likely to have POAG (OR, 3.00; 95% confidence interval, 1.09-8.24) than those in quartile 1.
This population-based study in Singapore shows an association of moderate myopia and increasing AL with POAG independent of other factors, including CCT.
The relationship between refractive error and glaucoma has been investigated in several clinical trials and population-based studies.1- 4 Most studies have suggested that moderate to high myopia is associated with increased risk of primary open-angle glaucoma (POAG),5,6 low-tension glaucoma,7,8 and ocular hypertension.9- 12 In the Blue Mountains Eye Study in Australia, after adjusting for age, sex, and other risk factors, eyes with moderate myopia were 2 times more likely to have POAG.13 In the Barbados Eye Study, a myopic refraction was one of several risk factors for POAG in adult black people.14,15 The Beaver Dam Eye Study showed that, after taking into account the effects of age, sex, and other risk factors, persons with myopia were 60% more likely to have glaucoma than those with emmetropia.16 In Asian populations, the relationship of myopia and POAG was reported in the Beijing Eye Study in China (significant relationship with high myopia of less than −6 diopters [D])17 and the Meiktila Eye Study in Myanmar.18 However, not all studies have found significant relationships; for example, no association between myopia and POAG was found in the Ocular Hypertension Treatment Study.3
The association between myopia and POAG has been thought to be due to a variety of mechanisms, including increased susceptibility of the optic nerve head to damage by raised intraocular pressure (IOP) and the increased effect of shearing forces in optic nerve head damage. Jonas et al19 showed that, for a given IOP in eyes with POAG, optic nerve damage appears to be more pronounced in highly myopic eyes with large optic discs than in non–highly myopic eyes. This may suggest a higher susceptibility for glaucomatous optic nerve fiber loss in highly myopic eyes compared with non–highly myopic eyes.18,20 However, a major gap in the current literature is a lack of studies on the relationship between axial dimensions and POAG. Thus, it is unclear whether the relationship between myopia and POAG is mediated by axial length (AL) or other factors (corneal curvature or lenticular changes with age). In addition, previous studies have not controlled for the effects of central corneal thickness (CCT), which is now known to strongly influence the measurement of IOP and is a risk factor for POAG.3
In this study, we investigated the association of refractive error and ocular biometry with POAG while controlling for the effects of CCT in a population-based study of Asian Malay adults in Singapore.
The Singapore Malay Eye Study is a population-based cross-sectional study of Malay subjects aged 40 to 80 years in Singapore. The study methods have been described previously.21,22 The sampling frame consisted of all Malays aged 40 to 80 years living in 15 residential districts across southwestern Singapore, selected using an age-stratified random sampling procedure from a list of Malay names provided by the Ministry of Home Affairs. Of the 4168 participants eligible to participate, 3280 (78.7%) were examined in the clinic and the remaining 888 (21.3%) were classified as nonparticipants. Nonparticipants tended to be older (aged 70-80 years) compared with participants, but there was no difference in sex, sampling location, and telephone ownership between the 2 groups.
Approval for the study protocol was granted by the hospital's institutional review board, and the study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all subjects before enrollment.
All participants underwent a standardized interview, examination, and ocular imaging at a centralized study clinic.23,24 Visual acuity was measured using a logMAR vision chart (Lighthouse International, New York, New York) read at a distance of 4 m. The refraction and corneal curvature of both of the participant's eyes were measured with an autorefractor (RK-5 Autorefractor Keratometer; Canon Inc Ltd, Tokyo, Japan). Final refraction was determined using subjective refraction by trained and certified study optometrists. Axial length and central anterior chamber depth (ACD) were measured by noncontact partial coherence laser interferometry (IOLMaster; Carl Zeiss, Jena, Germany). Central corneal thickness was measured in each eye with an ultrasound pachymeter (Advent; Mentor O & O Inc, Norwell, Massachusetts), and the median value of 5 readings was recorded.
A structured slitlamp examination (model BQ-900; Haag-Streit, Berne, Switzerland) was performed by study ophthalmologists before and after pupil dilation. A Goldmann applanation tonometer (Haag-Streit) was used to obtain 1 reading of IOP from each eye before dilation. After pupil dilation, the eyes were examined at the slitlamp again for an optic disc and retinal examination. Vertical cup-disc ratio (VCDR) and neuroretinal rim appearance were carefully recorded.
Refractive error was estimated using standardized subjective refraction techniques; if unavailable, autorefraction measurements were used instead. Myopia was defined as a spherical equivalence (SE) of less than −0.50 D; hyperopia, an SE of greater than +0.5 D; and emmetropia, an SE of −0.5 to +0.5 D. Moderate or high myopia was defined as an SE of less than −4.0 D; this cutoff was chosen because there were few eyes with high myopia (SE of less than −6.0 D) for an analysis with POAG.
