Association Between Myopic Refractive Error and Primary Open-Angle Glaucoma: A 2-Sample Mendelian Randomization Study

Hélène Choquet; Anthony P Khawaja; Chen Jiang; Jie Yin; Ronald B Melles; M Maria Glymour; Pirro G Hysi; Eric Jorgenson

doi:10.1001/jamaophthalmol.2022.2762

. 2022 Jul 28;140(9):864–871. doi: 10.1001/jamaophthalmol.2022.2762

Association Between Myopic Refractive Error and Primary Open-Angle Glaucoma

A 2-Sample Mendelian Randomization Study

Hélène Choquet ^1,^✉, Anthony P Khawaja ², Chen Jiang ¹, Jie Yin ¹, Ronald B Melles ³, M Maria Glymour ⁴, Pirro G Hysi ^5,^6,⁷, Eric Jorgenson ⁸

¹Division of Research, Kaiser Permanente Northern California, Oakland

²NIHR Biomedical Research Centre, Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology, London, United Kingdom

³Department of Ophthalmology, Kaiser Permanente Northern California, Redwood City

⁴Department of Epidemiology and Biostatistics, University of California, San Francisco

⁵King’s College London, Section of Ophthalmology, School of Life Course Sciences, London, United Kingdom

⁶King’s College London, Department of Twin Research and Genetic Epidemiology, London, United Kingdom

⁷University College London, Great Ormond Street Hospital Institute of Child Health, London, United Kingdom

⁸Regeneron Genetics Center, Tarrytown, New York

Accepted for Publication: May 31, 2022.

Published Online: July 28, 2022. doi:10.1001/jamaophthalmol.2022.2762

^✉

Corresponding Author: Hélène Choquet, PhD, Division of Research, Kaiser Permanente, 2000 Broadway, Oakland, CA 94612 (helene.choquet@kp.org).

Author Contributions: Dr Choquet and Mr Yin had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Hysi and Jorgenson jointly supervised this work.

Concept and design: Choquet, Melles, Hysi, Jorgenson.

Acquisition, analysis, or interpretation of data: Choquet, Khawaja, Jiang, Yin, Melles, Glymour, Jorgenson.

Drafting of the manuscript: Choquet, Jiang, Melles, Jorgenson.

Critical revision of the manuscript for important intellectual content: Choquet, Khawaja, Yin, Glymour, Hysi, Jorgenson.

Statistical analysis: Jiang, Yin, Hysi, Jorgenson.

Obtained funding: Choquet, Jorgenson.

Administrative, technical, or material support: Choquet, Khawaja.

Supervision: Choquet, Melles, Jorgenson.

Conflict of Interest Disclosures: Dr Choquet reported a grant from the National Eye Institute (R01 EY027004) during the conduct of the study. Dr Khawaja reported grants from Alcon and consulting fees from Abbvie, Aerie, Google Health, Novartis, Reichert, Santen, and Thea outside the submitted work. Dr Jorgenson reported being an employee of Regeneron Pharmaceuticals. No other disclosures were reported.

Funding/Support: Support for participant enrollment, survey completion, and biospecimen collection for the Research Program on Genes, Environment, and Health was provided by the Robert Wood Johnson Foundation, Wayne and Gladys Valley Foundation, Ellison Medical Foundation, and Kaiser Permanente Community Benefit Programs. Genotyping of the GERA cohort was funded by a grant from the National Institute on Aging, National Institute of Mental Health, and National Institute of Health Common Fund (RC2 AG036607). Drs Choquet and Jorgenson were supported by a National Eye Institute grant (R01 EY027004). Dr Khawaja is supported by a UK Research and Innovation Future Leaders Fellowship.

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: We are grateful to the Kaiser Permanente Northern California members who have generously agreed to participate in the Kaiser Permanente Research Program on Genes, Environment, and Health.

^✉

Corresponding author.

PMCID: PMC9335248 PMID: 35900730

This genetic association study uses a 2-sample mendelian randomization approach to evaluate shared genetic influences and investigate the association of myopic refractive error and primary open-angle glaucoma.

Key Points

Question

What is the nature of the association between myopic refractive error and primary open-angle glaucoma (POAG)?

Findings

In this genetic association study, which used 2-sample Mendelian randomization and included 154 018 participants from the UK Biobank and the Genetic Epidemiology Research on Adult Health and Aging (GERA) cohorts, myopic refractive error was associated with an increased risk of POAG.

Meaning

These results contribute to understanding the association between myopic refractive error and POAG; subsequent population POAG risk stratification potentially could be based on refractive error information, possibly enabling earlier diagnosis and preventative strategies for high-risk individuals.

Abstract

Importance

Refractive error (RE) is the most common form of visual impairment, and myopic RE is associated with an increased risk of primary open-angle glaucoma (POAG). Whether this association represents a causal role of RE in the etiology of POAG remains unknown.

Objective

To evaluate shared genetic influences and investigate the association of myopic RE with the risk for POAG.

Design, Setting, and Participants

Observational analyses were used to evaluate the association between mean spherical equivalent (MSE) RE (continuous trait) or myopia (binary trait) and POAG risk in individuals from the Genetic Epidemiology Research on Adult Health and Aging (GERA) cohort. To quantify genetic overlap, genome-wide genetic correlation analyses were performed using genome-wide association studies (GWAS) of MSE RE or myopia and POAG from GERA. Potential causal effects were assessed between MSE RE and POAG using 2-sample Mendelian randomization. Genetic variants associated with MSE RE were derived using GWAS summary statistics from a GWAS of RE conducted in 102 117 UK Biobank participants. For POAG, we used GWAS summary statistics from our previous GWAS (3836 POAG cases and 48 065 controls from GERA). Data analyses occurred between July 2020 and October 2021.

Main Outcomes and Measure

Our main outcome was POAG risk as odds ratio (OR) caused by per-unit difference in MSE RE (in diopters).