The diagnosis and classification of glaucoma cases have been reported previously25,26 and will be described briefly. Glaucoma cases were defined according to the International Society for Geographical and Epidemiological Ophthalmology criteria based on 3 categories.25 Category 1 cases were defined as optic disc abnormality (VCDR or VCDR asymmetry ≥97.5th percentile or neuroretinal rim width from the 11- to 1-o’clock position or the 5- to 7-o’clock position <0.1 VCDR) and glaucomatous visual field defect. Category 2 cases were defined as a severely damaged optic disc (VCDR or VCDR asymmetry ≥99.5th percentile) in the absence of an adequate visual field test. When diagnosing category 1 or 2 glaucoma, it was required that there be no other explanation for the VCDR finding (ie, dysplastic disc or marked anisometropia) or visual field defect (ie, retinal vascular disease, macular degeneration, or cerebrovascular diseases). Category 3 cases were defined as blindness in individuals who had no visual field or optic disc data (corrected visual acuity, <3/60) and who had undergone previous glaucoma surgery or had an IOP greater than the 99.5th percentile. Ocular hypertension was defined as the presence of IOP greater than 21 mm Hg in individuals who did not meet the criteria for glaucoma. Primary open-angle glaucoma was defined as an eye with evidence of glaucoma (as defined in this paragraph) with an angle appearance in which the posterior trabecular meshwork was seen for 180° or more of the angle circumference during dynamic gonioscopy.26
Statistical analysis was performed using commercially available software (SPSS, version 15; SPSS Inc, Chicago, Illinois). Any POAG was analyzed as a binary outcome variable, and all potential risk factors were categorized as defined in the previous section (eg, refractive error) or in quartiles (eg, AL). We used logistic regression models to estimate the odds ratios (ORs) and 95% confidence intervals (CIs) for POAG for each risk factor, adjusting for age and sex. In multivariate analysis, we further adjusted for education, height, hypertension, and hemoglobin A1c level, and subsequently IOP and CCT. We also adjusted for AL in models of refractive error and POAG to determine the relative contribution of AL to the relationship. The likelihood ratio test was used to assess the contribution of AL in the regression model. Finally, we estimated population-attributable risk for POAG associated with myopia and moderate or high myopia.
Because the correlation between the 2 eyes was high (correlation coefficient between the right and left eyes for AL, 0.93 [P < .001]) and the results of analyses using data from left eyes were similar, only the data from right eyes are shown in further analyses.
A total of 3280 subjects were recruited, giving a response rate of 78.7% for the study. From 3280, we excluded 154 with previous cataract surgery in both eyes, and 17 without lens status, leaving 3109 in this study. There were 104 cases of POAG. Table 1 shows the characteristics of participants with and without POAG. In general, persons with POAG tended to be older, but there were no significant differences for sex, education, previous hypertension, diabetes mellitus or smoking status, total cholesterol and triglyceride levels, or body mass index.
Table 2 shows the crude and adjusted mean of SE, AL, corneal curvature, ACD, and CCT based on right eye data. Eyes with POAG were not significantly different from normal eyes in any of these parameters.
Table 3 shows the relationship of refractive error, AL, corneal curvature, ACD, and CCT with POAG. Persons with moderate or high myopia were more likely to have POAG (OR, 2.87, after adjusting for age, sex, education, height, hypertension and hemoglobin A1c level; OR, 2.80 with further adjustment for CCT). Longer AL was associated with POAG (ORs, 2.49, 3.61,and 2.88, respectively, comparing quartiles 2, 3, and 4 of AL with quartile 1; P = .03 for trend). After further controlling for CCT, persons with AL in quartile 4 were 3 times more likely to have POAG (OR, 3.00) than persons with AL in quartile 1. Deeper ACD was significantly associated with lower risk of POAG (quartile 4 vs 1: OR, 0.31; 95% CI, 0.10-0.98). Corneal curvature and CCT were not associated with POAG.
In models of refractive error and POAG adjusted additionally for AL, the association between moderate or high myopia and POAG was no longer significant (OR, 2.77; 95% CI, 0.77-9.97), and AL contributed significantly to this association based on the likelihood ratio statistic (P < .001) (data not shown).
Finally, the population-attributable risk of POAG associated with myopia was 14.6% and for moderate or high myopia was 5.5% (data not shown).
This population study in Singapore Malays shows an association between moderate or high myopia (worse than −4 D) and POAG. Persons with moderate or high myopia had an almost 3 times higher risk of POAG compared with those with emmetropia. This finding confirmed previous reports of the association between myopia and POAG in other studies, principally the Blue Mountains Eye Study and the Beaver Dam Eye Study, the statistical data of which are given in Table 4 for comparison. Our study is unique, however, in that we were also able to demonstrate an association between increasing AL (measured using noncontact partial coherence laser interferometry) and POAG. We also demonstrated that the association of moderate or high myopia and POAG was no longer significant after controlling for AL, suggesting that axial myopia rather than other factors (eg, corneal curvature or lenticular changes) may be the main biometric constituent that underlies risk for POAG.
We found that a longer AL was associated with POAG, but also that a shorter ACD was also associated with POAG. This is an interesting observation, and we believe this could be due to ACD being shallower in older persons because of increased lens thickness27- 30 and increased prevalence of POAG with age.31- 33 Several theories have been put forward to explain a link between myopia and POAG. Myopia has been found to influence IOP, with myopia associated with a higher IOP than emmetropia and hyperopia.13 The optic nerve head in myopic eyes may be more susceptible than nonmyopic eyes to glaucomatous damage from elevated or normal IOP.8,34,35 Shearing forces exerted by scleral tension across the lamina cribrosa may be crucial to the mechanism of glaucomatous damage.36 Investigators37 have calculated that myopic eyes have higher scleral tension across the lamina than eyes with a shorter AL, even when IOP is the same. This difference becomes even more marked in eyes with thinner sclera. Similar connective tissue changes may also occur in glaucoma and myopia.38 Our finding that AL was significantly associated with POAG largely explains the association between myopia and POAG and may support a theory involving connective tissue changes being associated with longer axial dimensions as a potential mechanism for POAG.
The strengths of our study include that this was a population-based study with a high participation rate (78.7%) and with diagnosis of glaucoma based on optic nerve changes and perimetric findings according to International Society for Geographical and Epidemiological Ophthalmology criteria, unlike the Beaver Dam Eye Study16 in which IOP was used to define glaucoma. The standardized assessment of refraction, IOP measurement, and glaucoma definitions strengthen the validity of any conclusions. Our population-based design also minimizes selection bias, unlike some previous clinic-based studies2,4,7 in which the fact that more myopic patients would attend a clinic would increase the likelihood of detecting POAG. Furthermore, previous studies4,13,14,16- 18 have not examined the influence of CCT. Recent studies have shown the correlation of CCT with optic disc measurements,39,40 and thinner CCT has been identified as a risk factor for the development of POAG in eyes with ocular hypertension.3,41 In our study, CCT was not associated with POAG.
Longer AL however, does not correlate with CCT in myopic patients,42 in healthy populations, or in patients with glaucoma.43 We found similarly that AL was not associated with CCT (Pearson correlation coefficient, 0.081). Furthermore, we did not find that controlling for CCT influenced our results. The measurement of AL by noncontact partial coherence laser interferometry instead of ultrasonographic methods and the use of subjective refractions by trained optometrists instead of autorefraction further increase the robustness of our study.
A limitation of our study was its cross-sectional nature. The conclusions we can draw about the potential influence of myopia on POAG are not based on longitudinal data. A cohort study will help determine whether myopia is associated with subsequent risk of POAG. Another limitation is that these data were based on single measurements (of refraction, IOP, optic disc, and visual fields) during the course of the study. As shown in Table 3, the number of participants with higher degrees of myopia were small. Therefore, although the study had a large number of participants, once they were subdivided into myopia categories, more robust associations were hard to find because the study was underpowered to elicit this information. This could explain why the 95% CIs are quite wide in Table 3. Finally, the clinical diagnosis of glaucoma in myopia may be difficult. The optic discs of myopic patients are also notoriously difficult to assess. The discs frequently appear glaucomatous, with larger diameters, greater cup-disc ratios, and larger and shallower optic cups.20,44,45 Myopic discs are often obliquely inserted, which can result in an abnormal shape with horizontal ovalness or cyclotorsion. Visual field abnormalities have also been reported in highly myopic eyes and in eyes with tilted discs.46 It is possible that cases of high myopia were misclassified as POAG in previous population-based and clinical studies, leading to a spurious association between myopia and POAG. However, of a total of 59 cases of right eyes with POAG in our study, only 1 had high myopia of more than −6 D, so misclassification is not expected to be substantial.
Population-attributable risk estimates are best used to prioritize medical and public health interventions based on the magnitude of the potential effect of a risk factor on the disease outcome in the community. In determining the impact of myopia on POAG, we have found a population-attributable risk of 14.6% for myopia and 5.5% for moderate or high myopia. This finding suggests that myopia may account for 1 in 7 cases of POAG, at least in Malays. It is worth noting that the prevalence of myopia is higher in Chinese Singaporeans than in Malays. Furthermore, with increasing prevalence of myopia in younger Singaporeans, the contribution of myopia to POAG rates over time may be substantial.
In conclusion, in this population-based study in Singapore Malays, we found an association between moderate or high myopia and longer AL and POAG. Our findings suggest that axial myopia is a potential risk factor for POAG.
Correspondence: Tin Aung, MBBS, PhD, FRCS(Edin), Glaucoma Service, Singapore National Eye Centre, 11 Third Hospital Ave, Singapore 168751 (firstname.lastname@example.org).
Submitted for Publication: January 31, 2009; final revision received July 10, 2009; accepted August 16, 2009.
Financial Disclosure: None reported.
Funding/Support: This study was supported by grant 0796/2003 from the National Medical Research Council and grant 501/1/25-5 from the Biomedical Research Council of Singapore; by the Singapore Tissue Network; and by the Ministry of Health.