Results

Our observational analyses included data for 54 755 non-Hispanic White individuals (31 926 [58%] females and 22 829 [42%] males). Among 4047 individuals with POAG, mean (SD) age was 73.64 (9.20) years; mean (SD) age of the 50 708 controls was 65.38 (12.24) years. Individuals with POAG had a lower refractive MSE and were more likely to have myopia or high myopia compared with the control participants (40.2% vs 34.1%, P = 1.31 × 10⁻¹¹ for myopia; 8.5% vs 6.8%, P = .004 for high myopia). Our genetic correlation analyses demonstrated that POAG was genetically correlated with MSE RE (r_g, −0.24; SE, 0.06; P = 3.90 × 10⁻⁵), myopia (r_g, 0.21; SE, 0.07; P = .004), and high myopia (r_g, 0.23; SE, 0.09; P = .01). Genetically assessed refractive MSE was negatively associated with POAG risk (inverse-variance weighted model: OR per diopter more hyperopic MSE = 0.94; 95% CI, 0.89-0.99; P = .01).

Conclusions and Relevance

These findings demonstrate a shared genetic basis and an association between myopic RE and POAG risk. This may support population POAG risk stratification and screening strategies, based on RE information.

Introduction

Primary open-angle glaucoma (POAG) is the most common form of glaucoma, characterized by a progressive optic neuropathy.^1,2 Its etiology is complex, and well-established risk factors include age, family history, germline genetic variants, and elevated intraocular pressure (IOP).³ These factors may impart their effects on POAG risk directly or through the interaction of preexisting genetic risk and demographic and clinical factors.⁴

Previous observational studies have identified an association between refractive error (RE) and POAG risk.^5,6,7,8 Individuals with myopia are more likely to develop POAG,^9,10 and although individuals with high myopia had the greatest risk of POAG, those with low and moderate myopia also had an intermediate risk.⁴ These observations suggest an etiological link between RE and POAG beyond the diagnostic issues encountered in patients with high myopia.¹¹ However, the association between myopia and POAG is poorly understood at clinical, pathological, and mechanistic levels.^{8,12,13,14,15,16,17,18,19,20} Furthermore, although both RE and POAG are heritable,^3,21,22 it is unclear whether the link between RE and POAG operate through a shared genetic etiology.²³ It is also unknown whether this association represents a causal relationship, as associations from observational studies may be prone to confounding effects and therefore cannot answer these questions.

Mendelian randomization (MR) analysis can overcome some of the limitations of observational studies and establish a causal link between RE and POAG risk.^24,25,26,27 The MR approach takes advantage of the principle of random segregation of alleles, according to which genetic associations are likely to be independent of confounding factors. Findings from MR studies are generally consistent with those of randomized clinical trials.²⁸ However, to our knowledge, no MR studies have evaluated the observational associations between RE and POAG.

We use a 2-sample MR approach²⁹ to infer the potential causal relationship between myopic RE and POAG risk by comparing genetic effect estimates obtained through genome-wide association studies (GWAS) summary statistics. We are using results from (1) our previous GWAS of POAG,³ conducted using the Genetic Epidemiology Research on Adult Health and Aging (GERA) cohort, and (2) our recent GWA meta-analysis of mean spherical equivalent (MSE) RE,²² including UK Biobank (UKB) European participants. These large-scale GWAS provide an opportunity to examine shared genetic influences and potential causal effects between myopic RE and POAG risk. This study fills an important gap in our knowledge of the complex etiology of POAG.

Methods

Observational Analyses in GERA

We first conducted observational analyses to evaluate the association of MSE RE or age and sex (self-reported) with POAG risk in the GERA cohort, which consists of 110 266 adult members of the Kaiser Permanente Medical Care Plan, Northern California Region (KPNC), an integrated health care delivery system, and includes ongoing longitudinal electronic health records.^30,31 The institutional review board of the Kaiser Foundation Research Institute approved all study procedures. Written informed consent was obtained from all participants. Patients were not offered any compensation or incentives to participate in the study.

These analyses included 54 755 GERA non-Hispanic White individuals (4047 POAG cases and 50 708 controls). POAG status was defined using similar criteria as before.³ Briefly, patients with POAG were diagnosed by a KPNC ophthalmologist and were identified in the KPNC electronic health record system based on diagnosis codes from the International Classification of Diseases, Ninth Revision, or International Statistical Classification of Diseases and Related Health Problems, Tenth Revision. Patients with POAG were defined as having at least 2 diagnoses of POAG or 2 diagnoses of normal tension glaucoma or 1 diagnosis of POAG and 1 of normal tension glaucoma. Further, patients with POAG did not have any diagnosis of other subtypes of glaucoma. Our control group included all the noncases after excluding participants who have 1 or more diagnoses of any type of glaucoma.

Among those 54 755 individuals, 44 344 had at least 1 spherical equivalent value measured during routine eye examinations. Most individuals had multiple measures for both eyes. Spherical equivalent was calculated as the sphere + (cylinder ÷ 2). The spherical equivalent was selected from the first documented refraction assessment, and the mean of both eyes was used. As previously described,⁴ participants with histories of cataract surgery (in either eye), refractive surgery, keratitis, or corneal diseases were excluded from further analyses.

Among the 44 344 individuals with RE measurement, 15 295 (34.5%) had myopia and 2158 (4.9%) had high myopia. Myopia cases were defined as having a MSE of −0.75 D or less and control participants as having a MSE greater than −0.75 D. High myopia cases were defined as having a MSE of −5 D or less and controls as having a MSE greater than −0.75 D.

We first performed univariate model logistic regression analyses of POAG risk and each risk factor (age, sex, MSE RE, myopia, and high myopia). We then performed multivariable analyses including age and sex as covariates. As neurodegeneration at the optic nerve head and IOP are 2 possible pathways that may underlie the association between myopia and POAG,^{22,32,33,34,35,36} we also used a logistic regression model to investigate the association of MSE RE or myopia with POAG risk adjusting for vertical cup-disc ratio (VCDR) or IOP, in addition to age and sex. As previously described,³³ IOP was measured by standard Goldmann tonometry, which is the main standard equipment for measuring IOP in KPNC ophthalmology practices. In KPNC, VCDR was estimated by clinicians using ophthalmoscopy (using Topcon NW400 or Nidek AFC-330 cameras). Both IOP and VCDR were entered at each vision encounter into the electronic health records as smart variables. Among the 54 755 GERA individuals included in the observational analyses, 51 679 had at least 1 IOP value (average of 7.7 IOP measurements/individual) and 35 949 had at least 1 VCDR value (average of 4.6 VCDR measurements/individual) measured during eye examinations. An individual’s mean IOP (or VCDR) from both eyes for each visit was first estimated, and the individual’s median of the mean across all the visits was used for analyses.

Genetic Correlation Analyses in GERA

We assessed genetic correlations (r_g), which represent a measure of single-nucleotide variation (SNV) effects genome-wide, between POAG and MSE RE, myopia, or high myopia using cross-trait linkage disequilibrium score regression³⁷ using the LD Hub web interface.³⁸ For these genetic correlation analyses, we conducted 3 GWAS analyses (RE, myopia, and high myopia) in GERA non-Hispanic White individuals (eFigures 1-6 in the Supplement). We also used GWAS summary statistics from our previous GWAS of POAG conducted in GERA³ (including 3836 POAG cases and 48 065 controls who were non-Hispanic White). Genetic correlations were considered significant after Bonferroni adjustment for multiple testing (P < .017, which corresponds to .05 ÷ 3 traits tested).

Genetic Instruments for RE

Genetic variants as instrumental variables for RE (exposure) were extracted from a GWAS conducted in 102 117 UKB participants of European ancestry (eFigures 7 and 8 in the Supplement) with direct refraction measurement and part of our previously reported European ancestry meta-analysis.²² In UKB, RE was measured directly using the Tomey RC 5000 Auto-Refractor Keratometer. The spherical equivalent was estimated as the spherical RE (UKB codes 5084 and 5085) plus half the cylindrical error (UKB 5086 and 5087) for each eye. The MSE was then used as the outcome of the GWAS analysis.²² In the current study, we used the lead SNVs previously reported as genome-wide significant (P < 5.0 × 10⁻⁸) as a set of genetic instruments. Genetic instruments were then clumped using a window of 10 Mb and maximal linkage disequilibrium of r² = 0.001 between instruments to ensure that genetic variants were independent. After clumping, a total of 168 genetic instruments for MSE RE were used for the MR analyses (eTable 1 in the Supplement). The proportion of phenotypic variance in RE explained by those variants was calculated in the GERA non-Hispanic White sample to assess the strength of genetic instruments.

GWAS Summary Statistics for POAG

Genetic association data for POAG risk (outcome) were retrieved from our previous GWAS study conducted in GERA.³ In this study, the proportion of phenotypic variance in POAG risk explained by identified loci explained 3.0% in the GERA non-Hispanic White sample.

Two-Sample MR Analyses

All analyses were conducted in the R software version 4.0.1 using the TwoSampleMR package.²⁹ This package makes causal inference about an exposure on an outcome using GWAS summary statistics, generates linkage disequilibrium pruning of exposure SNVs and harmonizes exposure and outcome data sets. We used the inverse-variance weighted (IVW) method as our primary source of MR estimates. This IVW method essentially translates to a weighted regression of SNV outcome effects on SNV-exposure effects where the intercept is constrained to zero. Moreover, we reported the estimations from MR weighted median, weighted mode, and MR-Egger. Further, leave-one-SNV-out analyses were conducted (eTable 2 in the Supplement).

Sensitivity Analyses

The potential effect of pleiotropy was evaluated by the regression intercept from the MR-Egger method³⁹ and Cochran Q tests were used to evaluate the presence of global heterogeneity among the effects of the genetic instruments.⁴⁰ The MR-PRESSO^40,41 method was also used to provide an MR estimate that is robust against the presence of heterogeneity among SNV effects and to reassess the MR estimate after excluding outlier SNVs.

Replication MR Analyses

To validate our initial 2-sample MR findings, we conducted a replication analysis using data from the International Glaucoma Genetics Consortium large GWAS meta-analysis of POAG⁴² as the outcome. GWAS summary statistics for this study⁴² were publicly accessible via GWAS Catalog under study accession identifier GCST90011767 (corresponding to all European ancestry cohorts except UKB; n = 15 229 POAG cases and 177 473 control participants). The proportion of variance in POAG risk explained by identified loci was 3.4%. We used the same set of 168 genetic instruments for MSE RE as mentioned above.

Multivariable MR Adjusting for the Effect of IOP and VCDR on POAG Risk

Recent MR studies^22,43 suggested that higher IOP is associated with lower (more myopic) RE. Further, some genetic loci exhibited pleiotropic effects with VCDR, IOP, POAG, and myopia.³⁵ Because these associations can potentially bias the inferred relationship between MSE RE and POAG, we conducted multivariable MR analyses^44,45 to adjust for the potential effect of IOP and/or VCDR. GWAS summary statistics for these 2 eye traits were publicly accessible from the study of Springelkamp et al³⁶ and were obtained from study samples that did not overlap with those for MSE RE and POAG. SNVs achieving genome-wide significance for each trait (eg, IOP and MSE RE) were included in the multivariable MR analysis. After clumping, a total of 144 and 139 genetic instruments were used for the MR analyses adjusting for the effect of IOP and VCDR, respectively (eTables 3 and 4 in the Supplement). The multivariable MR analyses were performed by jointly fitting the SNV-MSE RE and SNV-IOP (and/or SNV-VCDR) effect sizes simultaneously in the regression model on the SNV-POAG association using the mv_multiple() function available in the TwoSampleMR package in R curated in the MR-Base platform.²⁹

We followed the Strengthening the Reporting of Genetic Association Studies (STREGA) reporting guidelines⁴⁶ for the current genetic investigation study.

Results

Observational Analyses

We conducted an observational analysis of data for 54 755 GERA non-Hispanic White individuals (31 926 [58%] females and 22 829 [42%] males), of whom 44 344 had RE measurements. The baseline characteristics of the study cohort by POAG status are summarized in Table 1. Among 4047 individuals with POAG, the mean (SD) age was 73.64 (9.20) years; the mean (SD) age of the 50 708 controls was 65.38 (12.24) years. Consistent with previous reports, POAG cases had a lower MSE RE compared with control participants (mean, −0.71; SD, 2.33, vs mean, −0.38; SD, 2.34; P = 7.16 × 10⁻¹⁴). Moreover, POAG cases were more likely to have myopia and high myopia compared with the controls (40.2% vs 34.1%, P = 1.31 × 10⁻¹¹ for myopia; 8.5% vs 6.8%, P = .004 for high myopia, respectively). Associations between MSE RE, myopia, or high myopia and POAG risk were stronger after adjusting for age, sex, and IOP or VCDR (Table 1).

Table 1. Characteristics of Participants With POAG and Control Participants From a Non-Hispanic White Sample of the GERA Cohort.

	No. of individuals (%)		OR (95% CI)^a	P value^a	OR (95% CI)^b	P value^b	OR (95% CI)^c	P value^c	OR (95% CI)^d	P value^d
	POAG cases (n = 4047)	Control group (n = 50 708)	OR (95% CI)^a	P value^a	OR (95% CI)^b	P value^b	OR (95% CI)^c	P value^c	OR (95% CI)^d	P value^d
Sex
Female	2163 (53.4)	29 763 (58.7)	0.81 (0.76 to 0.86)	7.70 × 10⁻¹¹	NA	NA	NA	NA	NA	NA
Male	1884 (46.6)	20 945 (41.3)	1 [Reference]	NA	NA	NA	NA	NA	NA	NA
Age, mean (SD), y	73.64 (9.20)	65.38 (12.24)	1.07 (1.066 to 1.073)	<1.0 × 10⁻³⁰⁰	NA	NA	NA	NA	NA	NA
MSE RE, mean (SD), D	−0.71 (2.33)	−0.38 (2.34)	0.94 (0.93 to 0.96)	7.16 × 10⁻¹⁴	0.87 (0.86 to 0.89)	9.46 × 10⁻⁶³	0.88 (0.86 to 0.90)	9.03 × 10⁻³⁰	0.91 (0.89 to 0.93)	2.08 × 10⁻¹⁵
No. of individuals with RE measurement	2983	41 360	NA	NA	NA	NA	NA	NA	NA	NA
Myopia, No. (%)
Yes	1199 (40.2)	14 096 (34.1)	1.30 (1.20 to 1.40)	1.31 × 10⁻¹¹	1.68 (1.56 to 1.82)	1.28 × 10⁻³⁸	1.66 (1.48 to 1.85)	2.61 × 10⁻¹⁸	1.50 (1.34 to 1.68)	3.45 × 10⁻¹²
No	1784 (59.8)	27 264 (65.9)	1 [Reference]	NA	NA	NA	NA	NA	NA	NA
High myopia, No. (%)
Yes	166 (8.5)	1992 (6.8)	1.27 (1.08 to 1.50)	.004	2.41 (2.01 to 2.86)	6.46 × 10⁻²³	2.40 (1.90 to 3.04)	2.35 × 10⁻¹³	1.76 (1.37 to 2.25)	8.39 × 10⁻⁰⁶
No	1784 (91.5)	27 265 (93.2)	1 [Reference]	NA	NA	NA	NA	NA	NA	NA
IOP, mean (SD), mm Hg	18.58 (4.11)	15.31 (2.74)	NA	NA	NA	NA	NA	NA	NA	NA
No. with IOP measurement	1907	49 772	NA	NA	NA	NA	NA	NA	NA	NA
VCDR, mean (SD)	0.59 (0.21)	0.31 (0.12)	NA	NA	NA	NA	NA	NA	NA	NA
No. with VCDR measurement	2938	33 011	NA	NA	NA	NA	NA	NA	NA	NA

Open in a new tab

Abbreviations: GERA, Genetic Epidemiology Research on Adult Health and Aging; IOP, intraocular pressure; MSE, mean spherical equivalent; NA, not applicable; OR, odds ratio; POAG, primary open-angle glaucoma; RE, refractive error; VCDR, vertical cup-disc ratio.

^{^a}

ORs or exp(estimate) and P values were assessed in univariate logistic regression models (without adjusting for any covariates).

^{^b}

ORs or exp(estimate) and P values were assessed in multivariate logistic regression models adjusting for age and sex.

^{^c}

ORs or exp(estimate) and P values were assessed in multivariate logistic regression models adjusting for age, sex, and IOP.

^{^d}

ORs or exp(estimate) and P values were assessed in multivariate logistic regression models adjusting for age, sex, and VCDR.

POAG Shares Genetic Determinants With RE and Myopia

To estimate the shared genetic basis of POAG and MSE RE, myopia, or high myopia, we conducted genome-wide genetic correlation analyses using cross-trait linkage disequilibrium score regressions^37,38 in GERA non-Hispanic White individuals. POAG was genetically associated with MSE RE (r_g, −0.24; SE, 0.06; P = 3.90 × 10⁻⁵) (Table 2). Consistently, we also found evidence of genetic association between POAG and myopia (r_g, 0.21; SE, 0.07; P = .004), as previously reported in the UKB European sample⁴² and between POAG and high myopia (r_g, 0.23; SE, 0.09; P = .01). These results suggest considerable shared genetic influences between POAG and myopic RE.

Table 2. Genetic Correlations Between POAG and MSE Refractive Error, Myopia, or High Myopia in a Non-Hispanic White Sample of the GERA Cohort.

Trait	r_g (SE)	P value
MSE refractive error	−0.24 (0.06)	3.90 × 10⁻⁵
Myopia (MSE ≤−0.75 D)	0.21 (0.07)	.004
High myopia (MSE ≤−5 D)	0.23 (0.09)	.01

Open in a new tab

Abbreviations: GERA, Genetic Epidemiology Research on Adult Health and Aging; MSE, mean spherical equivalent; POAG, primary open-angle glaucoma.

Mendelian Randomization Analyses Support Causal Effects of RE on POAG

To investigate whether RE causally influences POAG risk, we conducted 2-sample MR analyses using, as instrumental variables for the RE exposure, established genetic variants from a previous GWAS conducted in the UKB European sample and part of our previously reported European ancestry meta-analysis.²² We used 168 independent genetic variants as genetic instruments for RE and compared their effect sizes with those observed in the GWAS of POAG. We found evidence for a causal effect of RE on POAG risk, as a more negative MSE RE was associated with an increased risk of POAG (IVW model: odds ratio [OR] per diopter more hyperopic MSE = 0.94; 95% CI, 0.89-0.99; P = .01) (Table 3 and Figure). These 168-SNV genetic instruments explained close to 6.1% of the phenotypic variation in RE in the GERA non-Hispanic White sample.

Table 3. Mendelian Randomization Results of the Associations of MSE Refractive Error (UK Biobank European Individuals) With POAG (GERA Non-Hispanic White Cohort or IGGC European Individuals).

Exposure (UKB)	Outcome (GERA or IGGC)	No. of genetic instruments after clumping	MR method	OR (95% CI)	P value	Detected outlier SNV via MR-PRESSO
Initial MR analyses
MSE RE	POAG^a	168	IVW	0.94 (0.89 to 0.99)	.01
MSE RE	POAG^a	167	MR-PRESSO model	0.94 (0.90 to 0.99)	.02	rs10220706
Replication MR analyses
MSE RE	POAG^b	168	IVW	0.92 (0.89 to 0.95)	5.99 × 10⁻⁶
MSE RE	POAG^b	165	MR-PRESSO model	0.93 (0.91 to 0.96)	8.10 × 10⁻⁶	rs10220706, rs67362351, rs999951
MSE RE	POAG^b	144	Multivariable MR^c	0.94 (0.89 to 0.98)	4.89 × 10⁻³
MSE RE	POAG^b	139	Multivariable MR^d	0.94 (0.89 to 0.99)	.03
MSE RE	POAG^b	139	Multivariable MR^e	0.95 (0.90 to 1.00)	.05

Open in a new tab

Abbreviations: GERA, Genetic Epidemiology Research on Adult Health and Aging; IGGC, International Glaucoma Genetics Consortium; IVW, inverse-variance weighted model; MR, Mendelian randomization; MSE, mean spherical equivalent; OR, odds ratio; POAG, primary open-angle glaucoma; RE, refractive error; SNV, single-nucleotide variation; VCDR, vertical cup-disc ratio.

^{^a}

Non-Hispanic White sample from GERA.

^{^b}

IGGC European population.

^{^c}

Multivariable MR analysis, a regression-based MR method, to adjust for the effect of IOP.

^{^d}

Multivariable MR analysis, a regression-based MR method, to adjust for the effect of VCDR.

^{^e}

Multivariable MR analysis, a regression-based MR method, to adjust for the effect of IOP and VCDR.

Figure. — The x-axis shows 168 genetic instruments for MSE RE and their effect size estimates (β) with RE. The y-axis shows the association of the same variants with POAG risk. The Mendelian randomization inverse-weighted regression line is plotted (blue). The outlier genetic variant rs10220706 detected by the MR-PRESSO model is labeled in gold. OR indicates odds ratio.

Sensitivity Analyses

While no evidence of directional and/or horizontal pleiotropy was found (MR-Egger intercept P = .77), we observed significant heterogeneity among the effects of the genetic instruments using the computed Cochran Q statistics (Q = 246.6, P = 6.08 × 10⁻⁵). To address this evidence of heterogeneity, we conducted a MR-PRESSO model to identify potential outliers for RE. After excluding 1 outlier genetic variant (ie, intergenic SNV rs10220706) as the source of most of the observed heterogeneity, genetically determined MSE RE was confirmed to have a potential causal effect on increasing the risk of POAG (OR per diopter more hyperopic MSE = 0.94; 95% CI, 0.90-0.99; P = .02) (Table 3).

Replication MR Analyses

To validate our initial 2-sample MR findings, we conducted a replication 2-sample MR analysis using the data from our recent large GWAS meta-analysis of POAG⁴² conducted by the International Glaucoma Genetics Consortium as the outcome. We confirm that genetically assessed refractive MSE was negatively associated with POAG risk (IVW model: OR, 0.93; 95% CI, 0.89-0.96; P = 4.14 × 10⁻⁵ Table 3). Evaluation of MR associations under other MR models (MR weighted median, weighted mode, and MR-Egger) showed consistency with the IVW model (eTable 5 and eFigure 9 in the Supplement).

Multivariable MR Analyses

We also conducted multivariable MR analyses to assess the association of MSE RE with POAG risk after adjusting for the effects from IOP and/or VCDR. The association of MSE RE with POAG risk was essentially unchanged after adjusting for IOP or VCDR (OR, 0.94; 95% CI, 0.89-0.98; P = .005, or OR, 0.94; 95% CI, 0.89-0.99; P = .03, respectively) (Table 3).

Discussion

This study leveraged genetic methods to evaluate whether the reported observational associations between POAG and RE, myopia, or high myopia can be explained by shared genetic influences. We also investigated causality between POAG and RE. We found evidence for shared genetic influences between RE, myopia, or high myopia and POAG, as well as a potential causal effect of myopic RE on POAG, even after adjusting for IOP or VCDR effects. These associations were robust in sensitivity analyses that address significant heterogeneity among the effects of the genetic instruments.

Our genetic correlation analyses provide evidence for shared genetic influences between RE, myopia, or high myopia and POAG, with the strongest genetic correlation detected with MSE RE. Our results are consistent with a recent study that reported a nominal genetic correlation between POAG and myopia in the UKB European sample⁴² but stand in contrast to a previous study²³ that reported no evidence to support a genetic overlap between myopia and POAG. Using larger sample sizes and using a 2-sample MR approach, we were able to detect shared genetic associations between myopia and POAG and determine that lower MSE RE increases the risk of POAG.

Our MR analyses were conducted using a valid instrument for causal inference under the 3 assumptions required for MR studies^47,48: the genetic instrument must (1) be truly associated with the exposure, (2) not be influenced by any confounders of the exposure-outcome association, and (3) only be related to the outcome of interest through the exposure under study. The first assumption was satisfied by the use of SNVs previously reported as genome-wide significant in a large cohort, making those strong genetic instruments for MSE RE. The second assumption was partially satisfied by the fact that we found similar results using the MR-PRESSO method, which is a robust method for sensitivity analysis that first removes genetic variants for which variant-specific causal estimate differs substantially from those of other variants. The third assumption was satisfied because no evidence of horizontal pleiotropy was found even if recent MR studies^49,50 provide evidence that existing methods for detecting and accounting for horizontal pleiotropy are ineffective under some plausible conditions. Thus, our results were unlikely to be affected by the violation of MR assumptions.

Limitations

There are several limitations to the current study. First, although this genetic instrumental variable study of POAG on RE was able to detect that lower MSE RE is associated with increased risk of POAG, the mechanisms behind this association are not entirely clear. Future investigations may improve understanding of underlying molecular biology and exact pathways involved. Second, while we were able to determine that lower MSE RE is associated with increased POAG risk in European ancestry individuals, the generalizability of our findings may be not applicable to non–European ancestry populations. This limitation may be addressed in future investigations as genetic data in other ancestry groups become more available.

Our results support the use of RE as a metric to help stratify risk of POAG in the general population. Currently, general population screening for glaucoma is not recommended, in part due to the relatively low prevalence of undetected cases.⁵¹ Understanding which factors increase risk of POAG can help inform strategies for identifying glaucoma-enriched populations for targeted screening, which would lead to earlier diagnosis and preventative strategies for high-risk individuals.

Conclusions

Our study confirms previous observational studies showing that individuals with POAG have a lower MSE RE and are more likely to have myopia and high myopia compared with control participants. Our study also provides evidence of shared genetic etiology between POAG and RE, myopia, or high myopia. Further, using 2 large cohorts, our MR analysis suggests that myopic RE is associated with increased risk of POAG. Altogether, clarifying the nature of the association between RE and POAG may open new avenues of investigation into the specific mechanisms underlying these vision disorders.

Supplement.

eFigure 1. Quantile-quantile plot and genomic inflation factors (λ) observed for the GERA non-Hispanic white GWAS of mean spherical equivalent refractive error

eFigure 2. Manhattan plot of the GWAS of mean spherical equivalent refractive error in GERA non-Hispanic whites

eFigure 3. Quantile-quantile plot and genomic inflation factors (λ) observed for the GERA non-Hispanic white GWAS of myopia

eFigure 4. Manhattan plot of the GWAS of myopia in GERA non-Hispanic whites

eFigure 5. Quantile-quantile plot and genomic inflation factors (λ) observed for the GERA non-Hispanic white GWAS of high myopia

eFigure 6. Manhattan plot of the GWAS of high myopia in GERA non-Hispanic whites

eFigure 7. Quantile-quantile plot and genomic inflation factors (λ) observed for the UK Biobank European GWAS of mean spherical equivalent refractive error

eFigure 8. Manhattan plot of the GWAS of mean spherical equivalent refractive error in UK Biobank Europeans

eFigure 9. Replication of the association of MSE refractive error-associated variants with the risk of POAG.

eTable 1. Genetic instruments (168-SNP - P<5.0x10^-8) for mean spherical equivalent refractive error

eTable 2. Mendelian randomization leave-one-out analysis for MSE refractive error suggesting evidence of causality with POAG.

eTable 3. Genetic instruments (144-SNP - P<5.0x10^-8) for intraocular pressure

eTable 4. Genetic instruments (139-SNP - P<5.0x10^-8) for vertical cup-to-disc ratio

eTable 5. Evaluation of MR associations of MSE refractive error with POAG under different MR models

Click here for additional data file.^{(1.2MB, pdf)}

References

1.Jonas JB, Aung T, Bourne RR, Bron AM, Ritch R, Panda-Jonas S. Glaucoma. Lancet. 2017;390(10108):2183-2193. doi: 10.1016/S0140-6736(17)31469-1 [DOI] [PubMed] [Google Scholar]
2.Stein JD, Khawaja AP, Weizer JS. Glaucoma in adults—screening, diagnosis, and management: a review. JAMA. 2021;325(2):164-174. doi: 10.1001/jama.2020.21899 [DOI] [PubMed] [Google Scholar]
3.Choquet H, Paylakhi S, Kneeland SC, et al. A multiethnic genome-wide association study of primary open-angle glaucoma identifies novel risk loci. Nat Commun. 2018;9(1):2278. doi: 10.1038/s41467-018-04555-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Shen L, Melles RB, Metlapally R, et al. The association of refractive error with glaucoma in a multiethnic population. Ophthalmology. 2016;123(1):92-101. doi: 10.1016/j.ophtha.2015.07.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Mastropasqua L, Lobefalo L, Mancini A, Ciancaglini M, Palma S. Prevalence of myopia in open angle glaucoma. Eur J Ophthalmol. 1992;2(1):33-35. doi: 10.1177/112067219200200108 [DOI] [PubMed] [Google Scholar]
6.Czudowska MA, Ramdas WD, Wolfs RC, et al. Incidence of glaucomatous visual field loss: a ten-year follow-up from the Rotterdam Study. Ophthalmology. 2010;117(9):1705-1712. doi: 10.1016/j.ophtha.2010.01.034 [DOI] [PubMed] [Google Scholar]
7.Jiang X, Varma R, Wu S, et al. ; Los Angeles Latino Eye Study Group . Baseline risk factors that predict the development of open-angle glaucoma in a population: the Los Angeles Latino Eye Study. Ophthalmology. 2012;119(11):2245-2253. doi: 10.1016/j.ophtha.2012.05.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Marcus MW, de Vries MM, Junoy Montolio FG, Jansonius NM. Myopia as a risk factor for open-angle glaucoma: a systematic review and meta-analysis. Ophthalmology. 2011;118(10):1989-1994.e2. doi: 10.1016/j.ophtha.2011.03.012 [DOI] [PubMed] [Google Scholar]
9.Verhoeven VJ, Buitendijk GH, Rivadeneira F, et al. ; Consortium for Refractive Error and Myopia (CREAM) . Education influences the role of genetics in myopia. Eur J Epidemiol. 2013;28(12):973-980. doi: 10.1007/s10654-013-9856-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Hysi PG, Mahroo OA, Cumberland P, et al. Common mechanisms underlying refractive error identified in functional analysis of gene lists from genome-wide association study results in 2 European British cohorts. JAMA Ophthalmol. 2014;132(1):50-56. doi: 10.1001/jamaophthalmol.2013.6022 [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Chang RT, Singh K. Myopia and glaucoma: diagnostic and therapeutic challenges. Curr Opin Ophthalmol. 2013;24(2):96-101. doi: 10.1097/ICU.0b013e32835cef31 [DOI] [PubMed] [Google Scholar]
12.Mitchell P, Hourihan F, Sandbach J, Wang JJ. The relationship between glaucoma and myopia: the Blue Mountains Eye Study. Ophthalmology. 1999;106(10):2010-2015. doi: 10.1016/S0161-6420(99)90416-5 [DOI] [PubMed] [Google Scholar]
13.Wong TY, Klein BE, Klein R, Knudtson M, Lee KE. Refractive errors, intraocular pressure, and glaucoma in a white population. Ophthalmology. 2003;110(1):211-217. doi: 10.1016/s0161-6420(02)01260-5 [DOI] [PubMed] [Google Scholar]
14.Perera SA, Wong TY, Tay WT, Foster PJ, Saw SM, Aung T. Refractive error, axial dimensions, and primary open-angle glaucoma: the Singapore Malay Eye Study. Arch Ophthalmol. 2010;128(7):900-905. doi: 10.1001/archophthalmol.2010.125 [DOI] [PubMed] [Google Scholar]
15.Xu L, Wang Y, Wang S, Jonas JB. High myopia and glaucoma susceptibility: the Beijing Eye Study. Ophthalmology. 2007;114(2):216-220. doi: 10.1016/j.ophtha.2006.06.050 [DOI] [PubMed] [Google Scholar]
16.Grødum K, Heijl A, Bengtsson B. Refractive error and glaucoma. Acta Ophthalmol Scand. 2001;79(6):560-566. doi: 10.1034/j.1600-0420.2001.790603.x [DOI] [PubMed] [Google Scholar]
17.Kuzin AA, Varma R, Reddy HS, Torres M, Azen SP. Ocular biometry and open-angle glaucoma: the Los Angeles Latino Eye Study. Ophthalmology. 2010;117(9):1713-1719. doi: 10.1016/j.ophtha.2010.01.035 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Wu SY, Nemesure B, Leske MC. Refractive errors in a black adult population: the Barbados Eye Study. Invest Ophthalmol Vis Sci. 1999;40(10):2179-2184. [PubMed] [Google Scholar]
19.Saw SM, Gazzard G, Shih-Yen EC, Chua WH. Myopia and associated pathological complications. Ophthalmic Physiol Opt. 2005;25(5):381-391. doi: 10.1111/j.1475-1313.2005.00298.x [DOI] [PubMed] [Google Scholar]
20.Pan CW, Cheung CY, Aung T, et al. Differential associations of myopia with major age-related eye diseases: the Singapore Indian Eye Study. Ophthalmology. 2013;120(2):284-291. doi: 10.1016/j.ophtha.2012.07.065 [DOI] [PubMed] [Google Scholar]
21.Sanfilippo PG, Hewitt AW, Hammond CJ, Mackey DA. The heritability of ocular traits. Surv Ophthalmol. 2010;55(6):561-583. doi: 10.1016/j.survophthal.2010.07.003 [DOI] [PubMed] [Google Scholar]
22.Hysi PG, Choquet H, Khawaja AP, et al. ; Consortium for Refractive Error and Myopia; UK Eye and Vision Consortium; 23andMe Inc . Meta-analysis of 542,934 subjects of European ancestry identifies new genes and mechanisms predisposing to refractive error and myopia. Nat Genet. 2020;52(4):401-407. doi: 10.1038/s41588-020-0599-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Iglesias AI, Ong JS, Khawaja AP, et al. ; International Glaucoma Genetics Consortium (IGGC) and Consortium for Refractive Error and Myopia (CREAM) . Determining possible shared genetic architecture between myopia and primary open-angle glaucoma. Invest Ophthalmol Vis Sci. 2019;60(8):3142-3149. doi: 10.1167/iovs.18-26231 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Smith GD, Ebrahim S. ‘Mendelian randomization’: can genetic epidemiology contribute to understanding environmental determinants of disease? Int J Epidemiol. 2003;32(1):1-22. doi: 10.1093/ije/dyg070 [DOI] [PubMed] [Google Scholar]
25.Swerdlow DI, Kuchenbaecker KB, Shah S, et al. Selecting instruments for Mendelian randomization in the wake of genome-wide association studies. Int J Epidemiol. 2016;45(5):1600-1616. doi: 10.1093/ije/dyw088 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Paternoster L, Tilling K, Davey Smith G. Genetic epidemiology and Mendelian randomization for informing disease therapeutics: conceptual and methodological challenges. PLoS Genet. 2017;13(10):e1006944. doi: 10.1371/journal.pgen.1006944 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Davey Smith G, Paternoster L, Relton C. When will Mendelian randomization become relevant for clinical practice and public health? JAMA. 2017;317(6):589-591. doi: 10.1001/jama.2016.21189 [DOI] [PubMed] [Google Scholar]
28.Mokry LE, Ahmad O, Forgetta V, Thanassoulis G, Richards JB. Mendelian randomisation applied to drug development in cardiovascular disease: a review. J Med Genet. 2015;52(2):71-79. doi: 10.1136/jmedgenet-2014-102438 [DOI] [PubMed] [Google Scholar]
29.Hemani G, Zheng J, Elsworth B, et al. The MR-Base platform supports systematic causal inference across the human phenome. eLife. 2018;7:e34408. doi: 10.7554/eLife.34408 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Kvale MN, Hesselson S, Hoffmann TJ, et al. Genotyping informatics and quality control for 100,000 subjects in the Genetic Epidemiology Research on Adult Health and Aging (GERA) cohort. Genetics. 2015;200(4):1051-1060. doi: 10.1534/genetics.115.178905 [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Banda Y, Kvale MN, Hoffmann TJ, et al. Characterizing race/ethnicity and genetic ancestry for 100,000 subjects in the Genetic Epidemiology Research on Adult Health and Aging (GERA) cohort. Genetics. 2015;200(4):1285-1295. doi: 10.1534/genetics.115.178616 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Hysi PG, Cheng CY, Springelkamp H, et al. ; BMES GWAS Group; NEIGHBORHOOD Consortium; Wellcome Trust Case Control Consortium 2 . Genome-wide analysis of multi-ancestry cohorts identifies new loci influencing intraocular pressure and susceptibility to glaucoma. Nat Genet. 2014;46(10):1126-1130. doi: 10.1038/ng.3087 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Choquet H, Thai KK, Yin J, et al. A large multi-ethnic genome-wide association study identifies novel genetic loci for intraocular pressure. Nat Commun. 2017;8(1):2108. doi: 10.1038/s41467-017-01913-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Khawaja AP, Cooke Bailey JN, Wareham NJ, et al. ; UK Biobank Eye and Vision Consortium; NEIGHBORHOOD Consortium . Genome-wide analyses identify 68 new loci associated with intraocular pressure and improve risk prediction for primary open-angle glaucoma. Nat Genet. 2018;50(6):778-782. doi: 10.1038/s41588-018-0126-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Han X, Qassim A, An J, et al. Genome-wide association analysis of 95 549 individuals identifies novel loci and genes influencing optic disc morphology. Hum Mol Genet. 2019;28(21):3680-3690. doi: 10.1093/hmg/ddz193 [DOI] [PubMed] [Google Scholar]
36.Springelkamp H, Iglesias AI, Mishra A, et al. ; NEIGHBORHOOD Consortium . New insights into the genetics of primary open-angle glaucoma based on meta-analyses of intraocular pressure and optic disc characteristics. Hum Mol Genet. 2017;26(2):438-453. doi: 10.1093/hmg/ddw399 [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Bulik-Sullivan BK, Loh PR, Finucane HK, et al. ; Schizophrenia Working Group of the Psychiatric Genomics Consortium . LD score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. 2015;47(3):291-295. doi: 10.1038/ng.3211 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Zheng J, Erzurumluoglu AM, Elsworth BL, et al. ; Early Genetics and Lifecourse Epidemiology (EAGLE) Eczema Consortium . LD Hub: a centralized database and web interface to perform LD score regression that maximizes the potential of summary level GWAS data for SNP heritability and genetic correlation analysis. Bioinformatics. 2017;33(2):272-279. doi: 10.1093/bioinformatics/btw613 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44(2):512-525. doi: 10.1093/ije/dyv080 [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Burgess S, Davey Smith G, Davies NM, et al. Guidelines for performing Mendelian randomization investigations. Wellcome Open Res. 2020;4:186. doi: 10.12688/wellcomeopenres.15555.2 [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693-698. doi: 10.1038/s41588-018-0099-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Gharahkhani P, Jorgenson E, Hysi P, et al. ; NEIGHBORHOOD consortium; ANZRAG consortium; Biobank Japan project; FinnGen study; UK Biobank Eye and Vision Consortium; GIGA study group; 23 and Me Research Team . Genome-wide meta-analysis identifies 127 open-angle glaucoma loci with consistent effect across ancestries. Nat Commun. 2021;12(1):1258. doi: 10.1038/s41467-020-20851-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Patasova K, Khawaja AP, Tamraz B, et al. Association between medication-taking and refractive error in a large general population-based cohort. Invest Ophthalmol Vis Sci. 2021;62(2):15. doi: 10.1167/iovs.62.2.15 [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Sanderson E, Davey Smith G, Windmeijer F, Bowden J. An examination of multivariable Mendelian randomization in the single-sample and two-sample summary data settings. Int J Epidemiol. 2019;48(3):713-727. doi: 10.1093/ije/dyy262 [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Burgess S, Thompson SG. Multivariable Mendelian randomization: the use of pleiotropic genetic variants to estimate causal effects. Am J Epidemiol. 2015;181(4):251-260. doi: 10.1093/aje/kwu283 [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Little J, Higgins JP, Ioannidis JP, et al. ; STrengthening the REporting of Genetic Association Studies . Strengthening the Reporting of Genetic Association Studies (STREGA): an extension of the STROBE statement. PLoS Med. 2009;6(2):e22. doi: 10.1371/journal.pmed.1000022 [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Zheng J, Baird D, Borges MC, et al. Recent developments in Mendelian randomization studies. Curr Epidemiol Rep. 2017;4(4):330-345. doi: 10.1007/s40471-017-0128-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Davies NM, Holmes MV, Davey Smith G. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. BMJ. 2018;362:k601. doi: 10.1136/bmj.k601 [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Qi G, Chatterjee N. A comprehensive evaluation of methods for Mendelian randomization using realistic simulations and an analysis of 38 biomarkers for risk of type 2 diabetes. Int J Epidemiol. 2021;50(4):1335-1349. doi: 10.1093/ije/dyaa262 [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Brown BC, Knowles DA. Welch-weighted Egger regression reduces false positives due to correlated pleiotropy in Mendelian randomization. Am J Hum Genet. 2021;108(12):2319-2335. doi: 10.1016/j.ajhg.2021.10.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Hamid S, Desai P, Hysi P, Burr JM, Khawaja AP. Population screening for glaucoma in UK: current recommendations and future directions. Eye (Lond). 2022;36(3):504-509. doi: 10.1038/s41433-021-01687-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials